[00:00] TheMuso, oh, I did? [00:01] nice :) [00:01] TheMuso, no more bugging you for uploads :) [00:13] yay! :p [00:13] hey [00:13] robert_ancell, there? [00:13] seb128, hey [00:13] robert_ancell, hello [00:13] seb128, how's it going? [00:14] seb128, I had an attempt of updating GTK+ yesterday but the packaging is doing my head in :) Do you know how debian is going? [00:15] robert_ancell, out of some IRC issues and launchpad being down when I wanted to upload a glib fix that's going alright there ;-) [00:16] robert_ancell, not really, I've watched their svn but not a lot of activity, you meant update 2.22 or 3? [00:16] 2.22 [00:16] robert_ancell, the empathy build failure is my fault but I can't upload the fixed glib [00:16] seb128, oh, I was just looking at that :) [00:16] robert_ancell, congratulation for your new uploads rights btw [00:16] yay! [00:17] robert_ancell, the fix is commited and pushed but I got a "connection refused" on upload [00:17] np [00:17] and checked the clock it was 3 minutes after "the launchpad will be down for update" [00:17] robert_ancell, if you can upload already feel free to sponsor the vcs version [00:18] they like to bring it down when the people this side of the world start work [00:18] cool, will do [00:18] robert_ancell, Josselin moved the commands from the bin to the lib and I forgot the compat part of the changes [00:18] oh, I'll have to update the gcalctool dependencies [00:18] there is still a bug that the gio .pc has the wrong path for the schemas register command [00:19] robert_ancell, well the -bin should work the way it uses [00:19] really want the new GTK+ as there are a number of packages that depend on it [00:19] it has compat entries in bin [00:19] ok, gseal work? [00:19] I can try to get things moving in debian tomorrow [00:19] not sure, just required in configure.ac for some packages [00:19] robert_ancell, is there any update you want to get in ubuntu? [00:20] .31 update? [00:20] robert_ancell, btw I'm not sure what should pull dconf in [00:20] robert_ancell, yeah, what is blocked on the gtk update? [00:20] I think it should be ubuntu-desktop? [00:20] robert_ancell, I made glib recommends it for now [00:20] vinagre at least [00:20] ok [00:21] robert_ancell, well ideally dh_gsettings or whatever is the new dh_gconf is should do it [00:21] the same way dh_gconf was doing [00:21] set a misc;Depends on libdconf0 is there is a gsettings schemas [00:21] I thought we didn't need that now we have the triggers? [00:21] * TheMuso just noted libdconf was pulled into his maverick chroot. [00:21] well, technically having a gsettings schema does not require dconf. You could have another backends [00:22] backend [00:22] ok [00:22] so let's keep the recommends in glib for now [00:22] it should do the job [00:22] yeah, practically it works [00:22] robert_ancell: congrats on becoming core-dev [00:22] I'm not sure dconf is working though there [00:22] gcalctool doesn't seem to write any settings [00:22] micahg, thanks [00:23] ie the mode is not stored and dconf-viewer is empty [00:23] seb128, yeah, I've noticed that too. I'll investigate [00:23] Ah. libglib2.0-0 recommends libdconf0 [00:23] thanks ;-) [00:23] TheMuso, yeah, see backlog [00:24] seb128, oh, can you update versions? I've made it track the 2.30 packages [00:24] yeah just did [00:24] I have also noticed the system bell seems to be disabled under compiz in maverick. Is this intensional? [00:25] I doubt it but robert_ancell did a merge on debian which was non trivial [00:25] so one change my have skipped on the way [00:25] robert_ancell, done [00:25] groan. The system bell always seems to cause trouble... [00:25] robert_ancell, I should add an update in the cron job ;-) [00:26] robert_ancell: Happy to dig into it [00:26] seb128, yes please :) My sprint list has an item "work out some common webspace for our scripts" [00:26] robert_ancell, speaking of which if you read the meeting log there is a sprint wikipage [00:26] robert_ancell, so feel free to put items there [00:26] seb128, oh I will mwuhahaha! [00:27] ;-) [00:27] seb128, not here: https://wiki.ubuntu.com/DesktopTeam/Meeting/2010-07-06, do you mean the IRC log? [00:27] robert_ancell, I'm very tempted to update nautilus to 2.31 [00:27] seb128, what's the cost? [00:28] I need to check what gconf use it does [00:28] they will likely port it to gsettings this cycle [00:28] but they have tons of nice changes [00:28] like a stack of hundredpapercut issues fixed [00:29] redesigned the copy dialog for conflicts [00:29] that would be a big win [00:29] right [00:30] robert_ancell, IRC logs, yes, seems that rick didn't update the wiki [00:30] robert_ancell, https://wiki.canonical.com/UbuntuPlatform/Sprints/Maverick/Desktop [00:31] it's empty so far but feel free to start adding things ;-) [00:31] robert_ancell, otherwise do you have any opinion about going for the new gobject-introspection? [00:32] seb128, I'm not aware of anything in particular that has changed [00:32] well they changed the abi version [00:32] so that will require renaming our gir debs at least if we do it [00:34] it's yet another transition but it should be doable if we want to go for it [00:34] seb128, is debian going that way? [00:34] the guy handling the gnome-shell daily ppa asked if we were going to update since gjs 0.8 will require it [00:34] right [00:35] dunno, I will have to figure that as well [00:35] but I think we should try to get at least our platform uptodate [00:35] and that's sort part of the platform nowadays [00:35] seb128, are the gir debs just going to merge into the standard lib debs? [00:35] not that I know no, why? [00:35] I think introspection is transitioning from an optional system to a core part of the library interface [00:35] debian created gir-... binaries on purpose [00:37] robert_ancell, well the issue is that installing non soname-versioned files in libraries break co-installability of versions [00:40] robert_ancell, do you think it's wrong? [00:40] in any case I don't want to divert from debian on that [00:42] robert_ancell, ok, time to go to bed there [00:42] seb128, but aren't the gir files matched to the so files? match debian anyway [00:42] seb128, cya [00:42] robert_ancell, no they are not, /usr/lib/girepository-1.0/Wnck-1.0.typelib [00:43] with /usr/lib/libwnck-1.so.22 [00:43] seb128, I'll add it to the sprint list for discussion :) [00:43] we would need to have something in the typelib path that