[00:06] I see Patrick is still online - is that correct? I am still in the bzr chat room. [02:33] Am I doing something dumb, or is BzrDir.create_branch_and_repo('foo') supposed to raise a KeyError? [02:33] (Seems the format argument is less than optional.) [02:33] Peng: I imagine you're doing something unusual [02:33] or its broken [02:36] D'oh. Of course it works now when I try to reproduce it. :| [02:47] * Peng watches get_parent_map try to figure out that the repo is empty. === ivan`` is now known as ivan === ivan is now known as Guest60494 [02:59] I can't reproduce it now. :\ [02:59] The target *may* have already had a .bzr directory in an odd state. [03:00] No, wait, that raises a different error. [03:01] Here's the traceback, FWIW: http://paste.pocoo.org/show/152176/ === Guest60494 is now known as ivan [03:04] untested code path I imagine [03:05] file a bug please [03:16] lifeless: filed bug #486535, fwiw [03:16] Launchpad bug 486535 in bzr "BzrDir._find_or_create_repository KeyError (bzr+ssh + stacking policy?)" [Undecided,New] https://launchpad.net/bugs/486535 [05:29] Anyone there? [05:35] pidgin tells me there are 116 people in room, but there does not seem to be any activity, viz people coming or going, nor do I see my nick on the list... [05:39] it's usually pretty quiet on the weekend [05:40] OldAl: it doesn't matter who is here or not [05:41] OldAl: everyone might be here, and noone able to answer your question. Just Ask. [05:41] ah, geat! [05:42] I only want to ascertain that I have connected to #bzr chat room. [05:42] pidgin is correct, you are connected ;) [05:43] where are you from, lifeless? [05:43] I am from Canberra [05:44] May I put you down as a buddy? [05:46] IRC isn't like IM networks [05:46] there isn't the same social-network / buddies aspect [05:46] Here is a question - when I connect, I get a message: (15:51:49) NickServ: (notice) OldAl is not a registered nickname. So, should I a) ignore it b) register and if so where? [05:47] http://freenode.net/ [05:47] But I have registered on freenode.net - else I would not be here, would I? [05:48] what do you mean by upgrade :) [05:48] sorry [05:48] by registered [05:49] Well, I did fill out some details and was able to connect under an OldAl nick, but it showed my real given name... [05:50] so I suspect that it was not complete or something. [05:50] you register your nick by talking to nickserv using your irc client [05:51] pidgin is not a very useful irc client [05:51] OldAl: so each time you connect you need to login [05:51] and that message says you haven't actually registered a usercode [05:52] OK, so what is the button to press to start registration? [05:53] http://freenode.net/faq.shtml#nicksetup [05:53] and, lifeless, did you have to type in "OldAl:" in your previous message? [05:53] no [05:54] so what is the smart way of getting that? [05:54] getting what ? [05:54] irc clients tab complete nicks, so you type o [05:55] lifeless: OK : [05:55] lifeless: That's a great help. [05:55] lifeless: Many thanks for that. [05:56] lifeless: I do appreciate your help. Will read up the FAQ on registration of a user name. [05:56] so actually, I was answering 'did I need to have OldAl present at the start of the message' with 'no'. [05:57] not how I went about putting it there :)\ [05:58] lifeless: You wrote parts of FAQ for pidgin? [05:58] nope [05:59] lifeless: Oh, I was at cross purposes there - I am now using the tab - great "invention". [06:00] lifeless: Since my nick is apparently not registered. how can I talk to you? Is it a temporary "registration"? [06:01] this isn't IM [06:01] registration just means that a) you're allowed in some channels that forbid unregistered users and b) you can kick off other people who use your nick [06:02] lifeless: Well, I don't know IM. [06:02] for a) bob2 meant 'not allowed in'... [06:02] OldAl: imagine that this is a table in a restraunt [06:02] when you are connected, you are at the table. And everyone here can hear everyone else. [06:02] lifeless: yap [06:02] fine [06:03] *if* you get an 'official' name badge, then you can grab someone and talk to them privately, and sit at some tables