2012/09/24/#bzr.txt

[03:19] <_kbulgrien> Ok, distributed newb here.  I have a distibuted setup and have pushed/pulled as needed to take a mod from a subordinate branch up to the parent branch.  bzr status shows nothing.  I expected it to show I needed an update.  bzr update got the change.  What should have I done to see that an update was required?
[03:20] <bob2> missing
[03:20] <bob2> you're pushing commits around
[03:24] <_kbulgrien> Thanks.  I guess I read `bzr help missing` bzr missing [OTHER_BRANCH] as indicating a branch other than the one I am checked out from... hmm...
[03:24] <_kbulgrien> So if I pulled or pushed something into my checked out branch, I would not expect, from help, to need `bzr missing`.
[03:25] <bob2> well
[03:25] <bob2> it depends what 'checked out' means
[03:26] <fullermd> In a light checkout, I believe status tells you when the branch has moved ahead.  In a heavy it won't because that would require connecting up to the remote branch every time, and avoiding that sort of stuff is why you have heavy in the first place.
[03:26] <bob2> if it's a bound branch/lightweight checkout, be super careful
[03:28]  * _kbulgrien has forgotten how to tell if a co is lightweight or not
[03:28] <fullermd> If you didn't explicitly make a light one, it's heavy.
[03:29] <lifeless> bzr info
[03:29] <lifeless> if it says 'checkout of' its heavy.
[03:29] <_kbulgrien> bzr info -v doesn't seem to make it clear.   (Yeah, I forgot if I did --lightweight or not)
[03:29] <lifeless> paste the output ?
[03:29] <_kbulgrien> ok. gotcha.  heavy.
[03:32] <_kbulgrien> ok, I  can see the heavy/light thing.  Will just need to try to assimilate this mentality.  Am used to non-distributed where getting a status tells you everything.
[03:34] <_kbulgrien> well, successful night then... finally got bzr working to help keep a web host and a local test server tied together via version control so I can develop and push up, etc. to roll out changes.
[03:34] <lifeless> cool
[03:34] <_kbulgrien> So now I have something real to practice with.
[03:36] <_kbulgrien> I sure appreciate irc folks helping here and there to make it happen.  I think I shall call it an evening.  BBL
[03:42] <bob2> bound branches:(
[08:40] <mgz> hey
=== jam2 is now known as jam
=== mmrazik is now known as mmrazik|lunch
=== mmrazik|lunch is now known as mmrazik
=== dpb_ is now known as Guest75974
[15:17]  * SamB_MacG5 wonders why bzr-svn has to fetch so much revision info in order to grab -rsvn:1
[15:39] <flea> Hi, I wish to convert a current/live working directory into a bzr-controlled branch
[15:39] <flea> i'm a bit confused about the paths involved compared to when starting a repo from fresh
[18:13] <SamB_MacG5> anyone know jelmer's preferred TAB width?
[18:16] <SamB_MacG5> oh, it looks like the number I was looking for was "3"
[18:18] <SamB_MacG5> hmm, no, wait...
[18:31] <jelmer> SamB_MacG5: ideally, tabs in the C code, 4 spaces for python code
[18:32] <jelmer> SamB_MacG5: though subvertpy isn't sticking to that everywhere in the C code I think
[18:32] <SamB_MacG5> yeah, I hit a place where it doesn't line up without setting `tab-width' to 3
[18:33] <SamB_MacG5> obviously I didn't mean for the Python code
[18:33] <jelmer> SamB_MacG5: where is that?
[18:35] <SamB_MacG5> jelmer: adm_add() in wc.c
[18:35] <jelmer> SamB_MacG5: that's got all tabs here
[18:36] <SamB_MacG5> maybe I'm pulling from the wrong branch...
[18:37] <jelmer> SamB_MacG5: http://people.samba.org/bzr/jelmer/subvertpy/trunk/ ?
[18:37]  * SamB_MacG5 was lazy and used lp:subvertpy
[18:38] <SamB_MacG5> hrmm
[18:38] <SamB_MacG5> no new revisions
[18:40] <SamB_MacG5> what the heck, trying to pull from the https: version dies on me ...
[18:40]  * SamB_MacG5 blames Subversion
[18:41] <SamB_MacG5> or not, the traceback was from the middle of the smart protocol code :-(
[18:44] <SamB_MacG5> jelmer: check the line after this again:
[18:44] <SamB_MacG5> RUN_SVN_WITH_POOL(temp_pool, svn_wc_add2(
[18:45] <jelmer> SamB_MacG5: right, so there is one misalignment.
