[01:14] anyone know how to launch the kde panel from commandline (i have kwin installed, i've run that from within a fvwm session) [01:35] ah, startkde, i have evryything now, great === Ursinha` is now known as Ursinha === Ursinha is now known as Guest53271 [01:56] * RAOF is not sure why ceti331's trying to start a KDE session from within a fvwm session rather than just starting a KDE session. [02:56] i didnt have a kde session option displayed [03:04] If you installed kubuntu-desktop, you should have. [03:14] kde seems pretty good overall just flaky screen corner triggering which is a shame ... i like that shortcut [03:14] (i like screen corners bringing up desktop expo) [04:56] Good morning [05:31] good morning [06:16] hey didrocks [06:16] guten morgen pitti. Happy Friday! [06:16] didrocks: happy Friday 13th to you too! [06:16] pitti: oh indeed, it is! [06:48] hey didrocks, pitti [06:48] wow, Gtk3 sure left a lot behind [06:48] bonjour rickspencer3 [06:48] for example, I cannot find how you are supposed to do pixbuf_new_from_data now [06:49] maybe it's just my_image.set_data now? [06:49] grrrrr [06:50] I thought they pretty much just gutted pixbufs in their entirety? [06:51] In favour of cairo surfaces? [06:51] RAOF, oh, I don't know [06:51] there are still pixbuf's in the documentation [06:56] I'll just have nathan write the paste data into a temp file, and load that into the image [06:56] in the meantime ... [06:56] * rickspencer3 shakes fist at Gtk Python programming [06:57] I'm sure you can create a cairo image from data; maybe that's not enough, though. [06:57] RAOF, if just sound unnecessarily hard, and there is zero sample code [06:57] rickspencer3: yeah, it's cairo in the new world :) [06:58] I'm supposed to be introducing someone to how fun programming Ubuntu apps is [06:58] it's not going quite as planned [06:58] :/ [07:07] hey rickspencer3, how are you? [07:07] hi pitti [07:07] I'm super grumpy [07:08] trying to help someone make a simple copy and paste image example with Python and Gtk3 [07:08] and the documentation is a combination of missing and wrong :/ [07:08] which sucks considering none of my pygtk code is relevant anymore [07:08] rickspencer3: did you import GdkPixbuf and try GdkPixbuf.new_from_data ? [07:08] I was fairly sure I tried that already, but I don't remember any more whether it succeeded [07:10] pitti, ah! [07:10] sweet [07:10] As always, I'm sure the C documentation will be more useful. That's what I always look at when I'm doing GTK development. [07:10] :( [07:10] rickspencer3: gdk-pixbuf is a separate module, so it comes with a separate gir [07:11] I had no f'ing idea [07:11] and none of this is http://developer.gnome.org/gdk-pixbuf/unstable//gdk-pixbuf-The-GdkPixbuf-Structure.html#GdkPixbuf [07:12] $ mkdir /tmp/pixbuf; g-ir-doc-tool -o /tmp/pixbuf /usr/share/gir-1.0/GdkPixbuf-2.0.gir; yelp /tmp/pixbuf [07:12] FYI [07:12] pitti, is there a symmetrical function call that gets the data from the pixbuf? [07:12] is that get_pixels? [07:12] I'm waiting for the day when you can do yelp /usr/share/gir-1.0/GdkPixbuf-2.0.gir [07:12] it seems like we are this --><-- close [07:13] that's the Pythonic documentation, BTW [07:13] rickspencer3: almost [07:14] rickspencer3: new_from_data gets a string, while get_pixels() returns an array of bytes [07:14] oh geez [07:14] so how do I get a string? [07:14] you can convert the byte array into a string, hang on [07:28] oh geez, new from data takes 9 arguments :/ === zyga-afk is now known as zyga [07:30] new_from_data takes a *string*? Why doesn't it take a byte array?! [07:33] RAOF, no, it takes a byte array [07:33] does it not? [07:33] it used to === ubuntu is now known as Nathan_S [07:33] msg Nathan_S http://stackoverflow.com/questions/7906814/converting-pil-image-to-gtk-pixbuf [07:33] rickspencer3, RAOF: b'' _is_ a byte array [07:34] Ah, python. [07:34] rickspencer3: I got a DSL reconnect, what was the last line you saw from me? [07:34] get_pixels() is working fine here [07:34] it's the new_from_data I have trouble with [07:35] pitti, I saw that were going to look to convert get_pixels() return value to a string [07:35] but I don't think that is actually necessary [07:35] rickspencer3: no, it's fine [07:35] * pitti replays [07:36] rickspencer3: actually, I lied; new_from_data also takes a bytes array, so they are symmetrical [07:36] rickspencer3: so, get_pixels() seems fine [07:36] this works: [07:36] pb = GdkPixbuf.Pixbuf.new_from_file('/usr/share/pixmaps/debian-logo.png') [07:36] print('data:', pb.get_pixels()) [07:36] data: b'\xbf\x03H\x00 ...' [07:37] pitti, may I assume that what is on the Gtk.clipboard when someone