15:01:44 <slangasek> #startmeeting
15:01:44 <meetingology> Meeting started Wed Jun  6 15:01:44 2012 UTC.  The chair is slangasek. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
15:01:44 <meetingology> 
15:01:44 <meetingology> Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired
15:01:49 * stgraber waves
15:01:52 <ogra_> moop
15:02:14 <slangasek> [TOPIC] lightning round
15:02:15 * xnox turns off countdown theme tune
15:02:33 <slangasek> $ echo $(shuf -e barry doko stgraber jodh ev bdmurray slangasek ogra infinity cjwatson xnox stokachu)
15:02:36 <slangasek> ogra ev stokachu stgraber slangasek xnox cjwatson bdmurray infinity doko barry jodh
15:02:46 <highvoltage> nice
15:02:57 <ogra_> geez, i'm first !
15:03:00 <ogra_> done:
15:03:00 <ogra_> * flash-kernel fixes and merges and more findings for more fixes and merges next
15:03:00 <ogra_> week
15:03:00 <ogra_> * image tests ... and more image tests.. and even more ahead
15:03:00 <ogra_> * researching image build issues
15:03:01 <ogra_> * discovered a bunch of bugs with the recent panda kernels and worked with paolo on them
15:03:03 <ogra_> todo:
15:03:05 <ogra_> * corpus christi tomorrow in my state (not all of germany has it as a bank holiday, my state does though)
15:03:08 <ogra_> * fix live-build to actually produce ac100 images with the new flash-kernel
15:03:10 <ogra_> * finally switch to live arm images now that A1 is done
15:03:12 <ogra_> * look into how to integrate flash-kernel into debian-cd to make images bootable
15:03:14 <ogra_> * upload pending merges
15:03:16 <ogra_> ..
15:03:25 <ev> pass (still typing)
15:04:13 <xnox> stokachu: has been working hard this week as well?!
15:04:17 <slangasek> stokachu: hiya :) anything you'd like to share re: this week's activity?  (autofs, autofs, and more autofs)
15:04:49 <xnox> slangasek: there was dpatch envolved as well =)
15:06:25 <slangasek> stgraber: go ahead
15:06:33 <stgraber> - SRU
15:06:33 <stgraber> - Went through the list of pending SRUs for Precise and helped verify these that didn't need any weird setup
15:06:37 <stgraber> - First 12.04.1 team meeting, worked on team wiki page, announced team on mailing-lists
15:06:40 <stgraber> - While testing LXC nesting, found a pretty nasty bug in euca2ools (making /etc/ssl/certs a static file if installed before openssl), fixed and SRUed
15:06:43 <stgraber> - Release
15:06:45 <stgraber> - Quick training on cdimage handling and alpha-1 preparing with Kate
15:06:48 <stgraber> - Alpha-1 preparation
15:06:50 <stgraber> - Tracked down a python3-dbus bug making the installer crash when connecting to wifi, worked around for alpha-1, fix is now upstream
15:06:53 <stgraber> - Containers
15:06:56 <stgraber> - Added LXC backend to auto-upgrade-testing
15:06:58 <stgraber> - Implemented early version of the python module, currently wrapping the C commands, will be ported to liblxc soon
15:07:01 <stgraber> - Started testing LXC's SRU, found a bug, fixed in quantal and SRUed the fix
15:07:04 <stgraber> - ISO tracker
15:07:07 <stgraber> - New roles for product management and testcase management
15:07:09 <stgraber> - Add a new buginstruction text field to products
15:07:12 <stgraber> - Improve ACL checking
15:07:14 <stgraber> - Some more work on testsuites management, moving columns between tables, updating UI
15:07:17 <stgraber> - Other
15:07:20 <stgraber> - Booked flights to Plumbers
15:07:22 <stgraber> - TODO this week
15:07:25 <stgraber> - More alpha-1 fun
15:07:27 <stgraber> - Continue with ISO tracker changes based on the list from QA and work items
15:07:30 <stgraber> - Continue going through the pending-sru and help for these that are stuck on verification-needed
15:07:33 <stgraber> - Look at the isc-dhcp SRU for Oneiric (asked for some information in the bug)
15:07:36 <stgraber> (DONE)
15:07:57 <slangasek> ev: ready yet?
