-
Logged In: YES
user_id=128703
I wrote a demo code which finally showed me where I was
wrong (automatic selection of the interaction style). Sorry
for opening this bug and thanks to SCIM to have raised this.
Nevertheless, in my environment Gnome/Fedora Core 5, when I
input some characters in chinese, for example, and I select
the character in the proposed list by typing the...
2006-04-24 11:34:28 UTC in Smart Common Input Method platform
-
When I create an Input Context with the pre-edit
interaction style XIMPreeditNothing (and the Status
with XIMStatusNothing), XCreateIC() returns NULL.
I discovered that if I create the Input Context with
the sub-attribute XNSpotLocation defined, then it succeeds.
But the only requirements when creating an Input
Context with XCreateIC() are to provide the attribute
XNInputStyle and the...
2006-04-21 18:38:29 UTC in Smart Common Input Method platform
-
Logged In: YES
user_id=128703
I've had these upgrades since July the 28th. But I don't
remember if it made the crashs stop.
I just can say that my current gnubiff process has been
running for 8 days without a crash.
2005-08-16 15:15:22 UTC in gnubiff
-
Logged In: YES
user_id=128703
Here's another and different backtrace with gdb, but it ends
in the same function. It crashed after 3 days.
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1211040848 (LWP 20503)]
0x00321977 in __gnu_cxx::__pool::_M_get_thread_id ()
from /usr/lib/libstdc++.so.6
(gdb) bt
#0 0x00321977 in...
2005-07-25 12:34:35 UTC in gnubiff
-
Logged In: YES
user_id=128703
vitaly,
The pref' bug (which in fact concerns local mailbox) doesn't
fix this issue, I still had on a pop3 mailbox.
For the CVS sources, I don't know how to compile them,
there's no configure script and I didn't succeed in building
one with automake (that I'm not used to use). So I had to
untar the lastest source tarball and insert the modified
source files...
2005-07-22 15:28:50 UTC in gnubiff
-
Logged In: YES
user_id=128703
I also have some crashs with 1 POP3 box, and I'm using the
same Gnome et FC4.
It occurs within several days. Each time I discovered the
crash, I was coming back after I was away (morning, after
lunch...), and having the screensaver activated.
I decided to attach my gnubiff with gdb on Monday, and it
has finally crashed today.
Here's the backtrace:
Program...
2005-07-22 12:23:00 UTC in gnubiff
-
Logged In: YES
user_id=128703
Perhaps FAM doesn't like multithreading.
Moreover we may have different FAM libraries. The FAM
library from SGI was used until Fedora Core 2 (on which
gnubiff works). Since Fedora Core 3, this has been replaced
by the gamin project: http://www.gnome.org/~veillard/gamin/
Check where your FAM library comes from, I wouldn't be
surprised it's the SGI one. If yes,
2005-07-08 09:37:21 UTC in gnubiff
-
Logged In: YES
user_id=128703
Yes, it helps a lot!
FAMNextEvent is a blocking function. It should return when
there's an upcoming FAM event, I don't know why this doesn't
return then. The man says "If there are no events pending,
FAMNextEvent blocks until an event occurs."
But the synchronization done with FAMPending makes it work.
Instead of doing a sleep in a while loop, I suggest...
2005-07-04 08:55:04 UTC in gnubiff
-
Logged In: YES
user_id=128703
Sorry, I have the same behavior with the CVS code. Except
that the frame #11 is shifted to line 209 because you
modified local.cc:
#11 0x0809ae66 in Local::stop (this=0x8d11fd0) at local.cc:209
Don't you reprocude the issue?.
2005-07-01 08:33:30 UTC in gnubiff
-
Logged In: YES
user_id=128703
From the shell, the same problem exists. I've made a bunch
of tests, what I can say, is that it's not linked with Gnome
nor FC4, but it's a problem of thread when having at least
one mail in the local spool /var/spool/mail/mylogin. I also
discovered that I can open the preference screen and
configure gnubiff before it checks the local spool (and
freezes), so...
2005-06-29 09:46:04 UTC in gnubiff
