[05:02]  * thumper afk and gym willing, back alive later
[05:57] <mwhudson> huh, codebrowse for private branches is broken on staging...
[05:58] <mwhudson> and maybe in production?
[05:58] <mwhudson> surely someone would have noticed by now
[06:20] <thumper> mwhudson: you would have thought so
[06:20] <thumper> mwhudson: how broken?
[06:20] <mwhudson> thumper: entirely
[06:20] <thumper> do you know where in the chain it is broken?
[06:20] <thumper> mwhudson: found this out testing your changes?
[06:22] <mwhudson> thumper: yeah
[06:23] <mwhudson> thumper: it's also important to change codebrowse to use safe_open
[06:23]  * thumper nods
[06:23] <thumper> well, on the plus side, I'm not dead
[06:28] <mwhudson> thumper: hooray!
[06:36]  * thumper back after dinner to review mwhudson's branches
[08:22] <adeuring> good morning
[09:14] <mrevell> Morning
[10:22] <thumper> phew
[10:22] <thumper> that's the last five reviews done for today
[10:23] <thumper> hi mrevell
[10:23] <bigjools> yo thumper
[10:23] <thumper> hi bigjools
[10:23] <thumper> bigjools: can you please chase the bzr-builder issue today
[10:23] <thumper> bigjools: once we have that sorted, we can enable build to archive on dogfood
[10:23] <bigjools> thumper: yeah, I'm going to speak to lamont later
[10:23] <thumper> bigjools: and do some real tests
[10:23] <mrevell> Hi thumper
[10:24] <thumper> I'm done, but mrevell, we should have a chat sometime this week
[10:24] <thumper> not now
[10:24] <thumper> mrevell: could you do wednesday evening?
[10:25] <thumper> mrevell: that's your local time
[10:25] <mrevell> thumper, My Wednesday evening should be fine. What time are you thinking?
[10:25] <thumper> mrevell: 8:30pm for you is 9:30am for me, which would be good
[10:25] <thumper> mrevell: before that I have the stand up
[10:25] <thumper> mrevell: but could push it the other way if you like
[10:26] <mrevell> thumper, 9pm UK time or after works better for me
[10:26] <thumper> mrevell: 9pm UK time is good for me
[10:26] <mrevell> great, speak to you then thumper
[10:26] <thumper> ok
[10:26] <thumper> night all
[11:06] <deryck> Morning, all.
[11:30] <jml> good morning all.
[12:10] <mrevell> Hey, intellectronica do you have a moment?
[12:11] <mrevell> or deryck
[12:14] <intellectronica> mrevell: sure, what's up?
[12:15] <mrevell> Hi intellectronica. the bug-heat-view test is failing on my bug heat help branch. Mind if I pastebin you the output? I'm unsure how to adjust the test to take account of my changes.
[12:15] <intellectronica> mrevell: sure, paste away
[12:16] <mrevell> intellectronica, Cheers! http://pastebin.ubuntu.com/422705/
[12:18] <intellectronica> mrevell: i guess the output changed a bit because now you're surrounding it with an anchor or something?
[12:20] <intellectronica> mrevell: try changing "for img in soup.span.contents:" to "for img in soup.span.a.contents:"
[12:22] <mrevell> intellectronica, Oh, cool, thanks. I've been trying to work out how to make it happy with the anchor. Cheers, I'll re-run the test now.
[13:07] <huf> hi. is anyone successfully using the lp software other than you guys? :)
[13:08] <Ursinha> huf, well, ubuntu? :)
[13:08] <huf> i meant a separate installation
[13:08] <Ursinha> ah
[13:08] <Ursinha> that I don't know
[13:10] <Ursinha> bigjools, hi
[13:17] <maxb> huf: No-one's ever mentioned successfully creating another running installation. The image license terms are probably a notable part of that
[14:54] <jml> oh
[14:55] <jml> I just saw a comment from mwhudson, "all subversion-via-cscvs imports seem to be failing"
[15:48] <jml> gary_poster: Is there any discernable difference between a read-only webservice API and a static JSON file?
[15:49] <jml> gary_poster: I'd ask leonardr, but he doesn't seem to be here.
[15:51] <jml> auth & wadl, I guess.
[15:51] <gary_poster> jml, leonardr is out today.  Will be back tomorrow.  So, I'm not crystal clear on your question, but the only distinction I'd make is that, of course, a query string will quite probably matter for a webservice API, but be ignored if a file is server statically.
[15:52] <jml> gary_poster: ok, thanks.
[16:23] <deryck> hmmm, the bugtask table is missing its colors for status/importance now.