15:08:15 <ev> yes
15:08:29 <ev> - Very short week. Packing Millbank desk for Blue Fin move on Thursday, party
15:08:30 <ev> on Sunday, public holiday Monday and Tuesday.
15:08:30 <ev> - Mail to Andy, Martin, and James Westby on a plan for how we can get the
15:08:30 <ev> kernel crash collecting work finished and reporting to errors.ubuntu.com.
15:08:30 <ev> - Wrote debugging guide for whoopsie-daisy:
15:08:30 <ev> https://wiki.ubuntu.com/ErrorTracker/Debugging
15:08:30 <ev> - Uploaded a new Wubi for 12.10.
15:08:35 <ev> - Worked in some performance improvements to errors.ubuntu.com requested by
15:08:35 <ev> webops. We've been seeing some timeouts getting hit by some of the
15:08:35 <ev> components talking to Cassandra. As we don't have statsd and Graphite up yet
15:08:36 <ev> (RT awaiting action from webops), I can only pick at low hanging fruit. This
15:08:36 <ev> change moves from getting a column count from several wide rows (which reads
15:08:36 <ev> the entire row into memory) for the "average crashes per user" graph to
15:08:36 <ev> using counters.
15:08:42 <ev> - Added txstatsd and Graphite code to lp:errors and lp:daisy. Added JMX to
15:08:42 <ev> Graphite via JMXTrans. The latter gives us Cassandra operational stats in
15:08:42 <ev> Graphite. None of this is deployed yet as we need a new Graphite server
15:08:42 <ev> (RT #53325).
15:08:42 <ev> - Some clean up of the tastypie branch of lp:errors on advice from James
15:08:42 <ev> Westby. Now implements the list protocol to provide lazy evalution of the
15:08:42 <ev> query, emulating behavior found in evil ORMs.
15:08:43 <ev> - More discussion with Martin around uniting the Launchpad and crashdb
15:08:43 <ev> retracers.
15:08:44 <ev> (done)
15:08:47 <slangasek> good thing it's a short week ;)
15:08:56 <ev> heh
15:10:02 <ev> - Read on Wandsworth Common
15:10:03 <ev> - Played catch with the dog
15:10:03 <ev> - Watched pretty much the entire running of It's Always Sunny in Philadelphia on Netflix
15:10:03 <ev> (done)
15:10:19 <ogra_> legally ?
15:10:36 <ev> ogra_: it is not yet illegal to play catch on the common
15:10:37 <infinity> How do you illegally play catch with a dog?
15:10:44 <ogra_> ha
15:11:00 <slangasek> * lining up the python3 sprint
15:11:00 <slangasek> * DebConf planning
15:11:01 <slangasek> * discovered a lingering startup issue with rpcbind/statd/nis; fix in progress (bug #1007293)
15:11:03 <ubottu> Launchpad bug 1007293 in rpcbind (Ubuntu Precise) "nis doesn't work anymore after upgrade to 12.04" [High,In progress] https://launchpad.net/bugs/1007293
15:11:03 <slangasek> * SRU processing, and some verification (update-notifier, rpcbind)
15:11:05 <slangasek> (done)
15:11:06 <njin> who win, the dog ?
15:11:07 <slangasek> * helping here and there with A1 release this week
15:11:23 <ev> njin: always
15:12:11 <njin> Hurra for the dog
15:12:47 <slangasek> xnox:
15:12:53 <xnox> * autofs merge is in progress, some patches accepted into debian
15:12:53 <xnox> * applied for core-dev, due on 18th of June
15:12:53 <xnox> * working on foundations-q-ubiquity-lvm-luks, attempting to pre-seed now
15:12:54 <xnox> * mpt is working on the luks design (no mockups yet)
15:12:54 <xnox> * got quad hdd usb3.0 dock, helping kernel team to bisect when we lost
15:12:54 <xnox> 3.0 support in kernel cause I have the
15:12:54 <xnox> hardware. https://bugs.launchpad.net/ubuntu/+source/linux/+bug/966248
15:12:55 <xnox> * short week due to queen's diamond jubilee
15:12:55 <xnox> * this week: ubiquity, ubiquity, ubiquity...
