| 2012-09-20 |
|
delinquentme
| drop a particular file back to the version in the previous commit? | 00:00 |
| → reshtnk7 joined | 00:02 |
| ← kerrick left | 00:03 |
| ← cjz left | 00:04 |
| → tmcmahon joined | 00:05 |
| ← ecksit left | 00:05 |
|
Anders_J
| git checkout filename | 00:05 |
| ← serbang left | 00:05 |
| ← pppaul left | 00:06 |
| → juvenal joined | 00:07 |
| → scelis joined | 00:08 |
| ← rubious left | 00:11 |
| → serbang joined | 00:11 |
|
frogonwheels
| delinquentme: You mean in the last commit you did, or the one before that? first is git checkout -- filename the second is git checkout HEAD~1 -- filename (the -- is optional but disambiguates the filename) | 00:11 |
|
delinquentme
| frogonwheels, git checkout <filename> did it :D | 00:12 |
|
| TY VM though :D | 00:12 |
| ← serbang left | 00:13 |
|
frogonwheels
| delinquentme: np. 'previous commit' is a bit ambiguous :) | 00:13 |
| → jperry2_ joined | 00:15 |
| ← glosoli left | 00:15 |
|
delinquentme
| HEAD^ | 00:17 |
| ← Anders_J left | 00:17 |
|
frogonwheels
| delinquentme: same diff. I use windows cmd shell, so I avoid using ^ 'cause they have to be doubled. | 00:17 |
| ← jperry2_ left | 00:17 |
|
frogonwheels
| delinquentme: HEAD^ and HEAD~1 are equivalent otherwise | 00:18 |
| ← hagen left | 00:18 |
| → serbang joined | 00:19 |
| → w09x joined | 00:19 |
| → rohan32 joined | 00:20 |
| → setmeaway joined | 00:20 |
| → hellopat joined | 00:20 |
| ← thiago left | 00:22 |
| ← aruiz left | 00:24 |
| ← mmc1 left | 00:25 |
| ← jdunck left | 00:25 |
| ← chuckharmston left | 00:29 |
| → patrick99e99 joined | 00:29 |
| ← wasanzy left | 00:29 |
| → skyf joined | 00:29 |
|
patrick99e99
| hi buddies.. I am working with submodules and have a few questions.. If I have a submodule within a git repo, if I go into that submodule's subdirectory, shouldn't git branch show me only branches for that submodule? | 00:30 |
|
| because I just went in there, created a test branch and then cd .. to get back to the root (the main repo), and git branch shows the branch I created inside the submodule's dir... | 00:30 |
|
| is that the way it is supposed to be? Or is this an indication that something is not set up properly? | 00:30 |
| ← goshawk left | 00:30 |
| → rubious joined | 00:31 |
| ← rubious left | 00:31 |
| ← dzonder left | 00:32 |
| ← hwrd|work left | 00:32 |
|
patrick99e99
| anyone alive here? | 00:33 |
| ← adamben left | 00:33 |
| → goshawk joined | 00:34 |
| ← dsantiago left | 00:36 |
| ← ipalaus left | 00:36 |
| → Chib_ joined | 00:37 |
| ← Chib left | 00:37 |
| ← iocor left | 00:38 |
| ← hyperair left | 00:40 |
|
cmn
| then it's not really a submodule | 00:43 |
|
| patrick99e99: ^^ | 00:43 |
| ← Chillance left | 00:43 |
| → macabre joined | 00:44 |
| → wasanzy joined | 00:47 |
| ← gfontenot left | 00:52 |
| → dsantiago joined | 00:53 |
| ← aspotashev left | 00:55 |
| ← Droolio left | 00:55 |
| ← BiggFREE left | 00:56 |
| ← scelis left | 00:58 |
|
frogonwheels
| patrick99e99: in your 'submodule' is there a .git file or folder? you sure it's not some other kind of !subproject ? | 00:59 |
|
gitinfo
| patrick99e99: [!subprojects] So, you want to add git repositories inside of other git repositories? Well, you have four main options. First is to just do it, add the repo to the outer project's .gitignore, and treat them entirely separately. Best if they are entirely seperate. Otherwise your best options are "!submodule" "!gitslave" and "!subtree" Try typing those commands into this IRC channel. | 00:59 |
| ← jalama left | 00:59 |
| → scelis joined | 01:01 |
| ← reshtnk7 left | 01:01 |
| ← serbang left | 01:03 |
| → acrocity_ joined | 01:04 |
| ← devth left | 01:04 |
| → AlbireoX joined | 01:05 |
| ← kermit left | 01:05 |
| → iori joined | 01:07 |
| ← trungqtran left | 01:08 |
| ← w09x left | 01:10 |
| → adamben joined | 01:10 |
| → trungqtran joined | 01:11 |
|
patrick99e99
| frogonwheels: hi... there is no .git in the submodule... | 01:11 |
|
| !submodule | 01:11 |
|
gitinfo
| git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://book.git-scm.com/5_submodules.html | 01:11 |
| → Yuuhi``` joined | 01:11 |
| ← aantix left | 01:11 |
| ← acrocity_ left | 01:13 |
| ← Yuuhi`` left | 01:13 |
| → LeMike1 joined | 01:14 |
| → acrocity_ joined | 01:14 |
| ← esto left | 01:14 |
| ← LeMike left | 01:15 |
| ← scelis left | 01:17 |
| → alunduil joined | 01:18 |
| → xiangfu joined | 01:19 |
| ← maletor left | 01:19 |
| → esto joined | 01:20 |
| → hyperair joined | 01:20 |
| ← jesseFromYVR left | 01:22 |
| → Layke joined | 01:23 |
|
cmn
| patrick99e99: then it's not a submodule | 01:25 |
| ← wasanzy left | 01:25 |
| → cooloney joined | 01:26 |
| ← notVert left | 01:27 |
| ← storrgie left | 01:29 |
| → jalama joined | 01:30 |
| ← Mocramis left | 01:30 |
| → Mocramis joined | 01:31 |
| ← TheObfuscator left | 01:31 |
| → GeorgeJ joined | 01:33 |
| ← AlbireoX left | 01:35 |
| → AlbireoX joined | 01:35 |
| ← gusnan left | 01:35 |
| ← Milossh left | 01:36 |
| → wasanzy joined | 01:37 |
| ← ncr100 left | 01:38 |
| ← eletuchy_ left | 01:40 |
| → drl joined | 01:41 |
| ← kcj left | 01:41 |
| ← kpreid left | 01:41 |
| → kpreid joined | 01:41 |
| ← tmcmahon left | 01:41 |
| → thierryp joined | 01:44 |
| ← Mocramis left | 01:46 |
| → Mocramis joined | 01:47 |
| ← t0rc left | 01:47 |
| → jackh joined | 01:48 |
| → shane_ joined | 01:49 |
| ← thierryp left | 01:49 |
| ← gtrak` left | 01:51 |
| → maletor joined | 01:51 |
| ← juvenal left | 01:52 |
| ← blorbx left | 01:53 |
| → gtrak joined | 01:54 |
| ← etcetera left | 01:55 |
| → reshtnk7 joined | 01:56 |
|
rking
| I wonder what I'm doing, lately, that is leaving my .git/config's without a '[branch "master"]…'. I also wonder what I'm supposed to do to get it back. =\ | 01:56 |
|
cmn
| deleting it? | 01:57 |
|
rking
| I could (continue) to just edit .git/config, but I'm sure I"m misunderstanding something. | 01:57 |
|
| cmn: How would I be doing that? | 01:57 |
|
cmn
| git branch -d master | 01:57 |
|
rking
| I sure didn't run that, but hrm. | 01:58 |
|
| Also, is the recommended way to get it back this?: git config branch.master.remote origin; git config branch.master.merge=refs/heads/master | 01:58 |
|
cmn
| some version of that, if you branch is getting deleted | 01:58 |
|
| use git branch --set-upstream | 01:59 |
| ← artnez left | 01:59 |
|
cmn
| or push -u | 01:59 |
| toothr → toothrot | 01:59 |
|
rking
| Aha, push -u, that's it. =) | 01:59 |
| ← acrocity left | 01:59 |
| acrocity_ → acrocity | 01:59 |
| ← ehsan left | 01:59 |
| ← reshtnk7 left | 02:00 |
| ← jgoss left | 02:01 |
|
rking
| I don't get what that config item really means. E.g. why it's not git pull -u origin master | 02:01 |
| ← maletor left | 02:02 |
|
cmn
| what would that command even do?\ | 02:02 |
|
| and what config item are you talking about? | 02:02 |
| → juvenal joined | 02:03 |
|
rking
| branch.master.{remote,merge} | 02:03 |
| ← CollinPDX left | 02:03 |
|
rking
| All I really know is that it prevents "git pull" from working, so I don't see why "set upstream" isn't an option for the pull command. | 02:03 |
|
| There's some basic thing I don't understand about all of this. | 02:05 |
| → zhouxiukun joined | 02:05 |
| ← simesy left | 02:05 |
| → Bosox20051 joined | 02:06 |
|
cmn
| !fetch4 | 02:06 |
|
gitinfo
| [!fetchfour] Never use the four-word version of git-fetch or git-pull (e.g. git fetch remote refspec). It always ends in tears. Yes, if you understand the implications of FETCH_HEAD it can technically be done, but really it is easier to just fetch the whole remote (or perhaps edit the fetchspec if you never want other bits). If you must, see !fetch4why | 02:06 |
|
cmn
| even for push it's a bit iffy | 02:07 |
| ← iori left | 02:07 |
| ← duckxx left | 02:09 |
| → eletuchy joined | 02:10 |
| ← YoMomma left | 02:10 |
| rohan32 → rohan32|sleep | 02:13 |
| → ProLoser|Work joined | 02:13 |
|
ProLoser|Work
| how do i see the current svn revision #? | 02:13 |
|
cmn
| svn info, but you might want to ask in #svn | 02:14 |
| ← opus left | 02:15 |
| → mtlife joined | 02:17 |
| → maqr joined | 02:18 |
| → reshtnk7 joined | 02:18 |
|
ProLoser|Work
| i'm using git svn | 02:18 |
|
| i figured it out | 02:18 |
|
cmn
| it's still a svn thing | 02:18 |
| ← metcalfc left | 02:19 |
| ← skyf left | 02:20 |
| ← mtlife left | 02:21 |
|
patrick99e99
| I am trying to deploy a submodule with capistrano.. it keeps saying " Unable to resolve revision for 'HEAD' on repository" | 02:21 |
| → an0maly_ joined | 02:21 |
|
ProLoser|Work
| cmn: so am i to understand that all my svn commands are available under `git svn` | 02:22 |
| → maletor joined | 02:22 |
| → skyf joined | 02:22 |
| ← GeorgeJ left | 02:22 |
| → jesseFromYVR joined | 02:22 |
| → GeorgeJ joined | 02:23 |
| → jaredoconnor joined | 02:24 |
|
jaredoconnor
| can anyone point me toward some reading on good branching models for release management, hot fixes, etc ? | 02:25 |
|
cmn
| ProLoser|Work: what? | 02:25 |
|
| jaredoconnor: !flow | 02:25 |
|
gitinfo
| jaredoconnor: [!gitflow] The description of the gitflow branch workflow model is at http://nvie.com/posts/a-successful-git-branching-model/ while a tool to help implement this workflow is at https://github.com/nvie/gitflow See http://sethrobertson.github.com/GitBestPractices/#workflow for other workflow suggestions/references | 02:25 |
|
jaredoconnor
| thanks cmn | 02:25 |
| ← ProLoser|Work left | 02:26 |
|
jaredoconnor
| i saw the first one, but wanted to review a few others before i decided | 02:26 |
| → etcetera joined | 02:26 |
|
cmn
| you need to use whatever fits your system, not what someone says on the Internet | 02:26 |
| ← jason237 left | 02:27 |
| ← wasanzy left | 02:31 |
| → tjholowaychuk joined | 02:34 |
| → rchavik joined | 02:34 |
| ← rchavik left | 02:34 |
| → rchavik joined | 02:34 |
| ← Jay_Levitt left | 02:36 |
| → rurufufuss joined | 02:36 |
| ← cybersphinx left | 02:38 |
| ← eletuchy left | 02:39 |
| ← shane_ left | 02:40 |
| ← friskd left | 02:42 |
| ← intransit left | 02:42 |
| ← osse left | 02:42 |
| ← steevel left | 02:42 |
| ← zamba left | 02:42 |
| ← jast left | 02:42 |
| ← Shadeness left | 02:42 |
| ← toothrot left | 02:42 |
| ← netj left | 02:42 |
| ← iivvoo left | 02:42 |
| ← SirCmpwn left | 02:42 |
| ← hashpuppy left | 02:42 |
| ← markatto left | 02:42 |
| ← rue left | 02:42 |
| ← RJ3000_ left | 02:42 |
| ← AndreSomers left | 02:42 |
| ← bddn left | 02:42 |
| ← dimka left | 02:42 |
| ← EugeneKay left | 02:42 |
| ← smw left | 02:42 |
| ← Evanlec left | 02:42 |
| ← w0m left | 02:42 |
| ← The24z_away left | 02:42 |
| ← blathijs left | 02:42 |
| ← aef_ left | 02:42 |
| ← worch left | 02:42 |
| ← Tabmow left | 02:42 |
| ← quadz left | 02:42 |
| ← project2501a left | 02:42 |
| ← idoru left | 02:42 |
| ← ajpiano left | 02:42 |
| → iivvoo joined | 02:42 |
| → EugeneKay joined | 02:42 |
| → smw joined | 02:42 |
| → Tabmow joined | 02:43 |
| → markatto joined | 02:43 |
| → netj joined | 02:43 |
| → Evanlec joined | 02:43 |
| → zamba joined | 02:43 |
| → RJ3000_ joined | 02:44 |
| → toothr joined | 02:45 |
| → duckxx joined | 02:46 |
| → SirCmpwn joined | 02:47 |
| → idoru joined | 02:48 |
| → friskd joined | 02:48 |
| → intransit joined | 02:48 |
| → osse joined | 02:48 |
| → steevel joined | 02:48 |
| → jast joined | 02:48 |
| → hashpuppy joined | 02:48 |
| → Shadeness joined | 02:48 |
| → aef_ joined | 02:48 |
| → rue joined | 02:48 |
| → AndreSomers joined | 02:48 |
| → dimka joined | 02:48 |
| → w0m joined | 02:48 |
| → The24z_away joined | 02:48 |
| → blathijs joined | 02:48 |
| → worch joined | 02:48 |
| → quadz joined | 02:48 |
| → project2501a joined | 02:48 |
| → ajpiano joined | 02:48 |
| calvino.freenode.net set mode: +v | 02:48 |
| → kerrick joined | 02:49 |
| ← jaredoconnor left | 02:52 |
| ← jesseFromYVR left | 02:53 |
| ← FernandoBasso left | 02:54 |
| → converge joined | 02:56 |
| ← converge left | 02:56 |
| → converge joined | 02:56 |
| ← converge left | 02:56 |
| ← orafu left | 02:57 |
| → orafu joined | 02:58 |
| → ehsan joined | 02:58 |
| ← maletor left | 02:58 |
| → jesseFromYVR joined | 02:59 |
| → boombatower joined | 03:01 |
| → metcalfc joined | 03:01 |
| → fireh joined | 03:05 |
| ← gtrak left | 03:05 |
| → shane_ joined | 03:05 |
| ← Yuuhi``` left | 03:08 |
| ← shane_ left | 03:12 |
| ← towski left | 03:14 |
| ← rohan32|sleep left | 03:14 |
| → Yuuhi joined | 03:15 |
| ← tjholowaychuk left | 03:15 |
| → towski joined | 03:16 |
| ← towski left | 03:16 |
| → luc_ joined | 03:19 |
| ← Eridius left | 03:19 |
| → jdunck joined | 03:24 |
| ← luc_ left | 03:24 |
| → Esa joined | 03:25 |
| ← goshawk left | 03:26 |
| ← jalama left | 03:28 |
| ← Esa_ left | 03:28 |
| → eletuchy joined | 03:29 |
| ← patrick99e99 left | 03:33 |
| ← trungqtran left | 03:34 |
| ← alexisabril left | 03:35 |
| → johnToledo joined | 03:36 |
| → sangi joined | 03:36 |
| ← shankarcabus left | 03:39 |
| ← johnToledo left | 03:39 |
| → johnToledo joined | 03:39 |
| ← johnToledo left | 03:40 |
| → cakehero joined | 03:41 |
| ← apok left | 03:41 |
| → johnPollmann joined | 03:41 |
| ← cakehero left | 03:42 |
| → scelis joined | 03:43 |
| → luc_ joined | 03:44 |
| ← scelis left | 03:44 |
| → thierryp joined | 03:45 |
| ← gregm left | 03:47 |
| → simesy joined | 03:49 |
| ← thierryp left | 03:50 |
| → maletor joined | 03:51 |
| ← maqr left | 03:52 |
| ← OOPMan left | 03:53 |
| ← skyf left | 03:55 |
| → e66 joined | 03:58 |
| → macrover joined | 03:59 |
| ← maleknet left | 04:15 |
| → a-l-e joined | 04:16 |
| ← maletor left | 04:20 |
| ← aca20031 left | 04:23 |
| ← EyesIsMine left | 04:27 |
| → maletor joined | 04:28 |
| → Eridius joined | 04:28 |
| → kcj joined | 04:28 |
| ← kukks left | 04:29 |
| → caseymcg joined | 04:31 |
| → iori joined | 04:31 |
| ← Targen left | 04:32 |
| ← jcao219 left | 04:37 |
| ← rchavik left | 04:37 |
| → pppaul joined | 04:39 |
| → beautifulmind joined | 04:40 |
| → shruggar joined | 04:41 |
| → irqq joined | 04:41 |
| → pretty_function joined | 04:42 |
| ← Davey left | 04:42 |
| ← eletuchy left | 04:44 |
| ← johnPollmann left | 04:46 |
| ← beautifulmind left | 04:51 |
| → brasom joined | 04:52 |
| ← macabre left | 04:54 |
| ← variousnefarious left | 04:57 |
| → variousnefarious joined | 04:59 |
| ← AlbireoX left | 05:00 |
| → xxc_michael joined | 05:02 |
| → shane_ joined | 05:02 |
| ← jdunck left | 05:04 |
| → dfr|mac joined | 05:04 |
| ← Olipro left | 05:05 |
| ← reshtnk7 left | 05:05 |
| ← etcetera left | 05:05 |
| → Olipro joined | 05:09 |
| → thirdknife joined | 05:09 |
| ← oogatta left | 05:10 |
| → Targen joined | 05:11 |
| ← macrover left | 05:11 |
| → tvw joined | 05:14 |
| → florin joined | 05:14 |
| florin → Guest57555 | 05:15 |
| ← Guest57555 left | 05:15 |
| → maleknet joined | 05:16 |
| ← shruggar left | 05:17 |
| ← philips_ left | 05:19 |
| ← sh4rm4 left | 05:19 |
| → sh4rm4 joined | 05:21 |
| ← pppaul left | 05:21 |
| ← luc_ left | 05:22 |
| → philips_ joined | 05:23 |
| ← tvw left | 05:25 |
| → tvw joined | 05:25 |
| ← irqq left | 05:26 |
| ← delinquentme left | 05:26 |
| → leeN joined | 05:27 |
| → irqq joined | 05:27 |
| → o] joined | 05:28 |
| → beautifulmind joined | 05:29 |
|
o]
| any good link with a cheat sheet for git? | 05:29 |
|
| I always forget basic things like creating a new repo or setting up a new client | 05:29 |
| → private joined | 05:30 |
| ← tvw left | 05:30 |
| private → Guest6654 | 05:30 |
|
frogonwheels
| o]: Here's one http://cheat.errtheblog.com/s/git | 05:31 |
| → tdlguiinstall joined | 05:31 |
| ← tdlguiinstall left | 05:31 |
| ← dfr|mac left | 05:32 |
| → tdlguiinstall joined | 05:32 |
| → shruggar joined | 05:32 |
|
frogonwheels
| o]: and another few http://bit.ly/PrKnM4 | 05:32 |
| → j0ran joined | 05:33 |
| ← madewokherd left | 05:34 |
| ← irqq left | 05:34 |
| ← e66 left | 05:35 |
| ← idletom left | 05:38 |
| ← shruggar left | 05:38 |
| ← ph^ left | 05:38 |
| → ph^ joined | 05:39 |
| → oogatta joined | 05:39 |
| ← quazimodo left | 05:39 |
| → idletom joined | 05:39 |
|
o]
| yeah! I have to ask for help | 05:40 |
|
| ok, I have a lot of source code on my machine | 05:40 |
|
| and a server that I can access via ssh | 05:40 |
|
| I want to create a repository on that server and commit the files of my machine | 05:40 |
|
| first I have to create a bare repository on the server, right? | 05:40 |
|
| I did that a lot of times and I've forgotten everything :( | 05:41 |
| → ecksit joined | 05:41 |
|
frogonwheels
| o]: for a start you might want to look at !gitolite for managing your server repo.. | 05:41 |
|
gitinfo
| o]: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html | 05:41 |
|
frogonwheels
| o]: you also might want to go over some of the !doc | 05:41 |
|
gitinfo
| o]: A list of useful documentation of all kinds is here: http://git-scm.com/documentation -- or try the wiki at http://git.wiki.kernel.org/. Also try typing "!book" "!cs" "!bottomup" "!parable" "!best_practices" or "!vcbe" or "!designers" here in IRC. !book is probably the most helpful. | 05:41 |
|
frogonwheels
| o]: especially !cs | 05:41 |
|
gitinfo
| o]: "Git for Computer Scientists" is a quick introduction to git internals for people who are not scared by phrases like Directed Acyclic Graph. http://eagain.net/articles/git-for-computer-scientists/ See also !concepts !bottomup | 05:41 |
|
ecksit
| why is my git repo riddled with commits like "Merge branch 'master' of github.com:organisation/repo-name" | 05:41 |
|
frogonwheels
| ecksit: probably because you use !pull | 05:42 |
|
gitinfo
| ecksit: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/ | 05:42 |
| → iliketurtles joined | 05:42 |
|
o]
| frogonwheels: that server already has some of my repositories :) | 05:43 |
| ← longbeach left | 05:43 |
|
o]
| but it was more than 8 months than I've created them | 05:43 |
| ← iori left | 05:45 |
| → _iori_ joined | 05:45 |
| → thierryp joined | 05:45 |
| ← whaley left | 05:47 |
| ← khaije|minder left | 05:47 |
| ← juvenal left | 05:48 |
|
frogonwheels
| o]: basically, you init your local repo and commit stuff to it.. then you will need to init a place on the server (gitolite will handle that) to push to | 05:48 |
|
| o]: the same server can do it with gitolite.. it just means that future repo names would be more like git@yourhost rather than yourusername@yourhost | 05:49 |
|
o]
| git init --bare | 05:49 |
|
| Initialized empty Git repository in /home/josediego/die-tk.git/ | 05:49 |
|
mjt
| `git init' is simpler than installing gitolite ;) | 05:49 |
|
o]
| hmm my memory is not so bad | 05:49 |
| ← thierryp left | 05:50 |
|
frogonwheels
| o]: It really dpends on if you're thinking of opening it up to other users. | 05:50 |
|
o]
| now I need to create a local repo on the files that are on my desktop and add a remote, right? | 05:50 |
| ← iliketurtles left | 05:50 |
|
o]
| frogonwheels: not for now | 05:50 |
| → mishok13 joined | 05:50 |
|
frogonwheels
| o]: yep. | 05:50 |
|
o]
| how is the add remote ? | 05:50 |
| → eletuchy joined | 05:51 |
| → bddn joined | 05:51 |
|
frogonwheels
| o]: git remote add origin user@host:repo | 05:51 |
|
o]
| wonderful :)~~~~~ | 05:52 |
|
mjt
| git remote add <name> <url> | 05:52 |
|
o]
| mjt: should I put ssh:// on the url ? | 05:52 |
|
mjt
| host:repo already assumes ssh | 05:53 |
|
| or user@host: for that matter | 05:53 |
| → shruggar joined | 05:53 |
|
cmn
| and misbehaves if you put ssh:// in front, see man git clone for what each url means | 05:53 |
|
gitinfo
| the git-clone manpage is available at http://git-scm.com/docs/git-clone.html | 05:53 |
| → pantsman joined | 05:54 |
| ← pantsman left | 05:54 |
| → pantsman joined | 05:54 |
| ← jesseFromYVR left | 05:55 |
| ← eletuchy left | 05:55 |
|
frogonwheels
| o]: if you do, you ave to use ssh://josediego@host/home/josediego/die-tk.git rather than josedigeo@host:die-tk.git | 05:56 |
|
mjt
| heh | 05:56 |
|
frogonwheels
| o]: I know which one I'd use | 05:57 |
|
cmn
| or you can use ~ | 05:57 |
|
| or change your ssh options so that it roots in your home | 05:57 |
| → MrMeh joined | 05:57 |
|
o]
| all right | 05:58 |
| ← warthog9 left | 05:58 |
| → acrocity_ joined | 05:59 |
| → ThomasLocke joined | 06:00 |
| → fisted joined | 06:01 |
| ← acrocity left | 06:02 |
| acrocity_ → acrocity | 06:02 |
| ← maletor left | 06:02 |
| → weiyang joined | 06:02 |
| ← ecksit left | 06:03 |
| → shane__ joined | 06:03 |
| ← shane_ left | 06:04 |
| ← nigelb left | 06:05 |
| ← lucapost left | 06:05 |
| → lucapost joined | 06:06 |
| ← ph^ left | 06:06 |
| ← duckxx left | 06:06 |
| ← metcalfc left | 06:09 |
| ← shane__ left | 06:09 |
| → BadDesign joined | 06:10 |
| ← fisted left | 06:10 |
| → Anders_J joined | 06:11 |
| → nigelb joined | 06:12 |
| ← nigelb left | 06:13 |
| → nigelb joined | 06:13 |
| → AdrienBrault joined | 06:15 |
| → fisted joined | 06:15 |
| → cbreak-work joined | 06:15 |
| → iliketurtles joined | 06:15 |
| ← beautifulmind left | 06:15 |
| → _iron joined | 06:16 |
| ← fisted left | 06:17 |
