=== thumper-afk is now known as thumper [03:50] Good morning [05:01] good morning [05:05] you are all up early :) [06:22] hah [08:04] ahoy [08:06] good morning desktopers [08:06] hey Laney [08:06] how are you? [08:09] seb128: not bad thanks, although it's tipping it down with rain :-( [08:09] you? [08:09] I'm good thanks [08:09] it's nice and sunny outside here [08:09] hah === alf_ is now known as alf|xmir_devel [08:15] we couldn't be bothered to bring the laundry in yesterday :P [08:17] that looks like a mistake today I guess? ;-) [08:17] it certainly does [08:18] oh yay, we got gstreamer [08:18] huzzah [08:18] seems so indeed [08:19] https://lists.ubuntu.com/archives/saucy-changes/2013-October/011525.html [08:19] rsalveti mangled the changelog :P [08:19] (thanks for uploading it) [08:19] who uploaded? [08:19] (not me) [08:20] oh, rsalveti [08:20] some days I forgot he has upload rights ;-) [08:20] seems there was another merge anyway [08:20] good, we got a settings upload as well [08:20] nice busy changelog [08:20] rsalveti: you should upload merges with -v btw [08:21] yeah, that's nice too [08:21] maybe we'll even get an image with working tz [08:21] 8 different commiters listed [08:21] go settings team ;-) [08:22] @image with working tz... yeah, once they manage to ack an image [08:22] seb128: Error: "image" is not a valid command. [08:22] meetingology, shup up stupid bot [08:22] seb128: Error: "shup" is not a valid command. [08:22] shut [08:26] seb128: Error: "shut" is not a valid command. [08:27] * seb128 kicks ogra_ [08:27] :) [08:27] stupid bot [08:27] oh, sorry :p [08:27] haha [08:27] ogra_, guten tag, wie gehts? ;-) [08:27] moin moin :) [08:27] gut [08:27] sonnig hier :) [08:28] here as well! [08:28] not in u.k though, according to Laney [08:29] bah [08:29] you'll have your turn :P [08:30] didrocks, so, I did the error to exit the download settings' panel while it was download, since then every time I re-enter it, it keeps spinning "checking for updates" in an endless way ... bug on our side on or the service side? [08:30] seb128: the UI is so dummy that I would say the service side [08:31] seb128: if it's spinning, it's getting no signal feedback from the service [08:31] (after the request is done) [08:31] didrocks, ok, thanks === vrruiz_ is now known as rvr [09:24] mdeslaur, ping? [09:27] mhr3, he's in .ca, it's like 5am for him [09:27] mhr3, e.g try again in some hours [09:28] seb128, realized now... anyone else from security? [09:28] * mhr3 doesn't see chriscoulson [09:28] mhr3, you can try chriscc [09:28] yeah... tab completion fail :p [09:28] try #ubuntu-hardening [09:28] try #ubuntu-hardener [09:28] or whatever they channel is called [09:28] neither :) [09:29] #join #ubuntu-hardened [09:30] almost :) [09:30] bah, I was close :p [09:46] hm I should be more careful opening libreoffice docs in a bug where it notices it crashes xserver [10:17] seb128: mpt_: http://ubuntuone.com/0b8BFGrg26HdZtIHzHXeUm [10:17] I don't know how to fix the slight misalignment of the uncategorised entries right now [10:17] that's pre-existing [10:18] and actually you can't tell on the device because they're not on the screen at the same time :P [10:19] Laney, looks ok to me, I still think the margins are a bit much and that it would look nicer with 4 icons ... but we probably need a better screen resolution and that's going to be solved ;-) === mpt_ is now known as mpt [10:20] Laney, agreed with seb128 [10:20] Either 4 per row, or larger icons [10:20] sigh [10:21] But yes, it does look better [10:22] * mpt guesses Laney is a wee bit tired of tweaking margins [10:23] david cameron is brightening this work up [10:23] * Laney coughs [10:26] * Laney tries one last time [10:26] With some nice red rectangles to show you how big the items really are [10:53] desktoppers: what handles non free driver installs in ubuntu these days? [10:56] Riddell, https://launchpad.net/ubuntu/+source/ubuntu-drivers-common/ [10:57] Riddell, the frontend part in Unity is in software-properties-gtk [11:06] seb128: hmm interesting, thanks [11:07] yw === MacSlow is now known as MacSlow|lunch [11:27] huh, I got four on the screen [11:28] seb128: mpt: http://ubuntuone.com/5Alpd5ysmYQsWDu7Uoy9Mw [11:28] I hope you can see past the rectangles ;-) [11:30] Laney, I can see only rectangles :p [11:30] heh [11:30] Laney, but looks like it should be fine to me [11:30] ;-) [11:32] mhr3: hi, what's up? [11:33] mdeslaur, hey, read the hardened backlog I guess, ssl question [11:33] mdeslaur, morning, already pinged chris in #ubuntu-hardened [11:34] mhr3: ah, cool [12:17] Laney, do you know if GPS was descoped for v1? [12:18] seb128: pretty sure [12:19] Laney, btw I see that mpt just edited the design this morning, https://wiki.ubuntu.com/TimeAndDate?action=diff&rev2=78&rev1=77 [12:19] yes I pinged him to prompt that [12:19] did you see the MP? [12:19] oh ok [12:19] yes, I was just surprised that you said the design was already covering that [12:19] but I didn't see it the other day [12:19] and I saw it was updated ;-) [12:20] yeah [12:20] don't know when it was added [12:20] https://docs.google.com/a/canonical.com/document/d/1Anaxk4gaIjAJYdgpKICwxRpEinCVprIHl01sOGDaCFs/edit [12:20] location -> defer [12:20] ok [12:20] the indicator's actions weren't hooked up to anything last I checked [12:21] which was when I did the stuff in battery [12:21] pinged ricmm and he told me to check with tvoss who I haven't seen online since [12:21] Laney, tvoss is on holidays until next week [12:21] guessed that [12:21] so probably not going to happen [12:21] hmm, my arm board seemed to have hung itself ... if you pinged me in the last 24 hours, please reping. [12:21] Laney, mpt: if we don't have location, I guess we should hide the label at the bottom of https://wiki.ubuntu.com/Power?action=AttachFile&do=get&target=phone-battery-settings.png ? [12:22] seb128, indeed [12:22] mpt, thanks, doing that [12:23] tie it to the visibility of the location checkbox [12:23] then it'll automatically hide even when enabled if you have no support for it [12:23] Laney, right [12:24] Laney, by location you mean GPS right? [12:24] whatever the one in there is [12:25] it's GPS [12:25] which is what the label mentions anyway [12:25] doing that [12:25] maybe it could check for a location backend instead of just gps hardware [12:25] either is fine for now imo [12:26] Laney, well, if we want to handle cases with location without GPS we need to update the text anyway [12:26] let's just do it the simple thing and tide it to the GPS control [12:26] k [12:26] we can be smarter later if needed [12:26] You'll just get "inaccurate" location detection [12:29] Laney, the "display only the line you clicked on as selected" seems a bit inconsistant, since that's not what you get when reopening the panel [12:29] well, I guess that's ok [12:29] yes [12:29] I think that's the idea [12:32] Laney, "property int n: 4" ... is that the number of columns? [12:33] it's the "factor" [12:33] the relative spacing of the columns to each other vs. the margins to the grid [12:33] if that makes sense [12:33] yeah, reading the code comment again, it does [12:33] doesn't really make a difference as the boxes themselves dominate the size really [12:34] but when we get coloured backgrounds it might make it easier to tweak stuff [12:34] so you fixed the size of the item [12:34] I guess that makes sense [12:35] I capped the width of the label [12:35] and set the width of the grid to be the width of the items [12:35] so that it can be horizontally centred [12:35] * seb128 tests build [12:36] let me know [12:36] * Laney lunches [12:36] Laney, enjoy! [13:30] Laney: seems to be working fine, just had to do another upload/merge because I got a power (big-endian) specific ftbfs [13:30] had to create a FFe and such, but seems we're happy now === m_conley_away is now known as m_conley [13:41] rsalveti: yep, thanks for that [13:41] I saw some people complaining about it ;-) [13:41] oh rsalveti is back ;) [13:41] RUN [13:42] rsalveti, HIDE [13:42] oh, Laney is back === MacSlow|lunch is now known as MacSlow [13:42] I replaced myself with a bot [13:42] :) [13:43] Laney, nice try [13:43] Laney, soooo, your grid changes... [13:47] Laney, it looks fine on a phone, but a bit weird on width layout (e.g portrait mode) [13:47] Laney, do you think it would make sense to left align the icons rather than center those? [13:47] Laney, e.g like https://wiki.ubuntu.com/SystemSettings?action=AttachFile&do=view&target=phone-settings-landscape.png [13:48] Laney, well, it's a detail, rotation doesn't even seem to be working atm on my n7 [13:48] Laney, I'm happy to merge the current code and revisit what to do in portrait next cycle [13:48] I think you mean landscape [13:49] Sweetshark_: http://lists.x.org/archives/xorg-devel/2013-October/037996.html -- enjoy [13:49] Not sure, should be easy to change back if we have to though [13:49] Laney, yes, sorry [13:49] one of the original complaints was having different margins on each side [13:50] so ... [13:51] yeah, and it looks good on normal layout [13:51] landscape just means lot of extra spaces on the side [13:51] oh, well, I'm happy enough, acking that one ;-) [13:51] Laney, thanks for the patience on this one! [13:51] np, thanks for feedback [13:51] hum, CI unhappy [13:51] oh, commit message [13:52] let me fix that while approving [13:52] wooops [13:52] Laney, oh, and yeah, maybe a bit less width on the default geometry would be nicer ... do you want to do that before I ack it? [13:53] does it definitely get ignored on the phone? [13:53] yes, we are fullscreen on the devices [13:54] ok, trying 40 [13:55] looks good, pushed [13:57] Laney, launchpad doesn't think you pushed [13:57] oh, wrong branch remembered [13:57] * Sweetshark_ gives mlankhorst a hug! [13:58] done [13:58] Sweetshark_, mlankhorst: seems like we should upload that fix to saucy/sru it [14:00] mlankhorst: I tried to triage down to the exact shape that is causing it, but didnt complete it. Does LibreOffice give invalid shapes to X11 or should Xorg handle that in theory? [14:00] Laney, weird, launchpad doesn't seem to pick those commit ... maybe it's being slow [14:03] seb128: http://paste.ubuntu.com/6183947/ ... [14:03] Laney, https://code.launchpad.net/~laney/ubuntu-system-settings/size-layout-tweaks/+merge/188823 doesn't list it [14:03] lp says updating diff [14:03] wtf [14:04] and the branch page says "Updating branch..." [14:04] I guess something is happening [14:04] ok, just slowness I guess [14:04] let's wait a bit [14:05] seb128: branching got me it [14:05] so you can at least check it that way [14:05] also, did you forget my critical fix mp? [14:05] oh, yeah, sorry [14:05] looking to that one as well [14:07] I noticed the total size is wrong on my nexus 4 too :( [14:07] it shows 2.1 GB [14:10] :-( [14:10] it works fine on desktop and on the n7 [14:10] yep [14:11] Laney, btw my small battery fix from this morning should be good to review (there was no bug after all) [14:13] roger, will look in a minute [14:13] building