Re: How to disable reading of ~/.config/gtk-2.0 ?
- From: Geoff Bache <geoff bache jeppesen com>
- To: gtk-list gnome org
- Subject: Re: How to disable reading of ~/.config/gtk-2.0 ?
- Date: Fri, 12 Oct 2007 11:40:38 +0200
Ed Mansur wrote:
It's an internal implementation detail, so your documentation is the
source - look at gtk/gtkfilechoosersettings.c.
Is there another way users are supposed to edit it then? It seems to
seriously affect
what filechoosers look like (in the case of the location mode, which is
mostly what I care about,
an entirely new widget appears), and it seems a bit odd to direct people
to read uncommented
C code when they want to configure the default appearance of their
filechooser dialogs.
In any case, you cannot understand location mode by reading that file,
other than to state
that it has two possible values and they set a variable called
location_mode...
The gtkfilechooser file seems to be automatically generated when running
GTK apps. However,
it seems to get given different default values for some people despite
nobody here having a clue it
even existed. Most people seem to have LocationMode=path-bar, but the
guy who recently built
GTK 2.10 for us had LocationMode=filename-entry. What steers this?
Would be good even with programmer documentation, I see nothing in the
FileChooser docs
about location mode. Is it possible to examine it programmatically?
Now I wouldn't even know or care about any of this stuff, but
LocationMode=filename-entry
seems to have some weird effects on our application. The filechoosers
seem to not emit various
signals that they do emit when in path-bar set-up, which our application
is expecting, and more seriously,
the modality of the dialog seems to break down in some way (certain
changes to the main GUI are
allowed while the dialog is up). I haven't at all got to the bottom of
these effects, and to what extent
they are my fault as opposed to GTK's fault, so this isn't a bug
report, I was just wondering if somebody
could shed some light on this and maybe give me some pointers.
At least our acceptance tests no longer depend on this file, which is
progress!
Regards,
Geoff
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]