Re: ISSUE: Possibility to use non-resolvable domain name as path-identity

[Frank Ellermann <[e-mail]>]

Charles Lindsey wrote:
 
>> Plausible, but without a single default address for the role account
>> it's pointless.  Go to the SoA and try abuse@, Tech-C, or hostmaster
>> would be about as promising as trying news@, newsmaster@, or usenet@.
 
> I think there is more use for a working domain name in a path-identity
> than just being able to send email to news@ that address.
 
> For example, you can look it up in whois.

If a domain example.com has a subdomain news17.news-servers.example.com
which happens to be a NNTP host, then it's not very likely that you can
lookup news17.news-servers.example.com or news-servers.example.com in a
whois server of example.com.  [[ Taking the example.com literally I end
up at whois.iana.org ]]

> If the DNS says it is an MX, you can try mailing to that

Using which local part ?  Only postmaster@ is "guaranteed" to work for
news17.news-servers.example.com if that domain has an MX somewhere.

> If the DNS gives you an IP address, you can try traceroute on it

Yes, and I can also portscan it, or try ftp / http / smtp / ...  Only
IPv4 from my POV unfortunately.  That's nothing we need to discuss in
USEPRO, it's no 1123bis.

> if it is the first thing in the Path you can see whether that was
> the IP address that sent the message to you.

Yes, I can do nslookup q=a / q=ptr / etc. whenever it pleases me.

And if we'd want a normative or informative reference to... 
I-D.ietf-dnsop-reverse-mapping-considerations
...it should go into USEPRO.  But I don't see the point yet, I think
you and Russ said that the peers of news17.news-servers.example.com 
can simply hardwire a list of expected IPs instead of using DNS. (?)

> It is all part of the process of making it harder for the Bad Guys
> to hide or to keep on morphing.

If the IP of news17.news-servers.example.com is NOT published in DNS
it might be some kind of "security by obscurity".  If it's only an
anti-spam gimmick it's about as stupid as my local part "nobody" :-(

> And all I am asking is that it SHOULD resolve.

For which query types (apart from soa and ns) ?  A type 99 record
"v=spf1 -all" won't help for the purposes of news (in fact it would
be very near to pointless without a corresponding SMTP server, MTAs
can reject MAIL FROM:<[e-mail]> if
there's no IP and no MX, without wasting time for SPF checks).

SRV or similar records could be interesting, if the "news-servers"
at example.com wish to enumerate their hosts news17, etc.  That's
only a future possibility mentioned in the "URI" I-D so far, and
it's unrelated to any "SHOULD resolve (some query types TBD)".

> Then if it doesn't, it immediately draws attention to itself as a
> cause for suspicion.

I don't recall a single case where I tried `nslookup -q=any` for a
path identity, and I looked into the peering database a few times
while trying to figure out path header fields.  Admittedly I'm more
interested in mail abuse today.

Frank






Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2VFhifU044350 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Sat, 31 Mar 2007 08:43:44 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2VFhi9B044349; Sat, 31 Mar 2007 08:43:44 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2VFhJ5c044313 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Sat, 31 Mar 2007 08:43:43 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HXfjN-00069b-8M for [e-mail]; Sat, 31 Mar 2007 17:43:05 +0200
Received: from 212.82.251.197 ([212.82.251.197]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sat, 31 Mar 2007 17:43:05 +0200
Received: from nobody by 212.82.251.197 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sat, 31 Mar 2007 17:43:05 +0200
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Date:  Sat, 31 Mar 2007 17:38:37 +0200
Organization:  <URL:http://purl.net/xyzzy>
Lines: 69
Message-ID:  <[e-mail]>
References:  <[e-mail]>	 <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]> <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: 212.82.251.197
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey wrote:
 
>> Plausible, but without a single default address for the role account
>> it's pointless.  Go to the SoA and try abuse@, Tech-C, or hostmaster
>> would be about as promising as trying news@, newsmaster@, or usenet@.
 
> I think there is more use for a working domain name in a path-identity
> than just being able to send email to news@ that address.
 
> For example, you can look it up in whois.

If a domain example.com has a subdomain news17.news-servers.example.com
which happens to be a NNTP host, then it's not very likely that you can
lookup news17.news-servers.example.com or news-servers.example.com in a
whois server of example.com.  [[ Taking the example.com literally I end
up at whois.iana.org ]]

> If the DNS says it is an MX, you can try mailing to that

Using which local part ?  Only postmaster@ is "guaranteed" to work for
news17.news-servers.example.com if that domain has an MX somewhere.

> If the DNS gives you an IP address, you can try traceroute on it

Yes, and I can also portscan it, or try ftp / http / smtp / ...  Only
IPv4 from my POV unfortunately.  That's nothing we need to discuss in
USEPRO, it's no 1123bis.

> if it is the first thing in the Path you can see whether that was
> the IP address that sent the message to you.

Yes, I can do nslookup q=a / q=ptr / etc. whenever it pleases me.

And if we'd want a normative or informative reference to... 
I-D.ietf-dnsop-reverse-mapping-considerations
...it should go into USEPRO.  But I don't see the point yet, I think
you and Russ said that the peers of news17.news-servers.example.com 
can simply hardwire a list of expected IPs instead of using DNS. (?)

> It is all part of the process of making it harder for the Bad Guys
> to hide or to keep on morphing.

If the IP of news17.news-servers.example.com is NOT published in DNS
it might be some kind of "security by obscurity".  If it's only an
anti-spam gimmick it's about as stupid as my local part "nobody" :-(

> And all I am asking is that it SHOULD resolve.

For which query types (apart from soa and ns) ?  A type 99 record
"v=spf1 -all" won't help for the purposes of news (in fact it would
be very near to pointless without a corresponding SMTP server, MTAs
can reject MAIL FROM:<[e-mail]> if
there's no IP and no MX, without wasting time for SPF checks).

SRV or similar records could be interesting, if the "news-servers"
at example.com wish to enumerate their hosts news17, etc.  That's
only a future possibility mentioned in the "URI" I-D so far, and
it's unrelated to any "SHOULD resolve (some query types TBD)".

> Then if it doesn't, it immediately draws attention to itself as a
> cause for suspicion.

I don't recall a single case where I tried `nslookup -q=any` for a
path identity, and I looked into the peering database a few times
while trying to figure out path header fields.  Admittedly I'm more
interested in mail abuse today.

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2V4HEtw006533 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 30 Mar 2007 21:17:14 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2V4HE4M006532; Fri, 30 Mar 2007 21:17:14 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2V4Gqdk006520 for <[e-mail]>; Fri, 30 Mar 2007 21:17:13 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster#pop3^clerew#man$ac*uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 460de0b2.5fbe.2be for [e-mail]; Sat, 31 Mar 2007 05:16:50 +0100 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2V4Gotr010529 for <[e-mail]>; Sat, 31 Mar 2007 05:16:50 +0100 (BST)
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2V4GnBX010526 for [e-mail]; Sat, 31 Mar 2007 05:16:49 +0100 (BST)
To: [e-mail]
Xref: clerew local.usefor:24569
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]>	 <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]> <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Fri, 30 Mar 2007 19:14:13 GMT
Lines: 36
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Frank Ellermann <[e-mail]> writes:

>Charles Lindsey wrote:

>> The issue raised was to make that practice a SHOULD NOT (but better
>> worded as "SHOULD be resolvable"). I proposed a wording to that 
>> effect, and Frank suggested a rather similar wording, hence my
>> presumption that he was seconding.

>Plausible, but without a single default address for the role account
>it's pointless.  Go to the SoA and try abuse@, Tech-C, or hostmaster
>would be about as promising as trying news@, newsmaster@, or usenet@.

I think there is more use for a working domain name in a path-identity
than just being able to send email to news@ that address.

For example, you can look it up in whois. If the DNS says it is an MX, you
can try mailing to that, If the DNS gives you an IP address, you can try
traceroute on it, or if it is the first thing in the Path you can see
whether that was the IP address that sent the message to you. It is all
part of the process of making it harder for the Bad Guys to hide or to
keep on morphing.

And all I am asking is that it SHOULD resolve. Then if it doesn't, it
immediately draws attention to itself as a cause for suspicion.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2SHEd4g078395 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 28 Mar 2007 10:14:39 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2SHEdml078394; Wed, 28 Mar 2007 10:14:39 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2SHEG84078377 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Wed, 28 Mar 2007 10:14:38 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HWbab-0002UN-Ju for [e-mail]; Wed, 28 Mar 2007 19:05:37 +0200
Received: from d252026.dialin.hansenet.de ([80.171.252.26]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Wed, 28 Mar 2007 19:05:37 +0200
Received: from nobody by d252026.dialin.hansenet.de with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Wed, 28 Mar 2007 19:05:37 +0200
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Date:  Wed, 28 Mar 2007 18:56:21 +0200
Organization:  <URL:http://purl.net/xyzzy>
Lines: 15
Message-ID:  <[e-mail]>
References:  <[e-mail]>	 <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]> <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126> <[e-mail]> <[e-mail]> <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: d252026.dialin.hansenet.de
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey wrote:

> The issue raised was to make that practice a SHOULD NOT (but better
> worded as "SHOULD be resolvable"). I proposed a wording to that 
> effect, and Frank suggested a rather similar wording, hence my
> presumption that he was seconding.

Plausible, but without a single default address for the role account
it's pointless.  Go to the SoA and try abuse@, Tech-C, or hostmaster
would be about as promising as trying news@, newsmaster@, or usenet@.

And there's at least a public database of defunct abuse@ addresses. 

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2S4GjA1030297 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 27 Mar 2007 21:16:45 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2S4GjuN030296; Tue, 27 Mar 2007 21:16:45 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2S4GNBY030258 for <[e-mail]>; Tue, 27 Mar 2007 21:16:44 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster$pop3&clerew#man^ac&uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 4609ec16.ec40.b1 for [e-mail]; Wed, 28 Mar 2007 05:16:22 +0100 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2S4GMaf010987 for <[e-mail]>; Wed, 28 Mar 2007 05:16:22 +0100 (BST)
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2S4GLOL010984 for [e-mail]; Wed, 28 Mar 2007 05:16:21 +0100 (BST)
To: [e-mail]
Xref: clerew local.usefor:24567
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]>	 <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]> <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126> <[e-mail]> <[e-mail]>
Date: Tue, 27 Mar 2007 18:06:15 GMT
Lines: 29
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Harald Alvestrand <[e-mail]> writes:

>Frank Ellermann wrote:

>> You've noted #1093 as "no change required".  If a new ticket
>> number helps to tackle the issue, let's test it.  I think the
>> WG worked better when Bruce and Henry were still around... :-(
>>
>>   
>the issue raised was the possiblity to use non-resolvable domains as 
>path-identity. I take it from the above that you are not seconding 
>raising that as an issue.

Actually, it was the other way around. That possibility is how the draft
currently stands. The issue raised was to make that practice a SHOULD NOT
(but better worded as "SHOULD be resolvable"). I proposed a wording to
that effect, and Frank suggested a rather similar wording, hence my
presumption that he was seconding.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2Q7Y3rh061728 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 26 Mar 2007 00:34:03 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2Q7Y399061727; Mon, 26 Mar 2007 00:34:03 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2Q7Y2gb061720 for <[e-mail]>; Mon, 26 Mar 2007 00:34:02 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id B585C2596EF; Mon, 26 Mar 2007 09:34:01 +0200 (CEST)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 14308-08; Mon, 26 Mar 2007 09:33:55 +0200 (CEST)
Received: from [127.0.0.1] (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 477C72596EC; Mon, 26 Mar 2007 09:33:55 +0200 (CEST)
Message-ID: <[e-mail]>
Date: Mon, 26 Mar 2007 09:33:55 +0200
From: Harald Alvestrand <[e-mail]>
User-Agent: Thunderbird 1.5.0.9 (X11/20070104)
MIME-Version: 1.0
To: Frank Ellermann <[e-mail]>
Cc: [e-mail]
Subject: Re: ISSUE: Possibility to use non-resolvable domain name as	 path-identity
References: <[e-mail]>	 <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]> <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126> <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Frank Ellermann wrote:
> Harald Tveit Alvestrand wrote:
>  
>   
>> Frank, did you intend to second the entering of this issue,
>> or to say why you didn't think a change was needed?
>>     
>
>   <http://permalink.gmane.org/gmane.ietf.usenet.format/31641>
> |> Well, since that's almost exactly what I want it to say, can
> |> I take it that you 'second' the Issue?
> |
> | As continuation of the #1093 saga, yes, it makes no sense to
> | track the in essence same issue twice.
>
>   <http://permalink.gmane.org/gmane.ietf.usenet.format/31629>
> |> I do not want to take any position on the relative merits of 
> |> "usenet@" and "news@".
> |
> | ...making it more interesting for admins and other affected
> | parties, tiny little mazes, all alike.  The proposed issue
> | could be joined with #1093.
>
>   
>> It was not clear to me that you were agreeing that it needed
>> tracking.
>>     
>
> You've noted #1093 as "no change required".  If a new ticket
> number helps to tackle the issue, let's test it.  I think the
> WG worked better when Bruce and Henry were still around... :-(
>
>   
the issue raised was the possiblity to use non-resolvable domains as 
path-identity. I take it from the above that you are not seconding 
raising that as an issue.

If you want to reopen #1093, that's another matter.

Harald



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2NMT5bZ098996 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 23 Mar 2007 15:29:05 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2NMT5ro098995; Fri, 23 Mar 2007 15:29:05 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2NMT39g098986 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Fri, 23 Mar 2007 15:29:04 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HUsEk-0004V2-7y for [e-mail]; Fri, 23 Mar 2007 23:27:54 +0100
Received: from d253247.dialin.hansenet.de ([80.171.253.247]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Fri, 23 Mar 2007 23:27:54 +0100
Received: from nobody by d253247.dialin.hansenet.de with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Fri, 23 Mar 2007 23:27:54 +0100
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Date:  Fri, 23 Mar 2007 23:26:51 +0100
Organization:  <URL:http://purl.net/xyzzy>
Lines: 29
Message-ID:  <[e-mail]>
References:  <[e-mail]> <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]> <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: d253247.dialin.hansenet.de
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Harald Tveit Alvestrand wrote:
 
> Frank, did you intend to second the entering of this issue,
> or to say why you didn't think a change was needed?

  <http://permalink.gmane.org/gmane.ietf.usenet.format/31641>
|> Well, since that's almost exactly what I want it to say, can
|> I take it that you 'second' the Issue?
|
| As continuation of the #1093 saga, yes, it makes no sense to
| track the in essence same issue twice.

  <http://permalink.gmane.org/gmane.ietf.usenet.format/31629>
|> I do not want to take any position on the relative merits of 
|> "usenet@" and "news@".
|
| ...making it more interesting for admins and other affected
| parties, tiny little mazes, all alike.  The proposed issue
| could be joined with #1093.

> It was not clear to me that you were agreeing that it needed
> tracking.

You've noted #1093 as "no change required".  If a new ticket
number helps to tackle the issue, let's test it.  I think the
WG worked better when Bruce and Henry were still around... :-(

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2M6omE6029971 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 23:50:48 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2M6omnD029970; Wed, 21 Mar 2007 23:50:48 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2M6oQS2029945 for <[e-mail]>; Wed, 21 Mar 2007 23:50:46 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 601A8259763; Thu, 22 Mar 2007 07:50:25 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 12278-08; Thu, 22 Mar 2007 07:50:20 +0100 (CET)
Received: from [10.0.0.174] (dhcp-4009.ietf68.org [130.129.64.9]) by eikenes.alvestrand.no (Postfix) with ESMTP id F3EF5259757; Thu, 22 Mar 2007 07:50:19 +0100 (CET)
Date: Wed, 21 Mar 2007 23:56:05 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: Charles Lindsey <[e-mail]>, [e-mail]
Subject: Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Message-ID: <EBA3768D856B40B90155D690@B50854F0A9192E8EC6CDA126>
In-Reply-To: <[e-mail]>
References: <[e-mail]> <9E8434B41B5C79F191D7574F@[10.0.0.174]> <[e-mail]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

--On 21. mars 2007 14:46 +0000 Charles Lindsey <[e-mail]> wrote:

>
> In <9E8434B41B5C79F191D7574F@[10.0.0.174]> Harald Tveit Alvestrand
> <[e-mail]> writes:
>
>> No second recorded. Issue will not be tracked.
>
> It was seconded by Frank Ellerman, AIUI, on Feb 19th.

Frank, did you intend to second the entering of this issue, or to say why 
you didn't think a change was needed?

It was not clear to me that you were agreeing that it needed tracking.

                  Harald



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHCPvL076387 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LHCP2L076381; Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHC2UZ076333 for <[e-mail]>; Wed, 21 Mar 2007 10:12:23 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster#pop3&clerew&man&ac^uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 46016762.ca94.126 for [e-mail]; Wed, 21 Mar 2007 17:12:02 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2LHC1ji014772 for <[e-mail]>; Wed, 21 Mar 2007 17:12:01 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2LHC1FI014769 for [e-mail]; Wed, 21 Mar 2007 17:12:01 GMT
To: [e-mail]
Xref: clerew local.usefor:24561
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]> <9E8434B41B5C79F191D7574F@[10.0.0.174]>
Date: Wed, 21 Mar 2007 14:46:03 GMT
Lines: 16
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <9E8434B41B5C79F191D7574F@[10.0.0.174]> Harald Tveit Alvestrand <[e-mail]> writes:

>No second recorded. Issue will not be tracked.

