[01:05] Would anyone mind pointing out some of the advantages of Bazaar over other dvcs's ? I primarily use Git, but due to its lack of Windows compatibility options I've been experimenting with Mercurial (and not liking it very much)...will Bazaar give me a workflow similar to Git ? [01:11] the major difference is that bzr doesn't bundle checkouts, branches and repositories into one thing [01:12] bob2: and git does? can you elaborate on that? [01:14] bob2: hmm, i'm just reading now, it appears that branches want to go in seperate directories...this is one thing about mercurial i didn't like [01:15] mercurial does it in a bizarre way, afaict [01:15] bzr's is very simple: a branch is a directory with some stuff in it [01:15] bob2: I don't think that's an accurate depiction of bzr vs. git [01:15] davidstrauss: fair enough [01:16] bob2: well people in #mercurial tell me they prefer a workflow where you have multiple clones of a reposiitory (each in their own dir) [01:16] bob2: Using shared branch storage with branches in immediate subdirectories is basically the same [01:16] hm [01:16] except more annoying [01:16] bob2: the depends on what you're doing [01:16] that* [01:16] I kind of liked git's "transparent" branches, where you could do everything from only 1 dir [01:17] anyway [01:17] bzr has repositories, and bzr switch [01:17] Multiple branches per directory support will likely be in core eventually because it's a foundation of the Looms module. [01:17] in fact thats the major thing keeping me from liking mercurial, i dont want my project folder littered with extra dir [01:18] well, that's unrelated to bzr [01:18] repository can go anywhere [01:18] one checkout can move amongst branches with 'bzr switch' [01:18] i run my development server out of the folder i'm using, so if i switch branches, i would have to restart the development server in that other folder, which is not ideal [01:18] bob2: Yes, and you can bzr switch using local branches as the source, which keeps your working tree in one directory [01:19] if that makes any sense [01:19] yup, which works well [01:19] Evanlec: nope [01:19] i do web development [01:19] Evanlec: bzr switch is the same as git co - change this working copy to point at another branch [01:20] bob2: oh [01:20] Evanlec: the only difference is the branches aren't hidden in .git [01:20] Evanlec: The more fundamental differences between bzr and git are the commit interface, the merge model, and file vs. content tracking [01:20] bob2: what do you mean hidden ? [01:20] Evanlec: might be easier to just try out bzr [01:20] bob2: yea, i spose [01:21] does bzr have an equivalent of the git rebase command? [01:21] that is also something sorely lacking in mercurial [01:21] Evanlec: yes, though an extension, though i argue it's unnecessary [01:21] they made an extension that does it, but its really not as good as gits [01:21] Evanlec: http://fourkitchens.com/blog/2009/04/20/alternatives-rebasing-bazaar [01:21] davidstrauss: i don't understand how it could be unnecessary [01:22] Evanlec: then read my post :-) [01:23] davidstrauss: the link you just posted u mean? [01:23] k [01:23] Evanlec: yes [01:23] davidstrauss: I don't find the latter two matter much ime, but the index is an important difference, true [01:23] reading [01:23] Evanlec: anyway, in bzr, a branch is a directory, not a magic file in .git/refs [01:23] Evanlec: also, bzr supports the checkout workflow, which git does not [01:24] Evanlec: a checkout points at a branch (and can move amongst them with bzr switch) [01:24] Evanlec: sometimes the checkout and branch are the same dir, sometimes not [01:25] a branch is a directory...but i can use bzr switch in the same dir and not notice any directory 'change', like as if i didn't 'cd /some-new-branch' [01:25] Evanlec: yes, though you would do so on a checkout directory [01:25] branch != checkout [01:25] checkout = working set of files [01:26] branch = pointer to a particular revision [01:26] ah k [01:26] repository = big bag of revision data [01:26] so a branch is just a modified version of the same checkout ? [01:26] so to speak [01:26] they can all be in the same dir (e.g. when you do 'bzr branch http://blah/blah') or not [01:26] bob2: checkout is possibly the most poorly named command. checkout can create a checkout or