[10:26] <eli_> How do I keep the unity toolbar from poping out every time I go to use the back button on maximized browser?
[12:07] <AlanBell> eli_: in unity3d you can configure itso the launcher says out
[12:40] <JanC> you can configure Compiz to wait a little bit before screen side/corner activation, which IME helps to avoid accidental activation
[12:41] <JanC> the tricky bit is finding how long to wait though; if you set it too high, it gets annoying too, of course  ;)
[13:55] <mhall119> mhr3: can I request a DBus method for querying the list of categories?
[13:55] <mhall119> a list of filters would be nice to, but right now I just need the category
[14:01] <mhr3> mhall119, what would be the use case?
[14:02] <mhr3> scopes are tied to lenses, the category list is considered given
[14:02] <mhr3> similar to filters
[14:08] <mhall119> mhr3: but scope code can exist outside of the lens
[14:08] <mhall119> but it needs to know what categories there are when building the results
[14:13] <mhr3> mhall119, what would it really do with the localized strings identifying categories?
[14:14] <mhr3> as i said they should be considered constants for a particular lens
[14:22] <davidcalle> mhr3, do you know what are Mikkel plans for the new home dash?  Right now, it separates categories from single lenses, without any reference to the parent lens (for example, I have categories named "Messages", "Replies", "Images", from the Gwibber lens, without any visual reference to the Gwibber lens)
[14:22] <mhall119> ok, so singlet does things a little differently, I'm looking at just building in a meta-data service to singlet lenses
[14:36] <mhr3> davidcalle, unfortunately design wanted to do "magic" with categories in the global search, so right now lenses register extra categories (with a title that suggest something aggregated) and scopes use this special category during global search
[14:37] <mhr3> like files lens has special "Files and Folders" category
[14:40] <davidcalle> Ok, i didn't noticed the new cat. So things need to be changed in lenses/scopes. I think it's pretty nice actually to pass several cats to the Home Dash, the cat name + icon just need to be adapted.
[14:42] <mhall119> mhr3: is there some documentation about how to define these "magic" categories?
[14:43] <mhr3> mhall119, there's nothing magic about them, it's just that you use them based on the search type
[15:06] <mhall119> ah, ok
[16:00] <mhall119> mhr3: are Lenses and scopes exported to the SessionBus, or something different?
[16:02] <mhall119> I'm trying to expose a dbus service object from singlet, and have it reachable by a scope from a different process
[16:20] <mhr3> mhall119, yea, they are
[16:20] <mhall119> ok, I'm doing something wrong then.... :(
[16:20] <mhall119> mhr3: know of any good tutorials for python-dbus?
[16:20] <mhr3> mhall119, that's what the export() method does :)
[16:21] <mhall119> .export is from the gobject thought, right?
[16:21] <mhall119> from Unity.Lens and Unity.Scope
[16:21] <mhall119> I'm making a python class that I want to expose via dbus
[16:25] <mhr3> i saw a comment somewhere suggesting that mixing python-dbus and gdbus is a very bad idea
[16:25] <mhall119> :(
[16:26] <mhall119> All I want is a way to reference a lens's category without having to know the index number for it
[16:26] <mhall119> which is easy when they're in the same code
[16:26] <mhall119> or when I can lookup the python code for the lens from the scope code
[16:27] <mhr3> mhall119, wouldn't that make singlet unusable with non-singlet lenses?
[16:27] <mhall119> yes
[16:28] <mhall119> well, you still could use a singlet scope with non-singlet lens, you'd just have to know the category index numbers for the lens
[16:28] <mhall119> rather than having a developer friendly name
[16:29] <mhall119> and you'd have to update your scope if the lens author changes category index order
[16:29] <mhr3> i'd say just provide a method to manually register the categories
[16:29] <mhall119> what do you mean?
