(11:50:23) Conversation with #gtk+ at 2008-06-01 115023 on pbor|out (11:50:23) You joined #gtk+ (11:53:49) You are now known as pbor (11:55:15) schumaml (schumaml@dslb-088-067-084-169.pools.arcor-ip.net) joined #gtk+ (11:55:51) Lethalman quit (Read error: 145 (Connection timed out)) (11:56:31) Lethalman (~lethal@host68-31-dynamic.27-79-r.retail.telecomitalia.it) joined #gtk+ (11:57:16) Lethalman quit (Leaving) (11:58:24) Lethalman (~lethal@host188-34-dynamic.27-79-r.retail.telecomitalia.it) joined #gtk+ (12:00:49) nacho quit (Remote closed the connection) (12:12:30) fdd quit (Peace Be Upon You!) (12:12:54) geaaru (~geaaru@host51-214-dynamic.17-79-r.retail.telecomitalia.it) joined #gtk+ (12:18:32) satis quit (Ex-Chat) (12:20:40) behdad (~behdad@CPE001217b19226-CM0012c9c84bc4.cpe.net.cable.rogers.com) joined #gtk+ (12:21:05) satis (~c-a@77.240.67.219) joined #gtk+ (12:24:09) jrb quit (Ping timeout: 600 seconds) (12:25:02) jrb (~jrb@bzq-179-150-194.static.bezeqint.net) joined #gtk+ (12:28:52) Lethalman quit (Ping timeout: 600 seconds) (12:32:20) leio (~leio@ip53.cab54.mus.starman.ee) joined #gtk+ (12:32:36) leio (~leio@ip53.cab54.mus.starman.ee) left #gtk+ (12:33:26) dotx (~christian@log77-4-82-246-228-78.fbx.proxad.net) left #gtk+ (12:34:06) bugbot quit (Excess Flood) (12:34:38) bugbot (~supybot@landfill.bugzilla.org) joined #gtk+ (12:35:02) bkor sets mode: +o bugbot (12:37:01) Lethalman (~lethal@host145-34-dynamic.27-79-r.retail.telecomitalia.it) joined #gtk+ (12:37:20) linuxx (~linuxx@port-83-236-206-154.static.qsc.de) joined #gtk+ (12:45:12) nacho (~nacho@252.120.218.87.dynamic.jazztel.es) joined #gtk+ (12:51:36) Lethalman quit (Ping timeout: 600 seconds) (12:55:52) tortoise (~tortoise@87-194-162-231.bethere.co.uk) joined #gtk+ (12:56:08) linuxx quit (Leaving) (12:56:13) linuxx (~linuxx@port-83-236-206-154.static.qsc.de) joined #gtk+ (12:59:33) harlan (~harlan@lesnuits.olf.sgsnet.se) joined #gtk+ (13:00:26) harobed quit (Ping timeout: 600 seconds) (13:01:03) Lethalman (~lethal@79.43.185.172) joined #gtk+ (13:05:15) behdad quit (Leaving.) (13:05:54) roosmaa_ (~roosmaa@84-50-21-146-dsl.kjj.estpak.ee) joined #gtk+ (13:12:08) roosmaa quit (Read error: 145 (Connection timed out)) (13:14:51) aike quit (Leaving) (13:16:22) Lethalman quit (Ping timeout: 600 seconds) (13:18:38) leio (~leio@ip7.cab54.mus.starman.ee) joined #gtk+ (13:18:55) kmaraas quit (Ping timeout: 600 seconds) (13:22:27) tim (~tpm@dsl-217-155-195-89.zen.co.uk) joined #gtk+ (13:25:02) Lethalman (~lethal@host93-51-dynamic.27-79-r.retail.telecomitalia.it) joined #gtk+ (13:25:38) roosmaa_ is now known as roosmaa (13:26:12) how do I run the gtk unit tests? (13:26:47) (make check doesn't seem to do the trick) (13:27:30) make test (13:27:44) Gtk+Tests:ERROR: Failed to start Xvfb environment for X11 target tests. (13:27:45) mmm (13:28:05) kris: can I run just one? (13:28:09) geaaru_ (~geaaru@host111-218-dynamic.31-79-r.retail.telecomitalia.it) joined #gtk+ (13:28:47) ah, I can just run the exec (13:28:49) cool (13:32:31) leio quit (Leaving) (13:34:41) geaaru quit (Ping timeout: 600 seconds) (13:34:42) herz1 quit (Leaving.) (13:36:12) leio (~leio@ip139.cab68.mus.starman.ee) joined #gtk+ (13:36:40) leio (~leio@ip139.cab68.mus.starman.ee) left #gtk+ (13:39:48) Lethalman quit (Read error: 104 (Connection reset by peer)) (13:39:50) wolfey (~wolfey@wolfey.si) joined #gtk+ (13:40:04) hello what is right way to update ListStore (13:40:21) it seems to try to 'refresh' when my update is still in progress (13:40:34) (strezba:12656): Gtk-CRITICAL **: gtk_list_store_get_value: assertion `VALID_ITER (iter, list_store)' failed (13:40:36) (strezba:12656): GLib-GObject-CRITICAL **: g_object_set_property: assertion `G_IS_VALUE (value)' failed (13:40:40) (strezba:12656): GLib-GObject-CRITICAL **: g_value_unset: assertion `G_IS_VALUE (value)' failed (13:41:12) i have a 'refresh' thread that monitors 'active list' (13:48:35) i see (13:48:41) gdk_thread_enter to get lock ;) (13:48:44) google friend (13:48:52) wolfey quit (Lost terminal) (13:54:01) Lethalman (~lethal@host200-118-dynamic.33-79-r.retail.telecomitalia.it) joined #gtk+ (13:56:44) paolo borelli added attachment 111884 to bug 535971 (13:56:44) Attachment http://bugzilla.gnome.org/attachment.cgi?id=111884&action=edit updated patch (13:56:45) Bug http://bugzilla.gnome.org/show_bug.cgi?id=535971 normal, Normal, ---, gtk-gtkbuilder-maint@gnome.bugs, UNCONFIRMED, Bring back ability to specify a "root" object (14:00:36) Cimi (~cimi@host15-151-dynamic.58-82-r.retail.telecomitalia.it) joined #gtk+ (14:00:50) tortoise quit (Ex-Chat) (14:07:10) neenaoffline quit (Read error: 104 (Connection reset by peer)) (14:08:22) rm_you (rm_you@pool-71-112-238-182.sttlwa.dsl-w.verizon.net) joined #gtk+ (14:08:26) owen (~otaylor@96.237.148.170) joined #gtk+ (14:08:44) jkroon (~jkroon@c83-254-41-184.bredband.comhem.se) joined #gtk+ (14:09:30) hpj quit (Ping timeout: 600 seconds) (14:09:47) Lethalman quit (Read error: 145 (Connection timed out)) (14:10:08) Lethalman (~lethal@host200-118-dynamic.33-79-r.retail.telecomitalia.it) joined #gtk+ (14:11:05) vadi2 (~vadi@H12.C192.cci.switchworks.net) joined #gtk+ (14:11:38) What would "assertion `GTK_IS_WIDGET (widget)' failed" mean when I don't have any widgets simply called "widget"? (14:11:57) is there a way to put a GtkHScale widget into a menu (I guess as a GtkMenuItem) and have it be *usable* in the menu? when I try it isn't usable, as clicking on it selects the whole thing as a menu choice and the click does not go through to the widget (14:15:29) vadi2: nothing; "widget" here is the variable name in the gtk code (14:15:56) not the name of your widget (14:16:12) fizz: ah. I figured it out, renamed something in glade but forgot to in my code. ty. (14:18:14) kmaraas (~kmaraas@30.80-202-165.nextgentel.com) joined #gtk+ (14:21:50) Goundy (~Goundy@234.17-240-81.adsl-dyn.isp.belgacom.be) joined #gtk+ (14:24:28) Fred_ (~fred@118.231.39.77) joined #gtk+ (14:25:00) harlan quit (Ping timeout: 600 seconds) (14:28:09) vadi2 quit (Remote closed the connection) (14:30:43) hpj (~hpj@189.190.11.219) joined #gtk+ (14:37:22) jpetersen quit (Remote closed the connection) (14:38:36) acidx (~leandro@201-93-229-72.dsl.telesp.net.br) joined #gtk+ (14:39:05) is there a way to put a GtkHScale widget into a menu (I guess as a GtkMenuItem) and have it be *usable* in the menu? when I try it isn't usable, as clicking on it selects the whole thing as a menu choice and the click does not go through to the widget (14:41:06) rm_you: No (14:41:21) heh (14:41:28) i guess that's a nice absolute answer :) (14:42:27) I wonder how Nokia did it in their Hildon UI <_< (their applets are closed source...) (14:42:31) prolly some sort of hack... (14:43:06) Lethalman quit (Read error: 145 (Connection timed out)) (14:43:25) rm_you: given a deep knowledge of the GTK+ internals I could come up with a demo that showed that working, but it would be complex unsafe code (14:43:40) rm_you: a) hildon is open source b) are you sure that the "menu" is a menu? (14:44:35) a) the only two applets that do it (the display applet and the volume applet) *are* closed source, and b) not 100%, but nothing else looks or behaves like those do besides GtkMenus :/ (14:44:38) herzi (~herzi@p578EAFEE.dip.t-dialin.net) joined #gtk+ (14:44:39) at least that i can find (14:46:42) rm_you: https://stage.maemo.org/svn/maemo/projects/haf/trunk/ has a lot of the