[00:15] bschaefer: cool, what was the issue? [00:15] the display [00:16] thomi, so when it was done the display wasn't getting closed, but we didn't actually need the display [00:16] ahh ok. Let me know if you want me to re-review your MP [00:16] yup, working on the autopilot test right now [00:16] coolio [00:17] so far I changed it to test_ctrl_a, test_ctrl_c_v and test_ctrl_x_v [00:17] which will test for that crash anyone [00:17] anyway* [00:17] bschaefer: cool [00:17] bschaefer: while you're at it, I think you should change the text to something nicer than "Segfault" :) [00:17] yeah [00:18] haha [00:18] also thanks for the novel, its very helpful [00:18] thomi, that test is getting removed anyway [00:18] its replaced with ctrl_c_v which just uses Copy as the text [00:18] heh, ok [00:18] cool [00:19] and Cut for the ctrl_x_v one [00:19] thomi, o yeah, should I be putting these under DashKeyNavTests? [00:20] bschaefer: hmm, I'd make a new class, something like DashClipboardTests [00:20] since it's not really keyboard navigation [00:20] alright sounds good [00:23] Is precise going to ship Unity 5.6.0? [00:35] Daekdroom: or later [00:41] jono: I've installed Ubuntu 12.04 on VirtualBox too so I can test "Ubuntu Accomplishments" and if it's not too hard I can help implementing other accomplishments [00:42] Andy80, awesome! [00:43] if you can install and run it that would be cool as a first step [00:43] and some contributed accomplishments would be great [00:43] * DebolazW likes this idea. [00:43] Andy80, for creating them, see https://wiki.ubuntu.com/Accomplishments/Creating [00:44] DebolazW, if you can help, that would be cool too :-) [00:44] I think I would fall quite short if attempting to think up new achievements. :) [00:45] jono: yes, I already installed and tested on my netbook. Now I'd like to try again on the desktop so I can work better on the code. And yes, I was referring to your blogpost for all the informations :) the wiki is written very well and all the steps worked without any problem [00:46] Andy80, cool, so you saw the trophies appear OK? [00:46] DebolazW, you can also help improve the docs on our existing ones: https://wiki.ubuntu.com/Accomplishments/GetInvolved#Documentation [00:48] jono: yes, the one I have appeared. There was just a delay on the "First bug filled" trophy, but I could not understand why. [00:48] Andy80, that is because it was unlocked until you got the registered on LP one :-) [00:48] they come in in waves :-) [00:49] * thumper waves [00:49] jono: ahhh, now I understand :) it's something like .deb deps ;) to have one you need to first have another one ecc... [00:49] indeed [00:50] it helps map out the community journey [00:50] would be awesome to have trophies for Unity team contributions [00:50] e.g. First Bug Fix [00:50] First Successful Merge [00:50] etc [00:52] jono: ok, I'll have to implement something anyway if I want to test how to create them. Any idea, please let's add it to the wiki, possibly something like this: "First Unity bug fixed" - IN PROGRESS - @yourlaunchpadId [00:52] so people can know if another one is already working on it and they won't duplicate code [00:53] indeed [00:53] good idea [01:10] ok, time to sleep now (2:00 AM here), I'll continue tomorrow! See ya guys :) [01:10] thomi, hmm setting the primary text doesnt seem to work [01:12] jono: By the way, while I have you here, which mailing list should I post on if I want to draw attention to a compiz-on-unity bug? [01:13] bschaefer: be with you ina minute [01:14] thomi, alright, dont rush, I pushed the changes I had so far [01:14] DebolazW, probably unity-devel [01:15] bschaefer: OK, so.. when you say it doesn't work... [01:15] you mean that GTK code snippet? [01:15] It's a regression bug that sneaked in for 12.04, it looks very simple to fix but I have no idea how to do it. :-) [01:16] thomi, I make a PRIMARY clipboard, then set the text and try pasting. Nothing is in the PRIMARY clipboard when I try to manual do it also [01:16] bschaefer: it works perfectly for me - where are you trying to paste to? [01:16] https://code.launchpad.net/~brandontschaefer/unity/unity.fix-middle-paste-reg/+merge/94441 [01:17] line 182 [01:17] in the diff [01:17] 192* [01:19] bschaefer: OK, a few things I forgot to mention in my previous novel: [01:19] the most important thing is that the values of classes (like SearchBar) don't automatically refresh when the unity state changes, so you need to call 'searchbar.refresh_state()' if you think it may have changed unity-side [01:19] OR you can just construct the searchbar when you need to read the state [01:20] so diff line 194 should have a search_bar.refresh_state() [01:20] you guys are on fire!! another unity update in precise? [01:20] I should do a searchbar = self.dash.get_searchbar() [01:20] otherwise you'll get the 'search_string' from whenever it was created [01:20] do both? [01:20] bschaefer: or just call refresh_state() [01:21] just one or the other [01:21] ok [01:21] they actually do the same thing - both trigger a dbus call, grab the new state, and set the object properties [01:21] whats weird is I just changed it from PRIMARY to CLIPBOARD and now the text is getting set in the clipboard [01:21] calling refresh_state() looks nicer, sinc eyou've already created the object [01:22] very true [01:22] bschaefer: also, you should really move the mouse to "searchbar.x + (searchbar.width/2)" instead of "searchbar.x + 100"... [01:24] yeah, I saw it used above in a different test [01:24] ok [01:24] saw it used that way, but that makes more sense to put it in the middle [01:24] so are you sure middle-mouse-btn paste works in the dash? [01:24] yeah, I also tried to see if it was in the clipboard just using gnome-termainl [01:24] terminal* [01:25] and I can get it to work for the normal CLIPBOARD but not PRIMARY [01:25] bschaefer: OK. everything I need is in that one branch, correct? I'll build the branch and try and see what's goingon [01:25] yes! [01:26] ok, thanks. I added test for just ctrl+c and ctrl+x also [01:26] using Clipboard, so those have test now! [01:26] cool [01:26] * thomi grabs the branch [01:27] thomi, yeah also sorry about the length of my name haha [01:27] that I used for the launchpad login [01:27] no worries - that's what copy+paste is for [01:27] haha, not if its broken! [01:27] building... [01:28] I put the wait_for_text in there since it should be a better timer then sleep [01:29] ok [01:29] I didn't realise you even had to wait :) [01:29] idk if you do or not [01:29] I guess it might take a while to convert to utf8 [01:30] I bet its un needed [01:36] bschaefer: wtf? [01:36] wow, that's odd [01:36] yeah [01:37] OK, when I run it manually myself it works... [01:37] on the command line? [01:37] or python interpreter [01:38] yeah it does thats weird! [01:40] bschaefer: in the python interpreter [01:40] bschaefer: also, your indentation is a bit messed up [01:40] all indents should be 4 spaces [01:40] yeah, I just tested it in the interpreter and it worked [01:40] in the python files, that is [01:40] ops, have it set to 2 for unity [01:41] I was wondering why the functions were 4 spaces away [01:44] hmmm, OK, that's interesting [01:45] if you set the clipboard text, then exit the interpreter, the primary clipboard gets cleared, which makes sense, since it's a client-side property [01:45] but that's not happening in our test... [01:46] I also tried cb.store() which should store it for when the app closes [01:46] the primary text doesn't get cleared when I exit from the interpreter [01:47] thomi, nerver mind it does [01:47] huh? does what? [01:47] it clears the primary text [01:47] oh [01:47] ok [01:47] sorry [01:47] I'm with you [01:49] hmm but it should still work since the program is still running! [01:50] ahhhh [01:50] * thomi has an idea [01:51] what is it! [01:58] I wonder whether it's because our Mouse emulator uses python-XLib and we create our own display instance, while gtk uses the gtk.gdk.Display class [01:58] hmmm... [01:58] o i bet if we used the main display for the clipboard it would work [01:58] hmm well [01:58] it should actually already do that [01:59] thomi, one thing I found odd is if you replace PRIMARY with CLIPBOARD [01:59] thomi, then do a cb.store() after the set text and run the problem [01:59] and if you paste anywhere its the set_text worked [01:59] ok, I'm getting closer [01:59] program* [02:00] I have a tiny function now that works when run from the interpreter, but