=== jono is now known as ubuntu-goldfish [09:50] good morning all [09:51] morning [09:52] morning ochosi. Good to have you here because I noticed that there's a correction that has to be made to the formatting of http://docs.xfce.org/apps/parole/usage [09:53] ochosi: right now the audio section isn't formatted as a title, it's in the continuation of the last paragraph of the DVD section [09:53] brainwash: i know that currently the screen gets blanked twice, it even gets blanked twice on every display. kinda busy atm so i don't know when i'll really get to looking at the background-setting you mentioned ... [09:55] slickymaster: thanks, good catch! fixed [09:55] ochosi: ^^^^ [09:55] ochosi: np [09:55] knome: ping [09:56] slickymaster: i just realised now, but we'll have to add a "playback" section to the keyboard shortcuts [09:57] "Right arrow" > skip ahead 10secs [09:57] etc [09:58] ochosi: you mean another table named playback to that section? [09:58] yeah, i'm currently working on it [09:58] thought i'd just quickly chuck it in, it's rather easy to do [09:58] ochosi, if you feel like there's already plenty in your hands to deal with, I'll be glad to do it for you [10:00] slickymaster: actually, yeah, if you can/want... :) [10:01] i'll have to link you to the section of parole's code though that explains that [10:01] ochosi: No worries, I'll do it [10:01] I would appreciate the link [10:03] they start here: http://git.xfce.org/apps/parole/tree/src/parole-player.c#n2441 [10:04] and: seek_short = 10, seek_medium = 60, seek_long = 600 [10:04] (seconds) [10:04] as you can see there are more shortcuts that aren't in the docs right now [10:04] (f for fullscreen) [10:04] (p for playpause) [10:05] ochosi: I'll add those to the already existing tables [10:07] ochosi: I'll need just another thing though, the link to bluesabre's dokuwiki. I've lost it :( [10:09] hehe, no worries [10:10] slickymaster: http://smdavis.us/doku/doku.php?id=usage&#keyboard_shortcuts [10:11] ochosi: thks [11:30] ochosi: mmh, simply move the first set_background call and use it as fallback if there is no user background set via accountsservice? [11:31] ochosi: well, currently I simply commented the first call :D [11:31] out [11:32] but you will have to fix this, because it does cause the strange unblank behavior on two different systems [11:33] bug report required? [11:40] looks familiar -> http://ubuntuforums.org/showthread.php?t=2188790 === brainwash_ is now known as brainwash [11:55] slickymaster, pong [11:56] knome, good morning [11:56] g'day [11:56] brainwash: ok, i'll try to check it a bit later today then ;) [11:57] so I've speak with balloons regarding https://bugs.launchpad.net/ubuntu/+source/xubuntu-docs/+bug/1251332 and his opinion is that he should go with [11:57] sudo -i to get to a root prompt and there launch whatever graphical application [11:57] Ubuntu bug 1251332 in xubuntu-docs (Ubuntu) "Use of gksudo in Chapter 7. Printing and Scanning" [Undecided,New] [11:58] knome: so, if there's no objections from you I'll go ahead with that to re-write the documentation [11:58] slickymaster, ok [11:59] knome: I'll do it then and later on I'll propose a merge [12:04] slickymaster, ok, thanks [12:06] knome: np [12:57] brainwash: wanna show me the patch that worked for you? [13:24] ochosi: there is none yet, I simply commented out the first set_background call [13:24] that one was actually added to help with nouveau drivers when setting the background... [13:24] that's obviously not the best solution [13:25] it's a bit of a tricky situation, not everything seems to work with all drivers [13:25] I mean the part in the main function [13:25] to set the background which is specified in the config file [13:26] which is why i'm a bit hesitant to touch this stuff, i only have one laptop to test things on [13:27] well, create an extra branch :) [13:27] I can test it [13:28] hmpf yeah, all rather time-consuming and much tapping in the dark [13:34] what's this bug about backgrounds? [13:38] ali1234: well it's an issue in lightdm-gtk-greeter [13:38] lightdm always faded nicely from the default background to the user specified one for me [13:38] i thought it was supposed to do that [13:38] what greeter are you using? [13:38] i guess unity-greeter, right? [13:38] yeah [13:38] (lightdm-gtk-greeter doesn't fade) [13:38] what's the difference? [13:39] i thought unity used lightdm [13:41] lightdm is the display server [13:41] the rest are greeters [13:41] so the greeter is like a modular component? [13:41] yeah [13:41] ok, got it [13:41] the visual component [13:41] so the lightdm-gtk-greeter was written by ligthdm's author as a reference greeter [13:41] so who else uses the gtk one? [13:42] bluesabre and me picked