/srv/irclogs.ubuntu.com/2009/08/27/#bzr.txt

lifelesspoolie: would you like to have lunch?00:06
lifelesspoolie: today I mean00:06
pooliemm00:06
pooliei would00:06
pooliei have to finish circa 500:06
poolierather than 1 :-O00:06
lifelessthis sounds compatible so far00:06
pooliethe bad 1 o'clock that is00:07
lifelessI'll be finishing ~ 2 I imagine00:07
lifelesswish I knew why i'm waking so early00:07
lifelessbut it sure is peaceful to be hacking - no trucks or traffic00:07
lifelessanyhow, if you'd like to have lunch, so would I. So we should00:08
pooliesure00:19
pooliepick a place00:19
pooliemeeting here would let me work more00:19
pooliemeeting somewhere at the end of a motorcycle would add joy to my life :)00:19
lifelesspoolie: you're not seriously asking me to choose between you working or having joy?00:21
lifelesspoolie: thats rather faustian :)00:22
lifelessuhm, I could go pie. Or vietnamese. or $other00:22
poolieone of us needs to choose :)00:22
lifelesslets say pie00:22
pooliei am enjoying my work00:23
poolieok, here?00:23
igcpoolie: yes, bzr-mac00:23
pooliewfm00:23
igcmorning all00:23
pooliegreat00:23
lifelessdone00:23
lifeless12ish?00:23
poolieigc, i was thinking about platform-specific bugs00:23
lifelessor is that too early for your stomach00:23
pooliei suspect non-core people aren't subscribed to bug mail, or don't read it00:23
pooliein fact core people probably don't read all of it00:23
poolieshould we maybe explicitly forward or subscribe platform teams to relevant bugs?00:23
poolielifeless: wfm, i woke about 6.3000:25
igcpoolie: maybe. I'm not sure if there would be any backlash against that or not. We could ask if anyone objects?00:28
poolieok00:29
igcpoolie: there's certainly value in asking the teams to test platform specific fixes00:29
pooliei've had some success in the past pointing them out on the main list marked with eg win3200:29
pooliei guess the question is whether they're platform test teams or platform dev teams00:30
igcpoolie, jml: can users/teams subscribe to bugs with certain tags?00:30
jmligc, I don't believe so. It's quite possibly a filed bug too00:31
lifelessvila: still here?00:34
poolieigc, i was wondering the same thing00:34
poolieit's probably better to just subscribe the team00:34
pooliei think it's probably not appropriate to subscribe the whole team to all relevant bugs00:35
poolie(maybe they want it)00:35
pooliebut we could make ~bzr-win32-bug-team00:35
lifelesspoolie: do you want developers or qa-on-fixes00:36
pooliedevelopers00:36
lifelesspoolie: if its qa on fixes I'd suggest ~bzr-qa-PLATFORM00:36
poolieprimarily00:36
pooliei think some of these bugs are shallew00:36
pooliebtw did you have any luck with wine in the case we mentioned yesterday?00:37
lifelesspoolie: no, it was on my todo to experiment today, but late last night bialix said it works00:37
pooliehow sad :)00:37
lifelessits terrible fixing bugs and stuff00:39
lifelessbialix is going to try and get a 1.18.1 prepared with the windows fixes00:39
lifelessWhats really a shame is that all of them were very shallow.00:40
lifelessdoes buildbot actually run ascii tests specifically at the moment?00:41
pooliei don't know00:49
pooliei don't think so00:49
pooliebut it sounds like vincent will add it soon00:49
poolielifeless: oh, actually unfortunately i can't do lunch today, i have to do something for s00:55
lifelessok00:55
lifelessno wakas00:55
pooliemonday could be good, let's talk then?00:55
lifelessfriday or tuesday would be better for me00:57
pooliek00:58
igcpoolie: can you clarify vila's email last night ...00:58
pooliemm, which one?00:58
poolieabout the rc? just done00:58
igcpoolie: the one that said rc1 vs beta100:58
igcand lrelease date early next week00:58
poolieah, i didn't see that about the release date01:00
igcpoolie: so to confirm, rc2 next week01:00
igcpoolie: and 2.0 the week after01:00
igcif 2.0 is next Monday, that impacts what I'll do today!01:00
lifelesspoolie: would tomorrow work foryou?01:04
poolieigc: yes01:04
pooliewas i insufficiently clear about what i wanted to happen or was vincent just a bit random here?01:04
poolieprobably both01:04
lifelesspoolie: I think unclear; I know I've been unclear01:05
lifeless*I've been unsure*01:06
lifelessI was asking you yesterday whether this should be done before we're finished with the critical bugs or not, and you said not-before, but this seems to have happened before they are fixed : I don't really know where we stand.01:06
igcpoolie: I thought it was clear-ish but it's a change in process and we'll still finding our feet wrt monthly vs otherwise01:07
igcpoolie: 1.18 is yet to be announced for example IIUIC01:07
poolietrue01:07
pooliewhat's "this"?01:07
poolie2.0b1? at that time all of the 2.0 critical bugs were closed, i believe01:08
lifelessdoing a 2.0 b101:08
lifelesspoolie: jam is working on the fragmentation issue, I'm working on the pushing of 1.9 stacking against 2a issue01:08
lifeless415508 and 402645 were still open01:09
pooliebug 41550801:09
ubottuLaunchpad bug 415508 in bzr "Content filtering breaks commit w/ merge parents" [Critical,In progress] https://launchpad.net/bugs/41550801:09
poolie!01:09
poolieit shouldn't be01:09
lifelesspoolie: also when you said critical I heard 'targeted to 2.0' not 'critical'01:09
lifelesswhich is a hearing defect01:09
pooliewe have non-2.0 bugs marked critical01:11
poolieum01:11
lifelesspoolie: I'd like to either untarget, or bump to critical, all the non-critical, non-fixed, bugs targeted to 2.001:11
lifelessI think it will make sure you and I are on the same page01:11
pooliei'd like to either degrade or target all the critical bugs01:11
poolieso the non-critical bugs targeted to 2.0 are meant to be things we'd like to fix, or would consider merging01:11
pooliethis might be an unhelpful distinction compared to just looking at the high/critical lists generally01:12
poolieso, um01:12
pooliei think both those changes are good01:12
pooliebut let's do them one-by-one not wholesale01:12
poolieand then if we find a case that seems to invalidate the approach we'll reconsider it01:12
lifelessI strongly desire that targeted means 'will delay release'01:12
pooliemm, me too01:13
lifelessnot having them critical seems to make them mean something different to you01:13
pooliei found that yesterday required more thinking than was optimal01:13
poolieor more thinking of a kind of overhead category01:13
lifelessok bug 41673201:13
ubottuLaunchpad bug 416732 in bzr "check reports "Missing inventory {('TREE_ROOT'..." for trivial non-rich-root branch" [High,Confirmed] https://launchpad.net/bugs/41673201:14
lifelesstargeted to 2.0. SHould block it IMO01:14
AfClifeless: that's what "blocker" means in GNOME. Actually, "blocker" (if confirmed as such) means «holy shit, stop everything, and there will be a release about 30 minutes after we figure out the problem»01:14
pooliebug 402645 you mentioned before01:14
ubottuLaunchpad bug 402645 in bzr "fetch for --2a formats seems to be fragmenting" [Critical,In progress] https://launchpad.net/bugs/40264501:14
lifelesspoolie: lets do this over the phone; it will be quicker01:14
pooliemm01:14
poolielet me finish something here first01:14
pooliethen i'll call01:14
lifelessok01:15
AfClifeless: so yeah, I guess that's what "critical" means there too01:15
AfCit's all very informal01:15
poolieso the key question is01:16
poolieshould there be a category of 'wannabe blocker' or 'kinda blocker'01:16
lifelessAfC: there are two dimensions here; one is 'before this release', the other is 'what should a dev look at next'01:16
poolieit seems to cause mental pain for the RM and may have no benefit01:16
lifelesspoolie: I don't think there should be; it should be a blocker, and if someone disagrees they stop it being a blocker.01:16
lifelessmake the decision and make it easy to change if wrong.01:17
mneptoklifeless: the answer to the second question is always "my manboobs"01:17
lifelessmneptok: no, never. I'm sorry, but goram-no.01:17
AfClifeless: well, the later is task prioritization, and I've never found bug trackers are good for that. Anyway, topic for beer.01:17
mneptoklifeless: so much for the adventurous and brave spirit of New Zealanders.01:17
lifelessmneptok: ba!01:17
poolielol01:17
poolielifeless: your phone seems engaged, call me when you're fere01:25
pooliefree*01:25
poolieas in beer01:25
lifelesspoolie: doing01:39
lifelesspoolie: your phone rang out01:40
* lifeless tries again01:43
emmajanepoolie, ping?02:06
emmajanebeer?02:06
emmajaneI like beer.02:06
emmajanemneptok, stay classy :)02:07
* emmajane reads backwards in time.02:07
emmajanepoolie, reply sent. :) I'll be back in a bit for a little bit.03:05
jamlifeless: you only get knit-delta-closure etc if you pass "include_delta_closure=True" which doesn't happen for normal fetches, only when someone wants fulltexts.03:17
jamso you end up getting every little bit of content as a separate 'substream' in a normal fetch03:17
jam2a would be different03:17
jamthough I think it would be 1 group per substream03:17
jamwhich still isn't great03:18
jamand means that the target repository won't really have a chance to do batching03:18