15:12:55 <xnox> * during other days did boost1.49, offlineimap, usb-creator for
15:12:56 <xnox> debian work, visited canonical offices.
15:12:57 <ubottu> Launchpad bug 966248 in linux (Ubuntu) "USB3.0 Ports Not Working" [Medium,Confirmed]
15:13:21 <slangasek> xnox: do you know an eta from mpt for the mockups?
15:13:33 <xnox> slangasek: no.
15:13:51 <slangasek> ok
15:13:58 <cjwatson> Three-day week due to public holidays.  (Hid up north and pretended not to be British.)
15:14:01 <cjwatson> Wrote an API client for removing packages and killed off the old shell-access-only script.
15:14:04 <cjwatson> Landed PaS parsing fix for any-arm.
15:14:07 <cjwatson> Fixed ubiquity bug 1003443 which was breaking automatic tests.
15:14:09 <ubottu> Launchpad bug 1003443 in ubiquity (Ubuntu Quantal) "Ubuntu Quantal Desktop 20120523 failed to install: ubi-timezone failed with exit code 2" [Critical,Fix released] https://launchpad.net/bugs/1003443
15:14:09 <cjwatson> Added syslinux-themes-ubuntu-quantal package to fix Chinese edition image builds.
15:14:12 <cjwatson> Attempted to fix up update-manager following work on relative imports; unfortunately I ran out of time and Michael had to clean up after me.
15:14:15 <cjwatson> *Just* managed to get https://code.launchpad.net/~cjwatson/launchpad/queue-api/+merge/108967 in for review, which should be enough to allow moving queue to an API client.
15:14:18 <cjwatson> ..
15:14:53 <bdmurray> investigation into the fix for bug 346386 not working in Precise
15:14:53 <bdmurray> queried for foundations bugs with the most duplicates that are actively receiving dupes
15:14:55 <ubottu> Launchpad bug 346386 in apt (Ubuntu Natty) "[MASTER] Update fails with invalid package files with "Encountered a section with no Package: header"" [High,Fix released] https://launchpad.net/bugs/346386
15:14:56 <bdmurray> review of foundations team bugs with the most duplicates
15:14:58 <bdmurray> wrote json reader / alerter for spikes in bug volume e.g. recent-package-bugs
15:15:03 <bdmurray> modified bug bot not to modify precise bugs saying they need a new update-initramfs
15:15:06 <bdmurray> modified search-bugs in ubuntu-bugpatterns to copy tags from duplicate to the master bug
15:15:09 <bdmurray> wrote lp-bug-dupe-props in lptools to prevent working on bugs with duplicates with the same reporter
15:15:12 <bdmurray> modified lp-bug-dupe-props to check description for apport key / value pairs
15:15:17 <bdmurray> reported bug 1007637 regarding preventing reporting multiple package install failures
15:15:19 <ubottu> Launchpad bug 1007637 in apport (Ubuntu) "ubuntu general hook should check to see if package install failure has already happened" [High,Triaged] https://launchpad.net/bugs/1007637
15:15:20 <bdmurray> apport modification to ensure that package_hook tags update-manager dist-upgrade bugs dist-upgrade
15:15:23 <bdmurray> investigation into apport-package failures not being sent to the error tracker
15:15:26 <bdmurray> launchpad dev environment setup
15:15:34 <bdmurray> ␗ done
15:16:02 <infinity> * Last week:
15:16:02 <infinity> - Did some SRU stuff and other random AA faff while in HK
15:16:03 <infinity> - Worked on non-public *mumble*mumble* stuff with jk
15:16:03 <infinity> - Reviewed and sponsored some simple uploads for xnox
15:16:03 <infinity> * Linaro Connect last week:
15:16:05 <infinity> - Good discussions around v8 progress in ARM and with upstreams
