=== Sargun_Screen is now known as Sargun === sikon is now known as lucidfox === jorge is now known as jcastro === Ursinha` is now known as Ursinha === Ursinha is now known as Guest70417 [03:17] Hi, how do I push an new package into Ubuntu precise-updates ? [03:17] Is https://help.ubuntu.com/community/UbuntuBackports ? [03:17] https://wiki.ubuntu.com/UbuntuBackports [03:19] !sru | FourDollars [03:19] FourDollars: Stable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates [03:19] micahg: thanks === Guest70417 is now known as Ursinha === Ursinha is now known as Guest78022 [04:56] Good morning [04:57] morning pitti [06:10] wgrant: ah, yay duplication :) well, it wasn't much lost [06:10] Indeed, there's not a huge amount of code to either of them :) === tkamppeter_ is now known as tkamppeter [07:00] pitti: Hey, how do you unwrite a bugpattern? https://bugs.launchpad.net/ubuntu/+source/chromium-browser/+bug/929219 is accidentally snaffling lots of dbus crashes. [07:00] Launchpad bug 929219 in gwibber (Ubuntu Precise) "chromium-browser, gvfsd-http and others using eglibc crash with SIGSEGV in __nscd_get_mapping() or gethostbyname2_r()" [Undecided,Confirmed] [07:00] RAOF: just change it in lp:~ubuntu-bugcontrol/apport/ubuntu-bugpatterns, the cron job pick it up in the next 15 mins [07:01] RAOF: but I guess in this case it should probably be refined instead of removed? [07:01] pitti: Probably. [07:02] good morning [07:06] pitti: Hm. Probably just switching from “ThreadStacktrace” to “Stacktrace” would do, I think. [07:06] Why can you match on ThreadStacktrace, anyway? That seems hugely prone to false-positives. [07:08] bug patterns don't care -- they have no logic of their own [07:08] they just apply whichever key/value regexp matches you specify [07:59] ogra_: the instructions for the current armhf desktop images make it sound like they'd boot right off USB -- is that really possible now? [07:59] erm, no [08:00] (from http://iso.qa.ubuntu.com/qatracker/milestones/232/builds/22473/testcases) [08:00] the panda ROM can only boot off SD or usbnetwork via the mini USB port [08:00] ogra_: ok, so still "burn .iso on SD card and install onto USB-stick and generate a small boot SD" [08:01] ogra_: merci [08:01] pitti, ah, i think thats just a copy/paste of the x86 instructions [08:01] pitti, sudo dd if=path_to.img of=/dev/your/sd/card bs=4M [08:02] yeah, been there, done that; I just wondered whether the instructions were deliberate [08:02] attach a USB disk as target device to your panda, plug in the SD (and monitor,mouse,kbd) and just boot [08:02] the rest is like any other desktop install (except that we dont run a live session by default) [08:03] and dont remove the SD after install, it becomes your "bootfloppy" [08:04] nice; I'm eager to see the progress since alpha-2, it's been a while since I touched the Panda board [08:04] well, we default to have the 3D driver installed now ... and you run the "real unity" :) [08:05] ogra_, 2 USB ports, ENOTENOUGH for keyboard,mouse,usb stick [08:05] there are still flickering issues and since we switched to 3D by default firefox got massively slow [08:05] Heh. I'll need to try that again. [08:05] seb128, yeah, you need a hub :( [08:05] seb128: You mean your monitor is *not* a usb hub? :) [08:05] RAOF, not only it's not an usb hub, but it also doesn't do hmdi, only vga and dvi... ;-) [08:06] seb128: You may have other problems, then :) [08:06] seb128, you *can* install to SD ... but need to prepare the partitioning in advance and make sure the partitioner of ubiquity doent touch the device at all during install [08:06] *doesnt [08:06] RAOF, I've been using the pandobard on the TV so far, it's my only hdmi capable screen [08:06] seb128, so put that TV in your office and expense a new one ;) [08:06] ogra_, I guess I can as well install with a keyboard and no mouse, or hotswap those :p [08:06] ogra_, ;-) [08:07] oh, yeah, indeed === smb` is now known as smb === henrix_ is now known as henrix [08:18] I have single USB dongle for wireless keyboard&mouse =) Win! =) [08:33] jibel: on the ubiquity screen about partitioning/formatting ("erase/lvm/encrypt/other"), the continue button does not work; did you see this on other platforms as well? (I'm trying current beta-1 armhf) [08:34] pitti: hmmm...? do you have external drive plugged in or are you doing 'pre-partioned' way? [08:34] pitti, it works on intel [08:34] pitti: and ogra was reporting success on panda's with yesterdays beta candidates I think.... [08:35] I tried back and next again, and now it works [08:35] I only plugged in the USB drive when the ubiquity start screen was already on, so perhaps it got confused due to that [08:36] yes it would, cause it locks down environment / blocks device events early. [08:39] do we plan to fix/silence out that annoying fontconfig warnings each time you start an application? === cpg is now known as cpg|away [08:41] tsdgeos: Laney was planning to mass upload / fix all the fonts. [08:41] hahaha [08:42] I think I actually said that I /might/ fix /some/ of them :P [08:44] three cheers for Laney, volunteering to fix all of them. hip, hip. [08:45] HORAY! =) [08:46] *clap* *clap* *clap* [08:46] /quit [08:46] heh [08:46] but really, which ones are there on a default install? [08:46] is it just the ones from l-s? [08:46] but look at all the support you're getting [08:47] yes... I could convert it into delicious Danish beer [08:48] dholbach: we really need developer initiative for bug 1034928 [08:48] Launchpad bug 1034928 in language-selector (Ubuntu) "Fontconfig warning: Having multiple values in isn't supported and may not works as expected" [Undecided,Confirmed] https://launchpad.net/bugs/1034928 [08:49] they annoy everyone and there are too many font-config based fonts with these warnings which spam stderr [08:49] hi [08:49] look another volunteer [08:50] I'm just running ubuntu-support-status on a server for a customer, and I see the kernel is showing up as only 18m support. [08:50] what's going on? o_O [08:50] this is on LTS obviously. [08:50] thanks ogra_ !!! I knew you would step up =) [08:50] erm [08:51] xnox, i promise to fix all fonts that fail on arm after Laney is done ;) [08:51] Laney: no-change rebuilds then? :) [08:52] Nafallo: Oh, that probably needs a seed tweak. What's the ABI version here? [08:52] Nafallo: And do you mean precise? [08:52] cjwatson: precise indeed. ABI 29 :-) [08:53] cjwatson: thanks for confirming my suspicion :-) [08:54] Hmm, right - let me prod at this, the SUPPORTED_HINTS stuff looks rather unloved in precise [08:54] heh [08:54] another thing for the checklists ;-) [08:58] Nafallo: Which package name *exactly*? [08:58] xnox, feel free to add it to https://wiki.ubuntu.com/UbuntuDevelopment/BugFixingInitiative :) [08:58] Supported until October 2013 (18m): [08:58] linux-image-server [08:59] Supported until March 2014 (18m): [08:59] linux-headers-3.2.0-29 linux-headers-3.2.0-29-generic [08:59] linux-image-3.2.0-29-generic [08:59] ^-- cjwatson [08:59] OK, thanks [09:00] I wonder what's going on with those dates [09:00] dholbach: ok. [09:00] they do seem a bit random, don't they? :-) [09:00] :-D [09:00] Because one of "until October 2013 (18m)" and "until March 2014 (18m)" is lying [09:00] Never mind anything else [09:02] So linux-image-server is kind of OK as it is - the supported thing is linux-image-generic-pae, -server is just a transitional package really [09:02] Oh, except this is amd64 isn't it [09:02] linux-image-generic then [09:02] So in that case it probably is just the usual business about bumping for new ABIs [09:07] server and generic are the same these days? [09:07] * ogra_ wonders if the back/forward buttons in the ubiquity slidesho work for anyone else or if thats a pandaboard specific bug [09:07] +w [09:09] Nafallo: I *think* this should be fixed after the precise-updates Packages files are next regenerated (so