avoid conflict if the library soname change [00:44] I will try to sync with debian guys on their plans for gir and gtk updates [00:44] but for now time to get some sleep [00:45] try to figure what is wrong with dconf if you can today and let me know if you figure something ;-) [00:45] bye === bjf is now known as bjf[afk] [01:29] Looks like LP is back. [01:44] I have an acer aspire 3810TZ that has an atheros wireless N built in and was running Windows Vista Home Premium. With Windows I got 130mb/s. Now I am running Ubuntu 10.4 and I get only 1mb/s or unknown. Does anyone know how to fix this problem? [01:46] from what version of ubuntu was apport intorduced? [01:48] mediacenter: This isn't a support channel; #ubuntu or ubuntuforums.org will be better. That said, it sounds like you're reading the numbers reported by “Connection information”, which might simply be wrong. This is likely to be a bug (probably in the drivers). [01:48] jenkins: looks like edgy [01:48] micahg: wow that long ago, cool thanks [01:49] RAOF: Thanks for your help. === bjf[afk] is now known as bjf === bjf is now known as bjf[afk] === bjf[afk] is now known as bjf [03:16] TheMuso, I want to release glib from bzr that seb did, how do I sign the package without changing the changelog? [03:16] rickspencer3, hey [03:16] robert_ancell: don't edit the changelog, surely :P [03:17] lifeless, yes, I don't want to edit the changelog, but when you run debuild -S it complains it can't sign as seb [03:18] DEB_SIGNING_KEY or something. hang on [03:18] robert_ancell: one sec, will check the repo myself. [03:18] robert_ancell: use -uc -us on the end of the debuild command [03:18] Then run debsign over the changes file. [03:18] TheMuso, thanks [03:19] or debuild -S -m robertancell [03:19] yeah that works too [03:19] actually [03:19] -k [03:20] is the one to use [03:20] too many options... [03:20] hmm, still not working [03:20] I'm running: bzr-buildpackage -S -- -k robert.ancell@canonical.com [03:21] TheMuso, how have you been sponsoring my uploads? [03:22] oh, got it [03:22] no space after -k [03:22] bad app [03:23] lifeless, thanks [03:26] heh yeah I sometimes get caught with debsign on that one. [03:27] robert_ancell: I tend to use -uc -us and debsign afterwards. [07:03] Good morning [07:17] Hey pitti. [07:18] * TheMuso -> slightly early EOD [07:23] hey TheMuso, sleep well! [08:11] mvo: hi [08:11] hey kiwinote [08:11] mvo: I've made a few changes to the appdetails class in my branch, mostly just a few renaming things while we still can [08:12] kiwinote: ok [08:12] kiwinote: I have a look [08:12] mvo: you may want to revert some of the changes, but I've put a little comment explaining the change were applicable [08:13] mvo: I haven't updated anything outside the appdetails.py and test_database to use the new naming [08:18] kiwinote: ok [08:20] mvo: any changes you don't want, or do I change the appdetailsview to use the new naming? [08:21] mvo: the only last thing I was wondering about was the summary and description. they both kind of sound the same, but I couldn't think of better names to distinguish between them.. [08:22] kiwinote: thanks, I just looked over it and I think the changes are fine and make sense [08:22] ok [08:23] kiwinote: I think summary/description needs good doc strings. maybe summary/long_description? but that then becomes a bit too long [08:24] mvo: yeah, probably best just to leave it as it is until someone comes up with really good names for them [08:24] * mvo nods [08:26] kiwinote: merged [08:27] hello [08:28] bonjour seb128 [08:28] hey pitti [08:28] how are you? [08:34] brb after upgrade restart [08:37] desrt, ping [08:38] hey robert_ancell [08:39] robert_ancell: congratulations for your new core-dev badge, well earned! [08:39] pitti, hey [08:39] thanks [08:40] mvo: my branch should have the remaining changes needed for appdetailsview-wk to use the new naming [08:41] mvo: I'll update my own branch and the appdetailsview-gtk to use the new naming, so I won't be touching your code for the next while ;) [08:41] kiwinote: ok [08:42] kiwinote: I work on the remaining bits and will merge into trunk today [08:43] mvo: yep, that code is quite sweet, so it'll be nice once it's in trunk for us all to use. thanks [09:05] slomo, hey [09:06] slomo, do you know if anybody is working on updating gtk 2.22 with the new gdk-pixbuf source in debian? [09:08] slomo, do you also know what are the plan for gobject-introspection? will debian update or stay on that abi version? [09:12] seb128: no idea, sorry [09:12] slomo, for both questions? [09:12] yes [09:13] ok [09:13] slomo, you don't plan to work on the gtk update yourself? ;-) [09:14] i do but i have too many other things to do right now :( [09:14] might take a week or two [09:15] slomo, ok [09:15] slomo, I might have a go to the update and packaging gdk-pixbuf today [09:15] slomo, I will ping you for review and debian sponsoring if I do ;-) [09:16] ok, thanks :) [09:16] robert_ancell, you didn't do any work on updating gtk today right? [09:16] hey rickspencer3 [09:17] hi seb128 [09:18] hey rickspencer3, had a nice party last night? [09:18] I'm sure the Dutch folks are happy :) [09:18] pitti, yes, they were quite happy [09:18] :) [09:19] let's hope we'll meet them on Sunday! [09:19] though I think several heart attacks occurred in the last 60 seconds [09:19] hehe [09:19] pitti, let's wait to see what it's going to be on sunday ;-) [09:19] rickspencer3: heart attacks> I didn't follow the game, they shot a goal in the last minue? [09:20] pitti, yeah, in extra time there was quite some action in front of the NL goal [09:20] rickspencer3: i. e. they were damn close to a 3:3? [09:20] pitti, correct [09:20] hehe [09:20] one good shot would have equalized [09:21] pitti, the uruguayans were going at them crazy, and 3 mins of extra time stretched into 4 and a half :) [09:21] I'm supposed to be on a train to Brussels atm [09:21] but someone used the tracks to do themselves in [09:21] :( [09:21] so my train was canceled [09:21] urgh [09:22] rickspencer3: but that usually takes just an hour or two to settle; got some urgent appointments? [09:22] pitti, yeah, I needed to transfer to EuroStar [09:22] so would have missed my train there, etc... [09:22] so I hit the reset button, will try again later [09:23] or maybe next week, even === almaisan-away is now known as al-maisan [09:56] kiwinote: let me know when you feel I should merge the apdetails-gtk