[16:30] <mhr3> you're not going to get a "safe" way to deal with it
[16:30] <mhr3> if the categories change, who's to say the filters didn't change
[16:30] <mhr3> or whatever
[16:31] <mhall119> I was going to do something similar for filters, actually
[16:31] <mhall119> I just want to let scope developers say a result belongs in self.lens.teams
[16:31] <mhall119> rather than a result belongs in 0
[16:31] <mhr3> what is it good for? the scope won't work properly if it wasn't implemented for the filters that are in the lens *now*
[16:31] <mhall119> it's strictly for developer convenience and tooling
[16:32] <mhall119> mhr3: see http://bazaar.launchpad.net/~singlet-developers/singlet/trunk/view/head:/test/singlescope line 21
[16:33] <mhall119> i want to do something similar in a scope that's outside the lens
[16:34] <mhr3> mhall119, my suggestion is "cat1 = InheritedCategory(3)"
[16:35] <mhr3> imo that's as far as you can get convenience-wise
[16:35] <mhr3> scopes can't register categories anyway
[16:35] <mhall119> yeah, but that still requires the scope developer to know what order the categories are defined in the lens, and to update their scope if the order of categories changes
[16:36] <mhall119> which is what I was hoping to avid
[16:36] <mhall119> avoid
[16:37] <mhr3> to reiterate my point - if the order of categories changes you're screwed no matter what, the lens changed and you need to look at the changes again
[16:37] <mhall119> why would you be screwed?
[16:37] <mhall119> if I changed it so cat2 came before cat1
[16:38] <mhr3> what if the category is now something completely different?
[16:39] <mhall119> in the code I posted above, as long as cat1 exists, the scope part continues to work as expected
[16:39] <mhall119> doesn't matter if it comes first or second
[16:39] <mhall119> because self.cat1 resolves itself to the index number at runtime
[16:40] <mhall119> so the scope code doesn'tneed to know the index at the time it is written
[16:40] <mhr3> i agree that the categories show have been looked at, i think it would make sense to create new ones from scopes, but that's not the case, and most likely won't change until unity-6.0
[16:41] <mhall119> I don't want to create new ones in the scopes, I just want to expose the singlet lens's name->index map
[16:41] <mhr3> you can't do it cleanly
[16:42] <mhall119> because of the python-dbus/gdbus problems?
[16:42] <mhall119> because I woould do it completely outside of libunity
[16:42] <mhr3> the latter
[16:42] <mhall119> I'm trying to do this in singlet
[16:43] <mhall119> I have a singlet Lens object that knows the name->index map
[16:43] <mhall119> I'm just trying to expose that over dbus, so that my singlet Scopes can query it
[16:43] <mhr3> so how is it going to deal with l10n?
[16:44] <mhall119> it's not, there are no user-viewable strings
[16:44] <mhall119> just a variable name and index number
[16:44] <mhr3> then it's again not going to work with all lenses
[16:45] <mhr3> therefore falls into un-clean category
[16:45] <mhall119> the convenience won't be there
[16:45] <mhr3> you're adding complexity then
[16:45] <mhall119> but you  can replace self.lens.cat1 with 0, and it'll work with non-singlet lenses
[16:45] <mhr3> singlet dev: why is this working here, but not there?
[16:45] <mhall119> yeah, I guess I'm trying too hard on this
[16:46] <mhall119> I just hate the idea of a scope author defining lens constants in their code
[16:47] <mhr3> yea, it kinda sucks... but it's needed atm
[21:17] <mhall119> hmmm, I somehow have unity3d and unity2d running at the same time...
[21:26] <thumper> :)
[21:34] <varikonniemi> hello. There seems to be a problem with unity not being capable of hiding the window decoration when maximizing a windowed spring
[21:35] <varikonniemi> this is how it looks when maximized: https://imgur.com/Nb5hz
[21:45] <JanC> varikonniemi: what's a "windowed spring"?
[21:46] <varikonniemi> spring rts engine running in windowed mode
[21:47] <JanC> looks like the "title bar" in that screenshot is application drawn, or is that your default theme?