whenever an SRU is next published there) [09:09] okay, kewl :-) [09:09] I'll keep testing. [09:09] thanks muchly cjwatson === mcclurmc_away is now known as mcclurmc [09:49] ogra_: now I understand why we had had preinstalled images .. this is taking half a day [09:50] pitti, with a decent usb disk or key its about 1h [09:50] (i use a 32G USB 3.0 key here, that finishes in 45min) [09:51] OMAP5 will have SATA ... that should finish in 10min then [09:51] USB 2.0 simply only allows 24M/s max. [09:52] * ogra_ reboots into his finished install [10:14] * Laney joins the panda-fest [10:23] ogra_: does unity actually work for you? I've let the blank wallpaper with a mouse cursor stay around for 10 minutes, and nothing happens [10:24] pitti, yes, works fine, i just reported my bugs on the running desktop [10:40] pitti, did it work now ? [10:41] ogra_: no, I just set it back to my usual mode -- disable lightdm, install openssh-server, and let it sit headless under my desk [10:41] hmpf [10:41] would be good to know what failed and why [10:41] do you know which model your panda is ? [10:42] (there should be a sticker on the bottom) [10:42] "ES Rev B1" [10:42] PCBA: 750-2-2170-002 REV B [10:42] hmm, that shouldnt have any issues [10:42] im currently testin on a B2 but all the ES'es should be fine [10:44] pitti, filing a bug with ~/.xsession-errors and Xog.0.log might be helpful [10:45] okay; will do after I finish packaging pygobject and having lunch [10:53] ev, jodh: you guys still have workitems on https://blueprints.launchpad.net/ubuntu/+spec/desktop-p-graceful-failure that were targetted at beta1, I'm moving them to beta2, could you set them to postponed if you think you will not get to them this cycle? [10:54] seb128: just saw - cheers [10:59] ogra_: oh god, the flickering! [10:59] Laney, fun, isnt it :) [10:59] bug 1045491 [10:59] Launchpad bug 1045491 in pvr-omap4 (Ubuntu Quantal) "Moving mouse messes up the desktop" [High,Confirmed] https://launchpad.net/bugs/1045491 [11:00] cheers [11:00] the mouse seems slightly laggy too, but that could be me [11:00] * ogra_ would more like to know why the server image doesnt give any screen output at all [11:03] gah ! === MacSlow is now known as MacSlow|lunch [11:04] seems the new framebuffer driver actually requires console= to be set [11:48] does it make sense for svn to still be in main? [11:49] That depends on whether we care about our userbase who are developers of things other than free software. [11:49] Or indeed free software in more corporate kinds of settings. [11:50] In fact we use it ourselves in our datacentre (think LP code imports) so we de facto support it anyway. [11:54] cjwatson: right, i'm not saying throw it out of the archive :).. just trying to work out if it still has a place in main. [11:55] cjwatson: Does Launchpad use 'svn' for imports? [11:57] ah, libsvn1 from svn [11:58] Daviey: What I mean is that we don't really gain much as an organisation by pretending not to support something we use ourselves [11:58] I knew you didn't mean to throw it out of the archive [11:59] Is svn much of a maintenance burden? I don't really like the idea of tools our developers use to fetch upstream sources and patches not getting security updates... [12:00] mdeslaur: fair point. [12:00] Quite. I suspect Daviey is asking because it has a pending MIR attached. [12:00] The only reason i raised this was becuas eit has a Recommends showing on C-M.. and wanted a quick health check to see if it is still suitable [12:00] svn2cl doesn't look that complex though. [12:01] (Well, I haven't looked at the code but how much complexity can you realistically fit inside 20KB ...) [12:01] * answers from information theorists not acceptable === doko_ is now known as doko [12:13] should removing /etc/dpkg/dpkg.cfg.d/multiarch be enough in disabling multiarch? doesn't seem to work on a quantal chroot, or I've missed something [12:13] dpkg --remove-architecture i386 [12:13] (dpkg --print-foreign-architectures) [12:14] thanks === _salem is now known as salem_ [12:33] cjwatson, sent another e-mail to u-d, would you (or someone else who's got access to the moderation queue) mind approving it? I've also got another one coming later on [12:34] dpm, you should subscribe to the list ;-) [12:34] Subscription doesn't help with u-d. [12:35] exactly :) [12:35] I've processed your mail. [12:36] cjwatson, what teams have access? I guess ubuntumembers doesn't? [12:36] It's supposed to be ubuntu-dev. [12:36] Laney: fyi, I'm working on that cryptocipher build failure [12:36] ah, excellent [12:36] cjwatson, ok, makes sense, it just feels weird that dpm is out of the set of people who can post there ;-) [12:37] Well, I can always whitelist him. I'll do that now. [12:37] dpm, you should become ubuntu developer, come on #ubuntu-desktop we will find you stuff you can help with ;-) [12:37] cool, thanks seb128 and cjwatson ;) [12:37] seb128, oh, didn't realize the whitelisting came at a price! ;-) [12:38] Whitelisted now. [12:38] step aside seb128.. dpm wants to make the cloud a better place. [12:38] dpm, I was suggesting you the "proper way", whitelist is cheating :p [12:38] hahaha [12:38] I feel pressured now [12:38] Daviey, heh, I was there first, I won fair and square, back off :p [12:39] dang. [12:39] Daviey, oi, you've already got jcastro, you only get one community guy per team! [12:39] or dpm could do desktop during the day and cloud at nights... ;-) === MacSlow|lunch is now known as MacSlow [12:41] seb128, yeah, I think it's a good idea, I've always wondered what to do with my ever growing spare time :P [12:41] dpm, ;-) [12:41] take example on dholbach, he keeps telling me how much he misses doing desktop work! [12:43] seb128, I know, he was just PM'ing me about how he misses the good old desktop days, you should assign them some work items :-) [12:43] *him [12:43] ;-) === henrix is now known as henrix_ [12:52] doko, hi, do you know which version of gcc are supposed to support -fuse-ld=gold ? Precise one does not seem to === henrix_ is now known as henrix [13:02] janimo, none upstream until now. the safe way is to use -B/usr/lib/gold-ld/ [13:02] dobey, thanks === davmor2_ is now known as davmor2 [13:03] sorry :) [13:17] seb128, haha [13:18] seb128, try to work with these guys https://launchpad.net/~canonical-community/+mugshots for a while and see what kind of other work you think of as worthwhile afterwards :-P [13:19] dholbach, ;-) [13:41] Laney: I'm getting ready to test this patch: http://hackage.haskell.org/trac/ghc/attachment/ticket/6156/0001-Fix-for-optimizer-bug-on-linux-powerpc-6156.patch [13:41] Laney: We're definitely hitting this bug: http://hackage.haskell.org/trac/ghc/ticket/6156 [13:41] The minimal test case for -O0 and -O2 shows the failure in the current powerpc ghc [13:46] BenC: groovy [13:46] feel free to upload that if it works [13:46] Excellent, thanks [13:46] but please do a binary debdiff to make sure none of the Provides change [13:46] I will cry if they do [13:48] * soren is slightly amazed that Haskell on PowerPC is an area of anyone's particular concern [13:49] If I were to come up with an example of "niche", it might very well involve the words "Haskell" and "PowerPC" :) [13:49] amazingly, there are actually at least *two* people that care! [13:50] jibel: don't want to disturb #ubuntu-release. Debug shows 16.9 TB drive.... is that really 16.9TB drive? === Guest78022 is now known as Ursinha === Ursinha is now known as Guest61980 [13:50] Where are X keymaps located. I've modified a map in /usr/share/X11/xkb/symbols/ but when I set the same map and the same variant that I modified, nothing changes. [13:50] jibel: partman might not handle that well. [13:51] xnox, well, I'd like to allocate 16TB to each VM I use, but no it's 16GB [13:51] jibel: ok. I wonder if there is magnitude error somewhere. [13:52] xnox, output of sfdisk http://paste.ubuntu.com/1185781/ [14:03] slangasek, hey, do you know if qemu-linary should build the spice binary on 32bits in quantal? [14:14] xnox, especially for you I attached the error message in russian. as I said in the comment, the key seems to be to start the live session in a non-english language [14:15] jibel: locale which has commas as the decimal separator and hence we get shell errors....?! [14:26] mterry: ping [14:26] zul, hello [14:26] mterry: can you have a look at the MIR for requests please [14:26] zul, yes, on my list. I'll look at it now [14:26] mterry: cool thanks [14:33] jibel: hmm.. no crash in the current test. but i did install instead of live-session -> install. will redo again now. === jorge_ is now known as jcastro [14:40] hey [14:40] I have a question about udisks and udisks2 being on our default desktop ISOs [14:40] udisks is being pulled by checkbox [14:40] and apparently nothing else on the CD needs it [14:40] so here's the question: [14:40] or questions: [14:41] 1) is it safe/sane to keep both running at the same time? I've observed some differences as compared to precise and while I don't know if they are caused by the two running in tandem it's certainly a possibility [14:42] 2) is there a strong desire to eject udisks from the ISO? If so we could bump the priority of the bug in checkbox and rewrite the relevant parts to use udisks2 [14:42] slangasek, hey, since I know you, do you know who might be the best person to ask about ^ ? [14:49] jono, hey, would you mind telling me who is the best person to talk to, from the platform team, about udisks 1 & 2 being on the ISO [14:50] zyga, pitti used to maintain it in the past, not sure who has taken over [14:50] (or if someone has) [14:50] oh, cool [14:50] pitti, ^^ [14:50] zyga: udisks 1 should disappear; hasn't it yet? [14:50] nope [14:50] urgh [14:50] it's right on the iso [14:50] because of checkbox [14:50] and it's running [14:50] :-) [14:50] so scroll up to see my questions above [14:50] ah, and usb-creator [14:51] pitti - usb-creator also pulls udisks? [14:51] yes, there is a strong desire; udisks1 is unmaintained, and just mirrors what udisks2 is doing [14:51] pitti, IIRC usb-creator was not in rdepends [14:51] it is [14:51] oh, sorry then [14:51] usb-creator-common depends on it [14:51] right usb-creator-common [14:51] bug 1024405 [14:51] ok [14:51] Launchpad bug 1024405 in usb-creator (Ubuntu Quantal) "Port to udisks2" [Medium,Triaged] https://launchpad.net/bugs/1024405 [14:51] so I'm working on a bug to rewrite checkbox to udisks2 [14:52] zyga: nice, thanks! [14:52] https://bugs.launchpad.net/checkbox/+bug/1016035 [14:52] Launchpad bug 1016035 in checkbox "Add udisks2 support to scripts/removable_storage_* scripts" [High,In progress] [14:52] pitti, we need to decide on priority [14:52] if udisks1 will absolutely go away from the iso we'll bump it an get rid of udisks1 from quantal checkbox [14:53] well, it can't go away as long as there are rdepends [14:53] so we need to port checkbox and u-c [14:53] then it'll fall off itself [14:53] right [14:53] is it hard to port between the two? [14:53] the concept is the same [14:53] xnox, the api has changed to some degree [14:53] the API is a bit differently structured [14:53] * xnox has a local port to python3 of the u-c which needs to land [14:53] xnox, it's not just a x/1/2/ in the strings you send [14:53] zyga: =( [14:53] the biggest issue with usb-creator is that it has a full udisks mock implementation for the test suite [14:53] * xnox lol [14:54] ah [14:54] ! [14:54] the actual parts that are being used at runtime are fairly small [14:54] ok [14:54] pitti, one last question, in case both end up on the CD [14:54] pitti, and we run both by default [14:54] pitti, is that safe? [14:54] yes, should be [14:54] pitti, as I've noted in my bug it seems to do less than it did in precise [14:54] (and udisks 1 doesn't run until something talks to it) [14:55] pitti, to be precise, it