| === SamB_ is now known as SamB | ||
| jelmer | mathrick: that's plausible | 08:37 |
|---|---|---|
| mathrick | jelmer: just to check, I'm correct in trying to use bzr rebase to do what I was trying to do, right? | 08:38 |
| mathrick | I actually just retried that with 2a separate dir branches, and it still doesn't like what I'm doing | 08:47 |
| mathrick | "No revisions to rebase" | 08:47 |
| mathrick | I'm beginning to suspect it's replay that I want, which is somewhat inconvenient as I now have to create extra branches to do that | 08:47 |
| mathrick | jelmer: so yeah, replay worked like a champ, and it seems to me you did it that way to make sure rebase can never lose commits (the way it can in git if you're not careful)? Which is an entirely reasonable choice to make, but it needs documentation badly | 09:06 |
| mathrick | ie. rebase can only ever forward commits onto a newer tip, whereas to change the branching point to a non-tip revision, you need to branch + replay explicitly | 09:07 |
| mathrick | the whole of rewrite is really, really sparsely documented given the complexity of the topic | 09:08 |
| mathrick | seems like bzr help should have 'rewrite' listing *all* the commands in the plugin (it doesn't currently), plus 'rebase', 'rebase-tutorial', 'replay', and 'rebase-for-git-users' | 09:09 |
| mathrick | I might write some if I manage to do the things I want to do without any undue delays | 09:10 |
| mathrick | oh, and a bug, bzr switch -b -r X doesn't actually work | 09:10 |
| === lool- is now known as lool | ||
| fullermd | I had to use replay sometime earlier this year (late last?) to excise a commit. It took a fair while figuring out rebase wouldn't do it, and a while further and several trial runs to figure out how to get replay to DTRT. | 09:18 |
| fullermd | Sadly, it seems pretty mired in the Tuitless Abyss :| | 09:18 |
| fullermd | (even in bzr terms) | 09:19 |
| jelmer | mathrick: rebase rebases the *current* branch on the one specified. Rebasing on the second to last revision doesn't really make sense. | 09:22 |
| jelmer | mathrick: thanks for the suggestions, though unfortunately bzr-rewrite is orphaned | 09:25 |
| mathrick | jelmer: yes, but that's because the meaning of "rebase" is different between git and bzr. In git I'd say "git rebase --onto pristine bugfix1 bugfix2", which results in both bugfix1 and bugfix2 branching out of pristine's tip, rather than the accidental bugfix2 branching out of bugfix1 | 09:25 |
| mathrick | but there's no such syntax in bzr rebase, because I cannot give it a separate target and comparison branch | 09:26 |
| mathrick | so in bzr, it's bzr switch -b tmp; bzr switch bugfix2; bzr uncommit -r branch:co:pristine; bzr replay co:tmp -r branch:co:bugfix1.. | 09:29 |
| mathrick | bzr rmbranch co:tmp | 09:30 |
| jelmer | mathrick: it might very well be true that bzr-rebase is unaware of colocated branches - colocated branches are still a development thing, don't use them for productiony things | 09:33 |
| mathrick | jelmer: nah, it's not any more unaware than of regular branches, because I get the exact same thing | 09:34 |
| mathrick | the problem is really that rebase is not supposed to do that, but it's not explicit about it | 09:34 |
| mathrick | bzr rebase != git rebase | 09:34 |
| mathrick | git's rebase is really bzr rebase + bzr replay + some things that just aren't there (ie. git rebase -i and edit/squash revisions) | 09:35 |
| jelmer | mathrick: I'm not disagreeing, but I also don't really care anymore :-) | 09:39 |
| mathrick | jelmer: why not? About bzr rewrite specifically, or bzr in general? | 09:40 |
| jelmer | mathrick: both | 09:40 |
| mathrick | that's a pity | 09:41 |
| mathrick | oh, I missed your comment about rewrite being orphaned | 09:48 |
| mathrick | jelmer: if I wrote the docs, would you review and push it though? | 09:49 |
| mathrick | fullermd: yeah, I can see why rebase wouldn't do that even if it results in more work to get the same result, but the docs just don't tell you anything to expect that | 09:50 |
| jelmer | mathrick: I'd be happy to hand reign of the project over to somebody else | 10:02 |
| mathrick | jelmer: out of curiosity, what do you use for personal VCS needs these days? | 10:13 |
| jelmer | mathrick: git | 10:22 |
| jelmer | mathrick: anyway, it'd be great to see these projects continued even if I am no longer interested in them myself | 10:31 |
| mathrick | yeah | 10:32 |
| === marcoceppi_ is now known as marcoceppi | ||
| Azrael_- | hi | 14:52 |
| jelmer | hi Azrael_- | 14:54 |
| Azrael_- | i always fail branching a repo. i get the information "Getötet 2427kB/s | Revisionen herunterladen:Inserting stream:Estimate 459382/587439" Getötet = killed | 14:55 |
| Azrael_- | any idea why this could happen and how to prevent this? | 14:56 |
| Azrael_- | nm | 14:56 |
| Azrael_- | never thought bzr would need so much ram. it was killed due to insufficient ram | 14:58 |
| Azrael_- | 1gb ram wasn't enough | 15:02 |
| Azrael_- | perhaps 4gb swap will do it | 15:02 |
| jelmer | Azrael_-: what version are you running? | 15:05 |
| Azrael_- | Bazaar (bzr) 2.6.0dev2 | 15:05 |
| Azrael_- | yeah, it's working now | 15:06 |
| Azrael_- | wow, it only needed 1,3gb ram to complete this action | 15:17 |
| === BasicMBP is now known as BasicOSX | ||
| Wiz_KeeD | hey guys | 17:39 |
| Wiz_KeeD | what does this mean? Conflict adding file aluart_product_configurator/wizard/sale_config.py.BASE. Moved existing file to aluart_product_configurator/wizard/sale_config.py.BASE.moved. | 17:40 |
| Wiz_KeeD | when going there i see no such file | 17:40 |
| Wiz_KeeD | maybe I deleted it | 17:40 |
| Wiz_KeeD | but why is the error still there and how to get rid of it? | 17:40 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!