[00:01] LarstiQ: thanks :) [00:02] LarstiQ: thanks for the tips though, I think it's important to mention. [00:02] james_w: last_revision_info also returns a revno, last_revision does last_revision_info()[0], and for Branch6 formats last_revision_info does not touch the revision history, since the information is stored in that branch format [00:02] LarstiQ: great, thanks. [00:02] james_w: it is especially fun and noticable when doing a revision_history() of Mozilla over the network [00:02] 9Mb of revision history just to print the revno, ugh [00:03] Zectbumo: I'll go to bed in a second, but I'll read backlog [00:03] Zectbumo: quite a bit of work has been done last summer and it is functional, it could use more polish [00:03] Zectbumo: I suppose you're a VS user? [00:04] LarstiQ: I just started [00:04] er, was forced [00:05] LarstiQ: well, I'm going home from work now, so no worries [00:06] Americas timezone? [00:06] yup [00:08] beuno: hi [00:11] mwhudson, hey [00:12] beuno: so i heard a rumour you had some kind of php app we might want to steal ideas from? [00:13] ideas/css/ajax/voodo [00:13] beuno: i have to say i'm a little tired of suggestions like "implement a nice interface with CSS and javascript to navigate history." [00:13] because the design is much, much harder than the implementation :) [00:14] mwhudson, right. I've actually been working on a web app for bzr for the last year [00:14] so I have quite some work on the interface bits [00:14] I should be arriving back home on the 20th, and my primary use of free time will be to get the interesting bits out of it [00:14] and open source them [00:15] and maybe let you take a peak beforehand to see if it's useful for you [00:15] the rest of the PHP should follow, just think the css + ajax might be of more use now [00:16] that would be very cool, yes [00:16] so, it wouldn't be impossible to slap a new interface on that? [00:16] s/that/loggerhead [00:18] shouldn't think so [00:19] cool. I'll be more than happy to work on this with you, using what I already have, or starting from scratch if there are good reasons to do so [00:19] is there a public facing version of this app anywhere? [00:19] mwhudson, no, that's the problem [00:20] but I will get it up somewhere I can let you look [00:20] beuno: as i've been posting to the list, it's developing a coherent and sensible model for navigation that's got me stumped at the moment [00:20] beuno: can you explain briefly what you did? [00:23] mwhudson, yes, let me try [00:24] basically, what I have right now, is PHP getting all bzr history into a database. All that is used as a part of our internal workflow software, so there are many unrelated bits. Among those bits, you have one, that is currently broken due to last minute changes, which is history navigation through an ajax interface [00:25] since most users that use the system are bzr-ignorant [00:25] I went through a few redesigns with a few of the graphic designers to make it easier to understand [00:26] that led me to a more comprehensive history navigation, but also to some bzr-unrelated things I have to strip out to open it up to the world [00:26] it's all CSS + mootools for ajax effects [00:26] so it should be easy to port on top of pyhton [00:27] ok [00:27] there are also some nice looking icons I have to make sure aren't used somewhere else [00:28] that still doesn't explain to me what the history navigation is like :) [00:28] (sounds pretty cool though) [00:28] ah, right [00:28] let me try and describe from memory [00:28] I don;t think it's too much far off what you have right now, at least what I saw on the last PNG [00:29] it shows all revisions, and you can expand them with ajax as many levels deep as there is [00:29] each one has all the relevant information on it, commiter, time, message, files changed/added/removed [00:30] everything related to that revision can be seen by clicking on it [00:30] and without refreshing the page [00:30] do you have separate views for annotate/file listing/revisions ? [00:30] revisions with diffs, rather [00:30] there are some additional links which take you to per-file histories [00:30] yes, although, if IIRC, we don't have annotate yet implemented [00:30] just file listing and revisions [00:30] ok [00:31] and you jump from one to the other if needed [00:31] of course, none of this seems like much unless I can show [00:31] which I