[00:03] hyperair - not that i'm aware of [00:03] i'd be more concerned about why it's using so much RAM though [00:03] hmm, i wonder [00:03] might be related to the stuff i had in the calendar [00:04] but it seems that after a while the ram usage just naturally rises >_> [00:04] possibly. e-d-s only uses 9MB here, and i have a gmail account with ~60000 mails, and several calendars [00:04] @_@ [00:04] i gave up using evolution, mainly because i just don't agree with its behaviour [00:05] when i open a new window, it spends a crapload of time re-loading every bloody thing in a new process [00:05] yeah, i find quite a few things in evolution annoying [00:05] when loading, it spends a crapload of time talking to a bunch of stuff over the network prior to actually showing me a bloody gui [00:05] and if you've got a slow network.... [00:06] evolution is hell. [00:06] yeah, that's a pain [00:06] there's no other word for it [00:06] =.= [00:06] and interface sucks ;) [00:06] i mean geez, didn't the evolution devels ever realize that hey, network is slow and unreliable and maybe a program shouldn't rely so much on it to just _start_?! [00:06] i don't mind the interface so much [00:06] i do mind memory leaks [00:07] and i do mind hanging [00:07] I can never make myself use Evo for a long time so memory leaks aren't really a problem :) [00:07] or completely not starting up because it's waiting for input from some remote server for hours on end that won't reply because it really timed out [00:07] kklimonda: even if you cloes evo, e-d-s runs. [00:08] kklimonda: e-d-s starts up when you open gnome-panel's clock [00:08] kklimonda: e-d-s also starts up when you start pidgin [00:08] true [00:08] and e-d-s is the one leaking memory like a sieve [00:09] hmm. i think i'll just get rid of evolution-data-server. [00:09] yes [00:10] i remember it used to remove ubuntu-desktop with it, but it seems it doesn't anymore [00:10] good. [00:10] i'll just have to sacrifice ekiga and tasque >_> [00:10] i don't even use the e-d-s component of tasque, so why does it depend on e-d-s? [00:11] it only recommends it [00:11] no, it depends [00:11] otherwise it wouldn't be going out with e-d-s [00:11] no wait [00:11] it depends on libevo-cil [00:12] which depends on e-d-s [00:12] even better - libevo-cil depends on evo which depends on e-d-s [00:12] er no, it depends on evo itself, which depends on e-d-s [00:12] heh [00:12] so what's the point of recommending e-d-s.. [00:12] indeed [00:13] oh well [00:13] i don't use tasque much anyway, but i'll try doing something about libevo-cil sometime === eeejay is now known as eeejay_away === eeejay_away is now known as eeejay === eeejay is now known as eeejay_away [05:24] I'm installing ubuntu-desktop from the lucid repositories, is this a way to get 10.4 running? [05:25] I noticed that ubuntu desktop depends on some erlang stuff. What is the erlang doing? [05:28] emma: You should probably ask in #ubuntu+1 for anything Lucid related [05:28] oh okay sure. [05:32] jmarsden: do you know what the erlang is being used for? That's always seemed like a somewhat esoteric language? === eeejay_away is now known as eeejay [05:33] emma: No; use of Erlang is interesting... you might be able to figure out what uses it by checking the Depends: of each package in Ubuntu Desktop looking for it there? [05:34] Or if you have a working Lucid machine, try apt-cache rdepends erlang inside that machine. [05:34] cool that's what I'll do :) [05:47] erlang is required for desktop couchdb === astechgeek is now known as the_r0cker [08:09] Good morning [08:37] good morning pitti o/ [08:38] bonjour didrocks! did you have a good weekend and first advent? [08:39] pitti: yes, very tiring though: we had our Ubuntu Party in Paris :) [08:39] pitti: you? [08:39] wonderful; we made tons of cookies, and my wife and I made christmas calendars for each other [08:41] you don't have weight issue, so you can eat as much chocolate as you want :) Cooking week-end is great, yeah \o/ [08:52] pitti: hmm, christmas calendars! That's a great idea! [08:52] http://www.piware.de/2009/11/nicht-mehr-aufzuhalten-weihnachten/ :) [08:53] nice! [08:54] heh, I love how the 24th one is bigger [08:55] pitti: sweet :) [08:56] good morning === seb128_ is now known as seb128 [09:43] hello there [09:44] salut seb128 [09:49] hey seb128 [09:49] hi baptistemm [09:49] lut baptistemm didrocks [10:13] mvo, hi, sorry if you've explained this to me before, but what's the difference in Synaptic between "Removal" and "Complete Removal"? [10:15] mpt: in the former case, files in /etc ("conffiles") stay around [10:15] the latter is "purge" and also removes conffiles, log files, etc. [10:16] thanks pitti [10:16] pitti, so the former case is for if you think you might want to reinstall it again later? [10:16] and you spent some time setting it up [10:18] mpt: correct [10:18] ok [10:18] mpt: for some package a purge sometimes removes user added data too [10:18] mvo, as in, deletes stuff from people's home folders? [10:18] mpt: not only config, a database that got created in /var for example [10:19] mpt: no, never in /home [10:19] I see [10:19] mpt: but system-wide data [10:20] purge is always a bit risky and generally there is not that much data that is left on the system [10:24] mvo, is it possible to purge a package after removing it normally? i.e. does apt know that the package was installed previously and has been removed but not purged? [10:25] mpt: yes, you can look into synaptic and check the "Not installed (residual config) section [10:26] oh, that's what that means [10:27] pitti, hey [10:27] hey seb128, had a nice weekend? [10:27] very relaxing thanks [10:27] recovering from some ubuflu still [10:28] what about you? [10:29] ugh, you got ubuflu? [10:29] seb128: we baked tons of christmas cookies, and a gingerbread house [10:30] seb128: oh :( [10:30] seb128: poor you :( [10:30] seb128: the same flu that I got? [10:31] mvo, no, coughing and sneezing [10:31] and filing a bit tired [10:31] but nothing really strong [10:32] ok, good [10:32] I spent the weekend playing with my new gadgets and watching tv [10:32] heh :) [10:32] nice! [10:32] which was good anyway since it's raining outside... [10:33] pitti, cookies! nice ;-) [10:33] we try to not start before december there *g* [10:33] too much food [10:33] sprint, uds, end of year holidays coming [10:33] I need at least a 2 weeks diet ;-) [10:35] pitti, bug #489791 do you know what information is required? [10:35] Launchpad bug 489791 in cups "Compilation fails because of changed interface in libpoppler" [Undecided,Incomplete] https://launchpad.net/bugs/489791 [10:35] it's a pretty obvious "libpoppler abi changed, cups needs rebuild" [10:36] rebuild + code change [10:36] /url 1 [10:37] pitti, note that pdftopdf crashes on undefined symbol too [10:37] ie no printing working in lucid right now [10:39] seb128: I don't see a reason for it to be "incomplete"; should be reproducible just fine [10:39] ok thanks [10:39] I will reopen then [10:39] I set it to triaged and assigned to till [10:40] pitti, thanks [10:46] libpoppler is knackered anyway [10:47] debian bug 558463 [10:47] Debian bug 558463 in poppler "cups: Cups fails to print after upgrade" [Grave,Open] http://bugs.debian.org/558463 [10:51] Laney, what do you mean? [10:51] Laney, they just changed the abi again [10:51] they don't have abi stability for libpoppler [10:51] they do the glib version though [10:52] it's nothing new [10:52] no, I mean it's broken anyway :) [10:52] mpt: *ping* [10:53] unless you mean that debian needs to rebuild cups against this [10:53] Laney, it's not broken, it's the same issue I pointed before [10:53] the abi changed and the lib customers need a rebuild [10:53] hm [10:58] hm, without a soname bump? [10:59] pitti, right, they don't want to give stability to libpoppler [10:59] they only do for the glib and qt versions [11:00] they consider that's what applications should be using [11:00] too much constrains to do it for libpoppler itself apparently... [11:00] I don't say I agree though [11:00] but that's what I've been replied in previous cycles [11:02] ok, thanks [11:02] seb128: hi.. i thought you and pitti had an idea how to solve the .txt behavior from vfat drives... i vaguely remembered you saying you'd work on it if you have time... if not ... nvm :) [11:03] mac_v, you