[01:06] <Peng> bzr: ERROR: xmlrpclib.Fault: <Fault -1: "Unexpected Zope exception: Unauthorized: (<zope.app.publisher.xmlrpc.metaconfigure.BranchSetAPI object at 0x2aaabec5bd50>, '__call__', 'launchpad.AnyPerson')">
[01:06] <Peng> Wheee!
[01:06] <Peng> Apparently the launchpad plugin really wants my password.
[01:22] <lifeless> thats launchpad wanting it
[04:43] <johnf> If I have a smart server whats the best way to have it send emails when commits are made? I was using bzr_hookless_email but it seems to have some issues with 1.0. Want to know if its the best way to do this sort of thing before fxiixing it
[04:46] <dato> johnf: what issues?
[04:46] <dato> johnf: it should work, but maybe you'd need to restart it if you upgrade your branches
[04:46] <johnf> I have it running out of cron
[04:47] <johnf> let me paste the bactrace somewhere
[04:47] <dato> yes, please
[04:50] <dato> !pastebin
[04:50] <ubotu> pastebin is a service to post large texts so you don't flood the channel. The Ubuntu pastebin is at http://paste.ubuntu-nl.org (make sure you give us the URL for your paste - see also the #ubuntu channel topic)
[04:53] <johnf> dato: http://paste.ubuntu-nl.org/50358/
[04:53] <johnf> Havn't looked into the problem myself yet. Was wondering firstly if this is the best tool to use
[04:54] <dato> johnf: well, I wrote it, so I may be a bit biased. I find it convinient when it's difficult or cumbersome to get every committer to install and configure bzr-email themselves.
[04:55] <johnf> hehe
[04:55] <johnf> yeah I have the same problem.
[04:55] <dato> johnf: hm, let me try reproducing your problem.
[05:02] <dato> johnf: ok, got it, thanks for reporting!
[05:03] <dato> johnf: you can pull a fix from http://chistera.yi.org/~adeodato/code/bzr/bzr-hookless-email/, the launchpad branch is a mirror and will take a bit before it gets the fix.
[05:03] <johnf> dato: no problems. Was just having a look myself. Looks like you might need to lock the repository?
[05:04] <dato> yeah, that's what I did. looks like bzr-email will need updating as well. (that part of bzr-hookless-email came from bzr-email)
[05:16] <lifeless> johnf: patch bzr to send a 'commit has occured' at the end of the transaction to the smart server
[05:16] <lifeless> johnf: its on the TODO list
[05:17] <lifeless> johnf: or perhaps if you really mean 'when the branch tip advances' patch bzr-email to hook into BranchHooks.set_revision_info and do emails then
[05:17] <lifeless> both will run on the server and don't need a cron job etc
[05:17] <lifeless> johnf: its really waiting for someone that wants it to do it :)
[05:17] <lifeless> also, jan 1st. HOLIDAY.
[05:18]  * dato is just waiting for friends to arrive to have breakfast :)
[05:24] <johnf> dato: thanks
[05:25] <johnf> lifeless: I shall add it to my long list of things to do :)
[05:52] <lifeless> johnf: we're back tomorrow; can we arrange to meet up say friday?
[05:52] <johnf> lifeless: sure or earlier. I can drop past anyday with everything
[05:56] <lifeless> cool - well thursday sometime would be great. Anytime, just ring a bit before so I'm not out at the shops.
[07:31] <indu> hi all
[07:31] <indu> happy new year
[07:31] <indu> i need help in bazaar implementation
[07:32] <indu> i created a repo using bzr init-repo, and then created branches under that, using bzr init
[07:32] <indu> i am finding that the sub branches are independent and are no where related to the main branch which was created using bzr init-repo
[07:33] <indu> am i going in right path ?
[07:34] <Peng> indu: bzr init-repo doesn't create branches.
[07:34] <Peng> indu: It's simply an optimization, so the data from all branches created under it will be stored in one place instead of duplicated in each branch.
[07:35] <indu> Peng, could u please explain this clearly
[07:35] <indu> how does duplication be saved ?
[07:36] <indu> Peng,  sorry,how does the duplication is not occured  ?
[07:37] <Peng> indu: If you have two branches of the same project, 99% of the history is identical, but all of it would be stored twice. Using a shared repo, it's only stored once.
