Re: Bugs in PGnus

[Brian May <bmay@csse.monash.edu.au>]

>>>>> "Kai" == Kai Großjohann <Kai.Grossjohann@CS.Uni-Dortmund.DE> writes:

    Kai> Have you tried setting nntp-record-commands to t, then
    Kai> compare traces of the communication between Gnus and the NNTP
    Kai> server at uni and at home?

Here is what I get when I try to fetch an article (by message-id)
at Monash Uni:

19991125T093917.704 newsserver.cc.monash.edu.au HEAD <38385A79.8410721@mcn.net>
19991125T093917.902 newsserver.cc.monash.edu.au ARTICLE <38385A79.8410721@mcn.net>

This is what happens when PGnus knows the message number:
19991125T093927.248 newsserver.cc.monash.edu.au ARTICLE 34480

At home, the first two commands would look more like (edited to
look like the expected form):
19991125T093917.704 newsserver.cc.monash.edu.au HEAD <38385A79.8410721@mcn.net>
19991125T093917.902 newsserver.cc.monash.edu.au ARTICLE 34479

ie PGnus is able to find the message number just by looking at the
header. My guess is it does this via the Xref header. So, the
obvious problem is that PGnus cannot interpret the Xref header here
at Uni - I don't know why, it looks OK to me:

Xref: towncrier.cc.monash.edu.au rec.aviation.ifr:52840 rec.aviation:34480

Hang on... Perhaps PGnus is confused with the "towncrier.cc.monash.edu.au"
bit - my newsserver is setup as "newserver.cc.monash.edu.au":

[504] [dec16:bmay] ~ >host newsserver.cc.monash.edu.au 
newsserver.cc.monash.edu.au has address 130.194.1.90
newsserver.cc.monash.edu.au mail is handled (pri=20) by mx.cc.monash.edu.au
newsserver.cc.monash.edu.au mail is handled (pri=10) by towncrier.cc.monash.edu.au
[505] [dec16:bmay] ~ >host towncrier.cc.monash.edu.au
towncrier.cc.monash.edu.au has address 130.194.1.90
towncrier.cc.monash.edu.au mail is handled (pri=20) by mx.cc.monash.edu.au
towncrier.cc.monash.edu.au mail is handled (pri=10) by towncrier.cc.monash.edu.au

(note: they both have the same IP address).

However, I tried changing gnus-select-method to use towncrier, but
it didn't help.

Is there anyway to debug why PGnus doesn't like the Xref
header (assuming my assumptions are correct)?
-- 
Brian May <bmay@csse.monash.edu.au>