Re: Injecting agents and From (was: Re: Protocol changes in draft-allbery-usefor-usepro-00)

["Charles Lindsey" <[e-mail]>]

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

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

>>>> 3. [-1] (3.3.1,3.4,3.9.2) From not omittable in proto-article

>>>> There is existing usage where the injecting agent fills in From header
>>>> (not possible in NNTP, of course)

>>> Intentional change.

>>> What injecting agent supports this?

>> INN apparently (see below).

>> Newsreaders such as rn, trn, nn, and others of that generation had (and
>> still have) the capability to interact directly with the newspool
>> (either on the same host, or more likely NFS mounted from some server)
>> rather than going via NNTP (which did not exist when they were first
>> written). They injected articles by calling a program 'inews' which, in
>> the absence of an explicit From:, assumed the poster was the user who
>> had called 'inews'.

>inews as distributed with INN or the common news readers is a (part of a)
>posting agent, not an injecting agent.

Ah! So it boils down to a matter off semantics. Surely, when the first
newsreaders were written (RN and earlier, before NNTP came on the scene),
their final act when posting an article was to hand it off to 'inews',
which came with the software of the server (which would be BNews in those
days). What does the 'i' in 'inews' stand for, if not for 'inject'. For
sure, people using UNIX never configured individual pieces of software
with their own name and id, because they naturally assumed that the system
would pick it up from /etc/passwd and the fact that they were logged in as
their own id.

>  You can see this by observing what
>actions it takes and what agent it talks to.  It sends messages to an
>injecting agent via POST and expects that agent to do the injection;

Not so. 'inews' existed long before the POST command was invented. AFAICT
from the O'Reilly Book, which was the definitive tutorial on administering
Usenet from 1986 onwards, the principal interfaces to BNews were 'inews'
and 'rnews'. Inews had an amazing collection of parameters (you could
even create groups locally with it or issue control messages).

>Our agent distinctions are largely based on how INN and similar news
>servers work.  C News, being a much earlier implementation, may not have
>the same distinctions, or they may be much less clear.  Some
>implementations written prior to our standard will combine different
>agents into one program, so it's possible that in C News inews combines
>functions of a posting agent and an injecting agent.

In CNews, 'inews' IS the injecting agent, which performs all the relevant
checks, sends it to the moderator if needed, and otherwise puts it into
the input queue alongside stuff received via 'rnews', whence it eventually
gets stored locally and relayed onwards. If stuff arrives via the NNTP
POST command, then it merely generates a call on 'inews'. It is certainly
not regarded as a posting agent (a crude posting agent 'readnews' is
provided - not really suited to serious newsreading) and there is a script
'postnews' which prompts you for Newsgroup and Subject, lets you use your
favourite editor to construct the article (adding other headers, even
From:, as you wish), and finally calls 'inews'. So 'inews' is not regarded
in any sense as a posting agent, although it would likely be called
directly by scripts which generated articles automatically (such as
modbots).

For sure, fashions have changed since then. But I am sure there are still
people using updated versions of those early newsreaders (such as nn and
trn) which still call inews, taking advantage of the automatic fillng in
of From, and those will all work happily with Bnews, Cnews, INN, and for
all I know any other newsservers which provide the 'inews' interface.

So there is no reason not to recognise that, even if the usage is not so
common as it once was. The wording I put into USEPRO (when discussing
headers that could be omitted from proto-articles) was:

"...(and even From if the particular injecting agent can derive that
information from other sources.)"

which seems to me an accurate description of the situation I have
described. I see no reason why a similar wording should not continue to be
used.

>INN's inews used to redundantly perform a few (although not all by any
>means) of the functions of an injecting agent, such as mailing posts to
>the moderators of moderated newsgroups.

Sure, 'inews' used to have all sorts of Bells and Whistles, and even CNews
provides a cutdown version 'injnews' which it recommends for routine use.

-- 
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