[12:48] <trah-lt> hey :)
[01:48] <dilys> Merge to devel/launchpad/: [r=BjornT]  Denastifies the presentation of error messages in Launchpad. Adds distinct notification for error, warning, informational, and debug messages, and simplifies the style sheet. Replaces error-notification, warning-notification, info-notification, debug-notification, portalError, and portalMessage. (r3334: Matthew Paul Thomas)
[01:55] <raphink> wow what happened with karma ?
[02:12] <dilys> Merge to devel/launchpad/: [r=jamesh]  Tidies and deUbuntifies the code of conduct pages, hopefully fixing bug 31933 (People don't understand how to sign the Code of Conduct). (r3334: Matthew Paul Thomas)
[04:41] <johncudd> hello
[04:42] <johncudd> ubuntu ROCKS!!
[04:43] <johncudd> opped?
[04:44] <johncudd> I've never used IRC amazingly. You'd think having been on linux for so long, and computers in general I would have learned how to use IRC.
[04:46] <johncudd> how to improve ubuntu launchpad
[04:46] <johncudd> can anyone see these messages?
[04:56] <johncudd> join debian
[06:59] <omega21> hi there, im not getting an activation email... is something wrong?
[07:01] <omega21> anyone???
[07:06] <omega21> anyone???
[09:31] <dilys> Merge to devel/launchpad/: [trivial]  Remove givennames and familyname columns from the Person table and librarian-gc.py refactoring (r3334: Stuart Bishop)
[09:51] <jordi> kiko!
[09:51] <jordi> kiko: is carlos around?
[09:52] <kiko> jordi!
[09:52] <kiko> not yet, but soon
[09:52] <jordi> kiko: did you run today?!?!!?!!!!!
[09:52] <kiko> YES
[09:52] <jordi> oh no oh no
[09:53] <jordi> I described the symptoms I have in my knee after running to a (non-specialist) doctor, and he fears I have one of those injuries in my "menisco"
[09:53] <jordi> that'd suck. I'll go to some sport doctor soonish
[10:13] <carlos> morning
[10:14] <dilys> Merge to devel/launchpad/: [trivial]  Remove givennames and familyname columns from the Person table and librarian-gc.py refactoring (r3334: Stuart Bishop)
[10:53] <kiko> mmmmmumble
[10:59] <bradb> Wow stub, did you *just* bump it up to 200? :)
[10:59] <dilys> Merge to devel/launchpad/: [trivial]  Remove givennames and familyname columns from the Person table and librarian-gc.py refactoring and ftq quote handling (Bug #33920) (r3334: Stuart Bishop)
[10:59] <bradb> I was /just/ about to ask you to do that :P
[11:00] <kiko> another 3334 landing, lifeless 
[11:01] <jordi> carlos: so
[11:01] <jordi> carlos: who do I need to discuss this article with? jdub?
[11:01] <carlos> jordi: let's pitti and I to read it
[11:01] <jordi> nod
[11:01] <jordi> I'll send it to you two later today
[11:02] <carlos> jordi: ok
[11:02] <stub> bradb: It was 200 on Mondays rollout, shortly reduced to 75 because there are still performance issues. But it went back to 200 when lifeless cherry picked SteveA's hanging fix as I had not committed the config change to the production config
[11:02] <stub> production branch
[11:03] <bradb> I thought I saw it at 75 about 10 mins ago
[11:03] <stub> I'll check the configs - maybe the appservers have different settings
[11:04] <bradb> er, I see it at 75 again right now
[11:04] <bradb> 1  75 of 348 results
[11:04] <lifeless> stub ping
[11:05] <bradb> kiko: interestingly, our "Previous" link is disabled: https://launchpad.net/products/malone/+bugs?batch_start=75&batch_end=150
[11:06] <stub> bradb: half the app servers were set to 200, the other half 75.
[11:06] <lifeless> stub - patch number for 'ALTER TABLE Branch add column mirror_failure_message text; please
[11:06] <stub> lifeless: pong
[11:07] <stub> That the entire thing?
[11:07] <stub> 40-42-0.sql
[11:07] <bradb> stub: ah
[11:07] <lifeless> yeah, all we are doing is adding the one field
[11:07] <Ubugtu> Malone bug 36390 in malone ""Previous" link is disabled even when there are previous results" [Normal,Unconfirmed]  http://launchpad.net/bugs/36390
[11:08] <bradb> mpt: I thought you had a branch that removes the quotes from bug titles. Is that still in review?
[11:10] <lifeless> stub - also did you get my email about the bzr services to be part of stock rollouts
[11:14] <stub> lifeless: Yup. Seems sane if the setups are now stable
[11:14] <mpt> bradb, it's one of those that PQM blew off
[11:14] <lifeless> stub they are. So I will tell elmo we are agreed on that ?