is not sending the signal on fs mounts anymore [14:55] pitti, ah, good point, it's service activated === dendro-afk is now known as dendrobates [14:55] I still see it in the udisks and gvfs monitors, and gvfs depends on tracking mounts [14:55] what do you see? [14:56] the job FilesystemMounted? [14:56] I don't see that in dbus-monitor anymore [14:56] 16:56:28.416: /org/freedesktop/UDisks2/block_devices/sdb: org.freedesktop.UDisks2.Filesystem: Properties Changed [14:56] MountPoints: /mnt [14:56] udisk2 yeah [14:56] udisk1 nope [14:56] yeah, sure -- since it doesn't run by default [14:57] we have a bug in checkbox where it fails to detect a thumb drive [14:57] unless you actually activate it? [14:57] hmmm [14:57] sure but I look at that attachment: [14:57] https://launchpadlibrarian.net/114785262/device-inserted-quantal.txt [14:57] udisk is running there as checkbox requests it [14:58] but the signal it sends are different from precise where it also reported all fs mounts [14:58] I'm not sure what's going on in the plumbing layer [14:58] who does the mounting actually? [14:58] is that udisks2 or udisks1 or something even lower? [14:58] udisks2 [14:58] but that would explain why there is no signal [14:58] (udisks2 was faster) [14:58] right [14:58] ok [14:58] that explains everything then [14:59] through gvfs, triggered by a gnome-settings-daemon plugin (but that doesn't matter) [14:59] so in quantal udisks is just as capable, just not used by the relevant desktop machinery that requests the mount [14:59] yes, it didn't really change [14:59] ok, then it all makes sense now, thanks pitti [15:00] I'll try to get rid of that dependency :) [15:00] cheers! [15:00] and with that, good night everyone! [15:00] barry: what is it about python3 that stops this working? http://paste.kde.org/543998/ importing pyqt should give you a qstring surely [15:03] Riddell: that might be a better question for ScottK? i don't know the pyqt4 api very well and i wouldn't be surprised if the api were different for python3 [15:07] yes, it's different [15:07] QString going away is a *good* thing, you get plain old python strings by default [15:08] cf. ubiquity r5465 and r5466 [15:08] Mostly the latter [15:09] If you don't have to worry about py2 compat then you can just throw plain python strings around and not worry about it [15:13] Laney: The patch worked, so upload eminent [15:14] seb128: my recollection is that spice only works on 64-bit; hallyn would be able to confirm [15:14] zyga: maybe pitti [15:14] oh, already sorted ;) [15:15] slangasek, some of the redhat people working on gnome-boxes says that's not true anymore and that spice got fixed to work on 32 bits [15:15] slangasek, should work in the version we have [15:16] :) [15:16] slangasek, I will check with hallyn, thanks === Guest61980 is now known as Ursinha [15:18] barry: yeah, just wanted to make sure I haven't come across some fundamental python3 change I didn't know about [15:18] Riddell: not that i know of :) [15:18] It's a change in pyqt, not a change in python3 [15:19] pitti: I want to make a change to apport to fix bug 1028984 and I'm lost in the twisty maze of branches, where do I commit it and how to upload? http://paste.kde.org/544016/ [15:19] Launchpad bug 1028984 in python-qt4 (Ubuntu Quantal) "apport errored when filing bug on test image alpha3" [High,Confirmed] https://launchpad.net/bugs/1028984 [15:19] oh he's gone to bed [15:41] smoser: you uploaded apport, can you tell me what the right way is to make changes to it in bzr and the archive? [15:52] hi, I'm getting certificate errors when I visit wiki.ubuntu.com, anyone know if an rt ticket has been opened for that yet? [15:54] what's the error? the cert looks valid [15:54] Riddell, i can tell you because i did it wrong :) [15:54] The certificate is only valid for the following names: [15:54] *.canonical.com , canonical.com [15:55] jbicha: I'm seeing it issued as a wildcard for *.ubuntu.com [15:55] the upstream branch is a core dev branch, so suerly changes are supposed to be made there before being uploaded. [15:56] https://code.launchpad.net/~ubuntu-core-dev/ubuntu/quantal/apport/ubuntu is the branch, but i'm not exactly sure how pitti wants it handled. bdmurray surely knows. [15:58] TJ-: what if you restart your browser? [15:58] jbicha: very strange! Firefox shows *.ubuntu.com but openssl s_client shows *.canonical.com [15:59] jbicha: And Chromium also shows *.ubuntu.com and I've never used Chromium on that site previously [16:01] jbicha: Could be an issue with front-end SSL load-balancers being out of sync [16:05] ok, filed rt ticket 20372 [16:11] jbicha: TJ-: the certificate there is blancket, i.e. cloud friendly. not all browsers / ssl implementations will accept it. [16:12] but most modern browsers should. [16:12] check which extensions the certificate is using. [16:12] afaik it is not bound to particular IPs for *.ubuntu.com [16:14] ok, I restarted Firefox & it's working now, thanks [16:18] xnox, for sound indicator integration you need mpris enabled in rb, not sure if you disabled that? [16:18] xnox, it's in rhythmbox-plugins [16:19] seb128: having that package uninstalled doesn't help me, does it?! [16:19] xnox, no it doesn't [16:19] ;-) [16:19] seb128: that's what I get for partially upgrading at one point, don't I =) [16:20] xnox, indeed, got what you deserved for using dist-upgrade :p [16:20] seb128: now If you could help me regain overlay-scrollbars that would be wonderful. As I managed to disable them (when glade was crashing) and can't seem to "re-enable" them any more. [16:20] xnox, did you export OVERLAY_SCROLLBAR=0 or something? [16:21] LIBOVERLAY_SCROLLBAR=0 [16:21] nope. plus that won't work anymore since it became gtk module instead of pre-load. [16:21] xnox, otherwise check that overlay-scrollbar and overlay-scrollbar-gtk3 are installed [16:22] yeap have all 5 of them (i.e. including dbgsymbols) [16:22] xnox, the first one has a /etc/X11/Xsession.d/81overlay-scrollbar which does append overlay-scrollbar to the gtk loader list [16:22] xnox, did you hack that to not doit? [16:22] yeah, that file is still there. [16:22] echo $GTK_MODULES? [16:23] $ env | grep GTK_MODULES [16:23] GTK_MODULES=canberra-gtk-module:canberra-gtk-module:overlay-scrollbar [16:23] env | grep -i scrollbar [16:23] seb128: same output, i.e. just in the GTK_MODULES..... [16:23] xnox, gsettings get org.gnome.desktop.interface ubuntu-overlay-scrollbars [16:24] (we have too many way to handle those ;-) [16:25] seb128: was set to false..... [16:25] seb128: set to true, but restarting apps doesn't make difference. I guess i need to restart a session. [16:25] seb128: thanks though =) [16:25] seb128: will see if I have scrollbars tomorrow. [16:25] xnox, you shouldn't... [16:26] seb128, yeah, we should all just move to SuSE ... they just have yast :) [16:26] (wrt too many places to configure) [16:27] haha [16:27] xnox, but yeah, doesn't seem to pick it dynamically... [16:29] xnox: in the future if you dist-upgrade you can just reinstall ubuntu-desktop to get default "stuff" back. [16:30] jcastro, does that bring back recommends? [16:31] jcastro: thanks. Apparently I don't have update-manager installed. No wonder using quantal was so peaceful for me =) [17:00] seb128: I think there's a flag for recommends? [17:00] usually just reinstalling it fixes it up for me, after I shoot myself in the foot anyway... [17:00] jcastro, yeah, things is when recommends quite uninstalled it's hard to flag if it's because you didn't want them or if that was an error or upgrade issue [17:01] quite->get [17:01] jcastro, so they are cases where they will not be reinstalled because they were flagged as something you didn't want [17:02] also I'm assuming that whatever conflict is going on in the archive that led him to a partial upgrade be resolved [17:46] Hi, where can I get a 12.10 netinstall cd image? [17:53]