copies an image is a byte array suitable for new_from_data? [07:37] rickspencer3: and I also checked that creating a new pixbuf with new_from_data() from an existing one with get_data() also works fine [07:37] but if I try to use it with a hardcoded one, I get strange results [07:37] pitti, do you not need to supply the 9 arguments? [07:37] yes [07:37] ok [07:38] data = b'\xFF\xFF\xFF\x00\x00\x00\xFF\xFF\xFF\x00\x00\x00' [07:38] pb = GdkPixbuf.Pixbuf.new_from_data(data, GdkPixbuf.Colorspace.RGB, [07:38] False, 8, 2, 2, 6, None, None) [07:38] data, colorspace, has_alpha, bits_per_pixel, width, height, rowstride, and two Nones [07:39] rickspencer3: but it seems it mistreats the data, when I do it like this the get_pixels() result has a text prefix [07:39] *sigh* [07:39] rickspencer3: but that doesn't happen if I get the data from an already existing object, like another pixbuf or clipboard etc. [07:39] rickspencer3: so I think it should work fine for you [07:42] pitti, so, it looks like chromium just puts a pixbuf on the clipboard [07:42] I wonder if we can just do that from python as well [07:44] pitti, omg [07:44] this is actually waaaaay easier than it used to be [07:44] you can just put pixbuf's on the clipboard [07:45] so, all the groveling around for the data and formatting it seems to no longer be needed [07:45] rickspencer3: oh, does GdkPixbuf has a method for putting on the clipboard? [07:45] or maybe it was always like that, and I was just doing it the hardway all along [07:45] or clipboard a method for putting on a GdkPixbuf object? [07:45] pitti, clipboard.set_image(my_pixbuf) [07:45] (probably the latter rather) [07:45] and [07:45] yay [07:46] my_image.set_from_pixbuf(clipboard.wait_for_image()) [07:46] they are now both 1 liners [07:48] I wonder if they always were, but I never went that route because I was setting targets and everything [08:16] hey desktopers === tkamppeter_ is now known as tkamppeter [08:24] It's a seb128! RUN! [08:25] lol [08:25] RAOF, hey! [08:25] * pitti runs like hell [08:25] I didn't do it! [08:25] hey pitti [08:25] * pitti hugs seb128 [08:25] what did you guys break again?! [08:25] ;-) [08:25] * seb128 hugs pitti [08:25] nothing, I swear! [08:25] yeah yeah, don't believe you :p [08:25] Hey seb128 :) [08:25] RAOF, oh, my SRU patch pilot friend ;-) [08:26] RAOF, how are you? [08:26] Tired. [08:26] Also, FRIDAY! WOOOOOOOO [08:26] RAOF, you should call it a week, get a good deserved dinner and enjoy your weekend ;-) [08:27] RAOF, btw well done to Robert and you to get the system compositor work in a testable shape, I plan to install it on my test netbook today ;-) [08:27] I'll just finish the due-diligance on smb's squid3 merge and then have a bath [08:29] Boo. piuparts should have schroot integration. === ubuntu is now known as Guest86821 === Guest86821 is now known as Nathan_S === fenris is now known as Guest73296 === ricotz_ is now known as ricotz === MacSlow is now known as MacSlow|lunch === zyga is now known as zyga-food === zyga-food is now known as zyga [13:13] seb128: session-migration in NEW on all archs [13:15] didrocks, NEWed [13:15] seb128: \o/ thanks :) [13:15] now, let's chase mterry once he's around :) === MacSlow|lunch is now known as MacSlow [13:48] mterry: hey, how are you? :) [13:48] didrocks, great. Yourself? [13:48] mterry: I'm fine, thanks :) [13:49] do you have some time for a quick MIR? It won't hurt! ;) [13:51] hey mterry, how are you? [13:51] didrocks, sure, throw it over [13:51] pitti, hi! good [13:51] didrocks, aren't you on the MIR team? [13:51] hey mterry [13:52] mterry: bug #1024381 [13:52] Launchpad bug 1024381 in session-migration "[MIR] session-migration" [Undecided,New] https://launchpad.net/bugs/1024381 [13:52] mterry: well, I am, but it's my package :) [13:52] seb128, good EST morning :) [13:52] I can ack it if you want :p [13:52] mterry, good GMT afternoon ;-) [13:52] it's excellent "stuff" [13:52] mterry, happy friday 13th as well [13:52] * kenvandine does the friday the 13th dance [13:53] hey kenvandine [13:53] hey seb128 [13:53] kenvandine, congrats on being officially our new chromium maintainer ;-) [13:53] you got it to build and updated! [13:53] NO WAY! [13:53] temporary :) [13:53] kenvandine, you just get to fix powerpc now ;-) [13:53] hehe [13:53] apparently it has never been build for powerpc [13:54] so not my problem :) [13:54] mterry: do you plan to fix the update-manager autopkgtest (missing python module dependency, etc.), or should someone else look into this? [13:54] pitti, I thought I did fix it. What's missing now? [13:54] pitti, is there a failure log? [13:54] mterry: https://jenkins.qa.ubuntu.com/view/Quantal/view/AutoPkg%20Test/job/quantal-adt-update-manager/9/ARCH=i386,label=albali/console [13:54] mterry: apparenty python-coverage is missing [13:55] mterry: also, https://jenkins.qa.ubuntu.com/view/Quantal/view/AutoPkg%20Test/job/quantal-adt-update-manager/9/ARCH=i386,label=albali/artifact/results/dsc0t-nose-tests-stderr has some stuff; autopkgtest fails the tests on any stderr output [13:56] pitti, guh, didn't know it did that [13:56] mterry: I recommend something like this: [13:56] unittest.main(testRunner=unittest.TextTestRunner(stream=sys.stdout, verbosity=2)) [13:56] pitti, python3-coverage doesn't exist yet [13:56] pitti, but the setup.cfg enables it, for when it does exist. didn't think the noise would hurt anything [14:01] mterry: for testing you can run sudo adt-run --no-built-binaries update-manager_*.dsc --- adt-virt-null locally (or in a VM) [14:01] and then check $? [14:02] I remember running the test suite in a pbuilder, but not under adt-run, which was my fatal error [14:14] mterry: tell me if/when you have some time to review it, I've a pending gnome-session upload for it [14:14] mterry: if you think you can't make it, Monday is fine :) [14:14] didrocks, k [14:24] mterry: I think you asked about udisks over night [14:24] mterry: I have no particular plans to update it again, unless someone asks me about something specific; it's obsolete, and we are transitioning to udisks2 [14:24] pitti, ah right [14:24] we need to fix update-notifier and usb-creator-common still, though [14:25] pitti, does adt-run not have internet access? [14:25] mterry: it does [14:26] (on jenkins) [14:26] ok [14:31] pitti, I've subscribed to u-m and u-r-u jenkins failure RSS [14:31] pitti, u-m should be better in trunk. looking at u-r-u failure [14:31] mterry: nice, thanks! [14:32] mterry: please note that autopkgtest runs currently need to be kicked off manually [14:32] we'll fix this on Monday when jibel is back [14:32] mterry: so if you want a manual test, please prod someone from the QA team to press the button [14:32] pitti, I'm assuming it tests the archive, not trunk of the packages? [14:33] mterry: right [14:33] mterry: whenever the pacakge itself, or any of its dependencies change [14:33] pitti, OK. I wasn't planning to release anew for the test fix, but I can if you'd prefer [14:33] mterry: not that urgent; next week is fine, when adt is back [14:52] good night everyone, have a nice weekend! [14:59] have a nice week-end pitti :) [15:00] didrocks, in session-migration, you want DESKTOP_SESSION not XDG_CURRENT_DESKTOP? [15:00] i.e. you want to be able to distinguish between unity-2d and unity-3d or gnome fallback mode etc? [15:00] mterry: yeah, DESKTOP_SESSION, because we maybe have separate gsettings database later on based on session, not on desktop [15:00] exactly [15:08] hmmmm, can people upgrade properly even with the latest python3-defaults? :/ [15:09] chrisccoulson, those issues were supposed to be fixed yesterday, is your mirror uptodate? [15:09] chrisccoulson: no, manually install the python3 package [15:10] it's fixed, but if you're in a broken state, you need to manually fix it [15:10] seb128, yeah, i can see the latest version but it won't upgrade with the broken state [15:10] mdeslaur, thanks. i'll try that :) [15:10] i don't want to spend too much time on my computer ;) [15:11] chrisccoulson, btw stop working, you are supposed to be on vac ;-) [15:11] chrisccoulson: sudo dpkg -i /var/cache/apt/archives/python3_3.2.3-3ubuntu1_all.deb [15:11] chrisccoulson: and then turn it completely off :) [15:12] seb128, yeah, i should just turn my laptop off ;) [15:12] chrisccoulson, good! ;-) [15:12] ah, that's better, it's upgrading now :) [15:13] chrisccoulson, you say that because you don't know how buggy those updates are ;-) [15:14] seb128, oh, there's not much to update here actually. i've got the new indicator stack already ;) [15:15] chrisccoulson, ;-) [15:22] mterry: thanks for the review! [15:23] yw! === ubuntu is now known as Nathan_S [16:55] * didrocks waves good evening and week-end! [17:29] desrt, what's the status of your wip/action_helper branch? [17:38] kenvandine: waiting for lars to get back from vacation [17:38] ok, i uploaded it to a ppa so we can hit the ground running when he returns [17:39] also uploading libindicator and libmessagingmenu [18:00] can someone test: change your presence with empathy 3.4.2.1 on precise with the empathy window, then try changing it with the indicator menu. Mine always goes to available and then does not change. [18:01] kenvandine: ^ if you have precise [18:03] hmm, maybe ken had to leave early today === Amaranthus is now known as Amaranth