[07:38] <indu>  Peng here u mean branch as the one created  using bzr init ?
[07:39] <Peng> indu: Sure.
[07:39] <Peng> indu: If you just created the branch, there's no history, so it's not so relevant.
[07:39] <Peng> I'm sorry, but I'm gonna go.
[07:40] <Peng> But yes, the branches are independent.
[07:40]  * Peng wanders off.
[07:40] <Peng> There isn't really a way around that at the moment.
[08:09] <deflated> Hello. What do you think of taking Bazaar for backups?
[08:12] <Peng> deflated: I think it's very useful to use a VCS on some of your files, but to back up your email or whatever is a bad idea.
[08:20] <deflated> Thank you for your response. It's meant for backuping my source codes. I plan to store them on a remote server. I only change approx. 10 files a day so that it's not that efficient to transfer *all files* via FTP to the server. What do you think of using Bazaar in this case or is there even a better way? If the application is stable I'll release it as open-source so I'll need to use a VCS but it's planned for the end of this year. For that time I ne
[08:34] <deflated> So Peng, what do you think? Are there any better alternatives? Here's an interesting article of someone who uses Bazaar as for backuping his data: http://www.singularity.be/node/6357
[08:36] <deflated> But he uses rsync for remote backups. The remote server connects to his computer and gets all data.
[08:40] <michaelkohwj> i have a web host that i can SSH into, but it doesn't have bzr installed. can I still push and pull from it over ssh?
[08:40] <lifeless> deflated: bzr for that is fine :)
[08:41] <lifeless> michaelkohwj: yes if they have python 2.5 you can dump a copy of bzr in your home dir and set BZR_REMOTE_PATH
[08:41] <lifeless> later all
[08:42] <michaelkohwj> hm i see
[08:42] <michaelkohwj> lifeless: thanks
[08:42] <deflated> Ok, then I'll give it a try. Thank you very much!
[08:46] <michaelkohwj> i can FTP over, but i can't seem to get it right:
[08:47] <michaelkohwj> bzr push --create-prefix ftp://user:pass@url.com/~/dir
[08:48] <michaelkohwj> oh, it works, but now: "but does not have a valid .bzr directory"
[08:48] <michaelkohwj> ... --use-existing-dir
[09:17] <michaelkohwj> hi, how should I push to a remote server over ssh, that doesn't have bzr installed?
[09:17] <michaelkohwj> i can't install bzr on it either
[09:21] <michaelkohwj> ok, i got it working with ftp, but the source files aren't transferred
[09:26] <michaelkohwj> bzr push --create-prefix ftp://user:pass@url.com/dir --use-existing-directory -v just transfers .bzr
[09:26] <michaelkohwj> what shoul i do to push to a remote dir?
[09:26] <michaelkohwj> thanks!
[10:02] <thumper> michaelkohwj: pushing doesn't create the working tree
[10:02] <thumper> michaelkohwj: just the branch and repo if needed
[10:02]  * thumper recalls someone working on pushing a working tree of files
[10:02]  * thumper walks away again
[10:09] <Peng> michaelkohwj: Use SFTP if you can.
[10:09] <Peng> MicahElliott: (It's an FTP-like protocol tunnelled over SSH.)
[10:10] <Peng> thumper: push-and-update plugin.
[10:35] <indu> hi, could someone give me a link for downloading loggerhead
[10:35] <indu> I am not getting any
[10:43] <michaelkohwj> Peng: but if I pull from it, will the effect be the same?
[10:43] <michaelkohwj> Peng: let me explain more clearly:
[10:43] <michaelkohwj> Peng: X makes changes to the project and pushes it to the server
[10:43] <michaelkohwj> Y then clones from the server, works on it, and then pushes it up.
[10:44] <michaelkohwj> (i'm not sure that i'm using the correct terms)
[10:44] <michaelkohwj> then when X pulls from the server and merges, will X get Y's changes?
[10:47] <Peng> michaelkohwj: Will doing what be the same?
[10:49] <michaelkohwj> Peng: um, such as pulling and pushing from and to a thumbdrive?
[10:50] <michaelkohwj> (sorry, i'm new to SCM)
[10:50] <Peng> michaelkohwj: It doesn't matter what you push and pull to.
[10:52] <michaelkohwj> Peng: oh i see- the -diffs- are stored in the place i push to
[10:52] <michaelkohwj> is that right?