jamsince it won't get all the groups in a single 'insert_record_stream'03:18
jamyou would have to batch between the streams...03:18
spivjam: it's still the same write group, though.03:18
jamspiv: sure... still sort of violates what I thought substreams were abouet03:19
AfCjam: wish you were here. Then I could buy you a coffee.03:19
spivWell, it's explicitly allowed for there to be multiple substreams of the same kind.03:20
spivIt would be more elegant to turn the stream from the request into a continuous object stream, but Python doesn't make that easy to do.  Well, we could add more threads...03:23
jamspiv: sure. but 1 record at a time is not really a bunch of *streams*, but more, a bunch of records... :)03:23
jamalso, Repositories know about write groups03:23
jamVF's know about record streams03:23
jamso you would have to figure out how to have the VF know about the current Repo's write_group03:24
jamwhich is a small inversion03:24
jamor have Repo.commit_write_group() tell the VF.flush_your_buffer()03:24
spivWell, for 1.9 pack repos I don't think there was any practical difference between insert_record_stream([2 records]) and 2 insert_record_stream([1 record]) ?03:24
jam(and dump_your_buffer for abort)03:24
spivBecause it's the write group that allocates the pack file.03:24
jamspiv: present day there isn't for 2a03:25
jamI would *like* to regroup on-the-fly03:25
jamgiven the discussion about fragmentation and recombination03:25
jamI can't regroup if everything I see is a single record-at-a-time03:25
spivAh, I see.  I was trying to figure out why you cared :)03:25
AfCbzr 1.18 is out or will be immenently right? (ie, it's ok for me to "advise" people that they really ought to be using bzr >= 1.18 on a website & in release notes for something I'm cutting today or so?)03:34
jamAfC: 1.18 is out, the announcement hasn't hit pending the building of all packages for all platforms03:35
jamthe code is cut, though03:35
AfCk03:36
AfCclose enough03:36
poolie 2009/8/27 Mary Gardiner <mary@puzzling.org>:03:45
poolie> Why say "can we upgrade this to critical" when you mean "can we fix this?"03:45
poolie>                                    -- Martin Pool, twitter.com03:45
pooliegreat sig :)03:45
jmlheh03:46
spivHeh.03:46
fullermdWait.  We're supposed to FIX them?03:47
lifelesspoolie: I've just mailed the list with I think the pith of our chat03:50
lifelessspiv: jam: it would be quite easy to daisy chain all the substreams of the same kind together in the insert_stream method03:51
poolie>  "Do not assign bugs to other people unless you are confident they will work on them." sounds like some people can avoid getting bugs assigned to them altogether :)03:55
pooliebut no one on this team :)03:55
lifelesspoolie: if you assign a bug to me, its == 'hey robert, please fix this for me'03:57
lifelessand I'm very unlikely to unassign it immediately.03:57
* igc lunch03:57
pooliemm03:57
pooliei'll reply in mail03:57
lifelessI've seen fairly regularly in other areas, assign-to-someone being used by users to get their bug looked at03:58
lifelessI think that was before the permissions on the assigned field were straightened out03:58
pooliewhat is 'staying stay' supposed to mean?04:00
lifelessblah, typo04:02
lifelessjust switching folders.04:02
lifelesspoolie: staying stale04:06
lifelesspoolie: as data:  https://bugs.edge.launchpad.net/~mbp/+assignedbugs :)04:07
mwhudsonwould bzr reconfigure --append-revisions-only be a useful feature?04:08
lifelessmwhudson: yes!04:08
fullermdOnly I'm not sure I'm sanguine with it in reconfig.04:09
mwhudsonlifeless: is there an easy way to set it at the moment?04:09
lifelessmwhudson: vim04:09
lifelessfullermd: got a better place?04:10
fullermdUnfortunately, no.  We should have a place for a lot of things like that.04:10
fullermdAnd reconfigure is a great command for it.04:11
fullermdBut it's currently fairly purely "reshuffle the arrangement of bzrdir components", and I don't like mixing that with "set branch config options".04:11
fullermdmwhudson: init has a flag for it, so check the cmd_init code.04:11
lifelessfullermd: and repository options04:11
lifelessmwhudson: api wise its easy04:11
poolielifeless: i know04:12
lifelessbranch.get_config().set_user_option()04:12
pooliethey're some of my favorite bugs04:12
fullermdYah, I know.  I put that one there, remember  :p04:12
poolieit may be too large of a queue but they're also things that i want to do in my slack time, if any04:12
lifelesspoolie: o/~ these are a few of my favourite things04:12
fullermdStill a bit uncomfortable with it being there, though.04:12
poolieit may be there's some better place to store that04:12
lifelesspoolie: Would you be happy having an mbp tag on the tracker?04:12
pooliemm04:12
pooliemaybe04:12
poolietags are still harder to use04:13
pooliethere is this 'bug list' feature that's meant to do exactly what i want here04:13
poolieno idea when that will be done though04:13
lifelessI get concerned when I do something with someone else assigned to it.04:13
lifelessto the point that I don't do those things.04:13
lifelessor do them less often04:13
poolieok04:15
pooliethat's definitely a problem04:15
pooliesome of this stuff is like working in python04:15
poolienice, but you have to work with what you've got04:15
lifelesss/.*, //04:15
poolieie encoding all the bugs as tuples :-/04:15
lifeless:P04:16
lifelessI like that we can file bugs on launchpad in launchpad.04:16
lifelessthumper: will pyconz be briliant or just ok?04:49
thumperlifeless: I have no idea04:55
thumperlifeless: I expect it to be at least ok04:55
lifelessvery close to uds04:55
lifelessI'm trying to decide if I should come to pyconnz and give a lightning talk on fixing unittest04:56
lifelessamongst other things04:56
poolielifeless: i think you have a good point about assignment actually05:03
pooliei've just had enough metadevelopment for today05:03
lifelesspoolie: thats fine, I have too05:04
lifelessI was expecting a 3 minute call :P05:05
poolie:-)05:05
pooliespiv, igc, either of you want to catch up?05:16
poolienow would be good for me05:16
spivpoolie: sure, I'll call05:17
lifelesspoolie: I've EOD'd05:17
RobOakesHi, I just started using bzr (and I've absolutely loved it, especially the bzr-svn module), but had a couple of new user questions.  Is this a good place to ask them?05:24
RobOakesI created a test branch from an SVN repository and then posted it to my server.  From there, I made a "working" branch that I could experiment with.  However, whenever I post anything to the working branch, it also posts to the main branch on the server via SFTP.  Is that normal?  And if so, is there any way to disable it?05:27
fullermdRobOakes: How did you make the working branch?05:45
RobOakesThe first branch I created by using bzr push from within the SVN working copy.  It took four days, but eventually it created a bzr branch of the code.05:46
RobOakesThe "working" branch was created from that initial export using bzr branch /path/to/original05:47
fullermdWhat does 'info' tell you in the working branch?05:47
mtaylorlifeless: ooh, when's pyconnz?05:47
thumpermtaylor: you mean kiwipycon?~05:48
RobOakesIt says "checkout of branch sftp://path/to/branch"05:48
RobOakesRelated branches push branch: sftp://path/to/branch05:48
thumpermtaylor: http://nz.pycon.org05:48
* mtaylor is always looking for excuses to go places...05:49
thumpermtaylor: where are you based?05:49
fullermdRobOakes: Right.  So you didn't make a branch, you made a checkout (either by running 'checkout' instead of branch, or later using 'bind' or one of its relatives).05:49
mtaylorthumper: seattle05:49
* mtaylor now notices that CFP is open for US pycon05:49
thumpermtaylor: we'd love to have you :)05:49
RobOakesOkay.  Is there any way to unbind it?05:49
fullermdRobOakes: You want to turn it into a separate branch if that's what you want, with e.g. 'unbind'.05:49
thumpermtaylor: it is just a weekend though05:50
thumpermtaylor: we're only just starting out05:50
RobOakesVery, very nice.  I just tried a test commit and it looks like everything is working.  Now, when I want to update the remote branches, I use bzr push, right?05:50
mtaylorthumper: I'd love to come! you know, I'm sort of low on flying miles this year...05:50
thumpermtaylor: need to keep that gold membership?05:50
fullermdRobOakes: Right.05:51
mtaylorthumper: yes!05:51
mtaylorthumper: if I lose it, they'll stop upgrading me places, and that would make me sad :(05:51
thumpermtaylor: you could talk to me in person about code reviews05:51
mtaylorthumper: oooh. now there's a business reason to come05:52
RobOakesGot time for a second question?  Because of reasons too complicated to go into, I do most of my development work on a linux virtual machine.  However, the actual files are stored on the physical hard drive and accessed through an interface.05:52
RobOakesFor reasons I don't understand bzr seems to think that the file permissions change, quite frequently.  This leads to very big (and pointless commits).  Is there a way to tell bzr to ignore changes in the file permissions?05:52
fullermdThe only permission bzr watches is the +x.  That really shouldn't be changing around unless you're flipping it...05:53