[11:16] <Znarl> stub : gandwana Apps server 1 and 2 down.
[11:16] <stub> eh? bugger
[11:17] <lifeless> Znarl: hung again ?
[11:17] <lifeless> stub: I'll grab a backtrace
[11:17] <stub> I stopped them, but got distracted before restarting them
[11:17] <lifeless> haahah
[11:17] <stub> lifeless: don't bother
[11:17] <bradb> mpt: ok
[11:17] <lifeless> yeah, I can see how that would be less than easy
[11:20] <stub> SteveA: So we wanted another production update today?
[11:22] <lifeless> I'll ping him one sec
[11:22] <carlos> stub: yes, please :-P
[11:22] <SteveA> stub: yes please.  any time you fancy
[11:23] <lifeless> stub: rollouts - I think an rsync from chinstrap + a "ssh machine cd .../ && make build " will work and let us remove the ssh from appservers to chinstrap facility
[11:25] <stub> lifeless: ok. That requires us to run ssh-agent or enable automatic login to the various production accounts using my chinstrap ssh key.
[11:25] <lifeless> so we need to decide which is lower risk.
[11:25] <stub> lifeless: Doing if from balleny could be an option
[11:25] <lifeless> makes sense to me
[11:25] <lifeless> no need to bounce it past chinstrap at all
[11:26] <dilys> Merge to devel/launchpad/: [r=BjornT]  Denastifies the presentation of error messages in Launchpad. Adds distinct notification for error, warning, informational, and debug messages, and simplifies the style sheet. Replaces error-notification, warning-notification, info-notification, debug-notification, portalError, and portalMessage. (r3335: Matthew Paul Thomas)
[11:27] <daf> stub: can I have some DBA pee on https://chinstrap.ubuntu.com/~dsilvers/paste/fileIWZLqX.html?
[11:28] <stub> I hope that was a typo
[11:28] <bradb> heh
[11:29] <daf> s/pee/annointment/
[11:29] <daf> they're right next to each other on the keyboard
[11:30] <lifeless> daf: got some comment.sql love
[11:30] <bradb> on a chord keyboard, perhaps
[11:30] <daf> lifeless: indeed, I forgot about that
[11:31] <stub> How are users going to search keywords? Are they typing in keywords into a text field, or selecting them from a list?
[11:32] <daf> er
[11:32] <stub> If it is a text field, it would be better implemented as a new TEXT column on the Bug table, searchable using the text search engine, giving boolean operations for free. 
[11:32] <stub> If it is select from a list, the table suggested is probably best as we are only doing exact matches.
[11:33] <daf> I wasn't planning for full text searches
[11:34] <daf> but I'm nervous about limiting future options
[11:34] <stub> We need a constraint on the keyword column. Probably the standard valid_name stuff, as we probably don't want keywords and punctuation.
[11:35] <daf> why not punctuation?
[11:35] <SteveA> stub: person name blacklist?
[11:35] <SteveA> stub: is that in any way implemented yet?
[11:35] <stub> If the keywords are nice tokens rather than strings of arbitrary text, we can switch the implementation easily.
[11:35] <stub> SteveA: Not implemented at all. Just specced.
[11:35] <SteveA> thanks
[11:36] <daf> token == [a-z] [a-z_-] * ?
[11:36] <daf> well, digits too, presumably
[11:37] <stub> yup
[11:37] <daf> I'm sure people will try inserting . and : and = and #
[11:38] <daf> I don't like stopping them from doing so without good reason
[11:38] <stub> We want them to be usable in URLs
[11:39] <SteveA> daf, stub: bug keywords?  sensitive issue...
[11:39] <SteveA> please talk with kiko before proceeding
[11:39] <daf> ok
[11:39] <daf> kiko: are you around?
[11:40] <stub> I'd go with valid_name, which allows lowercase, digits, -, . and +
[11:40] <daf> well, I guess we can expand it later if we change our minds
[11:40] <daf> easier to go from restrictive -> permissive than the other way around
[11:40] <stub> Or storing them as a text field, allowing boolean queries like (laptop and power) or dell
[11:41] <daf> I think that's a really niche use case
[11:41] <stub> (or both...)
[11:41] <daf> I think:
[11:41] <daf> 90% of keyword usage will be searches for (product, keyword) or just keyword
[11:42] <daf> 9% will be searches for multiple keywords, possibly in the context of product/package/distorelease
[11:42] <daf> 1% might include ORs or complicated boolean stuff
[11:42] <daf> and I'm not sure it's worth supporting that
[11:42] <daf> for the added implementation complexity
[11:44] <daf> AFAIK, similar features in other software (del.icio.us, flickr, etc.), the operate on an AND-only basis
[11:44] <stub> I don't think implementation complexity is an issue - using a single text field may be less since all the heavy listing is already done and the search UI will be simpler.