not from the command line [02:00] Mainly curious here: Why was the switch workspace shortcut changed? [02:00] and it's only 5 lines long [02:01] nice! [02:01] could you pastebin it? [02:01] or just paste it here [02:02] yeah.. [02:02] bschaefer: http://pastebin.ubuntu.com/854850/ [02:03] if you download that to a file called 'cp.py' in your tests/autopilot dir [02:03] and then open a python interpreter and do this: [02:03] from cp import test_middle_mouse_paste; test_middle_mouse_paste() [02:03] it works perfectly [02:04] but if you run to from bash: (python cp.py) it doesn't do anything [02:04] hmm thats odd [02:04] so what does th interpreter do differently? [02:05] exactly - I can't think of anything [02:07] ok, using the 'xsel' app, it looks like that gtk method isn't doing what we think it is [02:07] the set_text method? [02:09] yeah [02:10] is the set_text only for the CLIPBOARD? [02:10] as I think it is used for copy more so [02:11] no, otherwise the interpreter wouldn't work... [02:11] no, the docs indicate it should work [02:11] and yeah, it works interactivelky [02:17] hmm weird, so if you run that program "python cp.py" and while it is sleeping select and copy something into primary [02:17] and it works [02:21] wow, ok [02:23] hah! got it [02:24] what was it? [02:24] the gtk.Clipboard needs the gtk event loop running. check this out: [02:24] http://pastebin.ubuntu.com/854874/ [02:24] that works perfectly [02:24] awesome [02:25] do we want the gtk main loop running in the test? [02:25] OK, so we need a way to do this in pyre python-xlib. I'll take a gander at the gtk source code and try and see how they do it [02:25] bschaefer: no, we can't have it running :( [02:25] yeah, hmm well I wish selecting and coping in the unity dash set it as primary [02:26] hmm I should look into that [02:26] thomi, why was the gtk main loop running int he interpreter? [02:26] bschaefer: do you use ipython? [02:27] I do, and it runs a gtk loop for you I believe [02:27] hmm I just type python on the command line [02:27] oh ok [02:27] maybe pygtk is trying to be clever, and does it differently when it's being run interactively? [02:27] I could init something [02:27] it* [02:28] in either case, it'd be nicer to do it in XLib [02:28] yeah [02:28] well, I say nicer.... [02:28] ;) [02:28] haha [02:28] that was definitively interesting, what made you think we needed the gtk loop? [02:29] dunno... just a wild guess [02:29] I'm grabbing the gtk source [02:29] haha either way good job! I was just getting more and more confused haha [02:32] right, the answer we need is in here somewhere: http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/precise/xsel/precise/view/head:/xsel.c [02:33] ahhh, nicely written C code... such a pleasure to read [02:35] hmm, so we are looking for some sort of event loop? [02:36] and yes this is very nice C code... [02:36] line 1468 is what we need I think [02:37] I just saw line 1624 [02:38] which looked like it was waiting for the next event... [02:40] I'm hoping we can do a fire-and-forget [02:40] * thomi hacks something together [02:45] nice, im interested to see what you hack together! [02:53] i hate to sound like a broken record, but is anyone going to fix fglrx? [02:54] whenever i try to play a video with totem/gstreamer x just crashes.. and it still has the window title not being updated bug [02:55] im on the radeon driver atm, cos that works without these issues, but obviously that solution isnt the most ideal [02:58] i fear that 12.04 will be released with this issue, because nobody cares enough to fix it [02:59] bschaefer: I see how it works... it's a bit of a pain [03:00] thomi, :(, if it will take much time there is still a manual test [03:00] that will cover this [03:00] basically, you need to create a window (it can be unmapped, so that's OK), and whenever you want to set the clipboard contents, you raise an X event. Then when someone pastes something, you get a XSelectionEvent, which you need to respond to sending the selection data [03:00] bschaefer: yeah, let's go with the manual test. [03:01] bschaefer: delete the AP test for MMB paste I guess [03:01] thomi, ok, let me remove that test :(, but ill keep the other ones [03:01] yeah [03:01] remember