lifelessmtaylor: mid nov05:54
fullermd(I mean, depending on just what the 'interface' is, it may not mean much, but it shouldn't be _changing_)05:54
RobOakesI'm not, but the way vmware handles it's remote directories is weird.  Sometimes, they mount as 777 and sometimes they mount with other permissions.05:55
RobOakesI still haven't quite figured out the overall method to the madness.05:55
fullermdThat's...    pretty insane.05:55
fullermdAnyway, making bzr ignore it would really be hackaroundish.  Wouldn't be my choice of how to handle it.05:56
fullermdBut given that, no, AFAIK there's no existing config option to say "ignore +x totally plz".05:56
RobOakesAgreed.  My plan has been to watch the permissions and only commit when they are 777, this prevents the nasty problem.05:56
fullermdMaybe you could thunk something in with a plugin, but that's way outta my league.05:56
RobOakesI may think about doing that.  My first desire would be for VMware to tell me why the permissions in the vmfs module are all crazy.05:57
fullermdVague memory tells me there is (or was) a plugin to try and allow some +x handling on win32; that might be a place to start.05:58
RobOakesI'll take a look at that.  I'll look through the devel mailing list and see if I can find anything about it.05:58
spivMaybe write a plugin or shell alias that checks that a) a file that should not be +x is indeed not, and b) a file that should be, is.  Then abort/warn before commit as appropriate?05:59
fullermdOh, good, somebody competent is here.  I'll leave you with spiv    8-}05:59
RobOakesThat would help solve the problem, and wouldn't be too complicated.06:00
RobOakesAnyway ... thanks for the help.  That clears up most of the miscellaneous questions I had.06:00
spivThat would catch any sort of weird "all mounted files get +x"/"all mounted files lose +x" condition.06:00
spivAnd hopefully you can fix the root problem :)06:00
* spiv grabs a late lunch.06:00
RobOakesAgreed.  What's funny is that I don't seem to have the same issue with my svn working copies.  (But I also haven't spent much time testing it.  I figure it's a weird VMware thing.)06:01
vilahi all06:54
vilalifeless, poolie : the test are run with and without locale for quite a few days/weeks06:55
vilapoolie: argh, sorry about the rc/beta misunderstanding :-(06:55
* vila checks mail06:55
lifelessvila: thanks06:56
poolieit's ok06:56
poolievila, what was the source of the confusion?06:56
vilahmm, I think I fired qlog to search for examples of updates for version_info and could only found rc examples of course06:57
vilayou certainly mentioned 'beta' but it didn't print hard enough in my brain to override at that point :-/06:58
vilaadd to that some stress due to bugs encountered while setting up my 2.0 branch and apport masking me the real cause, etc, => small mistake, big consequences :-/06:59
poolieok07:01
pooliei thought that might have been it07:01
pooliethe examples in the docs are a bit of a two-edged sword07:01
pooliewithout them it would be vague but it's easy to mentally or literally copy & paste them07:01
vilawhile discussing with jam, I still wasn't sure about my mistake as 'cycle.txt' doesn't mention 2.0beta1 but mention 2.0rc107:02
vilaso it's still unclear whether 2.0 is different because it starts the new cycle07:02
pooliejam, you rejected your merge sort bug?07:12
vilapoolie: re-reading releasing.txt, memory comes back also: the document is clear about final releases, less so about candidates and says nothing about betas, I'll outline everything I found by mail07:13
poolies/bug/mp07:13
vilapoolie: I saw it too, but I think he wanted to work on a second part instead and I'm pretty sure he send a submission to pqm07:13
lifelesspoolie:07:19
lifeless  File "/home/robertc/source/baz/pending/working/bzrlib/tests/test_selftest.py", line 2645, in test_run_bzr_user_error_caught07:19
lifeless    'bzr: ERROR: Not a branch: ".*nonexistantpath/".\n')07:19
lifelessAssertionError: pattern "bzr: ERROR: Not a branch: ".*nonexistantpath/".07:19
lifeless" not found in07:19
lifeless"""\07:19
lifelessbzr: ERROR: Attempt to escape test isolation: 'file:///nonexistantpath/' []07:19
lifeless"""07:19
jml:(07:21
igcbbl - some errands to run07:21
lifelessjml: this is good07:21
lifelessjml: its a step towards not making 4 directories on every test07:21
jmllifeless, I was reacting to a costly mistake that bzr let me make07:22
jmlmake lots and lots of tedious changes, shelve them, merge in latest trunk, forget to commit, unshelve them, realize they're all mixed together, revert, sigh07:23
lifelessjml: oh :(07:23
lifelessjml: did you lose the changes?07:25
lifelessjml: perhaps you can uncommit, or maybe we kept a backup for you?07:25
jmlbackups would end with '~', right?07:27
jml... and the modified files aren't open in my editor.07:27
lifelessmmm, they would have it in the name07:27
lifelessis there a backup of the shelf in .bzr/checkout/**07:27
lifelessjml: we should have kept backups for sure, grep for **/*~*07:28
lifelessem, find . | grep .. or $something like that07:28
jmlno backup in .bzr/checkout07:29
jmlthe backups look like they are of the wrong changes07:30
jmlI want what was in the shelf07:30
lifelessplease file a bug about this; we may tease it into several bugs07:31
lifelesshowever, the backups are numbered07:31
lifelessare you looking at the right ones07:31
jmlthere's only ~1~07:31
jmlno others.07:31
lifelessok, two bugs07:32
lifelessrevert is meant to back things up07:32
lifelessunless you've set no backups07:32
vilapoolie: ouch, only 2.0rc1 appears on https://edge.launchpad.net/bzr in the download part, that's not true for https://launchpad.net/bzr though07:43
pooliei like the pedantry of the tooltip on RDF07:44
poolievila, you're saying it should show the last stable release too07:44
pooliesounds like an lp bug07:44
vilayes07:44
poolieoh, sigh07:44
pooliewill you file it or should i?07:44
vilawell, had I made a beta I think the problem would have been the same... please do, I'm already filing  another:-/07:45
poolieoh, which one?07:45
pooliei'm going to finish soon btw07:46
pooliefor a change07:46
vilaTooManyConcurrentRequests when pulling from lp in a lp bound branch07:46
vilalifeless: bug #309309 was the one I was referring too yesterday in my review07:47
ubottuLaunchpad bug 309309 in bzr "selftest can't be run in a bound branch" [Undecided,New] https://launchpad.net/bugs/30930907:47
AfCLaunchpad is pathetic. Again. "The URI scheme "bzr" is not allowed.  Only URIs with the following schemes may be used: bzr+ssh, ftp, http, https, sftp" when trying to give it a bzr:// branch.07:48
pooliesomeone just filed a bug on that07:50
poolieit may be a recent regression07:50
AfCI mean, shit, even Ohloh manages to get that right.07:50
AfCpoolie: (no, I don't think so. I recall trying to register the proper branch URI for my mainline about a year ago and it didn't work. Back then it _also_ crashed my browser when I tried to file a bug about it, which was exceptional. But now it just errors)07:51
pooliewell, sorry07:51
lifelessAfC: I think you're being a little harsh; that code tries to be very paranoid (and thats arguably a mistake), but pathetic is a little strong.07:51
AfCpoolie: {shrug} I'm not about to hold their behaviour against you.07:52
AfCpoolie: but this is yet another feather in the cap of why I think it so important to encourage people to realize they don't need to use Lauchpad to work with Bazaar.07:52
spivYeah, I think that's always been the case.  I think there was/is some reluctance to open up unnecessary ports.07:54
AfClifeless: perhaps, perhaps not. Given that I've registered an external branch as a project's mainline I don't really see it as necessary that Launchpad advertises "to get this do `bzr branch lp:myproject`. It's like? Huh? That's a mirror, and a 5 hour out of date mirror at that. How about Launchpad advertise `bzr branch bzr://...` like I asked it to.07:54
AfClifeless: things like that, again and again07:54
spivSo my recollection is that refusing bzr:// is not because LP doesn't understand that prefix, but because it knows it can't work.07:54
AfClifeless: me registering a release series at a certain well known foreign location and then Launchpad claiming it's where the releases are hosted? That's not cool either.07:55
lifelessAfC: for some folk its great, because they don't want lots of people pulling directly. Maybe there should be a toggle.07:55
AfCspiv: that's interesting. Bizarre to my eyes, but certainly informative. Of course, if bzr:// doesn't work maybe someone should take it out of bzr and I'd stop using it.07:55
lifelessAfC: it works, its just firewalled @ canonical.07:56
spivbzr:// works just fine, in general.07:56
lifelessAfC: is there a bug on this?07:56
AfClifeless: classic07:57
spivOperationally, the way Launchpad is currently deployed, Launchpad cannot access that port on arbitrary hosts.07:57
AfC[and you said I was being harsh?]07:57
spivIt's probably worth revisiting that decision.  At the time there were very few TCP-only bzr branches out there, so it wasn't a big deal.07:58