should have noted it, it's possible but I've hundred of ideas every week and forgot about most of those ;-) [11:04] I've neither interest to work on this issue nor time for it [11:04] seb128: hmm , let me find the -desktop logs... ;) [11:04] seb128: ah ok :) [11:04] to be honest I don't think I will have any time for hundredpapercut this cycle [11:05] I'm back to do GNOME without robert_ancell this cycle [11:05] and we have to tackle boot speed issues [11:05] yeah , too bad they left you alone :( [11:05] which are already higher workload that reasonable [11:10] oh, come one [11:10] I start hating this hundredpapercut project, zillion of useless bugs opened [11:24] hehe [11:24] seb128: dejavu from last cycle? ... i remember you constantly complaining back then ;) [11:25] hey asac [11:25] yes, those bug are bottom level quality [11:25] no information about ubuntu version, package version, no clear description [11:26] and most are random funky requests of softwares not making your coffee and not putting the number of sugars you want in there [11:26] "bottom level quality" - nice categorization ;) [11:26] yeah [11:26] i think strictly speaking its dxteam responsibility to monitor and do patches for them ;) [11:27] well ... we can forward upstream of course :) [11:27] excepted that they don't [11:32] typical example: bug #488129 [11:32] Launchpad bug 488129 in nautilus "Nautilus usability" [Undecided,New] https://launchpad.net/bugs/488129 [11:33] this is like 10 reports in one, and certainly not a papercut [11:35] right, I'm closing it now [11:35] got a zillion hundredpapercut bugs spamming my box this weekend [11:36] the average quality of those is just ridiculous [11:43] mvo: the notebook page is used because of the range(5) declarations at the start of the relative python files [11:44] mvo, those numbers are used to change the notebook page and so I need to notebook page as the separator takes up a number [11:44] mvo, otherwise the viewswitcher parts are not connected correctly to their relative page [11:51] good morning everyone [11:51] hey chrisccoulson [11:51] hey pitti, how are you? [11:52] I'm great, thanks; and full of christmas cookies and "Stollen" [11:52] pitti - i love stollen :) [11:52] so do I :) [11:52] we have a german market here throughout december, and i always stock up just before christmas [11:53] Birmingham? Apparently it's getting close to the side of the Köln one [11:53] Laney - yeah, Birmingham [11:54] nice [11:54] i go to birmingham city centre twice per year - once for an eye test and once for the christmas market [11:54] we have one in Nottingham too, but I reckon it's quite small [11:54] hello chrisccoulson [11:54] hey seb128, how are you? [11:54] good thanks [11:54] what about you? [11:55] yeah, not too bad. i did some hacking in gnome-session over the weekend to try and make it load faster [11:56] oh, nice [11:56] did you manage to find any issue or thing we could do better? [11:57] seb128 - from what i saw here, the biggest amount of time is wasted reading gconf keys, and the first read always seems to take ages, regardless of how late you do it. [11:58] so i made some changes which avoid reading any gconf keys until later on in the session [11:58] and some other stuff which avoid doing things which waste time (eg, querying the session type from consolekit when there is no saved session to restore from etc) [11:58] wasn't the "read the gconf keys early" a change in the recent cycles for optimization? [11:59] seb128 - it takes nearly 600ms to read the gconf keys on my virtualbox install [11:59] chrisccoulson: hrm [11:59] gconf is slow... [11:59] hi vuntz [11:59] chrisccoulson: why would reading the gconf keys be slow? [11:59] I'm not sure it will be faster doing it in several calls later [11:59] vuntz - i'm not sure, but it seems to just take ages [11:59] chrisccoulson: if it's because it's the first thing to read in gconf, then not doing it in gnome-session means that g-s-d will be impacted [12:00] chrisccoulson: (which just moves the issue) [12:00] vuntz - thats not so much of a problem. the issue is with it happening in gnome-session, is that it delays the whole session from starting [12:00] chrisccoulson: g-s-d will also delay the whole session [12:00] since g-s will block on it [12:00] :-) [12:01] possibly. i will try various things to see if i can speed it up though, as it takes far too long :( [12:02] i think our whole budget is 4 seconds isn't it? [12:02] seb128^^ [12:02] chrisccoulson: avoiding the CK call is good, though. How much do we win with this? [12:02] chrisccoulson, right [12:02] vuntz, I'm thinking about dropping the wait for g-s-d to be done to start other things, etc [12:03] vuntz - the consolekit call is fairly insignificant, but i'm just trying to save time whereever i can spot something which wastes it [12:03] it just slow down the login time [12:03] or we need to make things register much earlier and get gnome-session moving [12:03] seb128 - would you mind trying my build of gnome-session in a bit? [12:04] i'm just re-testing it to make sure it doesn't crash ;) [12:04] chrisccoulson, I would be happy to test your changes ;-) [12:04] seb128: oh, it should certainly not wait for all modules to be done before telling gnome-session is ready [12:04] seb128: there are "mandatory modules" and the ones that can be set later [12:04] vuntz - the other thing which seemed to take a while on my VBox install was creating the DkpClient object. so i made a change to only create that when it's needed [12:04] that took nearly 100ms [12:05] chrisccoulson: oh, that's probably because it starts DK-power [12:05] chrisccoulson: that's a good one, indeed [12:05] vuntz - yeah, possibly. it's better if that's started later in the session really [12:06] vuntz, http://people.canonical.com/~seb128/bootchart/seb128-dellmini-lucid-20091127-2.png is current gnome-session [12:07] see the delay between the session bar and the things starting there [12:07] vuntz, http://people.canonical.com/~seb128/bootchart/seb128-dellmini-lucid-20091127-8.png is gnome-session starting everything [12:07] vuntz - i will probably put all my findings on a wiki page somewhere shortly, which will detail where all the time seems to be going [12:07] chrisccoulson, feel free to use the spec page for those notes [12:08] seb128 - yeah, i can do that [12:08] chrisccoulson, https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-startup-speed [12:08] seb128 - thanks [12:08] and https://wiki.ubuntu.com/DesktopTeam/Specs/Lucid/StartupSpeed [12:08] * vuntz would recommend a page on live.gnome.org if you want to get upstream involved... [12:09] seb128: what do you mean with "gnome-session starting everything" ? [12:10] vuntz, I moved everything to the application stage [12:10] and cleaned the required component list [12:10] and copy autostart for gnome-panel, nautilus, compioz [12:10] compiz [12:10] just to see what difference it would make [12:10] it's a bit surprising that gnome-panel is really busy at the beginning for one and not for the other [12:10] I don't say it's what we need to do [12:11] some of the things might be blocking on gvfs to be started, etc [12:12] like in the first case they are ready before gnome-panel but not in the second case [12:12] lunch time there [12:12] bbl [12:14] * mpt wonders what "A text tool utiliy" is supposed to be [12:17] mvo, what was that separator and471 was talking about? [12:18] mpt: I send him a mail about this, I'm not sure what purpose it serves [12:19] mpt: there is also a change that makes the view list a tree with just one sub-element, I think we need to wait with that until we actaully have more than one sub-element [12:19] mvo, what do you mean by "with just one sub-element"? [12:20] mpt: a treeview with "Get software" as root and "Free software" as child [12:21] mvo, ah, probably he's looking at the 4.0 mockup. :-) I do want that for 2.0, though I haven't specced it yet [12:21] mvo, I agree it doesn't make sense to have it until we're showing packages from arbitrary repositories [12:22] seb128 - i've put my gnome-session work here for you to try: http://people.ubuntu.com/~chrisccoulson/gnome-session-startup-time/ [12:23] it might be worth testing it with the default phased startup first, as i've not tried it with your configuration [12:23] the changelog is quite detailed too, and explains