It was seconded by Frank Ellerman, AIUI, on Feb 19th.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHCPpl076384 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LHCP8H076380; Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHC3Dw076334 for <[e-mail]>; Wed, 21 Mar 2007 10:12:23 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster*pop3#clerew$man#ac^uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 46016762.13f19.26a for [e-mail]; Wed, 21 Mar 2007 17:12:02 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2LHC2JA014782 for <[e-mail]>; Wed, 21 Mar 2007 17:12:02 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2LHC2RT014779 for [e-mail]; Wed, 21 Mar 2007 17:12:02 GMT
To: [e-mail]
Xref: clerew local.usefor:24562
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: content of Approved header
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]> <1DCF4142A3387EEB2774DC8D@[10.0.0.174]>
Date: Wed, 21 Mar 2007 14:46:47 GMT
Lines: 16
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <1DCF4142A3387EEB2774DC8D@[10.0.0.174]> Harald Tveit Alvestrand <[e-mail]> writes:

>No second recorded. Issue will not be tracked.

Accepted.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHCPSb076388 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LHCPRW076385; Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHC2wx076332 for <[e-mail]>; Wed, 21 Mar 2007 10:12:23 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster$pop3*clerew*man*ac&uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 46016761.21ce.54 for [e-mail]; Wed, 21 Mar 2007 17:12:01 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2LHC1uD014763 for <[e-mail]>; Wed, 21 Mar 2007 17:12:01 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2LHC1uf014760 for [e-mail]; Wed, 21 Mar 2007 17:12:01 GMT
To: [e-mail]
Xref: clerew local.usefor:24560
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: Moderators SHOULD retain existing message-id
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]> <009F44C930814C09E347FA38@[10.0.0.174]>
Date: Wed, 21 Mar 2007 14:44:06 GMT
Lines: 81
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <009F44C930814C09E347FA38@[10.0.0.174]> Harald Tveit Alvestrand <[e-mail]> writes:

>Recorded as issue #1479.

>Which section of USEPRO do you believe it pertains to?

3.8.  Duties of a Moderator

   4.  Moderators are encouraged to retain the Message-ID header field
       if it is valid, and also retain the Date header field unless it
       appears to be stale (72 hours or more in the past) for reasons
       understood by the moderator (such as delays in the moderation
       process) in which case they MAY substitute the current date.  Any
       Injection-Date, Injection-Info, or Xref header fields already
       present (though there should be none) MUST be removed.

I want s/are encouraged to/SHOULD/.

And I am also dubious about that explicit mention of 72 hours - leave it
to the moderator, or else let him insert an Injection-Date (though that is
maybe for the Injection-Date thread)..

>--On 14. februar 2007 14:26 +0000 Charles Lindsey <[e-mail]> 
>wrote:

>>
>> Issue # requested.
>>
>> We discussed this earlier, and I had assumed Frank was going to raise it
>> (maybe he assumes I was). It still needs resolving.
>>
>> The issue is whether this needs normative wording, or whether it is merely
>> a USEAGE matter.
>>
>> The argument for normative wording is that in various situations confusion
>> can be caused if the Message-ID in the published article is not that
>> chosen by the poster. Various such situations have been mentioned, but
>> here are the two main examples:
>>
>> 1. The same message might be both posted and mailed (maybe to some mailing
>> list). If it is essentially the same messqage in either medium, then it
>> ought to have the same Message-ID [1].
>>
>> 2. Some posters keep a record of the Message-IDs of the articles they
>> post, so that their User Agents can flag any replies to them.
>>
>> There are doubltess other weird and unpredictable circumstances which
>> could casue similar problems. OTOH, situations where it is _necessary_ for
>> the moderator to change a Message-ID are much fewer (e.g. he has
>> substantially altered the article before posting it, or he is aware it has
>> already been multi-posted to other non-moderated groups). A "SHOULD" still
>> gives him leeway to change it in such cases.
>>
>> [1] This is separate from the related issue where that mailing list is
>> itself gated into Usenet; this situation is already well covered in our
>> Gayewaying section.
>>
>> --
>> Charles H. Lindsey ---------At Home, doing my own
>> thing------------------------ Tel: [phone] Fax: [phone]
>> Web: http://www.cs.man.ac.uk/~chl Email: [e-mail]      Snail:
>> 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint:
>> 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
>>
>>
>>





-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHCPYf076386 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LHCPlF076382; Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHC4e2076335 for <[e-mail]>; Wed, 21 Mar 2007 10:12:23 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster*pop3$clerew&man&ac&uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 46016763.5ab8.1fa for [e-mail]; Wed, 21 Mar 2007 17:12:03 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2LHC27p014790 for <[e-mail]>; Wed, 21 Mar 2007 17:12:02 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2LHC2qf014787 for [e-mail]; Wed, 21 Mar 2007 17:12:02 GMT
To: [e-mail]
Xref: clerew local.usefor:24563
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: consistency of Injection-Info
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]> <455DAE9C850D8A51E21DAF2C@[10.0.0.174]>
Date: Wed, 21 Mar 2007 14:47:43 GMT
Lines: 16
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <455DAE9C850D8A51E21DAF2C@[10.0.0.174]> Harald Tveit Alvestrand <[e-mail]> writes:

>No second found. Issue will not be tracked.

Accepted.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHCPBD076383 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LHCPCk076379; Wed, 21 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LHC2Q2076331 for <[e-mail]>; Wed, 21 Mar 2007 10:12:23 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster#pop3^clerew$man^ac#uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 46016761.2668.2e for [e-mail]; Wed, 21 Mar 2007 17:12:01 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2LHC0OP014755 for <[e-mail]>; Wed, 21 Mar 2007 17:12:00 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2LHC0un014752 for [e-mail]; Wed, 21 Mar 2007 17:12:00 GMT
To: [e-mail]
Xref: clerew local.usefor:24559
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: ISSUE: Format of batched news in response to sendme
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]> <CB85AAA8DDABB80888FC73E9@[10.0.0.174]>
Date: Wed, 21 Mar 2007 14:31:35 GMT
Lines: 16
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <CB85AAA8DDABB80888FC73E9@[10.0.0.174]> Harald Tveit Alvestrand <[e-mail]> writes:

>No second recorded. Issue will not be entered.

I had already agreed to withdraw that one, following discussion.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAV1XW044002 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 03:31:02 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LAV1Ca044001; Wed, 21 Mar 2007 03:31:01 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUZbc043906 for <[e-mail]>; Wed, 21 Mar 2007 03:30:56 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 60B6025973C; Wed, 21 Mar 2007 11:30:34 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 08686-07; Wed, 21 Mar 2007 11:30:25 +0100 (CET)
Received: from [192.168.1.108] (dhcp-125c.ietf68.org [130.129.18.92]) by eikenes.alvestrand.no (Postfix) with ESMTP id 4312A259731; Wed, 21 Mar 2007 11:30:25 +0100 (CET)
Date: Wed, 21 Mar 2007 11:07:59 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: Charles Lindsey <[e-mail]>, [e-mail]
Subject: Re: ISSUE: Format of batched news in response to sendme
Message-ID: <CB85AAA8DDABB80888FC73E9@[10.0.0.174]>
In-Reply-To: <[e-mail]>
References: <[e-mail]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

No second recorded. Issue will not be entered.

--On 14. februar 2007 14:35 +0000 Charles Lindsey <[e-mail]> 
wrote:

>
> In response to a sendme control message, an article is sent, normally
> using the news transport mechanism (typically UUCP in such cases). I.e.,
> it is on the same "wire" as the sendme control message. Therefore it is in
> order to document the (somewhat bizarre) format, as was done in
> Son-of-1036 and in Usepro-06. It is documented nowhere else.
>
> Note that, as compared to Usepro-06, some brief mention of variants of the
> format for compressed batches would be in order.
>
> --
> Charles H. Lindsey ---------At Home, doing my own
> thing------------------------ Tel: [phone] Fax: [phone]
> Web: http://www.cs.man.ac.uk/~chl Email: [e-mail]      Snail:
> 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint:
> 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
>
>
>






Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUwnd043992 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 03:30:59 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LAUwll043991; Wed, 21 Mar 2007 03:30:58 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUvkh043982 for <[e-mail]>; Wed, 21 Mar 2007 03:30:58 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 818B425971C for <[e-mail]>; Wed, 21 Mar 2007 11:30:57 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 08891-02 for <[e-mail]>; Wed, 21 Mar 2007 11:30:45 +0100 (CET)
Received: from [192.168.1.108] (dhcp-125c.ietf68.org [130.129.18.92]) by eikenes.alvestrand.no (Postfix) with ESMTP id 6A7A925973B for <[e-mail]>; Wed, 21 Mar 2007 11:30:28 +0100 (CET)
Date: Wed, 21 Mar 2007 11:29:33 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: [e-mail]
Subject: Ticket status, USEPRO
Message-ID: <C339D8FF55D9FD5B86C0CCAF@[10.0.0.174]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Based on my Feb 12 summary, and the responses to that, the status of
tickets is as summarized below:

The following tickets are closed since the last message:

#1083: USEPRO 5.3: Rules for generating message-ID
  No change required
#1093: USEPRO 2.3: Supported email addresses
  No change required.

The following tickets will be closed when a new I-D incorporating the fix
appear:

#1413: USEPRO 5.5: ihave/sendme syntax
  Text accepted
#1417: USEPRO 3.4: Injecting-agent modification of message-ID
  Text accepted

The following tickets are open:

#1412: USEPRO 5.3: Cancel newsgroups: Matching
  No support for a change. Unless further discussion occurs, I'm going to
close this with
  "No change required".
#1414: USEPRO 3.2.1: delimiter for multiple Path identities
  No discussion
#1415: USEPRO 3.2.1: Number of path entries per site
  Discussion, but no proposed text.
#1416: USEPRO 3.9: Reinjection and Injection-Date
  Controversial.
#1479: USEPRO: Whether moderators SHOULD retain message-ID
  No discussion. (Issue has been discussed in the past under other
headings, I believe)

When following up, please place the ticket number in the subject line, and
follow up ONE ticket per message, unless the message proposes that two
tickets should be merged.

                     Harald





 



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUvXO043980 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LAUvcM043977; Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUsK7043937 for <[e-mail]>; Wed, 21 Mar 2007 03:30:56 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id CB70A259735; Wed, 21 Mar 2007 11:30:53 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 08686-09; Wed, 21 Mar 2007 11:30:44 +0100 (CET)
Received: from [192.168.1.108] (dhcp-125c.ietf68.org [130.129.18.92]) by eikenes.alvestrand.no (Postfix) with ESMTP id 6AB9025973A; Wed, 21 Mar 2007 11:30:27 +0100 (CET)
Date: Wed, 21 Mar 2007 11:11:36 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: Charles Lindsey <[e-mail]>, [e-mail]
Subject: Re: ISSUE: consistency of Injection-Info
Message-ID: <455DAE9C850D8A51E21DAF2C@[10.0.0.174]>
In-Reply-To: <[e-mail]>
References: <[e-mail]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

No second found. Issue will not be tracked.

--On 14. februar 2007 15:37 +0000 Charles Lindsey <[e-mail]> 
wrote:

>
> Issue # requested
>
> Injecting agents have considerable freedom in choosing which of the
> available Injection-Info parameters to use. However, if they select at
> random from the available possibilities, or in the values they insert in
> the chosen parameter (and especially in 'posting-account', it is going to
> make life difficult for the rest of the net, particularly for those
> wishing to use Injection-Info in killfiles in order to catch people who
> keep morphing their From headers. So they should choose a consistent
> format, and stick to it.
>
> Usepro-06 had the following wording in the Duties of an injecting agent:
>
>    Each injecting agent SHOULD use a consistent form of the Injection-Info
>    header field for all articles originating from the same or similar
>    origins.
>
> That, or something similar, needs to go after Step 11 of section 3.4
>
> --
> Charles H. Lindsey ---------At Home, doing my own
> thing------------------------ Tel: [phone] Fax: [phone]
> Web: http://www.cs.man.ac.uk/~chl Email: [e-mail]      Snail:
> 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint:
> 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
>
>
>






Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUvKP043978 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LAUvcm043973; Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUj6M043914 for <[e-mail]>; Wed, 21 Mar 2007 03:30:56 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id DECAD259731; Wed, 21 Mar 2007 11:30:44 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 08891-01; Wed, 21 Mar 2007 11:30:35 +0100 (CET)
Received: from [192.168.1.108] (dhcp-125c.ietf68.org [130.129.18.92]) by eikenes.alvestrand.no (Postfix) with ESMTP id 8E528259739; Wed, 21 Mar 2007 11:30:26 +0100 (CET)
Date: Wed, 21 Mar 2007 11:09:58 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: Charles Lindsey <[e-mail]>, [e-mail]
Subject: Re: ISSUE: content of Approved header
Message-ID: <1DCF4142A3387EEB2774DC8D@[10.0.0.174]>
In-Reply-To: <[e-mail]>
References: <[e-mail]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

No second recorded. Issue will not be tracked.

--On 14. februar 2007 15:15 +0000 Charles Lindsey <[e-mail]> 
wrote:

>
> Issue # requested.
>
> This was discussed earlier, but never resolved.
>
> It is required that a moderator uses an address identifying himself in the
> Approved header. For control messages, nothing is said (so you could
> legitimately write
>
>    Approved: foobar
>
> which is not particularly helpful. USEFOR says
>
>    The Approved header field indicates the mailing addresses (and
>    possibly the full names) of the persons or entities approving the
>    article for posting. ...
>
> And the USEPRO wording for moderators ensures that requirement is met. I
> want to see the same thing said for group control messages, since senders
> of such message clearly have a "duty" to fulfil that USEFOR requirement.
>
> That is not to say the Approved will be identical to the From (one might
> be a personal name and the other a role address). But it should certainly
> be something that sites considering whether to honour the control message
> should be able to look at and expect to find something familiar.
>
> --
> Charles H. Lindsey ---------At Home, doing my own
> thing------------------------ Tel: [phone] Fax: [phone]
> Web: http://www.cs.man.ac.uk/~chl Email: [e-mail]      Snail:
> 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint:
> 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
>
>
>






Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUvHY043976 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LAUvrb043974; Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUZhh043907 for <[e-mail]>; Wed, 21 Mar 2007 03:30:56 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id EABFE25973E; Wed, 21 Mar 2007 11:30:34 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 08642-10; Wed, 21 Mar 2007 11:30:24 +0100 (CET)
Received: from [192.168.1.108] (dhcp-125c.ietf68.org [130.129.18.92]) by eikenes.alvestrand.no (Postfix) with ESMTP id 97F7525971C; Wed, 21 Mar 2007 11:30:24 +0100 (CET)
Date: Wed, 21 Mar 2007 11:07:26 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: Charles Lindsey <[e-mail]>, [e-mail]
Subject: Re: ISSUE: Moderators SHOULD retain existing message-id
Message-ID: <009F44C930814C09E347FA38@[10.0.0.174]>
In-Reply-To: <[e-mail]>
References: <[e-mail]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Recorded as issue #1479.

Which section of USEPRO do you believe it pertains to?

--On 14. februar 2007 14:26 +0000 Charles Lindsey <[e-mail]> 
wrote:

>
> Issue # requested.
>
> We discussed this earlier, and I had assumed Frank was going to raise it
> (maybe he assumes I was). It still needs resolving.
>
> The issue is whether this needs normative wording, or whether it is merely
> a USEAGE matter.
>
> The argument for normative wording is that in various situations confusion
> can be caused if the Message-ID in the published article is not that
> chosen by the poster. Various such situations have been mentioned, but
> here are the two main examples:
>
> 1. The same message might be both posted and mailed (maybe to some mailing
> list). If it is essentially the same messqage in either medium, then it
> ought to have the same Message-ID [1].
>
> 2. Some posters keep a record of the Message-IDs of the articles they
> post, so that their User Agents can flag any replies to them.
>
> There are doubltess other weird and unpredictable circumstances which
> could casue similar problems. OTOH, situations where it is _necessary_ for
> the moderator to change a Message-ID are much fewer (e.g. he has
> substantially altered the article before posting it, or he is aware it has
> already been multi-posted to other non-moderated groups). A "SHOULD" still
> gives him leeway to change it in such cases.
>
> [1] This is separate from the related issue where that mailing list is
> itself gated into Usenet; this situation is already well covered in our
> Gayewaying section.
>
> --
> Charles H. Lindsey ---------At Home, doing my own
> thing------------------------ Tel: [phone] Fax: [phone]
> Web: http://www.cs.man.ac.uk/~chl Email: [e-mail]      Snail:
> 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint:
> 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
>
>
>






Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUvJg043979 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2LAUvVP043975; Wed, 21 Mar 2007 03:30:57 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2LAUjA3043913 for <[e-mail]>; Wed, 21 Mar 2007 03:30:56 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 4B7A825971C; Wed, 21 Mar 2007 11:30:44 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 08686-08; Wed, 21 Mar 2007 11:30:35 +0100 (CET)
Received: from [192.168.1.108] (dhcp-125c.ietf68.org [130.129.18.92]) by eikenes.alvestrand.no (Postfix) with ESMTP id CACF2259735; Wed, 21 Mar 2007 11:30:25 +0100 (CET)
Date: Wed, 21 Mar 2007 11:08:54 +0100
From: Harald Tveit Alvestrand <[e-mail]>
To: Charles Lindsey <[e-mail]>, [e-mail]
Subject: Re: ISSUE: Possibility to use non-resolvable domain name as path-identity
Message-ID: <9E8434B41B5C79F191D7574F@[10.0.0.174]>
In-Reply-To: <[e-mail]>
References: <[e-mail]>
X-Mailer: Mulberry/4.0.7 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

No second recorded. Issue will not be tracked.