source code (14:46:55) yes, but not the display or volume applet :( (14:47:25) https://bugs.maemo.org/show_bug.cgi?id=3199 (14:47:27) Bug 3199: was not found. (14:47:36) bugbot: shh (14:47:36) rm_you: Error: "shh" is not a valid command. (14:47:39) >_< (14:47:40) acidx quit (acidx) (14:49:31) rm_you: I'm not very clued in. My understanding was that the closed-source bits were mostly things that came from outside sources, etc. (14:49:56) yeah, it's been an uphill battle getting this applet done... *sigh* (14:50:52) vadi2 (~vadi@H12.C192.cci.switchworks.net) joined #gtk+ (14:51:17) mathrick quit (Read error: 104 (Connection reset by peer)) (14:51:48) mathrick (~mathrick@users177.kollegienet.dk) joined #gtk+ (14:52:11) jwendell (~wendell@201.38.220.67) joined #gtk+ (14:54:12) jkroon quit (Remote closed the connection) (14:55:46) It seems that a GtkProgressBar has a min width of 20px, but I'm somehow bypassing that level and able to scale it down to a 1px width, at which point my whole desktop enviroment dies (mouse moves, compiz window previews work, but nothing else. Clicks don't go through). I tried replicating the problem with a test setup, but can't, it only works with a specific glade file I did. I looked at the api and didn't see anything in regards to the min width, (14:57:39) vadi2: The "size request" of a widget is it's min size. If you set the SHRINK flag on a table or use gtk_widget_set_size_request() that min size will be ignored and a widget allocated smaller. That's generally a bad idea. (14:58:03) vadi2: Are you intentionally shrinking the progressbar down to 1px? (14:58:22) owen: No, it's just something that I did accidently and now trying to prevent it. (14:58:29) vadi2: Not sure about "desktop environment dies" thing ... sounds like a stuck grab, but GtkProgressBar doesn't grab (14:59:56) dpm (~dpm@p54A10CC0.dip0.t-ipconnect.de) joined #gtk+ (14:59:57) vadi2: The 'shrink' flag for gtk_paned_pack[12] is another thing that could result in the miun size being ignored (15:00:27) owen: oh, yes, the bars are in a handlebox which is in a vpaned (15:01:28) giusef (~Zmax@ip-72-55-165-86.static.privatedns.com) joined #gtk+ (15:02:21) owen: I can't see I'm seeing the shrink flag option to set in glade. What would be the syntax to set it manually in the .glade? (15:04:47) whoppix quit (Leaving) (15:06:05) owen: hm. At best I got "unknown property `shrink' for class `GtkVPaned'". (15:06:32) vadi2: It's a child property, not a property (15:07:13) Oops. Got it. (15:07:45) wolfey (~wolfey@wolfey.si) joined #gtk+ (15:07:49) somebody here perhaps? (15:07:59) owen: thank you. (15:08:01) i dont have idea why i get segmentation fault at new gnome canvas text item (15:08:10) while rectangles work (15:08:17) vadi2 (~vadi@H12.C192.cci.switchworks.net) left #gtk+ (15:09:06) item = gnome_canvas_item_new(group, (15:09:06) gnome_canvas_text_get_type(), (15:09:06) "x", 280, (15:09:06) "y", 100, (15:09:06) "anchor", GTK_ANCHOR_WEST, (15:09:08) "font", "fixed", (15:09:11) "fill_color", "black", (15:09:13) "text", "Active queue", (15:09:16) NULL); (15:09:38) documentation aint GTK's strong side (compared to javadoc) :) (15:10:27) wolfey: x/y should be doubles not ints maybe? (15:11:52) ahh (15:12:12) i'm stupid :) (15:12:17) i think i'll stay here for a while (15:12:29) madewokherd (~urk@c-71-61-145-98.hsd1.pa.comcast.net) joined #gtk+ (15:13:25) if i want it centered, i also put width and height? (15:17:56) dotx (~christian@log77-4-82-246-228-78.fbx.proxad.net) joined #gtk+ (15:18:37) oleavr (~oleavr@ti0035a340-2582.bb.online.no) joined #gtk+ (15:25:26) frubru (~frubru@81.21.22.191) joined #gtk+ (15:30:14) jkroon (~jkroon@c83-254-41-184.bredband.comhem.se) joined #gtk+ (15:30:59) tko_ quit (Read error: 131 (Connection reset by peer)) (15:31:42) tko (~tko@host81-159-235-184.range81-159.btcentralplus.com) joined #gtk+ (15:35:57) Eric Piel_gnome changed the Status on bug 468933 from NEW to RESOLVED FIXED. (15:35:57) Bug http://bugzilla.gnome.org/show_bug.cgi?id=468933 normal, Normal, ---, gtk-bugs@gtk.org, RESOLVED FIXED, gdk_screen_is_composited returns false when using xcompmgr (15:38:44) vadi2 (~vadi@H12.C192.cci.switchworks.net) joined #gtk+ (15:39:36) Is it possible to make so a handlebox is able to "snap on" to the position where another handlebox was? (15:41:59) jkroon quit (Remote closed the connection) (15:43:06) oleavr quit () (15:44:18) Johan Dahlin marked bug 535971 as a duplicate of bug 447998. (15:44:19) Bug http://bugzilla.gnome.org/show_bug.cgi?id=535971 normal, Normal, ---, gtk-gtkbuilder-maint@gnome.bugs, RESOLVED DUPLICATE, Bring back ability to specify a "root" object (15:44:21) Bug http://bugzilla.gnome.org/show_bug.cgi?id=447998 enhancement, Normal, ---, gtk-gtkbuilder-maint@gnome.bugs, NEW, GtkBuilder does not support building parts of the xml tree (15:44:40) eix (~Goundy@75.64-136-217.adsl-dyn.isp.belgacom.be) joined #gtk+ (15:49:13) oleavr (~oleavr@ti0035a340-2582.bb.online.no) joined #gtk+ (15:49:33) whoppix (~whoppix@c8113BF51.dhcp.bluecom.no) joined #gtk+ (15:53:43) Goundy quit (Ping timeout: 600 seconds) (15:53:53) Johan Dahlin cancelled none+ for attachment 111884 on bug 535971. (15:53:56) Johan Dahlin granted needs-work for attachment 111884 on bug 535971. (15:53:58) Attachment http://bugzilla.gnome.org/attachment.cgi?id=111884&action=edit updated patch (15:53:59) Bug http://bugzilla.gnome.org/show_bug.cgi?id=535971 normal, Normal, ---, gtk-gtkbuilder-maint@gnome.bugs, RESOLVED DUPLICATE, Bring back ability to specify a "root" object (15:56:17) geaaru_ quit (Leaving) (16:02:08) jkroon (~jkroon@c83-254-41-184.bredband.comhem.se) joined #gtk+ (16:04:16) dtolj (~dtolj@CPE001b1173d173-CM0016924d79f4.cpe.net.cable.rogers.com) joined #gtk+ (16:06:33) nacho quit (Remote closed the connection) (16:14:26) jdahlin: d'oh... I just finished implementing the root-object property :-) (16:15:12) nacho (~nacho@252.120.218.87.dynamic.jazztel.es) joined #gtk+ (16:15:36) pbor: sorry :-) (16:15:46) no, prob (16:15:46) pbor: you, kalle and kiko actually convinced me I was wrong (16:16:01) pbor: but before you code any further, check out comment 2 in the bug (16:16:23) actually, you should read the whole discussion in bug 447998 (16:16:32) Bug http://bugzilla.gnome.org/show_bug.cgi?id=447998 enhancement, Normal, ---, gtk-gtkbuilder-maint@gnome.bugs, NEW, GtkBuilder does not support building parts of the xml tree (16:16:55) I did... As for the limitations of the comment 2, I think it's exactly the same as a file with starts with a vbox as the real root (16:17:56) Correct, but as GtkBuilder supports more features which require references to external objects (sizegroup, treemodel, uimanager, etc) (16:18:29) which will mean that using a root is going to be a lot less useful than it used to be in libglade (16:19:13) mmm, I guess I may be missing something then (16:19:44) but my use case is pretty clear: I want to use the builder to build the inside of a composed widget (16:20:09) the way to do that now is to have a file which starts with a e.g. vbox (16:20:23) but such files cannot be edited in glade (16:21:00) so I want a file that starts with a toplevel window (so that glade understands it), but ignore the main window in the actual usage (16:22:01) that's really a limitation in glade, right? (16:22:23) well, yes and no (16:22:30) I hate to sound unproductive, but I think it's better to actually fix that instead