it can populate the working tree of a branch [01:26] not at all [01:27] bob2: so, "checkout" in a branch creates a working tree, but you run "bind" if you want to turn it into a checkout :-/ [01:28] bob2: I think the former usage needs to go away [01:28] a branch is like a branch in git - it points at a partinaclr revision in the repository [01:29] yea [01:29] in git you just 'checkout' branches [01:29] and thats about all checkout does [01:30] altho, you can also check out a revision [01:33] i dont understand why this is saying that gits rebase throws away history [01:33] to me, it seem like ti just adds to history [01:34] and then sticks your commits in that branch on top [01:36] i'm only doing primarily 2-person workflow here, i've no need to share my local branches [01:36] so rebase works fine for me [01:38] outside of git-land, people just don't bother [01:38] you commit, commit, commit, then merge [01:38] what if the mainline trunk moves ahead while you're commiting ? [01:38] Evanlec: you merge it in and commit [01:39] merge the mainline into your exp branch ? [01:39] Evanlec: naturally [01:39] but then when you merge your exp branch into mainline you get these out-of-place commits (of the merges in the toher branch) [01:40] Evanlec: so, bzr has rebase, and you can use it. [01:40] does it behave exactly teh same as git rebase ? [01:40] close enough, as I understand it [01:40] I don't rebase in either tool, so I can't really say [01:40] i was dissapointed with mercurials rebase [01:41] to understand the concern some people have with rebase, consider a change in mainline that breaks your code [01:41] if you rebase on top of mainline, your commits are now broken [01:41] even though they may not have conflicts [01:41] you mean when i merge back into mainline from my branch ? [01:42] if instead you were to merge mainline into your branch, and fix it as you merge, you don't have that problem - every commit still works. [01:42] or just when i rebase from mainline [01:42] ah i see [01:42] Evanlec: when you rebase from mainline [01:42] yea that makes sense now [01:43] so when you rebase, and this happens, you have to go back to your first commit and rewrite it to work against whatever rev of mainline you grabbed [01:43] and the next and so on [01:43] so you dont haev any problem with these extra 'merge' commits when you merge mainline into your own branch ? [01:43] this is neither good or bad, its just something you have to do with a rebase workflow. [01:44] Evanlec: no problem at all; bzr puts them off to the side and [in recent versions] doesn't show (any commit that isn't itself a mainline commit) by default [01:44] well the case you speak of sounds bad to me [01:44] lifeless: interesting [01:45] i was thinking that mercurial was going to be a 'simpler' version of git [01:45] but come to find out, its actually more complex for me because of the strange way i have to do things im used to doing in git [01:47] bzr is also quite different to git [01:47] yea [01:47] and trying to use your git workflow in bzr won't be the easiest thing [01:47] i guess i was hoping i could find a git clone (no pun intended) that was easier to use on Windows (for my new partner) [01:47] but the workflow people use with bzr is very easy and natural [01:48] just different [01:48] lifeless: thats what mercurial people said, but it didnt seem very easy, and certainly not natural, to me [01:48] was git your first DVCS? [01:48] yea [01:48] in fact my first vcs really [01:48] havent even used svn really === dereine is now known as dereine[OFF] [01:51] so with bzr the general thing is to just do your development [01:51] and not to spend a lot of time making the history pretty [01:51] heh [01:51] well not that my git history was very pretty [01:51] rather we try to have bzr report effectively [01:52] but thats what trying to not record merges and so on is all about ;) [01:52] everytime my local branch diverged from the remote, would get these merge commits (without asking for them) for a silly merge when they diverged by only 1 or 2 commits [01:53] so someone suggested i just use rebase instead of merge [01:53] and that seemed to solve that problem [01:53] it does remove the merge commits [01:53] but it has other tradeoffs [01:54] anyhow, [01:54] we have rebase [01:54] I have to