--On 14. februar 2007 14:56 +0000 Charles Lindsey <[e-mail]> 
wrote:

>
> Issue # requested
>
> There were two alternative texts in Usepro-06, of which Russ arbitrarily
> chose one.
>
> This issue originally arose from a suggestion by Harald that a domain name
> (e.g. foo.com) could be used as a <path-identity> even though the only
> records in the DNS were at one level lower (e.g. news.foo.com,
> mail.foo.com).
>
> For sure, people are likely to do it whatever we say, but I regard it as
> an undesirable practice. I therefore wish to make it a SHOULD NOT (or
> rather to say it SHOULD be resolvable, to MX/A/AAAA/CNAME, in the DNS,
> which amounts to the same thing).
>
> Also, I wish to see a mention of RFC 2142, just to draw attention to its
> existence. It does have the status of a draft-standard, although we have
> agreed that we do not intend either to supersede/obsolete/commend it.
> There were two wordings in Usepro-06, of which I would suggest
>
>    NOTE: According to [RFC 2142], the forms "usenet@<path-identity>" and
>    "news@<path-identity>" are common addresses for a news server
>    administrator.
>
> which uses a little bit of each of each. I do not want to take any
> position on the relative merits of "usenet@" and "news@".
>
> --
> Charles H. Lindsey ---------At Home, doing my own
> thing------------------------ Tel: [phone] Fax: [phone]
> Web: http://www.cs.man.ac.uk/~chl Email: [e-mail]      Snail:
> 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K. PGP: 2C15F1A9      Fingerprint:
> 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5
>
>
>






Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2KHCPiu074835 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 20 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2KHCPTn074834; Tue, 20 Mar 2007 10:12:25 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2KHC3Y9074809 for <[e-mail]>; Tue, 20 Mar 2007 10:12:24 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster#pop3&clerew*man$ac&uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 460015e1.8e23.61 for [e-mail]; Tue, 20 Mar 2007 17:12:01 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2KHC0kp003321 for <[e-mail]>; Tue, 20 Mar 2007 17:12:00 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2KHC1Nd003315 for [e-mail]; Tue, 20 Mar 2007 17:12:01 GMT
To: [e-mail]
Xref: clerew local.usefor:24552
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: #1416 Injection-Date - Summary of options
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
Date: Tue, 20 Mar 2007 13:51:24 GMT
Lines: 181
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

I think we have now said everything that is to be said on this, so I have
produced a summary of where I think we have got to.

It lists rules for what various agents are supposed to do with
Injection-Date in various situations. Most of these are, I think, agreed
(and they differ quite a bit from our starting point in the two versions
of USEPRO and in assumptions made at the start of this thread, so that at
least indicates some progress).

There still remains the two conflicting possibilities (labelled IC and IR
in this summary) that we have to choose between, though I think we
understand the consequences of each better than we did.

So please can we read this summary and agree (or not) that it fairly
represents what we are agreed and disagreed on, and maybe tune it if it
doesn't. Hopefully, we can then choose between IC and IR and move on.

-------------------------------------------------------------------------

The following rules are mostly agreed (and indeed some are already in the
present USEPRO). They are all included to give a complete picture. Note that
these are indications of what is supposed to happen, and not wordings ready
for use in the document.

The chief issue is to choose between Rules IC and IR, for which the several
"Good/Bad News" paragraphs need to be studied. Neither Rule is perfect. It
is a question of whether a little pain should be endured for the benefit of
an ultimate gain. In either case, the Sky will not Fall In - the worst is
that an occasional article might be seen twice at some server.

Posting agents:
---------------

P1. MAY omit various headers from proto-articles (specifically Message-ID,
Date, Path).

P2. But should (with a small 's') not omit Date (because we are following
the RFC 2822 idea that Date reflects time of composition).

   NOTE P2: This practice is already widespread, and growing, but it works
   against the man whose injections are significantly delayed after
   composition time; hence P5 below.

P3. MUST include Message-ID and Date if article is to be injected at
multiple injecting agents (and the same at each injecting agent, of course).

P4. MAY include Injection-Date indicating time of actual injection (as
opposed to composition).

P5. SHOULD/should include Injection-Date if injection time is significantly
later than composition time.

   Good News P5: A certain cure for the man with delayed injection problems;

   Bad News P5: Posting agents tend not to get upgraded with such
   new-fangled ideas until long after servers have adopted them. But at
   least this man has a strong incentive to upgrade (once he is aware of the
   possibility).

P6. When injecting at multiple injecting agents, MUST include Injection-Date
which MUST then indicate time of earliest injection at any such agent.


Injecting Agents:
-----------------

I1. MUST insert Date and Message-ID when absent and insert/extend PATH.

I2. MUST/SHOULD remove/rename Injection-Info (plus NNTP-Posting-*, X-Trace,
etc.) and SHOULD provide fresh Injection-Info. MAY reject (reinjected)
articles already containing those headers.

I3. MUST/SHOULD insert Injection-Date when Injection-Date is absent AND
(either Date OR Message-ID is absent).

   NOTE I3: But see IC and IR below for further details.

I4. MUST NOT insert (or alter or delete) Injection-Date if it is already
present.

   NOTE I4: That is to prevent re-acceptance problems arising from
   unanticipated (and unpreventable) "leaks" between small private networks
   and major networks such as Usenet. But see E1 for how a gateway might do
   such tricks.
 
We disagree on the following two possibilities:
 
IC. Injecting agents MUST insert Injection-Date when it is absent (unless
it is apparent that this is a reinjection).

   Good News IC: This provides immediate beneffit to the posting agent which
   has delayed injection, even if P5 is not implemented.

   Bad News IC: Until multi-injecting sites can be assumed to have
   implemented P6 this can cause re-acceptance of articles at later relaying
   agents if one of the multi-injected versions suffers substantial delay en
   route. So the existing behaviour is changed to that extent.
 
IR. Injecting agents MUST NOT insert Injection-Date if both (Date AND
   Message-ID are already present).

   Good News IR: Inverse of the Bad News IC.

   Bad News IR: Posting agents with delayed injection see no benefit
   unless/until they implement P5.


Relaying and Serving Agents:
----------------------------

R1. MUST discard incoming articles whose Injection-Date (or Date if
Injection-Date is absent) predates the earliest articles of which it keeps
record (or if more than 24 hours into the future, etc.).

That "earliest articles of which it keeps record" might be reworded (see my
reply to Frank on March 1st). The essential question is "if this article had
arrived X days ago, would I still be recording that fact?" (which allows for
different retentions for different groups, etc). And maybe it records the
time of arrival, or the original Date header, or whatever (about which I
don't think we care).


Gateways:
---------

And especially Gateways from one Netnews system A to another one B, in
particular from small private one-node networks into Usenet:

G1. MUST retain all headers as received (including in particular Date and
Injection-Date), unless some EXCEPTION applies, as listed below.

G2. MUST hand article to an injecting agent, or else perform all the duties
on an injecting agent (except perhaps insertion of Injection-Date)
themselves and then relay.


EXCEPTIONs:
-----------

These are the tricky bits

E1. An existing Injection-Date MAY be removed, but ONLY IF

a) There is no possibility that another Gateway (direct or indirect) between
those networks exists (which includes all manner of "leaks" by other nodes
in A over which the gatewayer has no control).

b) The article had not been previously gatewayed from B into A (i.e. it had
been originally posted by a user of A).

   NOTE: That covers the man who carries his private server around on a
   laptop, and other users of suck/rpost, but if covers very little else.

   NOTE: It is safer to allow gateways (which hopefully understand the
   circumstances of the particular article) to do any tinkering with
   Injection-Date than for injecting agents to do it; Hence Rule I4.

E2. Other evidence of injection within network A (e.g. Injection-Info,
NNTP-Posting-*, etc) MAY be removed, but ONLY IF

a) The gatewaying is to use an injecting agent which would not otherwise
allow reinjection.

   NOTE: That screams of "cheating", "horrible hacking", etc. So be it.
   Perhaps it is better not mentioned.

E3. Are there any other exceptions we need to mention explicitly?




-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2K0EtGY003471 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 19 Mar 2007 17:14:55 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2K0Etf3003470; Mon, 19 Mar 2007 17:14:55 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2K0EWsO003460 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Mon, 19 Mar 2007 17:14:54 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HTRzZ-0008Px-AQ for [e-mail]; Tue, 20 Mar 2007 01:14:21 +0100
Received: from d252210.dialin.hansenet.de ([80.171.252.210]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Tue, 20 Mar 2007 01:14:21 +0100
Received: from nobody by d252210.dialin.hansenet.de with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Tue, 20 Mar 2007 01:14:21 +0100
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: Where are my Issues?
Date:  Tue, 20 Mar 2007 01:12:55 +0100
Organization:  <URL:http://purl.net/xyzzy>
Lines: 36
Message-ID:  <[e-mail]>
References:  <[e-mail]> <[e-mail]> <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: d252210.dialin.hansenet.de
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey wrote:

> What I don't like is one person with umpteen pseudonyms which
> rotates randomly between them, or which conduct pseudo discussions
> amongst themselves on a group.

Yes, I also don't like this.  But I think your technical solution
of this social problem is a legal nightmare.  If the server admins
want to enforce unique IDs per user they could focus their efforts
on enforcing unique From addresses and/or publishing a clear AUP.

The trolls (in the new sense of it as trollish persons) I've in
mind would probably know or learn fast that they need tricks like
more accounts or more news servers no matter what you try on the
technical side.

OTOH normal users of one server can have legit reasons why they
use an ordinary From address in a programming language group,
and something else in a "critical" group (I hope examples of
such "critical" groups are clear).

Now if the server inserts a unique ID in theory good enough for
"global" (over all groups) killfiling behind their back these 
normal users could be identified as posters in such "critical"
groups.

And IMO the normal users are more important than identifying a
few trolls too stupid to figure Injection-Info parameters out.

Besides implementing killfiles based on Injection-Info with all
its RFC 2231 features isn't straight forward.  I doubt that it
works for your purpose, and I think that it will be harmful for
legit users.

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2JHCUae072177 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 19 Mar 2007 10:12:30 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2JHCUC0072176; Mon, 19 Mar 2007 10:12:30 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2JHC9nt072155 for <[e-mail]>; Mon, 19 Mar 2007 10:12:29 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster&pop3#clerew&man$ac$uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45fec467.a1ae.399 for [e-mail]; Mon, 19 Mar 2007 17:12:07 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2JHC1Vf023698 for <[e-mail]>; Mon, 19 Mar 2007 17:12:07 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2JHC0x5023694 for [e-mail]; Mon, 19 Mar 2007 17:12:00 GMT
To: [e-mail]
Xref: clerew local.usefor:24550
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: Where are my Issues?
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <[e-mail]> <[e-mail]>
Date: Mon, 19 Mar 2007 14:36:42 GMT
Lines: 56
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Frank Ellermann <[e-mail]> writes:

>Charles Lindsey wrote:

>> 4. ISSUE: consistency of Injection-Info
>> 5. ISSUE: Moderators SHOULD retain existing message-id
>[...] 
>> that still leaves no action taken on #4 and #5, and I
>> see that Frank has recently raised that same matter again.

>I recall doing that for #5, but not for #4 as posted in
><http://permalink.gmane.org/gmane.ietf.usenet.format/31621>

Sorry, that sentence was intended to refer to #5 only as regards your
position on it.

>#4 is about the old rat-hole of "privacy vs. kill-file",
>and whether servers should have the ability to enforce
>rules behind the backs of their users.  Your proposal:

>| Each injecting agent SHOULD use a consistent form of the
>| Injection-Info header field for all articles originating
>| from the same or similar origins.

>I don't second this, actually I think SHOULD is a bad idea.

OK, that's an opinion (better than silence).

>It's interesting enough that UAs with more than a minimum
>of self-esteem will offer to forge Dates using 00:00 UTC
>or similar for privacy reasons - relevant in conjunction
>with a no nonsense From address, some servers like GMaNe
>or T-Online (used to) enforce this.

>We shouldn't recommend obscure alternatives to the From-
>address in the form of "public user IDs hidden in the
>Injection-Info in a form still working with killfiles".

If people want to remain anonymous, that is fine. They can adopt a
pseudonym and post from [e-mail]. What I don't like is one person
with umpteen pseudonyms which rotates randomly between them, or which
conduct pseudo discussions amongst themselves on a group. One just needs
some way to be aware when that is happening (not that consistent
Injection-Date will solve all such problems, but consistenly showing a
correct posting-host would help a lot).

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2GH87Xk051634 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 16 Mar 2007 10:08:07 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2GH87dV051633; Fri, 16 Mar 2007 10:08:07 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from relay00.pair.com (relay00.pair.com [209.68.5.9]) by balder-227.proper.com (8.13.5/8.13.5) with SMTP id l2GH85C6051627 for <[e-mail]>; Fri, 16 Mar 2007 10:08:06 -0700 (MST) (envelope-from [e-mail])
Received: (qmail 72758 invoked from network); 16 Mar 2007 17:08:02 -0000
Received: from 216.222.244.107 (HELO ?192.168.2.11?) (216.222.244.107) by relay00.pair.com with SMTP; 16 Mar 2007 17:08:02 -0000
X-pair-Authenticated: 216.222.244.107
Message-ID: <[e-mail]>
Date: Fri, 16 Mar 2007 13:07:53 -0400
From: "Forrest J. Cavalier III" <[e-mail]>
User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616)
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: [e-mail]
Subject: Re: Where are my Issues?
References: <[e-mail]> <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Frank Ellermann wrote:

> I like it when we only discuss one big issue like #1416,
> that's already very complex.  So far the discussion did
> not manage to convince me that more accuracy than the
> date (ignoring the time) is strictly necessary for news
> to survive.

That comment really frosts me.  This is supposed to be
ENGINEERING of a protocol not the lottery.

Is "survival" the criteria we use now to weigh choices on
the draft? Is that the goal you and Charles are trying to reach
here?  (I mean, after figuring out how things work now.)

Sorry. Some things written here in seriousness just make
my jaw drop.  I don't miss John Stanley's vitriol, but
his well written cluebats were a welcome read sometimes.

This WG is too small to finish, I fear, with current
participants.  Russ is trying to be careful, and you and
Charles are just, just, ... not.








Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2GGBGam048011 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 16 Mar 2007 09:11:16 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2GGBGtP048010; Fri, 16 Mar 2007 09:11:16 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2GGBDW0048001 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Fri, 16 Mar 2007 09:11:15 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HSF1M-0003F0-Ij for [e-mail]; Fri, 16 Mar 2007 17:11:12 +0100
Received: from d255126.dialin.hansenet.de ([80.171.255.126]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Fri, 16 Mar 2007 17:11:12 +0100
Received: from nobody by d255126.dialin.hansenet.de with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Fri, 16 Mar 2007 17:11:12 +0100
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: Where are my Issues?
Date:  Fri, 16 Mar 2007 17:01:37 +0100
Organization:  <URL:http://purl.net/xyzzy>
Lines: 52
Message-ID:  <[e-mail]>
References:  <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: d255126.dialin.hansenet.de
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey wrote:

> 4. ISSUE: consistency of Injection-Info
> 5. ISSUE: Moderators SHOULD retain existing message-id
[...] 
> that still leaves no action taken on #4 and #5, and I
> see that Frank has recently raised that same matter again.

I recall doing that for #5, but not for #4 as posted in
<http://permalink.gmane.org/gmane.ietf.usenet.format/31621>

#4 is about the old rat-hole of "privacy vs. kill-file",
and whether servers should have the ability to enforce
rules behind the backs of their users.  Your proposal:

| Each injecting agent SHOULD use a consistent form of the
| Injection-Info header field for all articles originating
| from the same or similar origins.

I don't second this, actually I think SHOULD is a bad idea.

It's interesting enough that UAs with more than a minimum
of self-esteem will offer to forge Dates using 00:00 UTC
or similar for privacy reasons - relevant in conjunction
with a no nonsense From address, some servers like GMaNe
or T-Online (used to) enforce this.

We shouldn't recommend obscure alternatives to the From-
address in the form of "public user IDs hidden in the
Injection-Info in a form still working with killfiles".

That's occultism already covered by USEFOR with not less
than three SHOULDs for the "posting-account" parameter,
plus a privacy link to USEAGE at the end of the section
(not to USEPRO).

Of course Harald and Alexey are busy at the moment, with
lots of drafts (and planned meeting agenda) in various
stages between "almost ready for last call" and "not yet
approved" - I guess nothing will happen with your issues
before April when IETF 68 is finished, and drafts for
resulting minutes have been posted.

I like it when we only discuss one big issue like #1416,
that's already very complex.  So far the discussion did
not manage to convince me that more accuracy than the
date (ignoring the time) is strictly necessary for news
to survive.

Frank





Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2FHC6cs075125 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 15 Mar 2007 10:12:06 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2FHC63d075124; Thu, 15 Mar 2007 10:12:06 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2FHC5jt075118 for <[e-mail]>; Thu, 15 Mar 2007 10:12:05 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster*pop3&clerew^man*ac#uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f97e64.e13c.a6 for [e-mail]; Thu, 15 Mar 2007 17:12:04 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2FHC0lH028370 for <[e-mail]>; Thu, 15 Mar 2007 17:12:00 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2FHC0rR028365 for [e-mail]; Thu, 15 Mar 2007 17:12:00 GMT
To: [e-mail]
Xref: clerew local.usefor:24546
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 15 Mar 2007 16:18:09 GMT
Lines: 35
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:

>> I just read the manpage for rpost, and indeed it recommends such practices.

>> But I cannot see why any legitimate ordinary user would need to do such a
>> thing. Why does he ever need to rpost any article other than one he
>> generated himself on his own system (i.e. why should he want to relay).

>Given how rpost works, they have to do this in order to successfully
>propagate a message that they posted themselves on their own system.

Ah! This is the case of a reinjection from a one-node private network
(which could be regarded as a gateway, as in your Usepro draft). So it is
reasonable to assume such a network has no other hidden 'leaks', and that
is fine. But it will need careful warnings to ensure people do not try
these tricks where they are not in a position to guarantee the absence of
leaks.

And I wish the man page for rpost had made it clearer that this facility
was most certainly NOT to be used for reinjecting articles received from
outside (i.e. such sites should noway be relaying, which is what that man
page seemed to imply).