a test program for this issue [14:16] Sweetshark_: according to the spec it's invalid === Ursinha_ is now known as Ursinha [14:17] seb128: report from bleeding-edge land: don't upgrade to bluez5 :p === desrt_ is now known as desrt [14:17] oh i could have told you that [14:17] things break badly with pulseaudio at least ;D [14:18] desrt, thanks for the headup, I already had that one on my "stay away from" list ;-) [14:18] seb128: well I'll put it in saucy, but hoping for a reply from upstream first [14:18] also: no support in NM for DUN internet connections [14:18] mlankhorst, thanks [14:20] desrt, that sucks :/ [14:20] desrt, bluez4 and 5 not being co-installable is going to make that transition "fun" in any case [14:20] desrt, GNOME is the only desktop that ported their stuff afaik [14:21] prematurely, in my opinion [14:21] but that's sort of what GNOME does [14:21] right [14:21] fedora users suffer for the benefit of the long-term good :p [14:21] which is why I'm glad we decide to stay once cycle behind [14:21] decided [14:21] one [14:22] I'm not sure how GNOME could handle those transition better [14:22] then should have a preview mode, like for wayland [14:22] rather than half land stuff [14:25] hmm [14:25] seb128: well, gnome will work with bluez4, i think [14:25] userdata gets "unknown" type on mako [14:25] at least NM will [14:25] maybe we could just count those anyway [14:25] fedora just decided to go crazy with bluez5 and break all the things [14:25] which, again, is good in the longrun.... [14:25] seb128: would it be too late to grab some fixes from fedora for xserver? [14:26] mlankhorst: no, fix away [14:26] mlankhorst, bug fixes are still fine this week [14:27] desrt, it's not that easy, GNOME ported their code to bluez5 and didn't keep ifdef codepaths for bluez4 [14:27] desrt, e.g https://git.gnome.org/browse/gnome-control-center/commit/?id=ac43647ac01cb990555c7b4ff7b8bf5fb8b17b83 [14:27] neat. [14:28] * desrt hates those kind of #defines [14:28] desrt, of course that's an easy one, others are like https://git.gnome.org/browse/gnome-bluetooth/commit/?id=0b0534924fb5ea781195e116a7dfbf956d16379c [14:28] desrt, or https://git.gnome.org/browse/gnome-bluetooth/commit/?id=1755d1bd2c35443792e157de3fc8d2972849e1ce [14:28] ya [14:29] it's clear this falls on the heads of the release team [14:30] to the future!!! [14:30] i'm sure NM will eventually get caught up === greyback_ is now known as greyback|shops [14:38] attente, hey [14:38] hi seb128 [14:38] attente, how are you? [14:38] i'm good and you? [14:38] good thanks [14:39] attente, so, I've been trying the language setting on the n7, now that lightdm/logind are there [14:39] attente, it seems to still not be working ... do you know offhand what backend is used (I think to remember it's something in accountsservice) and what file on disk should get updated when you pick a language? [14:39] attente, I'm trying to figure out what's happening/not happening [14:40] seb128, i remember it going through accountsservice and accountsservice in the end is running a helper script to set the LANGUAGE environment [14:41] seb128, i can look into it though [14:41] attente, for the user (e.g ~/.pam_environment) or for the system in /etc? [14:41] attente, that would be nice [14:41] should be in the pam_environment iirc [14:41] ok, that doesn't seem to happen for some reason [14:41] but my memory is fuzzy... [14:41] attente, let me know if you need help debugging [14:47] attente, oh, I know! [14:47] hmm? [14:48] attente, the locales are missing from the touch image, since no langpacks are installed [14:48] attente, [14:48] # /usr/share/language-tools/set-language-helper /home/phablet de_DE [14:48] en [14:48] attente, that seems to be the command that accountsservice call (from stracing the service) [14:48] so how is QLocale getting that list? [14:49] I'm not sure, likely from a static list [14:49] but it doesn't mean the locale is installed/generated [14:50] # locale -a [14:50] C [14:50] C.UTF-8 [14:50] POSIX [14:50] en_AG [14:50] etc [14:50] but only en_ locales [14:51] yep [14:51] .pam_environment [14:51] LANGUAGE=fr_FR:en [14:51] it works once language-pack-fr-base is installed [14:51] which does [14:51] Setting up language-pack-fr-base (1:13.10+20130926) ... [14:51] Generating locales... [14:51] fr_BE.UTF-8... done [14:51] ... [14:52] all the locale you need [14:52] :s [14:52] actually make it fr_NL, we've taken over belgium after they took over france [14:52] :> [14:53] attente, I'm going to comment about that on https://bugs.launchpad.net/ubuntu-system-settings/+bug/1214417 [14:53] Launchpad bug 1214417 in ubuntu-system-settings "cannot change language" [Undecided,New] [14:53] attente, btw, https://code.launchpad.net/~attente/gnome-control-center/keyboard-layout-panel/+merge/188251 doesn't build [14:53] seb128, thanks [14:53] attente, doh, ignore me [14:53] locale issue [14:53] attente: yes please, ignore him ^ [14:53] :) [14:53] attente, I'm building/review it [14:54] lol [14:54] * didrocks runs :p [14:54] haha [14:54] sorry for doing live debugging/thoughts dump on the channel ;-) [14:55] seb128, keep an eye on /etc/environment ... its full of mess on touch (including hardcoded locale and language settings) [14:56] ogra_, right, the tools write to ~/.pam_environment which is enough to overwrite the locale stuff [14:56] seb128, is there some easy way to make the panel install the correct langpack? [14:56] yeah, but once we have a working solution we need to drop the mess from the mess [14:56] :) [14:56] attente, no, mostly due to the fact that the system is a ro image in touch [14:56] attente, you can't install anything [14:56] ah, right [14:56] attente, that's a known issue, we need to work out an overlay for those [14:56] or preinstall langpacks [14:57] we have plenty of space ... [14:57] 512M is the upper threshod for the image ... we are at 300+something [14:57] ogra_, we should install some langpacks then... [14:58] attente, it might be nice to find a way to limit the list of languages to the available system locales in that context though [14:58] well, talk to pmgowan ... he is the disk-space-watcher :) [14:58] attente, I'm going to file a bug for that [14:58] seb128, sure [15:05] attente, https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings/+bug/1234236 [15:05] Launchpad bug 1234236 in ubuntu-system-settings (Ubuntu) "The languages list should only lists the ones available on the image" [High,Confirmed] [15:07] seb128, thanks [15:10] mterry, hi [15:11] i have the unity-greeter default layout thing working, can you look at this MP: https://code.launchpad.net/~attente/unity-greeter/entry-selected-signal/+merge/188755 [15:12] good [15:12] it just gives the interface a well-known bus name "com.canonical.UnityGreeter" [15:12] and adds a method and signal [15:12] attente, hello [15:17] attente, we didn't want a well-known name because in some configurations, multiple greeters can exist. We export the unique bus name as a variable for indicators... hold on, let me find it [15:17] attente, UNITY_GREETER_DBUS_NAME [15:17] mterry, ah, sorry, i didn't realize that === greyback|shops is now known as greyback [15:17] i'll use it [15:24] mterry, Are you using unity-greeter-session-broadcast, or is that also a v2 feature now? [15:34] tedg, I added some support in unity8 trunk for listening to the ShowHome signal, because it was useful for QA purposes. But the rest of its integration will land with the split greeter [15:34] tedg, but it's not on the image for example [15:35] mterry, Okay, so the new upstart breaks it, but I think we're good :-) [15:35] tedg, heh [16:19] attente, I'm assuming you tested the new signal and it worked fine for you? [16:24] mterry, yep, everything still works [16:26] attente, awesome [16:29] how do you restart indicator-messages-service ? [16:31] does it auto spawn? [16:32] hmm nvm i guess, sry for the bother [16:40] mlankhorst: ping? [16:42] ah [16:42] I think I see the bug in QStorageInfo::driveType [16:43] the implementation makes my eyes hurt [16:50] https://qt.gitorious.org/qt/qtsystems/source/26ed19484e85fb814d2735fa4a98f1ad4c51199b:src/systeminfo/linux/qstorageinfo_linux.cpp#L256 [16:50] this breaks if you have more than 9 partitions [16:55] attente, this will need a UIFe too I think [16:55] mterry, the default layout in unity-greeter? [16:55] attente, no the keyboard switch in general [16:56] attente, doesn't it change how the indicator looks in the greeter (thus affecting screenshots?) [16:56] mterry, yes, but that's not covered by the FFe? [16:56] attente, I guess it's probably assumed by the FFe, but it means that we should notify ubuntu-doc mailing list [16:57] mterry, ok [16:57] i will file one right now [16:57] attente, maybe don't need to file [16:57] attente, just send email [16:58] attente, https://wiki.ubuntu.com/FreezeExceptionProcess [16:58] attente, under UserInterfaceFreeze [16:58] it has links [16:58] mterry, thanks, i'll email and cc you [17:00] attente, awesome. I'm doing final testing on unity-greeter, then will release once your indicator-keyboard change hits saucy [17:02] mterry, great, thanks! [17:21] uh, modem-manager is ancient :/ [17:22] modemmanager that is [17:23] 0.6.0 while 1.0.0 was released in july [17:25] tjaalton, check with cyphermox if there is a reason [17:26] probably network-manager, which is probably blocked by something else [17:26] oh well [17:28] seb128: Hi Sebastien! [17:29] seb128: I have the impression that the im-config control in language-selector is now a no-op in Ubuntu. If that's the case, I suppose it should be hidden. At the same time, I assume that the im-config control still makes sense in Xubuntu and Lubuntu, and if so we should keep showing it in those derivatives. Your thoughts? [17:29] well, ModemManager was blocked by the fact that KDE stuff wasn't ready for the API change [17:34] tjaalton: ^ [17:34] I think that's fixed now, I just didn't get the time to update MM [17:35] * didrocks waves good evening [17:35] tjaalton: this is also why 0.7.991 is also only in experimental in debian [17:35] GunnarHj, hey, long time we didn't see you online! [17:36] GunnarHj, I'm sorry, I'm not sure if im-config is used or not, I still have a difficult time to wrap my head around that stack and how it works [17:36] seb128: Yeah, mostly due to health issues... [17:36] tjaalton: is there a specific new feature you'd need? [17:36] GunnarHj, happyaron would be a better perso to ask about that [17:37] GunnarHj, he's based in China though, so probably not up at this hour [17:37] GunnarHj, he's (co)maintaining ibus/fctix/im-config [17:37] seb128: Ok, I'll ask him about it. And do some testing as well. [17:37] GunnarHj, thanks [17:38] GunnarHj, could you open a bug as well with what you found out? it might be easier to keep track of the issue/discuss it, especially if not everyone is online at the same time [17:39] seb128: Will do that too, when I'm sure enough that I'm right. [17:39] thanks [17:50] cyphermox: no, i'm just getting bug 1099766 on every login, and was