15:16:07 <infinity> - ARM will be backporting their GCC patches to 4.7 for us
15:16:10 <infinity> - Binutils should be out "soon", but still ARM-internal for now
15:16:12 <infinity> - Need to chase up Linaro to get qemu-user-aarch64 working
15:16:15 <infinity> - Lots of other good discussions around big.LITTLE, v7 server
15:16:17 <infinity> concerns, single zImage (almost there!), and other fun bits
15:16:20 <infinity> * PlusOneMaint, starting this week
15:16:22 <infinity> - Cleaned up quantal_probs on Monday, leading to my first +1 beer
15:16:25 <infinity> - Training up Andy for PlusOneMinion tasks, currently teaching him
15:16:27 <infinity> the joys and pain of hunting build failures
15:16:30 <infinity> - Setting up my new laptop, as the old one's display unceremoniously
15:16:32 <infinity> died the day after I got home from Hong Kong
15:16:35 <infinity> (done)
15:17:10 <barry> py3 porting: lazr.authorization, wsgi_intercept, oauth, httplib2 (issue 210), lazr.restfulclient.  this did not go well (details available on request), so i explored alternatives w/pitti who made me very happy.  apport's dependency on launchpadlib is server side only, so he's going to juggle the packages/deps to remove the ubuntu-desktop task dep.  \o/
15:17:14 <barry> python issue 14673 and pep 421 implementation (review, sponsor).
15:17:19 <barry> provided some consulting for bug 1008898 (upstream dbus-python py3 bug).
15:17:21 <ubottu> Launchpad bug 1008898 in ubiquity (Ubuntu Quantal) "crash after inserting wireless password" [High,Fix released] https://launchpad.net/bugs/1008898
15:17:24 <barry> worked a bit on my dd application + debian uploads
15:17:28 <barry> todo: re-chase some dependencies and update the spreadsheet.  prep for next week's sprint
15:17:41 <barry> done.
15:18:13 <njin> translating debugging central for italian reporters, hardware testing A1 - in progress
15:18:58 <njin> done
15:19:31 <ogra_> no jodh today ?
15:19:48 <infinity> Not in the channel, at any rate.
15:19:51 <ogra_> yep
15:20:07 <slangasek> right
15:20:50 <slangasek> [TOPIC] Bugs
15:21:08 <slangasek> bdmurray: anything new? :)
15:21:25 <bdmurray> I mentioned bug 346386 being unfixed in precise
15:21:28 <ubottu> Launchpad bug 346386 in apt (Ubuntu Natty) "[MASTER] Update fails with invalid package files with "Encountered a section with no Package: header"" [High,Fix released] https://launchpad.net/bugs/346386
15:21:44 <slangasek> yep
15:21:46 <ogra_> from natty ?
15:21:49 <ogra_> wow
15:22:04 <slangasek> did you find the reason for its unfixedness?
15:22:15 <bdmurray> I found the relevant bit I think
15:22:35 <bdmurray> in acquire-item.cc there is now a if (Verify == true) check
15:22:48 <bdmurray> which didn't exist in oneiric
15:23:08 <slangasek> should the bug be reopened?
15:23:44 <bdmurray> yes probably
15:24:11 <slangasek> ok
15:24:28 <slangasek> next steps?  will you follow up with the apt maintainers (David/mvo)?
15:24:57 <bdmurray> yes, I've talked to mvo about it briefly
15:25:06 <slangasek> ok cool
15:25:58 <slangasek> anything else on the "actively receiving dupes" list that we should be aware of?
15:26:02 <bdmurray> as I was looking at bugs w/ the most duplicates I ran across bug 529714
15:26:04 <ubottu> Launchpad bug 529714 in eglibc (Ubuntu) "rhythmbox crashed with SIGSEGV in _nss_wins_gethostbyname_r()" [Low,Triaged] https://launchpad.net/bugs/529714
15:26:31 <slangasek> hmm, are there new dupes coming in?