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2FHC4RE075116 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 15 Mar 2007 10:12:04 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2FHC4R5075115; Thu, 15 Mar 2007 10:12:04 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2FHC2DR075107 for <[e-mail]>; Thu, 15 Mar 2007 10:12:04 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster&pop3^clerew#man&ac*uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f97e61.182c8.54a for [e-mail]; Thu, 15 Mar 2007 17:12:01 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2FHC14I028379 for <[e-mail]>; Thu, 15 Mar 2007 17:12:01 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2FHC1A0028375 for [e-mail]; Thu, 15 Mar 2007 17:12:01 GMT
To: [e-mail]
Xref: clerew local.usefor:24547
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Where are my Issues?
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
Date: Thu, 15 Mar 2007 16:39:14 GMT
Lines: 29
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

On February 14th, I requested five ISSUE numbers:

1. ISSUE: Possibility to use non-resolvable domain name as path-identity
2. ISSUE: content of Approved header
3. ISSUE: Format of batched news in response to sendme
4. ISSUE: consistency of Injection-Info
5. ISSUE: Moderators SHOULD retain existing message-id

Harald requested a seconder for the first three. It is my understanding that
Frank agreed to second #1 on Feb 19th (though maybe as a continuation of
Issue 1093, though I think it is better kept separate).

There has been no seconder for #2, so I accept it is dead.

There was discussion of #3, after which I withdrew it.

But that still leaves no action taken on #4 and #5, and I see that Frank
has recently raised that same matter again.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2E6NGDZ014479 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 13 Mar 2007 23:23:16 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2E6NGx5014478; Tue, 13 Mar 2007 23:23:16 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2E6NFGX014471 for <[e-mail]>; Tue, 13 Mar 2007 23:23:16 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 79A064C3D9 for <[e-mail]>; Tue, 13 Mar 2007 23:23:15 -0700 (PDT)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 59BEC4C239 for <[e-mail]>; Tue, 13 Mar 2007 23:23:15 -0700 (PDT)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 4E0ADE7930; Tue, 13 Mar 2007 23:23:15 -0700 (PDT)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Tue, 13 Mar 2007 20:42:50 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Tue, 13 Mar 2007 23:23:15 -0700
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:

> I just read the manpage for rpost, and indeed it recommends such practices.

> But I cannot see why any legitimate ordinary user would need to do such a
> thing. Why does he ever need to rpost any article other than one he
> generated himself on his own system (i.e. why should he want to relay).

Given how rpost works, they have to do this in order to successfully
propagate a message that they posted themselves on their own system.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2E5FdQQ011321 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 13 Mar 2007 22:15:39 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2E5Fdgv011320; Tue, 13 Mar 2007 22:15:39 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2E5Fbu8011312 for <[e-mail]>; Tue, 13 Mar 2007 22:15:38 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster^pop3*clerew^man*ac#uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f784f8.4ffd.355 for [e-mail]; Wed, 14 Mar 2007 05:15:36 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2E5FaY5019252 for <[e-mail]>; Wed, 14 Mar 2007 05:15:36 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2E5FZc6019249 for [e-mail]; Wed, 14 Mar 2007 05:15:35 GMT
To: [e-mail]
Xref: clerew local.usefor:24544
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Tue, 13 Mar 2007 20:42:50 GMT
Lines: 33
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:
>> Russ Allbery <[e-mail]> writes:

>>> Not when doing this sort of gatewaying.  This is how suck/rpost feeds
>>> pretty much universally work.  This type of gatewaying is common
>>> practice;

>> Sounds horrible! Especially if done deliberately (I can see it arising
>> easily from misconfiguration).

>No, it's deliberate.  It's the only way that rpost can work properly.

I just read the manpage for rpost, and indeed it recommends such practices.

But I cannot see why any legitimate ordinary user would need to do such a
thing. Why does he ever need to rpost any article other than one he
generated himself on his own system (i.e. why should he want to relay).

So I still think it is horrible. What is the intended application where it
is needed?

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2CHHSEu090096 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 12 Mar 2007 10:17:28 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2CHHSlK090095; Mon, 12 Mar 2007 10:17:28 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2CHHROD090088 for <[e-mail]>; Mon, 12 Mar 2007 10:17:27 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id E714E4CA69 for <[e-mail]>; Mon, 12 Mar 2007 10:17:26 -0700 (PDT)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id A3AB14CB9D for <[e-mail]>; Mon, 12 Mar 2007 10:17:26 -0700 (PDT)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 9C0CEE7CFC; Mon, 12 Mar 2007 10:17:26 -0700 (PDT)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Mon, 12 Mar 2007 11:41:19 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Mon, 12 Mar 2007 10:17:26 -0700
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:
> Russ Allbery <[e-mail]> writes:

>> Not when doing this sort of gatewaying.  This is how suck/rpost feeds
>> pretty much universally work.  This type of gatewaying is common
>> practice;

> Sounds horrible! Especially if done deliberately (I can see it arising
> easily from misconfiguration).

No, it's deliberate.  It's the only way that rpost can work properly.

> But it is common for Date to be removed in that practice?

All the setups that I'm aware of retain Message-ID and Date, although I've
not done a lot of investigation and haven't run such a configuration
personally.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2CCC9ec071036 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 12 Mar 2007 05:12:09 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2CCC9hi071035; Mon, 12 Mar 2007 05:12:09 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2CCC72n071028 for <[e-mail]>; Mon, 12 Mar 2007 05:12:08 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster*pop3^clerew^man*ac#uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f54396.cf96.222 for [e-mail]; Mon, 12 Mar 2007 12:12:06 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2CCC2mr002289 for <[e-mail]>; Mon, 12 Mar 2007 12:12:02 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2CCC1dF002286 for [e-mail]; Mon, 12 Mar 2007 12:12:01 GMT
To: [e-mail]
Xref: clerew local.usefor:24542
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Mon, 12 Mar 2007 11:41:19 GMT
Lines: 33
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:
>> Russ Allbery <[e-mail]> writes:
>>> Charles Lindsey <[e-mail]> writes:

>>>> But agents that break the standards (whether new or old) are going to
>>>> cause breakages in any system, including the current one (all they
>>>> have to do at present is to remove the Date header and all the evils
>>>> you have been discussing will promptly ensue).

>>> What existing standard does that behavior break?

>> Current best practice is certainly that changing anything in articles
>> other than in variant headers (Path, Xref, etc) is a big No-No.

>Not when doing this sort of gatewaying.  This is how suck/rpost feeds
>pretty much universally work.  This type of gatewaying is common practice;

Sounds horrible! Especially if done deliberately (I can see it arising
easily from misconfiguration). But it is common for Date to be removed in
that practice?

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5crfK015693 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 9 Mar 2007 22:38:53 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2A5cquk015692; Fri, 9 Mar 2007 22:38:53 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp3.stanford.edu (smtp3.Stanford.EDU [171.67.20.26]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5cqLn015686 for <[e-mail]>; Fri, 9 Mar 2007 22:38:52 -0700 (MST) (envelope-from [e-mail])
Received: from smtp3.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id A8F084C534 for <[e-mail]>; Fri,  9 Mar 2007 21:38:51 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.stanford.edu (Postfix) with ESMTP id 8C7B54C530 for <[e-mail]>; Fri,  9 Mar 2007 21:38:51 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 8828EE7BD5; Fri,  9 Mar 2007 21:38:51 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Fri, 9 Mar 2007 17:44:23 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Fri, 09 Mar 2007 21:38:51 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:
> Russ Allbery <[e-mail]> writes:

>> Why would a significant number of articles arriving with this header
>> cause a relayer to change?  I don't see how the situation is any
>> different for relayers either way.  The only time there will be any
>> pressure on relayers to adopt this is if they're discarding articles
>> because of not having this support, and that again limits us down to
>> only those articles posted by off-line readers.

> If relayers never see any significant number of articles with this "new"
> header, they will say "why should we bother to implement it".

I don't believe the mere existence of the header will cause any change.
Lots of new headers show up and no one does anything about them.  What
will make a difference is specifically the feature that the header
enables, which involves the same number of articles whether injecting
agents are adding it routinely or not.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5b3XS014739 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 9 Mar 2007 22:37:03 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2A5b3as014738; Fri, 9 Mar 2007 22:37:03 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5b24w014731 for <[e-mail]>; Fri, 9 Mar 2007 22:37:02 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 9F9D24C2FB for <[e-mail]>; Fri,  9 Mar 2007 21:37:01 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 812214BFA3 for <[e-mail]>; Fri,  9 Mar 2007 21:37:01 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 698BBE7BD5; Fri,  9 Mar 2007 21:37:01 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Fri, 9 Mar 2007 17:34:00 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Fri, 09 Mar 2007 21:37:01 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:
> Russ Allbery <[e-mail]> writes:
>> Charles Lindsey <[e-mail]> writes:

>>> But agents that break the standards (whether new or old) are going to
>>> cause breakages in any system, including the current one (all they
>>> have to do at present is to remove the Date header and all the evils
>>> you have been discussing will promptly ensue).

>> What existing standard does that behavior break?

> Current best practice is certainly that changing anything in articles
> other than in variant headers (Path, Xref, etc) is a big No-No.

Not when doing this sort of gatewaying.  This is how suck/rpost feeds
pretty much universally work.  This type of gatewaying is common practice;
it doesn't make sense to me to say that it's in violation of best
practice.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5FVTY013247 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 9 Mar 2007 22:15:31 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2A5FVcl013246; Fri, 9 Mar 2007 22:15:31 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5FTn0013234 for <[e-mail]>; Fri, 9 Mar 2007 22:15:30 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster*pop3&clerew^man^ac^uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f23ef0.5751.279 for [e-mail]; Sat, 10 Mar 2007 05:15:28 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2A5FRfo005724 for <[e-mail]>; Sat, 10 Mar 2007 05:15:27 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2A5FR9X005721 for [e-mail]; Sat, 10 Mar 2007 05:15:27 GMT
To: [e-mail]
Xref: clerew local.usefor:24538
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Fri, 9 Mar 2007 17:34:00 GMT
Lines: 26
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:

>> But agents that break the standards (whether new or old) are going to
>> cause breakages in any system, including the current one (all they have
>> to do at present is to remove the Date header and all the evils you have
>> been discussing will promptly ensue).

>What existing standard does that behavior break?

Current best practice is certainly that changing anything in articles
other than in variant headers (Path, Xref, etc) is a big No-No. I am not
sure how explicitly RFC 1036 forbids it, but our news drafts certainly lay
it down as a MUST NOT once an article has been officially injected.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5FVFn013249 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Fri, 9 Mar 2007 22:15:31 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l2A5FVK8013248; Fri, 9 Mar 2007 22:15:31 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-4.gradwell.net (lon-mail-4.gradwell.net [193.111.201.130]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l2A5FTCh013233 for <[e-mail]>; Fri, 9 Mar 2007 22:15:30 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster#pop3&clerew#man$ac*uk) by lon-mail-4.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f23eef.da66.1b8 for [e-mail]; Sat, 10 Mar 2007 05:15:27 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l2A5FSm3005732 for <[e-mail]>; Sat, 10 Mar 2007 05:15:28 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l2A5FRu2005729 for [e-mail]; Sat, 10 Mar 2007 05:15:28 GMT
To: [e-mail]
Xref: clerew local.usefor:24539
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Fri, 9 Mar 2007 17:44:23 GMT
Lines: 44
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:

>> Yes, they will have some reason to be early adopters. But we also need
>> relayers to adopt in order to get the benefit, and they are unlikely to
>> do so unless they see some significant number of articles actually
>> arriving with this header. And that will not happen unless injecting
>> agents are routinely adding it.

>Why would a significant number of articles arriving with this header cause
>a relayer to change?  I don't see how the situation is any different for
>relayers either way.  The only time there will be any pressure on relayers
>to adopt this is if they're discarding articles because of not having this
>support, and that again limits us down to only those articles posted by
>off-line readers.

If relayers never see any significant number of articles with this "new"
header, they will say "why should we bother to implement it".


>> Would there be some future in saying that injecting agents should add
>> it, but not if the current time was suspiciously later than the Date
>> (for some value of "suspiciously" more than 1 day, bnut not much more)?

>Isn't that exactly the case where adding it causes the most damage if
>adding it is wrong?

The longer the cutoff for what is "suspicious", the more damage might
ensue. With most articles, 99.9% of cases there is no difference, as you
say. When the cutoff is 1 day, that might come down to say 97%. Exactly
where a reasonable cutoff would be is an interesting question requiring
discussion, which is why I did not suggest a precise value.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28J7Snk005147 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 8 Mar 2007 12:07:28 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l28J7Smb005146; Thu, 8 Mar 2007 12:07:28 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp2.stanford.edu (smtp2.Stanford.EDU [171.67.20.25]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28J7Q7w005140 for <[e-mail]>; Thu, 8 Mar 2007 12:07:27 -0700 (MST) (envelope-from [e-mail])
Received: from smtp2.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 834304CAA8 for <[e-mail]>; Thu,  8 Mar 2007 11:07:26 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.stanford.edu (Postfix) with ESMTP id 1E7FC4CAB8 for <[e-mail]>; Thu,  8 Mar 2007 11:07:26 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id F0C1AE7914; Thu,  8 Mar 2007 11:07:25 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Thu, 8 Mar 2007 14:35:34 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 08 Mar 2007 11:07:25 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:

> But in the case you mention, the problem arose because of the delayed
> multi-injection. The delay at D (or any other delay due to a slow path
> through the network) merely exaccerbated it. But even after that,
> propagation of the duplicate article is still going to be patchy. I am
> not claiming that no server will ever see it, but that the number off
> occasions on which this causes significant upset will be small enough to
> tolerate when compared with the better propagation of delayed
> single-injections. There is NO 100% solution that will solve ALL of the
> possible situations.

Only if you insist on adding this new feature (or bug fix or what have
you) for off-line readers without making off-line readers change.  If you
remove that (IMO artificial) constraint, there is a 100% solution, namely
supporting the existing protocol while still specifying a transition path
for people to adopt the new header which offers improved behavior in that
particular situation once people have adopted it.

>> None of those methods of reinjection detection work with existing
>> suck/rpost feeds precisely because servers currently try to reject
>> reinjected articles and therefore rpost removes all of those trace
>> headers to ensure that its messages don't look like reinjection.

> But agents that break the standards (whether new or old) are going to
> cause breakages in any system, including the current one (all they have
> to do at present is to remove the Date header and all the evils you have
> been discussing will promptly ensue).

What existing standard does that behavior break?

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28J4Dlf004979 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 8 Mar 2007 12:04:13 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l28J4D4D004978; Thu, 8 Mar 2007 12:04:13 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28J4C8f004971 for <[e-mail]>; Thu, 8 Mar 2007 12:04:12 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 9F4314C5A8 for <[e-mail]>; Thu,  8 Mar 2007 11:04:10 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 6AEA04C3F7 for <[e-mail]>; Thu,  8 Mar 2007 11:04:10 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 5B5E3E7914; Thu,  8 Mar 2007 11:04:10 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Thu, 8 Mar 2007 14:43:41 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 08 Mar 2007 11:04:10 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:

> Yes, they will have some reason to be early adopters. But we also need
> relayers to adopt in order to get the benefit, and they are unlikely to
> do so unless they see some significant number of articles actually
> arriving with this header. And that will not happen unless injecting
> agents are routinely adding it.

Why would a significant number of articles arriving with this header cause
a relayer to change?  I don't see how the situation is any different for
relayers either way.  The only time there will be any pressure on relayers
to adopt this is if they're discarding articles because of not having this
support, and that again limits us down to only those articles posted by
off-line readers.

All the Injection-Date headers that could be added by injecting agents
that are close to existing Date headers are pure noise as far as adoption
goes, since ignoring them and using Date will have the same effective
results in 99.9% of all cases.

> Would there be some future in saying that injecting agents should add
> it, but not if the current time was suspiciously later than the Date
> (for some value of "suspiciously" more than 1 day, bnut not much more)?

Isn't that exactly the case where adding it causes the most damage if
adding it is wrong?

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28HC604048902 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 8 Mar 2007 10:12:06 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l28HC6w6048901; Thu, 8 Mar 2007 10:12:06 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28HC4OX048883 for <[e-mail]>; Thu, 8 Mar 2007 10:12:05 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster$pop3^clerew*man&ac#uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f043e1.2fdb.24b for [e-mail]; Thu,  8 Mar 2007 17:12:01 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l28HC2AQ009896 for <[e-mail]>; Thu, 8 Mar 2007 17:12:02 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l28HC2FB009893 for [e-mail]; Thu, 8 Mar 2007 17:12:02 GMT
To: [e-mail]
Xref: clerew local.usefor:24535
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 8 Mar 2007 14:43:41 GMT
Lines: 43
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:

>> Indeed. User agents are increasingly following the RFC 2822 line, and I
>> think that is both desirable and unstoppable. But one of the
>> consequences of your rule 'R' is that Injection-Date will now hardly
>> ever be added by injecting agents, which someewhat defeats the purpose
>> of having it in the first place.

>Yes, what I'm arguing for is that it be added by posting agents in the
>normal case and only by injecting agents when it's clearly safe to do so.


>This change specifically benefits people who compose off-line and post
>later.  Right now, their messages may be rejected or suffer spotty
>propagation because of the current Date semantics.  They're feeling the
>pain, and the Injection-Date feature is designed for them.  So put the
>burden of implementing the feature on them.  Off-line readers that queue
>posts for later would be early adopters of the Injection-Date feature,
>adding it to their messages if they were concerned that the Date header
>may otherwise be stale.

Yes, they will have some reason to be early adopters. But we also need
relayers to adopt in order to get the benefit, and they are unlikely to do
so unless they see some significant number of articles actually arriving
with this header. And that will not happen unless injecting agents are
routinely adding it.

Would there be some future in saying that injecting agents should add it,
but not if the current time was suspiciously later than the Date (for some
value of "suspiciously" more than 1 day, bnut not much more)?