[16:23] <deryck> sinzui, related to your CSS refactor maybe? ^^
[16:24] <sinzui> deryck, I am sure it is
[16:24]  * sinzui starts to investigate
[16:24] <deryck> sinzui, was just about to ask if you're doing follow up fixes, or if we need to have it scheduled. :-)
[16:25] <sinzui> I will be landing fixes everyday until I am certain all is well
[16:35] <abentley> mars, I have a bootstrapping problem: http://pastebin.ubuntu.com/422829/
[16:45] <mars> abentley, looking
[16:46] <mars> abentley, did you try 'make clean && make' ?
[16:46] <mars> abentley, and if this was not created with rf-branch, then I assume you ran utilities/link-external-sourcecode ?
[16:46] <abentley> mars, same problem.
[16:47] <mars> a bit confused as to why 'elisa' is in that build chain?
[16:47] <abentley> mars, this is the directory that link-external-sourcecode would use, as its target, so no.
[16:48] <mars> abentley, ah, so this is running 'make' in devel then?
[16:48] <abentley> mars, yes.
[16:48] <abentley> stable, actually.
[16:49] <jtv> intellectronica: we won't get around to that lazr-js upgrade this cycle, I'm afraid
[16:51] <sinzui> deryck, did you report a bug about the colour issue?
[16:51] <deryck> sinzui, I did not.  I can, though.
[16:52] <intellectronica> jtv: that's a shame, it's not that much work, actually.
[16:52] <mars> abentley, 'make' works in devel?
[16:52] <jtv> intellectronica: I was wondering about that... do we have it documented?
[16:52] <sinzui> I will report the bug. The issue is simple and universal. I moved the anchor colour rules to the new style sheet. I need to move all the colour rules new so that they take precedence.
[16:53] <sinzui> deryck, ^
[16:53] <deryck> sinzui, ok, thanks.  if I had filed a bug, this would be the sort of thing to file against launchpad-web?
[16:53] <intellectronica> jtv: i don't think we do (and i must admit i don't remember it exactly to describe it), but we can look it up together if you want.
[16:53] <abentley> mars, I don't have a devel on that box.
[16:53] <sinzui> deryck, In this case it is.
[16:54] <deryck> ok, thanks
[16:54] <jtv> intellectronica: it's getting late here, but I would appreciate a quick collaborative look...
[16:55] <intellectronica> jtv: excellent, let's do that tomorrow my morning?
[16:55] <mars> abentley, was it upgraded to lucid recently?
[16:55] <jtv> intellectronica: let's!  Thanks.
[16:57] <abentley> mars, this box was upgraded to lucid a while ago.  What's new is I ran update-sourcecode for the first time (instead of rsyncing from devpad), and I think it deleted twisted.
[16:59] <abentley> mars, Okay, I went into pymodules and moved elisa out.  That seems to have fixed it.
[17:00] <mrevell> sinzui, Hi! Has your branch changing the CSS files landed?
[17:00] <sinzui> It did
[17:00] <sinzui> mrevell, merge away
[17:00] <mrevell> thanks sinzui!
[17:01] <sinzui> mrevell, I just started a branch to move all colour rules. (this will fix an issue on edge). Does your branch move colours?
[17:01] <mrevell> no sinzui
[17:01] <mars> abentley, :/
[17:03] <mars> abentley, ok, so that was a code isolation failure of some sort.  Love to know where that came from.  I'll keep an eye out for similar failures in the future.
[17:17] <mars> gary_poster, jml, looking at the build-infrastructure list, and I am sorely tempted to use the Wishlist priority (much to sinzui's chagrin :)
[17:17] <jml> mars, priority schmiority -- all I care about is whether they are fixed or not :)
[17:18] <jml> mars, but which ones in particular?
[17:18] <gary_poster> mars, naah.  your wishlist is somebody else's low.  The question is what *you* do.  Other people might tackle some of the other ones you think of as wishlist
[17:18] <mars> gary_poster, true true
[17:19] <sinzui> mars, feature tag + low priority
[17:19] <mars> jml, anything that says "make it ...", "want", or "should" would be fair game
[17:20] <mars> jml, faults first, improvements after
[17:21] <jml> mars, I agree with the second thing you said
[17:21] <mars> jml, sinzui, that really does describe the problem best: looking at this list of bugs, I do not know which are faults and which are features
[17:22] <sinzui> mars, :( My example of a bug is one where we know the rules, this situation violates the rules, so it is a bug. Otherwise this is a feature to introduce new behaviour.
[17:22] <mars> but then you end up with the semantic problems that Gary stated.  Oh well.
[17:25] <sinzui> mars, I think that is an accidental great incite into triage. the semantics are really essential-now, important-next, useful-future
[17:25] <mars> sinzui, yes
[17:25] <sinzui> or insight
[17:26] <mars> sinzui, a fault is anything that keeps it from running today.  Anything else is an improvement or feature.