it up and started improving it bit by bit [13:42] lubuntu i think [13:42] and maybe even distros outside ubuntu [13:42] (i think lunar linux, not sure anymore) [13:43] any reason why we don't use unity-greeter? [13:43] it's pretty nice... [13:43] well, firstly it depends on indicators [13:44] which we couldn't use for ages now [13:44] basically since they were ported to gtk3 [13:44] i think -gtk-greeter fits better with our style anyway, at least now that we've worked with it [13:45] the other thing is that the gtk-greeter is a component we control, so we can also introduce changes that make sense for xubuntu [13:45] e.g. introducing screen-blanking in case it's used as a lockscreen would be somewhat harder to get into unity-greeter [13:45] random: if the greeter can read the user's background, can it also read their theme? [13:46] it can read whatever accountsservice allows [13:47] accountsservice documentation isn't online except in source form [13:47] yeah, it's indeed not very well documented (it's been a while since i dealt with it) [13:48] and the package doesn't include the built docs [13:48] oh, and the user-wallpaper isn't in accountsservice upstream [13:48] so pretty much no chance then [13:48] that's an ubuntu-specific patch that never made it upstream (for whatever reason [13:48] ) [13:49] so anyway... if we can't fix the double background bug, how about implementing the fade/transition instead? [13:50] why can't we fix it? o.o [13:50] well, it sounds like the code is a bit fragile with all different drivers === olli_ is now known as olli [13:55] so we should disable the accountsservice part? [13:57] i'm not sure why we should disable the accountsservice part [13:57] it's not fragile at all [13:57] fix it then :) [13:57] yeah, i said i'll try to do so once i have more time [13:57] and testing resources are quite limited for me also [13:58] i already added a patch recently that should improve things, but it seems it doesn't work for everybody [13:58] I can write a patch (the way I like it to be fixed) [13:58] sure, go ahead [13:58] we can do it in a branch and try to get some testing [13:58] i'm all for fixing things [13:59] ali1234: fwiw, transitions and fades tend to produce lots of code. be my guest if you wanna submit a patch ;) (you could even use parts of thunar-desktop i guess) [14:01] is this thing really only 2000 lines of C code? [14:02] hehe [14:02] well personally i think if we wanna do fancy stuff like that, we should switch to the html engine [14:03] or even just fork unity-greeter and make it behave/look the way we want... [14:03] so every time you change the user selection it changes the background too [14:04] yup [14:05] making it do a fade would probably add like 5 lines of code [14:06] i'm totally happy to test and merge patches ;) [14:08] i am confused what this is supposed to achieve: http://bazaar.launchpad.net/~lightdm-gtk-greeter-team/lightdm-gtk-greeter/trunk/view/head:/src/lightdm-gtk-greeter.c#L1310 [14:10] what do you mean specifically? [14:10] the if(!matched) part [14:10] (just as a note in advance: as i mentioned before, we're not the authors of the greeter, we're just maintaining it as it was dropped by robert_ancell a longer while ago) [14:10] if you type a wrong username, it uses the first user's background? [14:11] well [14:11] the thing is: in xubuntu we use the user-list (in e.g. debian they don't they use the user-entry) [14:12] with the user-list, you could only type a wrong username when using "other" [14:12] as a few of the last patches by our debian maintainer have shown, we didn't really take care of the user-entry enough [14:12] actually, i understand it [14:12] oh, ok [14:12] it's the initially selected user [14:14] what if there are no users though? [14:15] something to test: install in a vm, usermod your UID to <1000, see if greeter crashes [14:19] is there some kind of testing harness for this? [14:20] "lightdm --test-mode --debug" apparently [14:21] indeed, it shows the bug [14:23] hm [14:25] so where do you maintain it if upstream dropped it? [14:25] in the same place, bzr [14:25] (i meant launchpad) [14:25] ok [14:26] lp:lightdm-gtk-greeter [14:31] what do all the files in /etc/lightdm/*.conf do? [14:34] you can use them to select the greeter / configure lightdm / configure the greeter/s [14:34] so how do i make it load ~/Source/lightdm-gtk-greeter/src/lightdm-gtk-greeter [14:34] so e.g. in /etc/lightdm/lightdm.conf you can select the greeter [14:35] did you install the greeter? [14:35] or what setting do you want to change specifically? [14:35] i want to use the greeter i just built without installing it [14:36] ah... it needs a .desktop file specially named [14:38] hmm that's interesting [14:39] it looks