wondering if a new release would help [17:50] Launchpad bug 1099766 in modemmanager (Ubuntu) "modem-manager crashed with SIGABRT in g_assertion_message()" [Medium,Confirmed] https://launchpad.net/bugs/1099766 [17:51] cyphermox: doesn't the package in experimental also require newer n-m? [17:51] tjaalton: nah, you should be fine with the version we have [17:51] I am trying to get around to testing 0.9.8.4 before it's too late [17:51] there's a few nice fixes in there for this release [17:57] cyphermox: the changelog says it breaks n-m (<< 0.9.8.2-1) :) [17:58] ah, this must have been added since I touched it [17:58] I don't see what's likely to have changed to warrant that breaks [17:59] http://ftp-master.metadata.debian.org/changelogs//main/m/modemmanager/modemmanager_0.7.991-1_changelog [17:59] hahah :) [17:59] still seems wrong. :) [18:00] ok :) [18:01] seems very very wrong, actually [18:01] anyway, the laptop is my old one with built-in 3g modem and no sim card attached, and it's trivial to disable from the bios, just wanted to check if the bug got fixed since [18:02] ah [18:03] I guess this is because of http://cgit.freedesktop.org/NetworkManager/NetworkManager/commit/?h=nm-0-9-8&id=c126d97cb6666c714d75dc68d808b34c77a99398 [18:03] (I meant the breaks) [18:03] tjaalton: Gobi modem? [18:05] cyphermox: i guess it's sierra [18:06] ah [18:06] I have an ericsson with no sim, but I don't get that error [18:07] yep, sierra usb modem converter [18:07] 1199:6813 [18:20] attente, so... As the user scrolls down to the bottom of the list, they pass the last user, then hit the non-user items like *guest or the remote logins [18:21] attente, you might want to switch to some system default in such cases, rather than sticking with the default of the last user passed [18:21] mterry, sure, thanks [18:44] attente, OK, ready to push to saucy when indicator-keyboard lands [18:44] mterry, sorry, i'm just testing the last change now [18:44] yup [18:56] mterry, is it better to just push now as the unity-greeter change doesn't depend on indicator-keyboard? [18:56] attente, push unity-greeter? [18:57] yes [18:57] oh [18:57] sorry [18:57] attente, I'd rather wait because if I push now we'll have regressions until the i-k fixes land [18:57] ok [18:57] understood [19:04] tedg, indicator-sound changed names from "soundmenu" to "sound" when it stopped being an indicator library. Along with that change, it seems to have dropped its special support for being in the greeter? (i.e. it shows the "Sound Settings" option)... [19:05] mterry, I'm not sure it had special support for the greeter... [19:05] mterry, But wouldn't it just show volume and settings? [19:05] tedg, yeah it used to not show that option [19:05] Oh, I see. [19:05] tedg, so right now it's not showing up in greeter because of the name change [19:05] tedg, but I don't want to fix that as long as that menu option is there [19:06] mterry, So you're going to hold user's volume control hostage until you get the options you want? You should run for congress! ;-) [19:07] tedg, listen, Mr. Texas... [19:07] mterry, Yeah, it seems to be using the desktop menu in the greeter. [19:07] tedg, is it too late to squeeze in a fix for the option? [19:09] mterry, Looking to see how tricky it'll be. [19:10] mterry, Doesn't look bad, let me give it a hack and see. [19:25] tedg, mterry: some of the indicators have greeter profile I think, should be easy to drop the item from the sound menu for the greeter [19:25] seb128, yeah I'd think so. I don't remember exactly what they check for, but there are several env vars or could even look for what user they are running as... [19:27] mterry, no, it's really profiles, see /usr/share/unity/indicators/com.canonical.indicator.sound [19:27] mterry, the code export different models over dbus [19:28] mterry, so it's basically changing the indicator to not export that item in the desktop_greeter structure [19:30] seb128, oh that's right [19:31] seb128, I remember that now [19:31] mterry, one "similar" example is https://code.launchpad.net/~larsu/indicator-sound/lp1220215/+merge/185681 === MrKeys88^ is now known as MrKeys88 [19:32] tedg, mterry: larsu is off tomorrow (German re-unification day), but I'm sure he can look at that/do the change once he's back [19:32] I'm sure he can! [19:32] I think it's mostly done. [19:32] not sure what we're talking about [19:33] lol [19:33] larsu, GO AWAY [19:33] For some reason I lost all sound indicator on my machine now... [19:33] BUT THIS ONE BUG I NEED TO FIX IT [19:33] tedg: all sound indicators? [19:33] larsu, hiding the 'sound preferences' item from indicator-sound in the greeter profile ... but tedg is on it [19:33] larsu, we don't want a way to spawn g-c-c in the greeter [19:34] seb128: oh! [19:34] oops [19:34] larsu, no worry ;-) [19:34] we have mterry watching for such things [19:35] * larsu waves to mterry [19:37] larsu, hi! Enjoy unifying tomorrow! [19:37] :) [19:37] haha will do! [19:42] larsu, if you're still around: https://code.launchpad.net/~ted/indicator-sound/greeter-no-settings/+merge/188911 [19:42] I think that should do it. [19:46] tedg: it should. Thanks! [19:46] tedg: I don't have time to test this - please let someone else test it before top-approving [19:46] larsu, I think we need mterry for that, he needs to change unity-greeter [19:47] tedg, I can test [19:52] tedg, it works fine, but the volume slider feels tiny, because the label isn't making it wider. Can you force a minimum size? [19:53] mterry, Not from the indicator sound side