branch, I think the other bits are ready [09:58] kiwinote: I split the webkit stuff into appdetailsview_webkit now, so it should be easy [09:58] mvo: I've just finished updating the appdetailsview-gtk file, I've still got to update some of the code in the application.py (eg the pkg_state stuff) before appdetailsview-gtk will work nicely [09:58] kiwinote: ok, thanks. no worries, just ping me :) === DrPepperKid is now known as MacSlow [10:38] mvo, could I ask why you are not using webkit in software-center anymore? (sorry for asking twice, my pc crashed) [10:40] and471: hello, there were some integration and a11y issues, but mostly because mmcg wrote a super-beautiful gtk version of the category view widget [10:41] and471: I am still fond of webkit and think its a good way to make beautiful UIs without too much work [10:41] mvo, the reason I ask is that I am creating this http://whyareyoureadingthisurl.wordpress.com/2010/06/14/introducing-the-wasiliana-mail-client/ [10:41] my first idea was to use webkit to create the UI [10:41] and471: but matthew decided to go the extra mile (that is rquired with gtk) [10:41] but now wx-toolkit has come to my attention (toolkit in clutter, used by meego) [10:42] mvo, I was wondering which would be better [10:42] and471: nice! I think you will still need webkit to display the html mails [10:42] mvo, yeah, [10:42] mvo, it was whther to just use for the email, or for the whole app [10:42] and471: I think it depends a bit, I personally think it much easier to find html hackers than people implementing gtk widgets [10:42] and471: but gtk feels snappier and is better integrated (at least currently) [10:43] mvo, hmmm, there is my dilemna :) [10:43] and471: I would design a abstract class and implement the html version first [10:43] mvo, did you encounter any performance issues with webkit? [10:43] and471: and then you can experiment with gtk/webkit/clutter [10:43] mvo, and allow for a mx-toolkit version :) [10:44] mvo, the project I am working on is designed for netbooks, and if there any performance issues, then there is no point my using webkit [10:44] and471: no real performance issues, but in order to be sure that the page is rendered before executing JS it feels on some place not as snappy as with native gtk [10:45] and471: I think this problems can be solved by more careful work on the html/JS, not a big deal IMO [10:45] mvo, the one thing good thing about mx-toolkit in relation to webkit, is that it can be styled with CSS also :) [10:45] and471: cool! [10:46] mvo, mx-toolkit is based on clutter, does clutter require compositing? [10:46] mvo, :) [10:46] and471: I quite like webkit still, without matthew I still would use it, but he does such beautiful work on the native widgets that its just too good :) [10:46] mvo, I branched a recent version and it did look good :) [10:46] and471: yes, clutter requires compositing. we were thinking about using it as well, but that makes other stuff more difficult [10:47] mvo, so by using mx-toolkit, would that be limiting my audience, and how does clutter run on netbooks (i.e. well or not) [10:47] mvo, sorry for these questions, you are just the most experienced person that I know well [10:48] and471: netbooks should be fine, its really stuff like nvidia without nvidia-glx that is problematic [10:48] and471: compiz runs fine on netbooks, once HW acceleration is availalbe you should be fine [10:49] mvo, well I shall continue work on my other project and speak to the lead dev later (hopefully) to see if he has any more advice [10:49] mvo, thanks for your help [10:49] and471: have fun! and of course we always have open arms in software-center for you as well :) [10:50] mvo, :) maybe a bit later :) [10:54] sure .) [10:58] mvo, see ya [11:05] what is the best way to have a patch in a deb, that only applies to the desktop edition? (in this case, a change to the default /etc/sudoers files if installling desktop edition) [11:06] how come glib2.0 doesn't have a build record for sparc? [11:07] seb128, pitti ^ could you have a look? [11:07] and471, to what? glib? [11:08] and471, no my comment [11:08] Laney, I guess it's a lamont question [11:08] seb128, what is the best way to have a patch in a deb, that only applies to the desktop edition? (in this case, a change to the default /etc/sudoers files if installling desktop edition) [11:08] seb128: ok [11:08] Laney, I think it's taking the buildds down for some reason [11:09] so lamont set it in an ignore list or something until that sorted [11:09] and471, don't? [11:09] seb128, :) [11:09] seb128, in this case it kinda needs to be [11:09] and471, there is no way to apply something only to the desktop edition since the same deb is used on all editions [11:10] and471, what are you trying to do? [11:10] seb128, it is about showing visual feedback with sudo - but only on the desktop [11:10] (i.e. asterisks) [11:10] and471, change whatever is building your images to tweak it [11:10] why do people care so much about sudo on the desktop [11:11] seb128, there are about 100 comments on the bug report, vish asked me to see how it could be implemented [11:11] and471, that change is not likely to ever go in ubuntu [11:11] seb128, hey don't shoot the messenger [11:11] it's arguably a security issue [11:11] well that has been discussed several times [11:11] use the new "Opinion" status for the bug :) [11:12] Laney, :) [11:12] sudo is a command line poweruser thing [11:12] we will not set different behaviour depending of what you are using it [11:12] no normal desktop user should see a sudo prompt anyway [11:12] seb128, I think the overall opinion of the bug report was that the server team don't want it, but the deisgn team do, and the server team said the deisgn team can [11:12] and471: you can't apply that to sudo, you'd have to modify suders in ubiquity [11:13] I'm not sure whether ubiquity already modifies sudoers in some way [11:13] pitti: i'll take blame :D , wanted and471 to get a working patch ;) [11:13] pitti, so would that be the 'least hacky' of the hacky ways? [11:13] but it'd still be inconsistent [11:13] why do design team care so much about command lines? [11:13] shouldn't we improve the desktop rather [11:13] and471: honestly? fix our documentation to use gksu or pkexec instead of sudo [11:14] what pitti said [11:14] pitti: which is easier ? fix every documentation or.. ? [11:14] and471: the next best thing, change ubiquity to set the