that require people to have official names [06:05] lifeless: But I have chatted with people already, in separate "table" (or room), as it were. [06:51] lifeless: missed you at UDS :( [06:52] mneptok: Swings and roundabouts [06:53] had stuff going on I would have had to shove around, and no specific things I needed to be at UDS for this time [06:53] and i thought we had something *special* [06:54] *sniff* [06:55] we do [06:55] I hear you made an unrepeatable quote [06:56] so bad only its existence hit the Quotes page [06:56] it wasn't that bad. [06:57] Keybuk likes to paint me as a greater rogue than i am. i think it's part of his rich fantasy life. [06:57] it wasn't Keybuk :) [06:58] hrmf. he said he had added it. [06:58] trying to make himself more attractive to me, i s'pose. [06:58] I think I'd need to go check I guess. [06:59] not that his rugged good looks and commit rights don't make him attractive enough. [07:02] mneptok: i even clicked a link in your blog and didn't regret it too much! [07:02] mwhudson: oh nice! you got that Thorazine scrip and lobotomy you were asking for! [07:03] (or you're just insanely drunk) [07:03] c'mon, the police aren't that bad [07:03] (i am disturbingly hungover for 20:00 though :/) [07:04] mwhudson: hungover or drunk ? [07:04] lifeless: hungover [07:04] wow [07:04] i didn't even drink _that_ much last night :/ [07:04] getting old, it's terrible [07:05] sneaks up on ya [07:05] i have t-shirts older than both of you. [07:12] hi. i was wondering whether it would be possible / feasible / desirable to integrate trac-bzr-loggerhead together. I guess a post commit hook that posts to trac and an apache rewirte rule so that the commit linking to the bzr source viewer ties into loggerhead instead? [07:13] chx_: I don't know what that means [07:13] nor have any idea about whether it would be good or bad to do it. [07:14] well, currently we use svn and commit messages (re #1234) post followups to tickets [07:14] that's awesome [07:14] The other part of the svn integration in trac is the source code browser [07:14] that's what i was considering replacing with loggerhead. [07:15] so loggerhead showing svn? [07:15] no, we are migrating to bzr [07:15] oh now we are talking... anyone from Canonical here ? [07:17] yes [07:17] loggerhead can show svn btw [07:17] wonders of bzr-svn :) [07:27] lifeless: pm? [07:28] I guess [07:30] oh no! the *bad* Monty! [07:30] * mneptok hides [07:32] mneptok: does that make you the Full one? [07:33] lifeless: not if you can't be bothered to visit UDS. [07:37] mneptok: Sif a matter of botherment. [07:43] stupid python, stop making bzr complain about my locale :( [07:45] kgoetz: install the locale data for it :) [07:47] lifeless: I'm not aware of any en_AU.UTF-8 python packages ;) [07:51] kgoetz: its generally just missing langpacks, if you're on Ubuntu [07:51] or it can be a missing langpack on a server. [07:56] nothing leaps out at me from the debian stable repos, so I might just put up with it [07:56] kgoetz: a ew questions [07:56] is it local or when pushing/pulling? [07:58] its local. and the install is pretty minimal - its netinstall of debian+a handful of packages [07:59] check your local is being created [08:00] kgoetz: also please file a bug if this is happening with bzr installed from a package; we shouldn't behave wrong from a package [08:01] lifeless: I don't follow - 'your local is being created'? [08:01] oh, *smacks self* [08:03] kgoetz: I'll be afk for a bit, but it sounds like you've found a cluebat? [08:04] lifeless: checking local setup atm. === BasicPRO is now known as BasicOSX [08:08] dpkg-reconfigure locales to the rescue. [08:11] Why is it so quiet here? [08:12] nobody wants to chat about bzr? [08:13] OldAl: seems no one needs support ;) it is sunday... [08:13] kgoetz: No, it is not support - just signs of life! [08:14] kgoetz: and just tell me, is this the bzr chatroom? (I expect so, but just to be sure. yes/no will do :) ) [08:15] OldAl: I'm looking at my screen+scrollback, and i see lots of conversation [08:15] Well, it must be #bzr, I hope. [08:16] kgoetz: Sunday or not, I've just completed registering my nick - or