[00:09] vila: I think I have tried dvc before. [00:10] vila: I dunno, I guess I just find the productivity gains not worth the effort of learning a new interface. bzr in shell mode is already 95% of what I need, and it always Just Works. [00:12] dvc seems quite nice, but what you really need is for emacs to learn what a 'project' is, i think [00:12] or at least a branch [00:12] i hate having to C-x v all my buffers when i change branches [00:14] Always works fine for me. First, you fire up vim... [00:43] poolie: now a good time for me to switch on the "clean up the chroot" for bzr PQM? [00:44] yes please [00:44] cool [00:50] poolie: that's done - would be good if we were able to do a test commit [00:51] mthaddon: sure [01:23] poolie: any chance of that test commit so we can confirm the new PQM setup works? [01:24] mthaddon: yes, will send it soon... [01:24] cool, thx [01:35] lifeless: I don't think I ever got around to factoring out the changes from my "hacks" branch, http://bzr.arbash-meinel.com/branches/brisbane/hack [01:35] lifeless: but the specific changes are pretty small [01:37] jelmer: ping [01:37] lifeless: hello [01:37] your InterBzrDirTransport [01:37] is that in a patch for bzr.dev? [01:37] BranchBzrDirInter? [01:38] oh right [01:38] that's in a branch here, but not submitted yet (since it lacks tests) [01:38] uhh don't think that matches our needs - thanks [01:44] poolie: spiv and I have just had a use case pass, which is "'bzr push bzr+ssh://foo' uses a smart verb to send the revision data in one stream" [01:45] poolie: just don't ask what it does under the hood :P [01:50] igc, did you say you'll help OOo with their error? i'd like that. [01:53] poolie: jelmer has replied to Heiner explaining that it's a bzr-svn bug and he's hoping to work on it this weekend [01:53] oh, great! [01:54] thanks [01:54] poolie: also, I'm looking into a batch of fast-import bugs/patches today so that might help as well [01:55] poolie: did you see my command hooks mail? [01:57] not yet [02:15] lifeless, btw, I plan to submit a patch for InterBranch as well (for update_revisions/pull/push) [02:16] lifeless, not really fit for your use case either, probably, but might change some of the same areas of code [02:16] lifeless: also, would bug 272444 be appropriate for brisbane-core? [02:16] Launchpad bug 272444 in bzr "Support symlinks to non-ascii file names" [Medium,Triaged] https://launchpad.net/bugs/272444 [02:34] * igc lunch [02:37] jelmer: patchesyadayadayada [02:37] jelmer: I'm surprised we don't today [02:40] lifeless, well, that one I don't actually care about :-) Was just wondering [02:40] I might sent in one for the revision XML escaping [02:40] having workarounds for that in bzr-git and bzr-svn is annoying [02:40] and could potentially lead to corrupt revisions when roundtripping [02:41] since there's no way to tell escaped characters apart [02:41] I certainly think the core should be fixed [02:41] bbc is purely focused on scaling-and-less-work-to-diff-two-inventories [02:42] verterok: ping [02:43] if its not a scaling issue (size or performance), or a work-to-generate-diff issue, then its not going to get my attention at the moment, nor would I expect it to get John/Ian/Andrews (dataloss bugs and regressions are obvious exceptions to this rule of thumb) [02:43] patches to the bbc development branch, to fix things that interest you would be cool [02:44] jelmer: once the bbc focus is reached, obviously we can look for minor tweaks to do, but there is no reason to delay them if someone else is interested. [02:46] lifeless, I should rephrase that; I meant it would be nice to get a new revision serializer in if we're going to be introducing a new repository format and rewriting revisions when upgrading to it. [02:46] so not necessarily as part of bbc but rather of whatever format it will end up in [02:46] (or is there no difference?) [02:46] jelmer: there isn't a difference [02:46] k [02:46] lp:~bzr/bzr/brisbane-core [02:46] development4* formats in there === mthaddon_ is now known as mthaddon [02:57] mthaddon: if you're still here - i sent a trivial merge to pqm [02:57] poolie: yep, watching the logs [02:58] poolie: seems to be working nicely [03:05] spiv: http://paste.ubuntu.com/116673/ [03:48] poolie: and we now have an acceptance test for pushing to a stacked branch using a streaming format [03:49] poolie: passing [04:44] lifeless: bzr pqm seems to be hung on the sending the mail portion - should i just kill it, or is there any troubleshooting we can do? [04:52] mthaddon: strace [04:52] mthaddon: lsof [04:52] mthaddon: etc etc etc [04:52] (pqm doesn't do any magic here - if the 'mail' command is wedged, I'm not any more sure than you are whats causing it. [04:53] ok, nothing obvious there - thx [04:54] what syscall is the program in? [04:54] is the mail server itself wedged? [04:55] "read(0, " was all I got [04:55] mail server seems to be responding fine, though [04:57] read(0 is in mail ? [04:57] thats reading from stdin [04:57] what syscall is pqm in? [04:57] I have to go cathc a train [04:57] can you brain dump all the info you have to a bug please [04:58] get a python bt in pqm if possible [04:58] and then kill it [04:58] so people aren't blocked [04:58] was in "wait4(4294967295, " [04:58] have killed it - will keep an eye on it - have seem this every now and again before, just want to make sure it's not related to the config change I've just made [05:32] mthaddon: thats a deadlock for sure, pqm is waiting, mail is wating [05:32] mthaddon: so the question is, why is mail waiting [05:33] spiv: so, turns out the Branch tests were running with real branches [05:36] lifeless: even when they were supposed to be parameterised not to? [05:41] :P [05:41] remember the bug we found where we got a real object back? [05:41] thats why it expected a remote, because we fixed that bug [05:42] the reason it only saw a real was... drum roll, because that was the server side one raising the event [05:44] (RemoteBranch.set_revision_info doesn't invoke hooks) [06:03] lifeless: we've seen this before with pqm [06:03] congrats on your test passing btw [06:04] i suspect somebody relying on gc to close the pipe or something similar [06:07] poolie: its in wait4( [06:07] pqm is waiting for the response [06:07] and mail is in read(0 [06:07] so its reading from stdin [06:07] yes exactly [06:07] I can't remember what module pqm uses to send mail [06:07] pqm thinks mail should finish up but it hasn't actually closed the input pipe [06:07] i looked into this last time [06:07] erm, 'which mail' [06:07] so it might be bzr-mail [06:08] or it might be the pqm success mail [06:08] the backtrace should tell us [06:08] i think it runs it directly through subprocess [06:08] there are two mail events [06:08] the command arguments to mail made it clear which it was [06:08] ok [06:08] well, i'm assuming here that there are not two bugs with identical symptoms [06:08] well, I started well before 9 :) and its >5, so really calling it a day [06:10] sure [06:10] unfortunately we must have discussed it only on irc [06:12] ah bug 242262 [06:12] Launchpad bug 242262 in pqm "deadlock while sending mail" [Undecided,Confirmed] https://launchpad.net/bugs/242262 [06:12] complete with suggested patch [06:13] mthaddon: ^ [06:14] aha [06:14] complete with suggested patch all tidied up by launchpad [06:14] getting rid of all that messy indenting :) === jbalint_ is now known as jbalint [07:05] hi all [07:05] * fullermd waves at vila. [07:06] fullermd: hi ! [07:18] rats, Rejected: [07:18] PPA uploads must be signed by an 'ubuntero'. [07:18] A rare species of vole found only on Madagascar? [07:25] fullermd: should be that yes :) [07:26] poolie: I can't dput the same file anymore, which '1' should I change into a '2' in 1.12~rc1-1~bazaar1 ? [07:26] rc1-2 ? [07:26] hm i guess so [07:26] if it was rejected i'm surprised it cares [07:26] did you sign