[11:44] <daf> ok
[11:45] <daf> so if we add Bug.keywords
[11:45] <daf> and BugA.keywords = "not_laptop"
[11:45] <daf> will an exact search for "laptop" turn up bug A?
[11:46] <stub> You wouldn't allow underscores. If it was 'notlaptop', then no.
[11:46] <daf> say we use valid_name and allow -
[11:46] <daf> what about "not-laptop"?
[11:47] <stub> the search would match that (or if not, it will when hyphenation is sorted properly. I think it is half done)
[11:47] <daf> the semantics I expect is search_term in bug.keywords.split(' ')
[11:48] <stub> Sure. So don't allow hyphens in keywords.
[11:48] <daf> or else SELECT * FROM bug, bugkeyword WHERE bugkeyword.bug = bug.id AND bugkeyword.keyword = 'search_term';
[11:49] <daf> is the advantage of using Bug.keyword that is would allow substring matches?
[11:50] <stub> The advantage would be the automatic boolean search support, and some spelling correction (searching for 'laptops' would match 'laptop'). If we improve the text searching, the keyword search gets those improvements too.
[11:50] <daf> I think spelling correction is useful
[11:51] <daf> I don't think boolean search support is useful
[11:51] <stub> Spelling correction is only useful if the UI is a text box.
[11:51] <BjornT> stub: i sent you a db patch, could you please review it today?
[11:51] <stub> If we have a set of allowed keywords, the UI could be a set of radio buttons.
[11:52] <stub> eg. link a set of keywords to product or distribution. What keywords are available to search on would depend on your context.
[11:52] <daf> I don't think we can come up with an exhaustive list of useful keywords
[11:52] <daf> and I don't think we want the burden of having to modify the code to add keywords
[11:52] <daf> any more than we would want a fixed list of milestone names
[11:54] <daf> maybe I'm misunderstanding you here
[11:56] <daf> I'm anticipating a text entry for setting keywords on bugs
[11:57] <daf> initially, a search interface could be as simple as cliking on a keyword to take you to a list of bugs with that keyword
[11:57] <stub> BjornT: Is there any advantage to using the Message table at all if we don't use the BugMessage table to store notifications?
[11:57] <SteveA> stub: when are you planning that rollout?
[11:58] <stub> BjornT: Your original approach may be better
[11:58] <SteveA> cos me and kiko and mark are going through launchpad specs right now, using launchpad
[11:58] <daf> the only disadvantage to using a full text field is the possibility of false positives
[11:58] <stub> (17:20:13) stub: SteveA: So we wanted another production update today?
[11:58] <stub> I was going to get opinions on what patch level needs to land, and if it should be today or next week.
[11:59] <bradb> I'd love to see revno 3329 in there.
[12:00] <stub> daf: For the UI you describe, I think your data model is preferable
[12:00] <bradb> (Fix bug 33342 (Please make filtering on the component possible (main, universe)))
[12:00] <Ubugtu> Malone bug 33342 in malone "Please make filtering on the component possible (main, universe)" [Normal,In progress]  http://launchpad.net/bugs/33342
[12:01] <BjornT> stub: i want comment notifications to point to the actual comment, so that the notification get the same message-id as the comment.
[12:01] <daf> stub: what would be the cost of switching to your model later?
[12:01] <stub> BjornT: ok. That is a good idea.
[12:01] <dilys> Merge to devel/launchpad/: Fix https://launchpad.net/products/rosetta/+bug/31381 (POMsgSet.active_texts assumes POFile.pluralforms is an int) r=kiko (r3336: Diogo Matsubara)
[12:01] <kiko> congrats matsubara
[12:02] <stub> daf: If the keywords are sane (no whitespace or selected punctuation), easy
[12:02] <daf> ok -- I think your suggestion of using valid_name is good
[12:02] <daf> at any rate, I think disallowing whitespace makes the UI much simpler
[12:02] <daf> matsubara: woo!
[12:05] <SteveA> stub: later today please
[12:05] <SteveA> and we can probably skip next week's regular one
[12:05] <daf> interesting
[12:05] <daf> https://chinstrap.ubuntu.com/~jamesh/oops.cgi/2006-03-23/A485
[12:05] <stub> SteveA: Its 6pm now, so we can't leave it too long
[12:05] <daf> total time is 811ms -- why did it time out?
[12:06] <daf> or does that not include the time the query was running for before it was killed?
[12:07] <SteveA> stub: estimated downtime?
[12:07] <stub> 10 mins
[12:07] <SteveA> stub: can you do it in 53 minutes time ?
[12:07] <daf> stub: would you expect "SELECT project FROM ProjectBugTracker WHERE bugtracker = 2" to be slow?