so I hope. [08:16] it would seem your identified, so yes, your registered [08:17] kgoetz: Thanks for that. [08:18] *you're [08:18] np. /whois OldAl to see [08:18] kgoetz: I will log off and leave you in peace! Thanks for helping me to check what is going on [08:18] OldAl: catch you nex time :) === BasicPRO is now known as BasicOSX [13:08] me_too: Like your nick... [13:09] me_too: Are you watching the screen of bzr? [13:09] menesis: Hi there! [13:10] Actually it is a bit late for me - down under Canberra, the time is past midnight [13:19] Noldorin: Alex, where are you? (I am in Australia, Canberra) [13:20] Noldorin: irc informs me that you are on bzr, so ... === Peng___ is now known as Peng_ [14:42] Hi, where is a downloadable version of the User Reference? [14:43] It can be viewed as a HTML, but there is no PDF version... [14:45] matt_p: http://doc.bazaar-vcs.org/bzr.2.0/downloads/ [14:45] or http://doc.bazaar-vcs.org/bzr.dev/downloads/ [14:45] GaryvdM: http://doc.bazaar-vcs.org/bzr.2.0/downloads/pdf-en/ [14:46] GaryvdM: There is only the Quick Ref. [14:46] I need this: http://doc.bazaar-vcs.org/bzr.2.0/en/user-reference/index.html [14:47] http://doc.bazaar-vcs.org/bzr.2.0/downloads/pdf-en/bzr-en-user-guide.pdf ? [14:48] GaryvdM: It is User Guide, not User Reference [14:48] matt_p: It's the same thing as far as I can tell. [14:49] GaryvdM: So you mean that this http://doc.bazaar-vcs.org/bzr.2.0/en/user-guide/index.html and this http://doc.bazaar-vcs.org/bzr.2.0/en/user-reference/index.html are the same documents? :-) [14:50] matt_p: Sorry - I'm wrong. It's not the same. [14:51] matt_p: Sorry - I don't know then. [14:51] matt_p: igc will know. But he will probably be asleep now. He is in Austraila [14:52] GaryvdM: OK, thanks. [14:57] Oh, I have a second question... I use Bzr and I understand its concepts as a user. But what is the right relation between branches and repositories? (Shared) repo contains the revisons in their physical forms and the branches only point to the relevant revisions? [14:58] matt_p: Yes. A branch has just a pointer to the latest revision, tags, and config [14:58] matt_p: yes, repositories holds all the information: file contents, deltas, other common info [14:58] matt_p: A stand alone branch has a repository [14:59] *has it's own repository [14:59] matt_p: http://bazaar-vcs.org/CategoryTerm [14:59] gioele: Nice collaboration :-) [14:59] GaryvdM: :) [15:00] GaryvdM, gioele: great, thank you. [15:00] matt_p: in particular the first sentence of http://bazaar-vcs.org/Branch [15:09] hi, i'm trying to host a bazaar server [15:09] so i start the server with bzr-2.6 server --directory=/mnt/md1/public/Glen/Repos [15:09] and i can branch from it fine [15:09] but when i try to push it tells me operation not possible: readonly transport [15:10] pushing to bzr://ip/myrepo [15:10] i'm obviously missing a step.. [15:10] pitseleh: you need bzr server --allow-writes [15:10] GaryvdM: ah, thanks for that :) [15:11] works perfectly :D, wonderful === pitseleh_ is now known as pitseleh [17:14] Hi jelmer :-) [17:15] hey Lo-lan-do [17:15] For some reason my changes don't work anymore. [17:15] (Even given the stretched value of "work" I used) [17:18] Lo-lan-do: it might be nice to have some tests for this code first and figure out the right approach [17:20] Probably, yes. Is there a tutorial on that, or should I just read the existing tests? [17:21] Also, it might be a bit complex, since I'm trying to do git+ssh:// with a git client… [17:33] Lo-lan-do: There isn't anything like that at the moment, in fact I don't think we have any tests for the server side [17:34] you should be able to test with the git:// server though, since the protocol should be the same [17:35] testing with ssh will be trickier [17:36] I think maybe the "git proxy" configuration could help [17:49] Actually, $GIT_SSH pointing at a custom script can do the trick :-) [17:50] Lo-lan-do: isn't that kinda hard to set up in the tests though? [17:58] I don't think so. It's just an environment variable. [17:59] And a 30-line shell script. [18:03] http://pastebin.com/f1db8ac6c [18:03] Stick that in a script, point $GIT_SSH at