pwfeedback option in sudoers [11:14] why should any normal user run sudo [11:14] what's the bug #? [11:14] pitti, ok [11:14] rickspencer3: bug #194472 [11:14] Launchpad bug 194472 in sudo (Ubuntu) (and 3 other projects) "Entering password in Terminal gives no visual feedback (affects: 7) (dups: 2) (heat: 26)" [Wishlist,Won't fix] https://launchpad.net/bugs/194472 [11:14] I tried it here locally, and it works rwell [11:14] uh [11:14] * and471 can feel a heated debate coming on.... [11:14] I think passwords have worked that way in unix for ... [11:14] uh .., [11:15] ever? [11:15] the debate has already happened on several iteration [11:15] rickspencer3, yes, displaying visual feedback on the number of chars is sort of a security concern for some users [11:15] some command line users rather let's say ;-) [11:15] rickspencer3: yes , but we can improve the situation for new users who get confused , for example http://ubuntuforums.org/showthread.php?t=214393 [11:15] I would be in the opinion of leaving the command line as it is [11:16] uh [11:16] seb128, (it deso remove the asterisks after you have pressed enter, but yes, some users still feel it is a risk) [11:16] the real question is why forum recommends using sudo [11:16] *does [11:16] rickspencer3: or > https://answers.launchpad.net/ubuntu/+source/sound-juicer/+question/2046 , they dont really understand why it is happening [11:16] rather than gksudo or equivalent [11:16] wow, I can't imagine we would touch tried and true command line behavior [11:16] seb128, IIRC there are some commands which gksudo doesn't run with [11:17] rickspencer3: its a small preference , which can be reverted for a user who does not like it [11:17] it would be a bug and we should fix it then [11:17] vish, preferences = unpredictability and code complexity [11:17] rickspencer3: it is already there [11:17] don't we have real desktop issues to work on? [11:17] * rickspencer3 stays out of it [11:17] rather than trying to change the way the command line is working since the stone age or so? [11:18] rickspencer3: its just an extra word in thesudoers file ;) [11:18] * seb128 as well [11:18] we shouldn't lower our standards just because some dumb users write buggy recommendations on forums for other users [11:19] My biggest concern is that this wouldn't help at all [11:19] since on most boxes you'd still see no password, just on new maverick desktop installs [11:19] we can't magically and retroactively fix the world [11:20] but we can prevent new users from not being confused :) [11:20] vish, why are "new" users using sudo on the cli? [11:20] and if we do that, I'm already seeing the next guy coming along and wanting the same for ssh [11:20] and if this confuses them, then it's the least of our worries [11:20] right, let's address this [11:20] that's just how the CLI works *shrug* [11:20] the trouble with the bug is that no one wants to get involved, they keep handing it to the next person [11:21] yeah :s [11:21] the trouble is why would you want to change how the command line work [11:21] (and it gets way more attention than it deserves, really, FWIW) [11:21] just don't direct normal desktop users to it [11:21] what needs to be done if we want to keep the current "no asterix" behavior? [11:21] rickspencer3: nothing [11:21] rickspencer3, nothing [11:21] nothing is a good amount of work to take on [11:21] ;-) [11:21] ;p [11:22] rickspencer3, if we want asterisks, we need a one line debdiff that changes one word [11:22] rickspencer3, hehe [11:22] but we don't want asterixes [11:22] and471: not quite [11:22] at least I don't [11:22] and471, it would change sudo [11:22] and471: you need some code in ubiquity to add it to an existing sudoers [11:22] rickspencer3: it can be reverted if someone doesnt :) [11:22] and471, next you get that with ssh and other prompts as well [11:22] pitti, yeah sorry, for just desktop it might be a few more lines :p [11:22] and471: and then update tons of documentation to account for both cases (asterisks and none) and explain the difference [11:22] command lines have certain tried and true behaviors and I don't see a good reason for Ubuntu to deviate [11:22] vish, you adress the issue with one command only though [11:23] su still works without * [11:23] ssh still works... [11:23] etc [11:23] rickspencer3: upstream fixed the bug mainly becaused be requested and now we are not using it :( [11:23] the bug comments say that usability = asterix, and security = balnk [11:23] seb128: i agree :) [11:23] rickspencer3: I think we can ignore the security part here; my concerns are about consistency and existing documentation [11:23] I don;t agree [11:23] s/be/we [11:23] we didn't reply to why we bring those users to a command line to start [11:24] Don Norman says that if all else fails, standardize [11:24] I think deviating from standard behavior here will do more harm than good [11:24] ok [11:24] that's my $.02 [11:24] * rickspencer3 backs out again [11:24] same 0.2 [11:24] seb128: often hardware dont work and there needs tweaking.. we resort to these commands.. [11:24] I don't support the change and I will not contribute to any move to that direction [11:24] * pitti goes back to fixing upower bugs [11:24] hmm.. [11:24] ouch [11:25] * seb128 goes back to backport nautilus hundredpapercut bugfixes [11:25] vish: gksu/pkexec FTW [11:25] vish ... [11:25] things which will really improve our desktop [11:25] I'm sorry if we're coming off as not being appreciative of your efforts [11:25] seb128, pitti hey come on guys, don't get down on vish, he is only trying to help [11:25] we definately love you and your contributions are invaluable [11:26] and471, it's not against vish [11:26] he's doing great work [11:26] and471, trying to make that point now [11:26] and I'm working right now on backporting nautilus changes for bugs he got fixed upstream [11:26] in fact, i don't think anyone is blocking the * thing [11:26] but you need to pick your battles [11:26] changing the command line is an hard battle for little win since that's not something users should have to use [11:26] seb128, I realise that, it just when you say I am off to do more important work, it sounds as if you are devaluing his work [11:26] so, for the record, vish == totally awesome [11:26] cool [11:27] and471, as said you need to pick battles [11:27] I think we all know what seb128 meant [11:27] ;) [11:27] seb128, not a battle, just a comment [11:27] and471, I've