[12:07] <stub> ok
[12:07] <SteveA> ta
[12:08] <stub> daf: No
[12:08] <daf> it timed out 7 times yesterday
[12:09] <daf> each time on https://launchpad.net/malone/bugtrackers/ubuntu-bugzilla/+index
[12:10] <carlos> SteveA, stub: would that rollout be from HEAD or what we merged on Wednesday night?
[12:10] <carlos> SteveA: I need a couple of fixes in with next week roll out....
[12:10] <carlos> or merge them now and get them rolled out today
[12:10] <daf> stub: can't reproduce it myself, though
[12:10] <carlos> I have them ready to be merged
[12:12] <daf> stub: oh, with ?display-all-watches=1, I can reproduce
[12:12] <daf> stub: looks like a job for Page Batching Man
[12:12] <stub> yup
[12:13] <jamesh> daf: are you sure it isn't just a case of death by 1000 cuts?
[12:13] <daf> I got confused by https://chinstrap.ubuntu.com/~jamesh/oops.cgi/2006-03-23/A485
[12:13] <daf> there was no obvious cause of timeout, either code or SQL
[12:13] <jamesh> daf: with the next rollout, we'll have the duration of the request in the OOPS report
[12:13] <daf> excellent
[12:14] <jamesh> daf: it might be doing a lot of Python stuff,
[12:14] <daf> c.f. https://chinstrap.ubuntu.com/~jamesh/oops.cgi/2006-03-23/B451
[12:14] <daf> Total sql time  	810 ms
[12:14] <daf> Non-sql time 	11966 ms
[12:14] <daf> that's obviously doing lots of Python stuff
[12:14] <daf> but in this case:
[12:14] <daf> Total sql time  	784 ms
[12:14] <daf> Non-sql time 	27 ms
[12:15] <daf> and yet it got a RequestExpired
[12:15] <jamesh> daf: the non-sql time is inaccurate
[12:15] <daf> oh
[12:15] <daf> I see
[12:15] <daf> why's that?
[12:15] <jamesh> well, it is inaccurate for SoftTimeoutErrors
[12:15] <daf> this is not a soft timeout
[12:16] <jamesh> daf: it is using the time when the last DB statement was issued rather than end of request
[12:16] <HumanTorch> hey everybody :)
[12:18] <daf> hello HumanTorch 
[12:18] <HumanTorch> hello daf :)
[12:18] <daf> jamesh: can we make the OOPS machinery be less confusing?
[12:18] <jamesh> in what way?
[12:18] <daf> there is nothing in this OOPS that tells me where the time spent was
[12:19] <mighty`> grr
[12:19] <daf> it says Python time was small
[12:19] <daf> it says SQL time was small
[12:19] <daf> it says an SQL statement expired
[12:19] <kiko> daf, it's supposed to be fixed
[12:19] <kiko> so let's see tomorrow's reports
[12:19] <kiko> (jamesh fixed it)
[12:20] <daf> ok
[12:24] <jamesh> daf: Python time is a lot higher than the page shows -- a lot of time ellapsed between statement 5 and the point of the traceback
[12:25] <jamesh> we'll be able to modify oops.cgi to accurately display that time soon.
[12:25] <daf> excellent
[12:25] <daf> thanks for looking into it
[12:27] <jamesh> daf: if you look at the recent staging OOPS logs in chinstrap:/srv/asuka-logs, you'll see a new "Duration" header.
[12:28] <jamesh> that's the missing information
[12:29] <dilys> Merge to devel/launchpad/: [r=stub on db patch, trivial]  rename BugBranch.fixed_in_revision to BugBranch.revision_hint (r3337: Brad Bollenbach)
[12:36] <stub> Launchpad is going down in 25 minutes for updates. Downtime will be approx. 10 minutes. Wikis will be in read only mode during this time.
[12:37] <jordi> carlos: if you said anything else, I missed it
[12:37] <jordi> natura seems to have crashed
[12:37] <carlos> jordi: nothing after you said you need to leave
[12:37] <jordi> ok
[12:42] <seb128> hi
[12:43] <jordi> hi seb
[12:43] <seb128> https://launchpad.net/malone/bugs/36214 does a "oops" page 
[12:45] <daf> got an OOPS code for me?
[12:46] <jordi> carlos:    build-tree/gtk+-2.8.14/po-properties/te.po in gtk+2.0 in Ubuntu Dapper
[12:46] <jordi> carlos: that one is in the approved queue
[12:46] <jordi> shouldn't that be in blocked?
[12:46] <carlos> jordi: no
[12:47] <carlos> the blocked ones are like -static
[12:47] <carlos> jordi: that's the good one
[12:48] <jordi> aha
[12:48] <jordi> I guess I mixed udeb and build-tree up
[12:48] <jordi> ok
[12:48] <carlos> stub: what happened to the other 10 minutes .....