[21:47] <varikonniemi> yes its my default theme
[21:49] <varikonniemi> i can trick it to work by repeatedly clicking that maximize button. However immediately after it has drawn one frame properly maximized, spring crashes with sigsegv
[21:51] <JanC> it could be a bug in either Spring or the graphics driver, I guess...
[21:52] <varikonniemi> http://springrts.com/phpbb/viewtopic.php?f=20&t=27563&p=512643#p512643 here is a thread i posted about the issue, and a stacktrace if that could help you hunt the problem down.
[21:54] <JanC> varikonniemi: I'm a user myself mostly, but maybe someone can use that info
[21:54] <JanC> varikonniemi: maybe best to file a bug also
[21:55] <varikonniemi> yeah, thats why i came here, so the developers could be aware of this problem. And possibly fix it if the problem is on ubuntus side. Spring is probably the best rts engine for opensource at the moment, it would be a shame if something as trivial as this slipped through into precise.
[21:56] <JanC> well, first thing to determine would be who's "fault" this is  ;)
[21:57] <JanC> (without blaming anybody personally, of course, but determining where the fix should go)
[21:58] <varikonniemi> well, reportedly it does work in gnome-shell correctly
[21:58] <varikonniemi> i think i will install gnome shell on my computer so i can try it out for myself and possibly narrow it down
[21:59] <JanC> varikonniemi: "it works with ..." is no proof of correctness, of course...
[21:59] <JanC> and might even depend on hardware/drivers... :-/
[22:08] <mhall119> mhr3: any idea what this means:   libunity-WARNING **: unity-scope-proxy-remote.vala:104: Unable to connect to Scope (/unity/singlet/lens/community/locoteams @ unity.singlet.lens.community.locoteams): GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface `com.canonical.Unity.Scope' on object at path /unity/singlet/lens/community/locoteams
[22:09] <mhall119> davidcalle: ^^
[22:09] <mhr3> mhall119, did you call export on the scope?
[22:10] <mhr3> does the dbus path match?
[22:10] <mhall119> yeah
[22:10] <mhr3> the lens doesn't think so
[22:11] <mhr3> also is the scope running or are you utilizing dbus activation?
[22:11] <mhall119> my scope's __init__ is calling:
[22:11] <mhall119> self._scope = Unity.Scope.new ("%s" % self._meta.bus_path)
[22:11] <mhall119> I'm running the scope as a daemon
[22:12] <mhr3> so the daemon was running when you started the lens?
[22:13] <varikonniemi> yeah, in gnome-shell it works perfectly also on my computer, so i really suspect the problem lies with unity
[22:13] <mhall119> mhr3: I killed the locoteams-scope process and started it again, then got that error
[22:14] <mhr3> mhall119, so the scope is working properly now?
[22:15] <mhall119> nothing is quite working yet
[22:16] <mhr3> cause if you kill the daemon the lens will try to connect to it again
[22:16] <mhr3> if it can't do that you'll see ^^
[22:17] <mhr3> but it may also mean that you export the scope too late
[22:17] <mhr3> it should be up right after dbus connection is acquired
[22:18] <mhr3> but if you just call export soon enough, that shouldn't be an issue
[22:21] <mhall119> mhr3: how about GLib-GIO-CRITICAL **: g_dbus_connection_register_object: assertion `object_path != NULL && g_variant_is_object_path (object_path)' failed
[22:21] <mhall119> from the lens daemon
[22:21] <mhr3> yea, that's bad
[22:21] <mhall119> when I start it
[22:22] <mhr3> make sure you aren't using dashes in the object path
[22:22] <mhr3> dbus doesn't like that
[22:22] <mhall119> there shouldn't be
[22:22] <mhr3> what do you pass as the object path?