head off for a while [01:54] enjoy! [01:55] alright thanks === `6og is now known as Kamping_Kaiser [04:40] Evanlec: So, did you get your question adequately answered? [04:54] davidstrauss: yes i suppose, thank you [04:58] Hello, is there an impact if I upgrade the bzr branch format to --rich-root-pack or --rich-root ? [05:02] hey, I have a question about the bzr push [05:02] I am working behind the firewall , I need a proxy to access the web [05:03] xiaohui: ok [05:04] when I solved the problem of using ssh through the proxy and bzr push it [05:05] I use *bzr push bzr+ssh://bazaar.launchpad.net/~MY_LAUNCHPAD_ID/PROJECT/name* and it show the progress bar [05:06] But it always stopped at the "coping inventory texts 2/5" [05:08] and some minutes later , it said that *connection was closed by the remote host , pipe broken ERROR* [05:11] does ssh reliably work otherwise? [05:38] AnAnt: yes there is, its a one way upgrade *and everyone that pulls or merges from you has to do it too* [05:38] lifeless: doesn't the upgrade happen if they pull ? [05:44] now, the progress show that *coping content texts* === bob2 is now known as Guest22735 === Guest22735 is now known as bob2 [06:57] will be life fail to be enriched in any way if i don't read that huge rebase thread? [07:01] I just moved the VM I run Loggerhead on from UML to a faster machine with Xen. I should've done some benchmarks beforehand. :\ [07:02] Peng_: did you see my branches targeted at using less ram? [07:03] mwhudson: Yeah. [07:04] Well, one. [07:04] My eyes glazed over at the tuple unpacking and I didn't try it. :P [07:04] Oh, right, the other was already merged. [07:04] :) [07:07] Hmm. Your other changes might've brought my normal memory usage from almost 125 MB to almost 115. [07:07] But it's only been a day or two, and I've crashed the machine twice and rebooted it two more times, so I can't say anything for sure. [07:10] mwhudson: So, do you want me to test anything? If so, what? [07:10] saved for posterity. I found it a pretty good short discussion [07:10] http://www.deze9.com/jp/wiki/?p=to_rebase_or_not_to_rebase [07:32] could some one of you give an URL of ssh server for testing my ssh is work behind the firewall? Thank you [07:36] Peng_: i guess testing the branch works would be cool [07:36] Peng_: have you had problems with memory consumption? [07:40] mwhudson: Define "problems". I've never woken up to find it swapping to death. [07:40] Peng_: well we have on launchpad, so i guess you haven't :) [07:40] mwhudson: Which branch? less-stupid or persist-_rev_info? [07:41] Peng_: persist-_rev_info contains less-stupid, so testing it would be more revealing i guess [07:41] * mwhudson afk [07:41] Peng: email me any problems you find :) [07:45] mwhudson: Oh god. The AbsentContentFactory exception attacks. [07:45] Anyway, unimportant. [07:51] mwhudson: Well, I'm running persist-_rev_info now. So far, nothing is exploding. :) === dereine[OFF] is now known as dereine [11:48] you've made reddit: http://www.reddit.com/r/programming/comments/8fbk6/to_rebase_or_not_to_rebase_bzr_irc_log/ === NEBAP|work is now known as NEBAP|away [12:07] Hi there. I never used version control before, and after some investigation I'm now deciding whether I should try bazaar, mercurial or plain svn. I wanted to try out each before deciding, and found out that bazaar is already installed in my ubuntu machine. can someone help me getting started? note: I'm new to linux as well [12:07] Waldir: do you have a project or do you want to start from scratch? [12:08] * dereine also startet only a short time ago [12:08] I was wondering whether I could test it on my website [12:08] ok then you have a folder? [12:09] er... where, on the server, on my machine? i currently have the site both online and in my machine [12:09] you can do both [12:10] its decentral so you don't have to worry about remote, you can work offline without knowing anything about the www [12:10] so I read. I don't know how to put that theory in practice though :P [12:11] so you have a folder [12:11] http://doc.bazaar-vcs.org/bzr.dev/en/mini-tutorial/index.html [12:13] ok, i'm installing the bzr-gtk so I have a GUI. huf, thanks, but I prefer not using the command line because the other people who will be working on this are even less familiar with the vcs than