<<< Date Index >>>     <<< Thread Index >>>

Re: bug hunting party



* Derek Martin on Tuesday, February 17, 2009 at 15:22:02 -0600
> On Tue, Feb 17, 2009 at 08:47:13AM +0100, Christian Ebert wrote:
>> iconv -l would have saved me for the iso-8859-i case of that bug,
>> because it worked with the old version (1.10) of iconv I had then
>> on my machine. 
> 
> The question is, is it generally available in iconv implementations
> (i.e. those other than GNU)?  If not, then I think it's really not
> practical to try to include it in Mutt, particularly where all we're
> trying to accomplish is working around a buggy system library.

~/src/mutt$ hg grep -i 'work ?around .*bug'
ChangeLog:5681: * doc/Makefile.am: Work around pdksh bug in xml doc make target.
mutt_sasl.c:5497:  /* Work around a casting bug in the SASL krb4 module */
INSTALL:5346:   option enables a workaround to this bug.
install-sh:3385:          # this colon is to work around a 386BSD /bin/sh bug
intl/ChangeLog:2868:    * Makefile.in (install): Work around for another Buglix 
stupidity.
ChangeLog.old:3134:     DE's patch to work around other software's possible y2k 
bugs.

I am also sure that libiconv 1.11 has other bugs, only not those
that triggered the problem.

It's more a question of which bugs in the software Mutt depends
on are deemed worthy to work around.

Both examples were broken messages, and Mutt provides lots of
workarounds to read those. I thought it would be a good idea that
Mutt would give some kind of error message instead of stalling
silently. But if that's impossible, so be it.

> The right solution is to upgrade your broken library...

Of course.

>>>>  http://localhost:8000/mutt/main/rev/fb77465af534
> 
> By the way Rocco, that link isn't very useful. ;-)

hehe

http://dev.mutt.org/hg/mutt/rev/fb77465af534

>> From a user's perspective I still maintain that I needed to be a
>> Mutt fanatic to continue using it after that happened because at
>> the very moment that mutt gets silently stuck on a huge mailbox
>> without telling me anything. How could I've known
>> 
>> a) that my libiconv version might be to old
>> b) about any workarounds
> 
> I guess by asking this list... ;-)

Well, in the second case Mutt stalled when opening the mutt-dev
mailbox ;-)

> Or, take the opposite view: How is a Mutt developer supposed to know
> that you have a buggy system library?

I guess by my report to this list ... ;-)

> In the general case, there's really not much that can be done
> about that problem, other than running your system updates
> regularly.

At the time (MacOS 10.4 iirc) both Apple's and fink's iconv were
at version 1.10, so I would have had to install a third iconv.

> Essentially all software you run on your system will be
> compiled against some system library, and if it's buggy, you'll
> have trouble.  But the software vendor/dev isn't in a position
> to fix that problem... it's not their code.  

Sure.

Let's consider it solved then.

c
-- 
\black\trash movie    _C O W B O Y_  _C A N O E_  _C O M A_
Ein deutscher Western/A German Western
-->> http://www.blacktrash.org/underdogma/ccc.html
-->> http://www.blacktrash.org/underdogma/ccc-en.html