15:26:32 <bdmurray> which times out
15:27:03 <bdmurray> no, but it clutters the most duplicates list
15:27:11 <slangasek> right
15:27:14 <bdmurray> and I thought breaking out the 'suggesting use of RTLD_DEEPBIND when
15:27:20 <bdmurray> opening nss modules from eglibc' task might be a good idea
15:27:25 <slangasek> sounds good to me
15:27:29 <bdmurray> okay
15:28:06 <bdmurray> then infinity and I were looking at bug 1001904 yesterday evening
15:28:08 <ubottu> Launchpad bug 1001904 in sysvinit (Ubuntu) "package sysvutils (not installed) failed to install/upgrade: trying to overwrite '/usr/share/man/man1/mesg.1.gz', which is also in package sysvinit-utils 2.88dsf-13.10ubuntu11" [Undecided,Incomplete] https://launchpad.net/bugs/1001904
15:28:38 <bdmurray> the users had added hardy to their sources
15:28:49 <slangasek> and people are doing this because there's a java package still in the hardy archive?
15:29:16 <slangasek> we removed it from partner for all releases... I guess we overlooked hardy because it was in the main archive, not in partner?
15:30:35 <slangasek> cjwatson: ^^ do we want to remove sun-java from the hardy release pocket?
15:32:00 <cjwatson> We've never regenerated a release pocket Packages so far
15:32:10 <cjwatson> In the past I've said we'd only do it given a CnD
15:32:23 * slangasek nods
15:32:47 <slangasek> any other action to take here, then?  I'm not sure it makes sense to worry about the missing replaces per se
15:32:52 <xnox> do we have permission to continue to distribute sun-java though? I thought there was termination.
15:32:52 <cjwatson> So, well, it's possible, but it would have to be a tremendously compelling reason
15:32:58 <cjwatson> xnox: Only for later versions
15:33:05 <xnox> ok
15:33:21 <infinity> We could fix the replaces, but really, pulling in Essential packages from a 4 year old release is sort of a "you get to keep both pieces" scenario.
15:33:32 <slangasek> agreed
15:33:41 <infinity> If we fix the replaces, would it completely disappear the old package?
15:33:47 <infinity> Cause that's vaguely elegant, at least.
15:33:56 <infinity> Or... Maybe that would be hilariously bad.
15:34:27 <bdmurray> well is there something better we can do to detect people in this scenario?
15:34:48 <infinity> It's pretty clearly unsupported-but-allowed use of the tools.  I dunno how much hand-holding we can do.
15:36:20 <slangasek> infinity: right - so what about detecting that this is what a user has done, so we can disregard the bug reports (or block them beforehand)?
15:36:39 <slangasek> I'm not interested in us spending effort making this work for users
15:36:48 <infinity> Disregarding is a bit harsh without feedback telling them why.
15:37:00 <slangasek> there are many bug reports that go disregarded
15:37:10 <infinity> And the effort spent on making fancy hooks for that sounds like more than just doing the replaces and papering over it.
15:38:35 <slangasek> so we think this is a one-off circumstance, and we won't see other bug reports?
15:38:57 <infinity> There are two dupes.
15:39:10 <infinity> Not sure how many others haven't been found.
15:39:15 <bdmurray> I'll try and search for more 3 isn't really a large quantity
15:39:18 <slangasek> by "other" I mean "other kinds of bugs caused by trying to travel back in time 4 years"
15:39:30 <infinity> Oh.
15:39:44 <slangasek> if it's going to be a general class of issue, better to suppress bug reporting
15:39:47 <infinity> Nothing that I've seen.  But bdmurray's the one who knows all. :P
15:40:03 <cjwatson> I can imagine some, but auto-installation of Essential seems like the hardest to avoid.
15:40:27 <slangasek> if we really think this is the only bug of its kind, then that's easily done
15:41:48 * slangasek assigns the bug to infinity
15:42:03 <infinity> Just what I always wanted.
15:42:03 <slangasek> bdmurray: others?
