=== cyphase is now known as Guest44582 === duflu_ is now known as duflu === Zhenech_ is now known as Zhenech [07:11] hi, i'm running unity8-desktop-session-mir on ubuntu 14.10, where can i get the log for unity8? [08:19] penghuan, ~/.cache/upstart/unity8.log [08:20] does anyone know why I got "terminate called after throwing an instance of 'std::length_error'" when i try to a http request using "request->execute( [08:20] bind(&Client::progress_report, this, placeholders::_1))"?? [08:20] I am actually following the SDK scope template. In the past, the same code did not have the problem. [08:21] my code is at bzr branch lp:~liu-xiao-guo/debiantrial/chinaweatherbad [08:22] Saviq, thanks! === Malsasa_ is now known as Malsasa [09:46] bah, unity-dash segfault on current rtm [10:56] seb128, got .crash? === greyback__ is now known as greyback === MacSlow is now known as MacSlow|lunch [12:36] Saviq, yeah, got that [12:36] it worked for once! [12:37] msg=0xb3ff902c "UbuntuClientIntegration: connection to Mir server failed. Check that a Mir server is\nrunning, and the correct socket is being used and is accessible. The shell may have\nrejected the incoming connectio"...) [12:38] that happened after click "back" on the nearby lens, which was taking time to exit, while it was not done yet poping from that one I right edge swiped [12:42] not easily triggerable using those steps though === dandrader is now known as dandrader|afk [12:51] seb128, iiinteresting, wonder what that message actually means (could it time out, too) [12:51] ugh === alan_g is now known as alan_g|lunch === dandrader|afk is now known as dandrader === MacSlow|lunch is now known as MacSlow [13:39] does screen backlight control still work for you guys in vivid? [13:39] desktop, that is [13:40] both, screen and kbd backlight control are gone for me [13:46] mzanetti, doesn't work as well no [13:46] mzanetti, no notifications but I think I notice some dimming happening [13:47] no dimming here [13:47] makes me blind at night [13:47] mzanetti, oh btw, so what's my best solution to the tag nonsense from yesterday? Fix the script? [13:47] and no keyboard backlight. given I'm already blinded by the scree I then fail to find keys :D [13:47] mterry: yeah, fixing the script would be way to go I guess [13:48] mzanetti, Saviq: and we ultimately want to drop any tag with a question mark (any bad tag)? [13:48] mterry, yeah, if possible [13:49] mterry: hmm... not necessarily... there are situations where valid tags have question marks [13:49] although not sure if one runs the script in those situations === karni is now known as karni-lunch === alan_g|lunch is now known as alan_g [14:13] Saviq, mzanetti: OK, how about http://paste.ubuntu.com/8987706/ ? It deletes any bad tag, and avoids the hardcoded list [14:14] mterry, sounds like something I wanted to do soon, too ;) [14:14] looks reasonable [14:15] mterry, it looks much slower though? [14:18] Saviq, it's slower yes, but takes a second or two on my machine if doing nothing [14:18] mterry, I mean on remote branches [14:18] Saviq, oh goodness I bet that is slower [14:18] mterry, it took like 15s to do it on lp:unity8 [14:18] Saviq, I tend to run locally [14:18] mterry, but that doesn't help [14:18] mterry, once you pushed [14:18] mterry, you can't push a tag deletion [14:18] mterry, you need to delete it remotely [14:18] Saviq, well ok this is slower then. But it seems to be doing the minimum necessary... [14:18] At least I don't know enough about bzrlib to know a more optimal way [14:18] mterry, yeah, I wonder if you can do something to cache it or something... === JMulholland_ is now known as JMulholland [14:21] mterry, if you care, try and see what you can improve on the script, I'll try later as well [14:21] Saviq, or maybe just use this logic to expand the hardcoded list [14:22] "caching the hard way" [14:22] mterry, it's a per-branch thing, not sure what can we do there, your approach seems the most