to fix the indentation also :) [03:01] already did :) [03:01] oh, and if you can, trim trailing whitespace :) [03:01] ok [03:01] cheers [03:02] thank you for spending so much time working on this! [03:02] ahh well, I now know a bit more about how X works.... I'm not sure that's a good thing or not [03:02] heh, no worries [03:02] haha yeah I didn't know copy and pasting had so many standards before today [03:07] thomi, alright, just pushed the changes if you want to review [03:07] * thomi looks [03:09] bschaefer: are you satisfied that you haven't introduced a memory leak in the new clipboard code? [03:09] I notice that the old code used a glib::GObject<> smart ptr thingy [03:09] ...and now we don't [03:10] o, umm I should add that back in [03:10] I thought it was casting it rather then using a smart ptr [03:10] no hang on [03:10] the docs say: [03:10] Returns : [03:10] the appropriate clipboard object. If no clipboard already exists, a new one will be created. Once a clipboard object has been created, it is persistent and, since it is owned by GTK+, must not be freed or unreffed. [transfer none] [03:11] better not unref it then :) [03:11] yeah haha, so it should be good :) [03:14] bschaefer: approved. Do you have permissions to set the MP to approved? [03:15] yeah [03:15] cool - do it :) [03:15] alright! [03:15] nice, not copy/cut and pasting isn't crashing and works! [03:15] now* [03:16] thanks for your help [03:27] Is someone able to point me to a document/spec that explains why we needed to change workspace switching keyboard shortcuts? IMO the ones we had are fine and have been around for a long time. [06:22] What the hell just happened to my inbox? [06:24] hmm.. caps lock / num lock / scroll lock hints aren't showing in latest unity [06:24] im on a netbook which doesnt have led indicators for them [08:06] wow, interesting thread on the unity-design mailing list :( [08:16] AlanBell: Which one? [08:16] [Unity-design] Remove Pornographic Apps from the "Apps available for download" (51 messages) [08:17] not sure adding my opinion to that would improve matters, or state anything not already stated [08:18] I don't like the adverts at all, and I have seen p0rnview pop up in it too [08:18] though like most I find religion far more offensive than porn, but that isn't really the point [08:19] what really offends me is that the lens folds up the stuff I do have, to make room for adverts for stuff I don't have [08:21] OMG huge trolling there [08:29] i hate how my porn comes up in recently used documents [08:29] or downloads [08:29] its like.. err.. glad this is my home computer, and not my work one :p === alan_g is now known as alan_g|afk [08:32] snadge, do you know that you can prevent some folders to appear in here? [08:33] no [08:34] i thought alternatively i could create a seperate login .. and do all my dirty masturbation and porn from that account [08:34] but y'know, that involves logging out and back in again.. effort ;) [08:39] snadge, look into System Settings -> Privacy -> Files [08:46] im joking.. im a grown man, in a relationship.. with a 1 week old child [08:46] .. which is probably why i need to masturbate.. but anyway j/k [09:22] didrocks, hello, would you have a hand for a bug fixes upload of the video lens? [09:22] davidcalle: are they critical bugs? [09:22] didrocks, nope. One high. [09:23] https://bugs.launchpad.net/ubuntu/+source/unity-lens-video/+bug/939279 [09:23] Ubuntu bug 939279 in unity-lens-video (Ubuntu) "video lens slows down pc makes it unusable" [High,Fix committed] [09:23] davidcalle: hum, can be handy to upload this [09:23] davidcalle: we only upload high/critical bugs [09:23] fixes* [09:24] didrocks, it's a corner case, but still. [09:24] as we are in beta freeze [09:24] ah corner case [09:24] it's the one where Video doesn't exist? [09:24] and so fallback to $HOME [09:24] didrocks, no, this one is in since yesterday (kenvandine pushed it) [09:25] It's the one where the folder contains a lot of vids and they havn't been seen in Nautilus (so no thumbnails). [09:25] davidcalle: ok, corner case [09:25] davidcalle: can wait after beta1, isn't it? [09:25] didrocks, sure we can. [09:26] davidcalle: let's plan on that then! :) [09:27] I will use this