=== dholbach_ is now known as dholbach [15:01] hello [15:01] * stgraber waves [15:01] * bdmurray waves [15:02] * psurbhi waves [15:02] o/ [15:03] * slangasek waves [15:04] hi [15:04] hi [15:04] #startmeeting [15:04] Meeting started at 10:04. The chair is slangasek. [15:04] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [15:05] [TOPIC] lightning round [15:05] New Topic: lightning round [15:05] $ echo $(shuf -e cjwatson barry doko csurbhi stgraber jhunt mvo ev vorlon bdmurray) [15:05] ev stgraber mvo bdmurray vorlon doko barry jhunt csurbhi cjwatson [15:05] eep, can we come back around to me? Still typing [15:05] ok :-) [15:05] stgraber: [15:05] - Rally all of last week [15:05] - On holiday on Monday [15:05] - Yesterday: Did some ISO testing, email backlog, IPv6 testcase, admin stuff for the Oneiric Linux Containers Sprint and published a blog post on Arkose [15:05] http://www.stgraber.org/2011/07/05/busy-week-for-arkose/ [15:05] - Today: Some more ISO testing and fixes, triaged some LTSP bugs and commited a few fixes, started poking at debsums again. [15:05] - Tomorrow and Friday: Some work on oneconf, dbus proxy for Arkose and if I still have time, some work on firewalling support in Arkose. [15:05] - On holiday all of next week, then back to Canada. [15:05] (done) [15:06] did: [15:06] distro sprint last week [15:06] software-center: webkit/python bridge evaluation (using qt webkit lp:~mvo/software-center/html-experiment), look into recommender systems specifically "collaborative filtering", some webcatalog/oneconf work with didrocks, work on gtk3, remove leaking abstractions, merge the gtk3 branch into trunk(!) not default yet though, add video player widget based on gstreamer in gtk2, trying to do the same for gtk3, utterly fail, use webkit instead; br [15:06] anch merges, upload new version; [15:06] apt: SoC mentoring, debug/fix sha1 breakage in the experimental branch, python-apt SRU for xz support [15:06] app-install-data update for a2 [15:06] (done) [15:06] distro rally last week [15:06] Blocking of apport package installation failures that are the result of a | I wrote a merge-proposal blocking this for english at least [15:06] segfault. Worked with mvo regarding aptdaemon and writing something sensible | https://code.launchpad.net/~brian-murray/apport/package-segfaul/+me [15:06] to history.log. Worked with kernel team regarding tagging of kernel oops bug |rge/65967 [15:06] reports. Clean up of casper bug reports regarding update-initramfs failure | mvo will write some modifications to dpkg and how debconf is called to [15:06] on Live Media. Worked with pitti regarding bugpatterns for pre-Natty |make sure messages are in English [15:06] releases. Discussed bug heat with deryck from the Launchpad team. [15:06] hrm [15:07] let me guess, paste from vim with vsplit? :-) [15:07] cjwatson: winner! [15:07] Blocking of apport package installation failures that are the result of a [15:07] segfault. Worked with mvo regarding aptdaemon and writing something sensible [15:07] to history.log. Worked with kernel team regarding tagging of kernel oops bug [15:07] reports. Clean up of casper bug reports regarding update-initramfs failure [15:07] on Live Media. Worked with pitti regarding bugpatterns for pre-Natty [15:07] releases. Discussed bug heat with deryck from the Launchpad team. [15:07] there that's better [15:07] done [15:08] mvo: python-apt> that's bug 805389? [15:08] Launchpad bug 805389 in python-apt (Ubuntu Lucid) "Support xz compression inside debs" [Medium,Incomplete] https://launchpad.net/bugs/805389 [15:08] mvo: if you didn't notice, RAOF asked for an amended upload [15:09] cjwatson: indeed, I will fix that today [15:09] ta [15:10] slangasek: I think you're next ... [15:10] speaking of writing sensible things to history.log - the latest bug I duped over to debconf about /var/cache/debconf/config.dat being locked shows a "LiveMediaBuild" field (bug #806272) [15:10] Launchpad bug 806272 in debconf (Ubuntu) "package libpam0g 1.1.2-2ubuntu8.3 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1 (dup-of: 792262)" [Undecided,Incomplete] https://launchpad.net/bugs/806272 [15:10] Launchpad bug 792262 in pam (Ubuntu) "package libpam-modules 