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28HC612048900 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 8 Mar 2007 10:12:06 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l28HC6IK048899; Thu, 8 Mar 2007 10:12:06 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l28HC45O048884 for <[e-mail]>; Thu, 8 Mar 2007 10:12:05 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster&pop3#clerew^man*ac^uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45f043e1.3120.18b for [e-mail]; Thu,  8 Mar 2007 17:12:01 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l28HC1YR009888 for <[e-mail]>; Thu, 8 Mar 2007 17:12:01 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l28HC1GC009884 for [e-mail]; Thu, 8 Mar 2007 17:12:01 GMT
To: [e-mail]
Xref: clerew local.usefor:24534
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 8 Mar 2007 14:35:34 GMT
Lines: 56
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:
>> Russ Allbery <[e-mail]> writes:

>> Some server D in the middle of the network put articles into a queue and
>> does not release them until some days later.

>> In the meantime, other servers have already flooded the article around
>> D, so that when some server (say X) receives the article from D, it will
>> already have received it days earlier via routes that do not involve D.

>But the one queued at D may have a different, newer Date header because it
>got the article from the other injection path, and therefore it looks like
>a new article since the message ID had been dropped from history but the
>Date header is newer than cutoff.

ITYM "newer Injection-Date header" - let us try to be clear which headers
we are talking about, because altered Date headers, if they occur, would
be equally disruptive on the present network.

But in the case you mention, the problem arose because of the delayed
multi-injection. The delay at D (or any other delay due to a slow path
through the network) merely exaccerbated it. But even after that,
propagation of the duplicate article is still going to be patchy. I am not
claiming that no server will ever see it, but that the number off
occasions on which this causes significant upset will be small enough to
tolerate when compared with the better propagation of delayed
single-injections. There is NO 100% solution that will solve ALL of the
possible situations.


>> There are several ways in which reinjection can be detected, such as the
>> Presence of Injection-Info, NNTP-Posting-Host, NNTP-Posting-Date,
>> X-Trace, 'POSTED' in the Path, or a suspiciously long Path.

>None of those methods of reinjection detection work with existing
>suck/rpost feeds precisely because servers currently try to reject
>reinjected articles and therefore rpost removes all of those trace headers
>to ensure that its messages don't look like reinjection.

But agents that break the standards (whether new or old) are going to
cause breakages in any system, including the current one (all they have to
do at present is to remove the Date header and all the evils you have been
discussing will promptly ensue).

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27LJYcI090306 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 7 Mar 2007 14:19:34 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l27LJYRK090305; Wed, 7 Mar 2007 14:19:34 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27LJXWJ090298 for <[e-mail]>; Wed, 7 Mar 2007 14:19:33 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 25CA94C510 for <[e-mail]>; Wed,  7 Mar 2007 13:19:32 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 06FF54BE80 for <[e-mail]>; Wed,  7 Mar 2007 13:19:32 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 02899E7A0E; Wed,  7 Mar 2007 13:19:32 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Wed, 7 Mar 2007 11:51:11 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Wed, 07 Mar 2007 13:19:31 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:

> Indeed. User agents are increasingly following the RFC 2822 line, and I
> think that is both desirable and unstoppable. But one of the
> consequences of your rule 'R' is that Injection-Date will now hardly
> ever be added by injecting agents, which someewhat defeats the purpose
> of having it in the first place.

Yes, what I'm arguing for is that it be added by posting agents in the
normal case and only by injecting agents when it's clearly safe to do so.

I'm returning again to what I consider to be a fundamental rule of
protocol design around existing protocols: put the burden of new features
on the agent that benefits the most from the feature.  Right now, we're
putting the burden of the Injection-Date feature on injecting agents.
Injecting agents don't care at all, don't benefit from the change, and
aren't in a position to know the intent.

This change specifically benefits people who compose off-line and post
later.  Right now, their messages may be rejected or suffer spotty
propagation because of the current Date semantics.  They're feeling the
pain, and the Injection-Date feature is designed for them.  So put the
burden of implementing the feature on them.  Off-line readers that queue
posts for later would be early adopters of the Injection-Date feature,
adding it to their messages if they were concerned that the Date header
may otherwise be stale.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27LEKwA090124 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 7 Mar 2007 14:14:20 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l27LEKtk090123; Wed, 7 Mar 2007 14:14:20 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp2.stanford.edu (smtp2.Stanford.EDU [171.67.20.25]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27LEJOg090117 for <[e-mail]>; Wed, 7 Mar 2007 14:14:20 -0700 (MST) (envelope-from [e-mail])
Received: from smtp2.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id AC7F84C8CE for <[e-mail]>; Wed,  7 Mar 2007 13:14:19 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.stanford.edu (Postfix) with ESMTP id 650DF4C8BD for <[e-mail]>; Wed,  7 Mar 2007 13:14:17 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 57E32E7A0E; Wed,  7 Mar 2007 13:14:17 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Wed, 7 Mar 2007 11:18:40 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Wed, 07 Mar 2007 13:14:17 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:
> Russ Allbery <[e-mail]> writes:

>> Flooding around them and then having them release articles again later
>> is *exactly* what creates this problem, so of course we're concerned
>> with them.  Anything that creates a delay that causes an article to
>> reappear later can create duplicates unless every copy of the article
>> has consistent Date information, which will not happen if we grant a
>> license to injecting agents to change the effective Date information of
>> articles, either by changing Date or by adding Injection-Date to
>> override it.

> Eh? I don't follow that argument.

> Some server D in the middle of the network put articles into a queue and
> does not release them until some days later.

> In the meantime, other servers have already flooded the article around
> D, so that when some server (say X) receives the article from D, it will
> already have received it days earlier via routes that do not involve D.

But the one queued at D may have a different, newer Date header because it
got the article from the other injection path, and therefore it looks like
a new article since the message ID had been dropped from history but the
Date header is newer than cutoff.

> The case you originally raised did not concern a server causing a delay,
> but an ordinary user wih a suck feed who *reinjected* the article some
> days later. That case might indeed cause a fresh Injection-Date to be
> added.

My point is that the two phenomena reinforce each other, which is why the
argument in your previous message was not complete.  If the reinjected
article was only reinjected a day after the first article, it won't cause
problems in practice *unless* some server D grabs the second article and
sits on it for another six days or so and then kicks it out there, at
which point it falls into the window between cutoff and retention at more
servers.  But such servers like D are a routine occurance.

>> The only way that they can detect this case is to notice that the
>> article has a Date header already.  That's precisely when I think they
>> should not be permitted to insert an Injection-Date.  So you have
>> arrived at exactly my argument, if you also observe that there's no way
>> to tell the difference between a reinjection and a first injection for
>> existing suck/rpost setups (completely apart from the separate question
>> of how they *should* work).

> There are several ways in which reinjection can be detected, such as the
> Presence of Injection-Info, NNTP-Posting-Host, NNTP-Posting-Date,
> X-Trace, 'POSTED' in the Path, or a suspiciously long Path.

None of those methods of reinjection detection work with existing
suck/rpost feeds precisely because servers currently try to reject
reinjected articles and therefore rpost removes all of those trace headers
to ensure that its messages don't look like reinjection.

Remember, I'm talking about how this protocol change interacts with
*existing software*, which was not written to our draft.

> Whether injecting agents then allow reinjection at all is a matter of
> local policy, but if and when they do detect and allow it, then they
> clearly MUST NOT insert a fresh Injection-Date themselves.

I've already argued at length, back when I finished my draft, for why I
don't believe that putting the onus on injecting agents to detect and
negotiate permission to do reinjection is workable in practice.  Hopefully
I won't have to repeat all those arguments.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27CC7kN056971 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 7 Mar 2007 05:12:08 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l27CC7k1056969; Wed, 7 Mar 2007 05:12:07 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27CC6qA056948 for <[e-mail]>; Wed, 7 Mar 2007 05:12:06 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster$pop3&clerew$man$ac*uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45eeac15.3e3.12d for [e-mail]; Wed,  7 Mar 2007 12:12:05 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l27CC3pY007487 for <[e-mail]>; Wed, 7 Mar 2007 12:12:03 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l27CC3vI007481 for [e-mail]; Wed, 7 Mar 2007 12:12:03 GMT
To: [e-mail]
Xref: clerew local.usefor:24530
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]>	<[e-mail]>	<[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Wed, 7 Mar 2007 11:45:15 GMT
Lines: 89
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> "Forrest J. Cavalier III" <[e-mail]> writes:

>Harald Alvestrand wrote:
>> It seems to me that Russ's argument has now come fully around to the 
>> idea that injecting agents should try to find a single date field that 
>> stays constant across all copies of the message, and only filter based 
>> on that.
>> 
>> But making Injection-Date: that field seems to me to be a positively 
>> weird usage of this field, which is, after all, documented as "a 
>> replacement for NNTP-Posting-Date".

NNTP-Posting-Date was never more than a Trace header. Injection-Date was
intended to be far more than that - nothing 'weird' about what was
intended to be its primary purpose. Removing any need for
NNTP-Posting-Date was a convenient by-product.

> And in the case of multiple 
>> injections that travel different paths before hitting a news network, 
>> it's not *possible* to tell the holders of all copies what that 
>> injection-date should be.

It needs to be understood that multiple injections are pretty rare, and
are only done by sites with particular problems/situations/needs and by
people who know what they are about (in my case, for example, it was
because I had a choice of two sites where one was unreliable and the other
was restrictive in what it would allow and might possibly munge in a way
that could break PGPVERIFY).
>> 
>> Should the filtering be done on Date:, ignoring Injection-Date: completely?

>That is what is done now, by existing implementations.

>As I understand it, Charles is willing to accept duplicates and inconsistent 
>propagation during the transition period, when some servers are checking 
>Injection-Date and others Date.

Indeed. No Pain, No Gain!

And I am claiming the Gain is worth the Pain in this instance, and that is
the issue we should be concentrating on.

>I think this is unacceptable and unnecessary.  If people want the benefits,
>there must be a way to get them without degrading existing operation.

But that turns out to be impossible, so we need to assess the various
compromises.

>I think the benefits that people desire are
>    - Better propagation of delayed injected articles
>    - Display of the actual authoring date.

Correct.

>I propose:
>    Agents MUST use the Date header field and not Injection-Date or the
>    deprecated NNTP-Posting-Date to make article staleness checks.  In the case
>    of delayed injection, Posting agents MAY provide copies of the proto-article
>    with a Date:header field that indicates the time of injection (rather than
>    conforming to RFC-2822 time of authoring) only if they also supply a
>    Message-ID field.  Posting agents MAY offer proto-articles to multiple
>    servers at different times only if both fields are present and unaltered in
>    each proto-article posted.  All agents and gateways MUST preserve the
>    contents of both header fields, refusing or not propagating an article rather
>    than altering either field. User-Agents MAY supply and display an
>    "Authoring-Date" header field.

But that either loses you the "Better propagation of delayed injected
articles", or it loses you the "Display of the actual authoring date" (at
least it loses it in Date header which is the only header recipients are
likely to see and understand, and which reading agents are likely to use
as a sorting key).

As I said, there is NO solution which is perfect from all POVs.

The worst that can happen under algorithm 'C' is that readers will
occasionally see duplicates of articles they have already read before. But
no articles will get lost which IMHO would be a worse evil.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27CC7OH056967 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 7 Mar 2007 05:12:07 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l27CC769056966; Wed, 7 Mar 2007 05:12:07 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27CC31U056944 for <[e-mail]>; Wed, 7 Mar 2007 05:12:06 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster^pop3^clerew$man*ac$uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45eeac13.9821.92 for [e-mail]; Wed,  7 Mar 2007 12:12:03 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l27CC3Pi007495 for <[e-mail]>; Wed, 7 Mar 2007 12:12:03 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l27CC3mU007492 for [e-mail]; Wed, 7 Mar 2007 12:12:03 GMT
To: [e-mail]
Xref: clerew local.usefor:24531
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]>
Date: Wed, 7 Mar 2007 11:51:11 GMT
Lines: 34
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Harald Alvestrand <[e-mail]> writes:
>> Charles Lindsey wrote:

>> So 2822 Date: semantics is exactly that - 2822 semantics.

>Which means that the clause in RFC 1036 that says RFC 822 overrides
>doesn't come into play.  In this case, RFC 822 left something unspecified
>and RFC 1036 chose to make it more specific because of a Usenet-specific
>need.  In other words, for once, RFC 1036 played by the rules and RFC 2822
>changed the rules out from under it.

>Charles is correct that not all *client* software was written with this
>understanding, and I exaggerated.  What I meant to say was that all
>*server* software of which I'm aware was written assuming the semantics
>specified in RFC 1036.

Indeed. User agents are increasingly following the RFC 2822 line, and I
think that is both desirable and unstoppable. But one of the consequences
of your rule 'R' is that Injection-Date will now hardly ever be added by
injecting agents, which someewhat defeats the purpose of having it in the
first place.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27CC7qo056964 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 7 Mar 2007 05:12:07 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l27CC7Mh056959; Wed, 7 Mar 2007 05:12:07 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-3.gradwell.net (lon-mail-3.gradwell.net [193.111.201.127]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l27CC3u2056943 for <[e-mail]>; Wed, 7 Mar 2007 05:12:06 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster&pop3*clerew^man$ac&uk) by lon-mail-3.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45eeac12.da3b.334 for [e-mail]; Wed,  7 Mar 2007 12:12:02 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l27CC20D007476 for <[e-mail]>; Wed, 7 Mar 2007 12:12:02 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l27CC2ro007473 for [e-mail]; Wed, 7 Mar 2007 12:12:02 GMT
To: [e-mail]
Xref: clerew local.usefor:24529
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Wed, 7 Mar 2007 11:18:40 GMT
Lines: 64
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:

>> We are not concerned with servers that hold articles in a queue (they
>> are by definition relayers and will soon get flooded around). We are
>> concerned with multi-injecting posting agents/proxies/whatever that
>> maintain such queues.

>Flooding around them and then having them release articles again later is
>*exactly* what creates this problem, so of course we're concerned with
>them.  Anything that creates a delay that causes an article to reappear
>later can create duplicates unless every copy of the article has
>consistent Date information, which will not happen if we grant a license
>to injecting agents to change the effective Date information of articles,
>either by changing Date or by adding Injection-Date to override it.

Eh? I don't follow that argument.

Some server D in the middle of the network put articles into a queue and
does not release them until some days later.

In the meantime, other servers have already flooded the article around D,
so that when some server (say X) receives the article from D, it will
already have received it days earlier via routes that do not involve D.
Therefore, we are concerned only with what happens at X which will see the
article with Date and/or Injection-Date exactly as originally injected
(and noway altered by D).

The case you originally raised did not concern a server causing a delay,
but an ordinary user wih a suck feed who *reinjected* the article some
days later. That case might indeed cause a fresh Injection-Date to be
added.


>> which may be an argument that injecting agents should refuse apparent
>> reinjections, or at least omit inserting Injection-Date in that case.

>The only way that they can detect this case is to notice that the article
>has a Date header already.  That's precisely when I think they should not
>be permitted to insert an Injection-Date.  So you have arrived at exactly
>my argument, if you also observe that there's no way to tell the
>difference between a reinjection and a first injection for existing
>suck/rpost setups (completely apart from the separate question of how they
>*should* work).

There are several ways in which reinjection can be detected, such as the
Presence of Injection-Info, NNTP-Posting-Host, NNTP-Posting-Date, X-Trace,
'POSTED' in the Path, or a suspiciously long Path.

Whether injecting agents then allow reinjection at all is a matter of
local policy, but if and when they do detect and allow it, then they
clearly MUST NOT insert a fresh Injection-Date themselves.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26I3eqM085433 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 6 Mar 2007 11:03:40 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l26I3eIa085432; Tue, 6 Mar 2007 11:03:40 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp3.stanford.edu (smtp3.Stanford.EDU [171.67.20.26]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26I3dc1085426 for <[e-mail]>; Tue, 6 Mar 2007 11:03:40 -0700 (MST) (envelope-from [e-mail])
Received: from smtp3.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id CBD4B4C83D for <[e-mail]>; Tue,  6 Mar 2007 10:03:39 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.stanford.edu (Postfix) with ESMTP id C1F174C847 for <[e-mail]>; Tue,  6 Mar 2007 10:03:37 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id BCB35E790A; Tue,  6 Mar 2007 10:03:37 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Russ Allbery's message of "Tue, 06 Mar 2007 09:57:37 -0800")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Tue, 06 Mar 2007 10:03:37 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Russ Allbery <[e-mail]> writes:
> Harald Alvestrand <[e-mail]> writes:
>> Charles Lindsey wrote:

>>> RFC 1036 is, as usual, vague in its wording. It speaks of when the
>>> article was "originally posted to the network"; it does not make the
>>> clear distinction between "posting" and "injection" that we are now
>>> used to.

> What distinction is that?  I use the terms interchangeably.

Oh, wait, no, I'm sorry.  I see what you're saying.  The same distinction
as between a posting agent and an injection agent.  Although if you use
*exactly* the same meaning with headers, Injection-Date and Injection-Info
are actually Posting-Date and Posting-Info, since they record when and how
the posting agent gave the message to the injecting agent, not when the
injecting agent sent the message to a relaying or serving agent.

Anyway, all of those terms refer to handing the article to other agents
for relaying, not composition times.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26HvcbK084945 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 6 Mar 2007 10:57:38 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l26HvcbA084944; Tue, 6 Mar 2007 10:57:38 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp3.stanford.edu (smtp3.Stanford.EDU [171.67.20.26]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26Hvb9g084938 for <[e-mail]>; Tue, 6 Mar 2007 10:57:38 -0700 (MST) (envelope-from [e-mail])
Received: from smtp3.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 738654C6CB for <[e-mail]>; Tue,  6 Mar 2007 09:57:37 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp3.stanford.edu (Postfix) with ESMTP id 449EF4BDB3 for <[e-mail]>; Tue,  6 Mar 2007 09:57:37 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 34935E790A; Tue,  6 Mar 2007 09:57:37 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Harald Alvestrand's message of "Tue, 06 Mar 2007 15:55:59 +0100")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Tue, 06 Mar 2007 09:57:37 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Harald Alvestrand <[e-mail]> writes:
> Charles Lindsey wrote:

>> RFC 1036 is, as usual, vague in its wording. It speaks of when the
>> article was "originally posted to the network"; it does not make the
>> clear distinction between "posting" and "injection" that we are now
>> used to.