very different with default settings [14:39] yeah [14:39] it does [14:39] lack of icon-theme, gtk-theme etc are the usual issues [14:40] i considered providing a good fallback gtk-theme [14:40] it looks ok [14:40] yeah, could be better though [14:40] just different [14:40] and also some fallback icons, to be sure [14:40] otoh so far no-one has complained about it ;) [14:41] to be honest i think i prefer the defaults to the way xubuntu configures it [14:45] fine by me [14:45] we considered letting greeter-themes even change the UI file [14:46] in general, it wouldn't be hard to ship multiple configuration files and let the user choose [14:46] so why does it look different even after make install? [14:47] because the xubuntu config file isn't in the bzr repo [14:47] but it is still in my filesystem [14:47] not sure i get it [14:47] if you install the greeter, it overwrites the lightdm-gtk-greeter.conf file [14:48] as in: the one that was shipped with xubuntu-default-settings [14:48] it didn't [14:48] then you didn't add prefix=/usr maybe? [14:48] /etc/lightdm/lightdm-gtk-greeter.conf is unchanged [14:48] maybe wrong prefix, it should change that file [14:49] oh i see [14:49] so it's really reading the config from somewhere completely different because i built it with the wrong prefix [14:50] yeah, i'd say so [14:50] yeah, fixed [14:51] good === cyphermox_ is now known as cyphermox [17:11] I'm going to have to look at LP Bug 1252800 later today [17:11] Launchpad bug 1252800 in xubuntu-docs (Ubuntu) "/usr/share/xubuntu-docs/about/xubuntu-index.html packed in Trusty Daily still refers to Xubuntu 13.10" [Undecided,New] https://launchpad.net/bugs/1252800 [17:12] skellat, new doc package isn't uploaded to trusty yet. [17:18] knome: I'm not 100% sure, but apparently no [17:54] ok, so what exactly is so terrible about simply not making the first call to draw the background? [17:55] if it fixes some driver bug, what about if we draw a black background here instead [17:55] and then draw the user or default background later [17:56] tbh i think the call didn't really help with nouveau [17:56] so i think we can also look into restructuring that [17:57] i can't see a reason why just skipping that draw will hurt anything [17:57] but drivers are weird so who knows [17:57] yeah, they are... [17:57] well i'd suggest we work in a branch and then do testing with all drivers we can [18:00] it's important though to keep in mind that the greeter shall not fail [18:00] it's too important [18:01] well the easiest way to guarantee that is to not functionally change it [18:01] ie by drawing a black pixmap on that first call, and only the first call [18:01] which is what bluesabre and me have done mostly [18:04] i can also imagine doing functional changes, as long as there's someone at work who feels confident with that (i don't) [18:09] @2000 lines of code there's a limit to how many bugs there can possibly be [18:09] ali1234: Error: "2000" is not a valid command. [18:09] indeed [18:10] depending though on what kind of other stuff you involve [18:10] many of our bugs have been lightdm bugs === J21_ is now known as J21 [20:03] hi sergiobenrocha2 [20:03] knome: sergiobenrocha2 is looking to get more involved with us :) [20:03] hey sergiobenrocha2 [20:03] I've been moving him about freenode ... we've come here now [20:04] welcome [20:04] Hah. [20:04] Welcome to another channel! [20:04] quick line from another channel - which sums it up [20:04] why didn't i get that tour? [20:04] elfy: ok, is possible that I create an wishlist in launchpad, and appears in this page, if the feature is good? [20:04] knome: please join #ubuntu+1 [20:05] then when you've done that - can you come here :p [20:06] lol [20:06] thanks everyone [20:06] done [20:06] :) [20:06] *that* was all of freenode? [20:07] no - just the bits I dragged sergiobenrocha2 around to [20:07] he started in #xubuntu [20:09] sergiobenrocha2, we basically just finished creating roadmap [20:09] sergiobenrocha2, was there something specific you had in your mind for the wishlist? [20:09] humm, yes, a thing or 2... or more [20:10] to make xubuntu more friendly [20:10] would you like to share them with us here or want to keep secrets? [20:10] :) [20:10] haha [20:11] so: [20:14] 1 - create launchers in panel would be more friendly in right-click, it is better to pop-up a configuration window that create a complete launcher, and not only an empty gray icon [20:14] sergiobenrocha2, did you know you can drag-and-drop launchers from the menu to the panl? [20:14] *panel too [20:14] sorry for english... portuguese is my language [20:14] Translator standing there ^ ;) [20:15] knome: but it possible to drag-and-drop from whiskermenu? I'm using it [20:15] * knome