it, and there you go, git+ssh:// without ssh :-) [18:04] I'm off to dinner, but I'll probably start doing unit tests afterwards. [18:48] hi, when you do 'bzr init' and then 'bzr add' it should add all subdirs. correct? [18:49] the part is that when i do 'bzr status' i get a good chunck of unknown objects [18:50] Objects that were created later on? [18:50] even doing 'bzr add foldername' and then do 'bzr status' the foldername will still show up under unknown [18:50] nope, the objects are already there [18:51] it only adds two directories out of about 10 [18:52] that is not how bzr normally acts. right? [18:52] eli_, that is not expected behavior, no [18:53] it should either add or ignore, no unknowns [18:53] hmm [18:53] it even treats .bzrignore as an unknown [18:53] that can happen [18:53] i'm using version 2.0.1 [18:54] if you manually create the file [18:54] without using `bzr ignore` [18:54] eli_: it's not perchance already versioned? (Different vcs maybe) [18:55] you mean having .svn in the subdirs? [18:56] eli_: oh, that might explain it [18:56] eli_: if you also have bzr-svn installed that is [18:57] LarstiQ, i think your suggestion may have hit it. [18:57] It was not .svn, there were .bzr in each subdir [18:58] yeah [18:58] bingo, adding everything now :) [18:58] Thanks! [18:58] that is indeed a case I forgot that will result in unknown dirs :) [18:58] eli_: cheers! [19:00] is there a way to list just the unkowns in bzr? [19:01] `bzr ls --unknowns` [19:01] optionally with -R [19:02] excellent. bzr is already proving much more handy than hg and svn :) [19:03] I'm glad to hear that :) [19:03] though a bit surprised perhaps, how is hg lacking? [19:12] hg is creates larger histories than bzr. at least in my case [19:19] So yeah i was trying to research hg vs bzr vs git (not in speed) for recetn versions and ... hmm... [19:21] chx_: and ? [19:21] empty/ [19:21] i cant find anything [19:22] chx_: http://bazaar-vcs.org/Benchmarks ? [19:22] as said, not in speed [19:22] we can simply take it granted they are all fast. [19:22] chx_: no, this is disk space [19:22] chx_: http://doc.bazaar-vcs.org/migration/en/why-switch-to-bazaar.html [19:24] chx_: Both my url, and LarstiQ's are up to date documents. [19:26] oh [19:26] the latter is very nice [19:26] how come i couldnt find it this with google [19:27] chx_: what did you google on? [19:27] I found it with http://www.google.com/search?q=why+bzr , But I knew the name of the doc. [19:27] bzr git comparison [19:27] bzr vs git [19:28] I am reading this [19:28] and though i myself am using bzr since long, i find interesting pieces [19:28] how do i create a bound branch? [19:28] chx_ bzr checkout [19:29] oh that [19:29] i see. [19:29] chx_: or if you allready have a branch, bzr bind [19:29] nice [19:30] another question , hardlinks [19:30] --hardlink Hard-link working tree files where possible. [19:30] i see that [19:30] this hardlinks files inside .bzr i presume? [19:31] I'm not sure [19:31] It explicitly says "working tree files", so probably not. [19:31] 'working tree' suggest otherwise, but I don't know what it does [19:31] hey Peng ;) [19:31] huh that would be nice but challenging what happens if someone edits one such file...? [19:32] chx_: Yeah. If your text editor doesn't break links, it won't work for you. That's why it's not the default. [19:33] Peng: and which editors breaks...? [19:33] chx_: No idea! [19:33] vim breaks hardlinks, so that should work [19:33] I'm sure emacs can too [19:34] it would need to save the file under a diff name , rm the oriiginal and mv back [19:34] i presume that's not hard. [19:35] It's a nice way of making the write a bit safer, too. [19:39] is there a way to have bzr to not create histories of binary files? [19:39] rather than an outright ignore? [19:41] for example, i have several subdirs with binaries and scripts [19:41] i would like to version control the scripts [19:41] but not the binaries. [19:42] but, if i do push or pull with bzr i'd like to have the binaries move between systems [19:42] probably sounds like a dream, but that would well for me [19:42] nope, something is either versioned or