[22:24] <mhall119> http://paste.ubuntu.com/821946/ is from pdb
[22:25] <mhall119> I assume it's okay for a lens to be 'unity.singlet.lens.community' and a separate scope to be 'unity.singlet.lens.community.locoteams'
[22:26] <mhr3> mhall119, you're passing the path to the lens constructor, right?
[22:26] <mhr3> not the name
[22:28] <mhall119> well crap
[22:28] <mhall119> good catch
[22:30] <davidcalle> mhall119, I've seen this one.
[22:31] <mhall119> davidcalle: mhr3 got me (mostly) sorted, I was passing the wrong dbus info
[22:32] <davidcalle> mhall119, yeah, I've just seen it, Xchat wasn't scrolled to the bottom :)
[22:54] <mhall119> davidcalle: how sould we deal with a lens or scope that uses a separate python file?
[22:55] <mhall119> for example, I have locodir.py that's a thin client for the LTP json API
[22:56] <mhall119> I guess these will all go into wherever in /opt the lens executable goes, huh?
[22:58] <mhall119> mhr3: davidcalle: what signal should I connect to in order to do a search and fill results when the user clicks on the lens, but hasn't entered anything yet?
[23:00] <mhr3> mhall119, search-changed
[23:04] <mhall119> mhr3: I have that, but it's not calling my search
[23:07] <mhall119> can I no longer use http urls for the icon field in the response?
[23:12] <mhr3> mhall119, it should work fine, you'll get it just once though, not everytime you switch to your lens
[23:27] <davidcalle> mhall119, sorry, was out. Yes, everything but the .service, .lens or .scope, in /opt/folder
[23:42] <mhall119> mhr3: passing https://api.launchpad.net/1.0/~ubuntu-us-florida/mugshot in the icon's place of the result model doesn't put that image on the item in the dash
[23:43] <mhall119> it worked in oneiric
[23:45] <mhr3> that would be an issue inside unity itself
[23:45] <mhr3> it works with music lens though... and it's using http uris for the icons as well
[23:47] <davidcalle> I can confirm that I'm using http icons all the time, on Precise. Home Dash & lens view.
[23:47] <mhall119> mhr3: hmm, it doesn't seem to show cover art for anything but what I have locally on the music lens
[23:47] <mhr3> mhall119, what can i say... works here :)
[23:47] <mhall119> dang, I hate "can't reproduce" bugs
[23:48] <mhr3> same
[23:48] <mhr3> mhall119, are you behind a proxy by any chance?
[23:48] <mhall119> nope
[23:48] <davidcalle> mhall119, https://code.launchpad.net/~scopes-packagers/+archive/ppa/+packages can you try the grooveshark scope for Precise and tell me if you see the icons?
[23:49] <davidcalle> Ok... I don't see http icons anymore. It was fine 5 minutes ago :)
[23:50] <mhall119> heh
[23:50] <mhall119> davidcalle: will I need to restart unity to get the grooveshark scope?
[23:51] <davidcalle> mhall119, killall unity-music-daemon
[23:51] <davidcalle> mhall119, that should do it. No need to restart Unity.
[23:51] <mhall119> what should I try searching for
[23:52] <mhr3> bschaefer, ping?
[23:52] <davidcalle> anything you want in the Music lens.
[23:53] <mhall119> no pretty covers
[23:53] <davidcalle> mhall119, but results?
[23:54] <mhall119> yes
[23:54] <davidcalle> Ok
[23:54] <mhall119> not sure where they're coming from
[23:54] <mhall119> but I see results
[23:54] <mhall119> some results have cover art, most do not
[23:54]  * davidcalle has restarted Unity, http icons are back...
[23:55] <mhall119> how do you restart unity now, setsid unity, unity --replace, killall -9 compiz?
[23:55] <davidcalle> mhall119, I've noticed that when I search for the same thing a second time, cover art that wasn't here appears.
[23:56] <thumper> cover art is loaded asynchronously
[23:56] <davidcalle> mhall119, I alway do altf2 unity --replace