can't, because I'm 16.000km away from the office [00:32] when looking at the file history, do you show the actual revisions that change the file or the mainline revisions that merge the changing revisions? [00:32] (and, about 10 hours work to fix new bugs introduced when updating it) [00:32] (is there much of a mainline/non-mainline distinction?) [00:33] actual revisions that change the file is the one that sticked. Before, we showed the files, just in a very greyed out color to show that it wasn't actually touched in that revision [00:34] i'm not sure i understand [00:34] you tried more than one approach? [00:34] yeap [00:34] it evolved over time as people complained [00:35] it's been a long year :D [00:35] interesting :) [00:36] yes, I would of really liked to get this out in the open before, but I just kept postpoing til I could clean up the code [00:36] for which I ended doing things directly on bzrlib [00:36] which ended up with me at the sprint [00:36] heh, i know that one :) [00:37] so now I have to go back to the start [00:37] I'm sure you do much better ;) [00:38] my thought after the sprint where [00:38] if I can show what I have right now to you [00:38] and you feel it's worth working on [00:38] I can invest a few days in abstracting a bit more so we can share the same interface code [00:38] and make sure all future work gets merged back both ways [00:39] (or even for some other project, I suppose) [00:39] well, i'm definitely extremely interested in seeing the results of incorporating a year's worth of user feedback [00:39] also, we don;t use IE at all, so there might be some IE-specific CSS and Javascript bugs I don'y know about [00:40] well, i don't either (surprise, surprise) but i'm sure we can find some way of testing it [00:40] (my typing is awful today) [00:40] hehe [00:40] good to hear [00:41] and/or just paying someone to make the problems go away [00:42] i also don't know and don't really want to know php [00:42] right, we can talk about that later on, I might be able to put in some time from my web designers into it, but it depends on how bad it is :p [00:42] I'll make sure none leaks into the interface [00:42] has anyone written a plugin for p4 support? [00:42] beuno: so, when do you think you'll have something i can look at? [00:43] some time after the 20th? [00:44] mwhudson, exactly :) it should be before the end of march, it's high on my ToDo list, especially after the fine folks from Canonical fed me and gave me nice slippers for a whole week :p [00:44] nDuff: yes [00:44] nDuff: that is, I recall robey posting about it [00:45] beuno: i'm away for the last week of march (roughly speaking) [00:46] mwhudson, ok, I'll probably mail to the list than, start gathering feedback so you have a much bigger backlog to read when you get back [00:47] or, spend some more time polishing it [00:51] beuno: well, i don't want to imply anything at all about you, but i always get a little bit afraid when people say they will do something "the week after next" or so [00:51] that it will never happen [00:52] so the sooner the better, i can live without polish :) === RAOF_ is now known as RAOF [00:52] sounds reasonable [00:52] I'll just send it out as soon as I get it running again [00:52] at the very least. screenshots [00:53] thanks a lot! [00:53] thank me when it's sent :D === kiko-afk is now known as kiko-zzz [01:06] Odd_Bloke: ping [01:07] abentley: Pong. [01:08] I'm just submitting one of your patches. What shortname would you like in the log? Odd_Bloke? dwatkins? [01:10] abentley: Go for the latter. :) [01:15] Odd_Bloke: done [01:15] New bug: #199629 in trac-bzr "Expose bzr source code?" [Wishlist,Triaged] https://launchpad.net/bugs/199629 [01:17] http://wingolog.org/archives/2008/03/11/using-newfangled-version-control-systems-from-emacs [01:17] last paragraph is quite interesting. [01:24] abentley: Thanks. :) [01:31] what's the bug for renaming .bzr.backup? [01:32] jml, bug #124325 [01:32] Launchpad bug 124325 in bzr "Naming of .bzr.backup directory makes deleting .bzr easy" [Medium,In progress] https://launchpad.net/bugs/124325 [01:32] thanks [01:32] welcome' [01:33] beuno: you were working on renaming the bzr package at the sprint. right? [01:33] how's that going? [01:34] jml, sent a patch to the list during the sprint [01:34] sweet [01:34] let me fetch the merge request for ya' [01:34] (didn't get