[16:00] <mvo> hello
[16:00]  * barry waves
[16:00] <robbiew> o/
[16:00] <robbiew> #startmeeting
[16:00] <MootBot> Meeting started at 10:00. The chair is robbiew.
[16:00] <MootBot> Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE]
[16:00] <jhunt> hi
[16:01] <mvo> hi
[16:01] <csurbhi1> o/
[16:01] <ev> howdy
[16:02] <robbiew> okey dokey
[16:02] <robbiew> [TOPIC] Lightning Round
[16:02] <MootBot> New Topic:  Lightning Round
[16:02] <robbiew> mvo?
[16:02] <robbiew> joking
[16:02] <mvo> *weeeeeh*
[16:02] <robbiew> barry?
[16:02] <barry> short week due to holidays; sundry computer-janitor bugs fixed, branches merged, releases made; bug 702375 (launchpadlib upgrade to 1.9.7); much debugging of bug 710711 (nm crashes) w/cyphermox.
[16:02] <barry> done
[16:03] <robbiew> does c-j remove old kernels now?
[16:03] <barry> the latter should be fixed, but there's still a bug in the atheros driver some where
[16:04] <barry> robbiew: did it not before?  i didn't change anything around that
[16:05] <robbiew> hmm... I think it does
[16:06] <robbiew> maybe I'm thinking about update-manager
[16:06] <mvo> that does that as well
[16:06] <mvo> well, in release-upgrade mode anyway :)
[16:06] <robbiew> ah
[16:06] <barry> right, normally u-m skips the kernel i think
[16:06] <mvo> yep
[16:06] <robbiew> no worries...just wondering how to best remove the old kernel bloat
[16:06] <robbiew> and forgot about c-j :/
[16:06] <robbiew> heh
[16:07] <barry> cli: computer-janitor clean --all
[16:07] <barry> mvo: we *really* need to get the computerjanitor package out of update-manager-core :)
[16:08] <barry> i keep thinking about that every cycle when i touch c-j :)
[16:08] <mvo> barry: much agreed
[16:08] <barry> but it's never very high on the priority list :)
[16:08] <mvo> same here
[16:08] <barry> anyway... done.
[16:09] <robbiew> cool
[16:09] <robbiew> ev?
[16:09] <ev> Meeting with ivanka to discuss the package and /home preservation stuff.
[16:09] <ev> Rebuilding the automatic partitioning page thanks to some lovely design work by Daniel Fore and UI/copy from mpt (https://docs.google.com/View?id=dfkkjjcj_101gnkrpg5v&pli=1#4_5_1_Automatic_partitioning_o and http://people.canonical.com/~evand/tmp/partitioner-redesign-progress.png - I will fix the os-prober madness).
[16:09] <ev> Brief discussion with mvo about his wonderful apt-clone branch that will form the basis of the package preservation.
[16:09] <ev> Will most certainly be asking for a FF exception for at least the confirmation page for the package preservation.
[16:09] <ev> Discussion with pitti on jockey from a chroot, he amazingly came up with a dbus-free code path in jockey which I'll finish incorporating into the installer today to knock off that nvidia item.
[16:09] <ev> (done)
[16:10] <ev> incidentally, forrst is very cool: http://people.canonical.com/~evand/tmp/forrst.html
[16:10] <ev> (think tumblr for code snippets and design
[16:10] <ev> )
[16:11] <mvo> screenshot looks very nice!
[16:11] <barry> ev: that installer page looks great
[16:11] <csurbhi1> ev: looks really cool
[16:11] <robbiew> nice screenshots
[16:11] <ev> thanks!
[16:11] <ev> still a bit of work to do, but I think we've finally nailed that first page
[16:12] <ev> it's very explicit about what your options are and what it's going to do
[16:12] <robbiew> an \o/ on the jockey stuff too
[16:12] <ev> pitti is a machine
[16:12] <robbiew> thnx ev
[16:12] <robbiew> doko?
[16:12] <ev> sure thing
[16:12] <doko> bootstrap work, packaging of new libs, gnat bootstrap (not yet sucessful)
[16:12] <doko> python3.2 final packages
[16:13] <doko> add another vm to openjdk-6 (jamvm)
[16:13] <doko> done
[16:13] <cjwatson> ev: I love that partitioner design
[16:13] <ev> cjwatson: that warms the cockles of my heart
[16:13] <cjwatson> heh
[16:13] <barry> doko: did you see my email about python2.7 in unstable?