What distinction is that?  I use the terms interchangeably.

I have a hard time interpreting this as meaning anything other than
injection time:

    The "Date" line (formerly "Posted") is the date that the message was
    originally posted to the network.

I don't see how, using regular English meanings of those words, one could
interpret that as composition time.  The phrase "to the network" seems to
me to remove all possible doubt.

>> RFC 822 was silent on the matter. Whether RFC 2822 was merely codifying
>> an existing assumption or introducing a deliberate change I do not
>> know.

> I can speak to that, since I was there.

> DRUMS debated whether Date: should reflect composing time or injection
> time, agreeing that it was unspecified in 822, agreeing that it should
> be defined as one or the other, and landed with a strong (but not
> unanimous) consensus on composing time.

> So 2822 Date: semantics is exactly that - 2822 semantics.

Which means that the clause in RFC 1036 that says RFC 822 overrides
doesn't come into play.  In this case, RFC 822 left something unspecified
and RFC 1036 chose to make it more specific because of a Usenet-specific
need.  In other words, for once, RFC 1036 played by the rules and RFC 2822
changed the rules out from under it.

Charles is correct that not all *client* software was written with this
understanding, and I exaggerated.  What I meant to say was that all
*server* software of which I'm aware was written assuming the semantics
specified in RFC 1036.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26HRCtA083313 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 6 Mar 2007 10:27:12 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l26HRCHJ083312; Tue, 6 Mar 2007 10:27:12 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from TheWorld.com (pcls6.std.com [192.74.137.146]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26HR9JK083306 for <[e-mail]>; Tue, 6 Mar 2007 10:27:11 -0700 (MST) (envelope-from [e-mail])
Received: from shell.TheWorld.com ([e-mail] [192.74.137.71]) by TheWorld.com (8.13.6/8.13.6) with ESMTP id l26HQA4I004146; Tue, 6 Mar 2007 12:26:13 -0500
Received: from shell01.TheWorld.com (localhost.theworld.com [127.0.0.1]) by shell.TheWorld.com (8.13.6/8.12.8) with ESMTP id l26HQA3Y647273; Tue, 6 Mar 2007 12:26:10 -0500 (EST)
Received: from localhost (schlitt@localhost) by shell01.TheWorld.com (8.13.6/8.13.6/Submit) with ESMTP id l26HQ3HY647910; Tue, 6 Mar 2007 12:26:04 -0500 (EST)
X-Authentication-Warning: shell01.TheWorld.com: schlitt owned process doing -bs
Date: Tue, 6 Mar 2007 12:26:03 -0500
From: Dan Schlitt <[e-mail]>
X-X-Sender: [e-mail]
To: Harald Alvestrand <[e-mail]>
cc: Russ Allbery <[e-mail]>, [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]>
Message-ID: <[e-mail]>
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Spam-Status: No, score=-4.3 required=10.0 tests=ALL_TRUSTED,AWL,BAYES_00 autolearn=ham version=3.1.8
X-Spam-Checker-Version: SpamAssassin 3.1.8 (2007-02-13) on pcls6.std.com
X-Virus-Scanned: ClamAV 0.88.4/2745/Tue Mar  6 08:59:40 2007 on pcls6.std.com
X-Virus-Status: Clean
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

That is the conclusion I was coming to after reading all the back and
forth discussion.

/dan

-- 

Dan Schlitt
[e-mail]


On Tue, 6 Mar 2007, Harald Alvestrand wrote:

>
> It seems to me that Russ's argument has now come fully around to the
> idea that injecting agents should try to find a single date field that
> stays constant across all copies of the message, and only filter based
> on that.
>
> But making Injection-Date: that field seems to me to be a positively
> weird usage of this field, which is, after all, documented as "a
> replacement for NNTP-Posting-Date". And in the case of multiple
> injections that travel different paths before hitting a news network,
> it's not *possible* to tell the holders of all copies what that
> injection-date should be.
>
> Should the filtering be done on Date:, ignoring Injection-Date: completely?
>
>                   Harald
>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26F3cAC072007 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 6 Mar 2007 08:03:38 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l26F3cAS072006; Tue, 6 Mar 2007 08:03:38 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from relay03.pair.com (relay03.pair.com [209.68.5.17]) by balder-227.proper.com (8.13.5/8.13.5) with SMTP id l26F3aI0071998 for <[e-mail]>; Tue, 6 Mar 2007 08:03:37 -0700 (MST) (envelope-from [e-mail])
Received: (qmail 98267 invoked from network); 6 Mar 2007 15:03:35 -0000
Received: from 199.224.97.93 (HELO ?192.168.2.11?) (199.224.97.93) by relay03.pair.com with SMTP; 6 Mar 2007 15:03:35 -0000
X-pair-Authenticated: 199.224.97.93
Message-ID: <[e-mail]>
Date: Tue, 06 Mar 2007 10:03:38 -0500
From: "Forrest J. Cavalier III" <[e-mail]>
User-Agent: Mozilla Thunderbird 0.7 (Windows/20040616)
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]>	<[e-mail]>	<[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Harald Alvestrand wrote:
> It seems to me that Russ's argument has now come fully around to the 
> idea that injecting agents should try to find a single date field that 
> stays constant across all copies of the message, and only filter based 
> on that.
> 
> But making Injection-Date: that field seems to me to be a positively 
> weird usage of this field, which is, after all, documented as "a 
> replacement for NNTP-Posting-Date". And in the case of multiple 
> injections that travel different paths before hitting a news network, 
> it's not *possible* to tell the holders of all copies what that 
> injection-date should be.
> 
> Should the filtering be done on Date:, ignoring Injection-Date: completely?

That is what is done now, by existing implementations.

As I understand it, Charles is willing to accept duplicates and inconsistent 
propagation during the transition period, when some servers are checking 
Injection-Date and others Date.

I think this is unacceptable and unnecessary.  If people want the benefits,
there must be a way to get them without degrading existing operation.

I think the benefits that people desire are
    - Better propagation of delayed injected articles
    - Display of the actual authoring date.

I propose:
    Agents MUST use the Date header field and not Injection-Date or the
    deprecated NNTP-Posting-Date to make article staleness checks.  In the case
    of delayed injection, Posting agents MAY provide copies of the proto-article
    with a Date:header field that indicates the time of injection (rather than
    conforming to RFC-2822 time of authoring) only if they also supply a
    Message-ID field.  Posting agents MAY offer proto-articles to multiple
    servers at different times only if both fields are present and unaltered in
    each proto-article posted.  All agents and gateways MUST preserve the
    contents of both header fields, refusing or not propagating an article rather
    than altering either field. User-Agents MAY supply and display an
    "Authoring-Date" header field.







Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26Eu7j5071548 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Tue, 6 Mar 2007 07:56:07 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l26Eu7vw071547; Tue, 6 Mar 2007 07:56:07 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l26Eu5bn071541 for <[e-mail]>; Tue, 6 Mar 2007 07:56:06 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 906892596DD; Tue,  6 Mar 2007 15:51:38 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 07333-06; Tue,  6 Mar 2007 15:51:33 +0100 (CET)
Received: from [127.0.0.1] (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 494862596CA; Tue,  6 Mar 2007 15:51:33 +0100 (CET)
Message-ID: <[e-mail]>
Date: Tue, 06 Mar 2007 15:55:59 +0100
From: Harald Alvestrand <[e-mail]>
User-Agent: Thunderbird 1.5.0.9 (X11/20070104)
MIME-Version: 1.0
To: Charles Lindsey <[e-mail]>
Cc: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey wrote:
> In <[e-mail]> Russ Allbery <[e-mail]> writes:
>
>   
>> Frank Ellermann <[e-mail]> writes:
>>     
>
>   
>>> It's also not acceptable to redefine the 2822-semantics (and now also
>>> USEFOR-semantics) of Date, something has to give, and we've to figure
>>> out what causes the least damage.
>>>       
>
>   
>> Clearly it is acceptable to do that since Usenet has done that since at
>> least the publication of RFC 1036, which redefines Date to something
>> different than RFC 2822 and synonymous with our Injection-Date concept.
>>     
>
> RFC 1036 is, as usual, vague in its wording. It speaks of when the article
> was "originally posted to the network"; it does not make the clear
> distinction between "posting" and "injection" that we are now used to.
>
> RFC 822 was silent on the matter. Whether RFC 2822 was merely codifying an
> existing assumption or introducing a deliberate change I do not know.
I can speak to that, since I was there.

DRUMS debated whether Date: should reflect composing time or injection 
time, agreeing that it was unspecified in 822, agreeing that it should 
be defined as one or the other, and landed with a strong (but not 
unanimous) consensus on composing time.

So 2822 Date: semantics is exactly that - 2822 semantics.

                        Harald



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l266NWEb040162 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 5 Mar 2007 23:23:32 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l266NWuW040161; Mon, 5 Mar 2007 23:23:32 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l266NU6I040154 for <[e-mail]>; Mon, 5 Mar 2007 23:23:31 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 1359C2596CD; Tue,  6 Mar 2007 07:19:04 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 27231-03; Tue,  6 Mar 2007 07:18:59 +0100 (CET)
Received: from [192.168.1.54] (162.80-203-220.nextgentel.com [80.203.220.162]) by eikenes.alvestrand.no (Postfix) with ESMTP id 615B42596C6; Tue,  6 Mar 2007 07:18:59 +0100 (CET)
Message-ID: <[e-mail]>
Date: Tue, 06 Mar 2007 07:23:23 +0100
From: Harald Alvestrand <[e-mail]>
User-Agent: Thunderbird 1.5.0.7 (X11/20060921)
MIME-Version: 1.0
To: Russ Allbery <[e-mail]>
Cc: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]> <[e-mail]>	<[e-mail]>	<[e-mail]>	<[e-mail]> <[e-mail]> <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

It seems to me that Russ's argument has now come fully around to the 
idea that injecting agents should try to find a single date field that 
stays constant across all copies of the message, and only filter based 
on that.

But making Injection-Date: that field seems to me to be a positively 
weird usage of this field, which is, after all, documented as "a 
replacement for NNTP-Posting-Date". And in the case of multiple 
injections that travel different paths before hitting a news network, 
it's not *possible* to tell the holders of all copies what that 
injection-date should be.

Should the filtering be done on Date:, ignoring Injection-Date: completely?

                  Harald

Russ Allbery wrote:
> Charles Lindsey <[e-mail]> writes:
>
>   
>> We are not concerned with servers that hold articles in a queue (they
>> are by definition relayers and will soon get flooded around). We are
>> concerned with multi-injecting posting agents/proxies/whatever that
>> maintain such queues.
>>     
>
> Flooding around them and then having them release articles again later is
> *exactly* what creates this problem, so of course we're concerned with
> them.  Anything that creates a delay that causes an article to reappear
> later can create duplicates unless every copy of the article has
> consistent Date information, which will not happen if we grant a license
> to injecting agents to change the effective Date information of articles,
> either by changing Date or by adding Injection-Date to override it.
>
>   
>> There may be a problem if they reinject to an agent that now adds
>> Injection-Date,
>>     
>
> Exactly.
>
>   
>> which may be an argument that injecting agents should refuse apparent
>> reinjections, or at least omit inserting Injection-Date in that case.
>>     
>
> The only way that they can detect this case is to notice that the article
> has a Date header already.  That's precisely when I think they should not
> be permitted to insert an Injection-Date.  So you have arrived at exactly
> my argument, if you also observe that there's no way to tell the
> difference between a reinjection and a first injection for existing
> suck/rpost setups (completely apart from the separate question of how they
> *should* work).
>
>   



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25N3d6v019035 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 5 Mar 2007 16:03:39 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l25N3dFM019034; Mon, 5 Mar 2007 16:03:39 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp2.stanford.edu (smtp2.Stanford.EDU [171.67.20.25]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25N3bV0019027 for <[e-mail]>; Mon, 5 Mar 2007 16:03:38 -0700 (MST) (envelope-from [e-mail])
Received: from smtp2.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id 71AE94CA59 for <[e-mail]>; Mon,  5 Mar 2007 15:03:37 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.stanford.edu (Postfix) with ESMTP id 4FD064CBD2 for <[e-mail]>; Mon,  5 Mar 2007 15:03:37 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 4B0DFE79DA; Mon,  5 Mar 2007 15:03:37 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Mon, 5 Mar 2007 20:11:56 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Mon, 05 Mar 2007 15:03:37 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:

> We are not concerned with servers that hold articles in a queue (they
> are by definition relayers and will soon get flooded around). We are
> concerned with multi-injecting posting agents/proxies/whatever that
> maintain such queues.

Flooding around them and then having them release articles again later is
*exactly* what creates this problem, so of course we're concerned with
them.  Anything that creates a delay that causes an article to reappear
later can create duplicates unless every copy of the article has
consistent Date information, which will not happen if we grant a license
to injecting agents to change the effective Date information of articles,
either by changing Date or by adding Injection-Date to override it.

> There may be a problem if they reinject to an agent that now adds
> Injection-Date,

Exactly.

> which may be an argument that injecting agents should refuse apparent
> reinjections, or at least omit inserting Injection-Date in that case.

The only way that they can detect this case is to notice that the article
has a Date header already.  That's precisely when I think they should not
be permitted to insert an Injection-Date.  So you have arrived at exactly
my argument, if you also observe that there's no way to tell the
difference between a reinjection and a first injection for existing
suck/rpost setups (completely apart from the separate question of how they
*should* work).

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25MS0MU016523 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 5 Mar 2007 15:28:00 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l25MS0tx016522; Mon, 5 Mar 2007 15:28:00 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25MRwvQ016505 for <[e-mail]>; Mon, 5 Mar 2007 15:27:59 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster^pop3$clerew$man$ac&uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45ec996d.1e95.1c5 for [e-mail]; Mon,  5 Mar 2007 22:27:57 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l25MRvPO009744 for <[e-mail]>; Mon, 5 Mar 2007 22:27:57 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l25MRvX5009739 for [e-mail]; Mon, 5 Mar 2007 22:27:57 GMT
To: [e-mail]
Xref: clerew local.usefor:24520
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]>      <[e-mail]> <[e-mail]>    <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Mon, 5 Mar 2007 21:41:11 GMT
Lines: 29
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Frank Ellermann <[e-mail]> writes:

>Charles Lindsey wrote:

>> It is my understanding that R is (near enough) what Russ is currently
>> proposing (so no guessing ad to what 'C' stands for).

>I don't get it.  Injection agents are supposed to add a Message-ID when
>it's missing.  They can also add a missing Injection-Date.  For strange
>"multiple injection" issues with a significant delay between the first
>and the last injection this isn't ideal.  But for the far more common
>case of messages posted some time after they were composed it's fine.

Indeed. The real issue here is the relative frequency of problems
resulting from multi-injection and those resulting from late inmjection. I
tend to share Frank's opinion that the latter will predominate, certainly
so over the long term, although getting hards fact to support either view
will not be easy.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25MS0HS016528 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 5 Mar 2007 15:28:01 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l25MS0aB016527; Mon, 5 Mar 2007 15:28:00 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25MRw53016506 for <[e-mail]>; Mon, 5 Mar 2007 15:27:59 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster*pop3$clerew*man$ac^uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45ec996e.d50e.cc for [e-mail]; Mon,  5 Mar 2007 22:27:58 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l25MRwjP009752 for <[e-mail]>; Mon, 5 Mar 2007 22:27:58 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l25MRvOj009749 for [e-mail]; Mon, 5 Mar 2007 22:27:57 GMT
To: [e-mail]
Xref: clerew local.usefor:24521
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Mon, 5 Mar 2007 21:47:38 GMT
Lines: 38
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Frank Ellermann <[e-mail]> writes:

>> Maybe we could tackle this "multiple injection" problem with a "SHOULD
>> give up on unavailable servers after some hours of the first successful
>> injection, and ask for user intervention" (because the user might know
>> that the unavailable server is really not connected to any parts of the
>> net where the injections already worked).

That would indeed be a wise precaution, even without USEPRO.

>If we're going to require that user agents change, it would be a far
>easier change to just require that they add Injection-Date than add this
>more complex logic.

I am open as to whether normal posting agents should be encouraged to add
Injection-Date as routine.

But for sure, "normal" posting agents will be the last things to become
fully compliant with our new standard.

However, those who engage in multi-injection, proxies and the like tend
not to use "normal" posting agents (or they interpose other agents between
those posting agents and the actual network). People who do that sort of
thing are more clueful than the average poster, and therefore more likely
to adopt our new ideas.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25MS0j2016526 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Mon, 5 Mar 2007 15:28:00 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l25MS0ji016525; Mon, 5 Mar 2007 15:28:00 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l25MRwPH016504 for <[e-mail]>; Mon, 5 Mar 2007 15:27:59 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster^pop3^clerew^man^ac#uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45ec996d.7dba.3b2 for [e-mail]; Mon,  5 Mar 2007 22:27:57 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l25MRu1r009734 for <[e-mail]>; Mon, 5 Mar 2007 22:27:57 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l25MRuhv009731 for [e-mail]; Mon, 5 Mar 2007 22:27:56 GMT
To: [e-mail]
Xref: clerew local.usefor:24519
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]>
Date: Mon, 5 Mar 2007 20:11:56 GMT
Lines: 111
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Frank Ellermann <[e-mail]> writes:

>> It's also not acceptable to redefine the 2822-semantics (and now also
>> USEFOR-semantics) of Date, something has to give, and we've to figure
>> out what causes the least damage.

>Clearly it is acceptable to do that since Usenet has done that since at
>least the publication of RFC 1036, which redefines Date to something
>different than RFC 2822 and synonymous with our Injection-Date concept.

RFC 1036 is, as usual, vague in its wording. It speaks of when the article
was "originally posted to the network"; it does not make the clear
distinction between "posting" and "injection" that we are now used to.

