Re: [port-srv-reg] "xmp" service type and the unified IANA Service Name and Port Number Registry

[Joe Touch <[e-mail]>]

Hi, David,

On 8/24/2011 11:20 AM, David Harrington wrote:
> Hi all,
>
> I think it would be good not to give Bobby potentially conflicting
> advice.
> So is the agreed-upon advice for him to change his current
> direct-to-port mappings to indirection-by-service-name?

I don't think there ought to be a recommendation on this point from the 
IETF per se. I do think that getservbyname() is ubiquitous and its use 
is good - if not preferred - practice.

> (and to use his new name rather than his old name to resolve this
> naming conflict)

Yes - and he's agreed tot that.

> Can we get that advice documented on the TSV wiki (or in the IANA
> registry)? or is this advice in the iana-ports document already?

Neither of those seem necessary - it's out of scope for both the wiki 
and the registry. It would (and IMO should) be in the user-ports doc 
I've been working on, though.

Joe

>
> dbh
>
>> -----Original Message-----
>> From: [e-mail]
>> [mailto:[e-mail]] On Behalf Of Joe Touch
>> Sent: Wednesday, August 24, 2011 12:33 PM
>> To: Stuart Cheshire
>> Cc: Bobby Krupczak; [e-mail]
>> Subject: Re: [port-srv-reg] "xmp" service type and the
>> unified IANA Service Name and Port Number Registry
>>
>>
>>
>> On 8/23/2011 10:41 PM, Stuart Cheshire wrote:
>>> On 23 Aug 2011, at 19:46, Bobby Krupczak wrote:
>> ...
>>>> I use the name in getservbyname()
>>>
>>> My personal view is that getservbyname() is a flawed API with very
>>> limited usefulness. It simply consults the local
>> /etc/services table on
>>> Unix, and, in your case, should always return the fixed
>> constant 5270.
>>> So calling getservbyname("xmp", "tcp") is just a
>> less-efficient way of
>>> writing "htons(5270)".
>>
>> Hi, Bobby,
>>
>> This view is not held by all parties involved in this issue.
>>
>> My view is that getservbyname provides the same level of indirection
>
>> inside a host that SRV records provide between hosts. In specific,
>> modifications of the /etc/services tables does occur and is
>> valid. As a
>> result, I would not suggest that you change to using the port number
>
>> directly.
>>
>> My position, FWIW, is that IANA assignments were the only
>> official ones,
>> and that they will take precedent - that is what I have already
>> recommended to IANA. Whether the SRV name or port name is
>> changed, there
>> are changes to code required.
>>
>> The goal of this exchange is to minimize the impact of resolving the
>
>> unification of IANA port names and SRV names. My
>> recommendation to IANA
>> is that IANA assignments should take precedence if a true
>> conflict remains.
>>
>> On one final point:
>>
>>> And, getting back to your "xmp" example, "xmp" isn't even
>> in the /etc/services table on OS X
>> ...
>>
>> Stuart - perhaps you can encourage Apple to update their
>> /etc/services
>> to track the IANA list more closely. Regardless of similar flaws in
>> other OSes, this list should be updated regularly.
>>
>> Joe Touch
>> IANA Port Review Team Lead
>>
>> _______________________________________________
>> Port-srv-reg mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/port-srv-reg
>>
>
> _______________________________________________
> Port-srv-reg mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/port-srv-reg