[10:59] <Peng> michaelkohwj: Something like that, yeah.
[11:02] <dlee> Michael:  I think you're asking if A --push--> B and B <--pull-- A have the same effect.  I believe they do.  You might also be asking if changes pushed to server by one person will arrive at another person's local copy when that person then pulls from the server.  Again, yes, this should work.  Caveat:  I've been using Bazaar for about a week myself, but I've been using VC systems of various sorts for years.
[11:03] <Peng> (If one person pushing and another pulling didn't work, that wouldn't be much of a VCS...)
[11:03] <dlee> :)
[11:04] <dlee> I thought the question might be more about how push and pull were related, but at 5 AM after a lot of local activity, I don't trust my read :)
[11:05] <michaelkohwj> actually, i was just confused when pushing didn't place the files in the remote server, but instead placed .bzr/ in it only
[11:06] <michaelkohwj> so .bzr contains the diffs, and thus the actual working files are unnecessary?
[11:06] <Peng> michaelkohwj: Yeah, .bzr contains all of the data. It would usually be a waste of bandwidth to create a working tree (and there would be other issues, like what if there are merge conflicts?).
[11:06]  * michaelkohwj has been enlightened
[11:06] <michaelkohwj> thanks Peng and dlee
[11:07] <dlee> Ah ok... .bzr contains everything necessary to reconstruct the local files.  It's just that it's not so easy for your "push"ing system to make the "push" receiver update the local files on that end.  You can't push your local files to remote because they may contain changes.  push and pull work with what's in .bzr, not what's in all the other files in your working tree.  Does that help?
[11:09] <GaryvdM> If you push over a local transport, it will also push the working tree. But pushing over a remote transport will not push the working tree.
[11:09] <GaryvdM> Working trees can easily be recreated from the repository (contained in the .bzr dir).
[11:10] <dlee> Doesn't a local-transport push still push the *committed* rev, not the actual local (pushing-side) working-tree files?
[11:10] <michaelkohwj> dlee: yes =)
[11:10] <GaryvdM> yes
[11:10] <dlee> A "no" to that would have been a shock...
[11:11] <michaelkohwj> thanks for the help, i understand now =)
[11:11] <dlee> I really hope I can talk my office into Bazaar now, even though they jumped at Svn very recently before I knew what they were doing :P
[11:16] <michaelkohwj> dlee: perhaps bzr-svn is a short-term solution? but changing everyone to bzr is probably better, i guess..
[11:16] <dlee> right on both counts imo
[11:20]  * dlee wanders off perchance to sleep
[13:20] <kohwj> hi, when I try to push over sftp, i get a Permission deined:
[13:21] <kohwj> error (errno 13). hold on, let me google it first
[13:22] <kohwj> hm not much luck. ftp works, though
[13:45] <kohwj> never mind, i just solved it with rsync and bzr update
[13:45] <kohwj> bzr is really user-friendly, imo
[13:46] <kohwj> thank you bzr
[15:20] <jelmer> re
[15:20] <Peng> fwd!
[15:21] <TFKyle> bcc!
[19:34] <_markus__> hi
[19:34] <lifeless> yo
[19:37] <_markus__> I'm interested in bzr to use it. I'm currently using svn. I've gone through the user guide but some questions are not answered. I'm not sure if bazaar is the right tool here. I've a central svn server with the projects (can't change this) and would like to check a version out and work offline on it, do offline commits and later merge this changes back to the svn repository. now, svn doesn't...
[19:37] <_markus__> ...support something like local commits. Bazaar seems to do, but I'm unsure how I can work with both together ...
[19:41] <Peng> _markus__: You want to use both bzr and svn?
[19:44] <_markus__> Peng: well .. I don't know, honestly :) I mean, I'm forced to use svn on the central side. I thought bzr can fill me the gap of local commits and then push the changes to the live system. Maybe I understand this wrong.
[19:46] <Peng> _markus__: There's an excellent bzr-svn plugin that lets bzr interact with svn repos. (Minor pain to install though.)
[19:46] <Peng> _markus__: Of course, ideally your office would switch to bzr. ;)
[19:47] <_markus__> Peng: I've this plugin install and read about it .. but maybe I'm dumb but I've no plan in front how things work together :-/
[19:48] <_markus__> (I'm using debian testing, everything is packaged up)
[19:48] <_markus__> Peng: it's not only about the office, but also various open source projects, so it's not my choice, really
[19:48] <Peng> _markus__: I think you can just "bzr branch" from the svn repo, and then "bzr push" your changes back to it.