time to test it in a PPA on people affected by it. [09:27] didrocks, fine with me. :) [09:27] yeah ;) === om26er_ is now known as om26er [10:55] didrocks, ok found a way to tirgger the 'white on black numbers' getting stuck; using the new Super+Shift+ keys, makes it come out and get stuck pretty much all the time [10:55] hello, I just did a precise -> precise upgrade after a couple of weeks blocked on an apt bug. Unity seems different. Is there something summarizing the diffs between current precise and, say, oneiric? [11:01] apw: yeah, I saw that I can trigger that quite easily as well [11:01] apw: can you comment on your bug report? (and give me the number back) [11:02] apw: I'm adding it to the priority list now that it's easy to get it [11:04] didrocks, hmm its been dup'd against another bug which sounds different, my original was bug #933630 [11:04] Launchpad bug 933630 in unity (Ubuntu) "Launcher doesn't auto-hide after using Super+1..9 shortcuts" [Undecided,Confirmed] https://launchpad.net/bugs/933630 [11:04] arrgle no thats the dup [11:04] mine is bug #934084 [11:04] Launchpad bug 933630 in unity (Ubuntu) "duplicate for #934084 Launcher doesn't auto-hide after using Super+1..9 shortcuts" [Undecided,Confirmed] https://launchpad.net/bugs/933630 [11:06] apw: agreed, the dup makes no sense [11:06] didrocks, anyhow added comment to the dup [11:06] apw: I'm undupping it [11:06] apw: and taking yours on the list [11:06] and the original :) [11:07] didrocks, also just filed this one: [11:07] bug #940201 [11:07] Launchpad bug 940201 in unity (Ubuntu) "Super+Shift+ viewport switchers also trigger Launcher/Dash/Unity Help popups depending on duration" [Undecided,New] https://launchpad.net/bugs/940201 [11:07] apw: ah, this is a dup :) [11:07] one sec :p [11:08] apw: it's a dup of bug #939521 [11:08] Launchpad bug 939521 in unity (Ubuntu) "Shortcut overlay appears even if you Super + another key" [High,Triaged] https://launchpad.net/bugs/939521 [11:08] apw: I filed it, we will have compiz enabling what we need to do just after beta1 [11:08] I'll fix this afterwards [11:08] didrocks, ok thanks [11:08] apw: btw, I'm interesting to gather feedback on those new keybindings [11:08] apw: I can't get used to them after 3 days here [11:09] so if you are trying some for some days, I'm interested in the feedback [11:10] i thought that we wern't removing the old ones yet, that was a bit of a supprise [11:10] didrocks, and for me, i think i have reached the number of bindings i can remember which all use the same keys [11:11] apw: yeah, compiz doesn't allow multiple keybindings for an action :/ [11:11] 'morning [11:11] apw: ahah, indeed :) [11:11] hey Andy80 [11:12] apw: at least, I tried to be clear in the changelogs for compiz/c-p-m/metacity to document the changes [11:12] didrocks, we must have a few lines of change for the release notes, perhaps we could beef those up for this purpose [11:13] didrocks: is it ok for you if I temporary switch to another task (helping Jono with Ubuntu Accomplishments) while the u1db guys complete the migration to the new API system? [11:13] apw: agree, I'm noting that in a note [11:13] didrocks, as we'll have to release note a complete change of key bindings surely [11:13] Andy80: sure sure, there is no hurry at all one the OneConf side before UDS, take your time! :) [11:14] didrocks: cool thanks :) [11:14] apw: yeah, we definitively need to do that. Noting that down :) [11:17] didrocks, are we going to get a ticky in appearance to allow us to turn off the help, or at least move it out to like 10s [11:18] as its pretty annoying when waht you wanted was the white on black numbers to appear [11:18] and its timed at the exact same time [11:18] apw: yeah, the timeout will get longer after discussing with John, but also, Super + another key will not show the help [11:18] basically, it's the 2 planned changes [11:19] (maybe not 10s, but a little bit more than the one used for the numbers ;)) [11:19] didrocks, i'd like a slidey, so as i get better at them i can move it right, eventually only getting it if i am very persistant [11:21] apw: yeah, i'll discuss this with John, but yesterday he basically agreed to get a longer timeout (he's not around until Tuesday, so