shrugs [20:15] i don't use a menu at all, so i don't know [20:16] maybe somebody who uses whiskermenu knows [20:16] possibly [20:16] maybe you should try. :) [20:17] I can drag-and-drop only favorites apps from whiskermenu, like terminal and thunar. Others, in sub-categories, it not possible [20:17] you should file a wishlist bug against whiskermenu for that [20:17] because it works with the regular xfce menu [20:18] ok [20:18] but whiskermenu is from ppa, how can ubuntu-bug will report that? [20:19] i don't think it can... you have to report manually [20:19] do it manually - if you can [20:19] https://github.com/gottcode/xfce4-whiskermenu-plugin/issues?labels=bug&state=open [20:19] file a new issue. [20:20] humm, ok [20:20] probably tag it with "feature" [20:24] ochosi: as usual, when i start looking closely at some code, i start to notice it's kinda funky [20:25] i have a feeling this thing leaks a display-sized xlib surface every time it changes the background [20:25] the set_background code also does a bunch of surface setup, before which things may not work properly [20:32] I created the issue [20:33] but I can not see, if i click in "Issues" tab [20:33] it's the number 37 [20:34] oh, ok, not worry [20:40] confirmed, lightdm-gtk-greeter leaks memory like crazy. [20:43] ali1234: surprise? :) [20:43] never :( [20:43] so you plan to fix the background part? [20:43] i'm not sure i want to touch this code [20:44] all this background stuff is related [20:44] the only easy fix is to replace that first sat_background() with something like clear_background() which will just clear it to black but otherwise do the exact same as set_background [20:45] ie all the memory leaking and stuff [20:45] the alternative is to rewrite it so that allocation and drawing are separate [20:45] this would fix both problems [20:45] but it's a quite big refactoring [20:46] allocation of the background image? [20:46] and i'm not sure if i really care for something i see for 10 seconds twice a month [20:46] so, when the greeter starts up it makes a pixmap which is the background [20:46] actually, set_background() does this [20:47] it creates a xlib pixmap, then draws the background into it with cairo [20:47] but the problem is that the next time you call it, it makes another new pixmap and does not free the old one [20:47] yes, but the greeter does it twice, if accountsservice provides an user background [20:47] I accidentally created a duplicate, but I changed it to an another issue, it is Issue #37. Can someone take the duplicate label? [20:47] yes, it does it every time you select a different user from the list [20:48] right [20:48] if you sit and click different users you can make it leak hundreds of mb if you are patient [20:48] no one does that [20:48] ali1234, we would be grateful if you fixed that [20:49] brainwash: could be a security vuln, i dunno... [20:49] who knows what happens if the display is locked and you crash lightdm... [20:49] anyway the thing is, it has to allocate and leave allocated that pixmap [20:49] it might even be necessary that the pixmap exists before you start doing gtk stuff [20:50] the background won't change if the screen is locked via greeter [20:50] this might be the nouveau bug that was mentioned [20:50] so this means you can't just defer drawing the background, because it also *creates* the background [20:51] I simply want this bug to be fixed -> bug 1251431 [20:51] bug 1251431 in LightDM GTK+ Greeter "user background gets painted over background specified in config file" [Undecided,New] https://launchpad.net/bugs/1251431 [20:51] this also fixes the light-locker issue for me [20:51] what does? [20:52] ochosi made a change, so that light-locker blanks the screen after locking it [20:52] it only works for me, if the background is set only once [20:53] right yeah, that makes sense [20:53] you know what i said about those pixmaps getting leaked? [20:53] uhm, it's bad [20:53] well, i bet the patch he made expects there to be only one [20:53] so it hides the top one [20:53] but there's still a stack of leaked ones behind it [20:53] so to fix *that* [20:54] requires the "proper" (but harder) fix of refactoring this thing [20:54] I mean this here "XForceScreenSaver(display,ScreenSaverActive);" [20:55] called twice unblanks the screen [20:56] at least for me [20:56] on two different systems [20:56] yeah [20:56] not this particular line, but the code part [20:56] hmm [20:56] no sure what you mean really [20:56] the communication with X [20:57] the thing is that all those leaked background pixmaps persist after the greeter exits [20:57] yes, but that's not related to the blanking issue :) [20:57] well, why did you mention it then? [20:58] it's related to bug 1251431 [20:58] bug 1251431 in LightDM GTK+ Greeter "user background gets painted over background specified in config file" [Undecided,New] https://launchpad.net/bugs/1251431 [20:58] ok. i don't understand how it is related... [20:58] so we really need to cleanup the code [20:59] unless you mean it's related in a general sense of "this code sucks and has lots of problems" [20:59] which is fair enough [20:59] take a look at the create_root_surface function [21:00] this function contains the line to blank the screen (after locking it) [21:00] so it behaves like xscreensaver [21:00] yeah [21:00] if you call set_background, you also call create_root_surface [21:01] ok... [21:02] and this is bad, because my screen does not remain blanked [21:02] ok, see that's where you lost me [21:02] if I call set_background only once, it works [21:02] screen stays blanked [21:02] ok, so on startup you mean [21:02] right [21:03] startup after yu pressed "lock screen" that is [21:03] yes [21:03] the normal screen lock does no allow user switching anyway [21:03] yeah but that's irrelevant [21:04] the problem is it sets the background to default, then creates a bunch of widgets, then sets background to the user one [21:04] so... we can fix many problems with a code cleanup :D [21:04] but it can't get the user background until after it did all that widget stuff [21:05] can't we set the background to default only as fallback option, if the user background is not available? [21:05] yes [21:05] but that means defering it [21:05] until after the point where we usually load the user background [21:05] yes, won't hurt or? [21:06] well, it might do, because there's no root pixmap until after set_background [21:06] currently I've simply commented the default background call out [21:06] yes [21:06] actually what you need to do is move the set_background to... [21:07] yeah [21:07] if (lightdm_greeter_get_hide_users_hint (greeter)) [21:07] which is much further down the functions [21:07] around 1890 [21:08] the solution is quite easy [21:09] go on... [21:09] looks like you are very interested in fixing this :D [21:09] it's fairly simple to fix, just a lot of refactoring [21:10] you are more skilled, I would mess up and break the code :/ [21:10] that's the problem with refactoring things [21:11] that's why nobody ever wants to do it [21:11] usually you have some tests to verify the functionality [21:13] but I'll try to write a patch later (to at least provide a fix for my bug report) [21:13] part of the problem is that it doesn't just use one pixmap [21:13] it makes one per screen [21:13] so then you have to have an array to keep track of them [21:14] and you have to handle screens getting added and removed [21:14] ouch [21:14] currently it does none of this, it just blasts all over whatvever exists "right now" and then leaks it [21:15] what about the unity-greeter? is it more mature? [21:15] dunno. it's vala, and huge [21:15] it's written in vala [21:15] yes [21:15] i'd rather fix this code i think [21:15] i bet unity-greeter has just as many problems [21:19] actually it's even worse than i thought, because it disconnects from the xserver after dumping the pixmap on it [21:19] so if you keep a reference to it, it can go invalid without warning [21:20] this is really hairy... [21:20] elfy, ping [21:21] you should ge a cup of coffee [21:21] get [21:21] * genii readies the coffeepot [21:21] i just realised this while i was brewing some tea [21:21] s'up knome [21:21] elfy, you worked with the community wiki i assume? [21:21] elfy, the community *help* wiki [21:21] once or twice [21:22] elfy, aha, okay [21:22] elfy, i still might need your opinion.. [21:22] well - that was a figure of speech :) quite a bit :) [21:22] but i'll get back to you later on that [21:22] lol [21:22] haha, in that case [21:22] i'll get to you now :) [21:22] lol [21:22] noticed i took a work item on refreshing the frontpage? [21:23] https://help.ubuntu.com/community/FrontPageRefresh [21:23] that's a brief start [21:23] i'm starting with trying to see what kind of stuff would be best grouped together [21:23] ok [21:24] how up-to-date is the NewDocs page? [21:25] pretty much worked on constantly [21:25] and why that name and not MostPopular or sth [21:25] god only knows [21:25] lol [21:25] also, [21:25] do you think the alphabet-pages could have some kind of shared banner [21:25] last post in the thread a bunch of people working on newdocs was 4 days ago [21:25] to link back to the frontpage (NewDocs) and the rest of the alphabets [21:26] gtg, brb [21:26] talk to slickymaster about that newdocs thingy - he'll have an idea [21:41] okay [21:42] he's done a lot of work with those doing that page [21:42] but what was it you wanted from me? [21:45] well, comments