what i changed [12:23] mvo, so the first big tasks are (1) showing non-app packages, (2) showing packages from all your enabled repositories (maybe that follows automatically), and (3) showing those repositories separately in the navigation pane [12:23] hopefully it makes things a little quicker [12:25] mpt: (1) and (2) are will be the same code, (3) is new to me, was that discussed and I missed it (other than in the 4.0 mockup)? [12:26] mvo, not yet -- maybe we can discuss prioritization of new interface bits this afternoon [12:27] mpt: ok, we need to add work items for them too (and put them somewhere) so that the tasks show up in the burndown chart [12:28] mvo, you mean an overall Foundations burndown chart? [12:29] mpt: yes, or my personal one. it needs to appear somewhere so that time can be budget for it [12:29] yes [12:29] mpt: adding a repository in the navigation is not trivial as it needs backend db integration [12:29] ok [12:31] I guess that means budgeting time for fixing particular bugs too [12:32] mpt: I thnk that depends on the bugs, if its requires code rework, then yes, otherwise I think we can just fix them when we reach the deadline for feature additions [12:32] mpt: do you have particular ones in mind? [12:32] mvo, not yet :-) [12:33] ok === MacSlow is now known as MacSlow|lunch [12:53] chrisccoulson, ok thanks [12:58] one question, will 2.28.* updates will continue to be pushed to karmic ? [13:00] baptistemm, depends of the update [13:00] if it has interesting fixes yes [13:00] otherwise no [13:08] and If I do packages, can they be reviewed? [13:12] baptistemm, sure, which one do you want to do? [13:13] I need to check which one are to do. I just wanted to know; because I know you're now focused on Lucid [13:14] we are [13:14] you are welcome to help on lucid too though [13:16] hmm, yeah I did the fontconfig 2.7.3 packaging but I'm not happy with, it crashes during the upgrade all gtk+ apps [13:16] it seems this is solved in 2.8.0 [13:16] I'll propose it tonigh once I tested [13:17] I can do the nautilus + gvfs update on karmic if they fit in my free time [13:18] I need to stick with a stable version snow as I use ubuntu for real work :) [13:20] fontconfig 2.7.x is not proposed on http://piware.de/workitems/desktop/lucid/versions.html but on http://piware.de/workitems/desktop/karmic/versions.html [13:21] seb128: would you have a minute today to review https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-suspend-quirks-halsectomy ? [13:21] pitti, yes [13:22] pitti, is there any webpage listing specs waiting for a review? [13:22] the blueprint emails are in the middle of the launchpad bug noise there [13:22] I need filter for those too I guess [13:22] seb128: try https://blueprints.edge.launchpad.net/people/+me/+specs?role=approver [13:22] danke [13:22] then sort by "Design" [13:23] "Review" need approval too right? [13:23] I'm not sure what is the difference between review and pending approval there [13:23] I guess it's launchpad granularity being higher than what we use? [13:28] seb128: I was just asking myself the exact same question :) [13:28] (about difference between review and needs approval) [13:33] mvo, I've updated the department+section algorithm to use just "Category:" and "X-Ubuntu-Category:". https://wiki.ubuntu.com/SoftwareCenter?action=diff&rev2=249&rev1=247 [13:33] https://wiki.ubuntu.com/SoftwareCenter#Genre [13:35] mpt, what is X-Ubuntu-Category? === seb128_ is now known as seb128 === seb128 is now known as seb128_ [13:36] seb128_, it's a way of overriding Category: in .desktop files so that a package appears somewhere else in the Ubuntu Software Center === seb128_ is now known as seb128 [13:37] mpt, do you consider there are cases where upstream and package categories might be different? [13:37] or is that just a workaround to avoid having fixing broken category use? [13:38] either way could you document in the wikipage? [13:39] seb128, we might discover a need for a subcategory that isn't in the menu spec, and we need to use it before the maintainer updates the spec. [13:39] seb128, I think (though I'm not sure) that another use of it will be for non-application software that should be categorized in the Center, but shouldn't show up in the Applications menu. [13:39] I would suggest it's not the right way to deal with those issues [13:39] seb128, what do you suggest? [13:40] we should rather being proactive fixing spec issues than adding an extra layer as an excuse to no push for the correct changes [13:41] not sure what I would suggest, I just want to be carreful to not create yet another situation where we are categorized as bad citizen by upstream because we do hacks rather than trying to fix issues correctly [13:41] if there is a valid need for that new key good [13:42] seb128, it's not an excuse. It's the same basic approach used in CSS, and RFC(2)822, and HTML5 [13:42] but if that's just because it's easier to change a new key than fix the existant one I would suggest that pushing for the real fix benefits everybody [13:42] I've no clue about either of those or about web coding ;-) [13:42] Part of knowing whether a new key is a good idea is implementation experience [13:43] well we could use the existant Categories key and tweak the known categories for that [13:43] seb128, sorry, I don't know what you mean by "known" in that sense [13:44] http://standards.freedesktop.org/menu-spec/menu-spec-1.0.html [13:44] it has a list of official categories [13:45] seb128, yes, I linked to that exact list from the SoftwareCenter spec [13:45] known = those listed there === MacSlow|lunch is now known as MacSlow [13:45] we could add tweaks to this list and implement the changes in our gnome-menus [13:46] if we figure we need some extra categories [13:46] one problem with adding our own categories is that we need to be careful with regard to i18n [13:48] seb128, if we did that we might use a new category name in a different way from how it was eventually standardized, and then we'd be in a mess. [13:49] I don't understand why using the fdo categories wouldn't work correctly? [13:49] or why we couldn't tweak those in a way which works [13:49] seb128, for the two reasons I gave ten minutes ago [13:50] seb128, for example, where is the fdo category for IDEs? There isn't one. What if we realize we need an "IDEs" subsection in the "Developer Tools" department? [13:50] mpt: there is no risk of that, because your private categories start with X- while standardized ones don't [13:50] the new category one is not a valid reason, we could as easily modify the current key than make a new one work [13:51] seb128, so we propose an "IDEs" Category to fdo, and they argue for a few months about whether that's really just a variety of TextEditors, and meanwhile we're stuck [13:51] no [13:51] we use a X-Ubuntu-IDE meanwhile [13:51] oh, I see [13:52] So instead of "X-Ubuntu-Category: IDE", we have "Category: X-Ubuntu-IDE" [13:52] yes [13:52] mpt: let me read through the new algorithm, but it still strikes me as overly complicated [13:52] seb128, ok, that works for me [13:52] good ;-) [13:52] and that makes the algorithm a bit simpler too I think [13:53] excellent [13:54] mpt: I think seb128 is right, one key makes everything simpler [13:54] I didn't know there were X-Y-Z Category *values* [13:55] mpt: and the exact location can be a comination of "and" and "or". e.g. "Sound & Video": "AudioVideo AND Audio AND Video" [13:55] mpt: Science/Mathematics: "DataVisualization OR Math OR NumericalAnalysis" [13:56] mpt: (forget the AND in audio & video, that should be a OR) [13:56] I guessed that :-) [13:56] mvo, ok -- all the commas in my table are ORs, but I can make that explicit [13:56] mpt: that should be fine [13:56] mpt: I figure the "," means "OR" [13:57] mpt: I don't understand (5) - "# Use “Tech Stuff” as the primary department, with no primary subsection. " [13:58] mpt: does that mean there is some tech stuff that has a sub-section and some that does not ? [13:58] mpt: when I asked about this at UDS you said it will be "tech stuff", "sub-section", "packages" [13:58] mvo, as written in the version you're reading, "Tech Stuff" as no subsections [13:58] mpt: does that mean we have "tech stuff", "some packages" and "sub-section" with "more packages" ? [13:59] mpt: aha, so its just one flat long list [13:59] mpt: that fine then [13:59] mvo, for the first iteration, yes [13:59] then we start working out how to extract particular types of package [13:59] (without using debtags, alas) [13:59] mpt: please upate trunk, that should give you a first cut of this [13:59] oh cool [14:00] mvo, all departments that have subsections probably will have some items that are not in any subsection [14:00] its still a bit buggy and no de-duplication, but you get a idea [14:01] mpt: hm, so the UI needs to be split into "sub-section" and "apps that are under this section but have no sub-section" [14:01] chrisccoulson, you broke my GNOME! [14:01] mpt: sounds like the equivalent of a "other" sub-section to me, just without the explicit label [14:01] mvo, yes, that's the next thing I need to update in the spec, the department screen [14:01] chrisccoulson, I only get a spinning cursor with your gnome-session changes [14:01] mvo, exactly [14:01] mpt: why not use other than? [14:02] mvo, because there's no point -- no-one would want to explicitly go into a subsection labelled "Other" [14:02] mpt: that means the UI needs to be different for categories with item and subsections and categories with sub-sections but no items? [14:02] mpt: or will all categories with sub-sections also have items that are not part of those sub-sections? [14:02] mvo, no, I think it will be two-pane in either case [14:02] ok [14:22] rickspencer3: ping [14:23] hi fagan [14:23] I hear on the grapevine that F-Spot's viewer is after being fixed [14:23] Is it too late to talk about getting rid of EOG too? [14:24] * fagan wants a bit of streamlining [14:24] fagan, let's see how it goes with f-spot [14:25] if it truly works just as well as eog for viewing photos, perhaps it won't be necessary [14:25] hey rickspencer3, good morning [14:25] hi pitti, good morning [14:25] rickspencer3: cool just thought I should ask [14:25] hey rickspencer3 [14:25] * pitti hands rickspencer3 a christmas cookie [14:25] hey rickspencer3 [14:25] hi seb128 [14:25] had a good thanks giving weekend? [14:26] * rickspencer3 *munch* *munch* *munch* [14:26] seb128, yes, it was quite nice [14:26] * kenvandine ate way too much :) [14:26] hehe [14:26] hmmm Ireland doesnt have thanksgiving :( [14:26] fagan: but we do have a 10 day christmas and new years eve eating festival [14:27] lots of tasty email to go go through this morning too [14:27] czajkowski: true [14:27] * seb128 is somewhat glad to have a food diet between sprint, uds and december [14:28] there is no way I can review all that bug mail [14:28] any specific ones I should be aware of [14:28] ? [14:28] tseliot, hello and welcome to the desktop team :) ! [14:28] seb128 - hmmm, that's wierd [14:29] rickspencer3: hi, and thanks :-) [14:29] did you move the nautilus and gnome-panel desktop files? [14:29] 477127 [14:29] chrisccoulson, where do you store required components? [14:29] or 479156 [14:29] chrisccoulson, no [14:29] chrisccoulson, it's stock lucid [14:29] I'd love someone to tel me why my software center is gone [14:29] (if I restored everything properly which I think I did) [14:29] seb128 - the required components are in /etc/xdg/gnome-session/components [14:30] there might be some clue in ~/.xsession-errors why it went wrong [14:30] that is there [14:30] czajkowski: karmic? [14:30] yes [14:31] czajkowski: did you try to reinstall it? [14:32] chrisccoulson, WARNING: no provider registered for component WindowManager,Panel,Filemanager: the key file has no key " WindowManager,Panel,Filemanager " [14:32] fagan: reinstall Karmic, or ? [14:32] seb128 - oops. i've just realised that the components must be separated by a semi-colon [14:32] i made the change on my machine when i realised it, but didn't update the package [14:32] my bad ;) [14:32] chrisccoulson, I wonder why it works for you ;-) [14:32] czajkowski: the software center sudo apt-get install it and see if that fixes it [14:33] if it doesnt its a bug id say [14:33] seb128 - it works for me because i updated the file locally here, but forgot to change the package ;) [14:33] possibly, I know I'm not the only one who is missing it [14:33] fagan: ya don't say, hence I copied in the bug numbers bug 479156 [14:33] Launchpad bug 479156 in ubuntustudio-menu "gnome-app-install instead of ubuntu-software-center" [Undecided,In progress] https://launchpad.net/bugs/479156 [14:33] bug 47127 [14:33] Launchpad bug 47127 in ntp "The daily update couldn't update ntp (dup-of: 47097)" [Medium,Invalid] https://launchpad.net/bugs/47127 [14:33] Launchpad bug 47097 in ntp "upgrade ntp-server crash" [Medium,Fix released] https://launchpad.net/bugs/47097 [14:33] bah [14:34] chrisccoulson, what do I have to change? [14:34] seb128 - the file should look like this: http://paste.ubuntu.com/331731/ [14:35] czajkowski: it hasnt been updated since the 23rd of October in karmic so I have the slight feeling that its something very small [14:36] czajkowski: run software-center in terminal does it show or does it give out an error [14:36] fagan: as I said, I will when I get home. at work [14:37] Oh ok [14:38] brb, need to restart session === robbiew_ is now known as robbiew [15:01] chrisccoulson, your changes seem to make no difference on login [15:01] compiz starts exactly at the same point [15:21] mpt: expanding a treeview/listview row dynamically is in itself not a big deal, the problem is that we need the treeview in "all-rows-have-equal" height mode. otherwise its getting really slow (because gtk will query each row every time the treeveiw change). so having a "+" button to install is easy, but expanding is not that easy [15:22] seb128 - hmmm, that's wierd [15:22] have you got the bootchart? and also the contents of your ~/.xsession-errors? [15:25] chrisccoulson, sec [15:34] chrisccoulson, http://people.canonical.com/~seb128/bootchart/seb128-dellmini-lucid-20091130-1.png lucid stock [15:34] chrisccoulson, http://people.canonical.com/~seb128/bootchart/seb128-dellmini-lucid-20091130-6.png your changes [15:36] seb128 - gnome-settings-daemon starts about 1/2 second quicker with my change, but like vuntz predicted, the delay just gets pushed elsewhere (to gnome-settings-daemon) [15:37] but, i could probably find a way around that too [15:38] seb128 - there should be some timestamps in your ~/.xsession-errors too. those would be useful to keep [15:40] chrisccoulson, do you want a copy of those? [15:47] seb128 - yes please [15:47] it might be worth running my other build without the changes, to compare the timestamps [15:47] i'll put the info on the spec then === bjf-afk is now known as bjf [15:56] chrisccoulson, http://people.canonical.com/~seb128/bootchart/.xsession-errors [15:57] seb128 - thanks. i get a permission denied error though ;) [15:57] chrisccoulson, retry [15:57] thats better. thanks [15:57] chrisccoulson, what is your other build? [15:57] where rather [15:58] seb128 - i'm just trying it now to make sure it still works [16:05] seb128 - my other build is here: http://people.ubuntu.com/~chrisccoulson/gnome-session-startup-time/profiling/ [16:05] the only changes in that one is the logging of timestamps in ~/.xsession-errors [16:09] mvo: fixed [16:13] and471: many thnaks [16:13] chrisccoulson, ok, trying that [16:13] seb128 - thanks [16:20] chrisccoulson, [16:20] http://people.canonical.com/~seb128/bootchart/.xsession-errors-stock [16:20] and [16:20] http://people.canonical.com/~seb128/bootchart/.xsession-errors-speed [16:21] seb128 - thanks. that's pretty good actually - the largest delays seem to correspond with what I see here in virtualbox (albeit, they are much shorter on your hardware) [16:21] still, 500ms to load gconf keys though [16:21] 150ms to start DK-power [16:23] so, that's quite a time saving. i just need to figure out a way to not shift those delays elsewhere, where they still delay the session loading [16:24] is the dkpower win effective or is that delayed too? [16:25] seb128 - it's delayed until it is actually needed. on the default session, it will be g-p-m that starts it (with this change) [16:25] but the session should be usable by then [16:25] ok, so it's a good win ;-) [16:26] or gnome-session will start it if you call the session dialog (but that won't happen whilst the session is initializing) [16:26] seb128 - i'm thinking of writing a small helper which starts alongside gnome-session and reads some of the