[16:14] <ev> cjwatson: it's mostly functional to boot
[16:14] <doko> barry: not touching 2.7 in unstable. what about the robust-packaging for natty?
[16:15] <barry> heh
[16:16] <robbiew> jhunt left...doh!
[16:16] <robbiew> cjwatson?
[16:17] <cjwatson> done: most of week on bootstrap project again, now into stage2; deferred foundations-n-cd-boot :-(; started on biosdevname packaging, thinking about how to arrange for it to be selectable but off by default; he
[16:17] <cjwatson> lping out with upstart merge
[16:17] <cjwatson> todo: finish biosdevname packaging before FF; review and sponsor new upstart; beat a few more late merges into shape, e.g. localechooser
[16:17] <jhunt_> robbiew: sorry - it happened again; my kernel seems to take umbrage when I remove my headset! :(
[16:18] <robbiew> jhunt_: wow...open a bug ;)
[16:18] <jhunt_> will do.
[16:18] <robbiew> mvo, ready?
[16:18] <mvo> yep
[16:18] <robbiew> cool...go!
[16:18] <mvo> gtimelog report feature broken for me, so report is from (unreliable) memory
[16:18] <mvo> misc:review/merge lp:~pitti/software-properties/pygi; upload apt-btrfs-snapshot package with automatic snapshots and control cli (which is pretty cool IMO), add support to the release upgrade for it; work on apt-clone for packageselection-foundations-n-update-manager-improvements;
[16:18] <mvo> apt: merge from debian and upload new ubuntu version (good stuff in here, including the new SetCandidateRelease feature)
[16:18] <mvo> python-apt: add bindings for new SetCandidateRelease() feature;
[16:18] <robbiew> :P
[16:18] <mvo> auto-upgrade-tester: add 2.6 -> 2.7 upgrade import failures detection script/improvements; upload a bunch of python2.7 issues, report some bugs; prepare deployment in data-center (pending bigger hdd)
[16:18] <mvo> software-center: get submit-usefullness fully working; get offline-mode working;
[16:18] <mvo> branch review/merges
[16:19] <csurbhi1> mvo, what is the "control cli" about?
[16:19] <mvo> csurbhi1: list available snapshots, revert to a previous one, delete
[16:19]  * cjwatson wonders if we should install apt-btrfs-snapshot by default when / is on btrfs
[16:19] <mvo> csurbhi1: and manually snapshots
[16:19] <mvo> (which is not that useful I guess :)
[16:20] <csurbhi1> mvo, thats cool :)
[16:20] <mvo> cjwatson: that is a good idea
[16:20] <mvo> one prolbem is that dpkg is currently pretty slow on btrfs, I imagine because of the explicit syncs()/fsyncs() we do
[16:20] <mvo> but otherwise its working pretty well for me
[16:20] <csurbhi1> mvo, why is there a need of sync?
[16:21] <cjwatson> sorrow and doom
[16:21] <mvo> csurbhi1: cjwatson is in a better position to tell about this
[16:21] <mvo> csurbhi1: its to prevent data loss on ext4 under certain circumstances
[16:21] <cjwatson> it's mostly spelled "e" "x" "t" "4"
[16:21] <cjwatson> er, yes
[16:21]  * mvo hopes that is a accurate description
[16:22] <csurbhi1> but why is that related to btrfs -snapshots?
[16:22] <cjwatson> dpkg can't go round detecting the filesystem all the time so it has to be safe
[16:22]  * mvo also notes that his spelling is not the best today
[16:22] <cjwatson> it's not fsync any more though, it's sync_file_range, afaik
[16:22] <csurbhi1> aah, i see :) thanks!
[16:22] <cjwatson> I think you would want to profile before deciding that was the problem though ...
[16:22] <cjwatson> (I haven't)
[16:22] <csurbhi1> mvo, if takes too long, then its worth filing a bug
[16:22] <csurbhi1> i mean fsync
[16:23] <robbiew> maybe we need a dpkg-ext4 :P
[16:23] <robbiew> to handle that filesystems...um, unique features
[16:23] <mvo> heh :)
[16:23] <robbiew> mvo: anything else?