of adding limited apis to GtkBuilder (16:22:55) it would mean that glade would have to save that info in some other file (16:23:08) do you use any comboboxes, treeviews or iconviews in your programs? (16:23:15) which doesn't really sound attractive at all (16:23:32) yeah (16:23:54) you can't easily construct them using a builder with a root set (16:24:04) since they reference external tree models (16:24:13) can you explain a little better that limitations, because I think I am missing something (16:24:16) New gtk+ bug 536063 filed by muntyan@tamu.edu. (16:24:16) Bug http://bugzilla.gnome.org/show_bug.cgi?id=536063 normal, Normal, ---, gtk-quartz-maint@gnome.bugs, UNCONFIRMED, Z-order for sibling child windows (16:24:41) Yevgen Muntyan added attachment 111891 to bug 536063 (16:24:41) Attachment http://bugzilla.gnome.org/attachment.cgi?id=111891&action=edit patch (16:24:56) jdahlin: if glade saved the file starting with the vbox, how would that change things wrt to treeviews? (16:25:03) pbor: just look at demo.ui in gtk+ (16:25:07) ok (16:25:11) *pbor looks (16:25:22) pbor: because the treeview is an eventual child of the vbox, which references a treemodel outside of the vbox part of the tree (16:27:50) danielm (~daniel@r190-135-147-16.dialup.adsl.anteldata.net.uy) joined #gtk+ (16:28:25) jdahlin: can't I simply add the list store with another call to add_with_root specifing the treestore as the root? (16:30:04) pbor: you can, but then you have to parse the whole tree for each external object you reference (16:30:17) (parse, but not construct) (16:30:20) harlan (~harlan@lesnuits.olf.sgsnet.se) joined #gtk+ (16:31:26) jdahlin: it doesn't sound so bad... I could put all the stores etc in a separate file (16:32:11) though that would screw glade once again (16:32:19) right, glade cannot cope with that either. (16:32:45) GtkBuilder does not contain the work 'glade' for a reason (16:32:50) s/work/word/ (16:33:09) It should not be tied to a specific UI designer, and ideally not contain API to work around broken ones ;-) (16:33:45) right, but I think that we need to find a good way to support that use case (16:35:31) from an application developer point of view is very frustrating that features that get consolidates with project ridley or whatever always contain feature regression that make them hard or impossible to actually use in an app that was first using the old lib that's now deprecated (16:36:30) (it's not only builder, it keeps biting me in tha ass all the time, e.g. gio that doesn't have way to get the hostname or way to set the backup name and so on) (16:38:36) libglade is deprecated, use gtkbuilder instead; but gtkbuilder is not a replacement for libglade ;) (16:38:57) jmd quit (Remote closed the connection) (16:40:29) Enselic (~martin@c-d211e455.017-113-6c756e10.cust.bredbandsbolaget.se) joined #gtk+ (16:40:34) pbor: it is possible to use gtkbuilder in an application, I even submitted a patch for gedit a few years back (16:40:52) but I believe it was not used because of the lack of a proper UI designer (16:41:22) nacho_ (~nacho@50.120.218.87.dynamic.jazztel.es) joined #gtk+ (16:41:25) jdahlin: exactly, maintaining ui files manually is worse than maintaining them in code (16:41:45) jdahlin: and blaming that all on glade is not IMHO fair (16:41:49) glade-3 is terrible (16:41:50) pbor: why don't you do the same as gnome-terminal, sue the gtk-builder-convert script to convert them at build time? (16:41:56) s/sue/use/ (16:42:13) just keep glade files around in the svn repository (16:42:26) gtk-builder-convert even has an option to split up .glade files into multiple .ui files (one per toplevel) (16:42:42) once glade-3 can load all the ui files, stop using