[04:50] <d3jake> I've encountered this error while trying to commit my project to a bzr repository hosted by sourceforge.net. http://pastebin.com/ahW8U8Ve This leads me to think that my bzr version is too new to work properly, is this a correct analysis?
[04:51] <AfC> d3jake: no, it means that for whatever reason the data on the server is in a *really* old format. If you can upgrade the server branches you'll fix the problem
[04:53] <d3jake> AfC, Unless a change was made to bzr's format breaking backwards compatibility within the last three-four months, I don't see why this error would be happening. I was able to make commits without issues using a not-too-distantly-old version of bzr to the same repository.
[04:54] <AfC> It's not the Bazaar version. I can't tell you what the error is, but it is doing a conversion to something (WAY) older than 2a, and so you're exposing yourself to trouble. Upgrade the server branch & repository if it's yours. And if it's not yours, ask them to. There's no reason, none, to not be using 2a
[04:55] <d3jake> Ohh, okay. So the problem is likely on sourceforge.net's end as they are the one's hosting the repository?
[04:55] <AfC> "hosting" is just disk space
[04:55] <AfC> whose project is it?
[04:55] <AfC> yours or !yours?
[04:55] <AfC> and,
[04:55] <AfC> do you have SSH access to the server?
[04:55] <d3jake> The project is mine. And no.
[04:55] <AfC> hm.
[04:56] <AfC> Well, others can correct me if I'm wrong, but all you need to do is `bzr upgrade` the (branch and the) repository
[04:57] <AfC> I can't recall if you can do that over bzr+ssh. You might be able to, actually.
[04:57] <AfC> failing that,
[04:57] <AfC> you can bzr upgrade locally to 2a
[04:57] <AfC> (actually, it sounds like you have already)
[04:57] <AfC> and just overwrite completely what's up there.
[04:57] <d3jake> So, open a console, go into the root directory of my project, and type "bzr upgrade", and then some sort of command to force the changes onto the server somehow?
[04:58] <AfC> I'll let others take it from here; there are a few web pages out there on the topic I'm sure
[04:58] <AfC> but short answer, yes
[04:58] <d3jake> Okay, cool
[04:58] <d3jake> Thank you for your help :)
[04:59] <AfC> bzr info
[04:59] <AfC> will guide you on your way
[04:59] <AfC> and read the help for upgrade
[05:00] <d3jake> Does it mean anything if bzr info tells me that my project is a standalone tree, and in format 2a? Does this mean that I need to upgrade the repo?
[05:09] <AfC> d3jake: your local tree?
[05:09] <AfC> If it says that, then you're fine locally (and is the case as evidenced by the strange error you got)
[05:10] <AfC> d3jake: if it says that for the remote tree, you are finished
[05:10] <d3jake> AfC: Yes, it is my local tree.
[05:10] <d3jake> I'm wading through trying to see if I can et the remote tree fixed, but I'm nearing the time when I should shove off for the evening...
[05:11] <AfC> so just do
[05:11] <AfC> $ bzr upgrade bzr+ssh://path/to/remote/tree
[05:11] <AfC> that'll work, or not
[05:11] <AfC> failing that, just remove it and upload a new one
[05:11] <d3jake> Remove it? As in delete the remote tree, and then reupload?
[05:12] <d3jake> Wouldn't I loose all of my revision history?
[05:16] <d3jake> Ah well, I need to bail. Thank you for your help.
[05:59] <AfC> If you've got conflicts, can you use `meld` to sort it out?
[05:59] <AfC> (ie, "yes", but what's a good idiom?)
[06:12] <spiv> AfC: not sure, but I'd be interested to hear if you find one!
[06:12] <AfC> spiv: ie, `bzr diff --using=meld` is vaguely a step in the right direction;
[06:12] <AfC> it seems like
[06:12] <AfC> $ meld path/to/file.OTHER path/to/file.THIS
[06:13] <AfC> is about right, but still, it's a bit fidgety
[06:59] <TimMiao> hello everyone, could anyone please tell me where can I set the proxy for bzr? to add a new line "http_proxy = http://some.proxy.address:port/" in bazaar.conf file doesn't work for me. Thanks!