merged) [01:34] http://bundlebuggy.aaronbentley.com/request/%3C2422be180803040403y6873e4f0pa7fffe32551a12a1@mail.gmail.com%3E [01:35] feel free/incentived to review (approve?) it :D [01:36] I can't approve it -- don't have voting privs [01:37] well, at least you can make some noise! [01:37] beuno: and I'm running out of review juice for today :) [01:37] aw, it's fine [01:38] it's almost 3am, I should be sleeping anyway [01:38] beuno: Whereabouts are you ATM? [01:38] Odd_Bloke, still in Prague til thursday [01:38] beuno: Cool. [01:38] then, Madrid for a week or si [01:39] then, home! [01:39] anyway, I'm off to sleep [01:39] night everyone! [01:40] night beuno [02:31] How long does a PQM run take? [02:33] Also, it turns out that I'm the top contributor to bzr: https://launchpad.net/bzr/ [02:33] Who know? [02:33] *knew [02:40] Odd_Bloke: I think it's more than 30 minutes. [02:40] more if the queue is busy, but it shouldn't be at the moment. [02:41] although I can't remember if we still share a PQM with launchpad, which obviously slows things down. [02:41] There's a page somewhere to view the current queue state, but I can never remember where it is. [02:44] Odd_Bloke: it seems you get a lot of karma for https://code.launchpad.net/~daniel-thewatkins/ [02:46] james_w: Yeah, thumper said that that's the case currently. [02:46] Because few people are using LP Code. [02:46] Well, the current PQM run has been going for around an hour and a half. :p [02:47] (https://pqm.bazaar-vcs.org/) [02:48] james_w: register some more barnches and you too can bask in the karma glory [02:48] Odd_Bloke: aren't you in the UK? [02:49] thumper: I am, yeah. [02:49] Odd_Bloke: what the hell are you doing up then? [02:49] thumper: The work I should have been doing last week. :p [02:50] :) [02:51] thumper: :-) [02:52] james_w: bzr has a separate PQM instance these days. It may be running on the same box though. [02:53] Odd_Bloke: not necessarily few using LP code, just many less than use bugs [02:53] jamesh: thanks. [02:56] thumper: Yeah, that's probably true. Surely that's always going to be the case though, so those who do use it will always have high karma? [02:56] (Not that that is necessarily a bad thing) [02:56] Odd_Bloke: perhpas [03:46] it would be great if someone could review http://bundlebuggy.aaronbentley.com/request/%3C1204811147.4995.4.camel@ganieda.vernstok.nl%3E [03:46] I think it's a great thing to have, and is quite simple. [03:47] Odd_Bloke: you're clearly not doing work, I keep getting bur tracker emails triggered by you :) [03:49] james_w: I'm multitasking... *shifty eyes* [07:16] hi all [07:16] i want one simple help [07:16] i had setup bzr [07:17] also i know how to use bzr upto a extent for simple project containing only files. [07:17] but i dont know how to use this bzr for database related projects. [07:17] how this bzr synchronise databases from client to servre ? [07:17] server* [07:18] can anybody put light on this ? [07:20] what do you mean by 'bzr synchronise databases'? [07:21] if you want to upload/download revisions from the server, use bzr pull and bzr push [07:21] but you probably mean something different, as you said it in context of 'database related projects' [07:40] Humm. [07:40] How should lp: URLs be represented in branch.conf? [07:40] It seems they're expanded. [07:50] luks: im asking about databases synchronisation [07:50] consider i have a project on server which have mysql database [07:51] bzr versions files, not databases [07:51] (This is with bzr.dev, with the directory service stuff.) [07:51] luks: yes yes [07:51] thats only i was aksing [07:51] luks: okay means bzr ll not play any role for database synchronisation ? [07:52] right [07:52] luks: okay thank you [08:09] Ackpth! bzr-svn opened like a hundred connections. Finally, most of the way through, it bombed on some branch. [08:10] There was a + sign in the URL> === doko_ is now known as doko [08:48] luks: need one more help. im bit confuse in that [08:48] i ll explain you my scenario first [08:48] ok [08:48] i have bzr setup on server [08:48] i have one project called VTS on server [08:49] i have 2 clients [08:49] im checking out that project on client using bzr checkout sftp://192.168.100.96/var/www/VTS VTS_local [08:50] on client i do changes in file and after that i do ... [08:50] bzr commit and bzr update [08:51] but now the problem that i facing is that, i cant browse the