will be after beta1) [11:22] didrocks, gawd you know how to mess with my head and no mistake, changing every single window management keybinding, and not keeping the originals ... gurgle [11:23] apw: yeah, I totally agree and I was not supportive of the change for the record [11:23] isnt holding down super supposed to bring up the shortcuts menu? [11:23] or has that been disabled [11:23] yeah, it is [11:24] it does* [11:24] you can turn it off though [11:24] its not coming up on my netbook.. im running the latest unity build in precise [11:24] oh, I think it doesn't come up on short screens [11:24] maybe its a resolution dependant thing yeah.. i have 1024x600 [11:24] I seem to recall seeing something about that in the code [11:25] bummer [11:25] apw: I would even not speak about the 7 hours to dig into different compiz sources (and metacity) to ensure everything is working and picking the right value at the right place. You can confort yourself in knowing it gave me a real bad headache ;) [11:25] perhaps it could be shrunk or made wider on short screens? [11:25] it also gets in the way of enhanced zoom on super+mousewheel [11:25] yeah ive noticed that [11:25] didrocks, heh sounds like working with the kernel :) [11:25] the proper thing to do would be to have the shortcut screen turn off when zoom is activated [11:25] apw: ahah, some kind of surgery indeed :) [11:26] also you can't read it with orca which is a bit faily [11:26] either that.. or the shortcut screen should zoom along with the rest of the desktop.. maybe its too small and you want to blow it up so you can see it better === _salem is now known as salem_ [11:36] didrocks, i have more odd behaviour, i am no longer getting window titles in the menubar, and holding alt for a lot of windows does not reveal their menu, is this right ? [11:38] apw: hum, doesn't seem related to latest change but rather bamfdaeamon going crazy (for both issues) [11:38] hm, i have this too [11:38] apw: you can poke d-feet about bamfdaemon [11:38] didrocks, odd over time something i did has made weechat now show it, whereas before it did not, but pidjin still not [11:38] apw, does ctrl+alt+right arrow jump to right workspace? [11:38] apw: if you can dump all bamf data, then, DBO can look at them [11:39] ah, hud-service[2129]: segfault at 10 ip b74bbd24 sp bf859a2c error 4 in libglib-2.0.so.0.3118.0[b743f000+f6000] [11:39] didrocks, erm, what is bamf and how can i get to its brains [11:39] but hud works [11:39] rye, think that is super-shift-left/right now [11:39] smb, well, does not work either, brings hints overlay [11:39] rye, does both for me, and thats a known issue [11:39] I got that overlay thing too but it also shifts [11:40] bug 940198 [11:40] Launchpad bug 940198 in unity (Ubuntu) "Help screen activates when switching desktops" [Undecided,New] https://launchpad.net/bugs/940198 [11:40] apw: do you have d-feet installed? [11:40] nope [11:40] apw: bamf is what matches your opened apps to the window [11:40] apw: please install it :) [11:41] then, go the session bus (connect to it if needed) [11:41] click on org.ayatana.bamf [11:41] than, you will see Object Paths like: /org/ayatana/bamf/application<…> [11:42] opens org.ayatana.bamf.application interface for all of them [11:42] and run DesktopFile() Methods === mmrazik is now known as mmrazik|lunch [11:42] tell me if you have a desktop file containing pidgin [11:43] for weechat, you have a gnome-terminal icon, isn't it? <- /me uses weechat as well [11:43] ok i have a button which looks like an elephant and hitting that says "Bus Address:" any dea what the answer is [11:43] didrocks, actually no, i have a weechat icon :) [11:44] oh? [11:44] didrocks, i couldn't cope with unity merging everything i run under one icon, so i worked out how to convince it to keep them separate [11:44] ah, you added a weechat .desktop file? [11:44] apw: in the menu File -> Connect to session bug [11:44] bus* [11:45] didrocks, yeah have one for weechat and one for mutt, so i get separate launcher entries [11:45] Exec=gnome-terminal --class weechat --disable-factory --window-with-profile=Ambiance -e weechat-curses [11:45] the key bit is the class needs to be unique to get unity launcher to keep them appart [11:46]