flexible and future-proof, I just wonder if it can be made quicker [14:22] especially as I like to run it on all the MP'd branches from time to time [14:22] Saviq, do we know a bzrlib expert? [14:23] mterry, wgrant comes to mind [14:23] at least he could point to someone else hopefully [14:23] yar [14:24] Saviq, asked in #ubuntu-devel [14:25] tx [14:29] Mirv: what are "qtbase-abi-5-3-0" and "qtdeclarative-abi-5-3-0" ? [14:34] dednick, ping [14:34] dandrader: hey [14:34] greyback_: they are imaginary packages provided by libqt5core5a 5.3.0 and libqt5qml5 5.3.0 that get added as dependencies to a package automatically if the package uses private headers/symbols from anywhere qtbase/qtdeclarative. thus they reforce rebuilding all such packages (as included in my Qt landings) when Qt version is updated [14:35] greyback_: so the vivid's Qt 5.3.2 packages provide qtbase-abi-5-3-2 and qtdeclarative-abi-5-3-2 accordingly [14:36] optimally as little packages as humanly possible depends on them / uses private headers [14:36] so I hope eg. webbrowser-app will get rid of those [14:37] Mirv: I'm not sure what imaginary packages are - there are no such packages in the repo though [14:37] dednick, that fixed it http://bazaar.launchpad.net/~unity-team/unity8/shellRotation/revision/1445 [14:37] dednick, do you see any potential problems with it [14:37] dednick, it's the same approach used for the Launcher [14:37] Mirv: my problem is we've a PPA (unity-team/demo-stuff) and installing anything we built recently fails like "qtubuntu-android : Depends: qtbase-abi-5-3-0 but it is not installable" [14:38] greyback_, rebuild it [14:38] Mirv: what do you suggest we try? [14:38] dandrader: ^ you heard the man [14:38] greyback_, you need to rebuild against -5-3-2 that got released into vivid now [14:38] dandrader: yeah, i had in one of my initial branches. dont think there were any troubles. as long as it doesnt break anything else, i dont see any prob. [14:39] Saviq: "apt-cache search qtbase" doesn't list it though [14:39] greyback_, because it's virtual [14:39] is that because it's imaginary? [14:39] greyback_, yes [14:39] ok [14:40] greyback_, ok, I will trigger a rebuild of all the stuff in the PPA then... [14:40] ⟫ apt-cache show libqt5core5a | grep -i Provides [14:40] Provides: qtbase-abi-5-3-2 [14:40] greyback_, ↑ [14:41] Saviq: yep [14:41] just not obvious to me [14:42] greyback_, that's because when we link to private bits, we need more than just "version 5" as there's no ABI guarantees [14:43] Saviq: sure, I understand why. I just find it hard to detect when packages are virtual [14:43] greyback_, and well, yeah, `apt-cache search qtbase-abi-5-3-2` will return the correct package [14:43] greyback_, you were just searching for one that wasn't there any more [14:43] huh [14:44] greyback_, -5-3-0 vs. -5-3-2 [14:44] no I see that. But I didn't notice that "apt-cache search qtbase" actually returned libqtcore5a [14:44] so my bad [14:44] now you know :) [14:45] greyback_: that's why they are imaginary :) ie. they are "packages" that are "Provides:" from the real packages [14:45] greyback_: you need to rebuild qtubuntu [14:46] greyback_: if it's vivid and it was built before Qt 5.3.2 landed finally yesterday to release pocket [14:46] Mirv: yep all clear now. Rebuild being done [14:46] I see, Saviq explained it [14:47] Mirv, btw, where do those come from when building? is it some hook to dh_shlibs? [14:47] or in .symbols in fact? [14:48] couldn't this be solved via .symbols? [14:49] Saviq: it's black magic. yes, dh_shlibdeps/dpkg-shlibdeps using /var/lib/dpkg/info/*.symbols [14:49] Saviq: so this is defined in the symbols of qtbase + qtdeclarative - private symbols are marked as such, and they give the dependency [14:50] it gets even more complicated with rsalveti's true black magic of the gles twin packages on x86. I manage to mess it up a bit every time despite practice. [14:52] so certain GLES symbols are specially