spivAnd, co-incidentally, someone did file a bug about it recently as mentioned above, so probably someone will be revisiting that decision soon.07:58
AfClifeless: I can try and file one if you like. Launchpad's javascript doesn't seem to be crashing Epiphany anymore.07:58
spivOh hey, another Epiphany user.07:59
AfC:)07:59
AfCspiv: of course. GNOME integration, baby07:59
poolievila: i don't think you put the changelog in the2.0rc1 release on lp?08:01
poolievila: i filed bug 41973508:01
ubottuLaunchpad bug 419735 in launchpad-registry "downloads page harshly truncates series descriptions" [Undecided,New] https://launchpad.net/bugs/41973508:01
vilagood08:02
poolie(not so important)08:02
poolieand more importantly bug 41973308:02
ubottuLaunchpad bug 419733 in launchpad-registry "new Downloads portlet should show latest stable release as well as latest" [Undecided,New] https://launchpad.net/bugs/41973308:02
poolieand reopened bug 27567708:03
ubottuLaunchpad bug 275677 in launchpad-registry "downloads should be grouped by release, not series" [Low,New] https://launchpad.net/bugs/27567708:03
poolievila, also we should probably change the process not to upload source zips08:05
pooliei misunderstood the request when i added that08:05
lifelessvila: checking the safety net appears to be 1-2 ms08:05
lifeless(*20K = most of a minute)08:05
pooliepeople actually wanted zips of windows binaries08:05
vilachangelog... oooh, I just understood how it works, but that's not mentioned in releasing.txt and I think at the time I filed  http://bazaar-vcs.org/Download I was already worried because the bugs didn't appear08:06
poolielots of bugs appear :)08:06
pooliewhich ones did you miss?08:06
vilaI mean they weren't targeted at the milestone08:06
pooliemm08:07
lifelesspoolie: I'm wondering if your fix to show every test is the right way to fix the test progress bar08:07
vilaalso creating the milestone was a key time in my mistale08:07
poolierather than what?08:07
lifelesspoolie: simply because if tests are very fast the progress bar starts to dominate :)08:07
pooliesrsly?08:07
poolienice08:07
lifelessI'm closing in on that08:07
lifelessRan 203 tests in 2.975s08:07
* jml is reminded of Ubuntu boot discussions08:08
pooliepresumably because of the time to paint it, not to calculate the text08:08
pooliewhat's the delta if you run with =none08:08
lifeless=none ?08:08
lifelesshow would I spell that08:08
vilaBZR_PROGRESS_BAR=none ?08:08
poolieyes08:08
lifeless./bzr --no-plugins selftest selftest08:08
lifelessok08:08
lifelessRan 203 tests in 2.881s08:09
lifelessabout 0.5ms per I guess08:10
lifelessso not huge yet08:10
* lifeless ignores for now08:10
lifelessmmm, but 5% of test time08:11
lifelesswhen it gets up a bit higher I'll worry08:11
poolie3% according to my calculator08:12
vilalifeless: why not ./bzr selftest -s bt.test_selftest -s bb.test_selftest ?08:12
pooliebut it seems like these are atypically fast tests?08:13
lifelessvila: I have a lot of plugins I don't want loaded for this; and its fast enough without -s that I've never learnt to use it.08:13
pooliei reckon the overall overage is still above 50ms08:13
bialixdoes anybody knows why merge shows many files as modified in its output, but status after merge shows there is changed only file or two?08:13
lifelesspoolie: they weren't when I started08:13
lifelessbialix: you're merging in things that have patches you already have08:14
lifelessbialix: so the change ends up being nothing08:14
bialixI don't understand08:14
bialixwhy merge shows me incorrect info08:14
pooliethere's a bug that merge should only show files as modified if they actually changed08:14
bialix?08:14
pooliebecause they 'might have changed'08:15
lifelessbialix: because it shows the files that had changed on branch A08:15
pooliebut the changes didn't add up to anything08:15
lifelessbialix: but in your branch, B, they already have the same textual changes08:15
bialixlifeless: about your fixes for win32 and locks08:15
bialixif I merge with --weave I don't have conflicts08:15
bialixbut resulting diff after merge is scarely short08:15
lifelessbialix: for example; if you and I both change "print foo" to "print bar" in different branches, if I merge from you I will get your change, but end up with no difference in the file.08:16
bialixlifeless: I think this is a bug in merge08:16
lifelessas poolie says there is a bug open.08:16
bialixok08:16
bialixlifeless: back to your 2 revisions with lock fixes: do you want to review them?08:16
bialixafter I cherrypick them with weave?08:16
lifelessI'm not entirely sure that this is wrong though, I mean, it would be odd to merge what you thought were different branches and see no output.08:16
lifelessbialix: if you'd like me too, I'm very happy to.08:17
bialixlifeless: for me odd thing is that after merge status shows there is nothing changed, while merge shows there is a lot changes08:17
bialixit makes me nervous08:18
luksyou can always diff against both parents and see if the diffs look the same08:18
lifelessbialix: I think it should say that this has happened08:19
lifelessbialix: being totally quiet would be wrong; looking like a normal change and st showing nothing all looks wrong.08:19
lifeless/all/also/08:19
bialixluks: I can, but only if I know that bzr lies me08:19
bialixlifeless: revno 4635 NEWS has completelly unrelated entry:08:20
bialix* Fix a test failure on karmic by making a locale test more robust.08:20
lukslifeless: well, if merge doesn't change anything in those files, does it need to list them?08:20
bialix  (Vincent Ladeuil, #413514)08:20
lifelessluks: see above08:21
bialixlifeless: it's not related to your patch. may I  omit it while cherrypicking?08:21
lifelessluks: its merging a patch from someone, you'd expect the change08:21
lifelessbialix: sure08:21
* bialix start hates the fact that bialix adding bzr.ico to bzr.exe, grrr08:22
luksperhaps merge could output some diffstats08:22
luksso after each file you see how much of it is changed08:23
lifelessluks: that would be one way08:23
lifelessM foo.py +0 -008:23
luksyeah08:23
lifelessM bar.py +123 -45608:23
* fullermd goes back to dreaming of having that in log...08:24
lifelessfullermd: jfdi08:24
lifelessfullermd: log -p on 2a:)08:24
luksis it usably fast?08:24
lifelessfullermd: hey, how did you go with your src tree?08:24
lifelessluks: oh yeah08:24
fullermdOh, yeah, I'll just go design up a new repo format   :p08:24
lifelessfullermd: shouldn't need to now08:24
lifelessit should be able to keep ahead of you08:25
fullermdlifeless: Oh, I didn't.  It's on my "hey, this is probably good enough to try sometime" list.   Soon's work quiets down a little.  15, 20 years maybe.08:25
fullermdOh, I think it desperately calls for being cached at commit-time.  log bzr.dev takes 1.6 seconds (with a somewhat chilly cache).  log -p hasn't finished yet...08:26
lifelessfullermd: in 2a format!08:26
bialixjam: merge -cN --weave ROCKS!08:27
fullermdWell, I don't have a sizable 2a branch around.  But still.  I'll bet it's more than a factor or 2 or 3.08:27
bialixjam: you ** genius08:27
fullermd(which is my WAG for "pretty expensive")08:27
lifelessfullermd: so; I know someone that does log -v -p ~ in cron08:28
lifelessof a big arse branch08:28
lifelessdata >> WAGs08:28
fullermdSure, but who's got a hundred thousand revs of ~?08:28
lifelessfullermd: I mean approximately in cron08:28
lifelessnot homedir08:28
* lifeless watches screens of tests fail08:30
lifelesswell, I expected that08:30
fullermd"Crap, we just blew up half the moon."  "Well, I expected that."08:32
bialixlifeless: cherrypicked branch bzr+ssh://bazaar.launchpad.net/~bialix/bzr/bzr-1.18-oslocks-win32/08:33
bialixshould I send merge proposal to the list?08:33
lifelessbialix: please08:34
lifelessor just propose it in lp08:34
bialixpoolie knows beforehand that I will ask you him about 1.18.1 and ran away08:35
bialixghaa08:35
bialixpoolie: 1.18.1?08:35
bialixhi garyvdm08:49
vilapoolie: https://code.edge.launchpad.net/~vila/bzr/releasing-clarified/+merge/1079208:53
bialixlifeless: oops, I've just realized I had to use --author when commit your fixes. I can rework if you wish09:13
lifelessbialix: its fine09:13
lifelessbialix: NEWS has my name anyhow09:13
bialixsorry09:13
lifelessits fine09:13
vilalifeless: https://code.edge.launchpad.net/~lifeless/bzr/test-speed/+merge/10775 leads to 14 failures for a full test run (I was checking --parallel=fork was still working but it's unrelated)09:22
lifelessvila: thanks, could you attach --subunit output from that to the merge thread?09:28
vilalifeless: yup, trying to do that right now but the output looks far too verbose >-/09:28
lifelessvila: bzr selftest --subunit | subunit-filter --no-skip > failed.log09:29
vilasubunit-filter ?09:29
vilagot it09:30
vilahmm, known failures and unavailable features are back <SHUDDER>09:31
vilalifeless: re-running without fork, will be longer09:32
lifelessvila: thats because you only fixed it in bzr :)09:32
lifelessvila: ignore them09:32
lifelessvila: they'll get filtered out on my end09:33
vilalifeless: the file is still 2M want that by mail ?09:34
lifelessvila: sure09:34
lifelessbzip2 it09:34
vilalifeless: sent09:37
lifelessyou'll probably like my mail to the list ;)09:38