[19:48] <_markus__> hmm
[19:48] <_markus__> so easy? let me try ..
[19:48] <Peng> _markus__: Ideally, those FOSS projects would switch to bzr too . . . :)
[19:49] <_markus__> Peng: hmm .. is there a nice gui client such as TortoiseSVN? TortoiseSVN is really very useful
[19:50] <Peng> _markus__: There are a few GUI clients. QBzr (Qt), bzr-gtk (GTK+), and TortoiseBzr or something.
[19:54] <_markus__> Peng: hmm .. TortoiseBzr doesn't seem to be nearly somewhere were TortoiseSvn is ... anyway . bzr branch is running. Unexpectedly easy. This command seems to download the complete history information, am I right?
[19:55] <jelmer> _markus__: yes. It will be slow the first time but will cache some data for later use
[20:02] <_markus__> Hm .. seems to hang .. :/
[20:07] <_markus__> Jup, it hangs. Is there any other way without using bzr branch? I used strace, it now just sits there and hangs here "6043  poll("
[20:08] <Odd_Bloke> _markus__: Does it create anything in the expected directory?
[20:12] <_markus__> Odd_Bloke: no, it only created some files in ~/.bazaar/svn-cache/
[20:13] <_markus__> On the console I rea "fetching svn revision info  7/37". The complete svn repos is no more than 50MB
[20:23] <Odd_Bloke> jelmer: ^
[20:28] <LeoNerd> You're right. bzr+ssh is a lot faster than sftp over long-latency links
[20:28] <LeoNerd> Thanks to whoever suggested that earlier
[20:43] <lifeless> LeoNerd: you using a pack repo or knits still ?
[20:43] <LeoNerd> How do I tell?
[20:49] <dlee> LeoNerd: What's the top line from bzr info?
[20:52] <LeoNerd> Standalone tree (format: dirstate-tags)
[20:52] <dato> that's knits
[20:53] <LeoNerd> That's in my checked out branch, not the (rempte) repo on the server
[20:56] <Peng> Run "bzr info http://path.to/remote/repo".
[20:56] <Peng> If you're using a checkout, bzr info might've listed the server's format.
[20:58] <LeoNerd> OK.. Server says Repository branch (format: dirstate-tags)
[21:01] <Peng> If you don't need compatibility with bzr < 0.92, upgrading to pack-0.92 should really boost performance over dumb protocols (http, sftp).
[21:02] <LeoNerd> Right.. Well, it's only me that uses this really, so.. yeah.. I guess I can do that
[21:20] <jelmer> re
[21:20] <Peng> glob!
[21:21] <jelmer> _markus__: Sorry, not sure what could cause that. We're relying on the Subversion libraries there
[21:21] <jelmer> Peng: :-)
[21:21] <jelmer> _markus__: Does it always hang or does retrying work ok?
[21:28] <mwhudson> lifeless: can you email me the loggerhead.conf file you are using for https://bugs.edge.launchpad.net/loggerhead/+bug/179347 ?
[21:28] <ubotu> Launchpad bug 179347 in loggerhead "NotBranchError on startup" [Undecided,New]
[21:29] <lifeless> mwhudson: done
[21:36] <mwhudson> lifeless: how totally strange
[21:36] <lifeless> mwhudson: I thought so.
[21:36]  * mwhudson will look more tomorrow, too tired tonight
[21:36] <lifeless> thanks
[21:37] <lifeless> I grabbed your lp branch from bazaar.lp.net
[21:38] <mwhudson> that's the right place to go
[21:49] <igc> morning
[21:49]  * fullermd waves at igc.
[21:50] <igc> hi fullermd
[22:28] <lifeless> mwhudson: consider relative urls for codebrowse, kthxsavebybandwidth
[22:29] <lifeless> e.g. consider http://codebrowse.launchpad.net/~lifeless/squid/squid3-trunk/files/cvs-1%3Ahno-20071220081046-d59dmpbuvvyioekf?file_id=src-19960222132353-t43j7l5jst5z-28
[22:57] <poolie> good morning
[22:59] <spiv> Good morning.