I am... I want the easiest possible, painless way to get started (assuming there is one :) ) [12:13] dereine: yes, please continue [12:14] Waldir: uh, not commandline ;) thats hard i will start bzr-gtk [12:14] Waldir: that is the command line [12:14] lol... it's not for me, its more for the others. even though most of them will probably be using windows anyway.. [12:15] see, you have to understand what you're doing, and the gui tools only help to hide that [12:15] oh, well... is there a better client for windows than tortoisebzr? [12:15] shouldn't the gtk and the qt version also works? [12:16] ooh [12:16] huf: I know but I was given very little time to choose a vcs and I can't have them wasting too much time learning to wotk with it. some have already used svn+tortoise (including me) but that's all [12:17] huf: anyway I read a lot about vcs in the last two days so I do understand the basics [12:17] (i guess, lol) [12:17] ok [12:17] click to your folder [12:17] ok [12:17] then menu -> branch -> intialize [12:17] current folder [12:18] it's something like /var/www/mysite.com right? [12:18] you can use every folder [12:18] so if you did this you can add files [12:18] just making sure :) [12:18] arg gui, its so "hard" to use ^^ [12:19] lol :( [12:19] you mean the context menu, with right-click? [12:20] yes [12:20] there you can add files [12:20] cause I can't find "branch" anywhere... [12:20] if you are ready with adding some files [12:21] see http://imagebin.ca/view/x8PyaX.html [12:23] thanks. but that's not the context menu, it's the menu bar on top, right? (on eihter case, there's no "branch" anywhere... sould I start bazaar somehow first?) [12:23] do you have olive? [12:23] i found it now [12:23] (olive) [12:24] lol I thought it integrated with nautilus somewhat like tortoise [12:25] hmm I clicked the applications -> programming -> olive menu item it but nothing happened.. [12:26] nevermind i tried again and it worked this time :/ [12:29] ok, I initialized the directory. the files that are inside it are added by default? or do I have to add them manually? [12:29] no you have to add them [12:29] ok [12:29] oh, there's no multiple selection! I have to add them one by one right? [12:30] ok, I added three files. there are some folders, can I add them as well or do I have to open them and add each file? [12:31] no clue, I just do bzr add; bzr ci -m 'Initial import' [12:31] Waldir: you should be able to add folders [12:31] ci? [12:31] dereine: commit [12:31] ah i have to learn the shortcuts [12:31] LarstiQ: thanks :) [12:31] ok, what now? [12:33] Waldir: commit it [12:35] am I supposed not to see the output of individual files inside the folders I added? [12:37] * LarstiQ blinks at Waldir [12:37] Waldir: what do you mean? [12:37] the files and inside the folders are showing status "unknown". the others in the root level had "added", and after the commit, "unchanged" [12:38] Waldir: ah hmm, that seems to mean the gui does not recursively add files in a dir if you add the directory. [12:39] I mean that the commit dialog showed something like "=== added file 'index.php' and then a diff, for each file in the root level. but for folders only showed "=== added directory 'app'" and nothing else [12:40] yes, it seems so [12:40] the tortoise client is able to do the recursion, I hope? [12:41] I believe I'll be the only one using linux for this project so if I have to use command line it shouldn't be too much of a problem [12:44] hello? [12:45] try it out :) [12:45] Waldir: I don't know if it can. If you could try it out and report bugs that would be nice. [12:45] Waldir: it should, imho. [12:46] I would love to submit bugs, it's just that I have external pressure to do this fast :) [12:47] I'll try it in my windows VM [12:49] oops.. it seems this version of virtualbox wont work with jaunty :( [12:50] i'll ask my colleague to do it on his win machine [12:55] how can i checkout a branch to a ftp? [13:03] dereine: it isn't clear to me what that would mean. Can you elaborate? [13:04] LarstiQ, dereine: ok, we installed bazaar+tortoisebzr on the win pc. how can i update my own files here with what he has there? (note: they are exactly the same) [13:05] Waldir: for sharing committed revisions, pull/push/merge [13:06] yes but he heas to put that on somewhere I can