After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 532044 - fails to sign in at login
fails to sign in at login
Status: RESOLVED WONTFIX
Product: bigboard
Classification: Deprecated
Component: general
unspecified
Other Linux
: Normal normal
: ---
Assigned To: bigboard-maint
bigboard-maint
gnome[unmaintained]
Depends on:
Blocks:
 
 
Reported: 2008-05-07 20:41 UTC by Owen Taylor
Modified: 2012-02-25 17:49 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Owen Taylor 2008-05-07 20:41:46 UTC
Moved from: http://bugzilla.mugshot.org/show_bug.cgi?id=1277
Reporter: Luis Villa
Opened: 2007-08-07 07:07 

When I log in to my system, the bigboard comes up, but often fails to log in or
display any icons. Once in every... dunno, 10 logins or so it works?

If I kill it and restart, it usually seems to come up fine (modulo various
to-be-expected brokennesses), so maybe it is a problem with the lack of network
at login time.

[I'm using the F7 packages, currently 0.5.5, but happened with the previous
version as well.]


------- Comment #1 From Colin Walters 2007-08-09 12:53 [reply] -------

I think what may be happening here is that during startup, we may be hitting a
timeout on some D-BUS method invocation (~5 seconds?), because both the sidebar
and mugshot are starting up, may be blocking on disk for icons or something,
and not respond fast enough.

I guess the solution would be to basically catch any D-BUS reply error and
essentially restart, probably with bounded exponential fallback to be safe.


------- Comment #2 From Luis Villa 2007-10-23 08:41 [reply] -------

Note that in the latest packages I'm now seeing the sidebar fail to start at
all; I have to pkill bigboard and then restart it to get a board at all. Not
sure if this is related.
Comment 1 Owen Taylor 2008-05-07 20:44:21 UTC
I'm still seeing something like this as of 2008-05-07, though it's
much more intermittant than described here: maybe 1/10 times rather 
than 9/10 times.

I think Colin's theory is probably right... we should change pyddm
to set "infinite" timeouts for its method calls (10 minutes or so, say)
Comment 2 Owen Taylor 2008-05-07 21:46:07 UTC
Made that change in pyddm. If we don't get any duplicates for while
after that code ends up in packages, then we'll consider this fixed.

(Though it's inherently problem if 25 second timeout versus a 10 minute
timeout makes a difference. 25 seconds is a long time for the user
to wait.)

r7251 | otaylor | 2008-05-07 17:43:36 -0400 (Wed, 07 May 2008) | 4 lines

Set a long timeout (10 minutes) when making calls against the data
model engine. This matches what the C client library does.
(http://bugzilla.gnome.org/show_bug.cgi?id=532044)
Comment 3 André Klapper 2012-02-25 17:49:43 UTC
"Bigboard" will not see active development anymore according to one of its developers (Colin).
Closing this report as WONTFIX - Please feel free to reopen this bug report in the future if anyone takes the responsibility for active development again.


[I've also closed this product for new entries and moved it to the "Deprecated" classification. I've also filed a Git archiving request against sysadmin.]