1.1.1-2ubuntu5 failed to install/upgrade: subprocess new pre-installation script returned error exit status 1" [High,Incomplete] https://launchpad.net/bugs/792262 [15:11] is that new data that apport didn't send before? [15:11] cjwatson: ack :-) [15:11] slangasek: history.log? yes that's new [15:11] slangasek: LiveMediaBuild has been around for quite some time [15:12] bdmurray: ok - so if LiveMediaBuild isn't there in the other reports, they weren't running in a live env [15:12] so I can't pin this bug on ubiquity :-) [15:12] slangasek: that is correct [15:12] * was off Monday, US holiday [15:12] * rally [15:12] * upgraded to oneiric, switched to unity 2d from GNOME classic - started filing bugs [15:12] * some work to help fix a regression in eglibc multiarch handling from Debian [15:12] * cleaning up -dev packages to make multiarch cross-building possible in practice [15:12] (done) [15:12] - not working last Thu and Fri [15:12] - GCC Linaro merge [15:12] - binutils update [15:12] - eglibc update [15:12] - python3.2.1 rc2 upload, including profile/pstats with the now [15:12] finally free license (took just a couple of years ...) [15:12] done [15:13] rally. monday was usa holiday. bug 788514: (gnome-applet-data, gnome-doc-utils, python-gnomekeyring, python-pyatspi). sync request pending for python-oauth (bug 806103). worked on a few others, but they were more problematic, will come back to them. done. [15:13] [15:13] Launchpad bug 788514 in Ubuntu Oneiric "python packages on the CDs not using dh_python2" [Medium,Confirmed] https://launchpad.net/bugs/788514 [15:13] Launchpad bug 806103 in python-oauth (Ubuntu) "Sync python-oauth 1.0.1-3 (main) from Debian unstable (main)" [Wishlist,Confirmed] https://launchpad.net/bugs/806103 [15:13] O'Rally last week. Spent most of last 4 days working on customer [15:13] bug 802661. The immediate problem is fixable by a tweak to a single [15:13] Upstart job, but the invesigations I've performed so far suggest [15:13] we've found the cause for a whole class of bugs (good news!) These [15:13] are not Upstart issues but rather are arguably a limitation of the [15:13] Error: Launchpad bug 802661 could not be found [15:13] way a generic Linux system currently handles loading kernel modules. [15:13] Ended up writing a kernel module to test the customer fix. The bad [15:13] news is that there is no generic fix available right now - looks [15:13] like the full solution requires some modprobe changes (more work to [15:13] do there...) and possibly some new tooling to help users. Playing [15:13] with dh_installinit. Plan for tomorrow - Upstart coding! [15:13] ✔ [15:13] python-profile/pstats> oh, about time [15:13] \o/ [15:14] barry: so python-oauth was still using pycentral on the CDs? Do we know for sure that it's the last one now? (how was it missed before?) [15:14] I'll recheck the alpha2 images, including the kubuntu images [15:14] slangasek: it wasn't actually missed, but accidentally moved to the 'completed' slot. sync'ing from debian should resolve that one [15:15] yeah, I was going to hold that sync until after alpha 2 [15:15] but i also was going to request doko re-run his script after we think we're done :) [15:15] cjwatson: fair enough [15:15] jhunt: "playing with dh_installinit" - fixing it to install both init scripts and upstart jobs? :) [15:15] barry: ah, ok [15:16] psurbhi: [15:16] slangasek: I'm adding some extra checks. [15:16] *) rally last week. [15:16] *) submitted a merge proposal for an initctl equivalent code mountall. [15:16] *) working on the mkinitramfs script - working on the part "when something goes wrong". Rearranged mountall, udev, upstart-udev-bridge to have a execution flow as is in the current initramfs. Seeing an error right now - "directory not empty - while deleting" - working on fixing that. Also again, not seeing the "stopped" for a job but stopping instead. [15:16] (done) [15:16] * holiday on Tuesday [15:16] * carrying on with IPv6 work; mostly finished with netcfg aside from a few tweaks, now blocked on ifupdown 0.7, working on DHCPv6 support there to try to move things forward [15:17] * improved netcfg BOOTIF code some more to allow preseeding netcfg/choose_interface, which