[12:48] <jordi> heh
[12:48] <SteveA> stub: what carlos said
[12:48] <stub> Sorry. I've had to shutdown earlier than expected
[12:48] <jordi> aol :D
[12:49] <jordi> hmm
[12:49] <jordi> Copyright 2004-2005 Canonical Ltd.
[12:49] <jordi> from the 503 page
[12:49] <mpt> yeah
[12:49] <mpt> that page smells
[12:50] <mpt> but it's not in the Launchpad tree, which makes it hard to update
[12:50] <carlos> shouldn't have it updated automatically, so we define the first year and we get automatically current year?
[12:50] <carlos> that would save us work :-P
[12:51] <mpt> It's a static page, so doing that would require a yearly cronscript
[12:53] <seb128> daf: OOPS-83A194
[12:53] <Ubugtu> https://chinstrap.ubuntu.com/~jamesh/oops.cgi/83A194
[12:54] <daf> seb128: thanks
[12:54] <stub> Ok. Launchpad is back and upgraded. Downtime was -6 minutes.
[12:54] <seb128> daf: thank *you* for looking on it :)
[12:54] <carlos> stub: it's back online before the official down time... interesting :-P
[12:55] <stub> And we should have SSL certificates installed for the librarian, so we can look into serving librarian files over SSL to stop certificate complaints.
[12:55] <Znarl> Not librarian yet.  Just shipit.kubuntu.org and shipit.edubuntu.org
[12:56] <stub> Sorry about the early start to the upgrade - the database patches got applied early and too fast to stop 'em. So it was either full steam ahead or leave Launchpad half broken for 10 minutes.
[12:57] <carlos> stub: not a problem here
[12:57] <seb128> gna
[12:58] <seb128> malone is broken
[12:58] <dilys> Merge to devel/launchpad/: [r=jamesh]  BranchSet.getByUrl and associated db changes (r3338: David Allouche)
[12:58] <stub> seb128: URL?
[12:58] <seb128> https://launchpad.net/products/gedit/+bug/28555 , there is no way to change the upstream task to fixed
[12:58] <Ubugtu> Malone bug 28555 in gedit "Searched text remains highlighted even after search window has been closed" [Minor,Confirmed]  
[12:58] <daf> seb128: this OOPS is weird
[12:58] <seb128> the only box it has is the upstream bug
[12:58] <jordi> 12:51 < ibid> jordi: fixed
[12:59] <jordi> carlos: ^
[12:59] <seb128> daf: that's a bug jamesh imported from bugzilla yesterday
[12:59] <seb128> daf: https://launchpad.net/products/gnome-media/+bug/2999 is dupped of it
[12:59] <Ubugtu> Malone bug 2999 in gnome-media "gnome-volume-control segfaults" [Normal,Confirmed]  
[12:59] <daf> seb128: the page doesn't crash for me
[12:59] <seb128> daf: works now
[12:59] <daf> hmm
[12:59] <carlos> jordi: he forgot the .pot 
[01:00] <carlos> ?
[01:00] <seb128> daf: it was broken before launchpad rollout
[01:00] <daf> perhaps the rollous fixed it
[01:00] <seb128> yeah, maybe
[01:00] <daf> see how fast we fix bugs for you?
[01:00] <seb128> https://launchpad.net/products/nautilus/+bug/36252/+editstatus
[01:00] <Ubugtu> Malone bug 36252 in nautilus "desktop/nautilus right-click umount problem" [Normal,Unconfirmed]  
[01:00] <carlos> seb128: I guess it was due the db updates before having launchpad stopped
[01:00] <bradb> kiko: Do you have an email from mdz with specific use cases for release targeting?
[01:00] <seb128> how come than I can't edit an upstream task now?
[01:00] <jamesh> seb128: the status of the watched task can't be changed because it is supposed to be linked to the upstream bug tracker
[01:00] <daf> stub: seb's OOPS had a dbschema value of 999
[01:00] <daf> jamesh: ^^^
[01:01] <daf> I can't tell what the schema was though
[01:01] <jamesh> seb128, daf: wait one day and it will go away
[01:01] <seb128> jamesh: hum, so it's auto-updating now? 
[01:01] <jamesh> (if that is possible)
[01:01] <jamesh> seb128: soon.
[01:01] <daf> jamesh: ok :)
[01:01] <stub> I'll kick off the bugwatch script now
[01:01] <daf> jamesh: well, looks like the rollout we just had fixed it
[01:01] <seb128> ah, you removed the edit feature before doing auto status update
[01:01] <seb128> interesting
[01:02] <jamesh> seb128: Bjorn's status syncing code is now in rocketfuel
[01:02] <jamesh> seb128: when creating an upstream task linked to a bug watch now, the status will default to unknown
[01:02] <bradb> That it's not obvious why you can't edit that task is probably a bug.