not. [19:42] (likewise, there is no 'precious' class of unknowns) [19:43] eli_: you fear the binaries would take up a lot of space? [19:43] yep [19:43] tons [19:43] but i do need them synced between two computers [19:44] and keep version control on the scripts [19:44] can the binaries be generated from something else? [19:45] nope, the binaries are data files. they are not dynamically generated [19:45] ok [19:45] could you maybe version them in a separate branch? [19:45] and have symlinks in the original? [19:46] that would seperate out the binaries from the scripts [19:47] generating the symlinks would be the next complication. we're talking +10,000 binaries [19:48] woha [19:48] yeah, lots [19:48] not your typical usecase :) [19:48] nope ;) [19:48] would it be possible to do a layer of rsync with bzr? [19:49] rsync for the binaries [19:49] you could maybe hack up a plugin (based push-and-update?) to do something like that [19:49] and bzr for version control [19:49] eli_: yeah [19:49] even with the binaries in the same dirs as the scripts? [19:49] yes [19:49] hmmm [19:50] that would work then [19:50] it would be a plugin tailored to your workflow [19:50] (or more general if you wish) [19:50] i know quite a few people in my field who would use such a workflow [19:50] ill write it up then [19:51] thanks for your help in brainstorming on this :) [19:51] adios! [19:51] eli_: http://doc.bazaar-vcs.org/plugins/en/push_and_update-plugin.html is what I'd start with [19:52] Thanks [19:52] will looks into [20:33] I am not familiar with the bzr test system. Can anyone suggest what I should look into when a test 'fails' because it starts thinking normal progress messages are errors? [20:34] errors: [20:34] 'All changes applied successfully.\nCommitting to: /tmp/testbzr-LDqLMO.tmp/bzrlib.plugins.rebase.test_blackbox.TestRebaseSimple.test_rebase_merge/work/main/feature/\nCommitted revision 3.\nAll changes applied successfully.\nCommitting to: /tmp/testbzr-LDqLMO.tmp/bzrlib.plugins.rebase.test_blackbox.TestRebaseSimple.test_rebase_merge/work/main/feature/\nCommitted revision 3.\nAll changes applied successfully.\nCommitting to: /tmp/testbzr-LDqLMO.tmp/bzrlib.plugins.re [20:34] base.test_blackbox.TestRebaseSimple.test_rebase_merge/work/main/feature/\nCommitted revision 4.\nAll changes applied successfully.\nCommitting to: /tmp/testbzr-LDqLMO.tmp/bzrlib.plugins.rebase.test_blackbox.TestRebaseSimple.test_rebase_merge/work/main/feature/\nCommitted revision 5.\n' [20:35] that's a bit unreadable [20:36] maxb: but blackbox tests especially look at the output the cli produces [20:36] if that doesn't match with what the test expect, it flags [20:38] maxb: ask 'is the new output what I want'? [20:38] maxb: if yes, change the test. If no, change your code. [20:38] hmm. I wonder why I am getting stderr output when before there was none at all [20:39] All I've done is change rebase parent selection logic === chx_ is now known as chx === menesis1 is now known as menesis [22:13] morning [22:15] igc: good morning. [22:15] poolie: top of the morning, to you too. [22:15] hello [22:16] poolie: Strange, but have no questions to ask right now... [22:16] hi poolie [22:16] hi igc [22:18] spiv, hi? [22:18] OldAl: so pull, merge and push are all in handled in the one qbzr module IIRC [22:19] OldAl: bound branch or otherwise, pull is the way to update a local mirror [22:19] so it's pretty useful [22:21] igc: I am sure it is. [22:22] OldAl: so starting on a small change to qpull is a good, safe way of beginning qbzr hacking [22:22] Have 'they' (who?) used gtk or Qt for the GUI of qrun? [22:22] I would prefer to put my hands into Qt rather than gtk [22:23] OldAl: any command startng with 'q' is Qt based, not Gtk [22:23] igc: Ah, great! [22:23] igc: that puts my mind at ease on that score. [22:24] OldAL: one more social thing: in open source, we value self-selection over central assignment [22:24] so me assigning bits of work to people isn't the way to go [22:24] igc: I know tha well. [22:25] igc: Why not suggest stuff to start on - individually. Nobody should mind suggestions... [22:25] instead, I ask for volunteers, people enter changes in our