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 736449 - search-provider: quit sooner
search-provider: quit sooner
Status: RESOLVED NOTABUG
Product: gnome-control-center
Classification: Core
Component: general
unspecified
Other All
: Normal normal
: ---
Assigned To: Control-Center Maintainers
Control-Center Maintainers
Depends on:
Blocks:
 
 
Reported: 2014-09-11 02:53 UTC by Michael Catanzaro
Modified: 2014-09-11 17:19 UTC
See Also:
GNOME target: ---
GNOME version: ---

Attachments
search-provider: quit sooner (1.02 KB, patch)
2014-09-11 02:53 UTC, Michael Catanzaro 		none Details | Review

Description Michael Catanzaro 2014-09-11 02:53:46 UTC 
This doesn't matter very much, but one minute is longer than other search providers hang around for. I think most are using a 10s timeout.
Comment 1 Michael Catanzaro 2014-09-11 02:53:48 UTC 
Created attachment 285877 [details] [review]
search-provider: quit sooner

One minute is a little long for us to hang around.  Ten seconds is more
typical.
Comment 2 Bastien Nocera 2014-09-11 09:18:13 UTC 
(In reply to comment #0)
> This doesn't matter very much, but one minute is longer than other search
> providers hang around for. I think most are using a 10s timeout.

Citation needed.
Comment 3 Michael Catanzaro 2014-09-11 13:24:06 UTC 
Boxes, Contacts, and nautilus all disappear a couple seconds after my search.  Weather and seahorse both hang around for about a minute, so actually I guess we just have two different camps of search provider timeouts.

Epiphany (fixed just now) and gnote (bug #736472) seem to hang around forever.
Comment 4 Giovanni Campagna 2014-09-11 17:19:14 UTC 
All search providers I wrote have 1 minute timeout, yes. This is to make sure that the service doesn't disappear between a GetInitialResultSet and a GetSubsearchResultSet.
Also it allows to cache some info in memory, which makes the second search a lot faster.

I'm not sure the memory saving from using 10s instead of 1 minute is worth here: if you search often enough that 1 minute becomes significant, then you would be continously reloading the daemon from disk.