=== chihchun_afk is now known as chihchun [03:33] trainguards, it seem like silo vivid/0 is in a weird state. [03:33] It is packages/built and landed at the same time. [03:34] trainguards, can we just publish silo 0 then? [03:34] Think the spreadsheet is what is screwed up. [03:37] tedg: yes the spreadsheet is balls. [03:39] robru, Yeah, I thought it was EOL'd here soon, no? [03:40] tedg: yeah we've had a lot of setbacks. I'm hoping it dies so, so soon [03:41] Zombie spreadsheet, it eats brains and just won't die. [03:42] robru, Thanks for pushing it along. [03:47] tedg: I appreciate the recognition, thanks. [03:50] === trainguards: IMAGE 50 DONE (finished: 20141210 03:50) === [03:50] === changelog: http://people.canonical.com/~ogra/touch-image-stats/50.changes === [04:13] trainguards, can we get a rtm silo for line 65 please? [04:13] tedg: rtm 8 [04:14] robru, Great, thanks! [04:14] tedg: you're welcome === chihchun is now known as chihchun_afk [08:25] hi! Is there any way to see the test logs from https://launchpadlibrarian.net/192178675/buildlog_ubuntu-vivid-arm64.libsignon-glib_1.12%2B15.04.20141209-0ubuntu1_FAILEDTOBUILD.txt.gz ? [08:27] mardy, did you try to build it locally on vivid? [08:27] mardy, I was also having a failing test when I did (already with the archive version when trying to rebuild) [08:29] oh, that one is just on arm64? [08:29] seb128: I did with pbuilder too, and it works; in fact, it seems that it's the arm64 build which is failing [08:29] seb128: I found the test errors, and they should now be fixed upstream too [08:30] k [08:49] mardy, builds fine in a porter box, maybe get somebody to retry the build? [08:57] Does anyone know if the system-image importer is enabled back up? [09:18] sil2100: image #50 DONE in the morning for vivid [09:19] but no rtm. [09:19] I'd guess importer is up [09:21] cihelp: does anyone know if krillin builds for ubuntu-rtm will work now? Francis mentioned they're partially enabled, whatever that means [09:22] hey, could I please get a rtm silo for line 53 [09:22] 52 already has vivid silo [09:22] sil2100: I'm currently trying to assess what the current status is [09:22] these are ota-1 fixes [09:23] vila: thanks [09:23] Wellark: o/ [09:23] Wellark: sure [09:23] sil2100: thanks! [09:23] mardy: retrying the arm64 build https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-001/+build/6630068 [09:23] sil2100: robert mentioned something about spreadsheet being broken, not sure in which why and if still [09:23] Wellark: huh, I actually wonder why this wasn't assigned [09:24] Mirv: oh? Did he say what was broken? [09:24] sil2100: np. accidents happen :) [09:25] sil2100: I didn't figure that out [09:29] I suppose Robert also didn't modify the spreadsheet scripts to enable my version number storage for commitlogs ;/ [09:32] davmor2: ping [09:43] mardy: it seems it succeeded with a rebuild! so it's just a flaky test. [09:43] running watch_only [10:24] sil2100: ogra is never on holiday is he? [10:25] davmor2: never! [10:25] But today he has some technicians at home I think [10:25] sil2100: ah so his Internet is down again then ;) [10:26] Not sure if they're related to the internet, but he said he'll be available in some hours [10:26] sil2100: he only pretends to have holidays to make the company not feel bad I'm sure of it :) [10:26] But that guy should really get a holiday next week! [10:27] sil2100: no Christmas we are all off he'll have no one to talk to.....I bet he codes the entire Holiday though :D [10:27] Next week is still a work week, so I thought maybe he should take that off [10:27] Since damn, I don't remember ogra having even one week free this year! [10:27] :O [10:27] He's like a machine [10:29] sil2100: he hasn't he has had national days off and nothing else [10:29] This calls for an intervention [10:29] Let's gather all together and have a serious talk with him [10:29] ;p [10:29] sil2100: why do you think I took out his internet ;) [10:30] sil2100: he needs a stronger Woman in his life that tells him we are going on holiday NOW! [10:47] any news on krillin builds yet? [10:48] trainguards, Mirv hello, may I ask for a silo for #49 (please see my update/comment explaining the change) [10:50] oh yikes, I hadn't realised tachash was down so autopkgtests aren't getting triggered [10:50] I guess that's due to the 1SS move [11:01] vila: any news? [11:03] jibel: Is tachash known to be down at the moment, or might there be a firewall bug? [11:04] cjwatson, yes, and last new I have are "no ETA, it's in DCE/GSA