[08:18] <vila> hi all !
[08:20] <fullermd> Eh?  Again?  Didn't you just say that yesterday?
[08:54] <spiv> fullermd: it's like the days just keep coming and ocming!
[08:54] <spiv> TimMiao: set it as an environment variable
[09:05] <fullermd> spiv: I know!  I'm _sure_ I never authorized that!
[09:48] <vila> . o O (Did I just saw this groundhog *again* ?)
[09:50] <Peng> Groundhog? Does that mean I get six more weeks of insomnia? :D
[09:54] <spiv> Peng: you could always transform that into six weeks of late-night Bill Murray movie watching!
[09:56] <vila> I love this movie but... six weeks... wow :)
[09:57] <quicksilver> my daughter was born on groundhog day.
[09:57] <fullermd> Did she see her shadow on the way out?
[12:37] <svaksha> hi, while checking out code I get  "bzr: ERROR: extra argument to command checkout". How do I solve it? TIA.
[12:40] <beuno> svaksha, what command are you running
[12:41] <svaksha> beuno: http://www.gnuenterprise.org/developers/bzr.php <-- trying to get the source
[12:41] <svaksha> bzr co bzr branch http://bzr.savannah.gnu.org/srv/bzr/gnue
[12:41] <beuno> right
[12:41] <beuno> so you don't need the first part, "bzr co"
[12:41] <beuno> just:  bzr branch http://bzr.savannah.gnu.org/srv/bzr/gnue
[12:42] <svaksha> beuno: still throws this error, bzr: ERROR: Not a branch: "http://bzr.savannah.gnu.org/srv/bzr/gnue/".
[12:43]  * svaksha tried it with sudo too
[12:44] <beuno> svaksha, FWIW, that page has it wrong
[12:44] <beuno> not sure if it can be reported
[12:44] <beuno> http://bzr.savannah.gnu.org/srv/bzr/gnue 404s
[12:44] <svaksha> beuno: yes, but i dont know how else to get the source
[12:44] <beuno> and bzr co bzr branch is from
[12:45] <beuno> svaksha, try: bzr branch http://bzr.savannah.gnu.org/r/gnue/
[12:45]  * svaksha has been trying different brz permutations with no luck
[12:45] <svaksha> s/brz/bzr
[12:45] <beuno> no, the url in that page is wrong
[12:46] <svaksha> beuno: that worked, thanks :)
[12:46] <beuno> cool, np
[12:51] <svaksha> beuno: fwiw, I reported the error in their irc channel. So thanks!
[13:00] <vila> jelmer: ping
[13:48] <jelmer> vila: hi
[13:48] <vila> jelmer: hi there
[13:48] <vila> jelmer: saw my pm ?
[13:49] <jelmer> vila: no, thanks for the reminder
[13:52] <frathgeber> ?nested
[14:05] <vila> jelmer: did you get failure reports from pqm ?
[14:10] <vila> jelmer: if yes, can you try feed-pqm my submission about 321320-isolate-doc-tests ?
[14:14] <jelmer> vila: yep, I did get (strange) failure emails
[14:15] <vila> jelmer: don't resubmit mine ! Just got mails
[14:18] <vila> jelmer: what kind of strangeness ?
[14:29] <jelmer> vila: I don't remember actually, I'll retry.
[14:33] <vila> jelmer: already failed right  ?
[14:34] <vila> jelmer: meh, misread the page.. again...
[14:41] <vila> jelmer: or not ?
[14:41] <vila> jelmer: you sent 2 requests ?
[14:58] <jelmer> yeah
[14:59] <vila> jelmer: got mail for the first one ?
[14:59] <jelmer> Yeah - PQM test failure.
[14:59] <jelmer> Ran 0 tests in 0.000s
[15:00] <jelmer> And empty output files attached
[15:00] <jelmer> merge http://bazaar.launchpad.net/~jelmer/bzr/daily-ppa http://bazaar.launchpad.net/~bzr-pqm/bzr/bzr.dev
[15:01] <vila> ha, I feel less alone :-/
[15:06] <mgz> ...committing is broken again?