lifelessjml: are you on xmpp09:38
jmllifeless, I think so.09:39
lifelessas?09:40
lifelessjml: [gchat counts]09:41
jmllifeless, I like your recent bzr ML post09:45
lifelessthank you09:45
jmllifeless, have you seen http://dev.launchpad.net/FasterTests09:45
lifelessno09:46
lifelessor rather maybe09:46
CaMasonI want to fork one of my projects and develop elements of it seperate to the trunk. I will want to periodically merge changes from the trunk into this fork, but not overwite specific files. Is this possible?09:46
jmllifeless, it should probably cross-pollinate with your post09:46
lifelessjml: actually, yes I have09:46
lifelessI've even edited it.09:46
jmlpage history doesn't say so, but maybe that's pre-open sourcing09:47
jmlprobably was.09:47
lifelesslook for my name09:47
jmlahh yes09:48
jmlanyway, I'm off for the evening09:50
jmlg'night.09:50
lifelessciao09:50
lifelessLarstiQ: know where jelmer is?09:52
vilawow karmic booting in16s...10:14
=== timchen1` is now known as nasloc__
lifelessvila: have you tried --lsprof-tests?11:31
=== mrevell is now known as mrevell-lunch
lifelessvila: piiing12:02
vilalifeless: poong, was biking and lunching :)12:42
vilalifeless: not tried yet, but it's been a while since I wrote a test that took longer than 50ms :P12:43
CaMasonif I've got two branches, one of which is a fork, how do I merge specific changes into the fork?12:46
CaMasoni.e. I don't want to alter some of the files in the fork12:46
vilabzr merge trunk12:47
CaMasonwont that give me a load of conflict messages?12:47
CaMasonor even merge in some changes I don't want?12:48
vilaThe only way to know is to try12:48
vilaStart with a clean working tree (no changes nor pending merges), 'bzr st' should reports nothing12:48
lifelessvila: 50ms is pretty slow ;)12:49
vilalifeless: higher value, many are 0ms :D12:49
vilalifeless: highest value, many are 0ms :D12:49
lifelessvila: --parallel gives wrong per test timings, btw12:49
vilalifeless: I know, I don't rely on these, I use '-v' without --parallel when the set is small enough12:50
vilain fact I mostly use parallel for the whole suite12:50
vilamostly12:50
vilabut no always :)12:51
lifelessyea. its on my to-fix12:51
vilaI still use -s more than --parallel12:51
lifelesssubunit's parallel support has landed12:51
lifelesssorry, progress nesting12:51
lifelessanyhow, those tests all pass for me12:51
lifelesswith revno: 465912:52
lifelessmessage:12:52
lifeless  Detangle test listing: its more part of the ui layer not the execute-this-test-layer.12:52
vilaeerk, same revno here12:52
vilarevision-id: robertc@robertcollins.net-20090826233048-4yerdwqhvi2dqzi9 right ?12:53
lifelessyes12:53
vilawith extensions built ?12:54
lifelesshmm will check12:54
vilahmm, mine weren't up to date12:55
viladirstae_helpers12:55
vilatricked by reviewing repeatedly in the same branch may be...12:56
vilalol, unrelated but funny failure: test_crash.TestApportReporting.test_apport_report_contents AssertionError: pattern "(?m)^CommandLine:.*selftest" not found in CommandLine: ['./bzr', 'tss', '--no-plugins']13:01
vilatss is an alias for selftest --parallel=fork) :13:01
vilaso, sorry for the false alarm, the above is only failure I get once *I* recompiled my extensions13:02
lifelessso, +1 to land?13:06
vilare-reading now13:07
lifelessalso file a bug on that failure13:07
lifelesstag selftest please13:07
vilak13:07
vilaI don't follow the stopTestRun merged in python.. Is that in 2.7 ?13:08
lifelessyes13:08
=== mrevell-lunch is now known as mrevell
vilaI thought done() has been defined in subunit and testttools too... will they be updated too ?13:10
lifelesstesttools is13:13
lifelesssubunit I need to check but I think is already13:13
vilahmm, indeed testtools commit message explain it all13:14
vilais the stop_early/stop_on_failure transitional or really needed ?13:20
lifelesssubunit does need an update13:20
lifelessvila: it's contractually correct I think13:20
vilaso both are valid and need to be replicated ? stop_on_failure having higher priority ? Sounds a bit brittle13:21
lifelesserm13:21
lifelessI'll need to look closer, we may hav duplication13:22
lifelessI only shuffled around in this patch13:22
lifelessso whatever we had before, we have now13:22
lifelessI thought you meant stop_early -> self.stop()13:22
vilano :)13:23
vilaYou give us a far cleaner TestRunner,13:23
vilawhatever is left appears bigger now13:23
vilaso a loop on result_decorators to propagate a boolean looks highly suspicious13:24
vilathe injection of BZRTransformingResult is uglier too now :)13:24
vilaand I still don't like that self.verbosity handling here13:24
vilaI think that sum up my review, will put that online under approve cause it's better, how about fixing those later13:25
vilalifeless: reviewed13:29
lifelessthanks13:30
vilareagrding 2.7, xfail becomes standard but not UnavailableFeature ?13:30
lifelessyes13:31
vilak13:31
senderhello, can anyone confirm that it is not a problem to just "rm <dir> -Rf" in a shared repository (the shared concept makes me feel this is not the right way to go, maybe leaving information behind in the shared .bzr or, worse, removing data necessary for other repo's under shared), thanks.13:40
vilasender: if <dir> is a branch all that will be lost is the working tree, the changes not committed there and the easy reference to the branch tip13:42
vilasender: doing 'bzr log -l 1 --show-ids' before the rm should give you all the needed bits to recreate the branch if needed13:43
sendervila: thanks. that means that the removed branch will always remain in the shared repo?13:46
vilathe point of the shared repo is to share revisions between branches so the revisions will stay there yes13:47
senderok. so, I need the hash @ revision-id or parent to recreate it?13:47
senderis there a way to fetch those, even after hard-removing a branch under a shared repo?13:47
vilathe revid yes13:47
vilayes13:48
vilabzr branch -r revid:<revid> trunk resurrected13:48
senderhmm where does13:51
sendersorry, where does trunk points to?13:51
senderI've removed the branch, i've got the revid of that branch, now i am in the shared dir, and the command above gives me 'trunk' is no branch13:52
vilahmm, generally I use a shared repo for each project I'm working on, so I always have a branch which is the trunk13:52
senderand would trunk be the branch that was removed?13:52
vilaerr, you have a shared repo without any more branch ?13:53
sendercan I pick just any other branch?13:53
vilayes13:53
vilawe're cheating here anyway by using the branch as a entry point to the repo13:54
vilathe revid doesn't have to be part of the branch history13:54
senderah, nice it works :) yes exactly, it would make more sense to address the shared repo (as far as i can see)13:54
vilasender: well, we are searching in the garbage can here, so it's not often used either13:56
vilaanother way, a bit clearer may be is to do:13:56
vilabzr branch trunk resurrected ; cd resurrected ; bzr pull -r revid:<revid> --overwrite .13:57
vilammm, not sure it's clearer, I guess it depends who you ask :)13:57
senderinteresting. now this is fastforwarding/scrolling back the resurected branch to a certain rev, right?13:59
senderI guess this would never work with the regular revid, b/c those can diverge between branches, right?13:59
vilaYou mean the revno ?14:00
senderyes14:01
fullermdYou could just as well use "bzr init ; pull -rX .".14:02
lukswould that work?14:03
fullermdYah, I've done it a few times.14:03
vilafullermd: preceding that with mkdir resurrected of course :)14:03
vilaand a cd14:03
luksinteresting, as far as I know, the revno is calculated based on the last_revision_info file14:03
fullermdWell, init would make the 'resurrected' for you   :p14:03
fullermdWell, obviously specifying the revid rather than the revno.  But you'd have to do that anyway, unless you had a branch with that rev in its history.14:04
luksah14:04
luksI assumed that was a response to sender's last question14:04
vilaand in the future you may not have to specify 'revid:' once we get some DWIMery there :)14:04
fullermdSomeday, if somebody reviews the patch, it'll even DWIM the revid without having to qualify it   :p14:04
vilafullermd: hehe, I said it first :)14:04
lukswhat's blocking the patch, btw?14:05
fullermdThat means you volunteer to review it, right?   :p14:05
fullermdNothing.  People looking at it.14:05
vilawhen 2.0 is out, I already told you that there will be many discussions around and including that subject :-D14:05
vilafullermd: no sequitur, what is the latest bzr version available for FreeBSD and what is the command to install it ?14:09
senderok one more question: is there a way to find all the removed branches (head revids I guess) in a shared repo (or look into the garbage can - as vila said it)?14:10
fullermdOh, I got 1.18 in ports...   last week I guess.  I usually get 'em submitted within a day of the release, and it takes a few days to get committed.14:10
fullermdSo something like (cd /usr/ports/devel/bazaar-ng ; make all install clean)14:10
vilabzr heads --dead14:10
vilayou need the heads plugin14:10
vilaand to clean the garbage from the shared repo, you do 'bzr gc'14:11