[01:02] <jamesh> seb128: the existing bug tasks should get their status/priority/severity fixed as they get synchronised
[01:03] <seb128> ok, cool
[01:03] <seb128> less to do which is good :)
[01:03] <jamesh> that's the idea :)
[01:03] <stub> daf: What OOPS?
[01:03] <daf> stub: https://chinstrap.ubuntu.com/~jamesh/oops.cgi/2006-03-24/A194
[01:04] <jamesh> stub: when I imported some bugzilla bugs yesterday, some of the tasks were using dbschema values that weren't valid in production til the rollout you just did
[01:04] <stub> Traceback (most recent call last):
[01:04] <stub>   File "checkwatches.py", line 82, in ?
[01:04] <stub>     main()
[01:04] <stub>   File "checkwatches.py", line 58, in main
[01:04] <stub>     except externalsystem.UnsupportedBugTrackerVersion, error:
[01:05] <stub> NameError: global name 'externalsystem' is not defined
[01:05] <stub> jamesh: ok. so nothing to worry about.
[01:07] <BjornT> stub: i'll try to add a test and fix it. for now you can simply change externalsystem to externalbugtracker and it'll work
[01:07] <stub> BjornT: I just changed 'externalsystem' to 'externalbugtracker' on production and it seems to be running happily (well... running)
[01:07] <BjornT> cool
[01:08] <stub> BjornT: Good luck implementing a mock bugzilla so you can test it!
[01:08] <kiko> bradb, hang on there
[01:08] <kiko> BjornT, the next time you check in something without running pyflakes on it MURDER WILL BE THE WORD
[01:10] <jordi> uh
[01:10] <jordi> carlos: is anyone doing imports right now?
[01:10] <BjornT> i haven't gotten around installing pyflakes since i upgraded to dapper :)
[01:10] <carlos> jordi: what do you mean?
[01:11] <jordi> the number is shrinking apparently
[01:11] <jordi> 301  -> 350  of 392 results
[01:11] <jordi> oh, or was it like this before?
[01:12] <carlos> jordi: I did some approvals/blocking
[01:12] <jordi> nod
[01:12] <jordi> hmm
[01:12] <jordi> this plone request is suspicious
[01:14] <bradb> kiko: basically, there's not much point doing the magic-status-change fix until it's actually possible to create that situation in the UI
[01:15] <kiko> I agree, bradb 
[01:15] <kiko> and that needs some planning
[01:16] <carlos> jordi: I didn't import it because the name is not the same as the current imported one
[01:16] <carlos> jordi: and it has lot less entries...
[01:16] <carlos> jordi: btw, the potemplate +admin page had a regression and rosetta experts cannot see the extended fields on it
[01:17] <carlos> jordi: kiko gave me extra powers until next rollout so if you need to change anything there, ping me
[01:18] <bradb> kiko: So, can we defer m-s-c until we start talking about release targeting?
[01:18] <jordi> carlos: ok
[01:19] <jordi> carlos: I'll ask hannosch before importing this plone thing
[01:19] <carlos> jordi: cool, thanks
[01:19] <jordi> carlos: dctrl-tools should be in now
[01:20] <kiko> bradb, yeah, cheapest to do so for now.
[01:20] <kiko> bradb, I will spec it out with you for mark
[01:21] <mdke> i sent a bug email while LP was offline during the update, it doesn't seem to have registered. bug mails I've sent since the update have come through. Do mails get queued or should I try again?
[01:21] <kiko> that's a good question. stub?
[01:21] <bradb> kiko: ok
[01:21] <stub> they queue
[01:22] <mdke> stub, how long should I wait for before retrying?
[01:23] <stub> mdke: Should have gone through by now, so now?
[01:24] <stub> I think we poll every three minutes
[01:24] <mdke> stub, ok. Is it worth debugging the problem?
[01:25] <BjornT> mdke: you should re-send it. it seems like your email was lost due to the incoming email script running while the database schema changed or something like that.
[01:25] <BjornT> mdke: you should have gotten a failure message, i'll try to get around implementing that soon.
[01:25] <mdke> okey dokey
[01:26] <siretart> woah. nice work on the soyuz update. I like the 2 column layout in soyuz :)
[01:27] <carlos> jordi: cool, thanks
[01:27] <carlos> jordi: you should take a look from time to time to the FAILED entries 
[01:28] <carlos> jordi: just in case some file imported for productseries failed
[01:28] <carlos> jordi: I will try to add notification emails for those cases
[01:28] <carlos> but atm we are not sending any notification about it
[01:30] <stub> BjornT: If we can't fix the script to leave mail in the queue on nasty exceptions, let me know as I'll need to disable the script as part of the rollout.