[18:46] <SamB_MacG5> but, don't fix it now, because it'll conflict with what I'm working on ;-)
[18:50] <SamB_MacG5> sweet, it built
[19:06]  * SamB_MacG5 looks into the __future__ to figure out how to make the test suite run ..
[19:08] <lifeless> SamB_MacG5: which test suite ?
=== iBasic is now known as BasicOSX
[19:14] <SamB_MacG5> lifeless: subvertpy's
[19:19] <SamB_MacG5> darn, now it's segfaulting :-(
[19:23] <lifeless> Possibly make check
[19:23] <lifeless> or setup.py test
[19:24] <lifeless> failing that read the README :)
[19:28] <SamB_MacG5> lifeless: that's what's segfaulting ;-)
[19:28] <SamB_MacG5> (no, not the README ;-)
[19:28] <lifeless> jelmer may be around, he's in the US atm - jelmer: ^
[19:28]  * jelmer waves
[19:28]  * SamB_MacG5 is trying make gdb-check right now
[19:30] <lifeless> jelmer: oh hai
[19:31]  * SamB_MacG5 wishes Apple had provided debugging symbols for Python
[19:31] <jelmer> hey lifeless:-)
[19:31] <lifeless> jelmer: any thoughts on SamB_MacG5's subverypy segfaults ?
[19:31] <lifeless> jelmer: also, have fun and sambacamp :)
[19:31] <lifeless> s/and/at/
=== joey[a] is now known as joey
[19:59]  * SamB_MacG5 tries to puzzle out how to get a Python traceback without any debug symbols ...
=== jaimef is now known as ober
[20:55] <SamB_MacG5> oh, whoopsie ...
[20:56]  * SamB_MacG5 discovers com.mygreatcompany.pkg.python-fastimport in his receipt database
=== jordan__ is now known as jordan
=== ianbrandt_ is now known as ianbrandt
[21:42] <delinquentme> ok so I've got this dir:  BETY  ... and all of my project files within the BETY folder.  I would like to relocate all of the working files into BETY/trunk  .. .so that I might make another branch ( which contains a bunch of redesign stuff ) in BETY/redesign
[21:42] <delinquentme> right now I've attempted to move all the files from BETY/* to BETY/trunk/* .. but I get complaints in bzr
[21:44] <lifeless> delinquentme: what complains
[21:45] <SamB_MacG5> delinquentme: tried moving also BETY/.* to BETY/trunk/.* ?
[21:45] <delinquentme> well as far as bzr status is concerned ... every file was removed
[21:46] <delinquentme> and then I've just got an Unknown:  trunk/
[21:47] <delinquentme> NM!
[21:47] <delinquentme> for some reason the mv * didnt copy the bzr files
[21:47] <lifeless> well
[21:47] <delinquentme> niiice
[21:47] <lifeless> the problem is that trunk is part of BETY/*
[21:47] <lifeless> so I think you need to be more specific
[21:47] <lifeless> starting from scratch
[21:47] <lifeless> assuming BETY is the root of your branch
[21:47] <lifeless> bzr mkdir trunk
[21:48] <SamB_MacG5> lifeless: what?
[21:48] <SamB_MacG5> why would he want to add a directory called trunk to the branch, exactly?
[21:49] <lifeless> SamB_MacG5: thats the scenario given
[21:49] <SamB_MacG5> lifeless: I'm pretty sure he wants that to *be* a branch
[21:49] <lifeless> delinquentme: then bzr mv <file> trunk/ - don't use *, because it will pic up trunk as well
[21:49] <lifeless> SamB_MacG5: for that, he needs to move them and then use bzr split
[21:50] <lifeless> delinquentme: is BETY the root of a branch, or just a container that contains multiple branches ?
[21:51]  * SamB_MacG5 is pretty sure that delinquentme was just not expecting * not to match names with leading .s
[21:53] <delinquentme> yeah :D
[21:53] <SamB_MacG5> jelmer: so, when you "fix compatability with svn 1.4", do you remember to actually run the test suite?
[21:53] <delinquentme> what SamB_MacG5 said
[21:54] <jelmer> SamB_MacG5: it's been a long time since I've run with 1.4 myself
[21:54] <delinquentme> is trunk a bad name for the local version of master branch?