15:43:04 * ogra_ has an errors.u.c question ... does that fit in this topic ?
15:43:12 <bdmurray> slangasek: nope
15:43:18 <slangasek> bdmurray: ok, thanks
15:43:20 <slangasek> ogra_: let's say yes
15:43:23 <ogra_> k
15:43:48 <ogra_> so if distros like mint supress updates and security fixes ... bugs will re-occur over and over on the list ...
15:44:13 <ogra_> can erros be taught to hide bugs once a newer version was uploaded ?
15:44:17 <ogra_> *errors
15:44:28 <ogra_> thats an ev question i guess ...
15:45:04 <ogra_> that came up recently in a discussion i had and i promised to ask here :(
15:45:06 <ogra_> err
15:45:07 <ogra_> :)
15:45:10 <slangasek> well, errors is meant to know if a crash occurs with the current package or only with the older ones... at one point the headline graph showed "# of crashes" vs. "# of crashes we'd have if people were up to date"?
15:45:41 <slangasek> and the current display shows the first and last known package versions for a crash
15:46:18 <slangasek> I think hiding the bugs entirely might be too much at this point, though, since there might be other reasons people aren't getting the update that are in our control?
15:46:32 <cjwatson> We also want to know about unexpected regressions
15:46:45 <ogra_> which wouldn happen with a newer version
15:46:49 <ogra_> *would
15:47:12 <cjwatson> And about whether lots of our users are seeing bugs because for some reason they aren't up to date; I think to some extent this is a different-audiences question
15:47:20 * slangasek nods
15:47:39 <ogra_> k
15:47:53 <cjwatson> Many developers would only care about newest-version, but (say) release management might care about everything; I think I'd like to see it be controllable ...
15:47:53 <slangasek> ev: ^^ so what's the right answer for ogra_'s concern?  do we track these by sending you mail? :)
15:48:20 <bdmurray> I seem to recall a filter by release / distro feature coming to errors which may help
15:48:26 <ogra_> haha, sorry, i only channeled the question, i didnt mean to do anything evil to ev
15:48:52 <slangasek> well, I think it's a legitimate request for errors.u.c functionality
15:49:18 <infinity> I think ev died.
15:49:20 <ogra_> indeed, but i dont want him to drown in mail caused by mint users :)
15:49:46 <slangasek> oh, I wasn't suggesting having mint users email ev, I was suggesting *you* mail him ;P
15:49:52 <ogra_> oh, ok
15:49:55 <slangasek> anyway, ev seems idle, so best to take this offline from here
15:49:58 <ogra_> yeah, i can do that
15:50:01 <ogra_> right
15:50:03 <bdmurray> you can also report bugs a launchpad.net/errors
15:50:20 <slangasek> heh, namespace
15:50:30 <slangasek> [TOPIC] Python3 vsprint
15:50:41 <slangasek> hopefully everyone got the mail about the porting sprint next week
15:50:49 <slangasek> any questions about the plan there?
15:51:00 <slangasek> everyone know whether they're in an American or European timezone?
15:51:12 <infinity> Do I have to abandon my +1 minions to come be sprinty?
15:51:21 <infinity> (And I have no idea what timezone I'm in...)
15:51:26 <infinity> (ever)
15:51:40 <ogra_> cmon .... we know where to look that up !
15:51:58 <ogra_> just do a A1 test install, ubiquity will tell you !
15:51:59 <slangasek> infinity: yes - you can tell them you're "on call" available for those days if things come up?
15:52:09 <infinity> slangasek: Check.  Can do.
15:52:33 <infinity> ogra_: I do all my test installs on stgraber's server, that doesn't help.
15:52:43 <slangasek> hah
15:52:48 <ogra_> heh
15:52:51 <slangasek> [TOPIC] AOB
15:52:54 <slangasek> anything else?
15:53:04 <ev> sorry, got pulled into a side conversation
15:53:08 <ev> I have AOB
15:53:14 <infinity> Not just OB?
