| [00:17] <lifeless> jelmer: ping | |
| [00:17] <lifeless> jelmer: whats a good, small, example of using tdb from C? | |
| [00:18] <dash> Hmm | |
| [00:18] <lifeless> mmmm | |
| [00:18] <lifeless> bacon! | |
| [00:18] <lifeless> So dash, whats your bacon number? | |
| [00:18] <dash> I just merged from trunk into my feature branch and now it wants to undo all the changes I made | |
| [00:18] <dash> maybe not all, but enough for me to feel like I did something wrong | |
| [00:19] <dash> should I remerge? should I merge from my feature branch into a copy of the mainline? | |
| [00:19] <lifeless> dash: the merge command undid your changes? | |
| [00:19] <Peng_> 1.) What command did you run? 2.) You're not using anything weird like bzr-svn, are you? | |
| [00:19] <dash> yes, bzr-svn =/ | |
| [00:19] <dash> oh, you know what | |
| [00:19] <lifeless> Peng_: I bet its a backout merge from trunk; but we should track it down | |
| [00:20] <dash> yeah this is probably my fault | |
| [00:20] <lifeless> dash: was it the merge command that undid your changes? | |
| [00:20] <dash> so I did "bzr merge http://svn-server/..../" | |
| [00:20] <jelmer> lifeless: I think tdbdump might be a good example | |
| [00:21] <lifeless> jelmer: point me at the src? | |
| [00:21] <dash> 'bzr diff' now shows a diff that will remove a bunch of my changes | |
| [00:21] <lifeless> dash: ok. | |
| [00:21] <jelmer> lifeless: apt-get source tdb && $VISUAL tdb-*/tools/tdbdump.c | |
| [00:21] <lifeless> dash: next question. Had you on trunk, *or on a branch merged into trunk* undone those same changes | |
| [00:22] <dash> yes | |
| [00:22] <lifeless> dash: then bzr is correctly propogating those undoes. | |
| [00:22] <jelmer> lifeless: and tdbtool for an example of writes | |
| [00:22] <dash> yeah, i figured | |
| [00:22] <dash> hence saying it was my fault :) | |
| [00:22] <lifeless> dash: :) | |
| [00:22] <lifeless> dash: just being sure | |
| [00:23] <lifeless> dash: you should just reject the undo here. | |
| [00:23] <dash> i checked this code into svn, then later reverted it | |
| [00:23] <dash> lifeless: right | |
| [00:23] <lifeless> I loves it when tools DTRT | |
| [00:23] <dash> i love it when they don't surprise me :) | |
| [00:23] <dash> i guess i should do the merge in 2 stages then | |
| [00:23] <dash> one to reject the rollback done earlier | |
| [00:24] <dash> and one to pull in the rest of the changes | |
| [00:24] <lifeless> dash: no, you should merge, reject, commit | |
| [00:24] <lifeless> one step | |
| [00:24] <lifeless> dash: that will mean that the next merge from this branch to trunk will reinstate these changes | |
| [00:24] <dash> hm! ok | |
| [00:24] <lifeless> and [correctly] show that you wanted them present on this branch at every point. | |
| [00:29] * dash makes a copy of his branch to make sure he didn't misunderstand :) | |
| [00:33] <dash> lifeless: the difficulty is that the mainline contains changes I _do_ want | |
| [00:33] <lifeless> dash: thats fine | |
| [00:33] <lifeless> dash: the way you reject could be to use 'bzr shelve' | |
| [00:33] <dash> true. | |
| [00:39] <lifeless> dash: another way to reject is | |
| [00:40] <lifeless> bzr merge -r [revert-in-trunk]..[revert-in-trunk-1] $trunk | |
| [00:40] <lifeless> oh | |
| [00:40] <lifeless> bzr merge -r [revert-in-trunk]..[revert-in-trunk-1] $trunk --force | |
| [00:40] <lifeless> (because you already have a merge in your tree) | |
| [00:42] <dash> that's what i meant by doing it in two steps | |
| [00:42] <lifeless> dash: with only one commit | |
| [00:42] <lifeless> ? | |
| [00:42] <dash> oh, didn't know you could do multiple merges with a single commit. | |
| [00:42] <lifeless> so | |
| [00:42] <lifeless> specifically | |
| [00:42] <lifeless> 1) merge trunk | |
| [00:42] <lifeless> 2) merge -r X..X-1 trunk --force | |
| [00:43] <lifeless> 3) commit | |
| [00:43] <lifeless> so 2) == reject | |
| [00:43] <dash> hah. | |
| [00:43] <dash> even nicer than what i was trying to do. | |
| [00:43] <dash> that makes more sense too | |
| [01:03] <dash> okay! other than having to resolve the same conflicts twice, that seems to have done the job | |
| [01:03] <dash> lifeless: thanks for the help. | |
| [01:03] <lifeless> dash: odd that you had to resolve any conflicts | |
| [01:03] <lifeless> dash: but I'm glad you are happy | |
| [01:04] <dash> yeah well, i should just know better than to check stuff in and revert it like that. :) | |
| [04:36] <johnf> could someone please try out 1.16rc1 packages I just buit in the bzr-beta-ppa. I've split out the docs into a -doc package. Need someone to make sure that the upgrade is smooth | |
| [04:38] <GPHemsley> jelmer: Are you around? | |
| === `6og is now known as kgoetz | |
| [08:37] <wgrant> johnf: It seems you accidentally clobbered the Jaunty 1.16rc1 in bzr-beta-ppa with the third Dapper attempt. You also don't have the luxury of python-support or python-central on Dapper, so I wouldn't even try backporting bzr to it... | |