gconf keys that are needed to load the session [16:27] that might then avoid the delays of reading them later on [16:27] and that could happen in parallel with gnome-session reading the desktop files then [16:27] do you need an another process? can't you get gnome-session to do that? [16:28] is that time io busy or cpu busy? [16:28] and471: good timing, I did some merging today and hope to upload 1.1 today [16:28] could we try to do disk preloading before the session? [16:28] mvo, cool [16:28] seb128 - i could do, but it would need to at least be in another thread, as the gconf calls all block [16:28] hum [16:28] not ideal indeed [16:29] it's not. that's why i thought about just writing a small helper which starts gconf, and does a read of the keys that are needed by gnome-settings-daemon [16:29] pitti, desktop-lucid-suspend-quirks-halsectomy could you give details on what the other remaining usages are? [16:30] and then hopefully when gnome-settings-daemon is started, it won't take ages to read them [16:30] "usages"? [16:30] seb128: ah, the bits I sent upstream? [16:30] pitti, "drop other remaining hal usage from pm-utils: INPROGRESS" [16:31] seb128: only trivialities; I updated the work item for clarification [16:49] pitti, ok thanks, looks good to me, approving now ;-) [16:51] weirdness [16:51] if the external mixer control box is off when I boot, something sets the sound card output volume to 0 on boot [16:51] if it's on, it's whatever I left it at [16:51] oh, is that it? [16:52] I keep wondering why the volume is 0 most of the times, but not always [16:52] pitti: do you have an external mixer though? [16:52] Keybuk: just external speakers [16:52] if it happens for you too, I may be mis-triaging ;) [16:52] the internal ones don't actually deserve to be called "speakers".. [16:52] * Keybuk has a physical box that the sound card outputs to, that then amplifies it and sends it to the 8 speakers [16:53] (sometimes I wonder why I don't work for Gentoo I'm clearly a ricer) [16:53] ricer? [16:54] http://www.urbandictionary.com/define.php?term=ricer [16:54] Keybuk, the difference is you're doing most of your ricing for millions of people, whereas the average Gentoo ricer does it individually [16:54] lol [16:54] pitti: or http://funroll-loops.info/ [16:55] lol [16:57] * Amaranth waves [16:58] hey Amaranth [16:58] Amaranth, could you make compiz stop to depends on compiz-fusion-plugins-extra? [16:58] at least recommends it [16:58] so I can give a try to what boot difference it makes [16:59] sudo dpkg -P --force-depends [16:59] pitti, I should have added "without breaking my box or having it complaining every time I want to do upgrades" [16:59] ;-) [16:59] seb128: I think I need to talk to the guy handling the compiz settings spec [16:59] heh [17:00] Amaranth, he's djsiegel there when he's around [17:00] Yeah I always forget his IRC nick :) [17:00] Amaranth, and I was there at sprint and uds, so I can probably reply to questions too [17:00] I don't want to hijack it from him [17:00] Amaranth, what do you need to know? [17:00] he doesn't care about packaging [17:00] Well getting rid of that dependency means changing the default plugins [17:00] he's just looking at a good set of options for something visually nice [17:00] Although I really wish I was able to get to that meeting... [17:01] you said the other day that nothing from extra was used? [17:01] No, but I don't know if he has anything planned [17:02] Amaranth, https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-compiz-effects [17:02] Amaranth, there is a bzr with what he wants [17:03] it's in the spec whiteboard [17:05] ah he actually has some changes made now [17:06] Amaranth, well he asked some people to test his changes and did a demo at uds [17:06] so I guess it's working since uds [17:07] hehe, and he is changing settings that don't even work... [17:08] which ones? [17:08] the cursor ones [17:08] that's an Ubuntu patch that (last time I checked) doesn't even do anything [17:08] I think we've been forward porting it since hardy... [17:11] Amaranth: hey [17:11] djsiegel: ah, there you are [17:11] hello again [17:12] seb128, I'm picking gnome-bluetooth to update, I'll do gvfs / nautilus if possible [17:12] baptistemm, ok [17:12] djsiegel: I see you want to speed up some of the animations in compiz [17:12] yeah [17:12] djsiegel: I believe you were the same person fighting against doing that in gnome-do for the same reasons I'm about to give you :) [17:13] let me hear 'em [17:13] Amaranth, that was me actually [17:13] djsiegel: How many frames is 80 milliseconds assuming 60 fps? [17:14] A little under 5 [17:14] Is 60fps what we're working with? [17:14] You think showing 5 frames for a glide 2 animation looks smooth? [17:15] djsiegel: Well yes, most people have LCD screens [17:15] 60fps is the high end, most laptops screens are 50fps [17:16] DBO: Are you sure about that? You may just be seeing nvidia's dynamic twinview crap [17:16] intel [17:16] positive about it too, a lot of the higher resolution LCD's are 50-52Hz [17:16] That's even worse then [17:17] Amaranth: check http://launchpad.net/compiz-settings-lucid [17:17] I am pushing a branch right now [17:17] the file exported-settings is my compiz settings dump [17:17] (I have been having trouble hot-loading gconf settings) [17:18] hrm, you seem to be setting a shadow radius of 16 as well [17:18] Does your laptop use intel graphics? [17:18] mine does not [17:18] Amaranth: these are not being handed down from god [17:18] if you have any objections or suggestions, tell me and I will make changes [17:19] Alright well I know at one point setting shadow radius to a multiple of 2 resulted in solid white shadows with intel [17:19] Let me check if that still happens [17:19] Amaranth, cleaning extra wins us 0.5s cpu usage [17:19] seb128: Consider it gone [17:20] I'm wondering what compiz still does for 8 seconds [17:20] That's horrible [17:20] Amaranth: do you know of a way to load/dump compiz settings from the command line? I would like to automate it a bit, and messing with gconf has not worked well enough [17:21] djsiegel: something using ccs, I guess [17:23] kenvandine, hey [17:23] het seb128 [17:23] hey [17:23] kenvandine, would you be interested in doing the empathy 2.29 update in lucid? [17:23] sure [17:23] kenvandine, rebase on debian experimental for packaging, they have it [17:23] most of the work is to port the indicator changes [17:23] thanks! [17:23] ok 16 seems to not break anything with intel anymore, bug fixed [17:23] seb128, is tomorrow ok? [17:24] djsiegel: compiz (animation) - Error: Animation settings mismatch in "Animation Selection" list for Focus event. [17:24] kenvandine, next week is ok, it's early lucid, plenty of updates and merges still to do [17:24] (I ran python compiz_settings.py) [17:24] ok, it will be a bit of work to update the patch [17:24] kenvandine, ie whenever you can will do [17:25] kenvandine, thanks! [17:25] Amaranth: yes, I see that [17:25] no animation is set for the focus event [17:25] so instead of simply having no animation, compiz spews? [17:25] why do people keep blaming me for that? :) [17:26] we didn't have an animation for focus events before either [17:26] hmm [17:26] well, I will add an empty entry then [17:26] before there was an animation rule matching all windows, with animation set to None [17:27] now there is no rule, so compiz whines [17:27] I will restore [17:27] hmm [17:27] shoot [17:27] compiz profile does not remember some keybindings, and number/layout of workspaces [17:27] weird, gconf-editor says there is a Fade event set for all windows [17:28] for focus, I mean [17:28] oh, I think it's because you have two matches for the focus one but only one animation... [17:29] maybe not [17:29] adding an animation in ccsm didn't fix it either [17:30] weird, what did you do to it? :) [17:30] oh, I see [17:30] djsiegel: You left focus_durations having two values [17:30] I re-added the empty focus and shade [17:30] but there was only one focus animation set [17:31] Huh? [17:31] I had zero focus animations set [17:31] I think the settings are not applying cleanly? [17:31] then your code is failing to wipe all of my settings [17:31] It's not my code [17:31] but I didn't have any focus animations either [17:31] it's gconftool --load [17:32] So, is there any better way to load these