the thing? [07:27] yes [07:28] argh, wait, there is a warning: [07:28] gpg: WARNING: This key is not certified with a trusted signature! [07:28] gpg: There is no indication that the signature belongs to the owner. [07:28] just fater saying: [07:28] gpg: Signature made Wed 11 Feb 2009 07:43:03 AM CET using DSA key ID DEF6218F [07:28] gpg: Good signature from "Vincent Ladeuil " [07:28] gpg: aka "Vincent Ladeuil " [07:28] though... [07:28] wow [07:28] launchpad has translations into 272 human languages [07:28] or at least it thinks it does :) [07:33] poolie: ok, I fixed that warning, I'm ready to try to dput again unless you have a better idea to check first [07:35] um [07:36] so i see you are an ubuntero now [07:36] let's go [07:40] poolie: yeah, I did that and fix a glitch in my gpg config too [07:40] dput done, no warning [07:41] ooh [07:41] Package includes an .orig.tar.gz file although the debian revision suggests [07:41] that it might not be required. Multiple uploads of the .orig.tar.gz may be [07:41] rejected by the upload queue management software. [07:41] is harmless I presume ? [07:41] you should be able to see in the web ui if it's building [07:41] yes, that's normal [07:48] poolie: mail received, accepted, hurrah ? [07:49] https://edge.launchpad.net/~bzr-beta-ppa/+archive/ppa/+builds?build_text=&build_state=all [07:53] Built successfully [07:53] Purging chroot-autobuild/build/buildd/bzr-1.12~rc1 [07:53] is the relevant part ? [07:53] so Hassium, the element after which the machine that built your package is named [07:54] was not called that when i studied chemistry [07:54] yet another joke flying high above my head ? :-) [07:54] no [07:54] it changed name in only 1997 [07:54] http://en.wikipedia.org/wiki/Hassium | http://en.wikipedia.org/wiki/Element_naming_controversy [07:55] we have a machine for each element i think [07:56] * fullermd has sure never heard of 'hassium'... [07:57] I always feel bad when scientists lose their energy on such disputes :-/ (Not only scientists for what it's worth...) [07:58] How many builds should there be ? [07:58] If it makes you feel better, I rarely lose my energy for arguing about things nobody else cares about ;p [07:58] fullermd: thanks, that helps :) [07:58] vila, about 3 [07:58] * fullermd . o O (to most people's everlasting chagrin...) [08:00] right, the three built successfully [08:00] Is there a better way to check that than visual-grepping the 'Built succesfully' line ? [08:01] In https://edge.launchpad.net/~bzr-beta-ppa/+archive/ppa look at the 'build status' [08:01] * igc dinner [08:01] Ha, of course, thanks :) [08:02] so that was only jaunty? [08:02] i think now you should send the others [08:02] poolie: yup [08:04] yet another glitch to fix first (reverting to checkouts my cautious change to branches) [08:05] lol [08:05] my bzr.dev doesn't include the fix for httlib broken readline so I can't bzr launchpad-login, the irony... [08:05] :/ [08:05] it should [08:05] i thought john merged it [08:06] Yeah, I should pull :) [08:06] far better :) [08:24] ok that's enough [08:24] night all [08:33] How long should I wait between a successful build and its occurrence in synaptic ? [08:33] vila: for the main ubuntu archive that can be easily 2h [08:34] mvo: ok, thanks, nothing to worry about then ? [08:35] err, should bzr-beta-ppa be considered the same as ubuntu archive in that respect ? [08:36] hm, no (I missed that it was about a PPA). but it does take a bit too nowdays [08:37] mvo: ok, I'll just wait then [08:37] mvo: The Ubuntu mirrors are refreshed every 2 hours?! [08:40] Lo-lan-do: the main archive, I can't speak about the mirrors. might be a bit more than 2h nowdays though [08:41] Impressive. [08:43] yes, its pretty nice === kiko-phone is now known as kiko [08:47] mvo: the user experience regarding uploading to ppa is really nice too, congrats [08:48] mvo: showing up in synaptic ! That was fast too :) === apw`` is now known as apw [08:54] nice :) [08:54] poolie: ppa uploaded for dapper feisty gutsy hardy intrepid jaunty, build for jaunty and intrepid ok, others pending [08:56] spiv: entertaining... [08:56] [54/1114 in 17s] branch_implementations.test_branch.TestFormat.test_set_reference(RemoteBranchFormat-default) [08:56] Command terminated === thekorn_ is now known as thekorn [10:28] hello, I did a 'bzr branch https://username/server/pathtorepository' from my svn server. now I did some changes and 'bzr commit' to apply locally but 'bzr push' says: 'bzr: ERROR: No push location known or specified.' [10:28] so: How do I push my changes to the central svn repository? [10:28] theAdib: "bzr push :parent" or "bzr push https://username/server/pathtorepository" [10:29] The former is shorthand for the latter in more recent versions of bzr. [10:30] :parent works smoothly :-) Thx. [10:34] how to get the source code of bazaar ... bzr branch lp:amarok is giving me error( cannot pass through proxy) I think I should try to create a patch... [10:34] shankhs: 'bzr branch lp:bzr' presumably won't help. ;) Try 'bzr branch http://bazaar-vcs.org/bzr/bzr.dev/'. === Odd_Blok1 is now known as Odd_Bloke [10:35] Odd_Bloke: thanx [10:35] * Lo-lan-do likes "apt-get source bzr" [10:37] Lo-lan-do: Not necessarily the best way to get a bzr that's worth patching. [10:37] If you're running unstable that gives you the source for bzr 1.5. [10:39] shankhs: What do you get if you run 'bzr launchpad-login'? [10:42] Odd_Bloke: No Launchpad user ID configured. [10:42] * awilkins thinks an lp: resolver that worked through SSH would be great [10:43] I can't use the lp: convenience links, I have to work out the full SSH link because our proxy config is such a PITA [10:45] bzr branch bzr+ssh//bazaar.launchpad.net/~bzr/bzr/trunk bzr.dev # That should work, might have to add username, and of course, you need a public key registered on your LP account [10:46] Missing : [10:46] Darn [10:47] Using lp: resolves the detail through some kind of HTTP-RPC which doesn't work through a bad proxy config even if SSH will work [10:47] awilkins: yeah, bug in the python xml-rpc library, not trivial to work around [10:47] so what should I do? [10:48] shankhs: Register a public key on LP and use ssh, or if you can pull over plain http, use that [10:48] I registered and got some SSH key now??? [10:52] shankhs: no, you need to generate an ssh key yourself, and then upload the public part [10:53] shankhs: https://help.launchpad.net/YourAccount/CreatingAnSSHKeyPair [10:54] ya I got it a very long string i registered starting ssh-rsa [10:54] LarstiQ: How is that going to help me ? [10:55] shankhs: it won't use http, so no http proxy [10:56] LarstiQ: ok [10:57] LarstiQ: can you please tell me the commands that I can use with ssh and bzr to download the source codes? [10:58] shankhs: after you have followed that CreatingAnSSHKeyPair page, do `bzr launchpad-login `, and then the `bzr branch lp:bzr` command [11:00] although, maybe that only works for people in the bzr team, not sure [11:00] shankhs: you tried `bzr branch http://bazaar-vcs.org/bzr/bzr.dev` and that didn't work? [11:01] LarstiQ: bzr: ERROR: Unknown repository format: 'Bazaar RepositoryFormatKnitPack6 (bzr 1.9)\n' [11:01] LarstiQ: when i tried bzr branch http://bazaar-vcs.org/bzr/bzr.dev [11:02] shankhs: What version of bzr are you using? [11:02] The latest is 1.11. [11:02] Odd_Bloke: how to know that? [11:02] * LarstiQ leaves shankhs in Odd_Bloke's capable hands and goes afk [11:02] shankhs: bzr --version [11:03] Bazaar (bzr) 1.6.1 Python interpreter: /usr/bin/python 2.5.2 Python standard library: /usr/lib/python2.5 bzrlib: /usr/lib/python2.5/site-packages/bzrlib Bazaar configuration: /home/shankhs/.bazaar Bazaar log file: /home/shankhs/.bzr.log Copyright 2005, 2006, 2007, 2008 Canonical Ltd. http://bazaar-vcs.org/ bzr comes with ABSOLUTELY NO WARRANTY. bzr is free software, and you may use, modify and redistribute it und [11:03] shankhs: Right, that's quite an old version. What distribution/OS are you running? [11:04] ubuntu 8.10 [11:04] updated an hour ago [11:05] Odd_Bloke: by the way what is the in bzr launchpad-login is it the ssh key? [11:06] shankhs: No, it's your Launchpad username. [11:06] ok [11:06] The SSH key is associated with the user account somewhere in the Launchpad UI. [11:08] shankhs: I suggest you look at https://launchpad.net/~bzr/+archive/ppa [11:08] i am getting ERROR: Connection error: while sending GET /%7Eshankhs/%2Bsshkeys: (111, 'Connection refused') [11:08] It has more recent versions of bzr available. [11:13] Odd_Bloke: what about the login problem I tried `bzr launchpad-login shankhs` [11:19] shankhs: What login problem? [11:20] am getting ERROR: Connection error: while sending GET /%7Eshankhs/%2Bsshkeys: (111, 'Connection refused') [11:20] my login name is shankhs [11:20] i tried my email id which i used to login [11:21] every permutation still getting the same error [11:22] I tried bzr launchpad-login [11:24] shankhs: You'll have the same problem because it uses the same XMLRPC mechanism [11:24] I reinstalled bzr and the version is 1.11 [11:24] awilkins: oh! [11:24] Do it with [11:24] --no-check [11:25] awilkins: thanx I got through [11:25] awilkins: now how to log out [11:25] and how to download stuffs [11:26] is there any good tutorials [11:27] All launchpad-login does is configure the username it uses for LP branches. [11:27] Which OS are you using~? [11:27] ubuntu [11:28] Ok then, ssh-agent running? [11:28] awilkins: how to check? [11:28] Try ps -e | grep ssh [11:29] ya `4771 ? 00:00:00 sshd ` [11:29] Ok, just an ssh server running [11:29] Not sure if Ubuntu will automatically prompt you for a key password actually. Try. [11:30] Is your new private key on your ubuntu keyring? [11:30] I am a complete newbie... [11:30] how to check them...please explain thankyou [11:31] Ok, did you generate a keypair for Launchpad? === pasky is now known as TRALALALA === TRALALALA is now known as pasky [11:31] awilkins: ya [11:32] Ok, go to Accessories > Passwords and Encryption Keys [11:34] Is there "Secure Shell Key" on your personal keys tab? [11:34] ya I think i have SSH key and a GPG key [11:35] Ok, so your key is already installed. [11:35] awilkins: now? [11:35] Try bzr branch bzr+ssh://bazaar.launchpad.net/~bzr/bzr/trunk bzr.dev [11:37] its running [11:38] Big branch, may take a while :-) [11:39] awilkins: ok :) [11:39] I'm doing it too, just to slow you down some more [11:39] (mostly to check I'm not talking out of my rear) [11:40] ssh: connect to host bazaar.launchpad.net port 22: Connection timed out [11:40] :( [11:40] whats wrong now ? :( [11:41] awilkins: bzr: ERROR: Connection closed: please check connectivity and permissions (and try -Dhpss if further diagnosis is required) [11:41] Ah, so you didn't get the twirly baton and stuff [11:42] awilkins: nope :( [11:42] Sounds like your firewall policy doesn't allow outgoing connections on port 22 [11:42] You in a corporate office? [11:42] awilkins: no in a college [11:42] Only slightly better [11:43] awilkins: we have squid proxy if this might help [11:43] You know your proxy settings? [11:43] ya [11:43] (bzr.dev hit revision 4000, woohoo) [11:44] Is your HTTP_PROXY env variabl eset? [11:44] yes [11:44] try [11:45] trying again [11:45] I got r4000. \o/ [11:46] awilkins: its not working and I am about to give up on bzr [11:47] awilkins: hey wait look https://bugs.launchpad.net/bzr/+bug/241698 [11:47] Ubuntu bug 241698 in bzr "POST to authenticating proxy causes "necessary data rewind wasn't possible" error" [Undecided,Confirmed] [11:47] Is that what you're getting? [11:47] AFAIR that's a squid issue [11:48] I tried using -Dhpss and I got HPSS calls: 1 [11:49] svn commands are working fine... though [11:50] svn has a simpler protocol for http I think [11:50] I think so [11:51] It's backed up by a more complex server [11:51] anyways I got to go hope some