| [08:39] <johnf> wgrant: there were already dapper packages but they don't seem to want to build after I split out bzr-doc. I think I'll just revert all the changes for it and it can just have one package | |
| [08:42] <wgrant> johnf: Ah, there is indeed a prehistoric python-support in dapper universe that isn't used for anything. Oops. | |
| [08:44] <johnf> and I've royally mucked up the version numbers. I should have been changing ~bazaar not the rc :( | |
| [08:45] <wgrant> johnf: You didn't do too badly - you just incremented the Debian revision, not the RC number. | |
| [08:45] <johnf> ahh I did do the right thing. Well close enough anyway | |
| [08:46] <wgrant> ~bazaar[23] would be correct, but what you did isn't world-burning. | |
| [08:57] <LarstiQ> johnf: och, you didn't muck up like me introducing a wrong version (-bazaar instead of ~bazaar) that ranks higher than all debian versions we'd want to use | |
| [08:59] <johnf> wgrant: so when would you up -[23]? That is something I've never been quite clear on | |
| [09:00] <wgrant> johnf: The first number after the last - is the Debian version. | |
| [09:00] <wgrant> As in Debian the distribution, not Debian the package format. | |
| [09:00] <wgrant> Debian's first 1.16 version would be 1.16-1. | |
| [09:00] <wgrant> If they make a change, 1.16-2. | |
| [09:01] <wgrant> When Ubuntu makes a change to a package, we append 'ubuntu1'. The 1 there is the ubuntu version - we'll increment that as we make subsequent changes. | |
| [09:01] <johnf> what about in the case where it is say a native ubuntu package do you still always leave at -1 in case it makes it intu debian? | |
| [09:01] <wgrant> We'll use -0ubuntu1 | |
| [09:01] <johnf> ahh ok makes sense | |
| [09:01] <wgrant> That way when it makes it into Debian, it'll be -1, and -1 > -0ubuntu1 | |
| [09:02] <wgrant> So upgrades will work when -1 is synced into Ubuntu. | |
| [09:02] <johnf> ok new packages should be syncing shortly | |
| [09:03] <wgrant> Have you test-built the Dapper on locally? | |
| [09:03] <wgrant> s/on/one/ | |
| [09:04] <wgrant> You can reasonably drop Dapper packages in a couple of weeks, too (Dapper loses desktop support RSN) | |
| [09:06] <LarstiQ> wgrant: ah, I've been looking forward to that | |
| [09:27] <fullermd> Mmm. These DeprecationWarning's on py2.6 are gonna get real old real fast... | |
| === Kissaki^0ff is now known as Kissaki | |
| [09:35] <wgrant> fullermd: In bzr, or in general? | |
| [09:35] <fullermd> Well, from pycrypto technically, but since I see it running bzr... | |
| [09:36] <pthulin> is there a way to merge one project into another? I have two python projects, and now I want them to be the same without loosing history | |
| [09:36] <wgrant> fullermd: That's fixed in Jaunty and Karmic, at least. | |
| [09:37] <fullermd> Neither of which is much help to people who run neither Jaunty, Karmic, Ubuntu, or Linux ;) | |
| [09:38] <wgrant> http://launchpadlibrarian.net/23631444/python-crypto-2.6-337073.diff is useful, in those strange cases. | |
| [09:39] <fullermd> Inneresting. | |
| [11:06] <visik7> ok I'm in a strange situation: | |
| [11:06] <visik7> I've my working tree out of date | |
| [11:07] <visik7> how could I know at which version is it ? | |
| [11:28] <fullermd> Well, you can poke around manually in .bzr/. At one time, info gave you something you could do some math on. | |
| [11:28] <fullermd> Or you could use revno/revision-info --tree, but you have to do that in the future. | |
| [11:29] <lifeless> or just run bzr update | |
| [11:30] <fullermd> Well, that doesn't make it known. Just moot :p | |
| [11:31] <lifeless> sureit does. | |
| [11:31] <lifeless> it would be -1 | |
| === cprov is now known as cprov-afk | |
| [13:46] <jelmer> GPHemsley: hi | |
| [17:28] <fcorrea> hello there. Aw...using bzr 1.13.1 here and working against subversion on the server. Does anyone know why it takes forever to push changes to back to svn? I looked up and found that it used to happen with bzr 0.5- | |
| [17:29] <fcorrea> btw it also happens when doing bzr co | |
| [17:30] <jelmer> fcorrea: newer versions of bzr-svn should be a bit better in that regard | |
| [17:30] <fcorrea> jelmer: will try | |
| [17:33] <jelmer> fcorrea: also, the first time you do a push/pull it'll be a fair bit slower because it has to analyse the repository, after that it should be significantly faster | |
| [17:34] <fcorrea> jelmer: yeah, but it also happens when pushing it back to svn, so I assume all the analysis were already done, but I may be wrong | |
| [17:49] <jelmer> fcorrea: also, how slow is slow? | |
| [18:30] <Mez> Out of curiosity, any of the bzr people heading to EuroPython? | |
| [18:32] <Mez> other than you, lifeless | |
| [18:35] <Mez> (wait, 2 talks? someone with dedication!) | |
| [19:31] <visik7> I've my working tree out of date | |
| [19:31] <visik7> how could I know at which version is it ? | |