RFC 822 was silent on the matter. Whether RFC 2822 was merely codifying an
existing assumption or introducing a deliberate change I do not know.

>All the Netnews software written since has been based on that assumption.

But that is clearly untrue, since it seems that 70% (admittedly taken from
a smallish sample) of posting agents are already following the RFC 2822
usage. Moreover, our intention when we invented Injection-Date was to
adopt that RFC 2822 usage (and USEFOR reflects that), as well as to
improve matters when injection was delayed after composition.


>The guarantee is that, if you store a record of message IDs you have seen
>for at least N days from the Date in the article and you reject any
>article with a date older than N days, you can use the contents of the
>history file as an authoritative record for determining whether you can
>safely accept an article without creating a duplicate.

But even if that guarantee occasionally fails, the Sky does mot Fall In,
and in the case we are looking at I cannot see how looping could ensue.
The effect will only be seen by clients of the server concerned, who may
be surprised to see articles (with a different Xref number) that they have
already seen before. The question to be addressed, therefore, is the
disadvantage of the occasional failure of this guarantee against the
advantage of improved propagation of articles injected late with respect
to their composition times.

I tried to analyse these pros and cons in the message in which I
introduced the terms 'C' and 'R' for the two possibilities. It would be
better for people to comment on those detailed arguments rather than for
me to repeat them here.


>The problem exists regardless of the delay.  It is more noticable the
>longer the delay is.  A problem that grows continuously with a variable is
>not a "weasel word."

I do not see how that could be so. If the difference between the injection
times is short (an hour maybe), then no effect will be observed unless
there are severe delays (of the order of 3 days) in propagating one of the
injected copies to some site, and then only if it arrives there within one
hour of that site's cutoff time for its earlier copy of the article.


>The problem occurs when a specific article arrives at A with one Date
>header, expires from A, and then arrives at A again with a different,
>later Date header between the time the history entry is dropped and A's
>cutoff value.

No, the problem only arises if

1. The server in question has been upgraded to use Injection-Date (not Date)
2. The multi-injecting posting agent has not been so upgraded
3. At least one of the injecting agents has been so upgraded

You need to multiply together the probabilites of those events, together
with the probability that the window between the earliest and latest
injections, together with that the delays to that server,
was wide enough to encompass its cutoff point.

Yes, it will happen, buy only occasionally, and less often as more systems
become upgraded.


>Such delays happen routinely.  Servers go down and hold on to articles in
>a queue while they're down until an administrator notices and kicks them.

We are not concerned with servers that hold articles in a queue (they are
by definition relayers and will soon get flooded around). We are concerned
with multi-injecting posting agents/proxies/whatever that maintain such
queues.

>Personal news servers behind suck/rpost feeds pull something down from one
>source, hang on to it for several days, and then reinject it at some other
>point.

Then those servers are effectively relaying plus reinjecting; any articles
they receive with Injection-Date in them will be sent on with that
Injection-Date intact.

There may be a problem if they reinject to an agent that now adds
Injection-Date, which may be an argument that injecting agents should
refuse apparent reinjections, or at least omit inserting Injection-Date in
that case.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l24J4tdU006251 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Sun, 4 Mar 2007 12:04:55 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l24J4tsr006242; Sun, 4 Mar 2007 12:04:55 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp2.stanford.edu (smtp2.Stanford.EDU [171.67.20.25]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l24J4p5q006236 for <[e-mail]>; Sun, 4 Mar 2007 12:04:51 -0700 (MST) (envelope-from [e-mail])
Received: from smtp2.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id E92554C69E for <[e-mail]>; Sun,  4 Mar 2007 11:04:50 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp2.stanford.edu (Postfix) with ESMTP id AC3094C4FF for <[e-mail]>; Sun,  4 Mar 2007 11:04:50 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 97DC9E7A85; Sun,  4 Mar 2007 11:04:50 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Frank Ellermann's message of "Sun, 04 Mar 2007 17:04:28 +0100")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Sun, 04 Mar 2007 11:04:50 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Frank Ellermann <[e-mail]> writes:
> Russ Allbery wrote:

>> It's not acceptable to me to break this fundamental of an existing
>> protocol guarantee in order to introduce a new feature, without which
>> Usenet has survived for many years.

> It's also not acceptable to redefine the 2822-semantics (and now also
> USEFOR-semantics) of Date, something has to give, and we've to figure
> out what causes the least damage.

Clearly it is acceptable to do that since Usenet has done that since at
least the publication of RFC 1036, which redefines Date to something
different than RFC 2822 and synonymous with our Injection-Date concept.
All the Netnews software written since has been based on that assumption.
I think that answers the question of which is more acceptable.

> Obviously I don't get what you consider as "fundamental of an existing
> protocol guarantee".  It would help to specify what this "guarantee" is
> (in the draft), and why it's "fundamental" for the protocol.

The guarantee is that, if you store a record of message IDs you have seen
for at least N days from the Date in the article and you reject any
article with a date older than N days, you can use the contents of the
history file as an authoritative record for determining whether you can
safely accept an article without creating a duplicate.  (If you base
history retention on the length of time from when you saw the article
rather than on the contents of the Date header, you have to add to your
retention time whatever leeway for future Dates you're allowing to be
fully correct.)

To maintain this guarantee requires the property that if you post a
message and provide the Date and Message-ID, the injecting agent will not
change those header fields and thereby potentially introduce a copy of the
article with a different Date header.

That this is not correctly reflected in the draft is one of the
motivations for this long thread that we're currently having.

> - It's also related to "multiple injections" with a significant delay
>   between the first and the last successful injection.  "Significant"
>   is a weasel word.  You claim that something fundamental breaks, and
>   I need an example what that is, so let's pick a random number like
>   "12 hours" for "significant".  Please tune that as needed.

The problem exists regardless of the delay.  It is more noticable the
longer the delay is.  A problem that grows continuously with a variable is
not a "weasel word."

> - Temporarily we can reduce the problem to three points, the "MIP",
>   the server A of the first injection, and the server B of the last
>   injection.  For this simplification to work we've to asume that
>   A and B are peers.

> - AFAIK we can also assume that A relays the article more or less
>   immediately to B.  Let's say A's first Injection-Date is T+0, and
>   A notes the Message-ID with arrival T+0 in its history.  B does
>   the same with T+1.

The problem occurs when a specific article arrives at A with one Date
header, expires from A, and then arrives at A again with a different,
later Date header between the time the history entry is dropped and A's
cutoff value.  This can happen with any variation in Date headers if the
article is delayed exactly the right length of time in getting to A, but
obviously it's more likely if the difference in Date headers is on the
order of days rather than hours.  It's therefore far more likely to happen
with serial multiple injection than with parallel multiple injection, but
it can still happen with parallel multiple injection if an article gets
stuck in a queue due to a server being down for an extended period and
then shows up later.

The fundamental assumption made by the uniqueness algorithm in a news
server is that if its retention time is at least as long as its cutoff
time, any article it receives will either authoritatively be in the
history file or not or will be discarded by its cutoff time.  Allowing
injecting agents to change the Date header breaks this assumption for
existing implementations and means that the history file can no longer be
treated as authoritative.  This breaks the flood-fill algorithm at the
heart of the Netnews protocol.

Such delays happen routinely.  Servers go down and hold on to articles in
a queue while they're down until an administrator notices and kicks them.
Personal news servers behind suck/rpost feeds pull something down from one
source, hang on to it for several days, and then reinject it at some other
point.  People misconfigure their access permissions and their peers queue
until they fix those permissions and then send them all the old articles
that had been queuing.

You cannot assume that A relays the article more or less immediately to B,
as this is simply the normal case and is in no way guaranteed by the
protocol or by the way that Netnews works in practice.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l24G9XVB092955 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Sun, 4 Mar 2007 09:09:33 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l24G9XCc092954; Sun, 4 Mar 2007 09:09:33 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l24G9TVw092948 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Sun, 4 Mar 2007 09:09:32 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HNtH0-0002aM-VB for [e-mail]; Sun, 04 Mar 2007 17:09:23 +0100
Received: from du-001-055.access.de.clara.net ([212.82.227.55]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sun, 04 Mar 2007 17:09:22 +0100
Received: from nobody by du-001-055.access.de.clara.net with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sun, 04 Mar 2007 17:09:22 +0100
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Date:  Sun, 04 Mar 2007 17:04:28 +0100
Organization:  <URL:http://purl.net/xyzzy>
Lines: 96
Message-ID:  <[e-mail]>
References:  <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: du-001-055.access.de.clara.net
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Russ Allbery wrote:

  ["missing" Injection-Date added by injection-agent]
>> for the far more common case of messages posted some time after they
>> were composed it's fine.

> It's not acceptable to me to break this fundamental of an existing
> protocol guarantee in order to introduce a new feature, without which
> Usenet has survived for many years.

It's also not acceptable to redefine the 2822-semantics (and now also
USEFOR-semantics) of Date, something has to give, and we've to figure
out what causes the least damage.

Obviously I don't get what you consider as "fundamental of an existing
protocol guarantee".  It would help to specify what this "guarantee"
is (in the draft), and why it's "fundamental" for the protocol.

One fundamental guarantee I'm aware of is "never ever screw with the
Message-ID", and apparently you think it's okay to make unspecified
exceptions for moderators.

At least I should understand what precisely breaks wrt Injection-Date
when I try hard, so far in vain.  Here are the pieces of this puzzle
as far as I understood it:

- The whole issue is related to "dupes" (articles, i.e. Message-IDs,
  accepted more than once by a server) and "nopes" (articles which
  should arrive but don't make it).

- It's also related to "multiple injections" with a significant delay
  between the first and the last successful injection.  "Significant"
  is a weasel word.  You claim that something fundamental breaks, and
  I need an example what that is, so let's pick a random number like
  "12 hours" for "significant".  Please tune that as needed.

- If a "multiple injection proxy" uses its own unique Injection-Date
  (of say the first successful injection) there should be no problem,
  or rather not the discussed problem, therefore we can ignore this
  case.

- If the "MIP" can inject the article without Injection-Date within
  "12 hours" (or whatever maximal delay you pick as "significant")
  the claimed breakage of different Injection-Dates added by the
  injection-agents probably won't convince me, therefore we can also
  skip this case.

- Finally the "MIP" manages to inject the article without its own
  unique Injection-Date, the injecting-agents add different Injection-
  Dates, and the delay between the first and the last injection is
  more than "12 hours" (or what you pick to show that this is really
  harmful).

- For some serious trouble we have to assume that the first and the
  last injection affect a connected part of the net.  We can ignore
  all other successful injections between first and last, the biggest
  damage should show up with the worst case "12 hours".

- Temporarily we can reduce the problem to three points, the "MIP",
  the server A of the first injection, and the server B of the last
  injection.  For this simplification to work we've to asume that
  A and B are peers.

- AFAIK we can also assume that A relays the article more or less
  immediately to B.  Let's say A's first Injection-Date is T+0, and
  A notes the Message-ID with arrival T+0 in its history.  B does
  the same with T+1.

- "12 hours" (or what you pick) later the "MIP" tries to inject the
  same Message-ID at B.  B has it still in its history and will say
  "no thanks".  No problem.

- Variation:  A was somehow unable to relay the article to B within
  the "12 hours".  Then B will accept it from the "MIP", note its
  arrival time as T+12h in its history, and add a corresponding
  Injection-Date.  After that sooner or later A and B will connect.

- If that's within the retention time of A, say T+7d, then A won't
  accept the Message-ID from B.  Likewise B won't accept the same
  Message-ID from A again within B's retention time, say T+7.5d.

- Because this was a simplification we're actually talking about
  servers A' (where the article arrived via A) and B' (where the
  article arrived via B), and the transports from A to A' and from
  B to B' resp. can take some time.

Now it's your turn (anybody here understanding the problem), what
damage can happen if the A to A' path uses Injection-Date T+0 set
by A, while the B to B' path use T+12h set by B ?

Replacing "12h" and "12 hours" everywhere by any value picked by
Russ to get an instructive example, and using plausible retention
and cut off times at A' and B' (not necessarily identical).

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l23JqZgq012550 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Sat, 3 Mar 2007 12:52:35 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l23JqZfJ012549; Sat, 3 Mar 2007 12:52:35 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l23JqV8x012531 for <[e-mail]>; Sat, 3 Mar 2007 12:52:31 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id A8D574C44A for <[e-mail]>; Sat,  3 Mar 2007 11:52:30 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 708734C0AB for <[e-mail]>; Sat,  3 Mar 2007 11:52:30 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 65353E797D; Sat,  3 Mar 2007 11:52:30 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Frank Ellermann's message of "Sat, 03 Mar 2007 15:41:58 +0100")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Sat, 03 Mar 2007 11:52:30 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Frank Ellermann <[e-mail]> writes:

> I don't get it.  Injection agents are supposed to add a Message-ID when
> it's missing.  They can also add a missing Injection-Date.

I don't agree that they should do this if Date is already present.

> For strange "multiple injection" issues with a significant delay between
> the first and the last injection this isn't ideal.

And, indeed, potentially breaks things.

> But for the far more common case of messages posted some time after they
> were composed it's fine.

It's not acceptable to me to break this fundamental of an existing
protocol guarantee in order to introduce a new feature, without which
Usenet has survived for many years.

> I'd be more worried about "multiple injections" without a Message-ID,
> i.e. something resulting in a "multi-post" after the injection-agents
> added their own (different) Message-IDs.

We've already discussed this and this already won't be allowed.

> Maybe we could tackle this "multiple injection" problem with a "SHOULD
> give up on unavailable servers after some hours of the first successful
> injection, and ask for user intervention" (because the user might know
> that the unavailable server is really not connected to any parts of the
> net where the injections already worked).

If we're going to require that user agents change, it would be a far
easier change to just require that they add Injection-Date than add this
more complex logic.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l23F6Vpu028870 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Sat, 3 Mar 2007 08:06:31 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l23F6Vhh028869; Sat, 3 Mar 2007 08:06:31 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l23F6SFP028863 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Sat, 3 Mar 2007 08:06:30 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HNVoF-0003TC-9w for [e-mail]; Sat, 03 Mar 2007 16:06:07 +0100
Received: from 212.82.251.167 ([212.82.251.167]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sat, 03 Mar 2007 16:06:07 +0100
Received: from nobody by 212.82.251.167 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sat, 03 Mar 2007 16:06:07 +0100
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Date:  Sat, 03 Mar 2007 16:05:01 +0100
Organization:  <URL:http://purl.net/xyzzy>
Lines: 17
Message-ID:  <[e-mail]>
References:  <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: 212.82.251.167
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Russ Allbery wrote:
 
>> Maybe it's better if you define and use a term like "retention time"
>> or "cut off time" as you've done it elsewhere.  I'm not sure what the
>> difference is, "cut off => reject as too old, retention => could be
>> removed from the history" (?)
 
> Right.
 
> I think we're converging on a desire to rephrase things that way and
> introduce those terms.

Fine.  Based on that the definition could note that "retention" isn't
directly related to "expiration", without going into details.

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l23En9OM027855 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Sat, 3 Mar 2007 07:49:09 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l23En9wq027854; Sat, 3 Mar 2007 07:49:09 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l23En4uw027838 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=NO) for <[e-mail]>; Sat, 3 Mar 2007 07:49:08 -0700 (MST) (envelope-from [e-mail])
Received: from list by ciao.gmane.org with local (Exim 4.43) id 1HNVWl-0007ho-4C for [e-mail]; Sat, 03 Mar 2007 15:48:04 +0100
Received: from 212.82.251.167 ([212.82.251.167]) by main.gmane.org with esmtp (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sat, 03 Mar 2007 15:48:03 +0100
Received: from nobody by 212.82.251.167 with local (Gmexim 0.1 (Debian)) id 1AlnuQ-0007hv-00 for <[e-mail]>; Sat, 03 Mar 2007 15:48:03 +0100
X-Injected-Via-Gmane: http://gmane.org/
To: [e-mail]
From: Frank Ellermann <[e-mail]>
Subject:  Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Date:  Sat, 03 Mar 2007 15:41:58 +0100
Organization:  <URL:http://purl.net/xyzzy>
Lines: 37
Message-ID:  <[e-mail]>
References:  <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]>      <[e-mail]> <[e-mail]>    <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Mime-Version:  1.0
Content-Type:  text/plain; charset=us-ascii
Content-Transfer-Encoding:  7bit
X-Complaints-To: [e-mail]
X-Gmane-NNTP-Posting-Host: 212.82.251.167
X-Mailer: Mozilla 3.0 (OS/2; U)
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey wrote:

>>> We disagree on the following two possibilities:

>>> C. Injecting agents MUST insert Injection-Date when it is absent.

>>> R. Injecting agents MUST NOT insert Injection-Date if both (Date AND
>>>    Message-ID are already present).

>> I haven't seen anyone arguing for R. I thought that one was settled
>> in USEFOR.

> It is my understanding that R is (near enough) what Russ is currently
> proposing (so no guessing ad to what 'C' stands for).

I don't get it.  Injection agents are supposed to add a Message-ID when
it's missing.  They can also add a missing Injection-Date.  For strange
"multiple injection" issues with a significant delay between the first
and the last injection this isn't ideal.  But for the far more common
case of messages posted some time after they were composed it's fine.

I'd be more worried about "multiple injections" without a Message-ID,
i.e. something resulting in a "multi-post" after the injection-agents
added their own (different) Message-IDs.

Maybe we could tackle this "multiple injection" problem with a "SHOULD
give up on unavailable servers after some hours of the first successful
injection, and ask for user intervention" (because the user might know
that the unavailable server is really not connected to any parts of the
net where the injections already worked).

The implementors of "multiple injection proxies" need to know that
there's a problem that can't be solved by the injection agents without
dropping the ball for a far more common problem of ordinary 2822 UAs.