| ← zhouxiukun left | 06:17 |
| → marthinal joined | 06:18 |
| → zhouxiukun joined | 06:18 |
| → resmo joined | 06:18 |
| ← newcode left | 06:18 |
| ← drizzd left | 06:19 |
| → luc_ joined | 06:20 |
| → JPohlmann joined | 06:20 |
| → drizzd joined | 06:20 |
| ← caseymcg left | 06:20 |
| → skorgon joined | 06:21 |
| ← skorgon left | 06:21 |
| → skorgon joined | 06:21 |
| → jbub joined | 06:21 |
| → fisted joined | 06:22 |
| → soee joined | 06:22 |
| ← redrocket left | 06:23 |
| → ph^ joined | 06:23 |
| → redrocket joined | 06:24 |
| → senny joined | 06:24 |
| ← shruggar left | 06:24 |
| ← iliketurtles left | 06:25 |
| → Milossh joined | 06:25 |
| → JohnBat26 joined | 06:25 |
| ← _iron left | 06:25 |
| ← resmo left | 06:27 |
| ← unreal left | 06:28 |
| → EyesIsMine joined | 06:28 |
| ← TheMoonMaster left | 06:29 |
| → piggybox joined | 06:29 |
| → flijten joined | 06:29 |
| → papegaaij joined | 06:30 |
| ← Caelum left | 06:30 |
| ← arvind_khadri left | 06:30 |
| → Caelum joined | 06:30 |
| → TheMoonMaster joined | 06:31 |
| → dzonder joined | 06:31 |
| → unreal joined | 06:31 |
| → acrocity_ joined | 06:32 |
| → rcu joined | 06:32 |
| → fornext joined | 06:33 |
| ← acrocity left | 06:34 |
| acrocity_ → acrocity | 06:34 |
| → fisted_ joined | 06:34 |
| → shruggar joined | 06:35 |
| → maletor joined | 06:37 |
| ← mishok13 left | 06:37 |
| ← fisted left | 06:37 |
| ← o] left | 06:37 |
| ← nyxynyx left | 06:38 |
| → o] joined | 06:39 |
| → reshtnk7 joined | 06:39 |
| ← shruggar left | 06:39 |
| → resmo joined | 06:40 |
| → shruggar joined | 06:41 |
|
EugeneKay
| sitaram - are you about, by chance? | 06:42 |
|
sitaram
| yup | 06:42 |
|
EugeneKay
| Are you aware of any hook/etc to gitolite that can fetch from an "upstream" repo when I fetch a gitolite repo? | 06:42 |
|
sitaram
| sure. Curiously, it's called "upstream" | 06:43 |
|
| src/triggers/upstream | 06:43 |
| → mmc1 joined | 06:43 |
|
EugeneKay
| Oh god github redesigned user pages | 06:43 |
|
| EugeneKay kills it with fire | 06:43 |
|
sitaram
| documentation for this is in the code itself; you won't find it in the docs | 06:43 |
| ← idletom left | 06:44 |
|
cmn
| EugeneKay: use the API | 06:44 |
|
sitaram
| EugeneKay: I thought your firefighter's oath prevented you from doing stuff like that... | 06:44 |
|
EugeneKay
| We need to have a fire to put out first..... | 06:44 |
|
| It looks like "nice" is a time in minutes to wait between fetching from upstream? | 06:44 |
| ← mysmysery left | 06:44 |
| → dc5ala joined | 06:45 |
| → Leo1 joined | 06:46 |
| ← shruggar left | 06:46 |
| → beneggett joined | 06:46 |
|
EugeneKay
| The doc just says "elapsed time", no indication of units | 06:46 |
| ← JPohlmann left | 06:47 |
|
sitaram
| see last line but one | 06:47 |
| → shruggar joined | 06:47 |
|
sitaram
| # * the 'nice' setting is in minutes and is optional; it is the minimum | 06:47 |
|
EugeneKay
| I must need more booze | 06:48 |
|
sitaram
| ("ADDITIONAL NOTES" section) | 06:48 |
| ← shruggar left | 06:48 |
|
EugeneKay
| Indeed, I was reading the last line | 06:48 |
|
| EugeneKay goes back to doing rude things to his gitolite install | 06:48 |
|
sitaram
| tautology (re "I must need more booze") | 06:48 |
|
EugeneKay
| :-D | 06:49 |
| → dermatthias joined | 06:49 |
|
EugeneKay
| Sadly, I am at grandmother's house this week and must limit my intake for the sake of appearances | 06:49 |
| ← mmc1 left | 06:49 |
|
sitaram
| :( | 06:49 |
| → nyxynyx joined | 06:51 |
| → maxandersen joined | 06:52 |
| ← maxandersen left | 06:52 |
| → maxandersen joined | 06:52 |
|
EugeneKay
| Question, if I create a new repo and this trigger is set up to run, will it fetch on the first init? Or the first time I try to fetch from the new repo? | 06:52 |
| → JPohlmann joined | 06:52 |
| ← JPohlmann left | 06:52 |
| → JPohlmann joined | 06:52 |
| → giallu joined | 06:53 |
| ← jbub left | 06:53 |
|
sitaram
| first fetch | 06:54 |
|
EugeneKay
| Hrm. I guess I can cron it | 06:54 |
| ← rcu left | 06:54 |
|
EugeneKay
| Or just say "screw it" | 06:54 |
|
sitaram
| or add it to the POST_CREATE trigger list -- might work | 06:55 |
|
| (as well as PRE_GIT I mean) | 06:55 |
|
| try it and see | 06:55 |
|
EugeneKay
| I shall | 06:55 |
|
| The doc in the .rc file says that's for post-wild-repo init | 06:55 |
| → jbub joined | 06:56 |
| → jaseemabid joined | 06:56 |
| ← jbub left | 06:58 |
| ← reshtnk7 left | 06:58 |
|
sitaram
| that needs fixing; see "trigger-specific arguments and other details" in http://sitaramc.github.com/gitolite/triggers.html | 06:58 |
|
EugeneKay
| Aha. | 06:59 |
| → quazimodo joined | 06:59 |
| → sebrock joined | 07:00 |
|
EugeneKay
| Hrm. The trigger doesn't appear to be doing anything. | 07:00 |
| → tsdh joined | 07:00 |
| → mishok13 joined | 07:01 |
|
sitaram
| probably because I have to go to lunch :) | 07:01 |
|
EugeneKay
| Likely. | 07:01 |
|
| Oh I see why | 07:01 |
|
| Your docs are wrong! | 07:01 |
|
| src/triggers/upstream says to use "option upstream.uri =". It's config! | 07:02 |
| → mtlife joined | 07:03 |
| → dege joined | 07:03 |
| → beautifulmind joined | 07:04 |
| → florin joined | 07:04 |
| florin → Guest83383 | 07:05 |
| ← bartman` left | 07:05 |
| → mtlife_w joined | 07:06 |
| → timj joined | 07:06 |
| → bartman` joined | 07:07 |
| ← mtlife left | 07:07 |
|
EugeneKay
| NOW it's roen. | 07:08 |
|
| broken, too. | 07:08 |
| → lb joined | 07:08 |
| ← lb left | 07:08 |
| → lb joined | 07:08 |
|
Sajbar
| EugeneKay: gratz! | 07:08 |
| → LeMike joined | 07:08 |
|
Sajbar
| EugeneKay: just checkout an earlier version where it worked :P | 07:08 |
| ← kerrick left | 07:09 |
|
EugeneKay
| Uh.... not really applicable :-p | 07:09 |
| → Danielss89 joined | 07:09 |
|
Sajbar
| EugeneKay: not running version control? | 07:09 |
|
| jesus;) | 07:09 |
| ← timj1 left | 07:09 |
|
Danielss89
| Hi. I'm having trouble pushing a branch to my remote.. | 07:09 |
|
| http://ds.gl/3d4c94.png | 07:09 |
|
| EugeneKay beats Sajbar about the head with a gitolite | 07:09 |
| ← AdrienBrault left | 07:10 |
|
Sajbar
| EugeneKay: you could have your config files under version control ;) | 07:10 |
|
EugeneKay
| Danielss89 - git push is expecting a repo name, "origin", THEN the ref | 07:10 |
| ← LeMike1 left | 07:10 |
|
EugeneKay
| Or just 'git push' | 07:10 |
|
| Run 'git branch -avv' & pastebin | 07:10 |
| ← acl_ left | 07:10 |
| ← letharion left | 07:10 |
|
Danielss89
| EugeneKay: how can i push the branch i'm in right now, to the branch called "booking" on remote | 07:12 |
|
EugeneKay
| Danielss89 - `git push -u origin booking` | 07:13 |
|
| Which will automagically set up tracking if it doesn't exist(which you want) | 07:13 |
| → caddoo joined | 07:14 |
| → RaceCondition joined | 07:14 |
| → rchavik joined | 07:14 |
| → shruggar joined | 07:14 |
| → deffrag joined | 07:15 |
|
Danielss89
| EugeneKay: it says that everything is up-to-date | 07:15 |
|
| but it's not.. | 07:16 |
| ← papegaaij left | 07:16 |
|
Danielss89
| EugeneKay: i have a commit which doesn't get pushed | 07:16 |
|
EugeneKay
| sitaram - damnit, nevemrind my above. I can't read. | 07:17 |
| → leokaito joined | 07:17 |
|
EugeneKay
| sitaram - and putting it in the POST_CREATE does in fact work. Yay! | 07:18 |
|
| sitaram - is there a way to specify a default for an option on ALL repos? | 07:18 |
|
| eg, I want to set the nice option to 60 globally, instead of doing it in each repo I make an upstream or | 07:19 |
|
| Danielss89 - !repro | 07:19 |
|
gitinfo
| Danielss89: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session, or at least explain exactly what you did that led up to the problem. This will help immensely with troubleshooting. | 07:19 |
| → Erenss joined | 07:19 |
|
EugeneKay
| Danielss89 - I am inclined to believe that `git` is working correctly, you're simply invoking something wrong. | 07:19 |
| → whitman joined | 07:19 |
| → letharion joined | 07:19 |
|
Danielss89
| EugeneKay: i agree :D | 07:19 |
| → caseymcg joined | 07:20 |
|
Danielss89
| ok.. now i'll start over.. | 07:20 |
|
| EugeneKay: i've just cloned my repo | 07:20 |
|
| now i need to get a remote branch so i can work on it | 07:21 |
| → chimay joined | 07:21 |
| → irqq joined | 07:22 |
|
Danielss89
| so i do: git checkout -b booking origin/booking | 07:23 |
| → TheHaven joined | 07:24 |
|
EugeneKay
| Yup. | 07:24 |
| → jbub joined | 07:25 |
| → mhauri joined | 07:26 |
| → osxorgate joined | 07:26 |
| ← tdlguiinstall left | 07:26 |
| ← beautifulmind left | 07:27 |
| → tdlguiinstall joined | 07:28 |
| ← worstadmin left | 07:28 |
| ← dzonder left | 07:30 |
| ← weiyang left | 07:30 |
|
osse
| The other day I was in a hurry and was confused at the same time. So I stashed some bits, commited som, resetted others. Now, when I run git log --all, I see a special commit I've not seen before. It's a commit with message 'index on <branch>: <prev commit msg>' How did that happen? | 07:30 |
| → papegaaij joined | 07:30 |
| → simesy_ joined | 07:30 |
|
frogonwheels
| osse: that's your stash | 07:30 |
|
jast
| osse: git log --all shows *everything*. that includes stashes. | 07:30 |
| ← dhruvasagar left | 07:30 |
|
frogonwheels
| osse: the stash comes in two parts. the contents of the wd and the contents of your index at the time of the stash. | 07:31 |
|
cbreak-work
| not really everything | 07:31 |
|
| just everything in history | 07:31 |
|
jast
| yeah, yeah | 07:31 |
|
cbreak-work
| git log -g for more of everything :) | 07:31 |
|
osse
| frogonwheels: ahh. Yeah I have a commit decorated with "refs/stash" as well. | 07:31 |
|
EugeneKay
| Danielss89 - did you get it sorted? ;-) | 07:31 |
|
Danielss89
| i'm working on it :D | 07:31 |
|
esc
| either I am in an A/B Test, or Github got a major makeover last night... | 07:31 |
|
EugeneKay
| esc - the latter | 07:32 |
|
osse
| I guess I've never stashed staged stuff before. | 07:32 |
|
| Thanks for the clarification :) | 07:32 |
|
EugeneKay
| I was just remarking upon how ugly it is | 07:32 |
|
esc
| EugeneKay: what the fuck, it's like almost a different site??? | 07:32 |
|
frogonwheels
| osse: np. look in gitk --all if you can | 07:32 |
| → thiago joined | 07:32 |
|
jast
| I don't see a makeover on github | 07:32 |
|
| so it must be in A/B stage | 07:32 |
|
EugeneKay
| sitaram's page now looks like http://i.imgur.com/RsQZJ.png | 07:32 |
|
esc
| okay, github has been going down the drain for a while now | 07:33 |
| → dhruvasagar joined | 07:33 |
|
cbreak-work
| github still looks the same for me | 07:33 |
|
esc
| EugeneKay: yeah, i have the same | 07:33 |
|
jast
| no, it's all betterification, I'm sure | 07:33 |
|
EugeneKay
| It's Agile | 07:33 |
| ← Vile1 left | 07:33 |
| ← simesy left | 07:33 |
|
osse
| frogonwheels: 'command not found: gitk' but 'git log --all --decorate --graph' does a decent job. I can see that the commit labeled refs/stash and the commit with messaged 'index of...' have the same parent. | 07:33 |
|
jast
| *that's* what you call a major makeover? jeez | 07:33 |
|
cbreak-work
| If I wanted to be agile, I'd be a gymnast. | 07:33 |
| ← sente left | 07:34 |
| ← simesy_ left | 07:34 |
|
frogonwheels
| osse: if you have x available.. (assuming ubuntu/deb) then apt-get install git-gui | 07:34 |
|
esc
| jast: okay, maybe not major, but it certainly is more than just little tweaks ;) | 07:34 |
|
frogonwheels
| osse: I like using guis where they are faster/useful. | 07:34 |
| → drougge joined | 07:35 |
|
jast
| the tweaks were small enough that I wouldn't have noticed them | 07:35 |
|
frogonwheels
| osse: needless to say I also spend a lot of time in shells as they often aren't. :) | 07:35 |
|
osse
| frogonwheels: I'm on an old, crappy RHEL :/ I installed git through the EPEL repo. It's not even in red hat's standard repos | 07:35 |
|
EugeneKay
| osse - use the rpmforge-extras repo | 07:35 |
| ← Guest6654 left | 07:35 |
|
EugeneKay
| Set 'includepkgs = git *-git git-*' in your rpmforge.repo to make sure it doesn't stomp on anything else | 07:35 |
|
| Or *git*, if you're feeling adventurous | 07:36 |
|
osse
| EugeneKay: ahh, I've heard about clashes between EPEL and other similar repos, but if you can filter like that I guess it'll work | 07:36 |
|
jast
| you forgot about gitweb in your first draft !111!11! | 07:36 |
| → jkremser joined | 07:36 |
|
EugeneKay
| You have to add the line after the rest of the config optiosn in the .repo [section] | 07:36 |
|
jast
| also tig | 07:37 |
|
osse
| frogonwheels: what was it you wanted to show me in gitk? | 07:37 |
|
EugeneKay
| But it works MUCH better than yum-plugin-priorities. | 07:37 |
|
| Which is to say, your system will still boot afterwards | 07:37 |
|
jast
| ooh, that sounds useful | 07:37 |
|
esc
| jast: really, it looks so unfamiliar, I couldn't help but not noticing ;) | 07:37 |
|
jast
| they didn't even change the colour scheme | 07:38 |
|
| the recent redesign of git-scm.com... now *that* was major | 07:38 |
|
| especially since it murdered our beloved tree-eating monster | 07:38 |
|
esc
| yeah, i loved that guy | 07:38 |
| ← _ikke_ left | 07:38 |
| ← dhruvasagar left | 07:39 |
|
esc
| we used him on the cover of our book, it was such an awesome logo | 07:39 |
| ← alunduil left | 07:39 |
| ← BadDesign left | 07:40 |
|
EugeneKay
| I blame the Github goes | 07:40 |
|
| s/goes/guys/ | 07:40 |
|
| God, I really need booze. I can't concentrate | 07:40 |
| → dhruvasagar joined | 07:40 |
|
osse
| So I've check out the commit that is the parent of this weird 'index of ...' commit. git show <hash> of the weird commit only shows me the git log-esque summary of the commit, but not the contents. Does that mean I hadn't in fact staged any thing? If not, how can I see what I had staged? | 07:40 |
| → alunduil joined | 07:41 |
|
osse
| The new profile of git's webpage is to slick and professional-looking :/ | 07:42 |
|
EugeneKay
| It looks like ass. | 07:42 |
|
| Compare the amount of information now with http://web.archive.org/web/20110611035709/http://git-scm.com/ | 07:43 |
|
osse
| Tried 'git show --format={fuller,raw} <commit>' Still the same output. | 07:44 |
| ← Anders_J left | 07:44 |
|
jast
| osse: to view diffs in 'git log' output, try -p ('patch') | 07:44 |
|
osse
| Ahh. My bad. 'git stash' *always* creates those two commits, even if the index is empty | 07:45 |
|
| that's why nothing appeared | 07:45 |
| → _ikke_ joined | 07:45 |
|
EugeneKay
| git-stash also eats babies, so I think you can forgive the excess commits. | 07:46 |
| → thierryp joined | 07:46 |
|
osse
| jast: nice, thanks. | 07:46 |
|
| made a test repo now where I stashed different ("non-empty") changes in both the actual file contents and index. Now 'git show <index commit>' shows what I expect. | 07:47 |
|
| Finally, peace of mind | 07:47 |
|
sitaram
| EugeneKay: re "nice option to 60 globally"; yes. See http://sitaramc.github.com/gitolite/git-config.html for examples (toward the end) | 07:48 |
|
EugeneKay
| Shiny | 07:48 |
| → kkszysiu_work joined | 07:48 |
|
EugeneKay
| Remind me what charity I'm supposed to donate to in your name? The EFF? | 07:48 |
| → BadDesign joined | 07:48 |
|
EugeneKay
| Actually, I don't think they'd take my money. | 07:48 |
|
sitaram
| EugeneKay: perl, vim, or failing that any open source project not involving mono | 07:48 |
|
| s/involving/using/ | 07:49 |
|
EugeneKay
| Mono, like "The Kissing Disease" ? >_> | 07:49 |
|
osse
| oh, you're one of *those* :P | 07:49 |
|
sitaram
| and proud of it | 07:49 |
|
osse
| no that's mano | 07:49 |
|
| or is it mono? | 07:49 |
|
EugeneKay
| http://en.wikipedia.org/wiki/Infectious_mononucleosis | 07:49 |
|
osse
| I was wrong. Last time that happened was back in '67 | 07:50 |
|
sitaram
| EugeneKay: some disease anyway; spread by people like Miguel de Icaza | 07:50 |
| → domguard joined | 07:50 |
| ← thierryp left | 07:50 |
|
osse
| I've used some Mono applications before. I didn't even sneeze | 07:50 |
|
jast
| osse: don't worry, it's free the first three times | 07:50 |
|
| mono doesn't induce sneezing | 07:51 |
| → chisler joined | 07:51 |
|
osse
| I didn't cough either | 07:51 |
|
sitaram
| EugeneKay: finally, note that you can also use a generic value (like 60 as you said) for | 07:51 |
|
jast
| hmm, they do normally ruin your throat | 07:51 |
| → Pupeno_W joined | 07:51 |
|
jast
| they, it, whatever | 07:51 |
| → Leo2 joined | 07:51 |
|
domguard
| hi, is there a way to use "git rebase -i" to squash two distant commits ? I mean there is 20+ commits between the two of them... | 07:51 |
|
sitaram
| EugeneKay: finally, note that you can also use a generic value (like 60 as you said) for "@all" repos, tnen specify other values for specific repos later (physically later in the text) | 07:51 |
| ← Leo1 left | 07:51 |
|
| EugeneKay nods sagely | 07:52 |
|
jast
| domguard: you can reorder the commits in the list | 07:52 |
|
| so just move it up and change it to 'squash' | 07:52 |
|
EugeneKay
| Though I'm using "option", not "config".... which I didn't know were different until tonight | 07:52 |
| → defts joined | 07:52 |
|
domguard
| ok great jast i'll try this immediately | 07:52 |
| → amcsi_work joined | 07:53 |
| ← LeMike left | 07:53 |
|
sitaram
| EugeneKay: differences between option and config: (1) "option foo = bar" is a short cut for "config gitolite-options.foo = bar" (2) config lines starting with "gitolite-options" are always allowed; you dont need to set GIT_CONFIG_KEYS in the rc file (3) they don't go into the "bare-repo.git/config" file | 07:54 |
| ← dhruvasagar left | 07:54 |
|
EugeneKay
| Indeed. | 07:54 |
|
| Yay, seems to work right | 07:54 |
|
| EugeneKay sleeps, plots world domination, and scratches his belly | 07:55 |
| ← pantsman left | 07:55 |
| → dhruvasagar joined | 07:56 |
| ← maletor left | 07:56 |
| ← dpino left | 07:56 |
|
EugeneKay
| Maybe I'll finally make my git-library script work tomorrow | 07:57 |
| → LimeBlast joined | 07:57 |
|
jast
| does it vend yet? | 07:57 |
| ← Guest83383 left | 07:57 |
|
EugeneKay
| Nope, but the IRC stats do | 07:57 |
|
| !stats | 07:57 |
|
gitinfo
| [!irclog] Public logs of #git are kept at: http://colabti.org/irclogger/irclogger_log/git or look at stats for the last 30 days: http://itvends.com/irc/git.html | 07:57 |
| → eijk joined | 07:58 |
|
EugeneKay
| Bonus points: git config --global alias.vend '!echo "IT VENDS! "$(curl http://itvends.com/vend 2>/dev/null)' | 08:00 |
| ← Bosox20051 left | 08:01 |
| → j0ran1 joined | 08:01 |
| → mikecmpbll joined | 08:01 |
| → Gx4xx joined | 08:02 |
| ← j0ran left | 08:02 |
| → suy joined | 08:02 |
| → florin joined | 08:03 |
| florin → Guest83122 | 08:04 |
|
osse
| What's the best way of generating the list of files added by a commit? | 08:04 |
| → gerryvdm_mbp joined | 08:04 |
| → _Qman joined | 08:04 |
|
osse
| ahh, in my case t he commit had no other changes than new files so git diff HEAD~ --name-only works | 08:04 |
|
thiago
| git show --name-only, git log --name-only | 08:05 |
| ← Gx4 left | 08:05 |
|
osse
| even better | 08:06 |
| → abizern joined | 08:06 |
|
jast
| osse: --name-status --diff-filter=A | 08:06 |
|
| I suppose the first one is optional | 08:06 |
|
osse
| jast: --diff-filter! Brilliant! | 08:07 |
| → Axsuul joined | 08:10 |
|
Axsuul
| A gitignore line of... /kitchen/**/metadata.json should ignore something like... /kitchen/cookbooks/foo/metadata.json right? | 08:10 |
| → LeMike joined | 08:10 |
|
thiago
| Axsuul: gitignore doesn't handle ** specially | 08:11 |
| ← ivan\ left | 08:11 |
|
Axsuul
| thiago: how would I accomplish that ? | 08:12 |
| → angelsl joined | 08:13 |
|
thiago
| Axsuul: no support for that | 08:13 |
| → Tegi joined | 08:13 |
| ← tdlguiinstall left | 08:14 |
|
jast
| you could put a .gitignore file in kitchen/ that has 'metadata.json' in it | 08:14 |
| → afuentes joined | 08:14 |
|
thiago
| or that, yeah | 08:15 |
| ← boombatower left | 08:17 |
| ← maleknet left | 08:18 |
|
Axsuul
| jast: thanks | 08:18 |
| → solofight joined | 08:18 |
| ← solofight left | 08:19 |
| → arietis joined | 08:19 |
| → Element9 joined | 08:20 |
| ← quazimodo left | 08:21 |
| → Jippi joined | 08:26 |
| ← Eridius left | 08:29 |
| → quazimodo joined | 08:29 |
| → Sander^work joined | 08:30 |
| ← avpx left | 08:30 |
| → solofight joined | 08:31 |
| ← Chib_ left | 08:31 |
| → maleknet joined | 08:31 |
| → Vile joined | 08:32 |
| → reshtnk7 joined | 08:36 |
| → notVert joined | 08:38 |
| ← jaseemabid left | 08:38 |
| ← gerryvdm_mbp left | 08:38 |
|
charon
| osse: but to look at the "change done by a commit" in general, not just HEAD, you'd use 'git show <the same options>' | 08:40 |
| → marhaban joined | 08:41 |
| ← maleknet left | 08:41 |
| → polymar joined | 08:41 |
| → maleknet joined | 08:44 |
| → Ivo joined | 08:46 |
| ← thiago left | 08:46 |
| → tvw joined | 08:46 |
| → thiago joined | 08:47 |
| → raboof joined | 08:47 |
| → gerryvdm_mbp joined | 08:48 |
| ← gerryvdm_mbp left | 08:48 |
| → gerryvdm_mbp joined | 08:48 |
| ← raboof left | 08:48 |
| → raboof joined | 08:48 |