lets kickseed use it [15:17] * barry has to remember to poke zul again about python-image-store-proxy :) [15:17] * fixed kickseed parser to avoid backslash damage [15:17] * minor installer fixes for alpha 2 [15:17] -- [15:18] * barry thanks zul! [15:18] * jhunt wonders if we should standardise the "end of report" marker :) [15:18] let people be creative! [15:18] jhunt: let a thousand dones bloom [15:18] heh [15:18] ev: [15:19] - Post-sprint stuff (expenses, survey) [15:19] - PyGI/GTK3 porting work on Ubiquity - getting closer. Started porting [15:19] segmented_bar and hit a double free down the stack, though Neil has just [15:19] given me some workaround suggestions. [15:19] - Started a conversation with Amanda around geoip-selective installation of [15:19] codecs. [15:19] - Chat with Pete on pulling some of the Ubuntu Light work into Wubi (bind [15:19] mounting Documents, Music, etc). [15:19] - Integrated wubi ext3 image builds into live-build. This mainly involved [15:19] tricking the lupin-support versions of the grub scripts to think they were [15:19] operating on a Wubi system. Still need to finish up the cdimage [15:19] integration, but the rest works (building the image, generating the wubildr [15:19] bootloader) [15:19] - Worked with Matthew to nail down the design for the wireless setup page of [15:19] ubiquity. Hacked on that widget some more. Just need to combine multiple [15:19] APs using the same SSID, adding secrets, and integrating it into the [15:19] ubiquity window. [15:19] - Finally got around to the IronPython investigation work requested by [15:19] slangasek. IronPython application and supporting libraries + CefSharp [15:19] bindings + Chromium Embedded Framework comes in at about 28 MB, though I'm [15:19] hoping lzma and shaving down icudt42.dll (unicode support) will save the day [15:19] there. I have a running example [15:19] bdmurray: watch out for my "morse code" report next week! :) [15:19] (http://people.canonical.com/~evand/tmp/ironpython-cefsharp.png) compiled [15:19] down to .NET bytecode and now I'm just trying to create a bundle out of the [15:19] resulting executable and supporting DLLs. I'm thinking I'll use 7-zip's SFX [15:19] header with a custom configuration and a lzma archive. I'll eventually hand [15:19] this off to The Team (design agency) as a test harness for the slideshow [15:19] work they're doing for us. [15:19] - Had a two hour meeting with The Team on the slideshow redesign work. I'll [15:19] need to investigate the best way to have the WebKit view talk back to [15:19] ubiquity as we want to move the finished dialog into the slideshow, and this [15:19] means sending a reboot signal to ubiquity. [15:19] - Reviewed Matthew's crash tracker design. [15:19] - Had a meeting with Jono L to discuss the application developer program. [15:19] - Fighting to keep my desk in Millbank. :-/ [15:19] TODO: [15:19] - Learn Cassandra and talk to James Troup about how to best structure the [15:19] installation success / failure data collection. [15:19] - Build a serial controlled USB switch box using an Arduino board and the [15:19] innards of a manual switch box, in support of the new installer testing [15:19] harness design. Finally get around to expensing ITX hardware with Pete. [15:19] (done) [15:20] ev: this switch box - is that to cut the mains power? [15:20] jhunt: it's to have the test master image a USB drive with the image to test, then have it switch the USB drive over to the slave [15:21] the idea being that no normal person does installs using PXE + NFS [15:21] and we should test the USB path, as that's probably most common [15:21] ev: ok. [15:21] ev: :) [15:21] I had crazy ideas around using USB OTG, but they were crazy [15:21] this is simple(-r) [15:22] ev: lots to read. interessting stuff around the arduino, I love playing with those [15:22] ev: how soon will you know if lzma will save the day? We are effectively blocked from moving forward on getting a contractor if we don't know what we expect to contract him to work in [15:22] ev: where are they putting your desk if not in Millbank? [15:22] I need to remember how to use a soldering iron without starting a fire or setting off the alarms. [15:23] slangasek: I hope to have rough numbers on the IronPython approach by tomorrow's end [15:23] soldering> oh, is that why they want you out of the office ;) [15:23] slangasek: they want to move me to a hot desk / force me to work from home [15:23] ev: it's like riding a bike :) [15:24] which is not easy when I have two laptops, a monitor, and a netbook [15:24] heh [15:24] and a soldering iron [15:24] but I'm not considered valuable because apparently we only need to talk to our own teams, not designers or anything [15:24] [15:25] solution: solder the netbook to the desk, then no one will want it [15:25] ;) [15:25] hahahahahaha [15:25] yes [15:25] will do [15:25] thanks for the sign off on that [15:25] [TOPIC] AOB [15:25] New Topic: AOB [15:25] anything else to discuss before ev gets me fired? [15:25] lol [15:26] if anyone has some experience with recommendation algorithms, I'm all ears (or the google api on that) [15:26] mvo: ACM? [15:26] isn't that the kind of thing they normally publish [15:27] ev: yeah, I was hoping for a shortcut ;) [15:27] mind you, Netflix was willing to pay $1M for theirs: http://www.netflixprize.com/ [15:27] people who enjoyed this algorithm might also enjoy [...] [15:27] lol [15:27] lol [15:27] :) [15:27] if you liked merge sort, you'll love Tim sort! [15:28] tim sort ftw [15:29] sounds like that's everything? :) [15:30] #endmeeting [15:30] Meeting finished at 10:30. [15:30] thanks! [15:30] thanks, all! [15:30] thanks [15:30] thanks [15:30] thanks! [15:30] thanks! [15:31] thanks [17:01] o/ [17:02] ready for the meeting ? [17:04] Okay, I will come to the meeting, too [17:04] o/ [17:05] anyone else or that just a 1-1 ? [17:05] bdmurray, welcome ! [17:05] pedro_, hggdh ? [17:05] * hggdh wakes up [17:05] ~ô~ [17:06] okay, lets start then [17:06] hello [17:06] #startmeeting [17:06] Meeting started at 12:06. The chair is jibel. [17:06] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [17:06] Agenda: [17:06] Previous Actions (all) [17:06] Community Efforts/Testing [17:06] Automated/Systems Testing [17:06] Engineering Team Bug Status [17:06] Other Topics [17:06] Chair Selection [17:06] [TOPIC] Previous Actions (all) [17:06] New Topic: Previous Actions (all) [17:07] nothing I can remember of. [17:08] bdmurray, research of bugs with workaround was done ? [17:08] jibel: yes it was [17:08] thanks. moving on [17:09] [TOPIC] Community Efforts/Testing [17:09] New Topic: Community Efforts/Testing [17:09] We are currently testing Oneiric Alpha2 candidates, and here are the first results [17:09] * Image Coverage : 84.78% (39/46) [17:09] * Mandatory Test Cases : 56.21% (95/169) [17:10] Major issues with Ubuntu (arch != arm ) : [17:10] * bug 781076 [17:10] Launchpad bug 781076 in doc-base (Ubuntu Oneiric) "package doc-base 0.9.5ubuntu2 failed to install/upgrade: Byte order is not compatible at ../../lib/Storable.pm" [Critical,Triaged] https://launchpad.net/bugs/781076 [17:10] * bug 806349 [17:10] Launchpad bug 806349 in ubiquity (Ubuntu Oneiric) "OEM Install fails with - KeyError: "The cache has no package named 'python2.6-minimal'" - without network connection." [High,Fix committed] https://launchpad.net/bugs/806349 [17:10] * bug 791611 [17:10] Launchpad bug 791611 in ltsp (Ubuntu) "LTSP live doesn't work with new Network Manager" [High,Triaged] https://launchpad.net/bugs/791611 [17:10] amd64+mac untested: bug 702283 [17:10] Launchpad bug 702283 in usb-creator (Ubuntu) "usb-creator doesn't create EFI-bootable USB sticks" [Undecided,Confirmed] https://launchpad.net/bugs/702283 [17:10] and CDs are oversized [17:10] EC2 tests ongoing. [17:10] arm: waiting for next set of images [17:11] Edubuntu has critical issues but won't be fixed on time for A2 [17:11] Ubuntu Studio and Mythbuntu are not tested [17:11] Xubuntu has problems with lightdm (session doesn't start and uses wrong settings) [17:11] charlie-tca, do you want to add something about xubuntu ? [17:12] That about covers it. If the user remembers to select a session, it mostly works [17:12] oh, and no live session for xubuntu, either [17:12] That's all from me [17:12] thanks. [17:12] Kubuntu is skipping Alpha 2. [17:13] thanks ScottK [17:13] studio had to re-built today, the image