hands" I'll ask CI again about the status [11:04] news* [11:04] jibel: OK, thanks, I couldn't find anything relevant in RT from a simple search but https://rt.admin.canonical.com/Ticket/Display.html?id=76455 is clearly pretty large [11:09] pstolowski: ok! I don't even try to track your silo logic, just do what you do and don't conflict with each other :) [11:12] Mirv, thanks & sorry for any confusion [11:14] sil2100: replied in #phablet [11:37] sil2100, uh oh, I reconfigured vivid silo 3 to remove indicator-network, but it was not removed? [11:38] indicator-power that is [11:39] https://ci-train.ubuntu.com/job/ubuntu-landing-003-0-reconfigure/22/console says "Removing indicator-power from silo." but apparently it did not [11:39] Saviq: uh, hmm, I remember we had problems with that before as well [11:39] Saviq: let me remove it manually just in case [11:41] Saviq: done [11:41] sil2100, thanks [11:48] sil2100, btw, any idea about bug #1400502 [11:48] bug 1400502 in connectivity-api (Ubuntu) "Can't install libconnectivity-qt1-dev on multiarch" [Undecided,Confirmed] https://launchpad.net/bugs/1400502 [11:49] python3-xdg is an Arch: all Depends of indicator-network, that's a dep of libconnectivity-qt1-dev → libconnectivity-qt1 [11:49] which leads to it being uninstallable for the non-default arch as it tries to install e.g python3-xdg:i386 [11:49] I think it's a bit more than that [11:50] You don't want to be going down the indicator-network:target -> unity8:target path either [11:50] cjwatson, shouldn't python3-xdg:i386 be fulfilled by python3-xdg that's Arch: all? [11:50] no, because it has no multi-arch field [11:50] but I think that is a red herring [11:50] or should indicator-network have python3-xdg:any in its Depends (does it make sense for a Depends at all)? [11:50] shouldn't indicator-network be Multi-Arch: foreign instead? [11:51] ok, I'll shut up now [11:51] Wellark, ↑ [11:51] that would cause libconnectivity-qt1's dependency on it to be fulfilled by the native architecture [11:51] cjwatson, sure, if that's the right solution [11:51] * Saviq no gets Multi-Arch fully, yet [11:51] (https://wiki.ubuntu.com/MultiarchSpec#Dependencies_involving_Architecture:_all_packages) [11:52] It depends on the kinds of technical interfaces that indicator-network provides to its dependencies [11:53] If it provides only architecture-independent interfaces - that is, *from the point of view of the things depending on it* it doesn't matter what architecture indicator-network is as long as it's runnable - then it should be Multi-Arch: foreign [11:53] cjwatson, indeed, it's only accessed via DBus afaict [11:54] cjwatson, thanks, that makes sense [11:55] I'm not 100% clear on why that dependency is there since connectivity-api has no matches for "indicator" outside the control file and changelog [11:55] I guess maybe it's artificial in some way? [11:55] strange, i'm being asked for the zope username when trying to bzr branch on the phone [11:56] brendand, you're being hacked ;P [12:09] sil2100, ping [12:10] jgdx: pong [12:10] What's up? [12:10] sil2100, hi, could you give me write acc to the citrain spreadsheet? [12:10] jgdx: sure, I'll give you access to the build jobs too [12:11] sil2100, great, thanks! [12:13] * sil2100 off to the dentist [12:14] is s-jenkins back? [12:25] popey, yeah, but it looks like nusakan doesn't have access to it, so no images still [12:25] but it is back up and building stuffs [12:25] well, i cant get to it over the vpn either [12:25] PING mayura.ubuntu-ci (10.100.0.4) 56(84) bytes of data. [12:25] [12:26] yeah, same from nusakan [12:26] ok [12:26] though ping might be off in the firewall [12:26] popey, i couldn't get to it until i setup the company-wide vpn [12:26] port 8080 works over the VPN [12:26] despite ping not working [12:26] just not from nusakan :) [12:26] right [12:27] somebody should RT that then [12:27] not here it doesn't. [12:27] there is an rt for that ... [12:27] http://s-jenkins.ubuntu-ci:8080/job/music-app-click/ never loads [12:27] should all be in the works already [12:27] popey: wfm, perhaps I'm in different groups [12:27] popey, try setting up the company-wide vpn (that's what did it for me) [12:27] ugh [12:27] ok. [12:27] oh, yeah, I'm using the company-wide VPN [12:28] it's a vast improvement in just about all respects [12:28] popey, yeah, less than ideal for sure to have to go through it, but like cjwatson said it is quite a bit better [12:28] sorry I assumed that's what you meant when you said VPN, not the old thing [12:28] https://wiki.canonical.com/InformationInfrastructure/IS/HowTo/CompanyOpenVPN this thing? [12:28] ogra_: that's as may be but https://rt.admin.canonical.com/Search/Simple.html?q=nusakan doesn't show it? [12:28] popey: yes [12:28] thanks [12:29] cjwatson, https://rt.admin.canonical.com/Ticket/Display.html?id=77276 ... not sure it mentions nusakan explicitly [12:29] ogra_: do you have a ticket number? [12:29] ok, thanks [12:29] it was pointed out on #is that it is the same issue [12:29] ah, it misspells "nusakan" [12:29] hah [12:30] popey: from that ticket it sounds like this stuff is all behind rapid now rather than batuan, so it indeed wouldn't be surprising that the old batuan VPN no longer works for it [12:30] my understanding is that the intent is to kill off the batuan vpn/jumphosting stuff [12:31] (long-term, perhaps, don't know) [12:32] \o/ that works [12:32] thanks cjwatson cwayne_ [12:43] are image builds back up? [12:45] no, firewall problem apparently [12:45] pmcgowan: https://rt.admin.canonical.com/Ticket/Display.html?id=77276 === alan_g is now known as alan_g|a === alan_g|a is now known as alan_g|lunch === chihchun_afk is now known as chihchun [13:19] oSoMoN_, Hi! how do i reproduce bug 1384460 ? [13:19] bug 1384460 in dialer-app (Ubuntu RTM) "Delegate unhandled URL schemes to the system" [High,In progress] https://launchpad.net/bugs/1384460 === chihchun is now known as chihchun_afk [13:24] om26er, the high-level use case is to browse to a page that has a "tel:" link on it (e.g. google search results for a restaurant in your area), tap on it and verify that it opens the dialer-app with the phone number pre-filled [13:24] om26er, if that works, then the bug is fixed :) === chihchun_afk is now known as chihchun [13:28] oSoMoN_, aah, works then [13:28] oSoMoN_, it shouldn't be too difficult to automate, I think. [13:29] om26er, last I checked it wasn’t really possible to write autopilot tests that involved more than one app, or has this changed? [13:30] oSoMoN_, that has been possible for a while, though the tests need to be in an independent project [13:31] om26er, also, note that the browser side of things is already unit-tested, and so is the url-dispatcher side of things, so I would think that adding a telephony-app unit test (if there isn’t one already) should be enough [13:31] s/telephony-app/dialer-app/ [13:34] oSoMoN_, while unittests are great, integration tests have their own worth for these cases. We can go ahead with landing this I guess since we have a manual case added to the TestPlan [13:35] cjwatson, thanks for the ticket, are you monitoring it with them? or should I ping someone [13:36] pmcgowan: I'm not, I just heard about it here [13:36] pmcgowan: but there was literally just an update on #is [13:36] cjwatson, indeed [13:41] om26er, I agree, integration tests are valuable too in this kind of case, but indeed I wouldn’t block on it in this specific instanc [13:41] instance === alan_g|lunch is now known as alan_g [13:50] trainguards: can I get a reconfigure on vivid silo14 please. Added the -gles twin [13:50] pmcgowan,cwayne_: hm, were you expecting the system-image importer to immediately start doing something once the firewall was fixed? [13:51] greyback: o/ [13:51] sil2100: hey! [13:51] cjwatson, no, but i'd expect it to be able to reach s-jenkins, so could get the custom tarballs that already exist [13:51] cwayne_: had they changed? [13:52] greyback: reconfigured :) [13:52] cjwatson, no [13:52] sil2100: thank you kindly good sir [13:53] cwayne_: ok, thanks. I guess that means image building should be functional again now, anyway [13:53] greyback: you are most welcome [13:54] and sounds like autopkgtests should be back up too [13:57] yay [13:57] \o/ [13:57] sil2100, when do we kick a build [13:58] pmcgowan: well, autopkgtests will be back soon, krillin builds not yet [13:58] pmcgowan: I poked CI and they say they're still working on it with top priority... [13:58] sil2100, I just heard krillin builds should work now [13:59] pmcgowan: but I had a talk with jibel and they anyway would need to have a new image during our EU evening, so we still have some time [13:59] fine [13:59] just want to be sure its all working [13:59] pmcgowan: Ursinha in #ci mentioned that the ETA is today, but not sure if it's ok now [14:00] sil2100, so following in #is they fixed the fw configs [14:00] sil2100: pmcgowan, it's being worked on now, we have a series of RTs that are being processed to fix the network glitches that are left [14:00] Ursinha, I