[21:55] <SamB_MacG5> well, I think this segfault is beyond me :-(
=== yofel_ is now known as yofel
[21:59] <delinquentme> and pushing my newly branched "redesign" dir up onto the primary repo?
[21:59] <delinquentme> can I just do bzr push :parent  and have it make the redesign branch on the remote / primary repo?
[22:00] <SamB_MacG5> jelmer: could you try running the tests against https://code.launchpad.net/~naesten/subvertpy/1050949-svn-1.4-compat ?
[22:00] <SamB_MacG5> delinquentme: probably not ?
[22:00] <SamB_MacG5> delinquentme: try "bzr config" to see
[22:01] <delinquentme> bzr: ERROR: unknown command "config"
[22:01] <SamB_MacG5> huh
[22:01] <SamB_MacG5> works on Bzr 2.3!
[22:02] <delinquentme> 2.1.4
[22:02] <delinquentme> ahhhh
[22:03] <SamB_MacG5> wow that's old
[22:03] <SamB_MacG5> and I thought I was behind
[22:03] <jelmer> SamB_MacG5: trying
[22:04] <SamB_MacG5> test_committed_queue (subvertpy.tests.test_wc.AdmTests) ...
[22:04] <SamB_MacG5> Program received signal EXC_BAD_ACCESS, Could not access memory.
[22:05]  * SamB_MacG5 gets that when he runs "make gdb-check TEST_OPTIONS=-v"
=== vednis_ is now known as mars
=== Dwizar is now known as BY0LOG1C
=== wgrant_ is now known as wgrant
[22:33] <delinquentme> umm was there a way to add just specific files for a commit?
[22:33] <delinquentme> I know how to add just one file to a commit .. what if I just want to pick out 10 files ?
[22:34] <lifeless> bzr commit filename
=== hazmat` is now known as 5JTAACNC2
=== 5JTAACNC2 is now known as hazmat
[22:52] <delinquentme> lifeless, yeah i mean how can I add only a selected number of files
[22:53] <delinquentme> SamB_MacG5, I think I asked about this before .. but the adding multiple files .. but not all files to a commit?
[22:53] <delinquentme> or say .. how to add multiple files and stage them to be comitted
[22:55] <lifeless> delinquentme: there is no staging area in bzr
[22:55] <lifeless> delinquentme: bzr commit file1 file2 file3 etc
[22:57] <mark06> hi all, is it possible to overwrite committer in commit log for adding email info (that is, apply bzr whoami "Name <email>" to all places where bzr whoami "Name" has been used)
[22:58] <lifeless> there is a bzr rewrite patch in the bug tracker somewhere
[22:58] <lifeless> IIRC jelmer rejected it as not being a common enough use case :)
[22:59] <delinquentme> lifeless, can I add all files within a dir?
[23:00] <delinquentme> say $ bzr commit -m "testing" dir1/*  dir5/* dir7/file.txt
[23:00] <lifeless> bzr commit -m testing dir1 dir5 dir7/file.txt
[23:00] <delinquentme> to add everything in dir1 and dir5 .. as well as file.txt
[23:00] <lifeless> bzr add says *track these files for changes*
[23:00] <delinquentme> thanks!
[23:00] <lifeless> bzr commit says *commit everything*
[23:00] <lifeless> bzr commit path1 path2... pathN
[23:00] <mark06> some decisions may be painful in future, I should have set an email since beginning, I used to think that was privacy, now I think that was paranoia
[23:01] <lifeless> says commit anything in these paths (that is already tracked)
[23:04] <delinquentme> lifeless, so if I havn't added a file yet IE it comes up under "unknown" on a bzr status
[23:04] <lifeless> right
[23:04] <delinquentme> those files wont be committed ?
[23:04] <lifeless> right
[23:04] <delinquentme> ah ok ok
[23:04] <lifeless> have you gone through the tutorial ?
[23:06] <delinquentme> yeah I just get wonked up on some of the difference between this and git
[23:06] <delinquentme> what about something akin to git commit --amend
[23:06] <delinquentme> like if I want to add new information to a previous commit ( one which I only have locally )
[23:12] <lifeless> bzr uncommit; bzr commit
[23:21] <mark06> --amend? ugh. So lifeless, if I ever find that bug, I would need to patch bzr myself, right? (and also, change the patch as it's likely outdated)
[23:27] <lifeless> mark06: bzr-rewrite is a plugin
[23:27] <lifeless> mark06: but yes, with s/bzr/bzr-rewrite/
[23:27] <mark06> ok thanks lifeless