vm-reply-include-text with supercite broken in8.0.12
Bug #322735 reported by
muewi
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
VM |
Fix Released
|
Medium
|
Ulrich Müller | ||
Gentoo Linux |
Fix Released
|
Medium
|
Bug Description
I use GNU emacs 22.1.1 w/ supercite 3.54 (which is bundled with that emacs version).
When I tried to switch to vm-8.0.12, R (vm-reply-
[ATTACHEMENT: nil ...]
and then barks that the message format does not comply with RFC 822.
This must be caused by some change introduced after vm-8.0.9, which is the version I used before and switched back to.
mail-citation-hook is set to (sc-cite-original), mail-yank-hook is nil.
Wilhelm
Changed in vm: | |
status: | Fix Committed → Fix Released |
Changed in viewmail: | |
status: | New → Fix Released |
Changed in gentoo: | |
status: | Unknown → Fix Released |
Changed in vm: | |
milestone: | 8.1.0b → 8.1.0 |
tags: | added: reply |
Changed in gentoo: | |
importance: | Unknown → Medium |
To post a comment you must log in.
I also use VM with supercite (but in XEmacs, 21.4.21 as packaged by Debian), and I also found that supercite stopped working correctly when I upgraded to VM 8.0.12. The symptoms aren't exactly the same. For instance, I've never seen a buffer with just an attachment button line. But I did see the error about a message format not conforming to RFC 822. Other symptoms I've seen include supercite not being able to guess a citation name (substituting "Anon"), and the reply buffer getting left in a weird state where undos are disabled and filladapt doesn't seem to work, even if I load it explicitly (I have it in mail-send-hook).
I think all of my problems are somehow caused by the change in BZR revision 453.6.58, "Yanking of messages uses the same MIME decoding as the presentation now.". This is the same one that added vm-mime- yank-attachment s, but it also got rid of a lot of other code in vm-yank-message, including some things whose comments suggest they were added explicitly for supercite's benefit.
Gentoo bug #256886 looks related (http:// bugs.gentoo. org/show_ bug.cgi? id=256886). The patch attached there changes the same region that changed in the problematic revision, and adds code to copy message headers into the buffer even in the MIME case. Applying that change helps my symptoms substantially, though it doesn't quite recover the old behavior: in a multipart text and HTML message, it seems to give the raw HTML rather than the plain text to supercite.