am thinking they are fixed but trying to confirm [14:00] cjwatson: i-network does not provide any libraries [14:00] pmcgowan: the CI vanguard team is working on it [14:00] all of the interaction happens over D-Bus [14:01] so I will change the Multi-Arch to foreign then, right? [14:01] Wellark: right, so Multi-Arch: foreign should do the job then [14:01] thanks [14:01] (sometimes interfaces have architecture-dependent elements even if there are no libraries involved, but those are kind of special cases) [14:01] cjwatson: still it's weird that python3-xdg gets into uninstallable state [14:01] any idea what's causing that? [14:01] Wellark: makes complete sense to me [14:01] well, you are the master :) [14:02] Wellark: it's an architecture: all package with no multi-arch field ... [14:02] so see the spec I quoted earlier :) [14:02] Wellark: it really is a red herring though, even if you rearranged that one thing you'd still run into a whole pile of other issues trying to install i-network from a foreign arch [14:02] Ursinha, IS just fixed several issues and I was told we should be able to build krillin now, so what is it they are working on? [14:03] pmcgowan: yesterday other several issues were fixed as well, but that wasn't enough, it's been an interactive process to figure out what's left [14:03] Wellark: the art of fixing multiarch installation issues is generally in finding the right place in the stack to change; it often *isn't* the case that you want to go all the way to the bottom [14:04] (unlike with many single-arch installation bugs) [14:04] pmcgowan: what I know is that one remaining problem to access snakefruit was being worked on, and our team is checking if things are working as they're supposed to [14:04] Ursinha, thats what was just fixed [14:04] you need to think about each point in the stack and work out what makes sense at each step, and if you manage to fix something higher up (as in this case) you can entirely avoid having to deal with messes below that [14:05] pmcgowan: it's probably reasonable to assume that we don't know everything yet :) [14:05] ack. [14:05] pmcgowan: exactly, there's work in our side to ensure if things are working with the unblocked firewall [14:05] ok [14:05] cjwatson: exactly, thanks :) [14:05] Ursinha: would it help to run through a full test build? [14:05] right thats where I was heading [14:06] lets make sure its working [14:06] cjwatson: I wasn't working directly on this, let me check with the ci people that are, they are most likely on it already, one minute [14:06] pmcgowan: we're working on that [14:06] making sure it's working [14:06] right got it [14:07] autopkgtests are being triggered again, but are broken because they can't talk to ftpmaster.internal [14:07] cjwatson: so, this is the current debian/control of i-network: http://bazaar.launchpad.net/~indicator-applet-developers/indicator-network/trunk.14.09/view/head:/debian/control [14:07] jibel: ^- are you folks aware of http://d-jenkins.ubuntu-ci:8080/view/Vivid/view/AutoPkgTest/job/vivid-adt-bowtie/lastBuild/ARCH=amd64,label=adt/console and similar failures? [14:07] cjwatson: now I'm not completely sure what I should modify, sorry :D [14:07] grmbl ... dsl sucks ... [14:08] Wellark: insert "Multi-Arch: foreign" on a line of its own after line 44 [14:08] (or anywhere in that stanza, but I usually put it next to Architecture) [14:08] cjwatson: ack. [14:08] cjwatson: thanks! [14:08] np [14:09] cjwatson, this is new. another FW issue maybe [14:10] jibel: will you chase that or do I need to? I'm not sure which hosts are involved [14:10] cjwatson, I'm on it [14:10] cjwatson: another related question: this is the connectivity-api debian/control [14:10] http://bazaar.launchpad.net/~unity-api-team/connectivity-api/trunk.14.10/view/head:/debian/control [14:11] libconnectivity-qt1-dev pulls in libconnectivity-qt1 whích now Depends on i-network [14:11] which in turn pulls in unity8 (long story) [14:12] cjwatson: would changing the Depends indicator-network to Recommends indicator-network stop the full chain being pulled when installing the dev package ? [14:15] Wellark: depends on the exact context; recommends are installed by default, although they're not installed for example when satisfying build-deps in package builds [14:15] jibel: thanks [14:15] Wellark: I wasn't totally clear why that dependency was there in the first place [14:15] it does seem an odd thing for a library to depend on [14:16] cjwatson, is kicking off a krillin test build something you can