access from my machine, right? he has his own website, can we use that, via ftp or http? [13:13] Waldir: yes, any place you can both reach should do [13:13] Waldir: sftp is preferable to ftp though [13:14] Waldir: (you could also use `bzr send` to email revisions across) [13:14] LarstiQ: I think the centralized workfow could work best, since we are more than two [13:14] Waldir: he could do something like, `bzr push sftp://webhost/~/public_html/branch` and then you could `bzr pull http://webhost/~friend/branch` [13:15] Waldir: ok, then you will need a location all of you have write access to [13:15] gah, running out of power already [13:16] yes, that can be arranged. but doesnt sftp require authentication? will those commands work just like that? [13:19] Waldir: you'll need credentials. If you have an appropriate ~/.ssh/config or remote user same as local, and a key, it will work just like that. [13:19] otherwise it will prompt for a password [13:19] Waldir: you can also include the username in the url if you want [13:20] it isn't needed per se [13:22] LarstiQ: ok :) now, does the push command need to be done from the directory where the repository is right now in the window machine? that is, do i have to cd there first? [13:22] by the way, tortoise does the recursive add and commit ^^ [13:23] ah, good about the add [13:24] Waldir: bzr needs to know which branch you are pushing. Doing the push from within the branch would do that, or you could use -d to designate another branch. [13:24] Frequently you want to push after a bunch of commits, so that then turns into `bzr ci; bzr ci; bzr push;` from the same branch. [13:25] ok i didnt get this last one [13:25] why would i commit twice? [13:26] Waldir: `*hack hack*; bzr ci; *hack hack*; bzr ci; *decide to publish*; bzr push` would be less terse. [13:27] ok, I got it :) [13:27] I'll try this [13:34] LarstiQ: i just connected to my ftp server using the places menu (connect to server). However, using olive it keeps asking for my password. Can it be because I am already connected with this username? [13:37] LarstiQ: ok, I signed off from the ftp in nautilus, then tried the command line. the error returned was: bzr: ERROR: Target directory ftp://my-web-host/~/public_html/branch already exists, but does not have a valid .bzr directory. Supply --use-existing-dir to push there anyway. [13:38] Waldir: I pretty much only know the commandline interface well enough. [13:39] Waldir: for that last error, it is saying the directory already exists while it didn't expect that. You can tell it to go ahead by supplying that option. [13:39] perhaps it already existed because of the previous attempt using olive? even though it didnt accept my password :/ I'll try with that option [13:41] LarstiQ: FTP temporary error: 451 /httpdocs/branch-name/.bzr/repository/upload/7wwumk8piyp4jgva3o2k.fetch: Append/Restart not permitted, try again. Retrying. [13:41] Waldir: ick, you have a terrible ftp server there. [13:42] lol [13:42] too bad my friend went to lunch, I can't try his ftp server :/ [13:43] but this was already good for a start. I suppose tortoise will allow push and pull from the GUI so i should be able to convince the rest of them to use this :) [13:43] hmmm [13:43] Waldir: it certainly should. [13:43] perhaps I can try via http? [13:44] Waldir: note that with more people working on the branch, divergence will start to occur. At that point you will also need the merge command next to pull. [13:44] Waldir: pulling, yes. But http itself is not a writable transport so push doesn't work on that. [13:45] yes, i was afraid you'd say that :) about the merge, doesnt push/pull try automatic merge? [13:45] Waldir: if your http server supports WebDAV you can use the webdav plugin [13:45] Waldir: no, they do not. [13:45] but they'll tell me that I need to use merge if there is a conflict, right? [13:46] Waldir: yes. [13:46] Waldir: since you're going with a centralized style, you may also consider using checkouts instead: http://doc.bazaar-vcs.org/latest/en/user-guide/index.html#team-collaboration-central-style [13:46] I will try this out later when I get the pull and push working. about the WebDAV, I don't even know what that means :P === dereine is now known as dereine[OFF] [13:47] Waldir: that way you would not use pull/merge/push