pan-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Pan-users] Multipart pic display


From: Per Hedeland
Subject: [Pan-users] Multipart pic display
Date: Fri, 4 Aug 2006 11:00:51 +0200 (CEST)

Hi,

I've noticed that for some pictures posted as multiple parts, pan will
fail to display the complete picture - seems it only shows what is
contained in the first part, with the rest black or occasionally
containing garbled parts of other pictures (i.e. smells of uninitialized
memory).

If I "save attachments" for such a post, it always gets the complete
picture, i.e. it seems the logic for combining the multiple parts
differs between display and save. Also it doesn't happen for *all*
multipart posts - I don't know what the "trigger" is, but here are two
examples (totally random, only hunted down for the purpose of this
message:-) from alt.binaries.pictures.wallpaper: Both are simple/dumb
split uuencoding, which I guess is pretty common, but one fails to
display while the other works fine:

Fails:

$ egrep '^(Subject|Message-ID)' pan-multipart-bad
Subject: (4_Harley_Fan enjoy a few walls - [orange county choppers 
wallpaper.jpg (1/3) -- [file 18 of 18) - 908 KB
Message-ID: <address@hidden>
Subject: (4_Harley_Fan enjoy a few walls - [orange county choppers 
wallpaper.jpg (2/3) -- [file 18 of 18) - 908 KB
Message-ID: <address@hidden>
Subject: (4_Harley_Fan enjoy a few walls - [orange county choppers 
wallpaper.jpg (3/3) -- [file 18 of 18) - 908 KB
Message-ID: <address@hidden>

Works:

$ egrep '^(Subject|Message-ID)' pan-multipart-ok 
Subject: BitTorrent Biking - Aspen Ride, Telluride, Colorado.jpg (1/2)
Message-ID: <address@hidden>
Subject: BitTorrent Biking - Aspen Ride, Telluride, Colorado.jpg (2/2)
Message-ID: <address@hidden>

(And no, the difference isn't two-vs-more parts, two-part posts may fail
to display too.)

Is this a known/unavoidable limitation of the display code, or just a
plain bug? (Hoping for the latter...:-) FWIW, I've seen this both in
0.14.x and 0.10x (the above example was with 0.106), though in 0.14.x I
had applied a patch for improved content-type guessing (was posted to
the list) and so thought that this might be the culprit, but apparently
not.

--Per Hedeland




reply via email to

[Prev in Thread] Current Thread [Next in Thread]