they had was not installable [17:14] There's still time to test the remaining testcases. [17:14] Go to the iso tracker http://iso.qa.ubuntu.com/qatracker/build/all/notcompleted and pick a test. You're help is welcome. [17:15] Next week we'll be testing proprietary graphics drivers [17:16] and the week after it will be 10.04.3 [17:17] [TOPIC] Automated/Systems Testing [17:17] New Topic: Automated/Systems Testing [17:17] patrickmw doesn't seem to be around. [17:17] let skip it then. [17:17] [TOPIC] Engineering Team Bug Status [17:17] New Topic: Engineering Team Bug Status [17:17] bdmurray, pedro_ ? [17:17] so after doing iso testing if you have some free time help to test the Lucid proposed packages? [17:18] the list is not that big bug having a few there [17:18] http://people.canonical.com/~ubuntu-archive/pending-sru.html [17:18] LINK received: http://people.canonical.com/~ubuntu-archive/pending-sru.html [17:18] sorry not the same topic though [17:18] jibel, not much to share this week on the bug status [17:18] mostly catching up from last week and having an extra look to 'iso-testing' tagged bugs [17:19] thanks pedro_ [17:19] bdmurray, any news on the foundation side ? [17:19] I was off Monday and Tuesday but some notable things from the rally are: [17:19] Blocking of apport package installation failures that are the result of a [17:19] segfault [17:20] Worked with kernel team regarding tagging of kernel oops bug [17:20] reports. [17:20] Clean up of casper bug reports regarding update-initramfs failure [17:20] on Live Media. [17:20] Worked with pitti regarding bugpatterns for pre-Natty [17:20] releases. [17:20] done [17:21] thanks [17:21] any question or comment ? [17:22] moving to next topic [17:22] [TOPIC] Other Topics [17:22] New Topic: Other Topics [17:22] any topic you'd like to talk about ? [17:23] jibel: you'd asked about 'test case' bugs did you have a plan for those? [17:23] Yes I do. [17:24] there plan is to review the list and eliminate what is not test case [17:24] this is on the list of bugs bdmurray collected, correct? [17:25] then for what is really a test case, identify what is useful to be automated [17:25] hggdh, yes [17:25] there is something like 1500 test cases we can use to expand the regression testsuite [17:25] can you post the link to the list? [17:26] but before that there is a huge work of manual reviewing. [17:26] its a restricted url right now [17:26] bdmurray, do you have the link ? [17:26] oh, OK [17:26] https://devpad.canonical.com/~brian/descriptions-with-testcase.txt [17:26] and its not updated regularly [17:26] but it could be ;-) [17:27] heh [17:28] OK. so, so far only Canonical can work on this list [17:28] thanks, we'll publish a clean list for review but there's no schedule defined yet. I'll have a closer look and see if we can make something useful out of it. [17:29] okay, I was just curious about it [17:30] anything else ? [17:30] last topic [17:30] [TOPIC] Chair Selection [17:30] New Topic: Chair Selection [17:31] * hggdh slightly raises hand (hoping it will go unnoticed) [17:31] hggdh volunteers \o/ congrats [17:31] darn! [17:31] [ACTION] hggdh to chair next meeting [17:31] ACTION received: hggdh to chair next meeting [17:31] :-) [17:31] 3 [17:32] 2 [17:32] 1 [17:32] #endmeeting [17:32] Meeting finished at 12:32. [17:32] Thanks all for attending! [17:32] jibel: you *could* have waited some few more seconds to accept me as chair ;-) [17:32] jibel: thanks for chairing today! [17:32] thank you jibel [17:32] hggdh: thanks for volunteering to chair the next one! [17:32] thanks jibel! [17:34] hggdh, there was so many hands raised that I barely noticed yours. That's pure chance [17:34] LOL [17:34] * charlie-tca was going to raise his, but got busy right then [17:37] Ursinha, can you join next meeting and present bug status on server side ? [17:37] its every Wednesday at 1700UTC [17:56] jibel: sure [18:15] bdmurray|jibel: how do you intend to mark bugs that are false positives on the 'test case' list? [18:16] * sbeattie re-asks in #ubuntu-quality === oubiwann is now known as oubiwann-lunch === Quintasan_ is now known as Quintasan === yofel_ is now known as yofel === oubiwann-lunch is now known as oubiwann