Re: New syntax on message-IDs (was Re: AUTH48 is stuck)

Hi Charles,

> I am surprised that
> INN bothers to test them at all.  The normal implementation should just
> extract eveything within <...>, and see if it is matched in the history
> file.

Because first of all we check the syntax of NNTP commands:

IHAVE <te"st@test>
435 Syntax error in message-ID

(We *reject* the message with 435, not 501.)


which is also done for readers:

ARTICLE bad-id
501 Syntax error in message-ID

ARTICLE 12a
501 Syntax error in article number



Secondly, the Message-ID: header will be checked in the message:

IHAVE <valid@mid>
335 Send it
From: "Demo User" <[email protected]>
Newsgroups: fr.test
Subject: I am just a test article
Date: 5 Nov 2009 04:38:40 -0500
Path: me
Message-ID: nasty-one!

Test.
.
437 Bad "Message-ID" header




RFC 3977:

   If the server is also conforming
   to another specification that contains a definition of message-id
   compatible with this one, the server SHOULD use those message-ids.

I also see in USEPRO that:

   [A relaying agent] MAY reject any article that contains header
   fields that do not have valid contents.

so I believe that's OK.



>>CHECK <"test"@test>
>
> CHECK is not a standard NNTP command BTW.

The result would have been the same with IHAVE.  See my examples above.



>>Is it OK for news servers to switch to the new USEFOR syntax right
>>now?  or would it cause problems?
>>We can use that new syntax, starting with INN 2.5.2.
>
> Yes, because in practice the awkward cases will never arise.

OK, thanks.

-- 
Julien ÉLIE

« La science consiste à passer d'un étonnement à l'autre. » (Aristote)