[16:23] <mvo> (done)
[16:23] <mvo> no
[16:23] <robbiew> thnx
[16:23] <robbiew> jhunt_: ?
[16:24] <jhunt_> Worked on check-config upstart facility which builds on visualisation
[16:24] <jhunt_> work - not complete yet). Spent most of the time (with lots of help from
[16:24] <jhunt_> cjwatson!) getting keybuks branches to pass the "make check" phase.
[16:24] <jhunt_> Result of which is that we now have lp:~upstart-devel/upstart/0.9.
[16:24] <jhunt_> Currently working on ubuntu merge. Advised robbiew+skaet that unlikely
[16:24] <jhunt_> to have this ready for FF, so we're now plumping to land a featureset
[16:24] <jhunt_> for Monday.
[16:24] <jhunt_> EOT.
[16:25] <robbiew> coolio
[16:25] <robbiew> [TOPIC] Natty Development
[16:25] <MootBot> New Topic:  Natty Development
[16:25] <csurbhi1> :-/
[16:25] <robbiew> Feature freeze tomorrow
[16:25] <csurbhi1> i am still there
[16:25] <csurbhi1> :D
[16:25] <robbiew> lol
[16:25] <robbiew> sorry
[16:25] <robbiew> [TOPIC] Lightning Round - part 2
[16:25] <MootBot> New Topic:  Lightning Round - part 2
[16:25] <robbiew> csurbhi1: ?
[16:25] <csurbhi1> Got a short update: This whole week worked on making initramfs event based. Here is a small development:
[16:25] <csurbhi1> *) Thanks to cjwatson, finally got mountall - upstart communication in initramfs sorted. could get root "ro" mounted on initramfs::/root through mountall (using udevd)
[16:25] <csurbhi1> *) Now working on creating upstart-udev based jobs for handling md arrays, crypt devices.

[16:25] <csurbhi1> :)
[16:26]  * robbiew needs csurbhi1 to stick to one nick...so he doesn't forget ;)
[16:26]  * csurbhi1 notes! :-/
[16:26] <robbiew> \o/ on mountall - upstart communication in initramfs
[16:26] <robbiew> thnx csurbhi1
[16:26] <robbiew> [TOPIC] Natty Development
[16:26] <MootBot> New Topic:  Natty Development
[16:27] <robbiew> who has FFEs coming up?
[16:27] <robbiew> that I should be aware of
[16:27] <ev> o/
[16:27] <mvo> so … I have e.g. submit_usefulness already in natty
[16:27] <mvo> but the server is not fixed yet to actually expose it
[16:27] <cjwatson> csurbhi1: oh cool, what was the bug?
[16:27] <mvo> will that require a FFe?
[16:28] <mvo> technically the feature is in natty, its just the deployment that is lacking behind
[16:28] <csurbhi1> cjwatson, it was the missing libnih_* libs missing
[16:28] <robbiew> mvo: hmm...good question
[16:28] <csurbhi1> i got a trace of init (by making a upstart job that gets the strace -p 1)
[16:28] <barry> i will hopefully verify launchpadlib 1.9.7 today and get a sponsor to upload it
[16:28] <robbiew> barry: ack
[16:28] <robbiew> ev: ack..already discussed
[16:28] <mvo> barry: I'm patch pilot today
[16:28] <ev> right-o
[16:28] <robbiew> mvo: I guess technically, the answer is "no"
[16:29] <mvo> \o/
[16:29] <robbiew> because you're not adding code TO the release
[16:29] <robbiew> ...but
[16:29] <barry> mvo: excellent.  i'll run leonardr's verification script after my natty vm finishes updating
[16:29] <mvo> hopefully it sorts itself really quickly as they promised me a updated deployment for tomorrow or friday
[16:30] <robbiew> ok
[16:30] <robbiew> mvo: might be worth letting skaet know what users can expect to see once the server side is complete
[16:30] <barry> mvo: yep.  i had too many problems backporting to maverick (leonard doesn't have natty), so i gave up on that.