do? [14:16] fginther: I can kick off an ubuntu-rtm/14.09-proposed build for all devices it builds [14:17] I don't know if you'd call it a test build [14:17] (ok, sorry, I used that term earlier) [14:17] cjwatson: i-network provides the dbus service that connectivity-qt1 uses at runtime. so in order for the library to work the service (now provided by i-network) must be installed or the api is useless [14:18] Wellark: hm, I grepped for such things and couldn't find it, but maybe the dbus service doesn't contain the substring "indicator" [14:18] we are working on separating i-network and the connectivity-service [14:18] but now for $reasons they are provided by the same process [14:19] robru: hm, your changes haven't been redeployed yet in production? [14:19] fginther: (I won't do so without confirmation that I should, though) [14:19] robru: since I don't see pkgversionlists generated during publish ;/ [14:19] cjwatson: for now, it would be just fine if i-network would not be installed as build-dep when something depends on connectivity-qt1-dev [14:20] robru: it's really breaking everything right now [14:20] Wellark: Recommends would likely achieve that [14:20] robru: commitlogs are crucial for our analysis of the differences between ubuntu and ubuntu-rtm [14:20] cjwatson: ok. will do that for now. the real fix is to separate the indicator and the service into two different packages [14:20] cjwatson: thanks again! :) [14:20] cjwatson, understood. I asked because I thought that it may be just a 'throw away' build. If it actually generates a new build, maybe not a good idea [14:21] fginther: I possibly could do a throwaway cdimage build but I'd rather not confuse things, and we want to test the system-image side of things anyway; the cdimage side is not the side that's been at risk recently [14:22] fginther: I cannot currently think of any part of the image build process other than what I've already tested and confirmed to work that might have been broken by the lab move, though [14:23] The lab is not heavily involved in image building [14:23] cjwatson, ok, do we know that any part is broken? I was asleep and may have missed the part where a krillin build was started and failed [14:24] and we should really cache the tarballs locally on nusakan for the future [14:24] cjwatson, but I'll poke on the rest of the team [14:24] robru: phew, ok, scratch that [14:24] (so that it doesnt fail the build but uses the last good tarball instead) [14:25] fginther: the only thing I know to have been previously broken was the acquisition of custom tarballs from s-jenkins during system-image imports, and that should now be fine [14:25] robru: it just took a while to get the json blob synced [14:25] the importer is running atm ... we should know in a few mins :) [14:25] robru: thanks for getting it deployed! [14:25] ogra_: \o/ [14:25] in that "w3m http://s-jenkins.ubuntu-ci:8080/" works from nusakan [14:25] ogra_: ah yes, good [14:26] cjwatson, ack. I also see the RT to update the firewall for nusakan to s-jenkins should be working... and you confirmed [14:26] genearlly we need somme caching though ... so that this doesnt bite us anymore in the future [14:26] ogra_: last good tarball> from experience with cdimage, that can be a mixed blessing [14:26] cjwatson, well, still better than not being able to build images at all [14:26] ogra_: you then switch to a different problem, where you can fail to notice parts of the system being broken and end up using old image contents when you thought you should be using new ones [14:27] ogra_: my experience has been that that cure is worse than the disease [14:27] tedg I'll let rsalveti to stop looking at it then shall I :) thanks for the comment on the bug though :) [14:28] s/let/tell [14:28] Oh, yeah. It's a bash/dash difference :-( [14:28] I'll patch it. [14:30] tedg: so indicator disappears cause one thing is faster than another so the fact it worked for me at all is pure fluke right? [14:30] davmor2, Yeah, we were trying to pause when that race happens, which it turns out we didn't really do. [14:30] davmor2, Because dash doesn't do the expansion, it only checks once. [14:31] tedg: man that is sucks :( [14:34] yipiie ... after 6 days with wonky DSL they *finally* found the error [14:39] Wow, so on image 50 it seems Pulse is taking 1.7s to start. [14:50] === trainguards: RTM IMAGE 174 DONE (finished: 20141210 14:50) === [14:50] === changelog: http://people.canonical.com/~ogra/touch-image-stats/rtm/174.changes === [14:50] o/ [14:50] \o/ [14:50] YAAY [14:50] Yaaaaay [14:50] YAAAAAAAAAAY [14:50] hooray [14:50] back in business [14:50] oooh [14:50] pmcgowan: ^ builds work fine now [14:51] sil2100: is this the real build or the power build [14:51] and which of the three tarballs I have do you want next? [14:51] ;-) [14:51] asl QA :) [14:51] *ask [14:52] QA: which tarball do you want [14:52] :-P [14:52] john-mcaleely: NONE!!!!!! [14:52] somehow, I guessed that davmor2 [14:52] john-mcaleely: which tarball for what device on what distro? [14:53] davmor2, well, there are three queued for RTM, and I could queue one for vivid [14:53] john-mcaleely: ugh, 3? :o [14:53] since tarballs are cumulative, we could skip two for RTM [14:53] john-mcaleely: do those fix the critical issues for the OTA milestone? [14:53] sil2100, yes [14:53] Daaamn [14:54] :> [14:54] sorry :-) [14:54] Ok, we still have time for one I guess? [14:54] * ogra_` does a jedi wave ... [14:54] there are no issues [14:54] I'll finish up the transition after UE Live is over [14:54] sil2100, davmor2 will need a custom tar run soon [14:55] cwayne_: definitely not :P [14:55] AaaaaAAa [14:55] ha [14:55] sil2100: see this is what happens when you open you mouth about image builds working :P [14:57] Ok then, so I publish ubuntu-keyboard silo now, let's get at least one device tarball in (if possible) and we kick a new image [14:57] jibel: so it looks like I will be testing tarballs for the rest of the day :) [14:57] davmor2, balme imgbot, not sil2100's fault [14:57] *blame too [14:57] ogra_`: why the tick on the tail [14:57] Then I land the transition and we kick the 'promotion candidate' [14:57] davmor2, dsl outages on and off for the last days [14:57] davmor2: you have time now for a device tarball? When would you be done with it btw.? [14:58] ogra_`: no nobody notice the buildbot appart from me and sil2100 but he got excited about it :) === ogra_` is now known as ogra_ [14:58] haha [14:58] I noticed the buildbot ;-) [14:58] * ogra_ too ... since my nick is in the url it pastes i even get an audible ping [14:58] lol [14:59] still no notification on my krillin :( [14:59] me neither [15:00] ogra_: wiat till 16:06 [15:00] yeah, i will give it some time [15:00] it usuall ytakes 10-15min [15:01] ogra_: I thought the push notification system happens at 5 minutes passed the hour may I was reading the crontab wrong though :) [15:02] davmor2: tedg: thanks, will give that a try === plars changed the topic of #ubuntu-ci-eng to: Need a silo? ping trainguards | Need help with something else? ping plars | Train Dashboard: http://bit.ly/1mDv1FS | QA Signoffs: http://bit.ly/1qMAKYd | Known Issues: CI services coming back up, still fixing missing networks. CI/Phone testing disabled. RTM Archive open for ota-1 (and only ota-1 !!) landings. [15:02] no other reason for pulse to take more time to start though [15:02] well, the imgbot might be ahead of it (it also runs at 5min intervals) ... then i was told that i only get the notification once the image has actually downloaded to the phone already [15:02] so that will take some extra time [15:02] unless we have more stuff starting up at the same time [15:03] rsalveti, Yeah, it does seem odd. But we should fix this anyway. [15:03] yeah [15:03] rsalveti, Wonder if it's just the dbus module. [15:03] Does Pulse lazy load them? [15:03] 174 running on my device. nice [15:03] tedg: not actually, it starts when pulse starts itself [15:03] tedg: just blame thin air if all else fails ;) [15:03] but that might slow things down for sure [15:03] as it's yet another module to load [15:04] will migrate the indicator code to use the native calls, should help on that [15:04] Yeah, but we've had it for a while. No reason for it to change in the last few images. [15:04] yeah [15:04] crap, so many issues with latest firefox [15:07] rsalveti, So I'm playing with that code now, do you have a pointer on the native interface? I can do that conversion. [15:08] ogra_: I has notification now [15:08] notification arrived here [15:09] woot [15:09] davmor2, same for me [15:09] yay [15:09] ogra_: told you 5 minutes passed the hour :) [15:09] tedg: get the code for pulse and check src/pulse/ext-stream-restore.h [15:09] bah [15:10] rsalveti, Cool, thanks! [15:10] opening system-settings ... device hangs :( [15:10] argh [15:10] and now the session restarts [15:10] hmmm, maybe I needed to revert system-settings as well [15:10] Ah, wait, I did [15:10] nvm [15:10] note, i'm still on 173 [15:10] AH [15:10] ok [15:11] trying to upgrade [15:11] 2x nvm [15:13]