fullermdIt's in bzrtools.14:11
vilaerr, wait, no, we still haven't that one :)14:11
vilaright bzrtools (what was I thinking... wasn't it in another plugin like, years ago ?)14:12
fullermdYah, it got Assimilated a while ago though.14:13
senderthat would be sth like git compact?14:13
* fullermd dreams of the day when he has a bzr command to check, reconcile, pack, and gc, all at once...14:14
vilafullermd: you shouldn't need them :)14:17
vilafullermd: seems to install 1.13 and failing because 1.13 is already installed14:19
fullermd1.13??14:20
senderi've got two branches in my shared, but bzr heads gives only 1 tip of branch, is this b/c these branches are not diverged?14:21
senderBoth HEADs have a different revno and different parent and revision-ids.14:21
fullermdOh, I guess you probably have an old ports tree from a release.14:21
vilafullermd: 7.214:21
vilasender: one is in the ancestry of the other ?14:21
fullermdsender: 'heads --dead' only lists heads that DON'T already have a branch pointing at them.14:22
fullermdvila: 'k, so you'll need an updated one.  Clean out the old (e.g., rm -rf /usr/ports/*), and pull down and extract a new with portsnap (e.g., portsnap fetch extract)14:22
fullermdThen you can get manually rid of the old bzr (pkg_delete /var/db/pkg/baz<TAB>) and install the new, or install portupgrade and use that to catch everything up.14:23
sendervila: true. thanks all!14:25
vilasender: so the most recent one is reported, the other one is not a HEAD because it has children14:25
sendermakes sense14:27
senderso, if i understood correctly, right now there is no way to clean dead heads out of a repo?14:27
vilafullermd: ok, so from there, it's basically 1) grumble, too old, 2) cd /usr/ports/<magical>/<package> 3) make all install clean (gee don't try -j there >-)14:27
vilasender: There is: create a new shared repo, copy all the branches you care about from one to the other14:27
vilanot pretty, but works14:28
=== cprov-afk is now known as cprov
fullermdvila: 1.5) pkg_delete to get rid of the old version, yah.14:28
senderit will do, thanks14:28
fullermdvila: And it -j's internally already where vetted   :]14:28
fullermdvila: If you install portupgrade, you can also (after the tree update) just fire a 'portupgrade -a' to upgrade everything with upgrades available.14:29
fullermdvila: That initial portsnap fetch/extract takes a while, but later updates (`portsnap fetch update`) are reasonably quick.14:29
* fullermd really should sit down and write up a portupgrade replacement someday :|14:30
vilafullermd: and what about upgrading the base system ? Not needed ? (*I* don't really care it's for the test farm so your advice will be god speak)14:30
vilaoh, Ill make that 2.5 instead 1.5 if you don't mind :)14:31
fullermdNot needed right now, I'd say.  7.2's only a few months old.  I don't usually upgrade base but a few times a year, at most (sometimes much less)14:31
vilak14:31
vilain rome...14:31
fullermdMy server's still running a pre-7.2 RELENG_7.14:31
fullermdFreeBSD 7.1-STABLE #0: Tue Feb 17 18:07:44 CST 200914:31
fullermdI guess it's probably due for an update...14:31
vilafullermd: so to compare to Ubuntu, ports ~= main - base system +  universe + multiverse ?14:33
fullermd(of course, _I_ update base as the first thing I do after installing a system, before I even install any ports.  But I'm nuts  :P)14:34
vilafullermd: I tried that and got a nice brick yesterday night :)14:34
vilavirtual brick mind you, but yet, that wasn't my day :D14:34
fullermdMmm.  I don't know Ubuntu well enough to judge comparisons like that.14:34
fullermdBase is...  well, base.  Kernel, basic complement of utils needed to run the system (newfs, mount, ps, ls, kill, sh, ssh, that sorta thing), and everything else is in ports.14:35
vilafullermd: sounds correct14:36
fullermdDon't need X or a web server or any of those weird fringe languages like python to get the system running, frex, so they're all off in ports   :)14:36
vilathat's not that clear when installing, but I think I get it14:37
fullermdI should try clarifying those bits in my rant from $YEARS ago.14:37
fullermd'course, that's just a subset of "I should update my whole dang website"   :|14:38
vilahaaa 1.18 ! Thanks fullermd that was well spent minutes14:39
fullermdvila: An offhand shortcut is "ls /{s,}bin /usr/{s,}bin"; that's base.14:39
fullermdAnything ports does (with occasional exceptions for special cases) is under /usr/local/14:39
vilaand /usr/local in ports !!! That\s why I couldn't make bash my shell !!!14:40
vilas/in/is/14:40
fullermdbash?  You mean there are people who don't use tcsh?   :p14:40
vilapff, I get out of these wars 15 years ago and settle to a shell I can use both interactive and batch :D14:41
fullermdemacs?   :p14:41
vilathen I learned perl, then python :D14:41
lifelessvila: an os isn't a shell14:41
vilafullermd, lifeless: the shell is where I see most of an os from emacs :-D14:42
lifelessvila: http://paste.ubuntu.com/260346/ - sneak preview for you of my next selftest patch.14:42
lifelessvila: this one causes some fallout, so its not finishd yet.14:43
vilaho ho, now you're getting angry :D14:43
vilais that permit dir and below ?14:43
lifelessyes14:43
lifelessI'll think about better names14:44
vilapermit tree14:44
lifelessdoesn't take a working tree14:44
lifelesspossibly just permit_in_url14:44
vilaok14:45
vilathat will not catch the access via os.xx calls, do you catch lots of pirates already ?14:46
lifelesswith this patch, the time for the selftest only modules drops from ~3.4 seconds to ~2.9 seconds14:46
lifelessos.xx is in the pipeline14:47
vilaoohh, you remove test_safety_net !! wow :D14:47
lifelessthis catches bzrdir.open(), as it documents14:47
lifelessyes, test_safety_net triggers this code :)14:47
lifelessso it fails at commit, not @ arbitrary later time14:48
lifeless[minor lie there; TestCaseWithMemoryTransport cd's to TEST_ROOT, and TEST_ROOT *is* permitted14:48
lifelessbut it was failing until I realised I needed to permit TEST_ROOT, at least transitionally14:48
lifelessTEST_ROOT is only permitted for TCWMT, not for TCWT14:49
vilaTEST_ROOT is above self.test_dir right ?14:49
lifelessyes14:50
lifelessin TCWT, TEST_ROOT/NAME is permitted14:50
lifelessin TCWMT,  TEST_ROOT is permitted14:50
lifelessthis isn't global, so its far safer14:50
vilahmm, that's bad, at least it shouldn't after setUp14:50
lifeless"at least transitionally"14:51
vilasorry, just reading the pastebin, I lack context14:51
lifelessI said that in IRC ;)14:51
lifeless23:48 < lifeless> but it was failing until I realised I needed to permit TEST_ROOT, at least transitionally14:51
vilabut transitionally is vague :)14:52
vilaI read it, just couldn't put a precise meaning on  it :)14:52
lifelessvila: the actual bug is that TCWMT makes any dir on disk at all14:52
lifelesshave to work towards it14:52
lifelesshave to fix bzrlib.config14:52
vilaI was about to mention config files14:53
lifelessyah14:53
lifelessso, my approach is14:53
vilaI never clearly understand why it needed that14:53
lifeless - add preventative jails replacing detection jails14:53
lifeless - ix fallout14:53
lifeless(wow?!)14:54
vilaix ?14:54
lifeless - fix f allout14:54
lifeless - remove detection jails, things go faster14:54
lifelessanyhow, midnight, and I bet I wake @ 5 again.14:55
lifelessso I'd better go sleep. gnight.14:55
vilahow about pushing down home/work dirs to TestCaseInTempDIr14:55
vilagood night :)14:55
lifelessyes, for sure14:55
vilaand the log file too :)14:55
vilaeven if some tests needs to inherit from TCITD now instead of TCWMT14:56
lifelessmeh, just delete that14:56
vilaoh, and puting the def setUp just after the def __init__ just at the beginning of the class will get you a beer or two next time we meet :-D14:57
lifelesshmm?14:58
vilaTCWMT.setUp is lost in the middle of the class14:59
lifelessoh14:59
lifelessis it in alpha order?14:59
vilaI bang my head against the desktop each time I missed __init__ or setUp because of that14:59
lifelesshere's a naughty tests14:59
lifeless======================================================================14:59
lifelessFAIL: bzrlib.tests.blackbox.test_split.TestSplit.test_split14:59
vilait was in alpha order, once, years ago ?14:59
lifeless----------------------------------------------------------------------14:59
lifelessRemoteException: RemoteException: Traceback (most recent call last):14:59
lifeless  File "/home/robertc/source/baz/pending/working/bzrlib/tests/blackbox/test_split.py", line 34, in test_split14:59
lifeless    self.run_bzr_error(('.* is not versioned',), 'split q')15:00
lifelessAssertionError: pattern ".* is not versioned" not found in15:00
lifeless"""\15:00
lifelessbzr: ERROR: Attempt to escape test isolation: 'file:///tmp/testbzr-byGS2o.tmp/' ['/tmp/testbzr-byGS2o.tmp/bzrlib.tests.blackbox.test_split.TestSplit.test_split/', 'file:///tmp/testbzr-byGS2o.tmp/bzrlib.tests.blackbox.test_split.TestSplit.test_split/']15:00
vilaexcellent, you need permit_url('q') ?15:01
lifelesssplit a/q15:01
lifelessits running the command outside of a tree altogether15:02
lifelesswhich is walking up15:02
vilahmm, isn't 'a/q' less clear ?15:02
lifelessworking_dir=a is the other way15:02