the converter script. (16:42:56) jdahlin: sure, I can do that... but still (16:43:11) this whole exercise was to simplify things (16:43:23) not to make them more baroque (16:43:37) pbor: I can understand your frustration, but there's nobody really forcing you to use gtkbuilder (16:43:44) if you think libglade is good, just keep on using it (16:44:35) jdahlin: but I think gtkbuilder is good! it's in gtk, one less dep, one less unmaintained lib and all that (16:44:49) jdahlin: and I am willing to *help* fix it (16:44:53) for my needs (16:45:14) I'm afraid in this case, it's not really possible to fix GtkBuilder in a good way (16:45:40) I still think that my patch is not that bad (16:45:42) owen and mclasen had me rewrite the parser to be a one step one, instead of a two step one as it was in libglade (16:46:11) no, the patch is good, but the limitations of the new api is unfortunate (16:46:35) since it's not possible to reference objects built outside of the root (16:46:39) I mean, I can choose the tradeoff of calling add_with_root multiple times (16:46:55) vadi2 (~vadi@H12.C192.cci.switchworks.net) left #gtk+ (16:47:39) pbor: or adding the api to specify multiple roots.. (16:47:46) nacho quit (Ping timeout: 600 seconds) (16:47:46) sure (16:48:22) jdahlin: if that's decided to be the right way, I am ok with doing that (16:48:43) pbor: I'm uncertain, I'd like to get input from other people (16:48:48) jdahlin: okay (16:49:46) or we could add a function to set a "root regex" (16:50:21) only objects with id matching the regex (or even a simple prefix) are used as root (16:50:31) or other things like that (16:51:37) roosmaa quit (Leaving) (16:51:38) jkroon quit (Ping timeout: 600 seconds) (16:55:54) pbor: To quote the wise jwz: "You have a problem. Then you decide to use regex to solve that problem. Now you have two problems." (16:59:45) behdad (~behdad@CPE001217b19226-CM0012c9c84bc4.cpe.net.cable.rogers.com) joined #gtk+ (17:01:11) joyr (~joyr@dsl-tkubrasgw1-ffb7c100-219.dhcp.inet.fi) joined #gtk+ (17:01:14) *jdahlin makes some noise and attempts to start a gobject-introspection on gtk-devel (17:02:01) Yevgen Muntyan added attachment 111897 to bug 536063 (17:02:01) Attachment http://bugzilla.gnome.org/attachment.cgi?id=111897&action=edit updated patch (17:02:03) Bug http://bugzilla.gnome.org/show_bug.cgi?id=536063 normal, Normal, ---, gtk-quartz-maint@gnome.bugs, UNCONFIRMED, Z-order for sibling child windows (17:03:03) giusef quit (giusef) (17:04:27) jdahlin: does it need support from inside glib? (17:04:55) jdahlin: i.e. can one grab that GObject-Introspection thing and be happy with current glib? (17:05:15) oleavr quit (Ping timeout: 600 seconds) (17:06:16) muntyan: yes (17:06:32) You can use it as it is right now, it doesn't have to go into glib to be 'useful' (17:06:41) fitzsim (~fitzsim@CPE001d7ec32f09-CM001ac3154790.cpe.net.cable.rogers.com) joined #gtk+ (17:07:11) jdahlin: so i can go and implement a binding for gtk right now, without waiting for glib-2.38? (17:07:29) muntyan: yes you definitely can (17:07:35) awesome! (17:07:54) however, the GIR format is not completely stable yet (17:08:03) and it lacks things you will need to write complete bindings (17:09:26) you don't mention those things in your email (17:11:04) I don't mention the contrary either! (17:11:37) right :) (17:17:21) neenaoffline (~neenaoffl@59.162.168.11) joined #gtk+ (17:17:29) New gtk+ bug 536072 filed by mangosoft@comhem.se. (17:17:29) Bug http://bugzilla.gnome.org/show_bug.cgi?id=536072 normal, Normal, ---, gtk-bugs@gtk.org, UNCONFIRMED, Windows with saved positions always open at (0,0) (17:25:07) jdahlin: does gobject-introspection need python? (17:25:23) it says "checking