15:53:16 <ev> and then I'm happy to loop back and talk with ogra_
15:53:20 <ev> infinity: :)
15:53:34 <ogra_> ev not urgent, we can do it another time
15:53:49 <ev> so pitti and I were talking about having references to the error tracker in uploads
15:53:52 <slangasek> infinity: he has Astounding Oratorial Bacon
15:53:56 <ev> much like we do with LP bugs
15:54:19 <infinity> That seems messy.
15:54:23 <cjwatson> I'm going to have to go to catch a bus, I'm afraid - need to leave sharp today
15:54:26 <infinity> Given that error reports should align with bugs anyway, no?
15:54:31 * slangasek waves to cjwatson
15:54:37 <ev> what do people think of a full SHA-1 hash of a crash signature (the things you see in the function column on errors.ubuntu.com) for this, or the first half of the SHA-1 hash
15:54:41 <bdmurray> I thought bugs were going to contain information about errors?
15:54:51 <slangasek> ev: wouldn't we want any error we care about fixing to have a corresponding bug report?
15:54:59 <ogra_> ++
15:55:04 * infinity notes we all seem to be on the same page here. :P
15:55:11 <infinity> errors -> bug -> fix bug.
15:55:19 <infinity> Underpants, etc.
15:55:30 <slangasek> if nothing else, an error tracker ref doesn't give us anything to hang the developer's thoughts on
15:55:34 <xnox> statistically ev, you can trim it down to first 7 characters. But, I'd rather see lp: bug references only.
15:55:37 <ev> something needs to refer back to the error tracker
15:55:45 <infinity> ev: The bug can.
15:55:48 <xnox> ev: add errors as a bug tracker on launchpad.
15:55:49 <ogra_> and the bug cant ?
15:55:56 <cjwatson> We need a tool to promote error-tracker entries to bugs, I think
15:56:00 <xnox> ev: link to 'distributions errors.ubuntu.com' as a bug watch
15:56:15 <cjwatson> To make it easy and make it more likely that it's done in a vaguely standard way
15:56:16 <infinity> That would actually kinda work.
15:56:27 <infinity> (The external tracker link)
15:56:36 <infinity> But just a simple tag works too.
15:56:46 <ev> tag is probably easlier
15:56:47 <ev> easier even
15:56:48 <xnox> infinity: sha1 tag?
15:56:59 <xnox> or url tag?
15:57:11 <infinity> Don't care.  Just "a bug tag".
15:57:20 <infinity> errors:sha1 or something would work.
15:57:23 <ev> I just need to give that one some though, as I really don't want to involve a cron job on this
15:57:26 <slangasek> you want a separate bug tag for each error entry?
15:57:35 <slangasek> my understanding is that launchpad does not like scaling tags
15:57:37 * cjwatson suggests checking with the LP team in case a bazillion different tags would cause some kind of problem
15:57:51 <ev> oh no, it wouldn't be lots of tags
15:57:52 <infinity> Oh, right, tags are actually hashed as keys.
15:57:54 <infinity> Nevermind.
15:57:57 <ev> it would be the needs-dupcheck tag
15:57:59 <ev> or whatever it is
15:58:09 <cjwatson> Right, gone
15:58:12 <ev> which would then cause the LP retracer to pick it up
15:58:23 <ev> but that adds the latency of waiting for the LP retracer
15:58:25 <ev> hmm
15:59:47 <ev> we could just have the errors.ubuntu.com backend create the bug when you click on a "create bug" link and immediately write it in to the bug field in the BucketMetadata column family
16:00:05 <ev> the "create bug" link only being present when one doesn't exist in the bug column
16:00:20 <slangasek> any other acl on that?
16:00:21 <ev> this would mean any linked bugs would be created by a bot
16:00:39 <ev> well, the standard canonical+bugsquad or whatever it is
16:00:43 * slangasek nods
16:01:14 <slangasek> sounds reasonable enough to me
16:01:39 <ev> okay, I'll confirm with martin and hopefully go with that
16:01:42 <ev> thanks
16:02:48 <slangasek> cool
16:02:50 <slangasek> anything else?
16:03:23 <slangasek> #endmeeting