[01:36] <BjornT> stub: it's hard, since with some exceptions it's ok to leave it in the queue, but for some it would mean that the queue would stop at the same message every time if we left it in the queue.
[01:37] <stub> BjornT: psycopg.error would probably catch all the database exceptions. If you get one of them, the message should always stay in the queue.
[01:38] <stub> maybe
[01:38] <salgado> mpt, any suggestion to fix https://launchpad.net/products/malone/+bug/977?
[01:38] <Ubugtu> Malone bug 977 in malone "Commenting on bug should optionally subscribe you" [Normal,Confirmed]  
[01:39] <BjornT> stub: this has only happened once as far as i know, so i think it will be enough to simply send an error message to the user, containing the message he sent. until i get around implementing it, if you could easily disable it while rolling out it would be good.
[01:39] <daf> salgado: add a checkbox to the form?
[01:40] <salgado> daf, yep, that's what I thought.
[01:41] <salgado> I wanted to check with mpt and see if we can get that fixed soon, as it already got 4 dupes
[01:42] <daf> salgado-brb: I can take a look at it right now
[01:42] <daf> should be simple enough
[01:56] <jordi> carlos: ok
[02:18] <daf> bradb, BjornT:
[02:18] <daf> when is a bugtask not a bugtask?
[02:18] <daf>         # Make sure we always have the current bugtask.
[02:18] <daf>         if not IBugTask.providedBy(context):
[02:19] <daf> (BugTaskView.__init__)
[02:35] <daf> kiko: yo
[02:39] <daf> mpt: yo
[03:41] <BjornT> daf: BugTaskView is used both for IBug and IBugTask, that's why that check is there.
[03:42] <daf> BjornT: now, if only the comment was as good as your explanation :)
[03:49] <TomaszD> hello, can someone explain to me why are there triple textfields for translations https://launchpad.net/distros/ubuntu/dapper/+source/alacarte/+pots/alacarte/pl/+translate ?
[03:50] <TomaszD> this is a bug, it appears in nearly every package I've encountered, not with all strings, but with some.
[03:50] <TomaszD> here it seems it's with all
[03:50] <TomaszD> what are we (Polish translators) supposed to do with this?
[03:50] <TomaszD> enter the translated text three times?
[03:51] <TomaszD> and three fields appear *always* when there's a singular and plural form to translate. This is confusing and puts our work to a halt.
[03:51] <TomaszD> I hope someone will take a look into this.
[03:53] <daf> interesting
[03:53] <daf> definitely a bug
[03:53] <TomaszD> yes, it's very annoying.
[03:53] <TomaszD> I wish this was fixed asap.
[03:54] <TomaszD> I hope this message will reach someone who is responsible for this.
[03:54] <daf> could you please file a bug?
[03:54] <TomaszD> daf, where and against what?
[03:54] <daf> https://launchpad.net/products/rosetta/+filebug
[03:55] <daf> please include the URL you just mentioned
[03:55] <TomaszD> oh, ok. Thank you.
[03:55] <daf> thanks for reporting it
[03:55] <daf> hopefully we can work out why it's happening soon
[03:58] <TomaszD> daf, https://launchpad.net/products/rosetta/+bug/36428
[03:58] <Ubugtu> Malone bug 36428 in rosetta "Triple textfields for translated text in most Polish templates" [Normal,Unconfirmed]  
[03:59] <TomaszD> should I set it as Confirmed?
[03:59] <TomaszD> I think I'll just do it, anyone can confirm it.
[04:00] <daf> yes, it's confirmed
[04:00] <TomaszD> ok, just did.
[04:00] <TomaszD> goodbye and thanks again.
[04:00] <TomaszD> :)
[04:00] <daf> bye!
[04:00] <daf> you're welcome
[04:12] <carlos> TomaszD: daf; That bug is fixed
[04:13] <carlos> and will be on production with next update
[04:13] <carlos> next week
[04:13] <daf> which bug was that?
[04:14] <carlos> https://launchpad.net/products/rosetta/+bug/31146
[04:14] <Ubugtu> Malone bug 31146 in rosetta "Too many fields in Polish translation of Ubuntu Documentation (quicktour)" [Normal,Fix committed]  
[04:14] <daf> what was the cause?