|
raboof
| i seem to have confused git-svn (by letting my disk fill up :S) | 08:48 |
| → tarkus joined | 08:48 |
|
raboof
| 'git svn dcommit' now gives 'RA layer file already exists: File <somefile>' | 08:48 |
|
| now what? :) | 08:48 |
| → jjang joined | 08:48 |
| ← jjang left | 08:49 |
|
raboof
| indeed the commit that adds this file has not been pushed to SVN yet | 08:49 |
| ← chisler left | 08:49 |
|
tarkus
| Q. When I do 'git push origin master' does it copies all the repos from local repo to the remote one or just a master? | 08:50 |
| ← gerryvdm_mbp left | 08:50 |
|
thiago
| tarkus: it pushes the contents of the master branch to the remote master branch, that's all | 08:51 |
| ← abizern left | 08:51 |
| → Gorkyman joined | 08:52 |
|
tarkus
| Q. Let's say i have multiple small commits in a dev branch (dedicated to a bug fix or a small feature). Is it ok to merge it into the main branch as a single commit to keep the history clean and simple in the main branch? Or this is something nobody do? | 08:53 |
|
thiago
| it's ok | 08:53 |
| ← rurufufuss left | 08:53 |
|
thiago
| it's usually project policy or personal taste | 08:53 |
| ← beneggett left | 08:54 |
|
thiago
| some projects insist on small, almost atomic changes | 08:54 |
|
| other projects will accept a moderately large change that makes sense together | 08:54 |
|
tarkus
| thiago, which command should i look into to do this single merged comit? | 08:54 |
|
jast
| you'd better not do it if you want to merge from dev branch to master again later on, though | 08:54 |
|
thiago
| most don't like one big change doing many unrelated things | 08:54 |
| → und3f joined | 08:54 |
|
jast
| because a squashed merge (git merge --squash) sort of breaks future normal merges | 08:54 |
|
thiago
| my recommendation, rebase -i the dev branch on top of master and squash the commits | 08:55 |
|
jast
| I don't see a pratical difference | 08:55 |
|
tarkus
| for example you're working on an open source project, and want to have fewer commits in the public branch than in your local dev branches.. sounds normal? | 08:56 |
|
thiago
| no | 08:56 |
| → swair joined | 08:56 |
|
thiago
| once you squash, you get rid of the smaller commits | 08:56 |
|
| you don't keep them around | 08:56 |
|
| so remember to squash them in a form that makes sense for later | 08:56 |
|
jast
| in an open source project, I don't want *few* commits, I want *sensible* commits. sometimes that's more, sometimes that's less. | 08:56 |
|
thiago
| agreed | 08:56 |
|
jast
| in any other project, to,o really | 08:56 |
| ← polymar left | 08:57 |
|
masak
| it's like in SRP: it's not about making things as small as possible, it's about making things that belong together go together. | 08:58 |
|
tarkus
| how do you keep unfinished work during multi-tasking.. by using staging? | 08:59 |
|
| instead of small commits? | 08:59 |
|
thiago
| some people keep different branches | 08:59 |
|
| just make a work-in-progress commit and then change branches | 08:59 |
|
| I do that, except that everything is in one branch | 09:00 |
|
masak
| staging and different branches are good for work-in-progress things. which one I use depends on the size/lifetime of the WIP. | 09:00 |
| → blami_ joined | 09:00 |
| → pfFredd joined | 09:00 |
|
tarkus
| in general is it good practice to have a single commit for a small bug fix are feature implementation as opposed to commiting several times while you work on the feature or a bug fix? | 09:01 |
|
jast
| if changes are isolated enough, I just use 'git add -p' or 'git commit -p' | 09:01 |
| ← osse left | 09:02 |
| ← hauron left | 09:02 |
| ← dareign85 left | 09:03 |
| ← Gorkyman left | 09:03 |
| → Gorkyman joined | 09:03 |
|
Gorkyman
| hey guys... I need some help... somehowmy git is saying that I'm up-to-date, but remote develop clearly has 2 more commits | 09:03 |
|
| so... how to get out of this mess? :) | 09:03 |
|
Bombe
| Merge those, if that’s what you want to do. | 09:04 |
|
masak
| heh -- I remember being in such situations early on with git. | 09:04 |
| → hinst joined | 09:04 |
|
tarkus
| thiago which command should i look into fro work-in-progress commits? | 09:04 |
|
Gorkyman
| problem is, I dont even know how to get those commits from remote, since its saying that its up-todate | 09:04 |
|
thiago
| Gorkyman: don't doubt Git. You're probably up to date. | 09:04 |
|
| Gorkyman: how do you know that the upstream has those commits? | 09:05 |
|
| tarkus: git commit | 09:05 |
| ← Esa left | 09:05 |
| → worstadmin joined | 09:06 |
| SeoZ → SeoZ-work[AWAY] | 09:06 |
|
tarkus
| thiago, git commit in a working branch, then git commit --ammed on top of it? then after feature is done merge into the main branch as a single commit? | 09:06 |
|
thiago
| tarkus: if you're working on a feature and don't have time to finish it now, git commit and write a message indicating you're not done. Like "WIP" | 09:07 |
|
| tarkus: change branches, work on what you need to work on, etc. | 09:07 |
| → gdmfsob joined | 09:07 |
|
thiago
| when you get back to that branch, you can continue working on and --amend, or you can reset to a previous commit and re-commit | 09:07 |
|
Gorkyman
| thiago: hm... yeah it seems that git is right... | 09:08 |
|
tarkus
| thanks guys | 09:09 |
| ← mishok13 left | 09:10 |
|
tarkus
| is it ok to have private and public repos for open source projects? in a private repo i tend to have more branches for experimental features and stuff like that and in the public repo there is clean, good code available for public consumption? | 09:12 |
| → btree_ joined | 09:13 |
| ← luc_ left | 09:13 |
| ← j0ran1 left | 09:13 |
| → luc_ joined | 09:14 |
| ← thiago left | 09:15 |
| ← btree left | 09:16 |
|
dr_lepper
| tarkus: well, sure, why not? | 09:17 |
|
tarkus
| i'm curious is that something enybody is doing.. or just a overcomplication | 09:17 |
|
dr_lepper
| with git everyone has his own private repository | 09:18 |
| ← quazimodo left | 09:18 |
|
dr_lepper
| millions of people have cloned Linux repository, for example | 09:18 |
|
tarkus
| i wish git has an ability to set permissions on repos level.. | 09:18 |
|
dr_lepper
| each of those clones is a private repository, who knows what they are doing with it? :) | 09:18 |
|
| what do you mean "on repos level"? | 09:18 |
|
tarkus
| *branch level.. | 09:19 |
| ← arietis left | 09:19 |
|
tarkus
| for example i could have a public repo with dev branch restricted to developers onlhy | 09:19 |
| → Esa_ joined | 09:20 |
| gitinfo set mode: +v | 09:20 |
|
OMGOMG_
| how to run a three way difftool between the working directory and (e.g.) HEAD and HEAD^? | 09:20 |
| → ivan\ joined | 09:20 |
|
tarkus
| in tfs for example you can set different permissions on different areas of your repo if needed | 09:21 |
|
dr_lepper
| tarkus: git does not deal with access control | 09:22 |
|
| tarkus: you want probably something like gitolite | 09:22 |
| ← xxc_michael left | 09:22 |
|
FauxFaux
| tarkus: !gitolite can do per /branch/ access control; nothing can do per-tree access control because that makes very little sense in git. | 09:22 |
|
gitinfo
| tarkus: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html | 09:22 |
|
charon
| raboof: in the worst case you can always make a fresh git-svn clone, fetch the other repo into it as a remote, use cherry-pick or rebase to "copy" the commits that are not in svn yet, and then dcommit again | 09:23 |
|
tarkus
| thanks, i'll look into it | 09:23 |
|
charon
| raboof: but i suspect if you nuke (!backup) the .git/svn cache data, and run 'git svn fetch' to regenerate it, the problem will vanish | 09:23 |
|
dr_lepper
| FauxFaux: doesn't gitolite do per-directory access control? | 09:23 |
|
tarkus
| how do you do scrum with OSS git/github projects? by using thoughtworks' mingle or is there a good alternative? | 09:24 |
|
sitaram
| dr_lepper: it can do that also; branch level is more commonly used | 09:24 |
|
FauxFaux
| Oh, for writes. | 09:24 |
|
sitaram
| dr_lepper: unless you're an SVN refugee, in which case both probably sound the same to you ;-) | 09:24 |
|
dr_lepper
| sitaram: yeah, that's what I thought | 09:24 |
| ← Esa_ left | 09:24 |
| → quazimodo joined | 09:24 |
|
dr_lepper
| hehe | 09:25 |
|
sitaram
| and as FauxFaux said all this is only for writes | 09:25 |
|
dr_lepper
| well, yeah, obviously | 09:25 |
| ← mhauri left | 09:27 |
| → FreeHaegin joined | 09:27 |
| → ryez joined | 09:27 |
|
FreeHaegin
| hey, I've heard rebase is bad if the commits you're rebasing have been pushed, is this still true if nobody has done any work on the pushed commits? | 09:27 |
| ← ryez left | 09:28 |
| → tusbar joined | 09:28 |
| → Lord_DeathMatch joined | 09:28 |
|
jast
| FreeHaegin: the question is more whether they *have* the pushed commits | 09:28 |
|
dr_lepper
| FreeHaegin: !rewriting_public_history | 09:29 |
|
gitinfo
| FreeHaegin: Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to `git pull --rebase` and even worse things if they have tagged or branched, so you must publish your humiliation so they know what to do. You will need to `git push -f` to force the push. The server may not allow this. See receive.denyNonFastForwards (git-config) | 09:29 |
|
jast
| hmm, that trigger could use a link to a helpful document or something | 09:29 |
| → i42n joined | 09:29 |
|
tusbar
| is it not possible to remove a directory and add a submodule with the same path/name in the same commit? | 09:29 |
|
charon
| jast: like "coping with upstream rebase" in man git-rebase? | 09:29 |
|
gitinfo
| jast: the git-rebase manpage is available at http://git-scm.com/docs/git-rebase.html | 09:29 |
|
charon
| eh, "recovering from upstream rebase" | 09:30 |
| → mhauri joined | 09:30 |
|
jast
| I suppose | 09:30 |
|
| does that file have anchors? | 09:30 |
|
FreeHaegin
| ok, basically the company policy is push to github so you don't lose your work but I'm 99% sure nobody else has pulled it down. will the server still deny my push? | 09:30 |
|
jast
| it does | 09:31 |
|
| FreeHaegin: by default, yes. with --force... one way to find out | 09:31 |
| → arietis joined | 09:31 |
| → gerryvdm_mbp joined | 09:31 |
| ← gerryvdm_mbp left | 09:31 |
| → gerryvdm_mbp joined | 09:31 |
| ← mhauri left | 09:31 |
|
FreeHaegin
| nah, I'm not forcing a push. that might make people hate on me. I'll just merge it in and wrangle it that way | 09:31 |
|
| thanks for the advice guys | 09:31 |
|
jast
| I've finally rewritten !rewrite. I've been finding it weird for a long time... | 09:35 |
|
gitinfo
| [!rewriting_public_history] Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to jump through hoops (and you have to tell them to), so it's infinitely better to just move on without rewriting. That said, you can use `git push -f` to force rewriting (the server may reject it, though). See http://git-scm.com/docs/git-rebase.html#_recovering_from_upstream_rebase | 09:35 |
| → jernej joined | 09:36 |
| ← bloony left | 09:38 |
| → Esa_ joined | 09:39 |
| → bloony joined | 09:39 |
| → simesy_ joined | 09:40 |
| ← maleknet left | 09:40 |
| → chisler joined | 09:40 |
| → psoo joined | 09:41 |
| → chx joined | 09:41 |
|
chx
| how can i tell git to throw away a directory and all files in it and get the contents from another brnach? | 09:42 |
|
jernej
| quick question. does commit object points to one tree with multiple trees or to just multiple trees directly ? | 09:42 |
|
FauxFaux
| jernej: git cat-file commit HEAD | 09:42 |
| → nadim joined | 09:43 |
| → tfitts joined | 09:44 |
| → thierryp joined | 09:46 |
|
jast
| chx: rm -rf directory (you know that's dangerous, right?); git checkout otherbranch -- directory | 09:47 |
|
jernej
| FauxFaux, ? | 09:47 |
|
jast
| jernej: just do it and you'll see what's in a commit object | 09:47 |
| → diegoviola joined | 09:47 |
| ← simesy_ left | 09:47 |
|
jernej
| h | 09:47 |
|
| ok* | 09:47 |
|
jast
| in a format for consumption by humans :) | 09:47 |
|
jernej
| i see, thanks :) | 09:48 |
| → devotee joined | 09:49 |
| → this_is_caitlin joined | 09:50 |
| ← thierryp left | 09:51 |
| ← jackh left | 09:54 |
| gitinfo set mode: +v | 09:55 |
| ← caseymcg left | 09:55 |
| ← gavin_huang left | 09:56 |
|
nadim
| hi, is is possible, through hooks, to change things like $version$ to the commit within a file? I am not arguing that it is a good thing to . URL to some documentation highly appreciated | 09:56 |
| → tdlguiinstall joined | 09:56 |
| ← psoo left | 09:57 |
| ← jernej left | 09:57 |
| → thiago joined | 09:57 |
| → beautifulmind joined | 09:57 |
|
wereHamster
| nadim: there are better ways to embed the version into your source code than using $version$ | 09:57 |
|
nadim
| wereHamster: I wrote "I am npot arguing that it is a good thing" specifically to not get that answer, thank you anyway | 09:58 |
|
mtlife_w
| Where can I find some error log if somebody is trying to login? There is no /var/log/git or something | 09:58 |
| → Melrian joined | 09:58 |
|
FauxFaux
| mtlife_w: Log-in to what? git doesn't have logins. | 09:58 |
|
wereHamster
| nadim: man gitattributes -> ident. But that's the blob id, not the commit id | 09:59 |
|
gitinfo
| nadim: the gitattributes manpage is available at http://git-scm.com/docs/gitattributes.html | 09:59 |
|
FauxFaux
| nadim: Go google people abusing !smudge , if you have more complex requirements. | 09:59 |
|
gitinfo
| nadim: [!filters] Filters can be very useful for automatically making changes to files at checkout(smudge) and checkin/add(clean). See 'man gitattributes' for more information | 09:59 |
| → sbodo joined | 09:59 |
|
wereHamster
| nadim: alternatively, use smudge/clean filters | 09:59 |
|
mtlife_w
| FauxFaux: Well I know, the private/pub key seems to be working I think. But somebody is getting 'Exception caught during execution of ls-remote command' | 09:59 |
|
nadim
| thank you guys | 09:59 |
|
FauxFaux
| mtlife_w: That sounds like you're not even remotely using git for the server. gerrit, I guess? | 09:59 |
|
mtlife_w
| EGit for eclipse yes, gerrit is a part of that right? | 10:00 |
|
FauxFaux
| You are one very confused puppy. | 10:01 |
|
mtlife_w
| Yes :( | 10:01 |
| ← mikecmpbll left | 10:01 |
|
FauxFaux
| a) work out what client you're using. b) work out what server you're using. c) work out what transport you're using. d) work out which is erroring. e) direct the real error message to the right people. | 10:01 |
| → maleknet joined | 10:02 |
|
mtlife_w
| a) Egit for eclipse. b) ubuntu 12.04 with gitolite. c) SSH transport d) No idea? e) That's it. | 10:02 |
| ← nadim left | 10:03 |
| → mikecmpbll joined | 10:03 |
|
FauxFaux
| Sounds like egit is broken. WIndow -> show view -> error log, get the real exception? And, if it's not obvious or googlable, direct it to #eclipse? | 10:04 |
| → serbang joined | 10:05 |
| ← berserkr left | 10:07 |
| ← chx left | 10:07 |
| ← serbang left | 10:09 |
| ← maleknet left | 10:09 |
| ← hyperair left | 10:11 |
| ← Chuwiey left | 10:13 |
| ← drougge left | 10:13 |
| → frsk joined | 10:14 |
| → drougge joined | 10:14 |
| ← ph^ left | 10:16 |
| → gavin_huang joined | 10:16 |
| ← Guest83122 left | 10:16 |
|
mtlife_w
| FauxFaux I will check. Moment please :) | 10:17 |
| ← gerryvdm_mbp left | 10:17 |
| → iocor joined | 10:17 |
| → fbroasca joined | 10:18 |
| → madsy joined | 10:18 |
| ← madsy left | 10:18 |
| → madsy joined | 10:18 |
| ← Melrian left | 10:19 |
| → Melrian joined | 10:19 |
| mtlife_w → mtlife | 10:20 |
| → bfgun joined | 10:21 |
| ← drl left | 10:24 |
| → FrenkyNet joined | 10:24 |
| → ph^ joined | 10:25 |
| ← bigfg left | 10:25 |
| ← flaguy48 left | 10:27 |
|
FrenkyNet
| Hi all, does anyone have an easy guide to installing git on a dedicated server? Secure of course. I'm no sysadmin btw, just a code-monkey. | 10:29 |
| → XATRIX joined | 10:29 |
| → osse joined | 10:29 |
| → FasTTo joined | 10:29 |
| ← FasTTo left | 10:29 |
|
XATRIX
| what do i have to do on the server side to be able to "gitosis@git.yourserver.com:yourproject.git" ? istead of SSHing it | 10:30 |
| → j0ran joined | 10:30 |
| → maleknet joined | 10:30 |
|
canton7
| FrenkyNet, just git? use your favourite package manager | 10:31 |
| ← tdlguiinstall left | 10:31 |
|
canton7
| FrenkyNet, a git server? Take a look at !gitolite | 10:32 |
|
gitinfo
| FrenkyNet: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html | 10:32 |
|
canton7
| XATRIX, that is ssh...? | 10:32 |
|
FrenkyNet
| thanks canton7 gitinfo | 10:32 |
|
gitinfo
| FrenkyNet: you're welcome, but please note that I'm a bot. I'm not programmed to care. | 10:32 |
|
FrenkyNet
| wow | 10:32 |
|
| bitchin' | 10:33 |
| → ignotus joined | 10:34 |
| ← RaceCondition left | 10:34 |
| ← Leo2 left | 10:35 |
| → intripoon joined | 10:37 |
| → liktoj joined | 10:38 |
| ← sgambino left | 10:42 |
| → medwards joined | 10:42 |
| → sgambino joined | 10:42 |
|
medwards
| Is there a way to tell if commits have been rearranged in an interactive rebase? (I want to do a TDD workshop and see if someone rearranged his history to put their test first) | 10:43 |
| → gerryvdm_mbp joined | 10:43 |
| ← gerryvdm_mbp left | 10:43 |
| → gerryvdm_mbp joined | 10:43 |
| ← Pupeno_W left | 10:44 |
|
medwards
| I'm thinking its possible that the timestamp is from the original commit time... | 10:44 |
|
wereHamster
| medwards: you could look at the author time | 10:45 |
|
| medwards: but that can be also changed, so if the commits are test first, you can not tell if the author actually wrote the tests first | 10:45 |
| ← luc_ left | 10:45 |
| ← shruggar left | 10:46 |
| ← ignotus left | 10:46 |
|
solofight
| people - whats the difference between a branch and a clone from a use case perspective ? say i have A and i clone it to B and i make changes B+x and later i want changes to be reflected to A as well so i do a push. Now take the same with a branch i have A, branched to B and some changes added B+x and when needed to be integrated with A its a branch merge ! ! | 10:47 |
| → Droolio joined | 10:47 |
|
solofight
| kindly throw some light if am missing something | 10:47 |
|
wereHamster
| solofight: a repository contains branches. | 10:47 |
|
medwards
| wereHamster: thats good enough. Someone goes to enough trouble to create an empty commit or wipe the timestamp then they're working too hard not to learn for me to care. | 10:48 |
|
| thanks! | 10:48 |
| → Cros79 joined | 10:48 |
|
solofight
| wereHamster: so what difference does it make in the above said workflow ? | 10:48 |
|
wereHamster
| medwards: it's not that hard: https://gist.github.com/3344208 | 10:49 |
|
solofight
| if i want to go one step further with branches also i can create right ? B+x branched to B+x+y and then merged when needed | 10:49 |
|
| wereHamster: ^ | 10:49 |
| → mhauri joined | 10:50 |
|
wereHamster
| solofight: man gitglossary and read the explanation of 'branch' and 'repository' | 10:51 |
|
gitinfo
| solofight: the gitglossary manpage is available at http://git-scm.com/docs/gitglossary.html | 10:51 |
|
medwards
| wereHamster: few of our guys are that comfortable with shell, and if they're reading git man pages to know about --reset-author then thats still useful to me. But maybe I should keep this around for myself >.> | 10:51 |
| → ncd|leeN joined | 10:51 |
| → shruggar joined | 10:52 |
| → juvenal joined | 10:53 |
| ← leeN left | 10:55 |
| ← sgambino left | 10:57 |
| ← XATRIX left | 10:58 |
| ← this_is_caitlin left | 10:58 |
| → ipalaus joined | 11:01 |
| ← Anubisss left | 11:02 |
| ← jdyin left | 11:02 |
| → Anubisss joined | 11:02 |
| ← Anubisss left | 11:03 |
| → Anubisss joined | 11:04 |
| → danielpunt joined | 11:04 |
| → maxandersen1 joined | 11:04 |
| ← gxk left | 11:05 |
| ← bartman` left | 11:05 |
| ← maxandersen left | 11:06 |
| → bartman` joined | 11:06 |
| ← acrocity left | 11:07 |
| → zenx joined | 11:07 |
| → OOPMan joined | 11:08 |
| → acrocity joined | 11:08 |
| → dareign85 joined | 11:09 |
| ← adamben left | 11:10 |
| → RaceCondition joined | 11:10 |
| ← Fullmoon left | 11:12 |
| ← RJ3000_ left | 11:13 |
| ← tusbar left | 11:13 |
|
arex\
| Am I able to merge changes from "features/someFeature" branch in "green" repo, to "develop" branch in "gate" repo? | 11:14 |
|
canton7
| sure | 11:14 |
| ← LeMike left | 11:15 |
| ← Esa_ left | 11:15 |
| → Esa_ joined | 11:16 |
| → HavenMonkey joined | 11:16 |
| ← SJr left | 11:16 |
|
arex\
| canton7: ok, great. currently thinking through how the workflow would be with feature branching, ci and two repos ("gate" and "green") | 11:17 |
| → tjadc joined | 11:17 |
| ← TheHaven left | 11:17 |
| → SJrX joined | 11:18 |
| ← Esa_ left | 11:20 |
| → RJ3000_ joined | 11:21 |
| ← drougge left | 11:22 |
| → drougge joined | 11:22 |
| → Esa_ joined | 11:22 |
| ← letharion left | 11:23 |
|
zenx
| Hi can someone explain me the difference between git cloning the ubuntu linux.git repo. I am inexperienced with git and istead of cloning ubuntu-quantal.git I cloned the former thinking it would have everything. But now I am trying to checkout a head in the ubuntu-quantal.git repo and can't do it | 11:24 |