[15:06] <mgz> that's not good, last thing to land was by me,,,
[15:08] <vila> mgz: don't be so paranoid :)
[15:08] <vila> . o O (Or how would we get you...)
[15:08] <vila> mgz: I looked at it, but saw nothing wrong
[15:09] <vila> mgz: my pqmlike setup is fine with my patch too :-(
[15:09] <mgz> maybe it's not my fault and the pqm box just ate too much over xmas and hasn't recovered
[15:12] <vila> well, nothing has landed since... unless you count failed submissions, pqm is on diet :)
[15:35] <vila> mgz, jelmer: sounds like the pqm *host* have been upgraded to lucid while we still run in a hardy chroot and there are.... unknown issues :(
[15:36] <fullermd> See what happens when you use adjectives instead of version numbers?   :p
[15:38] <jelmer> vila: ah
[15:42] <vila> jelmer: and stop queuing (s) my submissions, email (e) them instead when using feed-pqm:D
[19:00] <Phoenixz> A co-worker committed a change, pushed it, and I pulled it.. the commit is rubbish and should NOT be committed... How can I remove or revert this entire commit, and redistribute it?
[19:01] <knighthawk> I'm confused I just tried to commit to my central repo and it gave me a message saying that "No WorkingTree exists" I thought I didn't need a working tree for a central repo
[19:02] <jelmer> Phoenixz: you can commit the revert of that commit, e.g. "bzr revert -r -2 && bzr ci -m 'revert broken commit'"
[19:02] <knighthawk> Phoenixz, can you just revert your copy then commit it?
[19:02] <jelmer> knighthawk: you shouldn't need a workingtree indeed - how are you committing to the central repo?
[19:03] <knighthawk> jelmer, 'bzr commit -m "update repo" sftp://path to repo'
[19:04] <jelmer> knighthawk: That will try to do a commit in that location
[19:04] <jelmer> knighthawk: You probably want to commit in your local branch.
[19:04] <jelmer> if your local branch is bound to the central branch the changes will automatically be pushed to the central branch. alternatively they can be pushed manually
[19:04] <jelmer> so you probably want:
[19:05] <jelmer> "bzr bind sftp://path to repo; bzr commit -m 'update repo'"
[19:05] <jelmer> bzr ci -m 'update repo' && bzr push sftp://path-to-repo
[19:05] <knighthawk> if I run bzr info It has the central branch set up for push branch and submit branch
[19:06] <jelmer> knighthawk: in that case you wouldn't need an argument for "bzr push", but you'd still need to run it if the branch is not bound
[19:06] <knighthawk> but I would like to commit on my local as well. Right now I'm more concerned about getting my team up to speed on BZR so we can all use it.
[19:07] <jelmer> knighthawk: if you run "bzr ci -m 'foo'" without any other arguments it will commit to your local branch
[19:07] <jelmer> knighthawk: if you've bound the branch to another branch it will automatically push the new commit there as well (hence the "bzr bind" command I mentioned earlier)
[19:07] <jelmer> knighthawk: you don't have a bound branch from what I can tell, so you'll need to bind it manually.
[19:16] <knighthawk> jelmer thanks. I'll try the bound then see if that works.
[19:19] <knighthawk> bzr: ERROR: sftp://repos.grdev.com/opt/bzr-repos/grws/.bzr/ is not a local path.
[19:19] <knighthawk> when I try to commit after the bind.
[19:22] <knighthawk> bzr: ERROR: Permission denied: "/opt/bzr-repos.grws": [Errno 13] Permission denied
[19:22] <knighthawk> when I try to push
[19:24] <knighthawk> I want to remove a file from my working tree but don't want to remove it from my file system. I *should* have set up bzr ignore before it got picked up by my working tree but I didn't.
[19:25] <knighthawk> now how do I remove it without using bzr remove. or is there a way to tell bzr remove to leave the file on my filesystem?