marked as depending on normal libs on non-x86, and the twin packages instead on x86 [14:53] hurray whenever Qt gets runtime switching between GL and GLES [14:54] indeed :) [15:07] mterry: https://code.launchpad.net/~mterry/unity8/we-are-at-least-beta/+merge/241615/comments/594728 === karni-lunch is now known as karni [15:08] mzanetti, that doesn't matter [15:08] mzanetti, haha.. hrm [15:08] Saviq: yeah, I know... but we should not use the script on trunk if that's what it does now [15:08] mzanetti, you said you didn't want tags! :) [15:08] :D [15:09] mzanetti, let me test latest version of script [15:09] mterry, already ran it on trunk ;) [15:09] and is fine [15:10] hmm ok :) [15:10] mzanetti, OK I think I updated tags on that branch correctly, using a merge from trunk [15:10] mzanetti, I was using that branch as a guinea pig. I could imagine one of my interim scripts borked it [15:10] mterry: no worries... I just wanted to raise it, in case the script really destroys it [15:10] ok [15:13] mterry: approved [15:14] mzanetti, w00t! we're out of alpha! :) [15:15] mterry: after all we're 8.01 already [15:15] first bugfix release after stable :D [15:15] which I guess is true for the phone === dandrader is now known as dandrader|lunch === alan_g is now known as alan_g|tea === alan_g|tea is now known as alan_g === dandrader is now known as dandrader|bbl [16:46] * vila_ is vila from a laptop setup for testing [16:46] swift list --debug hangs [16:47] meh, damn xchat, wrong channel, sorry for the noise >-} [16:48] lp:~mzanetti/unity8/performanceoverlay: would delete 7.85+14.10.20140428.2-0ubuntu1 [16:48] lp:~mterry/unity8/greeter-profiles: would delete 0.1.16 [16:49] fixed [16:49] the new script? [16:49] or something I should fix in the branch? [16:50] mzanetti, I think just tags that the new script caught [16:53] mzanetti, I'm just uploading the new script [16:54] mzanetti, http://people.canonical.com/~msawicz/unity8/strip-tags.py [16:54] Saviq: so I should delete that tag? [16:54] mzanetti, it's an invalid tag on your branch, so yeah [16:55] I wonder how that came in though, it's not something we got from uity7 [16:55] om26er, can you please run ↑ on lp:~om26er/unity8/show_password_on_label_tap, and any local unity8 checkouts you have (just pass the different branches as arguments) [16:56] done [16:56] om26er, note you need to run the script on them all, pushing won't help to delete tags [17:01] @unity ↑ new strip-tags.py script, one that works for any branch (not unity8-specific any more, thanks mterry!) [17:02] Saviq, what you mean not unity8 specific? [17:02] Cimi, you can use it for any random branch to clean invalid tags from it [17:03] Saviq, how do we know when tags are valid/not? [17:03] Cimi, the invalid ones don't relate to a revision [17:04] Saviq, moving this to CI? [17:04] Cimi, not a good place for it, no [17:04] we've had this discussion and have some ideas [17:05] yeah, doing it manually for every single branch is not ideal too... [17:05] Saviq, maybe just having CI checking for tags [17:05] Cimi, that's the thing, you don't need to do it manually, we should be able to not do it at all once all of us get cleaned up [17:05] it does not have to remove them, just complain [17:05] Cimi, basically, don't worry, I'll tell you [17:05] i like that :P === alan_g is now known as alan_g|EOD === dandrader|bbl is now known as dandrader [21:35] so since gdm doesn't work in 15.04, and lightdm has never worked for me, does anyone have any ideas? My login screen is just black now... [22:17] josharenson: kdm? [22:17] * zbenjamin runs [22:21] josharenson: there is also XDM, Slim,lxdm [22:21] josharenson: ssdm also sounds interesting, its using Qt/QML [22:22] zbenjamin, ill try another dm [22:32] josharenson: you sure that this is not a xserver config/driver problem [22:34] I can get a desktop to show if I run startx (Xorg was already running), however unity doesn't start when I do this [22:35] meh :/