about my page [21:45] and if you had other ideas and insight what would be useful on the front page, and if there is something you think could be dropped [21:45] ok [21:46] do you think the alphabet-pages could have some kind of shared banner [21:46] i'm definitely not forcing any layout or changes that people who are familiar with the wiki do not think are useful, and open to all suggestions [21:46] what did that refer to? the newdocs thing? [21:46] yep [21:47] all the pages could have a banner that said "Popular posts" and then list all the alphabets [21:47] I'd imagine the only reason it hasn't is that no-one has either not bothered yet - or can't [21:48] lol [21:51] I've bookmarked your new page - I'll look tomorrow afternoon when I get back [21:52] but - the current page definitely needs a revamp :) [21:54] ali1234: there is another greeter based on lightdm-gtk-greeter [21:54] elfy: on the subject of the wiki and specifically that triage page - i think the descriptions of what bug statuses mean should be removed from that page and merged into Bugs/Status (which is linked near the bottom) [21:54] it might have some of those issues fixed [21:55] eewh, surge protection [21:55] ochosi: oh? where is it? [21:55] ali1234: here https://www.google.at/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CDIQFjAA&url=https%3A%2F%2Fgithub.com%2Fkalgasnik%2Flightdm-another-gtk-greeter&ei=TN6LUo0WjNfsBoibgegI&usg=AFQjCNHD3XUgYZOkNz9mE_foxPUMSomhlg&sig2=L2mGPe5-4j1kK_ohW2SGaQ&bvm=bv.56643336,d.ZGU [21:55] gah [21:55] sory [21:55] ochosi, well done [21:55] yeah :( [21:55] ali1234: https://github.com/kalgasnik/lightdm-another-gtk-greeter [21:56] ali1234: i've also gotten in touch with the author, because the changes seemed quite vast and i'm not sure we'd want all of that [21:56] hmm yes looks like it has some changes with how screen backgrounds work [21:56] he replied he might help but he was busy and haven't heard back since [21:56] well, i can try to port his fixes [21:57] did you read all my ranting? [21:58] do you happen to know what revision he forked from? [21:59] https://github.com/kalgasnik/lightdm-another-gtk-greeter/commit/a21a7e15e77d54fc36042f3589a34193c7d65841#diff-803c5170888b8642f2a97e5e9423d399L129 [21:59] typing in wrong window is bad... [22:01] wait, is this even a fork? [22:02] yeah, i read all your rants [22:02] this code has no resemblance to lightdm-gtk-greeter at all [22:02] but as i said, i'm not the author and didn't feel comfortable with fiddling with the bits and pieces too much [22:03] (or at least with the larger bits and pieces) [22:03] yeah i can't blame you for not wanting to change anything in this code [22:03] *i* don't want to either [22:03] it mostly worked so far [22:03] elfy, the wiki is impossible to edit! [22:03] the way it works is a nasty brutal hack [22:03] it works... no more, no less [22:03] so we started ironing out some stuff [22:03] and improved the looks [22:03] that's pretty much where it is now [22:03] knome: why do you think I said once or twice [22:04] ;) [22:04] I hate the wiki :) [22:04] +1 [22:04] ochosi, i'm talking about a different wiki! [22:04] i've started to consider making small previews of the background-submissions just to make the page-loading more bearable [22:04] and don't whatever you do logout of it - it takes a week of sundays to log back in ... [22:04] ochosi, i considered the same thing actually [22:05] ochosi, and maybe we simply should do Incoming and Accepted [22:05] wouldn't take very long to do it [22:05] yeah, as i suggested once... [22:05] no, not really [22:05] ok, how then? [22:05] and i said go ahead if you think that's what we need [22:05] and then you were "meh" [22:05] yeah, but i wanted to wait for pleia2 actually [22:05] "wouldn't take very long to do it" -- "no, not really" [22:06] knome: can you at some point soon - check the LTS pending blog, fiddle with it if necessary and get it published :) [22:06] not too worried about the other one [22:06] knome: the good part of that distinction would be that the not-yet-accepted ones would be more visible [22:06] ochosi, yes [22:06] right - I'm off now - back tomorrow [22:07] ali1234: fwiw, i never got around to really testing that other greeter, feature-full as it might be [22:07] ochosi, and we could have a naming scheme for the files that makes it easier to handle them [22:07] ok, i'll start setting up an accepted-page [22:10] knome: ok, so are there any submissions you wouldn't accept so far for some reason? (apart from the ones with license-issues) [22:10] i don't think there is any reason to decline any other at this point [22:10] yeah, i agree [22:10] so, naming-scheme [22:11] easiest is title. but title_author would also work [22:12] could even be author_title [22:12] yeah [22:12] true, there are many that submitted more than one [22:12] titles are just arbitrary... titles [22:17] ochosi, i would also make a note to the accepted page that those wallpapers are accepted *submissions*, and that there is no promises of them getting into the release, and that people shouldn't edit that page directly [22:18] yeah [22:20] knome: feel like doing a few thumbnails? ;) [22:21] send me all the files and i'll batch-convert them. [22:21] that'll take ages with my connection [22:22] ochosi: did you add all the "is_locked" stuff? [22:22] ali1234: yeah, in the last commit (or the one before that) [22:22] ochosi, heh. [22:22] is it's only purpose to trigger the XForceScreenSaver in create_root? [22:23] ochosi, what do you want me to do then? [22:23] ali1234: yeah. the main issue was that i didn't want to write another function to cycle through all displays... [22:23] knome: download the pics yourself, man! :) [22:24] ochosi: well my concern is it can only ever be called on gdk_display_get_default() anyway [22:24] which should always be the same thing [22:24] so why not just call it once, from main()? [22:24] sounds quite plausible :) [22:24] then every instance of is_locked can be removed [22:25] ochosi, all of them? [22:25] knome: all of the accepted ones [22:25] uh [22:25] what have you done so far? [22:25] knome: i'm still in process of moving/renaming them [22:25] okay [22:25] so do you want me to do that once you've moved? [22:26] as you wish, i mean basically you already know which ones i'm gonna move [22:26] well if the moving is fine for you, it's easiest for me to just grab * from one page [22:27] that's also ok [22:27] ok, good [22:27] let me know when you're done [22:27] basically the renaming is the more annoying part [22:27] sure [22:31] ok does someone want to try lp:~a-j-buxton/lightdm-gtk-greeter/experimental [22:31] i think this will fix the locking thing, but not the double drawing background [22:31] i might be totally wrong though [22:31] sure, i'll do that as soon as i've finished the monkey job in the wiki... [22:32] hmm... want a script for that? [22:33] hm, if you can include creating thumbnails, sure! :) [22:33] not a problem. what exactly do you need to do? download all submitted images and thumbnail them? [22:34] that's the one part, the more annoying part is moving some of the submissions to the accepted-page [22:34] and rename them to author_title [22:34] so that's hard to script [22:35] probably best to suffer of it once [22:36] after that it's one or two per time [22:37] yeah, i mean the script for downloading images and creating thumbs might still be handy [22:38] well blah :) [22:38] i also hate the plaintext links [22:38] might fix that later [22:49] did you just delete a bunch of stuff from the page? [22:49] ochosi is destroying my inbox [22:51] pleia2: sorry ;) [22:51] maybe it's time to set up some filter-rules? [22:51] why is there an application_octet-stream attached to the page? [22:51] ali1234: yeah, not sure, it's just now getting easier to see all the attachments [22:51] knome: i think i'm done [22:52] ochosi, i'll get to that soon. i need to fix something else first ;) [22:52] ochosi: normally I want to see :) I just keep getting notifications, whee [22:52] pleia2: sorry, but the cleanup was direly needed... [22:53] I know, I kid anyway [22:55] ali1234: hehe, that octet stream looks like a debian wallpaper [22:55] http://www.zimagez.com/zimage/screenshot-11192013-115531pm.php [22:56] pleia2, i'm the next one to spam your inbox [22:56] i like it [22:58] yeah [22:58] nice'n'spacey [22:58] pity that there's no author etc [23:00] pleia2, ready for another round of spam? [23:00] pleia2, ready or not, here it comes [23:01] si [23:04] ali1234: ok, locking and blanking still works [23:04] so far so good [23:07] ali1234: the fun thing is i wanted to implement it like that in the beginning, but then i somehow thought i have to force the screensaver for each display separately [23:08] display/screen confusion [23:08] easily done [23:08] yeah [23:08] :/ [23:08] well, thanks! [23:08] if you could submit it as a merge-request, i can merge it in [23:09] though i'm not sure why you get away with calling it multiple times for each screen, but calling the whole loop more than once breaks [23:09] unless of course you never tested on multiscreen :) [23:09] yeah i can mr it [23:09] i have a dualhead setup here [23:10] so that might fix blanking for brainwash hopefully [23:10] http://paste.ubuntu.com/6445334/ [23:10] knome: ^ [23:11] they are already uploaded [23:11] this only downloads and thumbnails them anyway :) [23:11] i'm just laying out the page [23:12] i think i'll keep the download-part