bug/wishlist tracker and assign them to themselves instead [22:25] igc: not if they want to work on it. [22:25] OldAl: sure, individual suggestion for people keen to get started in excellent [22:26] that's why I'm suggesting you working on qpul first :-) [22:26] OldAl: I'll raise the minor bug I'm thnking of now [22:26] igc: I gather that and am happy to look at. But then there is a bigger question of how to get [22:27] igc: things to speed up in the "documentation" group and I thought I would make a suggestion - to you :) [22:28] igc: But I don't want to do it in public before I suss out if you would be agreable. [22:28] igc: No conspiracies - but just that I don't want to be too big a PITA with suggesting something that my appear as extra work. [22:30] igc: If you are satisfied with the response, that is fine: if not, then there is a suggestion that I want to make, if agreeable to YOU, as it would affect you if you accept it. [22:32] OldAl: here's the bug you could work on: bug 486843 [22:32] Launchpad bug 486843 in qbzr ""learning help" in qpull is too verbose" [Undecided,New] https://launchpad.net/bugs/486843 [22:32] OldAl: don't worry about public suggestions being a PITA [22:33] OldAl: even if I don't agree, someone else may step up and do it or refine your idea [22:34] OldAl: if in any doubt, favour open communication over private emails and IRC chats [22:34] ubottu: Thanks for the bug 486843. "Being too verbose" is my problem, usually. I do feel that verbosity is superior to lack of clarity. [22:34] Launchpad bug 486843 in qbzr ""learning help" in qpull is too verbose" [Undecided,New] https://launchpad.net/bugs/486843 [22:34] Error: I am only a bot, please don't think I'm intelligent :) [22:35] ubottu: So who is the midget inside of you? (Reminds me of a chess playing "machine") [22:37] OldAl: ubottu really is code, not a person [22:38] igc: I am in agreement with "public" preference to "private". Alas, some things, like disabilities, may want to be kept private. [22:38] OldAl: very useful though - just say "bug xxx" and ubutto will display the bug summary [22:38] OldAl: right. Personal stuff is best kept personal [22:38] and private [22:38] igc: xxx is the number? [22:39] yep [22:39] just #xxx might be enough as well [22:39] iirc [22:40] igc: Actually, I don't mind to talk about *my* disabilities, but what I do - or do not do - affect the disabilities of others and that I want to keep private. [22:42] igc: also things like my experience and qualifications - it is rather a long list and requires explanation, but I feel it would be helpful to know for people that may try to get something useful done by me... [22:43] igc: Basically, again I try not to be a PITA, seldom successfully. [22:44] igc: Clarify: my tries at avoiding being a PITA are seldom successful. [22:45] OldAl: so my suggestion is to create a Wiki page or blog with your experience and skills. See http://ianclatworthy.wordpress.com/about/ for an example [22:50] igc: Just looked at your blog. Impressive! I guess for me it would be easiest to make something vaguely similar in my home page on PCUG and just pass on the open address, which, however, is not obvious to casual visitors (like there are no links to it in the face page.) [22:54] igc: clarification: PCUG is the PC Users Group in Canberra, very supportive of ONE OS, though not from the bunch of people who run the Internet Service, the admins. (The admins are very supportive of their independence of any elected bodies of PCUG :) ) [22:55] igc: My last sentence is clear as mud... [22:55] OldAl: sure. Or add http://bazaar-vcs.org/OldAl [22:57] igc: OK, that is simpler! Easier to see for anyone that may be interested in my experience and qualifications. Will give it a good run! [22:57] igc: Thanks - have a call from my real boss! [23:04] Good morning. [23:08] hi spiv [23:18] hello spiv [23:19] spiv, enquiring minds want to know about the patch pilot project :) [23:19] poolie: well I'm on this week [23:19] in my personal time [23:19] i know [23:20] i am obliquely asking spiv to send a mail about how the first week went [23:20] poolie: composing mail about that :) [23:20]