simpler remedy comes [11:51] Bazaar should work if you just serve the branch out of a dump http server [11:51] awilkins: thanx for your precious time... [11:52] awilkins: dump http server....!!! [11:52] dumb [11:52] sorry [11:52] ohh [11:52] ya i also think so [11:52] noo... it does work [11:52] svn has it's own WEBDAV server [11:53] It integrates as an Apache module [11:53] does that mean bzr wont work behind proxy servers? [11:53] awilkins: I see [11:53] It should work behind proxy servers ; but not all proxy servers are equal [11:53] awilkins: i agree [11:53] It didn't work out of IIS to start with either [11:53] awilkins: have you ever tested on squid...? [11:54] Because both IIS and Python ignore RFCs [11:54] shankhs: I know there there is at least one protocol bug related to a bug in squid [11:54] awilkins: I am login using SSH , how to log out? [11:54] launchpad-login doesn't "log in" it just configures the username [11:55] awilkins: ok [11:55] awilkins: does bzr implements its own protocol or some existing one? [11:56] awilkins: thanx once again for your help [11:56] Bazaar will use ; direct file reads, FTP, SFTP, HTTP, bzr protocol (raw, over SSH, or HTTP) [11:57] file/FTP/SFTP/bzr+http are writable [11:57] plain HTTP is read only [11:58] The smart protocol tries to reduce bandwidth consumption, the dumb ones all read indexes and read regions of files [11:59] You can also send merge bundles as files and have them treated as first-class branches as long as you have access to the branch they are based on [12:00] (well, for pulling/merging purposes) [12:00] So email also works (this is generally how patches to Bazaar are accepted, by mailing a merge bundle to the mailing list) [12:00] And i've been talking to myself.... [12:01] * awilkins slaps head [12:01] awilkins: I hadn't noticed. :p [12:05] awilkins: we're all very impressed, honest. [12:08] * awilkins hides under his comfort blankie [12:27] Hi; how can you resume an interrupted „branch“ command? I assume you can't easily, since there's bug #125607 open [12:27] Launchpad bug 125607 in openoffice.org "package openoffice.org-evolution 2.2.1-5ubuntu2 failed to install/upgrade: Abhängigkeitsprobleme - lasse es unkonfiguriert (dup-of: 125400)" [Undecided,New] https://launchpad.net/bugs/125607 [12:27] Launchpad bug 125400 in openoffice.org "[MASTER] package openoffice.org-common 2.2.1-5ubuntu2 failed to install/upgrade: subprocess post-installation script returned error exit status 1" [High,Fix released] https://launchpad.net/bugs/125400 [12:27] awilkins: welcome to http://bazaar-vcs.org/Quotes :) [12:28] mmm.... #125067 [12:28] * vila watches the DOS failure in ubottu.... [12:28] bug #125067 ? [12:28] Launchpad bug 125067 in bzr "allow 'bzr checkout' to resume" [Wishlist,Confirmed] https://launchpad.net/bugs/125067 [12:31] Hi all! [12:31] Youssef: Hello. [12:32] Odd_Bloke: can you help me please? [12:32] im new to bazaar [12:32] Youssef: I'm about to go for lunch, actually. [12:33] awilkins might be around (PING!). [12:33] hmmm... okay thank you very much! [12:33] Ptthp [12:33] Go on then [12:33] awilkins I need you [12:33] i have a problem with bazaar [12:33] It's that or implement an XML/XSLT layer for a shell script [12:34] awilkins? [12:34] Ask [12:34] hhaa okay [12:35] so, i'm trying to push my project but it says =>This transport does not update the working tree of: bzr://localhost/. See 'bzr help working-trees' for more information. [12:35] why? [12:35] i checked it out [12:35] Because it's pushing to a smart server - it concerns itself with getting the revision data across but it doesn't update the remote tree. [12:36] If you want that tree updated, do `bzr up` in it [12:36] in it? [12:36] You also seem to be pushing to a branch on your own machine [12:36] yeah thats it [12:36] With a shell whos present working directory is inside the tree are pushing to [12:37] It's not necessary to push to a server if you are pushing locally [12:37]