| [05:14] <FailBit> can I use launchpadlib to grab a commit message from a specified revno of a bazaar branch? | |
| [12:01] <shadeslayer> out of curiousity, how does launchpad deal with incoming uploads over ftp? Does it react when it detects new .changes files ( in which case, what happens if the .changes file is uploaded first, and the orig is uploaded at the end, and the processing might happen before the orig gets fully uploaded ) | |
| [12:01] <shadeslayer> or is it a cron job? | |
| [12:02] <shadeslayer> wgrant: cjwatson ^^ ideas? | |
| [12:30] <shadeslayer> wgrant: cjwatson ah ah, nvm | |
| [12:44] <FailBit> Anyone get my question from yesterday? | |
| [13:09] <cjwatson> FailBit: Not directly. You can use launchpadlib to get the branch URL, but you'd then need to use bzr or bzrlib to get individual commit messages. launchpadlib doesn't wrap bzrlib. | |
| [13:09] <FailBit> Ok | |
| [13:12] <cjwatson> shadeslayer: It's a cron job; but the FTP upload server is a custom one that moves the whole upload into place in one go at the end of the upload transaction. | |
| [13:14] <shadeslayer> roger | |
| [13:14] <shadeslayer> cjwatson: it's just that I was using the rsync/scp backend | |
| [13:14] <shadeslayer> and that does fancy chmod stuff at the end | |
| [13:15] <shadeslayer> and inotincoming had already triggered reprepro and reprepro had already deleted the files, so chmod failed | |
| [13:15] <shadeslayer> hurray for things being too fast | |
| [13:15] <shadeslayer> :P | |
| [13:15] <shadeslayer> anyway, lunch :) | |
| [13:21] <cjwatson> shadeslayer: Right, the upload server has to be quite careful if it wants to avoid that class of problem. On the other hand it's a trade-off; it means that you can't resume an incomplete upload if it fails, and other similar problems (https://bugs.launchpad.net/launchpad/+bug/30415) | |
| [13:21] <ubot5> Launchpad bug 30415 in Launchpad itself "soyuz upload system pays attention to ftp sessions" [Low,Triaged] | |
| [13:41] <FailBit> oops. | |
| [13:41] <FailBit> how do I get a information about a revno with bzrlib | |
| [13:41] <FailBit> ? | |
| [13:44] <FailBit> bzr log --line -c <revno> <branch> | |
| [13:44] <shadeslayer> cjwatson: mmm fun | |
| [13:44] <shadeslayer> anyway, I'm writing my own solution with ruby inotify now | |
| [13:45] <shadeslayer> because for some reason sftp + inotincoming doesn't work | |
| [15:29] <FailBit> what's "GhostRevisionsHaveNoRevno" ? | |
| [15:32] <FailBit> $ bzr log --line -r 13337 lp:inkscape/experimental | |
| [15:32] <FailBit> bzr: ERROR: Server sent an unexpected error: ('error', 'GhostRevisionsHaveNoRevno'… | |
| [15:32] <FailBit> what does it mean? | |
| [15:40] <FailBit> please, I'm so impatient… | |
| === Spads_ is now known as Spads | |
| [21:05] <FailBit> anyone? | |
| [21:17] <wgrant> FailBit: That's bug #1049124. What exactly are you trying to do? | |
| [21:17] <ubot5> bug 1049124 in Bazaar "bzr log -rxxx crashes out with 'ghost' in branch, but not in trunk" [Medium,Confirmed] https://launchpad.net/bugs/1049124 | |
| [21:18] <FailBit> wgrant: I was able to use nosmart+branchurl to get my bot to work | |
| [21:18] <wgrant> It specifically affects asking a stacked branch for a revision that isn't in the branch itself, so it's often simple to work around. | |
| [21:18] <wgrant> Right, that works too. | |
| [21:19] <wgrant> The smartserver revision_id_to_revno method doesn't know to look at the stacked-on repository. | |
| [21:19] <FailBit> The only problem is that bzr is slow =/ | |
| [21:19] <FailBit> 10 seconds after the request, usually it finds it | |
| [21:19] <wgrant> Well, at least you can perform remote operations at all :) | |
| [21:20] <wgrant> With eg. git you have to clone the repository locally to get the log. | |
| [21:21] <FailBit> Compare | |
| [21:21] <FailBit> /msg InkBot !bug 12345 | |
| [21:21] <FailBit> /msg InkBot !revno 12345 | |
| [21:21] <ubot5> bug 12345 in isdnutils (Ubuntu) "isdn does not work, fritz avm (pnp?)" [Medium,Fix released] https://launchpad.net/bugs/12345 | |
| [21:21] <wgrant> Sure. If you want quick, regular access, just keep a copy of the branch locally. | |
| [21:22] <FailBit> !revno takes another (optional) argument | |
| [21:22] <ubot5> FailBit: I am only a bot, please don't think I'm intelligent :) | |
| [21:22] <FailBit> /msg InkBot !revno 1234,lp:lib2geom | |
| [21:23] <FailBit> You can't have every branch always up to date locally unless you are the server :P | |
| [21:23] <FailBit> taotie or something like that. | |
| [21:23] <wgrant> That's the one. | |
| [21:24] <wgrant> The current codehosting infrastructure doesn't let us provide direct access like that in an efficient manner. Upcoming changes should improve that, but it's all at least a year off. | |
| [21:46] <FailBit> wgrant: the launchpad automailer is kinda odd — for example, comments in merge proposals have a mail sent instantly, while branch revision emails take a few minutes | |