|
| this relating to the ubuntu kernel tree | 11:24 |
|
cbreak-work
| zenx: different repositories can have different histories | 11:26 |
| ← hellopat left | 11:26 |
|
cbreak-work
| chances are they have nothing in common | 11:26 |
| → stodan joined | 11:26 |
| → adamben joined | 11:26 |
|
masak
| general query: when do you people tend to do interactive rebasing? I know when I do it myself, but I'm curious about other people's use cases. | 11:27 |
| ← RJ3000_ left | 11:27 |
|
OMGOMG_
| <+OMGOMG_> how to run a three way difftool between the working directory and (e.g.) HEAD and HEAD^? | 11:27 |
|
| thoughts? | 11:27 |
|
canton7
| masak, normally when tidying up a feature branch before merging it into some other, 'main' branch | 11:27 |
|
zenx
| cbreak-work: Thanks. I am cloning the ubuntu-quantal one now | 11:27 |
|
canton7
| masak, also when I make a stupid mistake in some commit I haven't yet pushed, in any branch | 11:28 |
|
OMGOMG_
| masak: yeah, that | 11:28 |
| → johnny_rose joined | 11:29 |
| ← sbodo left | 11:29 |
| ← chisler left | 11:30 |
|
masak
| ok. tidying up known-untidy commits to make a better history before publishing/merging it. | 11:30 |
|
| or fixing a mistake somewhere much earlier in unpublished history. sort of like 'commit --amend' but further back. | 11:31 |
|
_ikke_
| masak: yeah | 11:31 |
|
| But you can even remove entire commits | 11:31 |
| ← danielpunt left | 11:31 |
| → sgambino joined | 11:32 |
|
masak
| right. | 11:32 |
| → jaseemabid joined | 11:33 |
| ← dhruvasagar left | 11:34 |
| → tellone joined | 11:34 |
| → rgsteele joined | 11:34 |
|
canton7
| tbh, I mainly use 'git commit --fixup' to create fixup commits, then apply them with 'git rebase -i' just before merging | 11:34 |
| → LeMike joined | 11:34 |
| → bodom joined | 11:35 |
| → letharion joined | 11:35 |
| ← solofight left | 11:35 |
| → pandeiro joined | 11:36 |
| ← BadDesign left | 11:37 |
| → polymar joined | 11:37 |
|
tarkus
| is there a way to mark a remote as read-only, so i woudn't push something there accidently | 11:37 |
|
canton7
| tarkus, something like git config remote.origin.pushurl "Don't push here your oaf" | 11:39 |
|
| s/your/you | 11:39 |
| ← i42n left | 11:41 |
| → mmc1 joined | 11:43 |
| ← und3f left | 11:45 |
| → BadDesign joined | 11:45 |
| → cannonball joined | 11:45 |
| → Chuwiey joined | 11:45 |
| → logik-bomb joined | 11:46 |
| ← adamben left | 11:47 |
| → thierryp joined | 11:47 |
| → Tegi_ joined | 11:47 |
| → donri joined | 11:48 |
| ← _iori_ left | 11:48 |
| ← Tegi left | 11:49 |
|
logik-bomb
| Say I created a local/branch from my remote/master. That master got changed several times. How do I update my local/branch? Do I cut a new one and merge both (brach+master) or do I merge master into my branch? | 11:49 |
| → ahferroin7 joined | 11:49 |
|
Sajbar
| logik-bomb: fetch from remote and merge master into your branch | 11:50 |
|
canton7
| you can update your local master from origin/master and then merge your local master into your local branch, or just merge origin/master into your local branch | 11:50 |
| ← dr_lepper left | 11:51 |
|
logik-bomb
| will it create any problems when merging that branch into local/master ? | 11:51 |
|
| once the work is finished.. | 11:51 |
|
| ? | 11:51 |
|
canton7
| nope | 11:51 |
| ← thierryp left | 11:52 |
| ← marhaban left | 11:52 |
|
logik-bomb
| thanks for your help guys | 11:52 |
| ← ehsan left | 11:54 |
| → aruiz joined | 11:55 |
| ← mjau^ left | 11:58 |
| → dr_lepper joined | 11:58 |
| ← dr_lepper left | 11:58 |
| → dr_lepper joined | 11:59 |
| ← Lord_DeathMatch left | 12:00 |
| → und3f joined | 12:01 |
| ← quazimodo left | 12:01 |
| ← chimay left | 12:02 |
| → Lord_DeathMatch joined | 12:02 |
| ← mikecmpbll left | 12:03 |
| → mikecmpbll joined | 12:04 |
| ← keller left | 12:05 |
| ← LeMike left | 12:06 |
|
Sander^work
| How can I track changes to a mysql database with git? | 12:06 |
| → nwest joined | 12:06 |
|
arex\
| http://i.imgur.com/H8zbT.png | 12:07 |
|
Sander^work
| Could I just do a dump every day, and add it automaticly to the repository? | 12:07 |
|
cmn
| very badly | 12:07 |
|
| if you want backups, use a backup software | 12:07 |
|
| if you need versioning in your data, use the database for that | 12:07 |
|
Sander^work
| Ok.. What I really want is to track development of changes to the structure | 12:08 |
|
cmn
| that's something completely different | 12:08 |
|
| you can track the schema, but you'll also need to track the migrations | 12:09 |
| → alextheger joined | 12:09 |
| ← Lord_DeathMatch left | 12:10 |
| → Lord_DeathMatch joined | 12:10 |
| ← alextheger left | 12:11 |
| ← alunduil left | 12:11 |
| ← domguard left | 12:12 |
| → geekbri joined | 12:12 |
| ← kcj left | 12:13 |
| ← deffrag left | 12:13 |
| ← sangi left | 12:13 |
| ← marthinal left | 12:17 |
| → glennpratt joined | 12:17 |
| → arkiver joined | 12:18 |
| → flippo joined | 12:19 |
|
antihero
| CRAP | 12:19 |
|
| how do I completely remove all traces of a file from a rep | 12:20 |
|
| I accidentally checked in a file with passwords in it | 12:20 |
|
Sajbar
| branch-filter | 12:20 |
|
CareBear\
| man git filter-branch | 12:20 |
|
gitinfo
| the git-filter-branch manpage is available at http://git-scm.com/docs/git-filter-branch.html | 12:20 |
| → jcao219 joined | 12:21 |
| → LeMike joined | 12:21 |
| ← totte left | 12:23 |
| → GFLJay joined | 12:23 |
| ← jargon- left | 12:24 |
| ← Lord_DeathMatch left | 12:24 |
| → skyf joined | 12:24 |
| → bartek joined | 12:25 |
| → keller joined | 12:27 |
| → RJ3000_ joined | 12:27 |
| → pretty_f_ joined | 12:28 |
| ← pretty_function left | 12:28 |
| → TheJH joined | 12:28 |
| → gusnan joined | 12:29 |
| → TeckniX joined | 12:29 |
| ← juvenal left | 12:31 |
| → cakehero joined | 12:35 |
| ← jcao219 left | 12:36 |
| → mgaunard joined | 12:36 |
|
mgaunard
| how can I find how old a branch is? | 12:36 |
|
arex\
| Anyone doing feature branching + continuous integration? | 12:38 |
|
cmn
| mgaunard: you start by defining what age means | 12:38 |
|
mgaunard
| arex\: not really, I'm setting up the tester by hand whenever there is a new interesting branch to test, but I could be interested in automated solution | 12:39 |
|
| cmn: I'd like to know the date a remote branch was first created. | 12:39 |
|
cmn
| there is no way to know for sure | 12:39 |
|
| partly because "created" can also have different meanings | 12:40 |
|
| do you mean when the root commit was created, when it first diverged from its base branch, when it was first uploaded to the server... | 12:40 |
|
mgaunard
| either would do | 12:40 |
| ← zhouxiukun left | 12:41 |
|
cmn
| those can be very different times | 12:41 |
|
mgaunard
| not that different in my case | 12:42 |
| ← mikecmpbll left | 12:42 |
| → marhaban joined | 12:42 |
|
cmn
| git log --format=%aD branch | head -n 1 should show you the date of the root commit | 12:42 |
|
mgaunard
| but I'd be happy to have to commands to find out both | 12:42 |
|
cmn
| both of what? I gave three different dates | 12:42 |
|
mgaunard
| the above command gives the date of the last commit. | 12:43 |
|
cmn
| --reverse | 12:43 |
|
mgaunard
| tail -n 1 works too | 12:43 |
|
cmn
| sure | 12:44 |
| ← BadDesign left | 12:44 |
|
mgaunard
| this seems to be the root commit of the whole repository, not the first commit on the branch | 12:44 |
|
cmn
| there is no such thing | 12:45 |
|
| and if it sounds like you don't know what a branch is | 12:45 |
| → Lord_DeathMatch joined | 12:45 |
|
mgaunard
| what you call "when it first diverged" I guess | 12:45 |
|
cmn
| for that you'd need to parse --graph or the parents | 12:45 |
|
| well, if you know the base branch, you can just hide it with ^basebranch | 12:46 |
|
| if you don't, you'll need to figure it out by parsing the output of log | 12:47 |
| ← j0ran left | 12:47 |
| → anddam joined | 12:47 |
| ← LeMike left | 12:47 |
|
mjt
| there's also git merge-base | 12:48 |
|
cmn
| which you then have to figure out | 12:48 |
|
mgaunard
| I know the base branch | 12:48 |
|
anddam
| I'm using 'git remote set-url origin my_new_url my_old_url' to change my fetch url (from https to ssh) but I get "warning: remote.origin.url has multiple values" instead | 12:49 |
|
| and Fetch URL value doesn't change on origin, what am I doing wrong? | 12:49 |
|
| I'm following syntax, from what I can see | 12:49 |
|
cmn
| you did | 12:49 |
|
| it's not this command | 12:49 |
|
| at some point, you ran some command that transformed the url into a multivar | 12:50 |
|
mgaunard
| cmn: how do I hide with ^basebranch? git log branch^basebranch doesn't work, I don't really know the syntax | 12:50 |
| ← maleknet left | 12:50 |
|
cmn
| mgaunard: see man gitrevisions | 12:50 |
|
gitinfo
| mgaunard: the gitrevisions manpage is available at http://git-scm.com/docs/gitrevisions.html | 12:50 |
|
mgaunard
| cmn: wouldn't it be simpler to just tell me if you know the right command? | 12:51 |
| ← RJ3000_ left | 12:51 |
|
cmn
| no | 12:51 |
| → mikecmpbll joined | 12:52 |
|
mgaunard
| well I don't see how the manual helps, it's only possible to specify one revision to git log, not multiple ones | 12:52 |
|
| so I don't know how your filtering is supposed to work | 12:52 |
|
anddam
| cmn: basically I cloned a repo, then forked it on github so I changed my fetch and push url to my fork using https scheme, now I realize I do want ssh scheme | 12:52 |
|
mgaunard
| or maybe you meant specifying a starting and end revision, I see | 12:53 |
|
| not what I'd call filtering | 12:53 |
|
anddam
| cmn: so possibly in the first change I inadvertently changed the url to multivar | 12:54 |
| → quazimodo joined | 12:55 |
|
anddam
| cmn: I see in .git/config there are multiple entries for url key in remote origin stanza, I can delete them from here but how would I deal with such an issue from the cli? | 12:55 |
|
| just for the sake of knowledge | 12:55 |
|
| nvm | 12:56 |
|
| TIAS git remote set-url origin --delete | 12:57 |
| ← Silex left | 12:57 |
| → mikepack joined | 12:57 |
|
anddam
| now 'git remote show origin' shows the URLs I want | 12:57 |
|
| thanks | 12:57 |
| ← anddam left | 12:57 |
| → sontek joined | 12:57 |
| → Silex joined | 12:57 |
| ← Silex left | 12:57 |
|
sontek
| if I tried to merge, it had conflicts, and I started trying to merge and realized it was going to be more work than it was worth, is there a way to just reset to what the other user submitted? | 12:58 |
|
| and then I'll just re-add my changes manually | 12:58 |
|
cmn
| you wouldn't call removing uninteresting data from the output filtering? | 12:58 |
| ← madsy left | 12:59 |
| → Silex joined | 12:59 |
| ← bodom left | 13:01 |
|
masak
| sontek: git merge --abort | 13:03 |
|
| sontek: and then something with git merge -s theirs | 13:03 |
| → ReekenX joined | 13:04 |
| → roychri joined | 13:05 |
|
masak
| hm, is there a 'theirs' merge strategy? I know there is an 'ours' merge strategy, but I don't find any indication that there's a 'theirs' one. should be, though. | 13:05 |
| ← bartek left | 13:06 |
| ← mhauri left | 13:06 |
|
sontek
| There is an octopus strategy | 13:06 |
|
| ;) | 13:06 |
|
| I don't see a theirs | 13:06 |
|
talexb
| masak, Yes, there is a 'theirs' merge strategy. Look at the man page. | 13:06 |
|
masak
| talexb: I did. it only talks about the 'theirs' sub-strategy of *recursive* merge. | 13:07 |
| → alunduil joined | 13:07 |
| → freeed joined | 13:07 |
|
_ikke_
| masak: there are two different things | 13:07 |
|
| masak: You have a ours merge strategy | 13:07 |
|
talexb
| masak, Oops -- I mis-read your question, sorry. | 13:07 |
|
sontek
| Available strategies are: octopus ours recursive resolve subtree. | 13:07 |
|
_ikke_
| masak: But you also have recursive strategoy in ours / theirs mode | 13:07 |
|
sontek
| Those are the only strategies I see | 13:08 |
|
_ikke_
| sontek: Look under recursive | 13:08 |
|
| masak* | 13:08 |
|
talexb
| Yes, as _ikke_ says, it's recursive / theirs that I'm thinking og. | 13:08 |
|
_ikke_
| masak: But it operates slightly different then orus | 13:08 |
| ← alunduil left | 13:08 |
| → ehsan joined | 13:08 |
|
charon
| "slightly"? o.O | 13:08 |
|
_ikke_
| understatement | 13:09 |
|
| they're two different things | 13:09 |
| ← Axsuul left | 13:09 |
| → alunduil joined | 13:09 |
|
_ikke_
| ours strategy just ignores any changes from theirs. While recursive ours only select ours when there is a conflict | 13:09 |
| ← tsdh left | 13:09 |
|
| charon nods | 13:09 |
|
masak
| right. | 13:09 |
|
charon
| masak: there is no '-s theirs' because invaribly the people asking for it should use a reset instread | 13:10 |
|
masak
| why isn't there a 'theirs' merge strategy, to mirror 'ours'? | 13:10 |
|
| ah! | 13:10 |
|
| thanks! | 13:10 |
|
sontek
| So, if I have a single file that didn't merge cleanly and just want to take theirs I use recursive theirs? | 13:10 |
| ← ReekenX left | 13:10 |
|
_ikke_
| sontek: That only select theirs when a conflict happens | 13:10 |
| ← project2501b left | 13:10 |
| → ReekenX joined | 13:11 |
|
_ikke_
| sontek: So the result is something completely new | 13:11 |
| → jargon- joined | 13:11 |
|
sontek
| Yeah, I don't care about my changes, they were very minimal, it doesn't even make sense that they conflict | 13:11 |
| → shoky joined | 13:11 |
|
sontek
| but the resulting merge file is way to crazy to even try | 13:11 |
|
| I modified the very top line (removed it) and that shift seems to have confused git | 13:11 |
| → slavka`2 joined | 13:12 |
|
sontek
| so I just want to take their changes and then I'll add my few little things back | 13:12 |
| → cilly joined | 13:12 |
| ← gerryvdm_mbp left | 13:13 |
| ← flippo left | 13:13 |
| → scelis joined | 13:13 |
| ← slavka` left | 13:13 |
| → dv310p3r joined | 13:13 |
| → macabre joined | 13:13 |
| ← mikepack left | 13:13 |
| → gerryvdm_mbp joined | 13:13 |
| ← gerryvdm_mbp left | 13:13 |
| → gerryvdm_mbp joined | 13:13 |
| ← shoky left | 13:14 |
| → anddam joined | 13:15 |
| → mhauri joined | 13:15 |
| ← fbroasca left | 13:16 |
|
anddam
| I have file aDir/aFile in branch master, committed, I create branch newBranch, and mv aDir/aFile to newDir/aFile, then commit this to newBranch | 13:16 |
|
genesis
| http://www.pasteall.org/35409/diff someone to review this little patch quickly ? :) | 13:16 |
|
charon
| genesis: just send it to the list like everyone else? :) | 13:17 |
|
anddam
| when I switch to branch master aFile is in aDir, as expected, but newDir is there too | 13:17 |
|
| anddam oopses | 13:17 |
| ← Womkes left | 13:17 |
| ← marhaban left | 13:17 |
|
anddam
| I put another file in newDir, so it's correct newDir is still there | 13:18 |
| ← RaceCondition left | 13:19 |
| → juvenal joined | 13:19 |
| → RaceCondition joined | 13:20 |
| ← scelis left | 13:20 |
| → AaronMT joined | 13:20 |
| ← pretty_f_ left | 13:20 |
| → chuckharmston joined | 13:21 |
| ← setmeaway left | 13:21 |
| ← ReekenX left | 13:21 |
| → Womkes joined | 13:22 |
| ← Womkes left | 13:22 |
| → Womkes joined | 13:22 |
| → RaceCondishun joined | 13:22 |
| → scarlisle joined | 13:23 |
| ← ehsan left | 13:23 |
| ← Esa_ left | 13:24 |
| ← RaceCondition left | 13:24 |
| RaceCondishun → RaceCondition | 13:24 |
| → rapsys_ joined | 13:25 |
|
sontek
| Is there a way to get a vimdiff view of a single file thats conflicting? | 13:27 |
| ← Sander^work left | 13:27 |
| ← TheJH left | 13:27 |
| ← rapsys_ left | 13:28 |
|
_ikke_
| sontek: git mergetool | 13:28 |
|
| ? | 13:28 |
| ← glennpratt left | 13:29 |
| → rapsys joined | 13:29 |
| ← OOPMan left | 13:29 |
|
rapsys
| hi | 13:29 |
| → chisler joined | 13:29 |
|
rapsys
| i am looking for a reloable way to update a git clone and fetch tags | 13:29 |
| → whaley joined | 13:29 |
|
_ikke_
| what? | 13:29 |
|
FauxFaux
| What's a "reloable"? | 13:29 |
|
rapsys
| is "git fetch --tags" enough to get all tags and modification from origin repository ? | 13:30 |
|
_ikke_
| reliable I guess | 13:30 |
|
rapsys
| s/reloable/reliable/g | 13:30 |
|
| sry | 13:30 |
| → chrisgeorge joined | 13:30 |
| → runaway joined | 13:30 |
|
talexb
| rapsys, If you've cloned the repo, git branch -r should show you all of the remote branches. | 13:30 |
| → klj613 joined | 13:30 |
|
rapsys
| talexb: i already use that for branches | 13:30 |
| → yatiohi joined | 13:31 |
|
rapsys
| but i need to get the tags localy | 13:31 |
| ← cagedwisdom left | 13:31 |
|
klj613
| is it possible to specify a branch whilst `git log`? i dont want to checkout a different branch just to check log? | 13:31 |
|
masak
| klj613: yes. | 13:31 |
|
sontek
| _ikke_: thanks, thats what I was lucking for, still doesn't help me figure out what git thinks happened with this merge | 13:32 |
|
talexb
| rapsys, how about git ls-remote --tags | 13:32 |
|
_ikke_
| sontek: git doesn't think anything happened | 13:32 |
|
canton7
| klj613, 'git log <branch>' ? | 13:33 |
|
rapsys
| talexb: interesting, looking onto it | 13:34 |
|
talexb
| rapsys, Google is your friend. ;) | 13:34 |
| → gfontenot joined | 13:34 |
| → danielfi_ joined | 13:35 |
|
rapsys
| i am working on a deploy script, and after is a git checkout (tag|branch) enough to get localy all the diff localy ? | 13:35 |
| → ReekenX joined | 13:35 |
| → tplaner_ joined | 13:35 |
|
rapsys
| my goal is to 1 (check if rev/branch exists) and switch to it localy after | 13:36 |
|
sontek
| yeah, I added the 1 line back that I removed and the merge went cleanly | 13:36 |
| → trungqtran joined | 13:36 |
|
sontek
| I think it was the whitespace that was confusing git | 13:36 |
| ← RaceCondition left | 13:36 |
| → RaceCondition joined | 13:37 |
|
rapsys
| talexb: i already looked on google but didn't get all i was looking for sadly :'( | 13:37 |
| ← danielfilho left | 13:37 |
| → Davey joined | 13:37 |
|
talexb
| rapsys, You chould just do git checkout $branchThatMayNotExist and see if the operation succeeded. | 13:38 |
| ← sebrock left | 13:38 |
| ← Lord_DeathMatch left | 13:40 |
|
tarkus
| how can i create an empty branch (without any files inside)? | 13:40 |
| → Lord_DeathMatch joined | 13:40 |
|
tarkus
| sort of deattached from the master branch | 13:40 |
|
canton7
| tarkus, checkout --orphan | 13:40 |
|
tarkus
| canton7 thanks | 13:41 |
|
| then i want to attach remote repo to this branch, make pull request and make baseless merge to the main branch.. | 13:41 |
| → AlbireoX joined | 13:42 |
| → les_sylvains joined | 13:43 |
|
talexb
| tarkus, now it sounds like you want to git remote add other $otherRepo and then merge other into your current repo. | 13:43 |
| → stewbydoo joined | 13:43 |
| → bartek joined | 13:44 |
| → dvide joined | 13:44 |
| ← bfgun left | 13:45 |
| → tsdh joined | 13:46 |
| → Psy-Q joined | 13:46 |
|
tarkus
| talexb yep, i tried this route.. pulling files from a `base` repo.. but this overrides my local changes, didn't work as a merge.. i'm thinking trying creating `base` branch with `base` repository attached to it, pull into this branch, then merge base into the master | 13:46 |
| → glennpratt joined | 13:46 |
| gitinfo set mode: +v | 13:47 |
| → jrajav joined | 13:47 |
| → thierryp joined | 13:47 |
| ← dermatthias left | 13:48 |
| ← xiangfu left | 13:48 |
| ← errr left | 13:48 |
| → pppaul joined | 13:48 |
| ← pppaul left | 13:48 |
| → shankarcabus joined | 13:49 |
| → pppaul joined | 13:49 |
| ← pppaul left | 13:49 |
| → adamben joined | 13:49 |
|
Psy-Q
| after an automatic merge, i can no longer clone my repository. the master branch can't be checked out ("nonexistant object ....") but the strange thing is that the object exists! | 13:49 |
|
| https://gist.github.com/3756034 <- here's a gist of the attempted clone on a remote machine, and ls -l of the object on the server hosting the git repo | 13:49 |
| → pppaul joined | 13:49 |
| ← pppaul left | 13:50 |
| → bfgun joined | 13:50 |
| → pppaul joined | 13:50 |
| ← jaseemabid left | 13:50 |
|
tarkus
| What if you have a `base` project in a dedicated repo, clone it and commit/push changes into a dedicate repo.. then if `base` repo is updated you want to merge those changes into your project.. how do you do it? Just adding `base` and `origin` repos? pulling from base, and pushing\pulling to origin.. | 13:51 |
| → jackh joined | 13:51 |
| → hellopat joined | 13:51 |
|
tarkus
| ..so that you could use your `starter kit` project as a base for all other projects and at the same time, being able update `starter kit` and merge those changes into your current projects if needed | 13:52 |
| ← thierryp left | 13:52 |
|
canton7
| Psy-Q, well, for one a loose object wouldn't be transferred, so the object's missing from the packfile which was sent to the client. why? don't know. maybe try repacking everything? | 13:52 |
|
| (on the remote side, that is) | 13:52 |
| → setmeaway joined | 13:53 |
|
canton7
| and a git fsck never goes amis | 13:53 |