expressed my opinion on the sudo change [11:27] right lets all cool off... :D [11:27] I don't know anybody is heated there [11:27] sorry if that came wrongly [11:28] and471, vish: I'm not "getting down" on vish, sorry if I came across like that; I just don't like that particular idea, but I still do like vish [11:28] but I think we all expressed our opinions and have to agree to disagree [11:28] * pitti hugs everyone [11:28] * vish didnt take it personally too :) [11:28] vish is doing great work and lot of hundredpapercut bugs will be fixed thanks to him [11:28] (and shiny icons!) [11:28] I just don't want to be involved in that sudo change [11:28] * and471 accepts the hugs :) [11:28] ;-) [11:29] okay, in 10 secs we have gone from heated to loveydovey :D [11:29] hehe [11:29] just funny seeing it is tough to make hackers think different ;p [11:29] I don't think anybody was heated, sorry if that came wrongly [11:29] I just things we have strong opinions and we are not ready to be convinced it's where we want to go [11:29] seb128, np, as I said it is a limitation of IRC, when speaking in real life, facial expressions etc. affect tone [11:30] and471, well, it's good to call people on it if you think someone is out of line ... so you did the right thing [11:31] we need to keep #u-desktop a good place to be [11:31] keep everything hunkydory... [11:32] rickspencer3: pitti: seb128: i do understand that the sudo commands being given out are wrong , but my final take on this is that: this was something we[i] could help solve within our powers than trying to convince every person on the planet to not mention sudo :) [11:33] vish, tbh, I think using the cli to help people is *good* [11:33] in some limited circumstances [11:34] vish, and471: but changing sudo doesn't fix the issue [11:34] it would just make the sudo on modern ubuntu different from upgrades [11:34] or from su or vt login or ssh or zillion other commands [11:34] or other distributions [11:35] it's a consistant unix command line behaviour accross the board currently [11:35] just changing one command just creates inconsistency [11:36] seb128: how many users work on more than one distro.. ? and this is something which can be reverted if not preferred.. :) [11:36] anyway... [11:36] it doesn't adress vt login [11:36] or su prompts or ssh prompts or other commands acting this way [11:36] still on ubuntu [11:36] seb128: we can fix them all , one by one ? ;) [11:36] yes we can [11:37] lets start with sudo .. [11:37] but with the same energy we can fix an hundred desktop hundredpapercut [11:37] so should we fix the command line and no desktop papercut this cycle? [11:37] we have limited manpower and you can't get everything [11:37] seb128: you keep mentioning energy , but energy is being used here to not fix it , than to fix. [11:38] not really, we just discussed it [11:38] I disagree that changing just sudo is fixing it, it's just creating inconsistency [11:38] if you want to change it you need to do it with all command line prompts [11:38] but anyway everybody expressed opinions [11:39] I'm rather interested to working on desktop changes not command line [11:39] but if somebody wants to step for this nice [11:39] so I will stay out of the way from now on ;-) [11:39] seb128: this is where i get confused : " but if somebody wants to step for this nice" ? what does that mean? [11:40] if someone gets the fix.. or.. [11:40] if somebody having the position to do the change want to do it and defend it in discussion etc and deal with bug reports and inconsistency etc [11:40] having the position = upload rights and wanting to assume the change [11:41] I'm not going to be the guy who changed sudo and who will get the heated discussions with all the community members who disagree with it [11:41] * vish can take blame , but would need upload rights ;) [11:41] in any case it's time to get something to eat [11:41] bbl [11:41] it could always be reverted, right/ [11:41] ? [11:42] rickspencer3: yes [11:42] it's still only Alpha 3 [11:42] rickspencer3, not sure we can undo it on installed systems [11:42] it's the ssh system config [11:42] we can tweak the default value at installation [11:42] it will also not change for upgraders [11:43] bbl [11:43] seb128, go eat! :) [11:47] see ya guys, thanks for the discussion :) [12:17] * ogra wonders ehy empathy breaks hios image builds, there doesnt seem to be any upload that could have caused that [12:20] gar, it waited for glib it seems :/ === al-maisan is now known as almaisan-away === MacSlow is now known as MacSlow|lunch === cking is now known as cking-afk === almaisan-away is now known as al-maisan [14:18] seb128: what are your plans for gnome 3.0 / gtk3 for ubuntu btw? and if you package gdkpixbuf/gtk 2.21.something, just write me a mail once you've put things into svn and i'll review/build/upload it for you :) [14:21] slomo, not so much plans so far for this cycle [14:21] slomo, basically we would like to settle the platform, ie uptodate gtk 2.22 [14:21] getting gtk3 in the archive and maybe doing gtk3 builds of required GNOME libraries [14:22] slomo, we didn't have any pressure to do those so far though and I don't want to duplicate work [14:22] slomo, so I would be happy to work on those directly in debian === MacSlow|lunch is now known as MacSlow [14:23] slomo, I would appreciate reviews as well for those changes ;-) [14:23] seb128: sounds like a good plan imho, i hope we get the same done for squeeze [14:24] brb [14:29] seb128: do you have a package for libpeas already? [14:30] no [14:30] ok [14:30] slomo, who is taking care of gobject-introspection in debian? [14:31] I would like to figure what we should be doing [14:31] ie updating to 0.9 or not [14:31] seb128: bigon and joss [14:31] it has an abi number update [14:31] seb128: and imho we should update asap if it's not going to change soon again [14:32] who has an opinion on that in debian to discuss it? [14:33] bigon and joss :) [14:34] slomo, I guess I should join on oftc to discuss those changes [14:35] yes, i stopped caring about g-i- some time ago because there are too many things that i don't like (missing support for nested namespaces being the worst...) [14:38] slomo, nice, d-conf got newed in Debian now [14:40] yes, now i can finally test the gstreamer gsettings elements with a useful backend ;) [14:40] mvo: I've pushed the changes to the appdetailsview-gtk branch [14:41] mvo: I am just left with the