but update/commit for regular operation [13:47] hmmmm [13:47] Waldir: I'm guessing no then ;) [13:47] lolol [13:47] thanks for the link, I'll read that [13:47] Waldir: your host does probably support sftp instead of ftp, sftp is a waaaay nicer protocol [13:47] hmmm. let's see then [13:49] ssh: connect to host ftp.webhost.com port 22: Connection refused [13:49] bzr: ERROR: Unable to connect to SSH host ftp.webhost.com; EOF during negotiation [13:50] Waldir: perhaps another hostname than ftp.webhost.com for ssh access? But maybe they don't offer ssh (boo!) [13:50] (obs: I'm editing off my website's url cause it's so embarassing right now :P I have a redesign almost ready that will hopefully not make eyes bleed xD) [13:50] I'll check the email they sent me when they set up my hosting [13:50] hah, I have a high tolerance for eye bleedage ;) [13:51] k [13:51] perhaps there's something there [13:51] Waldir: if you're going to deploy websites with bzr, I suggest you look into the bzr-upload plugin. (http://bazaar-vcs.org/BzrPlugins) [13:53] there seems to be something right for the needs of everyone :) I am guessing this could be put in an easier to find location, perhaps in the main page have a set of common "I am a xxx and I want to use bazaar for xxxx" with links that have all the correct instructions and resources... do you think it would be a good idea? [13:54] (just the most common combinations, especially newbies, experts, svn/cvs guys, something like that) [13:54] Waldir: does http://bazaar-vcs.org/Scenarios fit what you're thinking of? [13:57] LarstiQ: more or less. perhaps the entries could be grouped to allow easier browsing, and the "newbie getting started" guide is missing there :P [13:58] Waldir: would you like to contribute that guide? [13:59] lolol I really think I should first be able to get my project working :P [13:59] Waldir: well, the good thing is that you can note things as you encounter them [14:00] although, looking at the scenarios, 'newbie' isn't really a scenario [14:00] I suppose so. I think I'll save this first conversation for future reference [14:00] ok [14:01] yes, it doesnt fit very well there, but such a guide would be really good [14:01] Waldir: how does http://bazaar-vcs.org/BazaarForWebDevs compare? [14:08] LarstiQ: it's pretty nice. However I think something specific for say windows users with tortoise, using minimal (if at all) command line instructions, with screenshots, would really help the, you know, hardcore newbies (or people not very much wanting to learn to use a new tool). I'm not one of these but I know many, many people who are [14:09] It's nice having something powerful that an experienced user can use and learn really fast, but we all have to start sometime, and by then we're kinda clueless [14:10] Waldir: yes, I agree. [14:11] my friend is back, we'll try his ftp server now [14:12] Waldir: launchpad.net might also be appropriate space for hosting your branch [14:14] I thought it wasnt possible to host closed projects there [14:14] Waldir: ah, that would count as not appropriate then :) [14:14] carry on, nothing to see here! ;) [14:14] lolol [14:14] * LarstiQ frowns at his bike keys still being lost [14:28] well, the ftp server accepted the push, but the files don't seem to be there. does push only save the .bzr branch there? [14:29] hmm, pull seems to be working, so I'm assuming that's the case [14:31] LarstiQ: I got this with pull: bzr: ERROR: These branches have diverged. Use the merge command to reconcile them. [14:32] and then, running merge: bzr: ERROR: Branches have no common ancestor, and no merge base revision was specified. [14:33] I suspect this might have something to do with the fact that we both created a repository in our machines, so they don't recognize each other as the same even though the file contents are indentical... is that right? === dereine[OFF] is now known as dereine [14:37] Waldir: exactly [14:37] LarstiQ: "Subversion and CVS are good choices for implementing this [centralized] workflow because they make it easy." on http://bazaar-vcs.org/Workflows#head-22b1dac636e54e501e9b5a0a235f21bb686b80ac I wonder if I really should be trying to set up bazaar... 