changes through browser on server...means using http://192.168.100.96/VTS/ (this is server) [08:52] luks: im cuming in 10 min [08:53] kanhaiya_kk: well, if you want to browse the changes via web, you need to install something that allows you to do that on the server [08:54] https://launchpad.net/bzr-webserve/ and https://launchpad.net/loggerhead/ are two options [08:54] maybe there are some others, but I don't know of them [09:15] bzrweb. [09:25] luks: okay [09:26] but the procedure, im adopting is right na ? [09:29] luks: i dont want to browse bazaar repository. simply i want to browse my project. [09:29] not sure what does that mean :) [09:30] oh [09:31] are you getting what i mean to say ? [09:31] I probably repeat this too often in this chanel, bzr is a version control system [09:31] not a deployment tool [09:32] if you want the working tree on the server, you need to create it (but I assume you already have it) and update it everytime you push to it [09:32] that is right. but is it necessary that i have to ssh server and do "bzr update" for reflecting changes [09:32] there is push-and-update plugin, which does both in one step [09:32] okay [09:33] but I'd still consider a different approach [09:33] using 'bzr push' to update a live website just doesn't seem right to me personally [09:33] tell me what is your approach ? [09:33] okay okay [09:33] then ? [09:34] I'd use 'bzr checkout' for the website [09:34] and then run 'bzr update' every time I want to update it [09:34] but I'd push to a different location, never to /var/www [09:35] this is for what ? [09:35] means due to ? [09:36] it gives you separation between the source code repository and the running website [09:36] okay. [09:37] just like I'd never install a svn commit hook to update a website, I'd never push directly to a running website from bzr [09:37] then what do you do for browsing the changes from server ? [09:37] 'bzr update' [09:37] by doing ssh to server? [09:37] yes [09:38] that means i have to provide access to server to each developer ? [09:38] you would need to do that with the push-and-update plugin, too [09:39] okay fine. [09:39] is this comes with apt-get ? [09:39] there are other options, like running it regulary from cron or watching for changes in the branch [09:39] what? the plugin? [09:39] yes [09:39] no, I don't think so [09:39] okay [09:40] https://launchpad.net/bzr-push-and-update [09:40] okay. [09:40] now just i want your suggestion ? [09:41] my suggestion was to not do it automatically [09:41] which is not what you want :) [09:41] generally we do opensource development on CMS,PHP projects. [09:41] no thats not the case [09:42] I too am often surprised at the number of people who want the VCS to handle deployment... [09:43] Always seems weird to me. I did it once with CVS many years ago, and it was horrible. All sorts of reasons come to mind to avoid it. [09:43] so what procedure should we follow, so that each developer dont have to bother about time. means by simply giving 2-3 commands he can commit changes and also can browse the changes from server, without giving them ssh access to server [09:43] Not least among them, that what's in the VCS isn't directly what needs to be deployed. There's all sorts of things in there that shouldn't go on a live site, the arrangement of the files that should isn't the same, and there's often stuff like permission frobbing that needs to be done too. [09:46] Need some kind of continuous intgration server that runs a build / deploy script when you push to a certain branch [09:46] luks: ? [09:46] Shame there are no server-side commit hooks yet ; but you could probably adapt push-and-update given that it just uses SSH to run "bzr update" on the target branch after it runs [09:47] kanhaiya_kk: I don't know of anything that wouldn't involve scripting [09:47] okay [09:47] but as I said, the whole 'developer dont have to bother' seems wrong to me [09:47] the developer _should_ bother [09:48] and not all changes require just updating the working tree [09:48] for example sometimes you need to change the schema [09:48] (DB schema) [09:49] how would you do it if your main repository is the running website and automatic 'bzr update' is the only way to access it? === weigon_ is now known as weigon [10:02] luks: hmmm [10:02] I have an idea to solve this, but it would indeed need scripting [10:03] a lot of the scripting could