[17:26] <mars> it becomes easy to decide the grey-area ones then
[17:26] <mars> "Buildbot should be public" - does that keep it from working today?  No?  Improvement!
[17:27] <mars> "Sanitize and document logins used in in pagetests" - Improvement
[17:27] <mars> "LaunchBag.developer is not updated when ftest login() is called" - probably a fault
[17:27] <sinzui> I distinguish priority 1 from 2 based on the severity of the blockage. severity 2 can be worked around, while 1 cannot
[17:27] <mars> interesting
[17:28] <sinzui> An example of that would be ec2 land is broken, but ec2 test with -s works
[17:28] <mars> right
[17:29] <mars> so I call that is a high priority fault
[17:29] <sinzui> I have rated items with work arounds are 1 because the solution requires arcane knowledge
[17:29] <mars> if there were no workaround, it would be critical
[17:29] <mars> hehe
[17:32] <mars> sinzui, I like your idea for three classifications.  I'll see if I can work with that to prioritize.
[21:21] <mars> flacoste, were the stylesheet problems on edge last week fixed?
[21:29] <flacoste> mars: yes
[21:29] <mars> flacoste, thanks.
[21:30] <mars> awe... sinzui, ping.
[21:31] <sinzui> hi mars
[21:31] <mars> hi sinzui, are the links to bug statuses on this page blue for you? https://bugs.edge.launchpad.net/launchpad-foundations/+bug/570380
[21:31] <mup> Bug #570380: ec2test sometimes hangs on the first windmill test <build-infrastructure> <Launchpad Foundations:Triaged> <https://launchpad.net/bugs/570380>
[21:33] <sinzui> mars, not in the branch I just pushed.
[21:33] <sinzui> do you want to review it?
[21:33] <mars> sinzui, I would not feel confident doing so
[21:33] <sinzui> mars: bug 570259 is the global issue
[21:33] <mup> Bug #570259: css colours are wrong after style sheet reorganisation <css> <launchpad-web:In Progress by sinzui> <https://launchpad.net/bugs/570259>
[21:34] <sinzui> mars, all the colours in the old style sheet (mostly enums) lost precedence when I moved the anchor rules to style 3.0
[21:53] <mars> mwhudson, what do you think of this?  https://dev.launchpad.net/Debugging/GDB
[21:54] <mars> mwhudson, I added a section for backtrace.py to the bottom.
[21:56] <mwhudson> mars: looks good, you don't need the .gdbinit file though
[21:56] <mars> mwhudson, oh, so "the .gdbinit should also work" was saying that backtrace.py does roughly the same thing as the GDB macros?
[21:57] <mwhudson> mars: yes
[21:59] <mars> mwhudson, updated
[22:05] <thumper> morning
[22:06] <mars> morning thumper
[22:30] <mars> It would be nice to have an easy way to copy the short "http://launchpad.net/bugs/IIII" link, instead of having to type it out by hand.
[22:30]  * persia remembers an exceeding contentious change that removed just that feature
[22:32] <persia> I believe the argument was that if a user was already on a page, there was no point in having a UI element that would link back to the *same* page, and people who wanted to copy could always do it from the location bar.
[22:37] <mars> persia, fair enough.  https://bugs.edge.launchpad.net/launchpad-foundations/+bug/570380 is not fun to copy-n-paste though :)
[22:37] <mup> Bug #570380: ec2test sometimes hangs on the first windmill test <build-infrastructure> <Launchpad Foundations:Triaged> <https://launchpad.net/bugs/570380>
[22:38] <mars> persia, I am also a fan of StackOverflow's titles:  stackoverflow.com/QESTIONNUMBER/some-human-readable-question-title
[22:39] <persia> Worse yet, because of the embedded "edge" it fails completely to be useful for lots of folks.
[22:39] <mars> :)
[22:39] <persia> This is part of why the bugbots do what they do, so one can usually say bug #570380 and get a useful response.
[22:39] <mup> Bug #570380: ec2test sometimes hangs on the first windmill test <build-infrastructure> <Launchpad Foundations:Triaged> <https://launchpad.net/bugs/570380>
[22:39] <persia> Doesn't help so much with mail though.
[22:46] <jml> persia: the change I remember more is when we removed the non-linked bug number
[22:46] <jml> persia: which caused an outcry because there was no longer an easy way to copy-and-paste it.
[22:47]  * jml disappears
[22:51] <wgrant> persia: Fails completely? edge works for everyone.
[22:51] <wgrant> (but yes, it's suboptimal)
[22:51] <persia> wgrant: Fails completely *to be useful*
[22:52] <persia> This isn't "fails completely"
[22:52] <persia> jml: I thought I remembered complaints about it changing both times (and yes, current behaviour matches very old behaviour)