Frank




Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225LtMV098934 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 22:21:55 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l225LtP6098933; Thu, 1 Mar 2007 22:21:55 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225Lsr2098927 for <[e-mail]>; Thu, 1 Mar 2007 22:21:54 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id C13194C4BC for <[e-mail]>; Thu,  1 Mar 2007 21:21:53 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 8E9014C098 for <[e-mail]>; Thu,  1 Mar 2007 21:21:53 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 88A43E7D16; Thu,  1 Mar 2007 21:21:53 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Charles Lindsey's message of "Thu, 1 Mar 2007 18:09:44 GMT")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 01 Mar 2007 21:21:53 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Charles Lindsey <[e-mail]> writes:
> Russ Allbery <[e-mail]> writes:

>> Given that development of NNTPRelay has essentially stopped, the
>> Highwinds support forums appear to be down, and Highwinds hasn't issued
>> a press release on their site for nearly four years, I wouldn't hold my
>> breath.  Highwinds did just release new builds last November for the
>> first time in two years, though, so they don't appear to have
>> disappeared entirely.

> Aren't they going to take some action to get into line with RFC 3977?

I doubt it, honestly.  I'm prepared to be favorably surprised, but I think
that Usenet's doing a significant shrinking at the moment and there isn't
a lot of momentum behind such things right now.  My guess is that Usenet
is going to revert to being far more of a hobbiest net than it was, which
has both good and bad aspects, and that momentum on software changes will
only pick up again (if it does) after that's happened.

> And for sure there will be other things they ought to be doing to become
> compliant with USEFOR/USEPRO. OTOH, if there are such a small number of
> such implementations, it is practical for them to be leant on (which is
> less true of news servers in general, and virtually impossible for user
> agents).

One of the reasons why I want the various Usenet standards to be as
backward compatible as possible is that I'd like to work on
fully-conforming software (insofar as I can find time to do so), and I
don't want to be put into a situation where I have to choose between
following the standard and working well with software that's not been
updated.  Because, unfortunately, I think it's likely that most software
will not be updated.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CYk5098422 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 22:12:34 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l225CYeC098421; Thu, 1 Mar 2007 22:12:34 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CRYb098381 for <[e-mail]>; Thu, 1 Mar 2007 22:12:28 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster&pop3*clerew^man#ac*uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45e7b23b.c9d.33 for [e-mail]; Fri,  2 Mar 2007 05:12:27 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l225CRQs005731 for <[e-mail]>; Fri, 2 Mar 2007 05:12:27 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l225CR6q005728 for [e-mail]; Fri, 2 Mar 2007 05:12:27 GMT
To: [e-mail]
Xref: clerew local.usefor:24511
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 1 Mar 2007 19:16:12 GMT
Lines: 58
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Harald Alvestrand <[e-mail]> writes:

>>
>> So let us examine the two competing strategies.
>>
>> I think we are agreed that
>>
>> . Injecting agents MUST insert Date and Message-ID when absent.
>> . Injecting agents MUST insert Injection-Date when Injection-Date is
>> absent AND (either Date OR Message-ID is absent).
>>   
>We have agreement on these 2 points.
>> . Injecting agents MUST NOT insert Injection-Date if it is already
>> present.
>>   
>We do not have agreement on this point.

I think Russ and I are agreed on that. It comes with allowing posting
agents to write it, which has benefits in some situations (e.g.
multi-injection, and "posting agents" which are in reality
proxies/servers/other-special cases).

>> We disagree on the following two possibilities:
>>
>> C. Injecting agents MUST insert Injection-Date when it is absent.
>>
>> R. Injecting agents MUST NOT insert Injection-Date if both (Date AND
>>    Message-ID are already present).
>>   
>I haven't seen anyone arguing for R. I thought that one was settled in 
>USEFOR.

It is my understanding that R is (near enough) what Russ is currently
proposing (so no guessing ad to what 'C' stands for).

>I have not scanned your later argumentation, since I do not agree that 
>this question is open.

This (or rather these) questions are precisely what this issue/thread is
about (though the actual nature of the solutions under consideration has
changed somewhat since it started and hence there is less focus on
reinjection and more on other scenarios that might follow).

My "later argumentation" shought to analyse the consequences of following
'C' or 'R' - in some sense it has to be a compromise because you can
always find a situation in which each of them does not do what you might
ideally like.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CTfm098407 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 22:12:29 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l225CTcj098404; Thu, 1 Mar 2007 22:12:29 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CRLR098378 for <[e-mail]>; Thu, 1 Mar 2007 22:12:28 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster#pop3^clerew*man*ac^uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45e7b239.6326.7c for [e-mail]; Fri,  2 Mar 2007 05:12:25 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l225CQXL005707 for <[e-mail]>; Fri, 2 Mar 2007 05:12:26 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l225CP4p005704 for [e-mail]; Fri, 2 Mar 2007 05:12:25 GMT
To: [e-mail]
Xref: clerew local.usefor:24508
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx>  <[e-mail]>	<[e-mail]>  <[e-mail]>	<[e-mail]>  <[e-mail]>	<[e-mail]>  <[e-mail]> <FC8A5FB7F478A0F84B09D3E0@[172.28.60.203]>
Date: Thu, 1 Mar 2007 18:05:22 GMT
Lines: 50
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <FC8A5FB7F478A0F84B09D3E0@[172.28.60.203]> Harald Tveit Alvestrand <[e-mail]> writes:

>--On 24. februar 2007 11:33 -0800 Russ Allbery <[e-mail]> wrote:

>>
>> Frank Ellermann <[e-mail]> writes:
>>
>>> I don't believe in this "identity = ID + timestamp" theory.  The ID is
>>> the identity, the timestamp is a kludge because histories are limited,
>>> and because accepting stale articles isn't in the interest of the
>>> poster.
>>
>> I think that's a wording distinction without any semantic difference.
>> Kludge or identity, history files are limited and therefore the timestamp
>> is a component of the effective identity of the article.

>I still don't believe that this "effective identity" concept is a Good 
>Thing.

As Russ has said, it is not intended as a concept to be mentioned in the
document. It is not even a particularly good terminology. But or the
limited puspose of the current thread it is good enough, so not much
point in changing it.

>- Since some servers will use Date: anyway, injecting-agents should police 
>injection based on Date:

That is the only bit I do not agree with. For a start, they should
certainly NOT be doing any policing if Injection-Date is already present
(though they might police on the content of that Injection-Date).

But, more to the point, injecting agents do not have information as to
what the rest of the net is doing (though they might usefully take note of
the policies of their local relayers and servers). Policing is really a
matter of local policy for each agent. We seem to be agreed that at least
7 or 8 days stale is quite acceptable for injecting, but even that is
longer than the retention of some relayers (OTOH, such relayers will be
flooded around). So we can't lay down hard and fast rules (though we might
provide _advice_ - probably in USEAGE).

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CTvx098405 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 22:12:29 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l225CT5g098402; Thu, 1 Mar 2007 22:12:29 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CRqL098380 for <[e-mail]>; Thu, 1 Mar 2007 22:12:28 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster$pop3$clerew^man^ac^uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45e7b23a.181fe.99 for [e-mail]; Fri,  2 Mar 2007 05:12:26 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l225CQ5w005723 for <[e-mail]>; Fri, 2 Mar 2007 05:12:27 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l225CQvq005720 for [e-mail]; Fri, 2 Mar 2007 05:12:26 GMT
To: [e-mail]
Xref: clerew local.usefor:24510
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 		<[e-mail]> <[e-mail]> 		<[e-mail]> <[e-mail]> 		<[e-mail]> 		<[e-mail]> 		<[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 1 Mar 2007 19:07:05 GMT
Lines: 64
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Frank Ellermann <[e-mail]> writes:

>Russ Allbery wrote:

>>    3.  It MUST examine the Injection-Date header field or, if absent,
>>        the Date header field, and reject the article if that date
>>        predates the earliest articles of which it keeps record or if
>>        that date is more than 24 hours into the future.  It MAY reject
>>        articles with dates in the future with a smaller margin than 24
>>        hours.

>I'd split that - future timestamps can get their own (small) paragraph:

>3b: It MAY reject articles with dates more than 24 hours in the future.

>Normally magic numbers like "24 hours" need a justification, but here
>it should be obvious that this is about a plausible timezone confusion.
>(If that's not the real reason for "24 hours" better say what it is.)

Essentially, it would be stupid to take action on articles only a few
minutes into the future, because of clock differences, timezone
confusions, etc (my wife's mobile still thinks we are in Britich Summer
Time). Anything over two hours would be quite reasonable.

OTOH, if we allowed things infinitely far into the future, various
scammers would have a field day. But they would need to be days into the
future to have much effect.

So we have to choose sume number between 2 hours and a few days, and for
that "24 hours" seems about right.

>The "articles of which it keeps records" isn't clear.  It's not about
>the oldest available article (that could be months or more), it's the
>oldest unavailable article (only recorded in the history).  Maybe it's
>better if you define and use a term like "retention time" or "cut off
>time" as you've done it elsewhere.  I'm not sure what the difference
>is, "cut off => reject as too old, retention => could be removed from
>the history" (?)

>What I have in mind is a server keeping articles in some groups for
>long periods, e.g. for "answers" groups, and articles in other groups
>only shortly, e.g. in binary groups.

Yes, "of which it keeps record" was meant to imply that the cutoff applied
to articles of that type (e.g. in that group). Essentially, the time for
which records off that article would have been kept if it had arrived on
time.

But if you want the wording clarified (and Russ seems happy to do so),
then that is fine by me. But do note that USEPRO has never included
significant mention of "expiry" at all (though USEAGE does), except in
connection with the Expires header. And we don't want to write USEPRO in
terms of implementation details except where that cannot be avoided.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CTEY098406 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 22:12:29 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l225CT02098403; Thu, 1 Mar 2007 22:12:29 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from lon-mail-1.gradwell.net (lon-mail-1.gradwell.net [193.111.201.125]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l225CRxv098379 for <[e-mail]>; Thu, 1 Mar 2007 22:12:28 -0700 (MST) (envelope-from [e-mail])
Received: from [80.175.135.89] ([80.175.135.89] helo=clerew.man.ac.uk country=GB ident=postmaster&pop3*clerew*man&ac$uk) by lon-mail-1.gradwell.net with esmtpa (Gradwell gwh-smtpd 1.243) id 45e7b23a.4af9.207 for [e-mail]; Fri,  2 Mar 2007 05:12:26 +0000 (envelope-sender <[e-mail]>)
Received: from clerew.man.ac.uk (localhost [127.0.0.1]) by clerew.man.ac.uk (8.13.7/8.13.7) with ESMTP id l225CQ8c005715 for <[e-mail]>; Fri, 2 Mar 2007 05:12:26 GMT
Received: (from news@localhost) by clerew.man.ac.uk (8.13.7/8.13.7/Submit) id l225CQef005712 for [e-mail]; Fri, 2 Mar 2007 05:12:26 GMT
To: [e-mail]
Xref: clerew local.usefor:24509
Path: clerew!chl
From: "Charles Lindsey" <[e-mail]>
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
Message-ID: <[e-mail]>
X-Newsreader: NN version 6.5.2 (NOV)
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 1 Mar 2007 18:09:44 GMT
Lines: 31
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

In <[e-mail]> Russ Allbery <[e-mail]> writes:

>Charles Lindsey <[e-mail]> writes:

>> OTOH, one would hope that it will be those dedicated relayers who will
>> be the early adopters of Injection-Date.

>Given that development of NNTPRelay has essentially stopped, the Highwinds
>support forums appear to be down, and Highwinds hasn't issued a press
>release on their site for nearly four years, I wouldn't hold my breath.
>Highwinds did just release new builds last November for the first time in
>two years, though, so they don't appear to have disappeared entirely.

Aren't they going to take some action to get into line with RFC 3977?

And for sure there will be other things they ought to be doing to become
compliant with USEFOR/USEPRO. OTOH, if there are such a small number of
such implementations, it is practical for them to be leant on (which is
less true of news servers in general, and virtually impossible for user
agents).

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: [phone] Fax: [phone]   Web: http://www.cs.man.ac.uk/~chl
Email: [e-mail]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l21DveXo033322 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 06:57:40 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l21DveVg033321; Thu, 1 Mar 2007 06:57:40 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from smtp1.stanford.edu (smtp1.Stanford.EDU [171.67.22.28]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l21Dvdhu033315 for <[e-mail]>; Thu, 1 Mar 2007 06:57:39 -0700 (MST) (envelope-from [e-mail])
Received: from smtp1.stanford.edu (localhost.localdomain [127.0.0.1]) by localhost (Postfix) with SMTP id B79E64C0C0 for <[e-mail]>; Thu,  1 Mar 2007 05:57:38 -0800 (PST)
Received: from windlord.stanford.edu (windlord.Stanford.EDU [171.64.19.147]) by smtp1.stanford.edu (Postfix) with ESMTP id 8E59D4C070 for <[e-mail]>; Thu,  1 Mar 2007 05:57:38 -0800 (PST)
Received: by windlord.stanford.edu (Postfix, from userid 1000) id 870D8E790A; Thu,  1 Mar 2007 05:57:38 -0800 (PST)
From: Russ Allbery <[e-mail]>
To: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
In-Reply-To: <[e-mail]> (Harald Alvestrand's message of "Thu, 01 Mar 2007 10:16:11 +0100")
Organization: The Eyrie
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 01 Mar 2007 05:57:38 -0800
Message-ID: <[e-mail]>
User-Agent: Gnus/5.110006 (No Gnus v0.6) XEmacs/21.4.19 (linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

Harald Alvestrand <[e-mail]> writes:

>> R. Injecting agents MUST NOT insert Injection-Date if both (Date AND
>>    Message-ID are already present).

> I haven't seen anyone arguing for R. I thought that one was settled in
> USEFOR.

I'm arguing for R so that we don't break the guarantees the protocol
currently makes while introducing a new feature.

USEFOR doesn't really settle anything about Injection-Date, I'm afraid.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>



Received: from balder-227.proper.com (localhost [127.0.0.1]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l219GMYe005822 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 1 Mar 2007 02:16:22 -0700 (MST) (envelope-from [e-mail])
Received: (from majordom@localhost) by balder-227.proper.com (8.13.5/8.13.5/Submit) id l219GMBh005821; Thu, 1 Mar 2007 02:16:22 -0700 (MST) (envelope-from [e-mail])
X-Authentication-Warning: balder-227.proper.com: majordom set sender to [e-mail] using -f
Received: from eikenes.alvestrand.no (eikenes.alvestrand.no [158.38.152.233]) by balder-227.proper.com (8.13.5/8.13.5) with ESMTP id l219GIgi005789 for <[e-mail]>; Thu, 1 Mar 2007 02:16:18 -0700 (MST) (envelope-from [e-mail])
Received: from localhost (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 3C0112596E0; Thu,  1 Mar 2007 10:11:55 +0100 (CET)
Received: from eikenes.alvestrand.no ([127.0.0.1]) by localhost (eikenes.alvestrand.no [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 20145-09; Thu,  1 Mar 2007 10:11:50 +0100 (CET)
Received: from [127.0.0.1] (eikenes.alvestrand.no [127.0.0.1]) by eikenes.alvestrand.no (Postfix) with ESMTP id 613132596DB; Thu,  1 Mar 2007 10:11:49 +0100 (CET)
Message-ID: <[e-mail]>
Date: Thu, 01 Mar 2007 10:16:11 +0100
From: Harald Alvestrand <[e-mail]>
User-Agent: Thunderbird 1.5.0.9 (X11/20070104)
MIME-Version: 1.0
To: Charles Lindsey <[e-mail]>
Cc: [e-mail]
Subject: Re: #1416 Reinjection - an attempted summary, and a suggested resolution
References: <45D2C21C.7090600@xxxxxxxxxxxxx> <[e-mail]> 	<[e-mail]> <[e-mail]> 	<[e-mail]> <[e-mail]> <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-Virus-Scanned: by amavisd-new at alvestrand.no
Sender: [e-mail]
Precedence: bulk
List-Archive: <http://www.imc.org/ietf-usefor/mail-archive/>
List-Unsubscribe: <mailto:[e-mail]?body=unsubscribe>
List-ID: <ietf-usefor.imc.org>

>
> So let us examine the two competing strategies.
>
> I think we are agreed that
>
> . Injecting agents MUST insert Date and Message-ID when absent.
> . Injecting agents MUST insert Injection-Date when Injection-Date is
> absent AND (either Date OR Message-ID is absent).
>   
We have agreement on these 2 points.
> . Injecting agents MUST NOT insert Injection-Date if it is already
> present.
>   
We do not have agreement on this point.
> We disagree on the following two possibilities:
>
> C. Injecting agents MUST insert Injection-Date when it is absent.
>
> R. Injecting agents MUST NOT insert Injection-Date if both (Date AND
>    Message-ID are already present).
>   
I haven't seen anyone arguing for R. I thought that one was settled in 
USEFOR.
> 3.2.7.  Injection-Date
>
>    The Injection-Date header field contains the date and time that the
>    article was injected into the network.  Its purpose is to enable news
>    servers, when checking for "stale" articles, to use a <date-time>
>    that was added by a news server at injection time rather than one
>    added by the user agent at message composition time.
>
>    This header field MUST be inserted whenever an article is injected.
>    However, software that predates this standard does not use this
>    header, and therefore agents MUST accept articles without the
>    Injection-Date header field.
>
>    injection-date  =  "Injection-Date:" SP date-time CRLF
>
>
>    See the remarks under Section 3.1.1 regarding the syntax of
>    <date-time> and the requirements and recommendations to which it is
>    subject.
>
>       NOTE: Since clocks on various agents are not necessarily
>       synchronized, the <date-time> in this header field might not be a
>       later value than that in the Date header field.  Agents MUST NOT
>       alter a pre-existing Date header field when adding an Injection-
>       Date header field.
>
>    This header field is intended to replace the currently-used but
>    undocumented "NNTP-Posting-Date" header field, whose use is now
>    deprecated.
I have not scanned your later argumentation, since I do not agree that 
this question is open.

                 Harald