settings? [17:32] I've tried loading to gconf, and compiz doesn't seem to like it [17:32] I just tried compiz profiles, and they don't capture all settings [17:32] jono: just checked out the code for lernid, it looks good but it is got a few bugs that I can't seem to solve, I am getting frustrated by quickly's architecture :-/ [17:32] it is > it has === eeejay is now known as eeejay_away [17:33] djsiegel: I dunno, I've always just edited the xml files when I wanted to change default settings [17:33] compiz (animation) - Error: Animation settings mismatch in "Animation Selection" list for Minimize event. [17:33] Which xml files? [17:33] *sigh* [17:33] djsiegel: The ones in the source package... [17:34] Amaranth: so, is that how I should swing this? [17:34] Edit that file? [17:34] that might be easiest [17:34] well, several files [17:34] one per plugin [17:35] Maybe I should just express all the changes as separate bugs? [17:35] we have quilt patches for each source package that modify default settings [17:35] then you can comment on a case by case basis if you find something wrong? [17:35] that could get messy, you make a lot of changes :) [17:35] maybe one per package [17:35] one bug per package? [17:36] yeah [17:36] what do you mean by package? [17:36] compiz plugin? [17:36] oh, and don't change anything for the extra plugins [17:36] which are the extra plugins? [17:36] no, the plugins are split up into compiz-plugins (compiz source package), compiz-fusion-plugins-main, and compiz-fusion-plugins-extra [17:36] hmm [17:37] and we'll be doing further splitting soon [17:37] I am only applying changes to plugins that were already enabled by default in Karic [17:37] karmic* [17:37] asac, around? [17:37] compiz-fusion-plugins-extra is getting kicked out completely (we only use extrawm from that and don't set any settings for it) [17:37] Amaranth: can we set some time aside to go through the changes I want to make? [17:38] then we can figure out the best way to proceed? [17:38] the rest are going to get split into plugins we use or that are commonly used and plugins that are crack [17:38] well if that time isn't today it's probably going to have to be next monday [17:38] I was hoping we could just have branches with xml files that are easy to load [17:38] Amaranth: any time [17:38] it could even be in a few weeks [17:38] baptistemm: for two/three more minutes ;) [17:39] but the loading of new setting en masse doesn't seem to work [17:39] okay, asas I"m cleaning the patch of gnome-bluetooth to update 2.28.3, and I wanted to know if 03-fix-killswitch-memleak.patch was reported to bgo? [17:39] s/asas/asac/ [17:40] baptistemm: no. please forward [17:40] okay [17:41] okay thanks [17:41] djsiegel: as long as you aren't enabling or disabling any plugins I can wait [17:41] djsiegel: but I need to know that soon because I want to get the packages split [17:42] does anyone know how fast synaptic refreshes its speed estimation? [17:42] * hyperair thinks this might be the reason synaptic causes X's CPU usage to rise significantly [17:43] actually I'm thinking about splitting each plugin into a separate binary package and just not having plugin pack packages at all [17:43] just make the 'compiz' metapackage depend on the plugins we use directly [17:43] hyperair: mvo would be the one to ask [17:43] ah [17:43] right [17:44] well i'll just wait then [17:44] hehe [17:44] Amaranth: that sounds like the best approach, if not the most arduous... [17:44] Amaranth: if you do that, i think it'd be a good idea to at least have them around as metapackages for easy installation. [17:45] Amaranth: https://wiki.ubuntu.com/Lucid/CompizDefaults [17:46] Commands, Negative (Needed for ally?), and Display Info on Resize are the three plugins I think we can disable [17:46] I know some terminal users who would seriously disagree with that last one :) [17:46] and commands is needed for compatibility with metacity [17:47] really? the commands plugin is devoid of any settings [17:47] which is the only reason I thought we could deprecate it [17:47] they are synced from metacity [17:47] and explain the terminal users' use case [17:47] oh, ok [17:48] ah, does the display info on resize on display resize info on terminals? [17:48] only display* === eeejay_away is now known as eeejay [17:48] yeah [17:48] Weird, I missed that setting [17:48] although there is also a bug there that makes resizes really slow for some people [17:48] what does that plugin cost us? [17:48] for terminals and Qt [17:48] apparently all Qt windows set the hint that they want the resize info [17:49] Yeah, I think the terminal use case may not be a strong enough reason to keep the plugin enabled [17:49] I'm fine with dropping it, really [17:49] just saying we're going to get complaints [17:49] if it has any cost for the average user [17:49] yeah [17:49] People throw crazy fits when it doesn't work exactly like metacity [17:49] we will get complaints no matter what [17:50] and471, thanks for trying it! [17:50] did you file the bugs? [17:50] jono no not yet [17:50] and471, if you could file them that would be awesome [17:50] I am now in sheening mode for 0.2 [17:50] jono, can I just send you the patch :-) [17:51] djsiegel: The main reason I want to get rid of the "plugin pack" packages is because they are not even named correctly anymore [17:51] compiz fusion is gone, it's all just compiz [17:51] but we haven't figured out upstream how we're going to split things up [17:51] ok [17:51] and471, indeed, just make a branch and propose a merge :) [17:52] jono, okay :-) [17:52] :) [17:52] thanks! [17:53] Amaranth: can you think of reasons to keep viewport-switch-on-window-drag-to-edge or snapping windows? [17:54] I think the first is poorly executed, and could really confuse people, and is asymmetrical because it currently doesn't allow dragging up (could be fixed, of course) [17:54] and the second is not helpful [17:54] hyperair: pulse() is run ~30/sec or so, could be that [17:54] and can cause frustration when dragging windows around, especially with netbook trackpads [17:55] djsiegel: We _have_ to keep snap [17:55] Why? [17:56] djsiegel: Because otherwise I'll throw my computer trying to line things up :) [17:56] What do you try to line up? [17:56] djsiegel: Try using your computer without snap for a week [17:56] Or do you just maximize everything? [17:56] I have, it's great [17:56] no, I don't maximize [17:56] please, tell me your use cases [17:56] I don't doubt you [17:56] i just need to know [17:57] I stick videos in one corner, a terminal in another, a tomboy window next to that terminal etc [17:57] Now, wobbly snap is very annoying and I hate it [17:57] But regular snap is awesome [17:57] so, you use it to align windows? [17:57] ok, I think we can do window snapping [17:57] but I want screen-edge snapping off [17:58] Err, half my use case was screen-edge snapping :) [17:58] at least against the panel, at the very least [17:58] you already have y-constraint [17:58] so why do you need to snap to the top panel as well? [17:58] you just drag against it [17:58] and it snaps [17:58] Why do you want to disable it? [17:59] Ideally we'd have edge resistance and not snap and none of them would matter [17:59] Because users get frustrated when dragging windows around and they snap too aggressively [17:59] s/them/this/ [17:59] I think we can fine-tune it [17:59] jono: https://code.launchpad.net/~rugby471/lernid/lernid-andrew [17:59] there are two types of people, I think [17:59] (1) those who understand the snapping behavior and use it to their advantage [17:59] jono, if you have and questions about the merge, the info is in the commit message, otherwise don't hesitate to ask me now :-) [18:00] and (2) people who don't use snapping, but are still affected and annoyed by it [18:00] so, let's adjust the settings so people who benefit from snapping still benefit, and people who don't use it aren't bothered too much [18:00] djsiegel: Alright, I can see that [18:01] and>any [18:01] btw, looking at the wiki it seems you want to move everything to the super key [18:01] are you going to do that for metacity too? [18:01] Yeah, that's tricky [18:01] Actually you _have_ to do it for metacity too [18:01] Otherwise the compiz settings won't stick, metacity will override them [18:01] What are the metacity equivs of show all windows