vilaon the error hand, that's clearly a test that could be done at a lower level right ?15:03
lifelessquite likely15:03
lifelessnot a broad focus yet15:03
vilacathc NotVersioned instead15:04
lifelessI'm focusing on environmental stuff at the moment15:04
lifelessmake TestCase* as cheap as possible.15:04
vilasure, not in your focu.. exactly15:04
lifelessthis is going to be a fun one:15:04
lifelessERROR: bzrlib.tests.blackbox.test_ancestry.TestAncestry.test_ancestry15:04
lifelessRemoteException: RemoteException: Traceback (most recent call last):15:04
lifeless  File "/home/robertc/source/baz/pending/working/bzrlib/tests/blackbox/test_ancestry.py", line 58, in test_ancestry15:04
lifeless    self._build_branches()15:04
lifeless  File "/home/robertc/source/baz/pending/working/bzrlib/bzrdir.py", line 1172, in sprout15:04
lifeless    force_new_repo, stacked_branch_url, require_stacking=stacked)15:04
lifelessBzrError: Attempt to escape test isolation: 'file:///tmp/testbzr-byGS2o.tmp/' ['/tmp/testbzr-byGS2o.tmp/bzrlib.tests.blackbox.test_ancestry.TestAncestry.test_ancestry/', 'file:///tmp/testbzr-byGS2o.tmp/bzrlib.tests.blackbox.test_ancestry.TestAncestry.test_ancestry/']15:04
lifeless  File "/home/robertc/source/baz/pending/working/bzrlib/bzrdir.py", line 474, in determine_repository_policy15:04
lifeless    policy = self._find_containing(repository_policy)15:05
lifeless  File "/home/robertc/source/baz/pending/working/bzrlib/bzrdir.py", line 659, in _find_containing15:05
lifelessits walking to the root to find a stacking policy15:05
vilarepo policy, I told you about these ones !15:05
lifelessyah15:05
vilatriggered with TMPDIR too15:05
vilascary a t first, but then I realized it wasn't harmful15:05
lifelessso my hook based net catches them15:05
lifeless:)15:05
vilaexcellent15:06
lifelessbut need a pervasive answer to stopping them15:06
vilamay be we need dedicated branches for that kind of patches so that the effort can be spread15:06
lifelesswe may need that15:06
lifelessOTOH it may be a simple annotation like '@needs_sprout15:06
lifelessand that would permit_dir TEST_ROOT for now15:07
lifelessI'd  like to land things that let you and others that are interested chime in in trunk15:07
lifelessrather than accrue large branches outside15:07
vilaso it may landed somewhere even or *because* it leads to too many failures to be addressed in a single sumbission, right annotations may work too15:07
lifelessI suspect determine_repo_policy is 90% of these15:08
lifelessor more15:08
lifelessif they were running against memory:/// it would be easy15:09
vilafrom memory, maybe, many pirates there, but some more, I should retry the experiment, I think I had a script for that15:09
lifeless3K15:09
lifelesshere, I'll send you the fail.log15:09
vila3000 failures ?15:10
lifeless:!subunit-filter --no-skip < failed.log | subunit-stats15:10
lifelessThere are 266 lines with trailing white space in 84 files.15:10
lifelessThere are 1961 lines longer than 79 characters in 313 files.15:10
lifelessTotal tests:    321215:10
lifelessPassed tests:      015:10
lifelessFailed tests:   321215:10
lifelessSkipped tests:     015:10
lifelessTags:15:10
vilaouch15:10
* lifeless waves gnight15:12
* vila waves back15:12
fullermdvila: BTW, I have this perl script I threw together to munch up the selftest failures and throw out the summary I posted to the list.15:12
fullermdYou may find use for it, considering how many of those BSD failures seem likely to be the same root issue, whatever it may be.15:13
vilafullermd: sure15:15
vilathe mail you're refferring to is around june 22 ?15:15
fullermdSounds right.15:15
vilaactually you may be interested in subunit which provided the same kind of filtering...15:16
fullermdOn its way.15:18
vilacool15:19
fullermdNo cracks about code quality.  I was young, and I needed the money  :p15:22
vilafullermd: oh, come on, that's perl, write once, use intensively, forget about it :D15:25
vilapython is far harder to use for throw-away code :-D15:26
vilahu ho did I see gmake installed, wow, of course the famous BSD make ! :D15:27
fullermdWell, everybody's gotta use their own local variant.  'till I get my replacement finished, of course, and everybody flocks to it.15:36
vilawell, I'm convinced that the problem is not with tools themselves but rather the way the are use (and in make's case abuse), I was.... impressed by the way autconf render the compile commands totally unreadable a couple of years ago...15:38
fullermdWell, it had to do that.  Otherwise it couldn't add 3 minutes up front to the build time.15:39
samkramergood morning15:46
vilamornibg samkramer, jam15:47
jammorning vila15:48
vilafullermd: wow, pertty clean perl, at least relative to my standard, use strict, even use warnings ! use Data::Dumper , the perl  army knive :) Apart from the 8wide tabs, nice code :-D15:59
fullermdHey, just 'cuz a language CAN be dirty doesn't mean _you_ have to be in it.16:00
fullermd(your tab size is your own problem though; mine at 4-char  :p)16:00
vilafullermd: oooh, literal tabs ! yeah mine is 4-char too but translated in spaces, always16:01
fullermdYeah, I believe in tabs for indentation.16:02
igcnight16:02
fullermd(not, of course, for alignment.  I shouldn't _have_ to say that, but it seems like so many people get caught up in conflating the two...)16:02
* fullermd waves to igc.16:03
vilanight Ian16:03
vilafullermd: I believe in editors handling the indentation for me and providing a way to share settings inside projects16:04
LarstiQlifeless: he left HAR early, the 14th or 15th, to go on a biking trip16:18
LarstiQlifeless: from Utrecht to Switzerland iirc16:18
vilaLarstiQ: pff, and no wifi on his bike ?16:53
=== deryck is now known as deryck[lunch]
samkramerany good links on configuring bazaar from scratch as a centralized server?16:57
=== beuno is now known as beuno-lunch
=== deryck[lunch] is now known as deryck
montywianyone that could help to get bzr 1.17 to work with gtk 0.97.0-final ?18:03
montywiAfter upgrading from bzr-1.13, 'bzr pull' dies with:18:03
montywi  File "/usr/local/lib64/python2.5/site-packages/bzrlib/config.py", line 1246, in get_credential_store18:04
montywi    cs = cs()18:04
montywi  File "/usr/local/lib64/python2.5/site-packages/bzrlib/plugins/gtk/keyring.py", line 39, in __init__18:04
montywi    gobject.set_application_name("bzr")18:04
montywiAttributeError: 'module' object has no attribute 'set_application_name'18:04
* mneptok pokes poolie, igc, lifeless etc18:23
mneptokour work is stalled until Monty can commit. :)18:24
montywimneptok: I got things going by downgrading to bzr-gtk 0.96.2, but it would be nice if bzr-gtk-0.97 would work too18:28
fullermdWell, vila cut the release, so it's obviously his fault  :]18:41
bialixlifeless: I should admit that I don't test your fixes yet18:55
bialixlifeless: only cherrypicked them and planning to test them soon18:56
vilamontywi: that's bug #403430 and/or #397967, gobject API is a mess regarding the two functions needed there, worse than I thought it seems, please file a bug reporting what os/version you are using18:59
ubottuLaunchpad bug 403430 in bzr-gtk "bzr diff fails with traceback and error 'module' object has no attribute 'get_application_name'" [High,Fix released] https://launchpad.net/bugs/40343018:59
=== beuno-lunch is now known as beuno
bialixlifeless: quick test reveals that push now works ok, and shelve at least run without errors19:00
vilamontywi: in the mean time, just commenting out the offending line should do as the surrounding comments will tell you19:00
bialixvila: bonsoir19:00
vilabialix: hi !19:01
* vila fades 19:01
=== vila is now known as vila-dinner
bialixtoday everybody ran away when I want to ask someting19:01
* bialix watches for vila!19:02
vila-dinnerbialix: quick then, gf waiting :)19:02
bialixgf?19:02
fullermdI'll run away BEFORE you ask me anything, if that'll help break your streak.19:02
bialixbon appetite, I'll wait19:02
vila-dinnergirl-friend19:03
Takhell - what was the solution to "ERROR: Operation denied because it would change the mainline history" with bzr-svn?19:03
Takgirl...friend?19:03
bialixso, I'm just curious, today is karmic featurefreeze, and what version of bzr was included there? I think it should be 1.18, but core devs insist on 2.0. who won in the end?19:16
maxb!info bzr karmic19:19
ubottubzr (source: bzr): easy to use distributed version control system. In component main, is optional. Version 1.18-0ubuntu1 (karmic), package size 6593 kB, installed size 21304 kB19:19
bialixcool, you know black magic19:21
bialixthen I'll continue backport fixes for some annoying bugs19:22
bialixlike https://bugs.launchpad.net/bzr/+bug/41893119:22
ubottuLaunchpad bug 418931 in bzr "AssertionError: _remember_remote_is_before((1, 18)) called, but _remember_remote_is_before((1, 15)) was called previously" [High,Fix committed]19:22
* Tak `bzr push --overwrite --no-strict`, wait for horrible things to happen19:24
bialix!info qbzr karmic19:36
ubottuqbzr (source: qbzr): Qt frontend for Bazaar commands. In component universe, is optional. Version 0.14-0ubuntu1 (karmic), package size 281 kB, installed size 1612 kB19:36