[16:30] <robbiew> in terms of feature highlights for Alpha 3
[16:31] <mvo> robbiew: good point
[16:31] <robbiew> heh...I sometimes have them ;)
[16:31] <cjwatson> csurbhi1: ah, right, makes sense
[16:31] <csurbhi1> :)
[16:32] <cjwatson> csurbhi1: are you using copy_exec in yoour initramfs hook now?
[16:32] <cjwatson> *your
[16:32] <csurbhi1> cjwatson, yes
[16:32] <csurbhi1> but ldd on upstart does not say anything about libnih*
[16:32] <cjwatson> that's odd, it does here
[16:32] <cjwatson> $ ldd /sbin/init | head -n2
[16:32] <cjwatson>         libnih.so.1 => /lib/libnih.so.1 (0x00303000)
[16:32] <cjwatson>         libnih-dbus.so.1 => /lib/libnih-dbus.so.1 (0x00f3b000)
[16:33] <csurbhi1> oh i meant actually the libnsss
[16:33] <csurbhi1> not libnih (err)
[16:33] <csurbhi1> copied the libnss*
[16:33] <cjwatson> ah, ok
[16:33] <cjwatson> yeah, that's dynamically loaded
[16:33] <cjwatson> ok, and that was what we discussed last time - good
[16:33] <csurbhi1> yes
[16:33] <csurbhi1> :)
[16:34] <csurbhi1> by going through the code, i somehow thought that it was creating its own db (rather than looking at any standard system db)
[16:35] <csurbhi1> and fixing this one, fixed the missing sendmsg() and rcvmsg() returning a 0
[16:36] <robbiew> [TOPIC] AOB/GoodNews?
[16:36] <MootBot> New Topic:  AOB/GoodNews?
[16:37] <ev> Worked on generating an application bootchart (functions/blocks of code rather than processes) similar to the ones created by the DX team over the weekend, nearly have something useful for ubiquity.  Very simple stuff, but pretty and hopefully valuable.
[16:37] <barry> 2 weeks to pycon :)
[16:37] <ev> yay pycon
[16:37] <robbiew> ev: suhweet!
[16:37] <ev> so excited to be going for the full thing
[16:37] <doko> barry: language summit on Thu, or Fri?
[16:37] <barry> thursday i believe
[16:37] <barry> march 10
[16:38] <barry> there's a vm summit on wed march 9, but i'm not attending that
[16:38] <mvo> some prototype work with stgraber on weblive integration and s-c http://people.canonical.com/~mvo/tmp/weblive-and-software-center.png
[16:38] <doko> me neither
[16:38] <robbiew> mvo: VERY cool stuff
[16:38] <mvo> ev: application bootchart? will that be useful for softare-center as well ?
[16:40] <ev> ugh
[16:40] <ev> dear Pidgin, die.  No love, Evan
[16:40] <ev> mvo: possibly? This is what it looks like: https://wiki.canonical.com/DesktopExperienceTeam/UnityPerformanceAnalysis?action=AttachFile&do=view&target=unity-bootlog-20100319-1.svg
[16:40] <ev> all you need to do is feed the small python application in unity trunk a csv file of names and times and it goes to town on it
[16:40] <mvo> cool
[16:41] <ev> label,start(float seconds),stop(float seconds) - I believe
[16:41] <mvo> I have a "with ExecutionTime()" thing in s-c already to do the measurements easily
[16:41] <mvo> that looks like a good match
[16:41] <ev> nice
[16:42] <ev> weblive> that looks awesome!
[16:43] <robbiew> I think it's mostly useful for trying out apps you pay for
[16:43] <robbiew> would need to sort out the licensing logistics of it...maybe if the application has a demo
[16:43]  * mvo nods
[16:43] <robbiew> we could use that
[16:43] <robbiew> like World of Goo
[16:44] <mvo> I wonder how well the performance is for a game, its really awsome for normal apps (like ff)
[16:44] <robbiew> I've done youtube over it
[16:45] <robbiew> did "okay"
[16:45] <robbiew> depends on the bandwidth
[16:45] <mvo> woah
[16:45] <robbiew> at the Rally, the sound was flawless...the video a bit choppy
[16:45] <robbiew> but the network in the hotel wasn't that good
[16:45] <robbiew> this was with an EC2 backing too
[16:48] <robbiew> okay..anything else
[16:48] <robbiew> ?
