Opened 15 years ago

Closed 15 years ago

Last modified 13 years ago

#2821 closed task (fixed)

Mutt-1.5.19.

Reported by: Ag. Hatzimanikas Owned by: Ag. Hatzimanikas
Priority: normal Milestone:
Component: BOOK Version: SVN
Severity: normal Keywords:
Cc:

Description

Version increment to 1.5.19.

I am proposing to skip the update (although many fixes applied in this release) but also perhaps some regressions too, as in a case that reported by Kyle Wheeler; an infinite recursive loop that can occur under some configuration scenarios.

This is fixed, however I don't want to add a patch in an updated (dev) package like it is the case with mutt. 1.5.18 is rock solid,

Setting this ticket to feature to get out of the way (for the moment).

Attachments (4)

mutt-1.5.18-md5.h_fix-1.patch (2.0 KB ) - added by willimm 15 years ago.
Backported Mutt fix for the md5.h header.
mutt-1.5.19-imap_hook_fix-1.patch (1.0 KB ) - added by willimm 15 years ago.
MUtt IMAP hook fix. Works for both 1.5.18 and 1.5.19.
mutt-1.5.18-md5_fixes-1.patch (14.6 KB ) - added by willimm 15 years ago.
Mutt MD5 fixes. Use this instead of mutt-1.5.18-md5.h-1.patch.
blfs-mutt-update.patch (3.5 KB ) - added by willimm 15 years ago.
Mutt update patch.

Download all attachments as: .zip

Change History (17)

comment:1 by willimm, 15 years ago

This should be for 6.4. Why? Look at this expert from the Changelog:

2008-11-15 19:51 -0800  Alexey I. Froloff  <raorn@altlinux.org>  (af87aa1846be)

	* md5.c, md5.h: md5.h, md5.c: updated to latest version from gnulib.
	Buggy old md5.h causes problems with gcc 4.3 compiler. In md5.h
	__attribute__ is #define'd to no-op and causes mutt_md5 to enter
	inifinite loop while calling memcpy().

Now do you understand why? There is a gcc 4.3 fix in there. Or can we make a patch for it? The better option is to upgrade.

comment:2 by Ag. Hatzimanikas, 15 years ago

I am waiting for a new release. If it will get late then we can update and backport the patch I was mentioning in the description. As I said there many bug fixes in this release but don't forget that is a development version and not a formal release (it's an exception that we have a development release in the book) and though wasn't confirmed there was another report for a segmentation fault.

Users are encouraged to upgrade and report problems in the mutt bug tracker, though, what I can do is to try to keep updated this ticket and maybe after a small period of no sign of a new release, we will update.

in reply to:  2 comment:3 by willimm, 15 years ago

Replying to ag@…:

I am waiting for a new release. If it will get late then we can update and backport the patch I was mentioning in the description. As I said there many bug fixes in this release but don't forget that is a development version and not a formal release (it's an exception that we have a development release in the book) and though wasn't confirmed there was another report for a segmentation fault.

Users are encouraged to upgrade and report problems in the mutt bug tracker, though, what I can do is to try to keep updated this ticket and maybe after a small period of no sign of a new release, we will update.

I'll still love to update this mutt version, it is the best way to go.

comment:4 by willimm, 15 years ago

I am waiting for BLFS to update to this version.

in reply to:  4 comment:5 by willimm, 15 years ago

Replying to willimm:

I am waiting for BLFS to update to this version.

...Or maybe not, I back ported the md5.h fix. I am attaching it, apply with -Np1 -i.

by willimm, 15 years ago

Backported Mutt fix for the md5.h header.

comment:6 by willimm, 15 years ago

Another mutt fix. This time, it is a IMAP related fix for hook.c. Works for both mutt 1.5.18 and 1.5.19.

by willimm, 15 years ago

MUtt IMAP hook fix. Works for both 1.5.18 and 1.5.19.

comment:7 by willimm, 15 years ago

Oops, forgot to update md5.c as well. Will do so.

by willimm, 15 years ago

Mutt MD5 fixes. Use this instead of mutt-1.5.18-md5.h-1.patch.

comment:8 by Ag. Hatzimanikas, 15 years ago

Milestone: future6.4
Owner: changed from blfs-book@… to Ag. Hatzimanikas
Priority: lownormal
Status: newassigned

Thanks, For completeness, please add some headers (date, name, etc..), with a small description and a link to the commit.

It's late night here and I have to finish something. Could you create a patch for the book, so I can apply it tomorrow night? Here are some details, from my logs:

ARCHIVE SIZE   : 3475.337891 kbytes for mutt-1.5.19.tar.gz
MD5SUM         : 73b3747bc7f7c805921e8d24ebac693f
TOTAL SPACE    : 28.755859 kbytes disk space required for mutt
SBU: 0.348837

Also a new dependency for the header caching (meant as a replacement for qdbm): http://tokyocabinet.sourceforge.net/index.html

I didn't check the switches though.

Also give me some details for the second patch, isn't applied in 1.5.19?

in reply to:  8 comment:9 by willimm, 15 years ago

Replying to ag@…:

Also give me some details for the second patch, isn't applied in 1.5.19?

What do you mean? The MD5 fix, or the IMAP hook fix, if you mean the IMAP fix, it applies for 1.5.19. If it's the MD5 fix, it's for Mutt 1.5.18.

comment:10 by willimm, 15 years ago

Wow, I am relay panting harder than a mutt to make all these packages, and that includes the BLFS book patch. It will prob. Just work, except for the missing patch link, which you can get from this ticket. (Filename mutt-1.5.19-imap_hook_fix-1.patch, attached to this ticket.) Please use the book patch.

by willimm, 15 years ago

Attachment: blfs-mutt-update.patch added

Mutt update patch.

in reply to:  10 comment:11 by willimm, 15 years ago

Replying to willimm:

Wow, I am relay panting harder than a mutt to make all these packages, and that includes the BLFS book patch. It will prob. Just work, except for the missing patch link, which you can get from this ticket. (Filename mutt-1.5.19-imap_hook_fix-1.patch, attached to this ticket.) Please use the book patch.

Sorry, I meant patches, not packages.

comment:12 by Ag. Hatzimanikas, 15 years ago

Resolution: fixed
Status: assignedclosed

Qdbm is still valid. As it seems there is no active development, Tokyo Cabinet is just the successor of Qdbm.

Thanks for your work in this ticket.

Fixed in r7738.

comment:13 by (none), 13 years ago

Milestone: 6.4

Milestone 6.4 deleted

Note: See TracTickets for help on using tickets.