bialix!info bzr-explorer karmic19:37
ubottuPackage bzr-explorer does not exist in karmic19:37
bialix!info bzr-svn karmic19:37
ubottubzr-svn (source: bzr-svn): Bazaar plugin providing Subversion integration. In component universe, is optional. Version 0.6.4-1ubuntu1 (karmic), package size 234 kB, installed size 1540 kB19:37
jwhitleyJust popping in here to tell everyone that it *rocks* that "bzr upgrade bzr+ssh://..." works.  Just finished a repo conversion on a remote slice with too little memory to do the conversion locally.20:26
jwhitley(That with bzr 1.18rc1)20:26
bialixconversion to 2a?20:27
=== EdwinGrubbs is now known as Edwin-lunch
lifelessmoin20:58
lifelessbialix: cool21:00
bialixlifeless: hi?21:00
lifelessbialix: I'm glad the fixes worked for you21:00
bialixah yes21:00
bialixthough new shelve lacks some nice features from old one21:01
jwhitley@bialix: yes, conversion to 2a.  worked like a charm.21:12
bialixthat's good21:12
lifelessjwhitley: thats a really good data point. We've got a desire to make upgrade always run remotely; we should document how to force it to do the work locally.21:15
lifelessigc: ^ what do you think, just a line in the upgrade guide?21:15
jwhitleylifeless: For some reason, I'd assumed for ages that upgrade was an inherently local operation.  Running out of memory on a remote host forced new awareness.  ;-)21:16
lifelessjwhitley: oh, did I misunderstand... it ran remotely and that was good?21:20
lifelessjwhitley: I thought you were saying, you had a remote branch ona server that doesn't have enough memory to do upgrades on  the server?21:20
jwhitleylifeless: you're correct, sorry to randomize matters.  For no particular reason, I'd assumed that 'bzr upgrade' had to be done on the same host as the branch/repo's filesystem.  I'm terribly glad that isn't the case, since the server does not have enough memory for the ugprade.21:28
lifelessok21:28
lifelessyou may find you need more memory for regular push-pull operations though21:29
lifelessas we're moving more and more code into the server to eliminate round trips21:30
lifelessif you find this happens you can use 'nosmart+' at the front of a url to disable server side processing for that operation21:30
=== Edwin-lunch is now known as EdwinGrubbs
lifelessvila-dinner: gimme a shout when you return21:56
* bialix wonder when vila sleep22:00
fullermdSleep is for wimps.22:02
fullermdHappy, healthy, well-rested wimps.  But wimps nonetheless.22:02
bialixI'm sure vila is not22:03
fullermdWell-rested?  Yeah, probably not   :)22:04
bialix:-#22:04
CaMasonhow can I copy specific changes from one branch to another?22:06
CaMasonI have two branches, but they need to be kept slightly different22:06
bialixit depends22:07
bialixyou can use merge -rN..M22:07
bialixor rebase -r N..M22:07
bialixerr22:07
CaMasonwhat's the difference?22:07
bialixreplay -r N..M22:07
bialixmerge just merge22:08
bialixreplay will replay commit with all commit metadata22:08
bialixwith merge you can cherrypick changes to specific files22:08
bialixnot entire changeset22:08
CaMasonok. Don't I lose the 'link' with a cherrypick merge?22:09
bialixbut merge will not carry commit metadata22:09
bialixyes, you do22:09
CaMasonis my workflow wrong?22:09
bialixthat's why Daggy Fixes is better22:09
bialixnot necessary, it depends22:10
bialixbut at least I'd recommend to read about Daggy Fixes22:12
bialixto understand another way22:12
=== cprov is now known as cprov-afk
bialixCaMason: if you're OK to share history without sharing some changes, you can merge only istory to preserve "link", but revert wrong changes22:17
CaMasonso merge, commit, revert, commit?22:18
bialixno22:19
bialixmerge, st/diff22:19
bialixrevert $FILES22:19
bialixor shelve changes away22:19
CaMasonah now that's an idea22:19
bialixcommit22:19
CaMasonshelve changes away?22:19
bialixbzr shelve -h22:20
bialixshelve aka selective revert22:20
CaMasonI see22:21
CaMasonwhat is the purpose of shelving them over reverting?22:24
beunoCaMason, to bring back the change later on22:24
beunoso maybe you shelve, commit, un-shelve, continue working on that change22:24
bialixyou may want to revert some changes in the one file and retain other changes in the same one file22:25
bialixwell, you can deleted shelved changes22:25
bialixwithout bring them back22:25
bialixunshelve --delete does this22:25
CaMasonI see, great :)22:25
bialixbecause bzr require explicit commit after merge you can change everything you want before you commit your merge22:26
CaMasonI didn't realise that, thanks22:27
bialixso you can do merge only for history without merging changes22:27
bialixmerge; revert . ; commit22:27
CaMasonthat's precisely what I wanted22:27
bialixrevert . (dot is intended)22:28
CaMasonhow about if I had a lot of files, and only wanted to keep 2 of them? Could I shelve them, then revert, then unshelve, then commit?22:28
bialixok, now you're armed22:28
CaMasonarmed D:=22:28
bialixyes you can22:29
CaMason:)22:29
bialixwhat a misterious smilie?22:30
CaMasonD:== that one?22:30
bialixyeah. it's a cat?22:30
CaMasonlol no. My friend keeps using it22:31
CaMasonthing of that long faced guy from muppets22:31
CaMasonand it's reversed. a 'shocked' face22:31
CaMasonD:22:31
* bialix start s fear22:31
bialixabout armed: there is such russian proverbs: if you informed then you're armed22:32
CaMasonI figured. I wanted to double check22:33
CaMasonhttp://www.toymania.com/columns/spotlight/megabeaker.shtml22:33
CaMasonthat's what the smiley is based on :)22:33
bialixmaybe I've translated it bad22:33
CaMasonD:=====22:33
CaMasonmaybe it's just be being an idiot ;)22:34
bialixfunny muppets22:35
bialixI saw this show22:35
igcmorning22:44
bialixigc: hi22:45
bialixyou're early today22:46
lifelesshi igc22:46
bialixlifeless: btw about test. is there a way to test NotBranchError?22:47
lifelessbialix: in what regard?22:47
bialixI have some code that trying to open branch and show error dialog if there is none22:48
bialixI've mocked gui22:48
lifelessraise errors.NotBranchError("woo hoo")22:48
lifeless?22:48
bialixinside function?22:49
bialixhow?22:49
lifelesswell just like that22:50
lifelessraise the exception wherever you want it raised22:50
bialixlifeless: http://pastebin.com/mf2f751922:52
bialixI want to test the code path for except NotBranchError22:52
Sjorshey :)22:53
Sjorsjust reported this bug: https://bugs.launchpad.net/bzr/+bug/42018822:53
ubottuLaunchpad bug 420188 in bzr ""Svndiff contains a too-large window" error" [Undecided,New]22:53
bialixI can't create empty dir in test sandbox22:53
bialixlifeless: err, I can, but empty dir won;t raise NotBranchError because root of sandboxed temp dir for tests has guard tree22:54
lifelessbialix: try this22:54
lifelessmake a dir22:54
lifelessbzrdir.BzrDir.create('dir')22:54
lifelessthis will make an empty dir22:54
lifelessshould stop the search and raise NotBranchError22:54
lifelessalternatively you could hook into BzrDir.pre_open22:55
lifelessand raise NotBranchError from there yourself, for the path that you'll be opening22:55
lifelessthat would be a little harder to write the code for; but do less IO/be faster22:56
bialixI'll try to play with it22:56
lifelessits basically22:59
lifelessbzrdir.BzrDir.hooks.install_named_hook("pre_open", my_deny_function, "deny a directory")22:59
lifelesswith22:59
lifelessdef my_deny_function(transport):23:00
lifeless    raise errors.NotBranchError(transport)23:00
lifelessthis is a global hook, gets reset at the end of the TestCase automatically.23:00
bialixpre_open is new thing?23:00
lifeless1.1423:00
bialixcool23:00
bialixthat's nice23:00
bialixthanks23:00
bialixSjors: looks like the bug actually in bzr-svn plugin23:03
Sjorsbialix: yes, indeed23:03
bialixretargeted23:04
lifelessvila-dinner: ping23:04
=== bialix is now known as bialix-wimp
pooliehello bialix, lifeless, vila, igc,23:36
lifelesshi poolie23:36
lifelesspoolie: oh, btw - http://twitter.com/search/users?q=martin+pool&category=people&source=users23:37
pooliehuh23:38
poolieupdated, it may take a while23:38
lifelesshi jcastro23:47
lifelessfigured you should be around here as you deal with upstreams <-> launchpad, and we're a big part of that23:48
=== vila-dinner is now known as vila
lifelesspoolie: remind/confirm 12 for lunch23:56
vilaback from movie, lifeless ? I won't stay long though23:56
pooliestill fine23:56
pooliehello vila23:56
lifelessvila: hi23:56
vilahey :-D23:56
lifelessvila: what movie?23:56
vilapff, a story about the end of world, french actors, quite good, very strange story. a bit a la Bunuel :)23:57
vilas/the/an end/23:57
lifelessvila: I highly recommend district 923:57
vilak, noted, will check availability :)23:58
lifelessvila: I'm thinking of recording somewhere the test themes we have to work through to solve this problem23:58
lifelessa bit of a todo23:58
lifelesswe could use bugs23:58
lifelessor a wiki page23:58
lifelessor something23:58
vilaamazing, I was thinking about using a wiki page to handle bug lists as an alternative to assigning bugs to myself without working on them... But you're talking more about something like the url jml gave for launchpad I presume ?23:59

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!