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

Julien ÉLIE julien at
Tue Nov 3 18:17:26 UTC 2009

Hi Charles,

>> If I understand well the new syntax in USEFOR, the following message-IDs
>> are invalid:
>> <"test"@test>
>> <te"st at test>
>> <test@[te\st]>
>> <test@[te\\st]>
>> whereas the following message-ID is valid:
>> <test@[te"st]>
> Correct. And the same is true in RFC 5322 also.

INN is currently only seeing the second one as invalid.  The four other
are considered to be valid.

CHECK <"test"@test>
238 <"test"@test>
CHECK <te"st at test>
438 <te"st at test>
CHECK <test@[te\st]>
238 <test@[te\st]>
CHECK <test@[te\\st]>
238 <test@[te\\st]>
CHECK <test@[te"st]>
238 <test@[te"st]>
CHECK <bad mid at test>
438 <bad mid at test>

Note that 438 is sent (refusal to receive the article) instead of 501
(syntax error) because of backwards compatibility with existing news
servers, contrary to what RFC 3977 suggests.

> But some strange variants
> such as <" at test> are allowed in the present USEFOR syntax.

INN currently considers it to be invalid.  So that's fine because
the new USEFOR syntax disallows it.

CHECK <" at test>
438 <" at test>

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.

Julien ÉLIE

« La différence entre un chanteur et une paire de chaussures est que le chanteur
  doit partir avant de lasser. La paire de chaussures, il vaut mieux les lacer
  avant de partir. » (Philippe Geluck) 

More information about the inn-workers mailing list