| ← chisler left | 13:53 |
| → kermit joined | 13:53 |
| → chisler joined | 13:54 |
|
Psy-Q
| canton7: so i can do that in a bare repo? (the .git directory on the server)? | 13:55 |
|
canton7
| sure | 13:55 |
|
Psy-Q
| canton7: alright, i will try | 13:55 |
| → erichynds joined | 13:55 |
| → aller joined | 13:55 |
|
canton7
| make sur eyou run commands as the gitolite user, or you'll end up with fooey file permissions (obvious, but some people forget) | 13:55 |
| ← adamben left | 13:55 |
| ← erichynds left | 13:56 |
| → madsy joined | 13:56 |
| ← madsy left | 13:56 |
| → madsy joined | 13:56 |
|
Psy-Q
| canton7: thanks :) it found a bunch of dangling commits. i will see if the clone works now | 13:57 |
| → kvanderw joined | 13:57 |
|
canton7
| dangling commits aren't a problem | 13:57 |
|
Psy-Q
| canton7: damn, it didn't find anything else | 13:58 |
|
| canton7: i'll see about repacking | 13:58 |
| ← soee left | 13:58 |
|
aller
| Hello. I have somehow corrupted my repo (maybe via amend): I can see MYTAG in e.g. SourceTree GUI list of tags. However, clicking the tag gives nothing (while it should give list of affected). On the other hand, doing git diff MYTAG is working. | 13:59 |
| ← medwards left | 13:59 |
| ← chisler left | 14:01 |
| → chisler joined | 14:01 |
| ← logik-bomb left | 14:02 |
| → medwards joined | 14:02 |
| → stopbit joined | 14:02 |
| → axl_ joined | 14:02 |
| ← polymar left | 14:04 |
| → maqr joined | 14:04 |
| ← Gx4xx left | 14:04 |
|
Psy-Q
| canton7: do you think repacking on the bare repo on the server is safe as well? will try it on a backup first, but just to know (this is the first time in 3 years that i had issues with git, that's why i'm not quite sure how to handle stuff) | 14:05 |
| → Gx4 joined | 14:05 |
|
aller
| In other words, "git tag" lists the tag, but e.g. "git log --simplify-by-decoration --decorate" does not show the tag | 14:05 |
| → adamben joined | 14:05 |
|
canton7
| Psy-Q, yeah, no problem | 14:06 |
| → codeM0nK3Y joined | 14:06 |
|
Psy-Q
| hmm, it's still the same issue | 14:06 |
|
canton7
| aller, that log command will only show tags which point to commits in the current branch | 14:06 |
| ← johnny_rose left | 14:06 |
| ← dmorgan left | 14:07 |
|
canton7
| Psy-Q, hmm. I'm afraid I'm out of my depth here. Either wait until someone with more knowledge of git's packing process turns up, or post to the mailing list | 14:07 |
|
Psy-Q
| the object still exists, too. and if i update the branch on a machine where i previously cloned the repo, it works | 14:07 |
|
| just new clones don't work | 14:07 |
|
| canton7: alright, thanks | 14:07 |
|
canton7
| Psy-Q, just see whether you received *any* objects? | 14:07 |
|
aller
| canton7: AFAIK, the "ghost tag" that I am observing, should be on current branch. Is there some way to check at which branch is the problematic tag? | 14:08 |
| → daviddoria joined | 14:08 |
| ← pppaul left | 14:08 |
| ← adamben left | 14:08 |
|
canton7
| this is rining a very faint bell... I think it was to do with old git + http? You're cloning over ssh aren't you? | 14:08 |
|
Psy-Q
| canton7: it doesn't even create the target directory, and "receiving objects" gets to 100% | 14:08 |
|
daviddoria
| when using git mergetool - what is the goal? To get the LOCAL file looking how you want it to? | 14:08 |
|
Psy-Q
| canton7: yes, over ssh with the plain old gitolite default setup on debian | 14:09 |
|
canton7
| Psy-Q, yeah a clone populates .git/, and then checks out the selected commit to populate the working copy | 14:09 |
|
Psy-Q
| canton7: ah! i'll look for .git | 14:09 |
|
canton7
| so I'm asking you to check whether the contents of .git are sensible | 14:09 |
| → adamben joined | 14:09 |
| ← adamben left | 14:10 |
|
canton7
| aller, git branch --contains <tagname> | 14:10 |
|
Psy-Q
| i think git clone removes the directory when it gets to that fatal error, so i'll have to ls -l while it's cloning and see if i can catch the moment when the objects come down the line :( | 14:10 |
| → pppaul joined | 14:10 |
| ← hobodave left | 14:10 |
| → ThePendulum joined | 14:11 |
| → ehsan joined | 14:11 |
|
canton7
| Psy-Q, or, it was never created, which explains the error... | 14:11 |
|
ThePendulum
| Hey. How do I set my project back to a certain commit? | 14:12 |
|
canton7
| (although it should moan about something else... so we're still in bugland) | 14:12 |
| ← lb left | 14:12 |
| → johnny_rose joined | 14:12 |
| → fbroasca joined | 14:12 |
| → lb joined | 14:12 |
| ← lb left | 14:12 |
| → lb joined | 14:12 |
|
Psy-Q
| canton7: it might never have been created, i added a list of files that appear during the clone to this gist: https://gist.github.com/3756034 | 14:12 |
|
canton7
| ThePendulum, define 'set back'. Rewrite history so the commits after that point never happened? Take a read-only look at a past commit? Create a new commit that undoes changes since a given commit? | 14:12 |
| ← ph^ left | 14:12 |
|
aller
| canton7: git tag | grep MYTAG -> MYTAG; git branch --contains MYTAG -> "" | 14:13 |
| → flaguy48 joined | 14:14 |
|
canton7
| aller, so looks like no branches reference the commit you're looking at. 'git log --oneline --decorate --graph MYTAG --branches' | 14:14 |
| → FernandoBasso joined | 14:14 |
| ← jrajav left | 14:14 |
| ← macabre left | 14:15 |
| → pfFred joined | 14:15 |
|
canton7
| aller, if you rewrote the commit which the tag points to at any point, the tag won't have been rewritten | 14:16 |
|
Psy-Q
| hrmf. refs/heads/master points to the same object still, and that object still really exists in the repo. and also on any clients that have cloned the repo before | 14:16 |
| ← pfFredd left | 14:16 |
| ← pfFred left | 14:16 |
|
aller
| canton7: as far as I understand: git commit ; git tag (set up a tag on last commit) ; git commit / amend the last commit -> "ghost" tag ? | 14:17 |
|
canton7
| Psy-Q, in the remote repo, 'git cat-file -p <object sha>' | 14:17 |
|
| aller, the tag will still point to the un-amended commit | 14:18 |
| ← runaway left | 14:18 |
|
canton7
| tags are supposed to be immutable and unchanging. they're not automatically rewritten. | 14:18 |
|
| that said, if two branches point to the same commit, and you rewriten one of them, the other remains unchanged | 14:18 |
| → runaway joined | 14:18 |
| → griotspeak joined | 14:19 |
| → marthinal joined | 14:19 |
| ← tgummerer left | 14:20 |
|
Psy-Q
| canton7: works fine, it shows the last commit | 14:20 |
| → tgummerer joined | 14:20 |
|
aller
| canton7: result of the command you gave me: http://i.imgur.com/wlNEg.png | 14:21 |
| ← brasom left | 14:21 |
|
aller
| so there is that "side path" pointing to OPTI_23, which is dangling | 14:22 |
| → Destos joined | 14:22 |
| → brasom joined | 14:22 |
| → dilfridge joined | 14:22 |
| ← cooloney left | 14:22 |
| → marhaban joined | 14:22 |
| ← ThomasLocke left | 14:22 |
|
aller
| canton7: how should I fix the repo? I also pushed the tag to origin. But I have direct acces to it | 14:23 |
|
canton7
| aller, delete + re-create the tag? | 14:24 |
|
aller
| OK. Thanks | 14:26 |
| ← bartek left | 14:27 |
|
ThePendulum
| canton7: I'd like to get to that commit like I never made any changes afterwards | 14:27 |
| → mabrand joined | 14:28 |
| → macabre joined | 14:28 |
|
canton7
| ThePendulum, have you pushed te history following that commit anywhere public? | 14:28 |
|
ThePendulum
| canton7: It's on bitbucket | 14:28 |
|
canton7
| might anyone you know/like have fetched that history? | 14:28 |
| ← juvenal left | 14:29 |
|
ThePendulum
| canton7: Probably not, it's a personal project. If they have, I don't feel bothered by it. | 14:29 |
|
canton7
| ThePendulum, cool! git checkout thebranch; git reset --hard thenewhead; git push -f; Warnings about nuking uncommitted stuff etc etc apply | 14:29 |
| → mva joined | 14:30 |
| → sbodo joined | 14:30 |
|
canton7
| (the real risk is "might people with commit access have fetched that to-be-rewritten history?", as they have the power to royally screw up your repo by accident) | 14:30 |
|
mva
| hi there! can somebody remind me link to "don't use git for websites" article? | 14:31 |
| → runaway_ joined | 14:31 |
|
ThePendulum
| The new head is the rev code that looks like 983244c59304, right? | 14:31 |
|
canton7
| !deploy maybe? | 14:31 |
|
gitinfo
| Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html | 14:31 |
|
canton7
| ThePendulum, sure, looks aright | 14:31 |
|
mva
| canton7: not exatly | 14:32 |
|
Psy-Q
| someone here claims it could be because someone checked into master with a newer git version than the gitolite server has | 14:32 |
| → jaseemabid joined | 14:32 |
|
Psy-Q
| but i can't just go and upgrade git on the server :( | 14:32 |
| ← piggybox left | 14:33 |
| ← runaway_ left | 14:33 |
|
ThePendulum
| canton7: That worked, thanks a lot! Haven't used git in a while, apparently | 14:33 |
|
canton7
| unless the server's git is utterly ancient... ooh, jgit problems maybe? | 14:33 |
| ← aller left | 14:33 |
| → runaway_ joined | 14:33 |
|
canton7
| ThePendulum, in general, !fixup is a very good resource | 14:33 |
|
gitinfo
| ThePendulum: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history! | 14:33 |
| ← swair left | 14:33 |
|
ThePendulum
| Thanks, I'll look into that | 14:33 |
| ← arkiver left | 14:33 |
|
ThePendulum
| After eating my grandfather, brb | 14:33 |
| pdelvo|afk → pdelvo | 14:33 |
|
Psy-Q
| canton7: i'm trying just with plain git from the git-core debian package, 1.7.10.4 on the client where it doesn't work, 1.7.2.5 on the server | 14:33 |
|
| i can try getting 1.7.10.4 on the server | 14:34 |
| ← jaseemabid left | 14:34 |
|
ThePendulum
| *eating cake with my grandfather, my god. | 14:34 |
| → gtrak joined | 14:34 |
|
canton7
| Psy-Q, I'm not convinced that'll be the problem. 1.7.2 isn't *ancient* | 14:34 |
| → jazz264 joined | 14:34 |
| jazz264 → jaseemabid | 14:34 |
|
mva
| anyway, thanks | 14:34 |
| ← mva left | 14:34 |
| → TheJH joined | 14:35 |
|
canton7
| mva, hmm can't think of the link you might be thinking of then. What else did it say? | 14:35 |
| ← ReekenX left | 14:35 |
|
canton7
| damn, missed him | 14:35 |
| ← runaway left | 14:35 |
| ← dr_lepper left | 14:35 |
| → bartek joined | 14:35 |
| ← pppaul left | 14:36 |
| → derpops joined | 14:36 |
| ← und3f left | 14:36 |
| → shane_ joined | 14:37 |
| ← papegaaij left | 14:38 |
| → tchan joined | 14:39 |
|
Psy-Q
| canton7: you won't believe this, but upgrading the server to 1.7.10.4 worked | 14:40 |
|
| i had to pull in a package from wheezy/testing for that :( | 14:40 |
|
| but it makes it work on all clients, old and new (!!) | 14:40 |
| → walters joined | 14:40 |
| ← tchan left | 14:41 |
| → piggybox joined | 14:41 |
|
Psy-Q
| oh. no, now it's broken again | 14:42 |
| ← RaceCondition left | 14:42 |
|
Psy-Q
| it worked to clone once, but the other virgin machines still can't clone | 14:43 |
| → RaceCondition joined | 14:43 |
|
canton7
| ah, so not fixed? | 14:43 |
|
Psy-Q
| no :( | 14:44 |
|
| i wonder why exactly one clone worked | 14:44 |
|
canton7
| I'd definitely go to the list with this | 14:44 |
|
Psy-Q
| okay | 14:44 |
| ← gavin_huang left | 14:47 |
|
ahferroin7
| see ya | 14:50 |
| ← ahferroin7 left | 14:50 |
| ← irqq left | 14:50 |
| → apok joined | 14:50 |
|
charon
| Psy-Q: random idea: are you sure the user who does the clone has sufficient permissions (also check the dirs!) to read the object? | 14:51 |
| ← danielfi_ left | 14:51 |
|
charon
| (apologies if canton7 brought this up already, i'm too lazy to read all of the backlog) | 14:51 |
| ← beautifulmind left | 14:51 |
| ← mikecmpbll left | 14:51 |
| → irqq joined | 14:52 |
| → mikecmpbll joined | 14:52 |
|
Psy-Q
| charon: well, the file is owned by the gitolite user and all interactions with the repo are over ssh as the gitolite user | 14:52 |
|
| charon: i'm not quite sure how gitolite works, it seems to give itself only read (no write) on all the repositories, but so far this has worked and is the same for all other repositories (which work fine) | 14:52 |
|
charon
| Psy-Q: ah, ok. | 14:53 |
|
| Psy-Q: actually, git itself makes the object files read-only, since no two objects have the same hash (unless you break sha1) | 14:53 |
|
Psy-Q
| i think gitolite does do some voodoo magic with those repos. we never had this issue on github or on plain ssh servers without gitolite | 14:53 |
|
| charon: ah, right! | 14:54 |
| ← rchavik left | 14:54 |
|
charon
| oh well, i'll go back to lurking. report it to the list :) | 14:54 |
| → duckxx joined | 14:54 |
|
Psy-Q
| am writing a post now :) | 14:55 |
| ← Shadeness left | 14:55 |
| → the_cybersphinx joined | 14:55 |
| → brianjlandau joined | 14:56 |
| ← brianjlandau left | 14:56 |
| → brianjlandau joined | 14:57 |
| ← teotwaki left | 14:57 |
| ← gdmfsob left | 14:57 |
| ← brianjlandau left | 14:58 |
| → kerrick joined | 14:58 |
|
charon
| i guess it would also help if you could strace the process launched on the remote end | 14:58 |
|
sitaram
| Psy-Q: gitolite does NOTHING with permissions except set umask to something strict | 14:58 |
|
| file system permissions I mean | 14:58 |
| ← caddoo left | 14:59 |
| → arkiver joined | 14:59 |
|
charon
| Psy-Q: perhaps you can proceed like this | 14:59 |
| ← whitman left | 15:00 |
| → erichynds joined | 15:01 |
| ← glennpratt left | 15:01 |
|
charon
| bah, why is this so hard to intercept! | 15:02 |
| ← senny left | 15:02 |
| → glennpratt joined | 15:02 |
| ← lb left | 15:03 |
| SirCmpwn → SirCmpwn_ | 15:04 |
| → zastern joined | 15:04 |
| → hobodave joined | 15:05 |
| → senny joined | 15:05 |
| ← johnny_rose left | 15:05 |
| → johnny_rose joined | 15:05 |
| ← Lord_DeathMatch left | 15:06 |
| ← johnny_rose left | 15:06 |
| → johnny_rose joined | 15:06 |
| ← tsdh left | 15:06 |
|
Psy-Q
| i'd have to trap the git that ssh invokes while cloning, get the pid and strace that, no? :| | 15:06 |
|
| the clone only takes a few seconds, i don't think i can be fast enough | 15:07 |
|
| i could slow down the bandwidth on the client and make it take longer :) | 15:07 |
| ← fbroasca left | 15:08 |
|
Psy-Q
| there are 5 processes that are started when i clone, git shell, git upload-pack, git pack-objects being the possible culprits | 15:08 |
|
| i'll try to strace the mother of those guys | 15:09 |
|
charon
| Psy-Q: i think i have an angle | 15:11 |
| ← freeed left | 15:11 |
|
charon
| (echo -e '0067want SHA1_OF_MASTER multi_ack_detailed side-band-64k thin-pack ofs-delta\n00000009done'; cat) | strace -f -o upload-pack.trace git upload-pack . >/dev/null | 15:12 |
|
| you can even leave out the cat | 15:12 |
| → fly9 joined | 15:12 |
|
charon
| i had that for debugging | 15:12 |
|
| then search for bits of the sha1 in upload-pack.trace | 15:12 |
|
Psy-Q
| wow | 15:12 |
|
charon
| with any luck, it will tell you that it tried to open the xx/yyyyy... file | 15:12 |
|
| perhaps we can learn from the error on that open | 15:13 |
| ← mhauri left | 15:13 |
|
charon
| (you'd run that on the server side obviously) | 15:13 |
| ← medwards left | 15:14 |
| ← cbreak-work left | 15:14 |
| → freeed joined | 15:14 |
| → ph^ joined | 15:15 |
| → soman joined | 15:15 |
| ← apok left | 15:15 |
| gitinfo set mode: +v | 15:15 |
| → medwards joined | 15:16 |
| → apok joined | 15:16 |
| → etcetera joined | 15:16 |
|
soman
| Hi all. What does next string '@@ -349,9 +350,8 @@' mean? I see it in diff | 15:16 |
| → e66 joined | 15:16 |
|
Psy-Q
| you're going to hate me for this, but after that one successful clone after i upgraded the client and a new commit and push from that client, the object HEAD is pointing at really does not exist | 15:16 |
|
FauxFaux
| soman: relating to like 394 for 9 lines in the left file.. | 15:17 |
| → kevwil joined | 15:17 |
|
charon
| Psy-Q: heh. now how did you manage that? :) | 15:18 |
| → Synthead joined | 15:18 |
|
Psy-Q
| charon: i have no clue :) but now i added a blank line to a file, pushed that, and now both the *previous* non-existing object exists | 15:18 |
|
| and the new object created by the blank-line push exists as well. though clones don't work. but now i can try that debug trick at least | 15:18 |
| ← skorgon left | 15:19 |
|
charon
| something is very wonky there | 15:19 |
|
| how can objects just disappear? | 15:19 |
|
Psy-Q
| charon: immediately after running your debug command, i get: "fatal: git upload-pack: not our ref b2820d5b67afdc400e47d5bf03d2cb91df18f779" | 15:20 |
|
| charon: it continues running, however | 15:20 |
|
| the object does exist, though: objects/b2/820d5b67afdc400e47d5bf03d2cb91df18f779 | 15:21 |
|
charon
| Psy-Q: that's the message you get if that object is not pointed at by any ref it advertises. check the sha1, check whether it is in the list of sha1s offered up when you simply say 'git upload-pack .' | 15:21 |
|
| and finally dig in the strace output file for the filename | 15:21 |
|
Psy-Q
| okay | 15:21 |
|
charon
| (or perhaps just the 820d5b67 part to make sure you can find it) | 15:21 |
| ← arkiver left | 15:22 |
| → beneggett joined | 15:22 |
| ← soman left | 15:22 |
| ← flijten left | 15:22 |
| → ReekenX joined | 15:23 |
|
Psy-Q
| charon: when i do "git upload-pack ." it does not show that sha1 for refs/heads/master. all the sha1's look unfamiliar, actually | 15:23 |
|
charon
| what does it offer for HEAD? | 15:24 |
|
| are you in the correct directory? :) | 15:24 |
|
Psy-Q
| also, upload-pack stalls after the 10th or 11th ref, prints "0000" and leaves the cursor there -- how long should this take on a 400 MB repo? | 15:24 |
| ← bartek left | 15:24 |
|
charon
| Psy-Q: it's waiting for input, just hit ctrl-d since you don't plan on actually getting anything from it | 15:24 |
| ← defts left | 15:24 |
|
charon
| Psy-Q: it advertises its capabilities and all of its refs, and waits for you to say which ones you want | 15:25 |
|
Psy-Q
| ooooh, okay | 15:25 |
| → defts joined | 15:25 |
| → bartek joined | 15:26 |
| ← ph^ left | 15:26 |
| → arkiver joined | 15:26 |
| → SirCmpwn joined | 15:26 |
|
Psy-Q
| HEAD from 'git upload-pack .' points at aa293a2f2dd8, which really is the culprit: "Trying to write ref refs/heads/master with nonexistant object aa293a2f2dd897c8fd8dabbe17417cc42bfdf4f5" | 15:27 |
|
| sorry, someone pushed to master while i was doing this and i copied the ref freshly after that, that's why the shas were confusing for a second | 15:27 |
| → aselder joined | 15:28 |
|
charon
| now i'm confused | 15:28 |
|
Psy-Q
| doing that again with the debug thing | 15:28 |
|
charon
| can you paste some of this? do HEAD and master disagree, or was that just a momentary fluke? why are people pushing while you're debugging this? | 15:28 |
| ← FernandoBasso left | 15:29 |
| → ph^ joined | 15:29 |
|
Psy-Q
| charon: i'm coping from foo.git to foo.git.bak and working inside foo.git.bak, and apparently i'd pulled (to see if the error on pull still exists) and of course that sha no longer agrees with what's in the .bak repo | 15:29 |
|
| i'll pull locally from now on | 15:29 |
| → kenperkins joined | 15:29 |
| ← ph^ left | 15:29 |
|
charon
| pull?! | 15:29 |
|
| !pull | 15:29 |
|
gitinfo
| pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/ | 15:29 |
|
Psy-Q
| the debug with the strace is still running, can i kill it (if it's waiting for input)? | 15:29 |
|
| argh! cloned | 15:30 |
|
charon
| Psy-Q: hit ctrl-d | 15:30 |
|
Psy-Q
| damn | 15:30 |
| → serbang joined | 15:30 |
| → ph^ joined | 15:30 |
|
Psy-Q
| ah, nice | 15:30 |
|
| i can give the strace | 15:30 |
| ← ph^ left | 15:30 |
|
charon
| it'll be a bit big, maybe you can cut off where it starts writing the pack | 15:30 |
|
| should be easy to see | 15:30 |
| ← _Qman left | 15:31 |
|
Psy-Q
| https://gist.github.com/3756628 <-- this is grep a2f2dd897c8f upload-pack.trace | 15:31 |
| → _iron joined | 15:31 |
| ← freeed left | 15:31 |
| → slavka` joined | 15:31 |
| ← jbub left | 15:31 |
| → i42n joined | 15:32 |
| → ph^ joined | 15:32 |
| ← ph^ left | 15:32 |
|
charon
| from a run that bailed, complaining that aa293a2f2dd8 does not exist? | 15:32 |
| → ph^ joined | 15:33 |
| ← ph^ left | 15:33 |
|
Psy-Q
| charon: ooh, i need to do the upload-pack bit on the *client*? because only the clients over ssh complain. if i do any of this in the repo on the server, it works. cloning locally on the server works as well | 15:33 |
| → tchan joined | 15:33 |
|
charon
| Psy-Q: hmm | 15:33 |
| → PapaSierra joined | 15:33 |
|
charon
| no, on the server. | 15:33 |
|
| but maybe it's not upload-pack that has a problem | 15:33 |
|
Psy-Q
| bleh, i can't do it on the client since i can't clone there in the first place, so the dir to do this in never exists | 15:34 |
|
| :) | 15:34 |
| ← serbang left | 15:34 |
|
Psy-Q
| i wonder what would happen if i copied my working clone to another machine and did a pull there | 15:34 |
|
charon
| get pull out of your head :p | 15:34 |
|
| there's more to clone than the pack, so maybe the resulting pack doesn't even have the object or some such | 15:34 |