and expo? [18:01] djsiegel, Amaranth : bear in mind some people don't have a super key [18:01] doesn't exist [18:01] but show desktop [18:01] yes, I know [18:01] yeah [18:02] thanks and471 [18:02] gonna take a look [18:02] djsiegel: also if you switch to 2x2 workspaces we might as well toss cube out completely [18:02] jono, are you a first time pygtk programmer or veteran? [18:03] cube is already not enabled by default [18:03] and471: jokosher... [18:03] so what do you mean "toss it out completely" ? [18:03] djsiegel: Right but I mean don't make the Extra visual effects enable it [18:03] Amaranth, ah yes [18:03] and471, I have written a bunch of apps, but still consider myself a novice [18:03] don't even install it [18:03] shove it in universe [18:03] etc [18:03] and471, you need to propose it for a merge [18:03] Amaranth: ah, is that what the extra effects do now? [18:03] yes, we will need to still have the cube... [18:04] jono, just curious, and see above :-) [18:04] I was wondering if there's any way to enable the cube with 2x2? [18:04] I'm pretty sure that'd what macslow made extra do [18:04] jono, oh sorry [18:04] No, you can't have vertical workspaces with cube [18:04] re plugin pack packages [18:04] jono, can you not just merge manually? [18:04] I don't see anything wrong with switching to 4x1 workspaces when you enable extra effects [18:04] and471: I still have not merge, sorry :( tomorrow, promised! [18:05] and471, I can do, just want to keep everything in place [18:05] mv [18:05] I am checking out your branch now [18:05] * hyperair coughs [18:05] djsiegel: That gets...tricky [18:05] jono, never mind, I have proposed now [18:05] mvo, hehe np [18:05] mvo, take notice though, jono is merging something I wrote 5 minutes ago :-) [18:05] mvo, hehe joking [18:05] and471: *weehhh* ;) [18:05] djsiegel: Are we going to automatically change a users 6x6 workspaces to 36x1 when turning on cube then change it back to 6x6 if they do back to wall? [18:05] and471: what percentage of people do not have a super key? AFAIK, it's only some thinkpad models which do not have it? [18:06] and471: its because the branch contains too much good stuff [18:06] djsiegel: Or are we just going to toss their settings for that when switching? [18:06] and471: that give me ideas [18:06] Amaranth: 99% of people will stick with the default settings and never change them [18:06] and471: like the config thing is great, I would like to make it a singelton and use gconf optionally [18:06] djsiegel, absolutely no clue, but up until around a year ago I didn't (it bugged me when I had to fiddle with gnome-do to change it, ironic you were a dev on that :-] ) [18:06] oh, the most we can do is 32x1 anyway :) [18:07] djsiegel, anyway it is not just thinkpad, some old desktop keyboard just don't have them [18:07] and anyone who has configured 32x1 can figure out how to get that back [18:07] i don't have super on my desktop [18:07] and471: yeah, we won't put anything essential behind super [18:07] like, you need to use Super to turn off the machine :) [18:08] cube is worthless with more than 4 workspaces anyway [18:08] djsiegel, hehe :-) [18:08] mvo, what to you mean make a singleton? [18:08] Amaranth: I think it's perfectly reasonable to think of the action of setting desktop effect settings as applying a blanket of new settings [18:08] when you choose "extra effects", you take what we give you [18:08] djsiegel: I so want a superpower button... [18:09] that's why there's a custom option [18:13] and471, looks good, the only bug I found was Lernid hanging when you exit the app [18:13] could you take a look at fixing that in your branch and then I will merge it in [18:13] jono, yeah, there is a weird thing going on with gtk [18:14] jono, sorry I don't have time, was just telling mvo, I have exams :-(, plus I am tired :-) [18:14] Amaranth: https://edge.launchpad.net/hundredpapercuts/+milestone/lucid-round-9 [18:14] and471, ok, I will merge this in, we can fix that later, it is not critical right now [18:14] Amaranth: I am going to file bugs there describing the changes we need to make [18:14] jono, see ya [18:14] thanks for the patch :) [18:14] I will aim for 1 bug per affected plugin? [18:14] And I will exhaustively list the settings [18:14] jono: np, btw great idea for an application :-) [18:14] in a spec document [18:14] djsiegel: alright, that'll be fine then [18:14] thanks and471 [18:14] btw, what is your name and email - for the credits? [18:14] dammit [18:15] djsiegel: hopefully I'll have compiz reading GtkSettings for the colors, btw [18:15] then you or others can make patches out of those specs? [18:15] yeah [18:15] Amaranth: what does that mean? [18:15] it will fix the viewport switcher bug? [18:15] djsiegel: Meaning following your theme [18:15] where will it follow? [18:15] djsiegel: Like the themed docky dots patch I made :) [18:15] yes, where exactly will the colors be used? [18:16] everywhere compiz uses blue right now [18:16] chaotic (lead visual designer at canonical) and I sat down and manually styled all compiz visuals enabled by default [18:16] iirc only resize [18:16] ah [18:16] well, so long as they look good with Human and NewWave that works too [18:16] yeah [18:16] Resize, Viewport Switch Preview [18:16] I think that was it [18:17] that's another reason to ditch cube, actually [18:17] Amaranth: https://bugs.edge.launchpad.net/compiz-settings-lucid [18:17] Why? [18:17] the OSD when switching workspaces looks awesome :) [18:17] but cube doesn't have it [18:17] it could look awesomer: https://bugs.edge.launchpad.net/compiz-settings-lucid/+bug/484523 [18:17] Launchpad bug 484523 in compiz-settings-lucid "Compiz is not using the Outline color set for the viewport switch preview in the Wall plugin" [Undecided,Confirmed] [18:18] hrm, white and gray? really? [18:18] yeah, to match the look of notify-osd [18:18] notify OSD is black [18:19] right, well, black + alpha, gray + alpha, white + alpha [18:19] that sort of see-through grayscale look [18:26] jono: 'help-and-run' I may have accidentally deleted a signal handler in a glade file, and that would be the cause of the problem. Check the revision diff and you will probably find it. Bye! [18:29] * hyperair coughs [18:29] well then, now that X has been OOM-killed and everything's back to normal... [18:29] mv-oh he disappeared >_> [18:30] * hyperair sighs [18:38] Hello. Does anybody know if behind the choice of putting nautilus scripts in /usr/share/nautilus_scripts and assuming they are installed only in that place there was some discussion? [19:00] So, how do you guys feel about turning the Nautilus location bar and possible status bar off by default? [19:02] kenvandine: https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-lucid-social-from-the-start -> note that you should have one "work items:" (you just renamed it to -alpha 2) [19:03] oh [19:03] ok [19:05] how does that look? [19:06] kenvandine: I think it'd be better to swap around [19:06] "Work items:" -> for alpha-2 [19:06] and keep the a3 ones in "Work items (alpha-3)" [19:06] the blueprint is targeted for alpha-3 [19:06] kenvandine: asac and I recently discussed how to clean this up and provide tool support [19:06] so won't it default to that? [19:07] but right now it just shows "work items:" stuff [19:07] kenvandine: we don't generate a report for alpha-3 [19:07] oh [19:07] ok [19:22] kwwii: btw, you can create an fd.o wiki acct on http://www.freedesktop.org/wiki/UserPreferences [19:28] hello [19:29] seb128, I did some packaging, gnome-bluetooth, gvfs and nautilus [19:30] seb128: if gdm fails to start X, does it exit with a useful exit status? === robbiew is now known as robbiew-afk [19:52] Keybuk - the reference machine for the startup time work is a Dell Mini, isn't it? [19:53] chrisccoulson: mini 10v ssd [19:53] chrisccoulson: iirc [19:53] ccheney - thanks [19:53] iirc it started out as mini 9 ssd but became hard to source so switched to mini 10v ssd [19:55] * ccheney headed to eye doctor, bbl [20:01] yeah same hardware, different case [20:07] pitti, does the script that update http://piware.de/workitems/desktop/karmic/versions.html is running? [20:07] baptistemm: it should [20:08] (although the canonical link is /lucid/ now) [20:08] pitti, Hi by the way :) [20:08] pitti, you mean http://piware.de/workitems/desktop/karmic/versions.html is