[19:25] <knighthawk> just found the --keep option. Thanks
[19:31] <jelmer> knighthawk: when you commit you shouldn't specify the remote repository
[19:33] <achiang> hello, is there a way to completely revert a commit, the same way you would say, "git reset --hard HEAD^" ? the git command erases the -1 commit from history, forever as if it never existed
[19:33] <achiang> bzr merge . --revision -1..-2 kinda gets me close
[19:33] <achiang> but i still see the -1 commit in my log
[19:35] <james_w> achiang, bzr uncommit
[19:35] <achiang> ah!
[19:35]  * achiang goes to read the man page
[19:36] <achiang> james_w: perfect, thank you
[19:36] <knighthawk> jelmer, thanks I think that's what I was doing wrong.
[19:37] <james_w> achiang, you'll want a bzr revert as well to match reset --hard
[19:37] <achiang> james_w: yeah, i just discovered that
[20:26] <mgz> bug 700000!
[20:26] <mgz> ...is a dupe.
[20:46] <Peng> Well, better luck with 800000.
[21:10] <jelmer> hehe
[21:10] <jelmer> mgz: I managed to get 600000 IIRC :-)
[21:10] <jelmer> bug 600000
[21:11] <jelmer> yup :)
[21:11] <mgz> go jelmer :)
[21:11] <jelmer> my greatest accomplishment of 2009 :-P
[21:11] <mgz> had a look at apache-bzr source.
[21:12] <mgz> nice idea.
[21:12] <jelmer> mgz: the code does indeed need some more work :-)
[21:13] <mgz> ...that... wasn't the implication >_<
[21:13] <jelmer> :)
[21:14] <jelmer> There was a reason the announcement said it was a PoC :-)
[21:17] <mgz> avoids the wsgi mess and confusion neatly at any rate.
[22:42] <knighthawk> I'm trying to remember the syntax to include the user name and password in the URL.
[22:42] <knighthawk> something like sftp://username:passwd@example.com
[22:42] <knighthawk> only that doesn't seem to be working, and I can't think of anything else.
[22:43] <bob2> don't do that for sftp
[22:43] <bob2> sftp://user@example.com/ does work, though
[22:43] <bob2> er "don't put passwords in there for sftp"
[22:44] <knighthawk> I've got members of my team using Windows. and I have no idea how to tell eclipse what password to use without putting it in the path.
[22:44] <bob2> nah, use ssh keys
[22:44] <knighthawk> perhaps in bzr I can tell it the passwd then eclipse wouldn't have to know but again. Don't know how to do that in windows.
[22:45] <knighthawk> hmmm that sounds like a good solution actually.
[23:23] <Phoenixz> A coworker of mine added, comitted, and pushed files into a repo that should not be there.. I pulled those files, and now I'm stuck with them. I want these files not only out of the WT, I want them out of the repo, I want that commit gone (the files are copyrighted, the repo is open source).. How can I undo / revert / remove this commit?
[23:23] <maxb> bzr uncommit
[23:24] <knighthawk> okay I'm confused again. I just did a commit to a central repos of revision 7. However whenever other members of my team try to update they just get revision 6 what am I doing wrong.
[23:25] <maxb> thats not much info to go on but the logical assumption would be that you've only committed locally
[23:33] <bob2> Phoenixz: exporting and importing are your only options afaik
[23:35] <Phoenixz> bob2: ok.. On a non-copyright issue then, is it possible to "uncommit" or unmerge those things away?
[23:35] <Phoenixz> bob2: like, remove commit 34... 34 will still be there in the repo, but at least won't affect files.. ?
[23:35] <bob2> yes, 'bzr rm' :)
[23:36] <Phoenixz> bob2: bzr rm is on file level.. there isn't anything like that on commit level?
[23:36] <bob2> kinda
[23:43] <Phoenixz> bob2: kinda... ? as in?
[23:44] <bob2> as in I'm not sure what the point is - bzr rm leaves stuff in the rev history the same way as bzr revert would
[23:47] <maxb> Phoenixz: the important question is: is the problem commit the last commit in the branch?