[04:15] <carlos> daf: completely broken code on DummyPoMsgSet class
[04:15] <daf> ah
[04:15] <carlos> it was returning the number of plural forms no matters if it was a singular form
[04:15] <daf> nice :)
[04:16] <carlos> daf: related to the [None]  * self.pofile.pluralforms OPPS
[04:16] <carlos> that one is also fixed
[04:16] <daf> yeah, I saw the merge
[04:16] <daf> rocking
[04:31] <mpt> daf, pong
[04:32] <daf> I have two screenshots for you
[04:32] <daf> 1: http://people.ubuntu.com/~daf/broken-notification.png
[04:32] <daf> 2: http://people.ubuntu.com/~daf/subscribe.png
[04:33] <daf> for the second one, I'd like (a) your opinion on the UI and (b) your opinion on what to do when the user is already subscribed
[05:03] <mpt> daf:
[05:04] <mpt> (1) please report a bug on the first one
[05:05] <mpt> (2) I've been telling anyone who would listen that we should have a checkbox for subscribing on the bug page, but I think it would be bad to have it under the comment field because that would mislead people into thinking that they needed to comment to subscribe, which would fill bug reports with spam
[05:06] <mpt> (3) Please e-mail me a link to the code you used to create graphs of bugs over time
[05:06] <daf> I don't have such code, I think
[05:06] <daf> only for OOPSes over time
[05:06] <daf> I did have bugs-over-time way back when we were using Bugzilla
[05:06] <daf> I guess I should have snapshotted my bugs DB
[05:09] <mpt> daf, I thought you'd presented it for bugs at one meeting, maybe I was imagining it
[05:09] <mpt> never mind
[05:10] <daf> I wonder if it would be possible to generate it from Malone
[05:10] <daf> given that we have BugActivity
[05:14] <mpt> Well, what I want to do is for every product, project, package, distro, and distro release, cache how many bugs were in each status and each importance level each day over the past year
[05:27] <mpt> that's not much to do with BugActivity (or at least, using BugActivity to calculate it would be excruciating)
[05:28] <bradb> sabdfl: bug 36448, fyi
[05:28] <Ubugtu> Malone bug 36448 in malone "Remove ability to search by milestones when release bug management is implemented" [Normal,Unconfirmed]  http://launchpad.net/bugs/36448
[06:01] <dilys> Merge to devel/launchpad/: [r=lifeless]  Subclass twisted.trial.unittest.TestCase to not leave signal handlers behind. (r3340: Andrew Bennetts)
[06:26] <dilys> Merge to devel/launchpad/: [r=jamesh]  Make the librarian calculate MD5 sums for files, for PersonalPackageArchivesStageOne. (r3341: Andrew Bennetts)
[06:38] <daf> mpt: where should the checkbox be, in your opinion?
[06:53] <dilys> Merge to devel/launchpad/: [r=jamesh]  Tidies and deUbuntifies the code of conduct pages, hopefully fixing bug 31933 (People don't understand how to sign the Code of Conduct). (r3342: Matthew Paul Thomas)
[07:02] <slomo_> are new uploads currently not processed? one of my uploads almost 2 hours ago still doesn't exist in launchpad somewhere else than on my uploads page (gtk-gnutella)
[07:08] <Kinnison> slomo_: one sec
[07:09] <Kinnison>   0.96.1-0ubuntu1  ?
[07:09] <Kinnison> If so, it was published about 4 minutes ago and will be hitting the mirrors etc in the next 30-40 minutes
[07:10] <slomo_> Kinnison: heh ok... sorry for the noise then :) i was just confused because normally things get published faster
[07:10] <Kinnison> I expect you managed to catch it in a cycle before
[07:11] <Kinnison> and we had an upgrade earlier which may have thrown the timing off a bit
[07:11] <Kinnison> it'll settle
[07:11] <slomo_> ok, np :) thanks for your time
[07:11] <Kinnison> No problem, take care
[07:21] <dilys> Merge to devel/launchpad/: [r=jamesh]  fix bug 1350, batch together related bug notifications into a single email. change the mail formatting according to MaloneEmailMessages. (r3343: Bjorn Tillenius, Diogo Matsubara)
[07:22] <jordi> carlos: #canonical
[07:45] <lamont> Kinnison: any clue why /var/run/utmp is in the chroot tarball for the main architectures, but not for hppa (or any other ports architecture, it appears)
[07:46] <lamont> Kinnison: nm
[07:46] <Kinnison> lamont: nope, I'd say "ask infinity" but he's going to bed
[07:46] <lamont> Kinnison: actually, he answered me..
[07:46] <lamont> but yeah, hez sleepy
[07:48] <cprov> Kinnison: how was the rollout ?
[08:03] <dilys> Merge to devel/launchpad/: [trivial]  fix bug 36411 (Layout messed up for "Target Milestone" on the advanced bug search page) (r3344: Brad Bollenbach)
[08:05] <jordi> daf: around?
[08:20] <dilys> Merge to devel/launchpad/: [r=jamesh]  add an email address which will accept moin notifications about wiki changes, and resend them to the corresponding lp spec's subscribers. (r3345: Bjorn Tillenius)
[09:28] <sladen> how do I un-mark a duplicate?
[09:36] <seb128> launchpad should allow to change upstream tasks not pointing to an upstream bugs, other way we have some stucked to a wrong value ...