be standardized [10:03] mlh: hmm tell me [10:04] keep the metadata has ordinary textfiles [10:04] im totally confused. Our team is thinking to use bzr for development server [10:04] generate or sync to/from the text file with scripts [10:04] version the textfile with bzr (or any other vcs) [10:06] okay [10:07] The rest is just a SMOP :p [10:07] :-) [10:32] hi, I'm using a "bound branches" model...but on the server repository I always get "working tree out of date"...can you help me? [10:34] rlubert1: Pushes do not update the remote tree [10:34] I wonder if there is some kind of FAQ page on the wiki [10:34] * awilkins vote to have this emblazoned in letters of fire at the top of all bzr documentation pages. [10:34] There is. [10:35] Whether it includes Q's that are really FA is another matter... [10:35] awilkins: ok, thanks [10:35] can i get examples for bzr enabled projects ? [10:35] awilkins: If you think it would help; it only tells you so every time you push :p [10:35] No offense to yuo, rlubert1, it just constitutes about 90% of the questions asked in here :-) [10:36] kanhaiya_kk: Projects that use bzr? [10:36] http://bazaar-vcs.org/WhoUsesBzr [10:36] awilkins: I was following this example: http://bazaar-vcs.org/BzrUsingBoundBranches, no problem [10:37] * Peng notes that WhoUsesBzr is linked to from the homepage. [10:44] Peng: yes [10:54] New bug: #201362 in bzr-svn (universe) "Bzr-svn cannot be loaded " [Undecided,New] https://launchpad.net/bugs/201362 [10:57] awilkins: "bzr init-repo --no-trees sftp://centralhost/srv/bzr/X-repo/ " that is even better...as clearly mentioned on the doc page...... /me [12:31] how do I create a branch and check it out of an old revision? [12:32] bzr branch -r 10 path/to/parent [12:32] what is path to parent? [12:32] Well, wherever the parent.. the thing you are branchingh [12:33] I'm branching my local copy [12:33] do I need to create a new dir in bzr to do this? [12:33] Well it'll need to be stored somewhere new [12:33] wow [12:33] ok [12:33] cd .. ; bzr branch -r 10 whereIjustwas [12:34] Or bzr branch -r 10 . ../some-new-branchname-here [12:34] Or other variations on the same theme [12:34] got it === mrevell is now known as mrevell-lunch === FreeNode is now known as herb === kiko-zzz is now known as kiko === FreeNode is now known as herb [13:24] Moin. [13:26] When I'd like to start over with "bzr init", can I just rm -rf .bzr to remove any old bazaar traces from a project? [13:28] Or how do I else remove bzr out of a project? [13:28] speakman: You can, yes. However, 'mv .bzr .bak.bzr' might be better, to ensure you don't lose any data accidentally. [13:28] idea; add an "uninstall" option to "bzr"; "bzr uninstall" which removes bazaar from the project. A more clean way than rm -rf .bzr. [13:28] Odd_Bloke: ok thanks! [13:30] (can I post this idea as a bug, or how do I post feature requests?) [13:30] New bug: #201399 in bzr "bzr commit filename commits all files" [Undecided,New] https://launchpad.net/bugs/201399 [13:32] i have one question: can i install Bazaar on usb drive and then connect to server from different clients? [13:33] anybody? [13:33] I'm not sure I udnerstand the question asked. [13:34] oh my dumn english :( [13:34] can i install bazaar on USB? [13:35] I think that's more of a Python question [13:36] you could check this site out: http://www.portablepython.com/ [13:36] on the bazaar-vcs.org i read: "Bazaar: ... Embeddable. A key design feature of Bazaar is support from the ground up for pluggable storage formats..." [13:37] i want embed Bazaar into my application. Is it real? [13:39] nonamekiller: bzr is implemented mostly as a Python library, so especially if your application is written in Python, it's very easy to use it as a library. === mrevell-lunch is now known as mrevell [13:43] radix: can i embed it in my java application? [13:44] nonamekiller: that's harder :) the easiest for *simple* embedding outside of the Python world is probably by spawning system commands to run bzr. [13:45] I don't know if bzrlib works in Jython, but if it does that could also be a possibility. [13:45] Jython is only compatible with python 2.2 [13:45] It would be possible to write Python programs that deal with bzrlib to provide an easier interface. [13:45] It's pretty good with IronPython though! [13:45] Although I'm told it will be better after the Google Summer of Code [13:46] Peng: You are