of things, that'd be ido. [19:53] tedg, well, it works [19:55] tedg, mterry: ah right, idoscalemenuitem should probably request a minimum size [19:56] I'll note it and fix it on Friday [19:59] larsu, Okay, you can do it as long as the MR description is: "Use this one CRAZY tip to make your slider larger." [20:00] * desrt just now notices seb's gtk thread [20:00] lol [20:00] desrt: it's getting harder and harder to convince him to take the new stuff [20:01] * desrt finds himself caring less and less [20:01] * larsu makes a sad face [20:01] honestly, for ubuntu in the LTS cycle, not upgrading gtk is one less headache to worry about [20:02] and the stock deprecation mess in 3.10 sucks [20:02] fair enough [20:02] also: the argument of syncing up with rhel7 on the release of gtk is a reasonable one [20:03] you have two very large companies using the same version of a major library in each of their respective 5-years-stable releases [20:03] there's a good chance for cooperation there [20:17] desrt, I don't think Canonical is "large" when used in the same sentence with Red Hat :-) [20:18] tedg: ya... i thought about that after i hit enter :) [20:18] tedg: oh, c'mon...redhat only has 5900 employees :) [20:19] what's an order of magnitude, between friends? [20:19] I'm sure we all work 10x harder :) [20:19] It's okay because I'm 10x better looking. [20:20] lol [20:20] I thought RHEL releases were more than 5 years, no? [20:20] rhel is supported for like 200 years [20:20] but i think they do a new release every 5 [20:21] huh. maybe more like 2-3, actually. not too much unlike us. [20:21] I think it's release ~2yrs and support for 10. [20:22] it's less regular [20:22] tedg: bugfixes only for 5 years though [20:22] between 5 and 6 was 3.5 [20:22] and it's already been 3 years since 6, and 7 isn't ready yet [20:23] It just blows my mind that there are people running, and supported, RHEL 4 [20:23] walters runs RHEL :) [20:23] (not 4, i think) [20:23] Heh, yeah, hopefully 6. [20:24] always fun to see gnome 2 on his laptop at conferences [20:25] tedg: rhel 4? _rhel 3_ is still supported [20:25] with good old kernel 2.4 [20:25] tedg: yeah, 10 by default, but you can pay for longer https://access.redhat.com/support/policy/updates/errata/ [20:25] 2.4. wow. [20:25] Wow [20:25] desrt: well you know my plan is still use ostree to many-boot [20:26] I missed the start of the troll [20:26] seb128: oh. you're still online. [20:26] desrt, is that a reaction to my "use 3.8 for the lts"? ;-) [20:26] seb128: ya [20:26] seb128: i don't disagree with you, btw [20:26] haha, I guessed so [20:26] desrt, btw you bluez5 story convinced me a bit more I was right today ;-) [20:26] your point that you would share the same release with rhel 7 is actually sort of a good point [20:26] your* [20:27] desrt, if it was not for that I would probably lean more toward the "update the stack libs" [20:27] seb128, you got a "sort of a good point" out of desrt, quick, print and frame for your wall! ;-) [20:27] (exception of the GTK 3.10 deprecations) [20:27] seb128: feel free to update the libs... just nothing that depends on the new bluez [20:27] desrt, GTK 3.10 deprecations are making that hard [20:28] seb128: deprecations are just deprecations.... they're not mandatory [20:28] I don't want to start in user flames about https://git.gnome.org/browse/gtk+/log/?qt=grep&q=Deprecate+and+ignore [20:28] desrt, they are [20:28] desrt, read the "ignore" in those commits [20:28] well, that's another story [20:28] well, that's my issue with GTK 3.10 [20:28] no surprise who committed all of those :( [20:28] yeah :/ [20:29] but I don't want to fight the user slashback of those for the LTS [20:29] man... that's brutal [20:29] breaking backwards compatibility is job security! [20:29] or a security job [20:29] * desrt didn't realise he had done that much damage [20:30] * tedg is still stuck trying to imagine 2023 when RHEL 6 dies [20:30] most of these are of questionable usefulness [20:31] but ya... that's a big chunk of incompatible changes and almost every single one of them is going to make some small group of people angry [20:31] a few of them large groups :/ [20:31] desrt, they are, but you still have a part of the userbase that cares and are going to be angry at you [20:31] desrt, and I'm not sure it's worth the update (added to the other reasons for staying on 3.8) [20:31] no argument [20:33] i have to admit, though, doing periodic housecleaning _is_ sort of important [20:34] I'm fine with it, but after the LTS [20:34] nod [20:34] when it's on the start of a long cycle [20:34] so we have time to deal with the fallouts [20:34] I wish things would be deprecated a cycle before being ignored though [20:35] it's a bit harsh to deprecate and ignore at the same time [20:35] this is upshot from the gtk hackfest, fwiw [20:35] a lot of it is caused by the fact that the only real way to get these settings into gtk is via xsettings and xsettings is going away soon [20:35] (for obvious reasons) [20:36] the LTS will still be based on X, i'm guessing [20:36] well, it didn't go away this cycle [20:36] seb128: true... [20:36] so they could have been "non ignored" this cycle still [20:36] and yeah, even if we were going full energy on wayland, that would be after the LTS [20:36] it would be crazy to land that during the LTS cycle [20:36] you keep saying wayland... [20:36] ;) [20:37] well, anything non-X [20:37] but even if Canonical was not doing Mir, we wouldn't have done wayland for the LTS [20:37] nod [20:37] fwiw, wayland