enable component code not working completely , as I get http://pastebin.ubuntu.com/460217/ [14:41] mvo: It seems to be some dbus or aptdaemon stuff, do you mind looking at that? === cking-afk is now known as cking [15:05] seb128: ok the only issue I see with the menu is the bullets showing up as dots, but that's triaged and assigned for A3. [15:05] jcastro, excellent ;-) [15:07] htorque was seeing problems yesterday with the tools menu in gedit but it looks fine to me [15:11] jcastro, I can confirm similar issues [15:11] ok so enough not to turn off the double menus yet, good, that's what I wanted. heh. [15:12] jcastro, you recommend keeping the double menus for now? [15:12] seb128: yeah, at least until we get the bullets [15:13] we haven't had a flood of doublemenu reports, so we know people are reading the wiki page and following the directions [15:13] oh [15:13] "bullets showing up as dots" [15:13] showing as checks marks you mean rather [15:13] no, as check marks [15:13] yeah, like in the document list in gedit [15:13] right [15:13] you wrote as dots before [15:13] I was getting confused on what the issue was ;-) [15:14] oh, I see I did, whoops [15:15] mvo_: when you are around, there are some sentences for you about 20 lines up [15:17] kiwinote: it appears I disconnected [15:17] kiwinote: could you please give them to me again? [15:18] mvo: I've pushed the changes to the appdetailsview-gtk branch [15:18] mvo: I am just left with the enable component code not working completely , as I get http://pastebin.ubuntu.com/460217/ [15:18] mvo: It seems to be some dbus or aptdaemon stuff, do you mind looking at that? [15:19] kiwinote: sure, I'm happy to have a look. thanks for your work on this! [15:23] seb128, also the application menu is still broken if you turn off double menus [15:23] for desktop users, not unity of course [15:24] kenvandine, how broken? [15:25] if you turn off double menus, you get no menu items [15:25] * kenvandine finds the bug [15:25] wfm [15:25] I just did unset the variable and run gedit [15:25] it's correctly exported and menus are working on GNOME [15:26] bug 596871 [15:26] Launchpad bug 596871 in indicator-appmenu (and 1 other project) "APPMENU_DISPLAY_BOTH=0 hides items in the panel menu applet (affects: 1) (heat: 11)" [Undecided,Invalid] https://launchpad.net/bugs/596871 [15:26] humm [15:26] i tried it last night [15:26] oh... unset the variable? [15:26] that's what I tried [15:27] i get the other menus fine [15:27] APPMENU_DISPLAY_BOTH=0 works fine as well [15:27] but the applications menu is empty [15:27] * kenvandine tries again... maybe it was fixed in yesterdays upload and the bug not closed [15:28] kenvandine, oh ok [15:28] i did try it early yesterday [15:28] and it was still broken [15:28] I didn't try the gnome-panel menus [15:28] ah [15:28] I don't want to restart my session ;-) [15:28] hehe [15:28] I though you said the appmenu applet was empty [15:29] ah, no [15:29] i'll make sure it is still broken, but i suspect it is :) [15:29] brb [15:32] yup, still broken [15:43] didrocks: are you back today? [15:44] jcastro, he's out until next monday, I think [15:44] he's at yet another French Ubuntu event ;) [15:44] probably speaking to 10,000 people right now === thekorn_ is now known as thekorn [16:12] jcastro, do you need anything? I can I help you? [16:13] seb128: nope I was just filing some bugs on the banshee-meego bits he landed, nothing I can't catch up with him later on [16:13] jcastro, ok [16:15] tedg, I got subclassing working, but I have one last problem [16:16] tedg, I can't add a custom signal to indicator, why is this? [16:18] and471, Cool, what was the issue? You should be able to add a signal... you can definitely do that in C. [16:18] and471, I'm not sure if you're running into a Vala-ism htere. [16:18] there [16:19] tedg, in vala, the issue was that I needed to set the indicator subclass as a property of my main class, it was getting lost before the gtk.main() [16:19] tedg, this latest problem is in python :) [16:20] tedg, I can add a signal using a gtk.StatusIcon, but not for an indicator [16:20] and471, Ah, makes sense. [16:20] and471, Are you trying to connect to a signal or add a new one? [16:20] and471, In your subclass. [16:21] tedg, http://pastebin.com/Zq6GhF6F [16:21] tedg, add a new one [16:24] and471, Sorry, I'm not sure. I'm guessing it must be something with how it's bound into Python. [16:24] kenvandine, ^^ [16:24] * kenvandine looks [16:25] tedg, sorry, just when it was working in one language, I have switched to another :) [16:25] tedg, it is for a different project though :) [16:25] * and471 thanks kenvandine [16:27] and471, so what isn't working in there? [16:27] kenvandine, so it appears to work (no error), but then when I try to connect the signal to a callback [16:27] self.indicator.connect("recording-done", self.on_indicator_recording_done) [16:28] http://pastebin.com/wU76HC6u [16:28] kenvandine, ^, but if change to using a gtk.StatusIcon (the fallback) it works [16:32] and471, maybe because your class is a subclass of appindicator.Indicator instead of GObject? [16:32] tedg, ^^ [16:33] kenvandine, That, in theory, shouldn't be an issue as appindicator.Indicator is a subclass of GObject. But I don't know how the Python bindings handle that. [16:34] ok [16:34] So it should go GObject -> AppIndicator -> His Class instead of GObject -> StatusItem -> His Class [16:35] I'd be curious if the Python bindings aren't properly communicating that to PyGTK or something? [16:35] kenvandine, Could you do a check with PyGI -- it might work better? :) [16:36] and471, can you try inheriting from gobject.GObject as well? [16:36] class KazamIndicator(appindicator.Indicator, gobject.GObject): [16:37] and471, or give me the rest of your code so i can try it myself :) [16:38] kenvandine, it didn't work [16:38] ok [16:38] kenvandine, I shall give you my code [16:38] thx [16:38] kenvandine, be warned though, it is very rough :-0 [16:38] sure :) [16:40] kenvandine, okay, it should be available here https://code.launchpad.net/~and471/+junk/kazam-temp [16:41] kenvandine, the indicator stuff is in kazam/indicator.py, you can manually use gtk.StatusIcon instead of the indicator subclass by changing line 35 to be some gobbledegook [16:41] (highly technical manual override) [16:41] :) [16:41] kenvandine, just run bin/kazam, click record, it should countdown and the indicator should appear, that is when you