9.9 [14:38] but still what should I do then, to fix that 2 repos problem? [14:39] Waldir: two main approaches: 1) drop one of them 2) graft them together [14:39] Waldir: seeing as you have just started, I would go with 1) [14:42] Waldir: the reason CVS/svn make the centralized workflow easy to set up is because you can't do anything else with them. [14:42] Waldir: given a svn setup and a centralized bzr setup, I find bzr easier to work with. [14:42] Waldir: YMMV [14:44] Well, I'm taking some time to get started, but that is not the problem. I think I won't be able to convince the rest of the guys to use this if they have to fiddle with the command line... [14:44] I see. [14:44] Waldir: Tortoise svn is certainly more mature than Tortoise bzr. [14:46] I'm afraid we'll have to go with that one for this project, then. I'd use bzr myself for my website if my host's ftp support was better :( [14:46] which sorely needs people using it and improving it if it wants to get to the TSVN level, but I digress. [14:47] Waldir: you can also use bzr with svn as a backend btw [14:49] my concern is just the interface. I'd rather use tortoisesvn with bzr as a backend instead, since it seems to be better for merging and stuff :P [14:50] * LarstiQ shudders [14:50] I really don't think so, but feel free to try :P [14:52] I don't know, I was just guessing :P [14:56] LarstiQ: Your solution worked btw: I just tried doing a change then commit then push and then pull on the windows machine (which is called "update" in the tortoise interface), and it worked. We were able to see the diff there, of the change I did here :) Only question now is how to revert a change, but that will be another day's lesson :) anyway, I already got started with this. Thanks a lot for helping with that. [14:56] I gotta leave now. See you next time! [14:57] Waldir: `bzr revert`? [14:57] Waldir: ciao! [14:57] LarstiQ: thanks :) bye! [14:58] LarstiQ: can i use bzr push for ftp to have a current version of a repo as checkout on ftp [15:02] dereine: in bzr terminology, push will not create or update working trees on ftp [15:02] or any remote transport [15:02] dereine: if you want to do something like deploy a website, the bzr-upload is a better choice ime [15:03] dereine: if you don't actually want a checkout, but just publish your branch so others can use bzr to interact with it, `bzr push` is all you need. [15:03] and it will work with ftp, presuming the ftp server isn't broken. [15:03] LarstiQ: no i need exact the checkout [15:03] dereine: sftp or bzr+ssh:// are usually better choices if available [15:03] i know :( [15:03] dereine: what do you need a checkout for? [15:04] the designer there don't use bazaar [15:04] so he edits the file with ftp [15:04] and i work on a local branch [15:05] thx anyway for the links [15:07] * SamB thiinks dereine should delette some of his files "by accident" [15:08] ^^ [15:08] i'm just wondering if its possible [15:09] dereine: ouch [15:10] ok i will force him to use bzr :) [15:10] dereine: it is possible, but very awkard. [15:10] dereine: you'll need ssh access to the ftp machine, and bzr installed on it [15:12] mh this will be hard :) i asced the support yesterday evening they don't support ssh even for really business accounts which costs too much :) === dereine is now known as dereine[OFF] === dereine[OFF] is now known as dereine === bob2 is now known as Guest62442 === jfroy_ is now known as jfroy [20:19] Hello, I set up bazaar on my mac for a project a couple of weeks ago. I'm still very much a bzr noob. When I initially set it up, I checked in the whole project. Since then, I've done a lot of work - added a lot of directories and files, edited a lot of files. [20:19] What's a recommended way for me to update my repository? [20:20] It's just a local repository. .bzr directory inside my top level directory. nothing more. [20:21] Should I just do "bzr add" then "bzr commit" [20:29] OK, well, that's what I did. And it seemed to work. bzr seems very simple and straightforward. I'm glad I met it. :) [21:49] when I bzr branch a copy of lp:maria, bzr 1.13 uses up >600Mb of memory during the process, which is a bit of a bummer. Any way to reduce this? [21:50] [this is a branch of the MySQL server tree. which has a lot of history, around 500Mb or so] [22:38] knielsen: you can probably pull in stages, bzr branch -r1000 lp:maria, cd maria; bzr pull -r 2000 etc === bob2 is now known as Guest81090