of that script around, we could attach a tarball with all backgrounds for convenience [23:13] ah [23:13] good ol BeautifulSoup [23:13] hey there bluesabre [23:13] it will probably crash if there's anything "not an image" attached to the page [23:13] when it works, it works well [23:14] or if someone uploads an image named "do=view&.jpg" [23:14] actually probably not since it would be escaped [23:14] that would be only slightly evil [23:14] in what package would i find that beautiful soup? [23:15] python-beautifulsoup [23:15] gah, i was looking in libpython [23:15] i'll start testing the wallpapers then [23:16] just have to add a rm *_250.jpg [23:17] ali1234: you wouldn't know anything about using the X11 dpms extension? [23:17] ochosi, https://wiki.ubuntu.com/Xubuntu/Roadmap/Specifications/Trusty/CommunityWallpapers/Accepted [23:18] knome: fail: juergendonauer_walkingfishermaninthesunset.jpg_250 [23:18] fixed already you stupid [23:18] :P [23:18] lol [23:18] oh sure, pin it on me! :D [23:19] * knome pins [23:19] damn [23:22] ochosi: i didn't know anything about any of these things before i started digging :) [23:23] ali1234: hehe, well same here ;) anyway, thing is that that would be the preferred blanking of the screen, the standby of the display [23:24] so you want the screen to turn off when it does XForceScreenSaver? [23:24] yeah [23:24] cause the VT-switch cause the screen to switch off for a second [23:24] so ideally it would stay that way [23:24] instead of "waking up" only to stay blank [23:26] DPMSForceLevel? [23:26] http://ftp.x.org/pub/X11R6.9.0/doc/html/DPMSForceLevel.3.html [23:27] knome: set xfdesktop to change the wallpaper to one of the submissions ;) [23:27] ali1234: yeah, i'd try DPMSForceLevel(display, DPMSModeOff); [23:27] ochosi, huh:P [23:27] but from what i heard, there are multiple dpms extensions for X11 [23:27] i have no idea how to test this stuff - screensaver/locking is horribly broken here, i had to remove all of it [23:27] so i started to be a bit doubtful of what would work, it also adds an extra dependency [23:28] what are you using now? [23:28] the monitor power switch [23:28] the proposed solution for xubuntu is light-locker [23:28] haha [23:28] that one also wouldnt need the dpms patch i guess [23:29] for some reason the display still goes blank after a time, but does not turn off [23:29] well those are different things afaik [23:29] blanking and standby [23:29] with potentially different timeouts [23:29] check "xset q" for your local settings [23:29] they're all disabled in the control panels [23:30] Standby: 0 Suspend: 0 Off: 0 [23:30] not sure what those control panels you're referring to control [23:30] and in the screensaver part? [23:30] me either [23:30] there is no screensaver part [23:30] rly? [23:30] funny [23:30] oh wait [23:30] timeout: 600 [23:30] that's it [23:30] 10min [23:31] a bit short for my taste, but ok [23:31] so if you enable standby, it'll shut off the backlight of the monitor [23:31] bbl [23:31] (or was it suspend that did that, i'm always confused by those two namings...) [23:34] bluesabre: btw, i noticed that the parole-docs don't cover the "hidden" playback-shortcuts yet (arrow-keys etc), slickymaster agreed to add them [23:34] oh yeah [23:34] :D [23:35] cause those are actually quite important in the docs, as they're not easily discoverable [23:35] right [23:53] ali1234: because you mentioned cross-fading would be only 5 lines, what method would you use exactly? [23:53] set a timer, draw the background multiple times [23:53] that was before i really looked at the code though [23:53] doing this would be a terrible idea since each time you draw the background it leaks a pixmap [23:56] that just means we have to make our default wallpapers lean ;) [23:56] no [23:56] the pixmap is monitor sized [23:56] contents don't matter, it's not compressed [23:56] this is what create_root does [23:57] true (i was kinda kidding there) [23:57] specifically the pixmap it creates doesn't get freed - by design [23:57] i'm reading all about RetainPermanent at the moment [23:57] yeah, that might've been robert_ancell's idea then [23:57] i can see why it's done [23:58] if you only set the background once it makes perfect sense [23:58] yeah, which originally was the case [23:58] because it stays around after the greeter exits [23:58] the background-switching was introduced by us (mr_pouit) [23:58] but if you change the background, you're making a new one each time, and they all stay [23:58] maybe best to check what unity-greeter does [23:59] as it also shows user-wallpapers [23:59] change the background 100 times and you leak 1GB [23:59] maybe it would be better to only do this RetainPermanent thing once when the greeter exits