[16:49] <robbiew> #endmeeting
[16:49] <MootBot> Meeting finished at 10:49.
[16:49] <robbiew> thnx all!
[16:49] <mvo> thanks
[16:50] <barry> thanks!
[16:51] <jhunt_> thanks
[16:52] <ev> thanks
[18:24] <highvoltage> Anybody around for an Edubuntu meeting?
[18:25]  * highvoltage goes ahead and spews out random updates for anyone who's willing to read then
[18:25] <highvoltage> Stéphane finished the application selector so that users can choose what to remove from their final installation: http://www.stgraber.org/2011/02/22/edubuntus-installer-ready-for-11-04/
[18:26] <highvoltage> that's pretty much the last of what we need for Natty feature-wise
[18:26] <highvoltage> LTSP Live is in a pretty bad state at the moment, but that's bugs and isn't directly affected by the feature freeze on Thursday.
[18:28] <highvoltage> I propose a bug day for 16 March, this page will be published to the frontpage if no one has a problem with the date (or the lolbug): http://edubuntu.org/2011-02-23/edubuntu-bug-day-16-march
[18:28] <czajkowski> highvoltage: would you not tie in with UGJ ?
[18:29] <highvoltage> czajkowski: admittingly, I haven't been following the UGJ stuff recently, is there one planned for that week?
[18:29] <highvoltage> czajkowski: I'd prefer it if it could tie in somehow with that
[18:29] <czajkowski> highvoltage: 2 weeks later
[18:30] <czajkowski> highvoltage: http://www.lczajkowski.com/2011/02/23/ubuntu-global-jam-coming-soon/
[18:31] <highvoltage> czajkowski: that would've been nice, but it's after beta freeze, ideally we should do it before beta freeze imho
[18:31] <highvoltage> (beta freeze is also UI freeze this time round)
[18:32] <czajkowski> :s
[18:34] <highvoltage> I think we should do something for UGJ anyway, it would be good for documentation, website/wiki updates and other close-to-release stuff
[18:35]  * stgraber waves
[18:35] <stgraber> and goes back to hide for 25minutes until the meeting actually start ;)
[18:37] <highvoltage> yeah I made a mistake. edubuntu meeting is in 20 minutes so I'll go back to lurking then :)
[18:39] <tumbleweed> highvoltage: you really are mixing up your timezones atm :)
[18:41] <highvoltage> tumbleweed: yeah the last week has been bad, and something is messes up in my chromium profile so all my times in my zimbra calender is also wrong :)
[18:41] <highvoltage> s/messes/messed/
[18:42] <charlie-tca> um, why not have a bug day on a thursday and get the bugsquad involved?
[18:44] <highvoltage> charlie-tca: that's a good idea, are you involved with the bugsquad?
[18:44] <charlie-tca> yeah
[18:44] <highvoltage> charlie-tca: so could we make it a bugsquad thing too then and do it on March 10th?
[18:44] <charlie-tca> let me look
[18:46] <charlie-tca> Yeah, I think we could do that
[18:47] <highvoltage> ok, I'm going to link to the bugsquad too then in that story..
[18:47] <charlie-tca> This is where we plan them - https://wiki.ubuntu.com/UbuntuBugDay/Planning
[18:47] <charlie-tca> I could help you set it up
[18:48] <highvoltage> charlie-tca: thanks. would it be better to host it in the ubuntu-bugs channel or in #edubuntu?
[18:48] <charlie-tca> bugs channel if we get the hug day set up for it
[18:49] <highvoltage> ok
[18:55] <highvoltage> mgariepy: besides what I pasted in #edubuntu, all that we discussed is bugday stuff, we'll be doing the edubuntu bugday with the bugsquad team on 10 March
[18:55] <charlie-tca> Adding it to the planning page
[18:55]  * highvoltage did already
[18:55] <highvoltage> (if that's ok)
[18:55] <charlie-tca> yup
[18:56] <highvoltage> charlie-tca: I wasn't sure if I should've added your name to it too, but I'll leave that to you
[18:56] <charlie-tca> nope
[18:56] <charlie-tca> you and pedro know I am involved. That's good enough
[18:58] <highvoltage> ok, I updated http://edubuntu.org/2011-02-23/edubuntu-bug-day-10-march
[18:59]  * highvoltage checks with the bugsquad docs whether there's anything else that needs to be mentioned
[19:00] <highvoltage> the dates on https://wiki.ubuntu.com/UbuntuBugDay will just need an update, it says the next bug hug day is on 10 Feb
[19:00] <charlie-tca> there will more to do about the 5th,
[19:00]  * stgraber waves
[19:00] <alkisg> ...we're declaring 16 March an Edubuntu Bug day!