get the error === MacSlow is now known as MacSlow|afk [16:47] * seb128 just upgraded empathy [16:48] desrt, it's working using dconf! [16:50] kenvandine, anything? [16:54] slomo, did you have time to work on the vala 0.9.2 ftbfs issue? [16:55] slomo, there is a -2 let me sync that! [16:56] slomo, thanks :-) [16:57] slomo, hum, it seems that one still fails to build [17:03] how do I downgraded a git checkout to a specific commit? [17:14] seb128: git checkout abcdef1234567890 [17:15] slomo, thanks, I was trying to give an argument to git clone [17:23] and471, it's looks like that should work fine, let me eat something and try to create a simpler test case to test the bindings [17:25] kenvandine, cool [17:39] well, i'm going to try and upgrade to maverick in a bit :) [17:39] finally! [17:40] so, if i disappear for a long time, then it didn't go too well ;) [17:41] seb128: awesome :) [17:42] seb128: so i'm on the CD now, or...? [17:44] desrt, yes you are! [17:44] sweet :) [17:45] and a bit scary.... [17:45] desrt, do you plan to do a dconf release btw? [17:45] probably after the next glib release [17:45] the current dconf-editor sucks [17:45] more broken APIs :p [17:45] yes. robert ancell wrote it. :p [17:45] well the git one doesn't [17:45] so give him some credit ;-) [17:45] oh [17:46] that's why you want a release :) [17:48] desrt, right [17:49] okay [17:49] i'll be dropping a glib release soon [17:49] by the end of the week [17:49] i'll do a dconf to go with [17:50] we've been doing glib releases about once a week and every time there is a list of changes too big to fit onto one screen of a NEWS file. it's kinda fun :) [17:51] desrt, it makes me glad we don't have a lot of softwares using ie new apis yet [17:52] we try to avoid breaking widely used APIs [17:52] even in the unstable release [17:52] nonetheless there are a few breaks this time around, still :/ [17:52] probably nothing you'll notice, though [17:53] btw... did the Bsymbolic stuff cause you any trouble? [17:54] * desrt is planning to rip that stuff out for good before the next release [17:54] desrt, what Bsymbolic? [17:54] good :) [17:55] if you didn't notice then it means it's working properly, i guess :p [17:55] we're dropping that IA__g_* stuff from glib and gtk [17:55] in favour of using the mostly-equivalent -Bsymbolic-functions linker flag [17:55] i did a release with a very minimal removal of the old stuff and addition of the new stuff [17:55] in the next day or so i going to do a more permanent removal of the old stuff [17:56] desrt, ubuntu uses BSymboic by default [17:56] just wanted to give people a chance to complain before we went ahead and did it for good [17:56] seb128: fascinating. [17:56] -Bsymbolic or -Bsymbolic-functions? [17:56] it's creating some issues every now and then [17:56] because -Bsymbolic will seriously break stuff [17:56] the second one [17:56] we do that for some cycles [17:56] cool [17:57] out of some stuff declaring symbols in two places [17:57] well, it's actually not even the fault of the software [17:57] usually in the binary and in a .so [17:57] it works without issues [17:57] turns out the linker itself is really damn stupid [17:57] say you have some library that has like: [17:57] lib.c: [17:57] int somevar; [17:57] lib.h: [17:57] extern int somevar; [17:57] then you have an app [17:57] app.c: [17:57] a() { somevar = 5; } [17:58] (after including "lib.h", of course) [17:58] this doesn't work like you might think it works [17:58] when linking, the linker adds a 'copy' relocation [17:58] it copies the contents of the original 'somevar' variable out of the library's data segment and into the application's data segment [17:58] so now 'somevar' is a variable declared by the application [17:59] it then sets up the application's copy of 'somevar' to interpose over the one in the library [17:59] so if the library is using -Bsymbolic it will use its own local copy [17:59] and never see that it got set to 5 [17:59] totally stupid [17:59] interesting [17:59] this is exactly the mechanism by which the glib thread vtable gets filled in [18:00] so if we linked glib with -Bsymbolic, threads don't work [18:00] i'm really unsure why it does that [18:00] the only benefit is that the application gets to use its own copy directly without a relocation [18:00] and no PLT lookup [18:02] desrt, https://bugzilla.gnome.org/show_bug.cgi?id=609505 is the sort of issues we get with -Bsymbolic-functions [18:02] Gnome bug 609505 in general "Segfault on run with cheese 2.29.90" [Critical,Needinfo] [18:02] ah ya [18:02] of course [18:02] that's sort of like an application bug, though [18:02] it is [18:03] strictly speaking, it's your fault [18:03] but in a practical sense..... [18:03] I already once with hadess though we argued it's our fault for using this flag for our builds ;-) [18:03] desrt, why? [18:04] because technically cheese is allowed to do what it's doing [18:04] and, in fact, it works [18:04] it's just buggy to have the same symbols defined in the binary and the so [18:04] technically not true :) [18:04] ELF was designed to deal with this situation in a specific way [18:05] so it's no bug, in theory [18:05] but in reality it's probably not what they meant to do... [18:05] in reality, they just messed up their build system [18:20] pitti, is there a quick way of figuring out which keymap is being used by udev on my machine? [18:21] cking_: quickest is probably "sudo udevadm test /class/input/eventXX [18:21] cking_: with XX being the event number of your keyboard [18:21] cool. [18:30] good night everyone; time for dinner and then watching THE GAME [18:34] ivanka, link to fonttest.design.canonical.com on the blog post is broken [18:35] ivanka, in fact all links except the last one [18:35] and471: thanks! [18:35] ivanka, no problem [18:36] and471, ok... figured it out [18:36] you can't subclass appindicator.Indicator [18:36] kenvandine, yay! [18:36] kenvandine, oh... [18:36] it takes a parent class of gobject [18:37] http://pastebin.ubuntu.com/460323/ [18:37] that is an example [18:38] kenvandine, so I have to have an indicator property within a generic gobject? [18:38] yes [18:38] kenvandine, thanks for your help, why is it that it doesn't work? [18:38] also, in retrospect, i can't see any reason to subclass it :) [18:38] at least in your case [18:39] kenvandine, true... [18:39] the first arg in the constructor is the parent gobject class [18:39] ivanka, as a prize, can