|
| and it might fail at the ref-update stage or at the checkout stage | 15:35 |
|
canton7
| Psy-Q, cloning locally over file:// or with a relative path? | 15:35 |
| ← thirdknife left | 15:35 |
|
charon
| oooh, brilliant. | 15:35 |
| → ph^ joined | 15:35 |
|
Psy-Q
| canton7: ah, let me try that! i cloned locally using git clone /foo/bar.git /tmp/blah | 15:35 |
|
charon
| yeah, clone locally over file://, see if that reproduces | 15:35 |
| ← ph^ left | 15:35 |
| ← babilen left | 15:35 |
| ← slavka`2 left | 15:35 |
|
| canton7 is still better there's a corrupted pre-generated pack which doesn't contain that object (which does exist loose) | 15:35 |
|
canton7
| *betting | 15:35 |
| → babilen joined | 15:35 |
|
canton7
| ..and which claims it does contain that object | 15:36 |
| → ph^ joined | 15:36 |
| ← JohnBat26 left | 15:36 |
|
canton7
| (although my knowledge of packs is sketchy at best) | 15:36 |
|
charon
| i like that theory | 15:36 |
| ← ph^ left | 15:36 |
|
Psy-Q
| cloning with file:/// on the server worked fine | 15:36 |
|
| that even rhymes | 15:36 |
|
canton7
| damn | 15:36 |
| ← amcsi_work left | 15:37 |
|
charon
| Psy-Q: try splitting the clone in parts on the client. roughly: mkdir foo; cd foo; git init; git remote add origin url-to-foo; git fetch foo # let's skip checkout for now | 15:37 |
|
canton7
| actually, does file:// transfer a pack or loose objects? I know it doesn't hardlink, but.. | 15:37 |
|
Psy-Q
| some people on the web claim to delete master | 15:37 |
|
charon
| canton7: file:// spawns upload-pack | 15:37 |
|
canton7
| charon, aha cool | 15:37 |
|
charon
| just not over ssh | 15:37 |
| → arvind_khadri joined | 15:38 |
|
charon
| Psy-Q: if all of those parts worked, run 'git fsck' etc. in it | 15:38 |
| ← RaceCondition left | 15:38 |
| → Ecesis joined | 15:38 |
| → freeed joined | 15:38 |
|
charon
| wait, the last fetch should be 'git fetch origin' | 15:38 |
| → WhereIsMySpoon joined | 15:38 |
| → RaceCondition joined | 15:38 |
| → dcs_ joined | 15:39 |
|
Psy-Q
| https://gist.github.com/3756628 <-- added output of git fetch origin after that | 15:39 |
|
| it lists the branches, gets to master, then fails | 15:39 |
|
charon
| aha | 15:39 |
|
canton7
| Psy-Q, what if you clone the repo locally (which succeeds), then clone that clone over ssh? | 15:39 |
|
WhereIsMySpoon
| Hi, im trying to use git on windows and i get as far as git push -u origin master (after adding the git url and putty key in the remote(origin) settings) and it says no supported auth methods available (server sent: publickey) | 15:39 |
|
| i know i have an ssh key for git for this comp,i just set it up | 15:40 |
| ← apok left | 15:40 |
|
WhereIsMySpoon
| just in case the old one broke for some mysterious reason | 15:40 |
|
Psy-Q
| canton7: hmm, it's a bit tricky to do that, i would have to add it to gitolite etc. and give the appropriate permissions (we don't have a test copy of the gitolite server to play with right now) | 15:40 |
| → torjus_ joined | 15:40 |
|
WhereIsMySpoon
| im using tortoisegit on windows fyi | 15:40 |
|
| but atm i just resorted to simple git commands via git bash | 15:40 |
|
canton7
| Psy-Q, how about cloning over "raw" ssh, rather than gitolite? Is that possible? | 15:40 |
| ← griotspeak left | 15:41 |
|
computerdruid
| WhereIsMySpoon: how did you set up your ssh key? what file is it in? | 15:41 |
| ← jackh left | 15:41 |
| ← FrenkyNet left | 15:41 |
|
Psy-Q
| canton7: i'll see if i can stuff that repo into a plain user dir, that should make it ssh'able | 15:41 |
|
charon
| Psy-Q: can you run 'git verify-pack foo.idx' on all the objects/pack/*.idx in the server repo? | 15:41 |
| ← beneggett left | 15:41 |
|
charon
| and also in the one that you just (unsuccessfully) fetched, if it left a pack index | 15:41 |
|
WhereIsMySpoon
| computerdruid: used puttygen, saved public and private key in a folder, added the private putty key to the origin of the folder and the ssh-rsa bit into git | 15:41 |
| → jdunck joined | 15:41 |
|
WhereIsMySpoon
| by origin of the folder i mean into the settings of origin in remote in tortoisegit settings | 15:42 |
|
canton7
| Psy-Q, ^^ do that first, it'll be quicker and more informative than my suggestion | 15:42 |
|
| WhereIsMySpoon, define "into git" | 15:42 |
|
Psy-Q
| canton7: ok | 15:42 |
|
WhereIsMySpoon
| canton7: github | 15:42 |
|
| sorry | 15:42 |
|
computerdruid
| WhereIsMySpoon: try "ssh -i path/to/private/key [email@hidden.address] to see if authentication is working | 15:44 |
|
canton7
| WhereIsMySpoon, you'll need an ssh-format private key for that. puttygen can export yout ppk in openssh format for you | 15:44 |
| → chuckf joined | 15:44 |
| ← RaceCondition left | 15:44 |
|
charon
| canton7: i like your delta reuse theory, but then it seems weird that a local file:// fetch would succeed... | 15:44 |
|
canton7
| charon, yeah, my heart sank when he said that worked | 15:45 |
|
WhereIsMySpoon
| permission denied, computerdruid | 15:45 |
|
Psy-Q
| canton7: cloning through plain ssh and gitolite both fail the same way with the 1.7.2.5 client, and both work with the 1.7.10.4 client | 15:45 |
|
canton7
| Psy-Q, work more than once? | 15:45 |
|
Psy-Q
| canton7: ah, let me see | 15:45 |
|
canton7
| and this is cloning the clone of the original repo? | 15:45 |
|
WhereIsMySpoon
| canton7: there are 3 options ssh1, ssh2 rsa and ssh2 dsa | 15:45 |
|
Psy-Q
| canton7: cloning over ssh with git clone [email@hidden.address] | 15:45 |
|
WhereIsMySpoon
| in puttygen | 15:46 |
|
canton7
| WhereIsMySpoon, conversions -> export openSSH key? | 15:46 |
| → beneggett joined | 15:46 |
|
WhereIsMySpoon
| canton7: ive never had to do this before | 15:47 |
|
| when did this become the norm | 15:47 |
|
canton7
| WhereIsMySpoon, it isn't the norm. it's just required if you want to follow computerdruid's debugging suggestion | 15:47 |
| ← runaway_ left | 15:48 |
|
canton7
| an alternative is to fire up putty, point it at your private ppk key, and see if you can ssh [email@hidden.address] | 15:48 |
|
Psy-Q
| charon: no errors from git verify-pack | 15:48 |
| → thierryp joined | 15:48 |
| → sKeiths joined | 15:48 |
| → jjbohn joined | 15:48 |
|
Psy-Q
| at least i can reliably reproduce it on the 1.7.2.5 client | 15:48 |
| → scelis joined | 15:48 |
|
charon
| Psy-Q: not even on the client side after fetching? | 15:48 |
|
WhereIsMySpoon
| canton7: ah, just did it, permission denied again | 15:48 |
|
Psy-Q
| charon: aaah! hold on :) | 15:49 |
|
WhereIsMySpoon
| computerdruid: highlight so you know i just converted to openssh format and did your ssh -i thing and it still failed | 15:49 |
|
canton7
| so. we have an object which exists on the remote but isn't transferred on clone. 1.7.2.5 always fails the clone, whereas 1.7.10 works once then fails? Cloning over local fs works for all versions, both relative and file:// ? Cloning a clone of the repo works over ssh for all versions? verify-pack and fsck report no errors | 15:49 |
|
Psy-Q
| charon: only one .idx file makes it to the client, and that one doesn't give errors from git verify-pack | 15:49 |
|
charon
| Psy-Q: now try actually looking for the object in that repo | 15:50 |
|
WhereIsMySpoon
| er wait hang on | 15:50 |
|
canton7
| WhereIsMySpoon, throw -vvv into the mix | 15:50 |
|
charon
| Psy-Q: in fact git verify-pack -v foo.idx | grep sha1 should tell you if it's in the pack | 15:50 |
|
WhereIsMySpoon
| canton7: it pasted \'s into bash instead of /'s | 15:50 |
|
| the ssh -i thing works | 15:50 |
|
canton7
| hmm? | 15:50 |
|
WhereIsMySpoon
| hoiwever git push -u origin master still doesnt | 15:51 |
| → descention joined | 15:51 |
|
canton7
| WhereIsMySpoon, git remote -v ? | 15:51 |
|
WhereIsMySpoon
| canton7: origin [email@hidden.address] (fetch) | 15:52 |
|
| origin [email@hidden.address] (push) | 15:52 |
|
canton7
| sensible | 15:52 |
|
| so 'ssh -vvv [email@hidden.address] works? | 15:52 |
|
Psy-Q
| charon: verify-pack shows the object in the object on the 1.7.10.4 client where everything works, but not after the fetch on the 1.7.2.5 client | 15:52 |
|
canton7
| ah, you passed -i | 15:52 |
|
Psy-Q
| s/in the object//g | 15:52 |
|
charon
| aha, so finally we got something | 15:52 |
|
canton7
| WhereIsMySpoon, stick your ssh private key in ~/.ssh, and try the push again | 15:52 |
|
WhereIsMySpoon
| canton7: no it doesnt | 15:52 |
|
| the -i thing does tho | 15:53 |
| ← thierryp left | 15:53 |
|
WhereIsMySpoon
| canton7: im using windows | 15:53 |
|
| where do i put it? | 15:53 |
|
| users/me/.ssh | 15:53 |
|
| ? | 15:53 |
|
charon
| Psy-Q: now my next step would be tracing the conversation about what objects to send, but that'll be somewhat difficult thanks to ssh | 15:53 |
| ← dege left | 15:53 |
|
canton7
| WhereIsMySpoon, git hash will be able to tell you where it is. probably there, yes | 15:53 |
| ← Synthead left | 15:53 |
|
WhereIsMySpoon
| canton7: ok moved there, git remote -v still doesnt work | 15:54 |
|
| er, i mean ssh -vvv | 15:54 |
|
canton7
| WhereIsMySpoon, oops, call it id_rsa too | 15:54 |
| ← GFLJay left | 15:54 |
|
computerdruid
| canton7: does that work if it isn't actually an RSA key (even though it probably is) | 15:55 |
|
Psy-Q
| charon: hmmmm. well, i could get the ssh server to give some debug output if that's necessary, i guess. or can it be done with verbose output from the client side? | 15:55 |
|
WhereIsMySpoon
| canton7: no extension? | 15:55 |
| ← john_anderson left | 15:55 |
| → RaceCondition joined | 15:55 |
|
canton7
| computerdruid, nope, it's a guess :P | 15:55 |
|
computerdruid
| WhereIsMySpoon: no extension | 15:55 |
|
WhereIsMySpoon
| canton7: that auth'd | 15:55 |
| → john_anderson joined | 15:55 |
|
canton7
| WhereIsMySpoon, correct. and like computerdruid said, if it's a dsa key, it should be called id_dsa | 15:55 |
|
| WhereIsMySpoon, ok, so the sshe side works fine. it's the putty side only that's failing? | 15:55 |
|
WhereIsMySpoon
| canton7: however git push still doesnt work | 15:55 |
| → joephilly joined | 15:55 |
|
canton7
| WhereIsMySpoon, what? How odd. Oh heck, git might be configured to use plink | 15:56 |
|
joephilly
| How can I stop a file from being tracked?... it's a DLL | 15:56 |
| → nickpresta joined | 15:56 |
|
canton7
| joephilly, git rm --cached | 15:56 |
|
charon
| Psy-Q: in a pristine repo (start with a new one, up to fetch): GIT_TRACE_PACKET=3 git fetch origin 3>pkt-trace.dump | 15:56 |
|
WhereIsMySpoon
| canton7: ive used putty with another repo fine | 15:56 |
|
charon
| git has all these awesome debugging hooks that nobody knows about | 15:56 |
|
canton7
| WhereIsMySpoon, echo $GIT_SSH | 15:56 |
|
joephilly
| canton7, do I need to run it on the particular file? | 15:56 |
|
WhereIsMySpoon
| canton7: C:\Program Files\TortoiseGit\bin\TortoisePlink.exe | 15:57 |
| ← shruggar left | 15:57 |
|
canton7
| WhereIsMySpoon, aha, so that's why the ssh is working but the push isn't. never mind, let's stick with putty | 15:57 |
|
charon
| Psy-Q: for now let's look at the start of that conversation, again until it starts sending the pack (or receiving, in this case) | 15:57 |
| ← Erenss left | 15:57 |
|
canton7
| WhereIsMySpoon, tortoisegit is definitely configured to look for the right ppk file? | 15:57 |
| → dzonder joined | 15:57 |
|
WhereIsMySpoon
| canton7: yes | 15:58 |
|
computerdruid
| sorry WhereIsMySpoon, I'm used to openssh, not Putty | 15:58 |
|
| didn't mean to waste your time | 15:58 |
|
WhereIsMySpoon
| computerdruid: no worries | 15:58 |
|
| computerdruid: should i not use openssh key then | 15:58 |
| → martinjlowm joined | 15:58 |
| ← Davey left | 15:59 |
|
WhereIsMySpoon
| or does it not matter | 15:59 |
|
Psy-Q
| charon: pkt-trace.dump was empty. is it really 3 i have to redirect? | 15:59 |
|
computerdruid
| why did putty invent their own key format, anyway? | 15:59 |
| ← arkiver left | 15:59 |
|
WhereIsMySpoon
| cause theyre mad balls | 15:59 |
|
computerdruid
| WhereIsMySpoon: I'm pretty sure either should work, they just have different configuration methods | 15:59 |
|
WhereIsMySpoon
| hm | 16:00 |
|
| canton7: would it be easier if i set up a new folder for use with git and you guide me through all teh steps you think i should need to get it working on a repo? | 16:00 |
|
| just in case you've forgotten something | 16:01 |
|
masak
| if git-cherry-pick didn't exist, what would be a way to emulate it? | 16:01 |
|
| (for the simple case of cherry-picking one commit, say) | 16:01 |
|
wereHamster
| WhereIsMySpoon: wouldn't it be easier to use https instead of ssh? | 16:01 |
|
Psy-Q
| GIT_TRACE_PACKET doesn't seem to do anything, perhaps it doesn't exist on 1.7.2.5 | 16:01 |
|
WhereIsMySpoon
| wereHamster: er? | 16:01 |
|
canton7
| WhereIsMySpoon, I barely know torgoisegit at all, so I won't be able to help with it. The bit that's failing is the auth, and you're convinced tortoisegit is correctly configured? Out of curiosity, load the key into pageant and try again | 16:01 |
|
Psy-Q
| it works on 1.7.10.4 | 16:01 |
| ← osxorgate left | 16:02 |
| ← rjmt_ left | 16:02 |
|
WhereIsMySpoon
| canton7: that worked | 16:02 |
| → steveoliver joined | 16:02 |
|
WhereIsMySpoon
| D: | 16:02 |
|
canton7
| WhereIsMySpoon, so it's the tortoisegit configuration *alone* which is the problem | 16:03 |
|
WhereIsMySpoon
| sorta stupid i have to load it into 2 different things but whatevs | 16:03 |
| ← tchan left | 16:03 |
|
canton7
| WhereIsMySpoon, the pageant stuff is separate to the tortoisegit config | 16:03 |
|
WhereIsMySpoon
| oic | 16:03 |
| ← tellone left | 16:03 |
|
WhereIsMySpoon
| one sec, brb | 16:03 |
| ← chuckharmston left | 16:03 |
| → shruggar joined | 16:04 |
| → rjmt_ joined | 16:04 |
| → apok joined | 16:04 |
|
Psy-Q
| there are many many bugs still posted against debian's 1.7.2.5 package | 16:05 |
|
| maybe it's just best if i tell all the devs to upgrade to 1.7.10.4 and things work again? | 16:05 |
|
| perhaps this is one of those bugs | 16:05 |
|
WhereIsMySpoon
| canton7: back | 16:05 |
|
computerdruid
| Psy-Q: I haven't been following entirely, how easy is it to reproduce this problem? | 16:06 |
|
| Psy-Q: is it something you could bisect to find out the cause? | 16:06 |
|
Psy-Q
| computerdruid: i don't know, it only happesn on one of our 12 repositories and "only" with the master branch. charon and canton7 have been trying to trace this for at least an hour already :) | 16:07 |
| ← apok left | 16:07 |
| → mritz joined | 16:07 |
| → GFLJay joined | 16:07 |
| → gxk joined | 16:07 |
| → apok joined | 16:07 |
|
Psy-Q
| computerdruid: i think i can't reproduce it with just any repository, though (?) | 16:07 |
| ← etcetera left | 16:07 |
| ← nwest left | 16:07 |
|
charon
| Psy-Q: well, for the users upgrading is definitely best ;-) i'm just wondering if the bug was really fixed or only hidden | 16:07 |
|
computerdruid
| Psy-Q: but it's the client version that changes things, right? so you could bisect on the client? | 16:08 |
| → stopbit_ joined | 16:08 |
| → chuckharmston joined | 16:08 |
| ← trungqtran left | 16:08 |
| ← stopbit left | 16:08 |
|
computerdruid
| with that one repository that breaks them | 16:08 |
|
Psy-Q
| computerdruid: yeah, we've been stepping through those bits with charon | 16:08 |
|
WhereIsMySpoon
| canton7: you got any idea why tortoise git isnt reading this putty key file then? | 16:08 |
|
Psy-Q
| oh! i remember something canton7 said earlier | 16:08 |
|
WhereIsMySpoon
| i mean, i have it input in the settings | 16:08 |
|
Psy-Q
| to copy the repo to a machine that locally has 1.7.2.5 | 16:08 |
|
| and see if that works to clone via file:/// on there | 16:08 |
|
canton7
| WhereIsMySpoon, nope. I dumped tortoisegit when I started to properly learn git, as it was getting in the way | 16:08 |
| → etcetera joined | 16:08 |
|
canton7
| Psy-Q, aah, the file:/// clone was with 1.7.10 ? | 16:08 |
|
WhereIsMySpoon
| canton7: ok ill see if theres a tgit channel | 16:09 |
|
canton7
| WhereIsMySpoon, ok cool | 16:09 |
|
WhereIsMySpoon
| canton7: computerdruid thanks for the help | 16:09 |
|
| s/:/, | 16:09 |
|
charon
| Psy-Q: but without the packet trace the only way i see to get at that conversation is writing a stub script that wraps around ssh on both ends, and dumps the conversation to files | 16:09 |
|
Psy-Q
| oh! | 16:10 |
|
charon
| Psy-Q: or an equivalent feature in ssh, but i doubt it exists | 16:10 |
| → ph^ joined | 16:10 |
| ← tjadc left | 16:10 |
|
Psy-Q
| damn, then i followed your instructions wrong at that step. the file:/// breaks as well on 1.7.2.5 (!) | 16:10 |
|
canton7
| Psy-Q, I see you've posted to git-users. In general the level of expertise on there is about on par with this channel, but with fewer people around. the main list, [email@hidden.address] is probably the best place to go | 16:10 |
| → mtlife` joined | 16:10 |
| ← fornext left | 16:11 |
|
Psy-Q
| canton7: ah, okay :) misunderstood that, then | 16:11 |
|
canton7
| yeah, sorry I should have specified which list | 16:11 |
|
Psy-Q
| but anyway! the repo also breaks locally with file:///, but not with a relative path | 16:11 |
|
| which is good | 16:11 |
|
canton7
| right! so the old theory can be resurrected | 16:11 |
| ← GFLJay left | 16:12 |
|
Psy-Q
| i like it if things break reliably | 16:12 |
|
canton7
| and it takes the blame off ssh :P | 16:12 |
|
Psy-Q
| so to reproduce the error, you need debian stable, git 1.7.2.5 and our repo :) | 16:12 |
| → Mantic0re joined | 16:12 |
| → devth joined | 16:12 |
|
computerdruid
| Psy-Q: is the repo publicly sharable? | 16:12 |
|
Psy-Q
| computerdruid: unfortunately, it's in our only non-FOSS project :( | 16:13 |
|
computerdruid
| Psy-Q: of course it is :P | 16:13 |
|
Psy-Q
| the FOSS projects are on github, which works perfectly :( | 16:13 |
| ← scelis left | 16:13 |
|
canton7
| Psy-Q, this is a bit tricky, but can you clone the broken repo with 1.7.10 (into a new bare repo), then clone that over file:/// with 1.7.2.5? | 16:13 |
|
Psy-Q
| canton7: can try! | 16:13 |
| ← mtlife left | 16:14 |
|
canton7
| Psy-Q, actually, thinking about it, that won't yeild as much info as I thought it would | 16:14 |
| ← mtlife` left | 16:15 |
|
Psy-Q
| canton7: oh no, not at all, only that it *completely fixed the problem* (!) | 16:15 |
| ← bfgun left | 16:15 |
|
canton7
| oh. that's not what I was expecting. damn | 16:15 |
| → maletor joined | 16:15 |
|
canton7
| what git version's running on the gitolite server? | 16:16 |
|
Psy-Q
| git clone --bare [email@hidden.address] blah.git <-- and then clone blah.git using file:/// on a host with 1.7.2.5 -- it works | 16:16 |
| → tchan joined | 16:16 |
|
Psy-Q
| 1.7.10.4 on the server, 1.7.2.5 on the affected clients | 16:16 |
|
canton7
| right | 16:16 |
|
Psy-Q
| but hey -- shouldn't i just clone --bare using 1.7.10.4 and stick that back on the server? | 16:17 |
| ← WhereIsMySpoon left | 16:17 |
|
Psy-Q
| it won't help trace the error of course | 16:17 |
|
charon
| yes, but do keep a broken copy | 16:17 |
|
canton7
| Psy-Q, that was what I was thinking. Do keep a copy of the faulty repo | 16:17 |
|
| charon, assuming pack reuse is the problem, would that process cause the pack to be rewritten? I was thinking not | 16:17 |
|
Psy-Q
| i can tarball that repo and send it to someone under some pseudo-NDA | 16:17 |
|
| not that there are secrets in there, we're planning to open this stuff up as well, but we haven't cleaned it properly yet | 16:18 |
| → hwrd|work joined | 16:18 |
|
charon
| canton7: evidently there is a client component to the problem, as it only triggers for 1.7.2.x | 16:18 |
| → thirdknife joined | 16:19 |
| → Davey joined | 16:19 |
|
charon
| canton7: so my pplan was to find out whether the client asks for something different and the server obliges, or whether the conversation is the same and the client just masks it | 16:20 |
| ← babilen left | 16:20 |
|
canton7
| charon, sure. I'm trying to remember two posts I saw on the ML about what I *think* were breakages in packfiles. One was with bad modes/bad padding of modes/bad something to do with modes, and the other was ordering of items in trees(?) (which doesn't sound likely here??) | 16:20 |
| ← Melrian left | 16:20 |
|
canton7
| charon, aha | 16:21 |
|
charon
| Psy-Q: can you do the init-fetch dance with both client vers, and then verify-pack each to see if they agree on not having the object/? | 16:21 |
|
| verify-pack -v | grep sha1, that is | 16:22 |
|
canton7
| Psy-Q, anyone with push access running jgit/egit ? | 16:22 |
| → Sky[x] joined | 16:23 |
| ← notVert left | 16:23 |
| ← Desproges left | 16:23 |
|
Psy-Q
| canton7: not that i know, no. eclipse/zend studio counts as jgit, right? | 16:23 |
|
canton7
| eclipse counts as egit yup. not sure about zend studio | 16:23 |
| ← _jmg left | 16:24 |
| → cwang joined | 16:24 |
| ← torjus_ left | 16:24 |
|
Psy-Q
| canton7: i just heard from a dev that no one used egit on that repo when the error already happened | 16:24 |