is working (with native wayland clients, no X) in F20 [20:37] with gnome-shell [20:38] it's a bit of a disaster, though... only barely works [20:38] well, it's a preview [20:38] but it's one of those death-by-a-million-cuts thing [20:38] I guess next cycle is going to be near parity [20:38] ya. indeed. [20:38] which is still not a strong argument to switch, especially for LTS [20:38] once they fix up the 50 or so "small problems" [20:38] it's clear that the bulk of the problem is already solved [20:38] which is very nice to see [20:39] yep [20:39] good sign for the next "generation" [20:39] also: this steam thing.... [20:39] valve lit a fire under nvidia's ass [20:39] so ya... the new generation is looking quite good :) [20:39] I'm not sure why [20:40] they could have shipped the nvidia binary drivers no? [20:40] seb128: nvidia's drivers for linux suck because "nobody wants to play games on linux" [20:40] no they don't [20:40] you suddenly get one of the biggest PC game makers saying "linux is our new primary platform" [20:40] their binary drivers are quite decent in performances no? [20:41] yes. they are... but they are a disaster for system integration [20:41] well, I'm not complaining [20:41] and they're X-only [20:41] i'm gonna assume that valve doesn't build a new gaming OS centred around X [20:41] right [20:41] they mostly need GL I guess [20:41] X is a very very complicated way to get a GL surface :) [20:42] yeah [21:35] attente, ping [22:54] bschaefer, hey [22:55] attente, hey, question about k-indicator :) [22:56] is there an easy way to add a languages to it, say from a test? As im running into problems where k-indicator is stealing the ibus shortkey :( [22:56] well it seems to be stealing it... [22:57] bschaefer, you can set the gsettings key directly, but you need to use a test dbus session [22:57] i mean assuming this is for a test [22:57] dang... hmm possibly ibus already has that session started [22:57] i mean [22:57] autopilot* [22:58] attente, and when you add a language that way, does it reflect onto ibus? As it normally would doing the "+"? [22:58] doing the add language your self through k-indicator? [22:59] attente, also, do you know when that dbus call in g-s-d is to be reverted :)? [22:59] (for starting ibus) [22:59] bschaefer, is this about the non-latin layouts problem? [23:00] i found that that problem was not fixed by disabling ibus [23:00] attente, its about g-s-d starting up an ibus-daemon starting up when im-config is set to none [23:00] attente, correct, its a different problem, but im talking about ibus not working because g-s-d is start up its own daemon [23:01] so there's a line in g-s-d [23:01] if one is not opened, which is has the option "--panel disable" which causes thing to not work... [23:02] attente, yeeah, seb128 and I talked about it and i never asked who was going to be doing that reverting :) [23:03] bschaefer, is there a bug report for this? [23:04] hmm im not sure :), I made one about the ibus ap test failing...let me pull that up [23:04] so this line here: https://git.gnome.org/browse/gnome-settings-daemon/tree/plugins/keyboard/gsd-keyboard-manager.c#n989 [23:04] https://bugs.launchpad.net/unity/+bug/1233739 [23:04] Launchpad bug 1233739 in unity (Ubuntu) "IBus Autopilot test failures" [High,Triaged] [23:04] * bschaefer looks [23:05] attente, right, which came in from this bug: https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/1194138 [23:05] if we move it into the condition so that it only executes on an actual ibus IM, it will prevent it autostarting on non-ibus input sources [23:05] Launchpad bug 1194138 in gnome-settings-daemon (Ubuntu) "ibus-daemon doesn't autostart" [Undecided,Fix released] [23:06] attente, that would be good, as upstart handles starting the ibus-daemon fine [23:06] IFF im-config if set to ibus... [23:06] but if im-config is set to xim/none then g-s-d is autostarting the ibus daemon with some options that are causing ibus to not for me at lease... [23:07] the "--panel disable" being the options that cause it to fail for me.... [23:07] from: /usr/share/dbus-1/services/org.freedesktop.IBus.service [23:07] bschaefer, what if we try to add that option into g-s-d when it tries to autostart it? [23:07] can we do that? [23:08] attente, that option is bad, but im also not sure why its there (or the point of it really) [23:08] at lease from what im looking at, if we change the /usr/share/dbus-1/services/org.freedesktop.IBus.service to start ibus with a: [23:09] ibus-daemon --replace --xim [23:09] vs the one that is failing atm: ibus-daemon --replace --xim --panel disable [23:09] attente, but another problem, is the ibus-daemon is always running (which is bad) [23:09] even for people who don't use ibus [23:09] which is why thought it should be reverted... [23:09] * bschaefer digs through lgos [23:09] log* [23:10] bschaefer, you can get it autostarting properly with upstart? [23:10] attente, if ibus is set up correctly... [23:10] i mean [23:10] im-config [23:11] if im-config -n ibus, then all if fine [23:11] the problem is the default is none [23:11] so upstart wont start ibus because its none [23:11] attente, Sep 25 13:28:33 happyaron argued that we should drop the g-s-d dbus activation of ibus [23:11] bschaefer, ok [23:11] attente, which I agree with, (that conversation was a bit ago) [23:12] bschaefer, so you just want me to disable that entirely? [23:12] attente, would be nice :) [23:12] it's fairly simple, i'm just not sure of the consequences of such a change [23:12] attente, well this bug would be a bug again: [23:12] https://bugs.launchpad.net/ubuntu/+source/gnome-settings-daemon/+bug/1194138 [23:12] Launchpad bug 1194138 in gnome-settings-daemon (Ubuntu) "ibus-daemon doesn't autostart" [Undecided,Fix released] [23:12] attente, well we were going to talk about it in Oakland [23:13] we are* [23:13] yep, i know, but it's a bit too late for this release :P [23:13] attente, i was just curious if that reverting was still happening :), we can wait to talk to seb128 tomorrow [23:13] yeah no worries [23:13] attente, its just causing some fun AP test failures which is causing unity to not get released (on top of other AP test failures...) [23:14] attente, well if its to late for a release, which is soon we can just wait until we talk about it in Oakland [23:14] that bug up there, it's a fairly big bug [23:14] that way we can try to get everything fixed at one time, or organize it a bit more [23:14] no, not that it's too late to release [23:15] i meant, that meeting in oakland would be too late for us to discuss this release [23:15] right [23:15] * bschaefer was thinking to late to release at all for this release :) [23:15] attente, well in the meantime ill try to come up with a workaround in the AP tests...as I don't want to cause a real regression [23:15] vs failing tests that don't mean much when it comes to actually using the system... [23:17] though the fact that the ibus-daemon is always running is no fun for users who don't want it on [23:18] so disabling dbus activation in g-s-d, does that help at all? [23:18] yes...but then it'll cause that bug I linked above to come up again :( [23:18] :( [23:19] yeeah, soo hmm we can talk with seb tomorrow possibly theres a work around for that bug without having the ibus-daemon always on... [23:20] yeah.. i didn't realize it was that big a problem :s [23:20] sorry about that [23:20] attente, also what happens when the user selects fcitx? [23:20] no worries, these things are hard to know about, unless you already knew about it :) [23:21] attente, does the k-indicator depends on ibus completely? [23:21] bschaefer, from what i remember, no [23:22] i think i asked this before :), and yeah... hmm I wonder how that bug works out when an IMe is already running... [23:22] cause you can't have the ibus-daemon open IIRC if fctix/gcin/hime is open [23:23] if fcitx is running it's completely independent of the gnome stack i guess [23:23] (although i guess there is some stepping on toes if it's running concurrently with ibus) [23:23] right, but it looks like: [23:23] (indicator-keyboard-service:11152): IBUS-CRITICAL **: ibus_bus_call_sync: assertion 'ibus_bus_is_connected (bus)' failed [23:23] Segmentation fault (core dumped) [23:23] will happen if ibus-daemon is not running [23:23] attente, the thing is ibus-daemon shouldn:t be running when fcitx is running [23:24] which is why it might cause that bug again...say in Kylin which has fctix as the default vs ibus [23:24] attente, i need to test it out though +~ [23:24] dam language switch... [23:24] :) [23:25] heh, ok [23:25] i'll MP a branch that disables the dbus activation and you can test it out i guess? [23:26] attente, alright, but it will cause that problem again :), I just need to test if i have im-config -n fcitx set [23:26] will that bug be present (and also if ibus some how starts up when upstart should have started fcitx) [23:28] so i think the main reason we had ibus start always is because it's required in g-c-c in order to be able to add ibus IMs in the first place [23:28] yeah, which makes sense, though what happens when you add a language in fcitx? I don't think that gets updated the same as ibus [23:29] which is fine, as far as i know ibus is the default for ubuntu? [23:29] ibus is the default for the gnome stack [23:30] there isn't any support in g-s-d or g-c-c for fcitx [23:30] i see, soo that might not play out so well for Kylin ... hmm [23:30] attente, well lets wait tomorrow before you push a branch to at lease talk with seb about it [23:31] i can do this fcitx testing on my own with out any changes [23:31] ok [23:31] i guess worst case scenario is to purge ibus [23:31] right, but i don't want to cause problems for gnome either :) (im just a simple unity/nux/compiz dev!) [23:31] although it's strange to me that i-keyboard crashes without ibus... [23:32] i thought i had handled that.. [23:32] yeah, im thinking it might have to handled that gracefully vs crashing [23:32] sorry, i meant purge ibus locally on your machine for testing :) [23:33] attente, well thats not an easy thing to do, cause my tests need ibus :) [23:33] oh, sorry, i misunderstood [23:33] as im testing ibus works in unity, but with the g-s-d dbus thing...it causes the ibus-daemon to start with odd options [23:33] attente, possibly i am... :) [23:33] right [23:33] i see [23:34] attente, but ideally ibus-daemon shouldn't start if its not configured to start,but seb128 wants to drop im-config off [23:34] and only start ibus-daemon when the local is set to CJK(V)? or something like that... [23:34] but its starting always on my machine which isn't set to that at all... [23:35] attente, but best to wait for some more input :) [23:35] doing it based on locale is not going to work for a lot of people i guess [23:35] yeah... [23:35] i suggested checking the languages in use...but that adds an annoying whitelist that has to be managed [23:36] as the user shouldn't have to do im-config -n ibus, the user should never have to know about that option...