the same than http://piware.de/workitems/desktop/lucid/versions.html [20:09] baptistemm: hey! [20:09] baptistemm: it's a symlink, yes [20:09] hmmm [20:10] I wanted to see the version to update for karmic, I'm interested to update gnome components === eeejay is now known as eeejay_away [20:19] pitti, hmm poppler has a CVE fixed since 0.12.0, would you agree to upgrade to 0.12.2? [20:19] baptistemm: backporting the fix is fine [20:19] 0.12.2 breaks ABI without announcing so, so the full upgrade isn't suitable for SRU [20:20] (it horribly breaks cups, for example) [20:20] seb128 - i've detailed the gnome-session findings at https://wiki.ubuntu.com/DesktopTeam/Specs/Lucid/StartupSpeed now === robbiew-afk is now known as robbiew [20:20] i will probably do some g-s-d work later, so i might ping you for some more testing if you're still around ;) [20:20] * pitti hugs chrisccoulson [20:20] pitti, and what about 0.12.1? [20:20] * chrisccoulson hugs pitti [20:21] I don't know [20:21] there is also some security fixes apparently [20:21] http://cgit.freedesktop.org/poppler/poppler/commit/?h=poppler-0.12&id=bb21f825fcaa49e149cc7a7eb9c68f15f11fff1d [20:22] * baptistemm fears the noise coming from its laptop HD [20:27] chrisccoulson, excellent, thank you [20:30] baptistemm, better to backport the security patches rather I would say [20:34] gnome-bluetooth-2.28.3-0ubuntu1 compiled [20:43] does someone understand why my build fails http://launchpadlibrarian.net/36268161/buildlog_ubuntu-karmic-amd64.nautilus_1:2.28.2-0ubuntu1_FAILEDTOBUILD.txt.gz ? [20:44] it seems there is some breakage === jono_ is now known as jono [21:00] seb128 - i had an idea for improving g-s-d startup time, but i'm not sure what you think about it. the plugins are currently loaded in order of priority. i'm thinking that we should pick some plugins based on priority (such as xrandr, xsettings) which can block the rest of the session from loading, and then load the rest (lowest priority plugins, such as media-keys, housekeeping etc) when the rest of the session is loading [21:00] what do you think? [21:03] chrisccoulson, when is session registration done right now? [21:05] chrisccoulson, if it's done after high priority plugins it seems only a matter of tweaking those numbers? [21:05] in any case seems a good idea [21:08] seb128 - currently, it fork()'s, and the parent process exits when all of the plugins are loaded and started. the next session phase starts as soon as the parent exits [21:09] so it would just be a matter of the parent exitting after a plugin of a pre-determined priority has started [21:09] oh, ok, not really optimal indeed [21:10] what's "gdm-new" ? :) [21:10] Keybuk, in what context? [21:11] ubuntu-desktop PPA [21:11] the new gdm before it landed in karmic [21:11] different name to make switch between versions easier [21:12] it's in the jaunty ppa only, we should clean it now, thanks for noticing! ;-) === eeejay_away is now known as eeejay [21:14] seb128 - do you know what gnome-settings-daemon-helper is for? (it's part of the gnome-session source) [21:14] chrisccoulson, seems we need xrandr xsettings basically [21:14] seb128 - yeah, that's what i was thinking [21:14] no, but vuntz most likely does [21:15] seb128 - it seems like it only sets up an environment variable (GTK_RC_FILES) [21:16] not sure if that's used anywhere. the value it's set to tends to suggest it's old GTK1 specific [21:16] in which case, we could probably drop from the session ;) [21:18] a grep of the GTK source suggests that it's not used [21:20] kenvandine: ugh, desktop-lucid-social-from-the-start is a heavy beast.. [21:21] yes [21:25] Amaranth: hey [21:26] Just played with snapping, and set edge-resistance to 18px [21:26] djsiegel: howdy [21:26] I think it's really good [21:26] with no snapping, it's really difficult to place a window against the side of the screen, which I imagine many people would try to do [21:26] more often that placing two windows exactly side-by-side [21:27] we will have to test that edge resistance; it may be too small [21:42] oh dear, clunking noises from my hard drive is surely not good :( [21:50] Amaranth: is it possible to make both wobbly and snap available ? ;) [21:50] mac_v: wobbly has its own snap [21:50] it has to due to what it does [21:50] but that is not good :( [21:50] djsiegel: 18px feels a lot more like metacity [21:51] djsiegel: things don't stick together so strongly [21:51] do you think that's good? [21:52] yeah [21:52] djsiegel: we should keep loading the image plugins though [21:53] why? [21:53] was just looking over your settings, random thought there :) [21:53] djsiegel: Because plugins that need them don't depend on them so things just fail [21:53] hmm [21:53] ok I see [21:53] but those plugins shouldn't depend on them because they may not need them (depends on what kind of image you use) [21:53] hmm [21:53] well [21:53] we are configuring compiz for the default settings [21:54] if people want to do things that aren't the default, then they can enable them [21:54] well cube will need at least one of these [21:54] unless your average person would do something that would necessitate those image plugins [21:54] I don't see why [21:54] the image on top of the cube [21:54] hmm [21:54] also we should disable the video plugin, it has _never_ been useful or used [21:55] we can do a solid color for the cube tops [21:55] or enable the image plugins when you enable the extra effects [21:55] right? [21:55] that could work, yeah [21:55] I don't want to ship image plugins enabled when 99% of people will never ever use them [21:55] oh, and we need to enable the titlebar info plugin [21:55] just to make it so the small number of people who customize won't have to check a checkbox [21:56] although maybe I should just patch the titleinfo stuff into gtk-window-decorator [21:56] but we need the feature one way or another [21:57] What does it do? [21:57] adds a string to the window title for root windows and remote windows (ssh -X) [21:57] hmm [21:57] we only need it for root [21:57] I don't understand [21:58] oh, I see [21:58] so people can see what apps are running as root [21:58] let me test this [21:58] we've had people complaining about this not existing for some time [21:58] eh [21:58] I don't like it [21:58] seems like a hack [21:58] so it says "ROOT" in the title bar, is that really going to stop me from messing up? [21:59] well if I put it in gtk-window-decorator we can tweak the theme [21:59] yeah, windows running as root should be a separate theme [21:59] Amaranth: can we do that? [21:59] select a different theme for root windows? [21:59] I'll try to do something there [22:00] that could be cool [22:00] so forget titleinfo then [22:00] yeah [22:00] there is a bug for that one too... for a separate root theme ;) [22:01] so, Amaranth, disabling the video plugin -- that won't interfere with window thumbnails with video playback in them, right? [22:01] so disable video, change snap to 18px [22:01] pushed [22:01] djsiegel: No, it's supposed to be a replacement for Xv [22:02] hrm, do we really only have scaleaddons for the close window option? [22:03] oh, no, we do the title stuff in there too [22:04] Amaranth: ah, we also need to put the highlight back a bit [22:04] highlight? [22:04] oh, I think we should try turning resize mode back to normal [22:04] yeah [22:05] hmm, maybe not [22:05] weird white flashing bug with intel [22:06] Amaranth: so keep as rectangle? [22:06] yeah, for now [22:09] good night everyone [22:09] night, pitti! [22:10] night pitti [22:30] * ccheney found out the reason he can't see well with his glasses, his vision has improved quite a bit [22:30] from -5.50 to -4.50 [23:05] hmmm, was the launchpad downtime planned? [23:06] i don't remember seeing any mail [23:09] yes [23:10] Keybuk - thanks [23:10] * Keybuk had a mail [23:10] I remembered having it when I got the "DOWN!" screen :p [23:10] the announcements must go to a list i'm not subscribed to then :-/ [23:10] or they just get lost in the noise [23:13] chrisccoulson: oh, no [23:13] the mail I have is about a different time entirely ;) [23:13] heh, thanks ;) [23:13] so it's not just me then! === eeejay is now known as eeejay_away [23:41] awesome I just got the update to the software center that has system packages, its great to see it working. [23:45] Wow slow as hell though