I also get an email (as i am not an ubuntu member?) :D === MacSlow|afk is now known as MacSlow [18:40] ivanka, (an email with an invite for the ppa that is) [18:44] and471: thank you! I think they are all good now. [18:45] ivanka, as a prize, can I also get an email with a ppa invite (as i am not an ubuntu member?) :D [18:45] and471: You have my eternal thanks - I'll have a little think about an appropriate prize ;-) [18:45] ivanka, hehe [18:45] ivanka, any idea when the font will be available to people besides ubuntu members? [18:48] august 8th [18:48] jcastro, :( [18:53] ivanka, jcastro : is this a secret plan to get more people applying to be ubuntu members... :) [18:53] and471: shhh - don't give it away! [18:55] :) [18:56] cassidy, hello, do you have a bug about gsettings migration issue? [18:56] cassidy, opening the preferences dialog displayed a "do you want to import the account listed there" dialog [19:09] heh, i think my daughter has broken the mobile in her cot [19:10] chrisccoulson: You got her an iPhone? [19:10] jpds - lol ;) [19:10] not just yet [19:11] it would have to be an android device anyway ;) [19:11] she grabs her mobile as it's spinning round whilst she sits in her cot, and all it does is click now [19:12] i think she must have stripped the teeth off the gears ;) [19:21] could make a mobile out of old discarded iphone3's, a mobile mobile if you will === al-maisan is now known as almaisan-away [19:27] lifeless: hey? [19:32] desrt: hi [19:33] lifeless: i seem to recall that you were using luks? [19:33] not atm but I did for a few years [19:33] ya. i remember this circa cambridge UDS [19:33] any particular reason you stopped? [19:33] it has overhead [19:33] I switched to an SSD based laptop a few months back - march [19:34] I decided to drop it at the same time, for a few reasons [19:34] and with the performance benefit of SSD the extra latency of the encryption became noticable? [19:34] I was worried the wear levelling and the block-not-used commands wouldn't filter down the layers [19:35] ah. fascinating. [19:35] zero evidence for or against [19:35] the drives are supposed to their own wear-leveling, aren't they? [19:35] apparently the quality of that built in leveling is a major point of differentiation [19:35] i went intel [19:35] but there is this ATAPI command to hint [19:36] mine is fujitsu I think, its what lenovo put in the laptop [19:36] samsung, maybe? [19:36] if I'd had a few more cycles I would have researched the issue [19:36] i called them up to ask them what they'd put in my laptop if i got the SSD option and they said samsung === fta_ is now known as fta [19:36] i looked up reviews and samsung always came out on the bottom with intel on top [19:36] so i just ordered with the smallest disk possible and bought the intel drive separate [19:36] [ 2.066790] scsi 0:0:0:0: Direct-Access ATA SAMSUNG MMCRE28G VBM1 PQ: 0 ANSI: 5 [19:37] yup [19:37] yeah [19:37] I had a bunch of time pressure [19:37] so I punted [19:37] me too [19:37] but it turns out i got fucked anyway [19:37] i ordered during the mobile core i7 shortage [19:37] and took some briefly considered risk reduction [19:37] desrt: same! [19:37] desrt: x201s ? [19:37] so my 1-2week delivery turned into me cancelling the order after a month and a half [19:37] ya [19:37] I got mine [19:38] i'm glad i didn't [19:38] 5 weeks I think it was [19:38] i ended up reordering on july 1 [19:38] the first day of the new sales tax regime in ontario [19:38] saved 8% due to that [19:38] nice [19:38] plus they had a 25% sale, plus 22% for using my visa card [19:38] so i ended up getting it about $500 cheaper than if my order had gone through before [19:39] thats cool [19:39] $1100 total [19:39] how much mem did you get [19:39] 4 [19:39] I got 8, really liking it [19:40] plus the bluetooth and top-of-the-line wifi [19:40] heh [19:40] 8 was too expensive for my blood :p [19:40] the wifilink 6000 series ? [19:40] lemme check the invoice [19:40] Intel Centrino Ultimate-N 6300 AGN [19:41] 02:00.0 Network controller: Intel Corporation WiFi Link 6000 Series (rev 35) [19:41] i think 'wifilink' is the low model [19:41] wasn't a selectable option in .au [19:41] :( [19:41] ya. it's the default one here [19:41] was like a $40 upgrade [19:42] a reasonable price to pay to be able to mock my friends when they can't get a signal with their powerbooks :p [19:42] :) [19:42] anyway.. i was going to LUKS it up [19:42] but i guess now i won't [19:43] well [19:43] like I say 0 evidence pro or con [19:43] well... your 'gut' reasoning resonates well here [19:43] I just know that new things generally take some time to integrate well vertically in the stack [19:43] yeha [19:44] it's not like i'm smuggling state secrets across international borders or anything [19:44] (everyone knows that i use the internet for that!) [19:44] :) [19:45] k. thanks for the infos [19:45] de nada [19:45] * desrt pretends to do real work now === fta_ is now known as fta [21:12] success \o/ [21:33] jcastro, are you there? [21:33] waltercool: hi [21:34] jcastro, hi there :P [21:34] *weep* [21:34] jcastro, can you give me a invitation for ubuntu font ppa? :B [21:36] hey pitti [21:36] hey kenvandine [21:36] pitti, is the game over? [21:36] :'-( [21:37] kenvandine: yes, Spain kicked us out, 1:0 [21:37] guess i know the results [21:37] bummer [21:37] but they were unstoppable today [21:37] they had a perfect technique, and our guys could hardly keep the ball -- they were in pretty bad mood today apparently [21:37] my german brother-in-law was making me watch all the germany games on vacation :) [21:38] not that i resisted much :) [21:39] heh [21:39] well, the previous ones were quite fun to watch indeed [21:39] beer and wild german soccer fan, good times! [21:39] didn't get to see today's though :/ [21:40] kenvandine: then it was you who was missing [21:40] :) [21:41] guess i'll still watch it tonight just to complete the series [21:41] pitti, germany had a good run though! [21:41] absolutely, and still a good chance to become 3rd === fta__ is now known as fta [21:55] pitti - i would tease you, but i can't really say anything after we lost so embarassingly to you ;) [21:55] hdh [22:04] didrocks: are you lead on quickly now? === tkamppeter_ is now known as tkamppeter === oubiwann is now known as oubiwann-away === oubiwann-away is now known as oubiwann === bjf is now known as bjf[afk] === tkamppeter_ is now known as tkamppeter