| → towski joined | 16:24 |
|
canton7
| ah ok | 16:24 |
| ← angelsl left | 16:24 |
| ← nickpresta left | 16:24 |
| ← jdunck left | 16:24 |
| ← afuentes left | 16:24 |
|
Psy-Q
| charon: didn't i do the init-fetch dance already on both sides? | 16:24 |
|
| aah, no, i just did it on 1.7.2.5 | 16:25 |
| → converge joined | 16:25 |
| ← klj613 left | 16:25 |
| ← LimeBlast left | 16:26 |
|
ThePendulum
| Psy-Q: You share your nickname with a large mental health organisation here | 16:26 |
| ← chuckharmston left | 16:26 |
|
Psy-Q
| ThePendulum: they got it after me :( | 16:26 |
|
charon
| i think snooping on the ssh conversation should be as easy as setting GIT_SSH to a script that does tee outgoing | ssh "$@" | tee incoming, so that would be next if 1.7.2 doesn't have it but 1.7.10 does | 16:26 |
|
Psy-Q
| ThePendulum: i had it from the psyq compiler from psygnosis (rest in peace) | 16:26 |
| ← gerryvdm_mbp left | 16:26 |
| → Synthead joined | 16:27 |
|
ThePendulum
| Psy-Q: See, that's what they are here for! They will help you get rid off that silly idea! | 16:27 |
| ← cwang left | 16:27 |
| → cwang joined | 16:27 |
|
canton7
| another possibility: there's a git version with GIT_TRACE_PACKET *and* the clone problem | 16:27 |
|
ThePendulum
| Who am I kidding, fuck, I lost 2 friends to mental health issues so that name depresses me :| | 16:27 |
|
| Anyway | 16:27 |
|
Psy-Q
| oh | 16:28 |
|
| i can change my nick if it helps :| | 16:28 |
|
| but i'd have to re-register | 16:28 |
| → tdlguiinstall joined | 16:28 |
|
canton7
| (GIT_TRACE_PACKET was introduced in bbc30f99, which was v1.7.5 | 16:29 |
|
Psy-Q
| canton7: damn. a few versions too far ahead for me | 16:29 |
|
ThePendulum
| Psy-Q: Nah nah, haha, don't worry about that. | 16:29 |
| ← mikecmpbll left | 16:30 |
|
canton7
| Psy-Q, might it be easier to stick v1.7.5 on a box / in a VM, than follow charon's approach? No guarantees the problem still exists in 1.7.5 of course | 16:30 |
|
Psy-Q
| well, verify-pack on 1.7.2.5 definitely says the sha doesn't exist there | 16:30 |
| → ksiazczyk-grzego joined | 16:31 |
|
Psy-Q
| and on 1.7.10.4 it definitely says that it exists | 16:31 |
|
| canton7: yah, well... | 16:31 |
|
| anyway, i think you've helped quite a lot with the --bare hint :) | 16:31 |
|
canton7
| another debugging possibility: take the clonecreated by 1.7.2.5, give it verbatim to 1.7.10, and see whether it can see the sha in the pack? | 16:31 |
| ← sam___ left | 16:32 |
|
Psy-Q
| i'll try to pipe that SSH stuff now | 16:32 |
| ← stewbydoo left | 16:32 |
|
Psy-Q
| canton7: ah, good, will do | 16:32 |
| ← daviddoria left | 16:32 |
|
canton7
| (that *should* tell us whether it's that the ref was never asked for, or was asked for, given, but can't be read for some reason) | 16:32 |
| ← tdlguiinstall left | 16:33 |
|
Psy-Q
| canton7: 1.7.4.10 does not see it either in the fetch that 1.7.2.5 did | 16:33 |
| ← marhaban left | 16:33 |
|
Psy-Q
| 10.4! damn | 16:33 |
|
| typing is harder every minute now :P | 16:34 |
|
canton7
| Psy-Q, right, that's my theory well out of the window -- roll on charon's suggestions | 16:34 |
|
Psy-Q
| going with the SSH eavesdropping | 16:34 |
| → dfr|mac joined | 16:35 |
| ← stodan left | 16:35 |
| → dreiss joined | 16:37 |
|
Psy-Q
| hm. incoming does list the ref with the sha that doesn't exist in the pack | 16:37 |
| → keepguessing joined | 16:37 |
|
Psy-Q
| as master, of course | 16:37 |
|
| did the clone --bare dance and everything's fine on 1.7.2.5 | 16:39 |
| ← ksiazczyk-grzego left | 16:39 |
|
Psy-Q
| if you want the broken repo, just let me know :) | 16:39 |
|
keepguessing
| Hi I am new to git. infact I find it difficult to grasp what git is at all ... I have been using perforce in my company and seriously hear a lot of hoopla about git being the best cvs | 16:39 |
| → jdunck joined | 16:39 |
|
keepguessing
| Any help in pointing out resources for a noob in git with perforce experience | 16:39 |
|
Psy-Q
| but i think it's not that much use to trace a bug in such an old git version, even if it is the one that debian stable installs right now? | 16:39 |
|
| because wheezy is right around the corner for debian | 16:39 |
| ← thirdknife left | 16:39 |
|
canton7
| keepguessing, all of !book are good | 16:39 |
|
gitinfo
| keepguessing: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable | 16:39 |
| → chimay joined | 16:40 |
|
Psy-Q
| if i can remotely buy you three guys a beer, let me know how :) | 16:40 |
|
keepguessing
| will they help me with my perforce background | 16:40 |
|
Psy-Q
| charon, canton7, computerdruid: you three guys, i mean ;) | 16:40 |
|
keepguessing
| canton7: I end up comparing with perforce and then loosing all grip | 16:40 |
| ← irqq left | 16:40 |
|
canton7
| Psy-Q, like charon said, the problem may still be there, but masked. I'd recommend adding all the new data you've got to your email, and sending it on up to the main mailing list. someone there might be interested | 16:40 |
|
| keepguessing, we recommend that people with svn experience forget everything they know about svn. I suspect the same goes for perforce | 16:41 |
| → tkellen joined | 16:41 |
|
keepguessing
| canton7: haha. I exactly feel the same that i should forget my git expereince. Because git is difficult to grab :( | 16:42 |
| → irqq joined | 16:42 |
| → griotspeak joined | 16:42 |
|
keepguessing
| canton7: I meant perforce is difficult to forget | 16:42 |
|
canton7
| keepguessing, it's not though. It's just different to what you're used to | 16:42 |
|
tkellen
| can someone tell me the shortest way to revert my last commit? i want all of the changed files to go back into my staging | 16:42 |
|
keepguessing
| canton7: hmm yeah I think I have a mental block :) Thanks for your advice | 16:42 |
|
canton7
| tkellen, is that commit public? has it been pushed? | 16:43 |
|
tkellen
| canton7: nope, it's local to me still | 16:43 |
| → tomislater joined | 16:43 |
|
tkellen
| canton7: i committed changes to 3 files instead of 1. i want to go back to the state where all three were in stagings so i can commit them individually properly | 16:43 |
|
canton7
| tkellen, git reset --soft HEAD^ | 16:43 |
|
charon
| Psy-Q: i'd be sort of interested just to do an automated bisection, and look at the result to judge whether it was a bugfix or unintentional | 16:43 |
|
| Psy-Q: you could of course just do the bisection yourself ;) | 16:44 |
|
Psy-Q
| charon: phew.. | 16:44 |
|
tkellen
| canton7: boy, that was easy. thanks! | 16:44 |
|
canton7
| !reset | 16:44 |
|
gitinfo
| A good resource explaining git-reset is http://git-scm.com/2011/07/11/reset.html | 16:44 |
|
Psy-Q
| well, i'll leave my fingers off this for now :) | 16:45 |
| → notVert joined | 16:45 |
|
charon
| Psy-Q: can you at least -- as a last step ;) -- compare the negotiation part of incoming and outgoing (in particular the 'want' lines in outgoing) from a run with 1.7.2 and 1.7.10, each? | 16:46 |
| → mtlife joined | 16:46 |
| → danielfilho joined | 16:47 |
| → boombatower joined | 16:47 |
| ← towski left | 16:47 |
|
charon
| it's not entirely obvious to me, from what you said above, if 1.7.2 listed a 'want' line for this | 16:47 |
|
| or just an advertisement for the ref on the server side | 16:47 |
| → juvenal joined | 16:48 |
| ← fisted_ left | 16:48 |
| → fisted joined | 16:49 |
| → ncr100 joined | 16:49 |
| → owen joined | 16:50 |
| ← tvw left | 16:50 |
| ← fisted left | 16:51 |
| ← FreeHaegin left | 16:51 |
|
Psy-Q
| charon: oh damn, i closed all my shells. let me restart them :) | 16:51 |
|
| should've had a tmux session.... | 16:51 |
| ← jargon- left | 16:51 |
| → trungqtran joined | 16:52 |
|
owen
| Hi all. I'm using git-svn to talk to a git repo at work. Unfortunately, I'm new to this process & forgot to give the --authors-file when I did `git svn clone'. Is there any way to transform the ugly svn authors (username only) to a nice name + email after the fact, or would it be easiest to set up the authors file & re-clone from SVN? | 16:52 |
| ← bgerber left | 16:53 |
|
Psy-Q
| i notice incoming on 1.7.2.5 includes binary crap at the end | 16:54 |
|
| https://gist.github.com/3757035 <-- incoming | 16:54 |
| ← patie left | 16:55 |
| ← gusnan left | 16:55 |
| ← kevwil left | 16:56 |
| → bgerber joined | 16:56 |
| → jluc_ joined | 16:56 |
| → kevwil joined | 16:56 |
|
Psy-Q
| that's incoming and outgoing there | 16:56 |
| → Protux joined | 16:57 |
| ← Protux left | 16:57 |
| ← bartek left | 17:00 |
| ← defts left | 17:00 |
| ← senny left | 17:00 |
| → robde joined | 17:01 |
| → mklappstuhl joined | 17:01 |
| ← shruggar left | 17:03 |
|
charon
| Psy-Q: umm, the lack of a pack at the end (and that's really the pack -- it basically contains everything in the repo) indicates that you didn't fetch in a newly init'd repo? :) | 17:04 |
|
Psy-Q
| charon: oh! | 17:04 |
| → tewecske joined | 17:04 |
| ← tewecske left | 17:04 |
|
Psy-Q
| argh, now the incoming is 50 MB large | 17:05 |
| ← dc5ala left | 17:05 |
|
Psy-Q
| will add the first bits of that | 17:05 |
|
charon
| Psy-Q: yeah, only the human-readable part is interesting | 17:05 |
| ← Error404NotFound left | 17:06 |
| ← jluc_ left | 17:06 |
| ← suy left | 17:06 |
| → apok_ joined | 17:07 |
|
Psy-Q
| there we go, gist updated | 17:07 |
|
| looks pretty identical | 17:07 |
| ← converge left | 17:07 |
| → stewbydoo joined | 17:07 |
| → converge joined | 17:08 |
| ← kevwil left | 17:08 |
| → nightcrawler786 joined | 17:08 |
| → scelis joined | 17:08 |
| → j416 joined | 17:08 |
| ← apok left | 17:09 |
| apok_ → apok | 17:09 |
| ← freeed left | 17:09 |
| ← aruiz left | 17:10 |
|
canton7
| except the line 0032want aa293a2f2dd897c8fd8dabbe17417cc42bfdf4f5 in outgoing 1.7.10.4 | 17:11 |
|
| which, surprise surprise, is the master ref | 17:11 |
|
| the equivaltent line in 1.7.2.5 is 0032want 4d7080de6d6269fc996d971fa319b4a612256279 | 17:11 |
| → freeed joined | 17:11 |
| → nickpresta joined | 17:12 |
|
canton7
| which is refs/for/refs/heads/master, whatever that is | 17:12 |
| ← john3 left | 17:12 |
| → skorgon joined | 17:12 |
| → Bosox20051 joined | 17:12 |
| ← MacGyver left | 17:12 |
| ← giallu left | 17:12 |
| ← Soliton left | 17:13 |
|
Psy-Q
| ooooh | 17:14 |
|
canton7
| is refs/for a gerrit-only thing? are you using gerrit? | 17:14 |
|
Psy-Q
| so it is a client problem, with the client requesting the wrong thing? | 17:14 |
| → whitman joined | 17:14 |
| → Soliton joined | 17:14 |
|
canton7
| yeah | 17:14 |
| ← jaseemabid left | 17:14 |
|
canton7
| looks it | 17:14 |
|
Psy-Q
| don't know what refs is :) | 17:14 |
|
canton7
| ar eyou using gerrit? | 17:14 |
| → guns joined | 17:14 |
| ← owen left | 17:14 |
| → fisted joined | 17:15 |
|
Psy-Q
| i don't know what gerrit is :) | 17:15 |
|
| no, i guess we're not | 17:15 |
|
canton7
| code review tool | 17:15 |
| ← mklappstuhl left | 17:15 |
|
Psy-Q
| we don't have code reviews, we just pair a bit | 17:16 |
|
| no clue where refs/for actually came from. hmm. | 17:16 |
| → tdlguiinstall joined | 17:16 |
|
spyro
| Is it possible to embed only a subdirectory of a submodule in your superproject? | 17:17 |
| ← Olipro left | 17:17 |
|
canton7
| spyro, no, I don't believe so | 17:17 |
| → Error404NotFound joined | 17:18 |
|
canton7
| aha, 1e7ba0f9: fetch-pack: match refs exactly | 17:18 |
| → serbang joined | 17:18 |
| ← dfr|mac left | 17:18 |
|
Psy-Q
| canton7, charon: if you're ever in the zürich area, i owe you dinner or lunch | 17:18 |
| ← tkellen left | 17:19 |
|
canton7
| so, fixed in 1.7.8.2 | 17:19 |
| → witul joined | 17:19 |
|
canton7
| so the clone --bare / copy worked because it cut out the refs/for | 17:19 |
| → Shusshu joined | 17:19 |
| ← Shusshu left | 17:19 |
|
Psy-Q
| hmmm | 17:20 |
|
| canton7: ah, is something mentioned about that in the changelog? | 17:20 |
|
witul
| Hi everyone. I'm looking for post-receive hook which is compatible with Trac 1.0 | 17:20 |
| → fbroasca joined | 17:20 |
|
canton7
| Psy-Q, https://github.com/git/git/commit/1e7ba0f9 | 17:21 |
| → GFLJay joined | 17:21 |
|
canton7
| (the great thing about git, is you can search commits very, very easily :P 'git log -S refs/for' brought that one up | 17:21 |
| ← medwards left | 17:22 |
| ← serbang left | 17:22 |
| ← shane_ left | 17:24 |
| → jaseemabid joined | 17:24 |
| ← Error404NotFound left | 17:24 |
| → pfFredd joined | 17:24 |
| → opus joined | 17:25 |
|
Psy-Q
| canton7: wow | 17:25 |
| ← avinashhm left | 17:26 |
| ← youngblood left | 17:26 |
|
canton7
| charon, so you were right on the money :) | 17:26 |
| → Olipro joined | 17:27 |
|
Psy-Q
| anyhow, i thank you again and my offer of free lunch or dinner stands :P | 17:27 |
|
| now we don't have to upgrade all the gits | 17:27 |
| → babilen joined | 17:27 |
|
canton7
| heh, thanks. charon did all the correct guessing | 17:27 |
|
Psy-Q
| but it was a team effort :) | 17:28 |
| → rgr joined | 17:28 |
|
Psy-Q
| bye then, i thing i'll return tomorrow (hopefully to help out a bit if i can, not with more problems) | 17:28 |
|
| s/thing/think/g | 17:28 |
| ← gtrak left | 17:29 |
| ← Psy-Q left | 17:29 |
| ← pfFredd left | 17:29 |
| ← Sky[x] left | 17:29 |
| ← mgorbach left | 17:29 |
| → gtrak joined | 17:29 |
| → maleknet joined | 17:29 |
| → torjus_ joined | 17:30 |
| → flavius joined | 17:31 |
| → mgorbach joined | 17:32 |
| ← fbroasca left | 17:33 |
| → marhaban joined | 17:33 |
| → Sky[x] joined | 17:33 |
| ← fisted left | 17:37 |
| → fisted joined | 17:37 |
| → titaniumNoob joined | 17:37 |
| → RJ3000_ joined | 17:37 |
| ← RJ3000_ left | 17:37 |
| → RJ3000_ joined | 17:38 |
| → shruggar joined | 17:38 |
| → CHC joined | 17:39 |
| → iliketurtles joined | 17:39 |
| ← maleknet left | 17:40 |
| ← etcetera left | 17:40 |
| ← thiago left | 17:42 |
| → defts joined | 17:42 |
| → MacGyver joined | 17:42 |
| ← MacGyver left | 17:42 |
| → MacGyver joined | 17:42 |
| ← fireh left | 17:43 |
| → jjido joined | 17:44 |
| → etcetera joined | 17:44 |
| ← tdlguiinstall left | 17:45 |
| ← shruggar left | 17:45 |
| → pfFredd joined | 17:45 |
| → bartek joined | 17:46 |
| ← osse left | 17:46 |
| ← iliketurtles left | 17:47 |
| ← CHC left | 17:48 |
| → shruggar joined | 17:48 |
| → rubious joined | 17:48 |
| ← bartek left | 17:48 |
| → serbang joined | 17:48 |
| → thierryp joined | 17:48 |
| ← leokaito left | 17:49 |
| ← chimay left | 17:49 |
| → leen^| joined | 17:49 |
| → bartek joined | 17:51 |
| → artnez joined | 17:52 |
| → towski joined | 17:52 |
| ← shruggar left | 17:52 |
| → revagomes joined | 17:53 |
| → jrajav joined | 17:53 |
| ← thierryp left | 17:53 |
| ← ncd|leeN left | 17:53 |
| → Eridius joined | 17:53 |
| ← mmc1 left | 17:53 |
| → shruggar joined | 17:54 |
| ← Chaser left | 17:54 |
| → panpainter joined | 17:55 |
| → Chaser joined | 17:55 |
| ← opus left | 17:55 |
| ← i42n left | 17:57 |
| ← shruggar left | 17:58 |
| → nwest joined | 17:58 |
| → mister_zombie joined | 17:59 |
| → shruggar joined | 17:59 |
|
mister_zombie
| Hi, any way I can ignore a specific remote svn path when doing git svn checkout? | 17:59 |
| ← griotspeak left | 17:59 |
| → FernandoBasso joined | 18:00 |
|
mister_zombie
| As in, ignore a branch? | 18:00 |
| → kevwil joined | 18:00 |
| → dogarrhea2 joined | 18:00 |
| ← kevwil left | 18:01 |
| ← scelis left | 18:01 |
| ← bartek left | 18:01 |
| → bartek joined | 18:01 |
| → idletom joined | 18:03 |
| → YoMomma joined | 18:03 |
| → Sliver joined | 18:03 |
| → fisted_ joined | 18:04 |
| ← shruggar left | 18:04 |
| → shruggar joined | 18:06 |
| ← bartek left | 18:06 |
| ← fisted_ left | 18:06 |
| → fisted_ joined | 18:06 |
| ← fisted_ left | 18:07 |
| → jason237 joined | 18:07 |
| ← jjbohn left | 18:07 |
| ← fisted left | 18:07 |
| ← titaniumNoob left | 18:07 |
|
milki
| there appears to be options in the man page | 18:09 |
| → bondar joined | 18:09 |
| ← serbang left | 18:09 |
|
mister_zombie
| Yeah I tried ignore-paths | 18:09 |
| ← apok left | 18:09 |
|
mister_zombie
| I think I shot myself in the foot somewhere down the road. | 18:09 |
| → apok joined | 18:09 |
|
milki
| mister_zombie: and !doesntwork? | 18:09 |
|
gitinfo
| mister_zombie: Sorry to hear it doesn't work. What happened? Did it fall asleep on your couch, with your sister? Eat your homework? What did you want it to do? What happened instead? Please be specific! | 18:09 |
| → jjbohn joined | 18:10 |
|
mister_zombie
| I created a new branch called trunk in my branches directory by accident. | 18:10 |
|
| So now everytime I checkout the thing with SVN, even though I deleted the branch, it ties master to the branches/trunk instead of the trunk. | 18:11 |
|
dogarrhea2
| i need to update one repository with the main repository, what's this called? | 18:11 |
|
milki
| mister_zombie: theres probably a related config option. /me doesnt know how the git-svn ties work | 18:11 |
|
| dogarrhea2: updating is usually a fetch | 18:12 |
| → john3 joined | 18:12 |
|
dogarrhea2
| hrm | 18:12 |
| → opus joined | 18:12 |
|
dogarrhea2
| i am always getting some kind of merge error out of date or some other error when i pull | 18:12 |
|
| i just want what's there. i don't care about my local repo | 18:12 |
|
milki
| merges can happen on merge or pull | 18:12 |
|
dogarrhea2
| when i do git reset --hard head then all of a sudden all my files are untracked | 18:12 |
|
| i thought the git reset --hard head would take everything from the remote and just dump it into local | 18:13 |
|
milki
| you probably dont want head | 18:13 |
|
| nope | 18:13 |
|
| head refers to your current commit | 18:13 |
|
| the commit on the top when you do a git log | 18:13 |
|
dogarrhea2
| how do i just get what ever is on git hub with no regard to my current repository | 18:13 |
|
| someone misinformed me earlier then I guess | 18:13 |
|
milki
| dogarrhea2: man git-reset [--hard] <refname> | 18:13 |
|
gitinfo
| dogarrhea2: the git-reset manpage is available at http://git-scm.com/docs/git-reset.html | 18:13 |
|
milki
| refname here will refer to your remote tracking branch | 18:14 |
|
dogarrhea2
| or an alias right? | 18:14 |
|
milki
| i dont know what an alias is | 18:14 |
|
dogarrhea2
| canonical name | 18:14 |
|
| git remote -v shows : someName http://someurl | 18:14 |
|
| milki hasnt seen that in the man docs for refs | 18:14 |
|
milki
| that someName is called a remote name | 18:15 |
|
dogarrhea2
| someName is an alias | 18:15 |
|
milki
| that does not constitue as a full refname | 18:15 |
|
dogarrhea2
| ok. so that's not a refname at all. i wonder what it would be | 18:15 |
|
milki
| your known refnames will be from what git branch -a says | 18:15 |
|
| and git tag --list | 18:15 |
|
| <remote>/<branch> are your remote tracking branches | 18:16 |
| ← liktoj left | 18:16 |
| → fisted joined | 18:16 |
| ← jaseemabid left | 18:17 |
| → titaniumNoob joined | 18:17 |
| ← jjido left | 18:17 |
| gitinfo set mode: +v | 18:17 |
|
dogarrhea2
| no refs in common and none specified; doing nothing | 18:18 |
| ← maxandersen1 left | 18:18 |
| → maxandersen1 joined | 18:18 |
| → OOPMan joined | 18:18 |
| ← ipalaus left | 18:19 |
| gitinfo set mode: +v | 18:19 |
|
maxandersen1
| i'm trying to convert my git svn tags/ "branches" into real tags. i've used git tag -m "svn branch tag" tagname tags/tagname | 18:19 |
|
| when doing git -d tags/tagname I get this error: | 18:19 |
|
| warning: deleting branch 'tags/jbosstools-3.http://0.0.ga/' that has been merged to | 18:19 |
|
| 'refs/remotes/origin/tags/jbosstools-3.http://0.0.ga/', but not yet been merged to HEAD. | 18:19 |
|
| tried google for something similar but very little info out there. | 18:20 |
| → dfr|mac joined | 18:20 |
| ← titaniumNoob left | 18:20 |
|
maxandersen1
| but as long as I have a tag on the same commit as the branch that tag should be just fine to checkout, right? | 18:20 |
| → titaniumNoob joined | 18:20 |
|
cbreak
| maxandersen1: stuff in refs/tags are real tags. | 18:20 |
|
maxandersen1
| cbreak: yes I get that - its just that warning that worries me. | 18:21 |
| ← chisler left | 18:21 |
| ← RJ3000_ left | 18:21 |
|
cbreak
| I don't see what you want to do | 18:21 |
|
maxandersen1
| cbreak: makes it sound like the data is "lost" when its not (afaics) | 18:21 |
|
cbreak
| you have tags | 18:21 |
|
| and want to make... tags from them? | 18:21 |
| → pppaul joined | 18:22 |
| ← pppaul left | 18:22 |
|
maxandersen1
| I have branches that are named tags/tagname, i.e. tags/jbosstools-3.2.0.Final | 18:22 |
|
| thats is how git svn converts svn tags | 18:22 |
| → kevwil joined | 18:22 |
|
cbreak
| where? | 18:22 |
| → pppaul joined | 18:22 |
|
maxandersen1
| to a branch (which they are) | 18:22 |
| ← pppaul left | 18:22 |
|
cbreak
| git branch -a, does it show them? | 18:22 |
|
maxandersen1
| cbreak: yes | 18:22 |
| → jargon- joined | 18:22 |
|
maxandersen1
| I want to convert what is in that branch into a tag. | 18:23 |
|
cbreak
| hmm... | 18:23 |
|
maxandersen1
| i.e. create tag for it and delete the branch was my thinking | 18:23 |