joking righ? [13:46] awilkins: I don't think so. [13:46] Also, didn't Sun just hire the main Jython devs? [13:46] awilkins: I'm pretty sure it's been updated beyond 2.2 [13:46] awilkins: it might not be fully compliant with any later releases, but I know it's got some features from later releases. [13:46] Peng: I've tried reasonably seriously to get it to work in IronPython, the test suite wont run because there's no "subprocess" module [13:47] And that's just the first stumbler [13:47] Well, not wuite the first, I patched around the first [13:47] Peng: they hired some guys. maybe jython will get better [13:47] Hum. [13:48] But IronPython introduces a whole bunch of issues, platform caps detection, missing modules, etc [13:48] I'm hoping some people will notice pypy-jvm and get that working :) [13:48] I thought you or someone else said IronPython was pretty close. [13:48] My mistake then. [13:48] IronPython can be cadged into loading bzrlib if you patch bzrlib [13:48] (this is the Lapha 8 build of 2.0) [13:49] 2.0 alpha 8 is much closer to python 2.4 / 2.5 than 1.1 [13:50] I wish I had the time and motivation to take it further, but I don't ; the first thing you'd have to do is write a subprocess module ; either as a lib, or as a builtin for IronPython [13:50] The win32 implementation of subprocess in lib doesn't work because it needs mscvrt which isn't a builtin of IronPython [13:51] Although I suppose it might be easier to make a mscvrt module for IPY than a subprocess :-) [13:51] I had a look, but I've not had the time to get around the Scary Attribute Jungle that's in there [13:52] is there exist portable python? [13:52] I was told Jython is pretty close (by Guillermo Gonzales, I think) [13:53] nonamekiller: CPython runs on pretty much all the platforms anyone cares for. [13:53] (not counting mobile ones, but it's pretty good there too :) [13:54] There's also a USB distro of the Win32 CPython if that's what you mean [13:54] yes! [13:54] thx! [13:54] http://www.portablepython.com/ [13:54] oh, is that what "portable" means these days? [13:57] You might have to write a quick batch file to open a command prompt with the right ENV variables set up (to make it less of a hassle typing t:\python25\python.exe t:\python25\scripts\bzr all the time) [13:59] the py2exe'd installer of bzr and setting BZR_HOME should work on an USB drive [13:59] True, but it's not as fabulously flexible [14:00] Although it's probably faster to load [14:00] Can you use plugins with that version? [14:00] sure [14:01] Apart from some of the GUI ones, according to the page [14:01] you can use even the GUI ones [14:01] Even ones that use GTK and things like win32com? [14:01] if you install them, then yes [14:02] but that applies also to bzr with standalone python, since they are not in stdlib anyway [14:03] there isn't really many differences between py2exe'd applications and standalone python [14:03] I've always just used the "Full Monty" (ahaha) because I knew I'd probably end up hacking the code anyway. [14:24] New bug: #201409 in bzr-loom "pulling a loom resets to last-recorded state" [Undecided,New] https://launchpad.net/bugs/201409 [14:32] anyone who can help a newbie ? [14:32] p_masho: sure, ask away. [14:32] have taken a snapshot of a repos.. have hacked two files and addes an image ..so [14:33] done > bzr add path/to/images/foo.png === thekorn_ is now known as thekorn [14:34] but.. my patch file which I assume I get off bzr diff > patch.file has got all sorts of stuff [14:40] p_masho: What do you have a patch for, and what stuff? [14:50] Anyone recommend a tool for drawing directed acyclic graphs for illustrating version control concepts? [14:50] visio :/ [14:50] * awilkins dry-retches [14:51] other than that, dot from graphviz [14:51] I have it, but I hate it because my *other* project is a Big Ball of VBA/VB6 Mud plugin for Visio [14:51] * awilkins will probably use Visio *spit* === Af1 is now known as AfC [15:03] xfig! :p [15:05] New bug: #201422 in bzr-gtk "'WorkingTree4' object has no attribute 'iter_changes'" [Undecided,New] https://launchpad.net/bugs/201422 === mvo__ is now known as mvo [15:09] p_masho: can you explain what you mean by "all sorts of stuff" please. [15:11] james_w: its perl (spits) so its the compiled files generated by make that turn up.. ;-( [15:11] p_masho: ah, so you want to ignore them rather than version them I assume? [15:12]