[19:00] <alkisg> ...The bug day will last all day on 10 March 2011
[19:00] <charlie-tca> Yeah, we have bug days weekly on thursday
[19:00] <alkisg> sorry just saw that
[19:00] <maco> alkisg: which date?
[19:00] <charlie-tca> 10 march
[19:00] <alkisg> highvoltage: that 16 needs a change to 10?
[19:01] <alkisg> And wed to thu?
[19:03] <highvoltage> alkisg: I got the date, where does it say wed?
[19:03] <highvoltage> (unless you're looking at the publish-date)
[19:05] <alkisg> highvoltage: i refreshed my browser and it's fixed now
[19:05] <highvoltage> alkisg: ok
[19:06] <highvoltage> so... is the lolbug ok?
[19:06] <highvoltage> I like it because it makes bug days sound less scary, but not sure if people would be ok with that
[19:09] <highvoltage> ok, I'll hit publish then
[19:13] <highvoltage> ah, I guess it's also worth mentioning, last week we changed the weekly meetings to monthly meetings. weekly meetings seem to be a bit too frequent for us
[19:13] <highvoltage> so it will be on the last wednesday of each month now
[19:14] <highvoltage> I also added a block on the edubuntu page that reminds of the next meeting, I'll just have to remember to update it after every meeting
[19:14] <highvoltage> anything else?
[19:15] <alkisg> http://www.stgraber.org/2011/02/22/edubuntus-installer-ready-for-11-04/ this is nice
[19:15] <alkisg> I hope in the future local edubuntu teams will be able to use PPAs to add additional local software to that installer :D
[19:16] <highvoltage> yeah, stgraber put quite some work into the logic of how selections affect other selections. it's quite cool
[19:17] <highvoltage> alkisg: for their own spins that should be possible, but for edubuntu specifically we can only use main/universe/restricted
[19:17] <highvoltage> the last 2 daily builds failed because I accidentally added 2 packages from multiverse (eek)
[19:17] <alkisg> highvoltage: isn't that negotiable? :D I mean, to add custom repositories from within ubiquity?
[19:17] <alkisg> (not in the dvd of course)
[19:17] <stgraber> alkisg: there's one line in the plugin where you defined what source package to look at for meta packages
[19:18] <stgraber> alkisg: so a local respin would need to have the PPA in sources.list, have apt-get update run and update that line in the plugin
[19:18] <highvoltage> alkisg: I think that's a "yes" to your question, then
[19:19] <alkisg> That's nice. But a local respin can also just directly add software to the CD.
[19:19] <highvoltage> :)
[19:19] <alkisg> I think that local teams would appreciate a more direct approach though...
[19:19] <highvoltage> alkisg: that's what's happening with edubuntu actually, those packages are already installed. ubiquity just removes the packages that get unticked
[19:19] <alkisg> But there's an option to add more packages from the DVDE
[19:19] <alkisg> *dvd
[19:20] <alkisg> So the code to add packages is there - it would only take an "add-apt-repository" in some ubiquity page
[19:20] <alkisg> Maybe under a "customize" button somewhere
[19:21] <alkisg> That would help in getting some communities better organized
[19:21] <alkisg> (not talking about natty specifically - just an idea)
[19:22] <highvoltage> alkisg: yeah, if someone wants to do that then the ubiquity plugins in the edubuntu-live package could be useful
[19:23] <highvoltage> well, I guess that's a wrap then, next Edubuntu meeting is on 30 March 2011. Make some noise about it. if we do it monthly we might as well do it properly :)
[19:23] <mgariepy> thanks everyone :)
[19:24] <alkisg> Thanks
[19:24] <highvoltage> thanks alkisg, charlie-tca, mgariepy, stgraber, maco and czajkowski!
[19:24] <highvoltage> *GONG*