whether Python support is requested", which sounds like it's optional, yet it aborts on that check (17:25:26) it == configure (17:26:25) or would it work with python-2.4? (17:32:17) jrb quit (Ping timeout: 600 seconds) (17:32:46) muntyan: yes, it depends on python which will be a build time dependency for projects using the scanner (17:33:03) muntyan: the C api to access the metadata is currently not installed, as it needs to be updated to the new format (17:33:13) and it depends on python-2.5 (17:33:19) python-2.4 doesn't work (17:34:22) correct, I use a couple of 2.5 specific features (17:35:33) I need to catch a bug, please follow up on gtk-devel or on my blog (17:35:35) *bus (17:35:45) jdahlin is now known as jdahlin-afk (17:37:55) KaL (~KaL@213.37.62.251.static.user.ono.com) joined #gtk+ (17:38:31) nekohayo (~jeff@206-248-171-113.dsl.teksavvy.com) joined #gtk+ (17:41:20) xkr47 quit (Ping timeout: 600 seconds) (17:47:25) manphiz (~dxy@218.244.247.198) joined #gtk+ (17:49:17) giusef (~Zmax@ip-72-55-165-86.static.privatedns.com) joined #gtk+ (17:52:49) frubru quit (Ex-Chat) (17:54:22) rvv (~rvv@111.84-48-17.nextgentel.com) joined #gtk+ (17:58:40) ADC Reference Library in devhelp \o/ (17:59:30) now to just figure out how to find out all classes automagically rather than having to convert them individually (18:05:26) I am having a problem with drag and drop. If I drop a file from my browser onto the widget, it gets a drag and drop signal, but I can't get a uri from it.. if I peak in ->data of the GtkSelectionData and I see the uri... How do I correctly get the uri? (18:05:27) jkroon (~jkroon@c83-254-41-184.bredband.comhem.se) joined #gtk+ (18:05:47) qball: gtk_selection_data_get_uris (18:06:17) yeah that _does_ not work (18:06:23) returns null (18:06:27) Cimi quit (Sto andando via) (18:06:42) then you messed things up? (18:06:48) what's "my browser"? (18:06:54) zwnj (~behnam@194.225.70.245) left #gtk+ (18:06:58) firefox 3 (18:07:33) maybe it doesn't provide text/uri-list target? then you should learn whatever firefox does (18:07:57) hmm (18:08:01) perhaps it's using some URL target or MOZFOOBAR or something (18:08:11) doublep (~paul@212.98.162.2) joined #gtk+ (18:08:15) hmm yeah, grrr annoying (18:08:23) is it plain text perhaps? (18:08:47) get_text() fails too (18:09:10) I get a target type _NETSCAPE_URL (18:09:26) Ibrahim_ (~Zugadi@196.200.94.248) joined #gtk+ (18:09:52) Ibrahim_ (~Zugadi@196.200.94.248) left #gtk+ (18:10:11) did you ask it for text/uri-list? in your targets list (18:10:16) *muntyan has no idea what mozilla does there (18:10:19) I have both there (18:10:24) let me remove that one (18:13:04) hmm if I don't include _NETSCAPE_URL it does not drop (18:13:21) but if it is a _NETSCAPE_URL I can do manual parsing (18:13:27) using g_uri_list_extract_uris (18:23:33) squentin (~squentin@bsb94-1-82-232-16-14.fbx.proxad.net) joined #gtk+ (18:28:41) manphiz quit (leaving) (18:29:43) jpetersen (~petersen@xdsl-81-173-146-67.netcologne.de) joined #gtk+ (18:37:04) dino (~dino@121-72-130-64.dsl.telstraclear.net) joined #gtk+ (18:44:05) fredmorcos (~fred@ip-141-31-187-86.nat.selfnet.de) joined #gtk+ (18:51:18) Rhamphoryncus (~rhamph@S010600146c3c91a1.ed.shawcable.net) joined #Gtk+ (18:54:59) jkroon quit (Remote closed the connection) (18:55:01) paolo borelli added attachment 111904 to bug 535971 (18:55:01) Attachment http://bugzilla.gnome.org/attachment.cgi?id=111904&action=edit patch with set_root() method (18:55:02) Bug http://bugzilla.gnome.org/show_bug.cgi?id=535971 normal, Normal, ---, gtk-gtkbuilder-maint@gnome.bugs, RESOLVED DUPLICATE, Bring back ability to specify a "root" object (18:55:29) * Disconnected (18:55:29) You are now known as pbor|out