|
cbreak
| just mv them all into refs/tags? :) | 18:23 |
| → pppaul joined | 18:23 |
| ← pppaul left | 18:23 |
|
cbreak
| mv .git/refs/heads/tags/* .git/refs/tags/ | 18:24 |
|
| or so | 18:24 |
| → pppaul joined | 18:24 |
| ← pppaul left | 18:24 |
|
cbreak
| (backup first :) | 18:24 |
|
maxandersen1
| cbreak: well that would be the same as doing "git tag -m 'blablah' tagname tagas/tagname" and then git branch -d tags/tagname | 18:24 |
|
| would it not ? | 18:24 |
| → hyperair joined | 18:24 |
|
cbreak
| no | 18:24 |
|
| no point making worthless annotations | 18:24 |
| → pppaul joined | 18:24 |
| ← pppaul left | 18:24 |
|
cbreak
| without the -m | 18:25 |
|
maxandersen1
| cbreak: just what I was recommended to make it a "real" tag …or is that just bs ? | 18:25 |
| → pppaul joined | 18:25 |
| ← pppaul left | 18:25 |
|
cbreak
| annotations are pointless if you don't actually make an annotation | 18:25 |
| → techhelp joined | 18:25 |
| → pppaul joined | 18:26 |
|
maxandersen1
| well the annotation would be "make svn branch into real tag" …is that not an annotation ? | 18:26 |
| ← pppaul left | 18:26 |
|
cbreak
| no | 18:26 |
|
| that's rather worthless | 18:26 |
| → pppaul joined | 18:26 |
| ← pppaul left | 18:26 |
|
cbreak
| an annotation would be "this is the first release of blah, changes include new foobar and a faster bazbaz algorithm" | 18:27 |
|
maxandersen1
| cbreak: http://stackoverflow.com/questions/4971746/why-should-i-care-about-lightweight-vs-annotated-tags | 18:27 |
| → pppaul joined | 18:27 |
| ← pppaul left | 18:27 |
| ← kenperkins left | 18:27 |
| ← pfFredd left | 18:27 |
|
cbreak
| something unclear? | 18:27 |
| → pppaul joined | 18:27 |
| ← pppaul left | 18:28 |
|
maxandersen1
| cbreak: sure - but this is for recording migration info. i thought git tag -m made it have a text plus record who/where the tag was made (by me during migration) | 18:28 |
| → bartek joined | 18:28 |
|
maxandersen1
| cbreak: but does it anyway "hurt" to have the msg vs not having it ? | 18:28 |
| → pppaul joined | 18:28 |
| ← pppaul left | 18:28 |
| → pppaul joined | 18:29 |
| ← pppaul left | 18:29 |
| → scelis joined | 18:29 |
|
cbreak
| it makes an annotated tag | 18:29 |
|
| a tag object | 18:29 |
|
| it does not hurt, as long as you write a useful tag message | 18:29 |
|
maxandersen1
| yes - something in future i would do for each final release - why not have that in the past tags too ? | 18:29 |
|
cbreak
| if you write the same message in multiple tags, then it's worthless | 18:29 |
| → pppaul joined | 18:29 |
| ← pppaul left | 18:29 |
|
maxandersen1
| cbreak: well I guess I can write "tag for tagname" :) | 18:30 |
|
cbreak
| also worthless | 18:30 |
|
| tags already have their name in their name | 18:30 |
|
maxandersen1
| i just thought it would be good for having them all show up in describe etc. | 18:30 |
| → pppaul joined | 18:30 |
| ← pppaul left | 18:30 |
|
cbreak
| they show up in describe if you use git describe --tags | 18:30 |
| ← bartek left | 18:30 |
|
cbreak
| besides, you are converting svn tags | 18:31 |
| → pppaul joined | 18:31 |
| ← Gorkyman left | 18:31 |
|
cbreak
| svn tags have their "tag message" in their latest commit | 18:31 |
|
| since they are created by copying files somewhere else in the filesystem and committing that | 18:31 |
| → Fandekasp joined | 18:31 |
| → archdori joined | 18:32 |
|
maxandersen1
| cbreak: yeah I can see that. | 18:32 |
| → patrick85 joined | 18:32 |
|
maxandersen1
| cbreak: but back to that warning on branch delete…anything I should worry about ? | 18:32 |
| ← titaniumNoob left | 18:32 |
|
cbreak
| depends. | 18:32 |
|
| you read the warning | 18:32 |
| → bartek joined | 18:32 |
|
cbreak
| do you care about that? | 18:32 |
|
| (about the branch not being merged?) | 18:33 |
|
Sliver
| Why does git return error: 'patch does not apply' for git add -e if I close the editor without changing anything in the diff? | 18:33 |
| ← blami_ left | 18:34 |
| ← HavenMonkey left | 18:34 |
| ← shruggar left | 18:34 |
| ← Fandekasp left | 18:34 |
| → Advocation joined | 18:35 |
|
plato
| is there a way for me to add a folder to a commit without adding the contents? | 18:35 |
|
cbreak
| no | 18:35 |
|
plato
| git add folder/ adds the contents | 18:35 |
|
cbreak
| git does not track folders | 18:36 |
| → chuckharmston joined | 18:36 |
|
plato
| damn | 18:36 |
| ← Chaser left | 18:36 |
| ← revagomes left | 18:36 |
|
Advocation
| can anyone help with a problem I'm having? I'm receiving this error when trying to push a branch to remote origin | 18:36 |
|
| fatal: https://git@github.com:Advocation/advocation.git/info/refs not found: did you run git update-server-info on the server? | 18:36 |
|
cbreak
| Advocation: what command? | 18:36 |
|
Advocation
| git push origin gh-pages | 18:37 |
| → d3vic3 joined | 18:37 |
|
cbreak
| that URL looks wrong | 18:37 |
|
| what does git remote -vv show? | 18:37 |
| ← mister_zombie left | 18:37 |
| → titaniumNoob joined | 18:37 |
| ← torjus_ left | 18:37 |
| ← o] left | 18:37 |
| → revagomes joined | 18:38 |
|
Advocation
| origin https://git@github.com:Advocation/advocation.git (fetch) | 18:38 |
|
| origin https://git@github.com:Advocation/advocation.git (push) | 18:38 |
|
scelis
| plato: when i want to "track" a folder i usually just 'git add folder/.empty' or something similar. just track a small dotfile that you never need to touch. or a readme or something | 18:38 |
| → mmc1 joined | 18:38 |
| → Chaser joined | 18:38 |
|
cbreak
| those are wrong | 18:38 |
|
Advocation
| ah ok! | 18:38 |
|
maxandersen1
| cbreak: I don't see why I would want a branch merged - I want to have a tag for that release made. | 18:38 |
|
cbreak
| Advocation: should look like https://github.com/vmware/rbvmomi.git | 18:38 |
|
Advocation
| ah cool, I'll change them over | 18:38 |
| ← titaniumNoob left | 18:38 |
|
cbreak
| Advocation: URL's have the format hostname:port/path | 18:38 |
|
| and Advocation is not a port | 18:38 |
|
| maxandersen1: so you shouldn't care about the warning. | 18:39 |
| → o] joined | 18:39 |
|
Advocation
| cbreak: hah, thanks | 18:39 |
| → iliketurtles joined | 18:39 |
|
cbreak
| Advocation: you can copy&paste the URL from the github page | 18:39 |
|
maxandersen1
| cbreak: okey :) | 18:39 |
|
Advocation
| cbreak: its the first repo I've tried creating without using the guides; didn't quite go as well as I hoped! | 18:39 |
|
| cbreak: thanks for your help :) | 18:39 |
|
cbreak
| np. | 18:39 |
| → timtim joined | 18:39 |
| → thiago joined | 18:40 |
| ← timtim left | 18:40 |
| ← OOPMan left | 18:41 |
| ← juvenal left | 18:42 |
| ← Bosox20051 left | 18:43 |
| → juvenal joined | 18:45 |
|
plato
| thx scelis | 18:46 |
| ← Beej_ left | 18:46 |
| → goshawk joined | 18:47 |
| → iamtakingiteasy joined | 18:47 |
|
maxandersen1
| cbreak: any good way of counting all commits ? (i.e. trying to check how big an impact my branch deletes are having on the commmit count before doing the repo splitting) | 18:47 |
| ← zenx left | 18:47 |
|
maxandersen1
| (sorry if weird question - just looking for git commit counts similar to what is written out when pushing) | 18:47 |
| ← iliketurtles left | 18:48 |
| ← aircan left | 18:48 |
| → jjido joined | 18:49 |
| maxandersen1 → maxandersen | 18:49 |
| ← maxandersen left | 18:49 |
| → maxandersen joined | 18:49 |
| brooks.freenode.net set mode: +v | 18:49 |
| ChanServ set mode: -v | 18:49 |
| ← revagomes left | 18:49 |
| ← Yuuhi left | 18:49 |
| ← rubious left | 18:49 |
|
cbreak
| maxandersen: that makes little sense | 18:50 |
| → Yuuhi joined | 18:50 |
|
cbreak
| you can just count the lines in git rev-list | 18:50 |
|
| with wc -l or so | 18:50 |
| → Beej_ joined | 18:51 |
| ← Targen left | 18:52 |
| ← nickpresta left | 18:53 |
| → tworkin joined | 18:53 |
| → C10H8N4O2 joined | 18:53 |
| → ShadeTornado joined | 18:54 |
| ← opus left | 18:54 |
| → Axsuul joined | 18:54 |
| ← ShadeTornado left | 18:55 |
| ← C10H8N4O2 left | 18:55 |
| → iliketurtles joined | 18:55 |
| → aantix joined | 18:55 |
| → ShadeTornado joined | 18:55 |
| → C8H10N4O2 joined | 18:55 |
|
C8H10N4O2
| hey all | 18:56 |
| ← maletor left | 18:56 |
| → opus joined | 18:56 |
| ← opus left | 18:56 |
| ← a-l-e left | 18:56 |
| ← doomMonkey left | 18:56 |
| → OOPMan joined | 18:56 |
|
Advocation
| so with github user pages, do I need to keep the gh-pages branch? | 18:57 |
|
| the docs say that it feeds from the master branch... | 18:57 |
|
C8H10N4O2
| I am setting up a git server. I have the initial repo set up. I develop modules for Wordpress, Joomla!, Magento and extend frameworks such as CI and Zend. I want to have each of these software pieces be its own project but done want to have 20 repos to manage them that I have to log into is thre an easy way to manage multiple projects from one central repo | 18:57 |
| ← Synthead left | 18:57 |
|
Advocation
| but in the creation it said to use the gh-pages branch... | 18:57 |
| ← rgsteele left | 18:58 |
| → anth0ny_ joined | 19:00 |
| ← kevwil left | 19:00 |
| ← tomislater left | 19:00 |
| → procrastubator joined | 19:00 |
|
procrastubator
| So I am using Bitbucket and I was wondering how I could go about downloading an older commit of my project files? | 19:01 |
| → nitind joined | 19:01 |
| → kevwil joined | 19:01 |
| ← patrick85 left | 19:01 |
| → rubious joined | 19:01 |
|
talexb
| procrastubator, Clone the repo, then rewind to the commit you want .. | 19:01 |
|
procrastubator
| talexb: How do I rewind tyhrough commits? | 19:02 |
|
talexb
| procrastubator, I just use something like gitk to point and click .. | 19:02 |
|
procrastubator
| Alright cool I will look into that. | 19:03 |
| ← les_sylvains left | 19:03 |
|
talexb
| I'd reset HEAD to the commit I want, then get the file. | 19:03 |
|
| Don't forget to put HEAD back where you found it afterwards .. | 19:03 |
| ← iliketurtles left | 19:03 |
| → sq-one joined | 19:03 |
| → FrenkyNet joined | 19:04 |
| → griotspeak joined | 19:04 |
| → pfFredd joined | 19:04 |
| ← ehsan left | 19:04 |
| ← arvind_khadri left | 19:05 |
| → jas4711 joined | 19:06 |
| ← jason237 left | 19:06 |
| → cagedwisdom joined | 19:06 |
| ← Tegi_ left | 19:08 |
| → delinquentme joined | 19:08 |
|
delinquentme
| OK im having what I think are .gitignore issues | 19:08 |
| → thirdknife joined | 19:08 |
|
delinquentme
| I've got a bzr project within a git VCed dir | 19:08 |
|
C8H10N4O2
| is there any way to remove files from teh untracked status without removing them from future commits | 19:08 |
|
delinquentme
| within the .gitignore I have BETYdb/.bzr/* ... which should ignore all files within that directory ... however when I run "git status" it showing a number of files within that directory as being deleted | 19:09 |
| ← dcs_ left | 19:09 |
|
SamB_MacG5
| delinquentme: huh, I would have thought this was a question for #bzr | 19:09 |
| → maletor joined | 19:09 |
|
delinquentme
| True or false ... if its in the .gitignore it shouldn't be showing up in a git status command at all | 19:09 |
|
SamB_MacG5
| delinquentme: why aren't you ignoring the whole bzr tree ? | 19:09 |
| ← marthinal left | 19:09 |
|
nitind
| With BETYdb/.bzr/ | 19:10 |
| ← aselder left | 19:10 |
|
SamB_MacG5
| delinquentme: oh, *deleted*? | 19:10 |
|
delinquentme
| i am | 19:10 |
|
SamB_MacG5
| I guess that directory existed before | 19:10 |
|
delinquentme
| however they're showing up when I run a git status | 19:10 |
| ← pppaul left | 19:10 |
| → ehsan joined | 19:11 |
|
SamB_MacG5
| you could either commit the fact that they aren't there anymore, or filter history to expunge their ever having been there | 19:11 |
|
delinquentme
| BETYdb/.bzr/ or BETYdb/.bzr/* to ignore everything in the .bzr dir? | 19:11 |
| ← defts left | 19:11 |
| → pantsman joined | 19:11 |
|
SamB_MacG5
| delinquentme: but why are you ignoring only the .bzr directory? | 19:11 |
|
delinquentme
| SamB_MacG5, and the .bzrignore file too? | 19:12 |
|
| there isn't a second dir is there?? | 19:12 |
| ← bartek left | 19:12 |
| → bartek joined | 19:12 |
|
SamB_MacG5
| delinquentme: I meant, why are you *not* ignoring that whole tree? | 19:12 |
|
| the BETYdb tree | 19:12 |
|
delinquentme
| nahh I want to vc on that | 19:13 |
|
| just not the stuff in bazaar | 19:13 |
| ← idletom left | 19:13 |
| ← diegoviola left | 19:13 |
|
| SamB_MacG5 wonders if bzr-git would let you push that branch into a git repository so you could use it as a submodule ... | 19:14 |
|
SamB_MacG5
| delinquentme: what? You mean you have some files that aren't tracked by bzr, and some that are, and you want the former to be tracked in the containing git repository? | 19:15 |
| → tomislater joined | 19:16 |
| ← procrastubator left | 19:16 |
| → serbang joined | 19:17 |
| ← kevwil left | 19:17 |
| ChanServ set mode: +v | 19:17 |
| ← pantsman left | 19:17 |
| → NcA^ joined | 19:18 |
| → pppaul joined | 19:18 |
| → pantsman joined | 19:18 |
| ← pantsman left | 19:18 |
| → pantsman joined | 19:18 |
| ← pppaul left | 19:18 |
| ← bondar left | 19:18 |
| → pppaul joined | 19:18 |
| → kevwil joined | 19:18 |
| ← pppaul left | 19:18 |
| → upper2bits joined | 19:19 |
| → pppaul joined | 19:19 |
| ← pppaul left | 19:19 |
| ← Vile left | 19:19 |
|
upper2bits
| hey. So I have a local file that wasn't under git and someone added a file with the same name so when I pulled their code it overwrote my local copy. Is there any way to recover that? | 19:19 |
|
| guessing not since it wasn't in git but it's weird git would overwrite it like that | 19:19 |
|
NcA^
| question: I have a remote pointing to a separate repo, and am looking to merge the QA branch off of that, into the master of my current repository, what's the cleanest way to perform this? | 19:20 |
| ← thiderman left | 19:20 |
| → pppaul joined | 19:20 |
| ← pppaul left | 19:20 |
|
talexb
| upper2bits, I'm guessing that file is gone .. | 19:20 |
|
upper2bits
| :-( | 19:20 |
| → pppaul joined | 19:20 |
| ← pppaul left | 19:20 |
| → pppaul joined | 19:21 |
| ← pppaul left | 19:21 |
| → pppaul joined | 19:22 |
| ← pppaul left | 19:22 |
| → opus joined | 19:22 |
| → pppaul joined | 19:22 |
| ← pppaul left | 19:22 |
| ← Sliver left | 19:22 |
| ← dogarrhea2 left | 19:22 |
| → Hasbro joined | 19:23 |
| → pppaul joined | 19:23 |
| ← pppaul left | 19:23 |
| ← keepguessing left | 19:23 |
| → drizzd_ joined | 19:23 |
| → dvdhrm joined | 19:24 |
| → pppaul joined | 19:24 |
| ← pppaul left | 19:24 |
| → pppaul joined | 19:24 |
| ← pppaul left | 19:24 |
| ← asakura left | 19:25 |
| → pppaul joined | 19:25 |
| ← pppaul left | 19:25 |
|
cbreak
| upper2bits: git should reject that merge | 19:25 |
|
| something like "Untracked file would be overwritten by merge" | 19:26 |
|
upper2bits
| that's what I thought but it didn't... | 19:26 |
| → pppaul joined | 19:26 |
| ← pppaul left | 19:26 |
| → pppaul joined | 19:27 |
| ← pppaul left | 19:27 |
| ← drizzd left | 19:27 |
| → pppaul joined | 19:27 |
| ← axl_ left | 19:27 |
| ← pppaul left | 19:27 |
| ← letharion left | 19:28 |
| → pppaul joined | 19:28 |
| ← pppaul left | 19:28 |
|
Chuwiey
| Hey all, I've created a repository on a remote server. Then cloned it, and now I'm trying to push my changes back to the master branch on the remote. When doing 'git push origin master' I get -> refusing to update checked out branch in a non-bare repository... Could someone please explain this to me? | 19:28 |
|
cbreak
| Chuwiey: you created client repositories on the server | 19:28 |
|
Chuwiey
| cbreak: meaning? | 19:29 |
|
cbreak
| you should have created them with git init --bare | 19:29 |
| → pppaul joined | 19:29 |
| ← pppaul left | 19:29 |
|
cbreak
| just delete/rename them and make new ones | 19:29 |
|
Chuwiey
| how do i delete a repo without removing the files... | 19:29 |
| ← dvdhrm left | 19:29 |
|
cbreak
| delete it with removing the files of course | 19:29 |
| → pppaul joined | 19:29 |
| ← pppaul left | 19:29 |
| → pppaul joined | 19:30 |
|
Chuwiey
| cbreak: Obviously I have code I don't want to lose in there... | 19:30 |
| ← pppaul left | 19:30 |
|
cbreak
| yeah, and? | 19:30 |
|
| you cloned, didn't you? | 19:30 |
| → pppaul joined | 19:31 |
|
Chuwiey
| on local, not on the server... | 19:31 |
| ← pppaul left | 19:31 |
| ← jjido left | 19:31 |
|
cbreak
| you don't need the code on the server | 19:31 |
|
| just delete or rename the repo there | 19:31 |
|
| then make a new one | 19:31 |
|
| and push the history into it | 19:31 |
| → pppaul joined | 19:31 |
| ← pppaul left | 19:32 |
|
Chuwiey
| ok, so you're saying that the code I have on my local, i should push to the remote | 19:32 |
| ← cakehero left | 19:32 |
| → pppaul joined | 19:32 |
| ← pppaul left | 19:32 |
|
cbreak
| no, the HISTORY | 19:33 |
| → pppaul joined | 19:33 |
| ← pppaul left | 19:33 |
|
cbreak
| the chains of commits | 19:33 |
|
Chuwiey
| yeah, ok | 19:33 |
|
cbreak
| git init --bare will give you a server repository | 19:33 |
|
| one without working directory or files | 19:33 |
|
| one you can push into | 19:33 |
| → pppaul joined | 19:34 |
| ← pppaul left | 19:34 |
|
Chuwiey
| ok... | 19:34 |
| → cakehero joined | 19:34 |
| → t0rc joined | 19:34 |
| → pppaul joined | 19:34 |
| ← pppaul left | 19:34 |
| → macrover joined | 19:34 |
| → Jay_Levitt joined | 19:34 |
| → benbergman joined | 19:35 |
| ← tworkin left | 19:35 |
|
Chuwiey
| cbreak, and from then on, anyone else who might clone the remote repository, will still be able to push to it? | 19:35 |
| ← serbang left | 19:35 |
|
cbreak
| yes | 19:35 |
| → pppaul joined | 19:35 |
| ← pppaul left | 19:35 |
|
cbreak
| the problem was the server repo | 19:35 |
|
| if you fix that, pushing is just the usual permission problems | 19:35 |
|
Chuwiey
| ok | 19:35 |
|
cbreak
| but those are unrelated to git :) | 19:35 |
|
Chuwiey
| :) | 19:35 |
| ← bartek left | 19:36 |
| → pppaul joined | 19:36 |
| ← pppaul left | 19:36 |
|
cbreak
| if you want different users to access the repo | 19:36 |
| → bartek joined | 19:36 |
|
cbreak
| take a look at man git-init --shared | 19:36 |
|
gitinfo
| the git-init manpage is available at http://git-scm.com/docs/git-init.html | 19:36 |
| ← imajes left | 19:36 |
| → pppaul joined | 19:36 |
| ← pppaul left | 19:36 |
| → imajes joined | 19:37 |
|
Chuwiey
| once I recreated a bear repository, what should I do locally to push the code to the remote? is it git push <fullhostpath> branch ? | 19:37 |
| → pppaul joined | 19:37 |
| ← pppaul left | 19:37 |
| → pppaul joined | 19:38 |
| ← pppaul left | 19:38 |
| ← archdori left | 19:38 |
| → les_sylvains joined | 19:38 |
|
EugeneKay
| Chuwiey - git remote add $REMOTE $URI; git push -u $REMOTE $BRANCH(or --all) | 19:38 |
| → Synthead joined | 19:39 |
|
EugeneKay
| You would use --all for the push to populate the remote initially with everything you have | 19:39 |
|
Chuwiey
| EugeneKay: thanks :) | 19:39 |
|
EugeneKay
| YOu only need the -u for pushes that involve "new" branches - it's a flag to set Upstream information so you can do `git push` without any further info | 19:39 |
| ← rubious left | 19:40 |
|
Chuwiey
| EugeneKay: and if i wanted to have a working production site off one of the branches from that repository, would I checkout the branch onto a public_html dir on the server, or should I point to somewhere specific in the repo? | 19:40 |
|
EugeneKay
| !deploy | 19:40 |
|
gitinfo
| Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html | 19:40 |
|
EugeneKay
| Read that ^ | 19:40 |
|
Chuwiey
| I did... | 19:40 |
|
EugeneKay
| I use(and wrote) the script outlined as the sixth(?) option | 19:41 |
|
| It can be abused to deploy different branches to different places, or even to remote hosts | 19:41 |
|
Chuwiey
| I'll reread.. thanks :) | 19:41 |
| ← cakehero left | 19:42 |
| → doomMonkey joined | 19:42 |
| → letharion joined | 19:43 |
| → tjadc joined | 19:43 |
| → pppaul joined | 19:43 |
| ← pppaul left | 19:43 |
| → pppaul joined | 19:44 |
| ← pppaul left | 19:44 |
| → pppaul joined | 19:44 |
| ← pppaul left | 19:44 |
| → dodo joined | 19:45 |
| → pppaul joined | 19:45 |
| → Anders_J joined | 19:45 |
| ← pppaul left | 19:45 |
| ← opus left | 19:45 |
| → pppaul joined | 19:46 |
| ← pppaul left | 19:46 |
| → cantoma joined | 19:46 |
| ← nwest left | 19:46 |
| SJrX → SJr | 19:46 |
| → Marcux83|away joined | 19:46 |
| → devces joined | 19:46 |
| → gozes joined | 19:46 |
| Marcux83|away → Guest89094 | 19:47 |
|
gozes
| hello everyone | 19:47 |
| ← converge left | 19:47 |
| → jperry2_ joined | 19:47 |
| → milligan joined | 19:48 |
|
milki
| hmmm, the link changed again | 19:49 |
| → adityamenon joined | 19:49 |
| → thierryp joined | 19:49 |
| ← adityamenon left | 19:49 |
| → rubious joined | 19:49 |
| → shruggar joined | 19:49 |
|
gozes
| I just did a pull from my repo, but when I do a submoulde update it give me this error No submodule mapping found in .gitmodules, hoever, when I look insede my .gitmodules all the path and urls are there | 19:50 |
|
| were did I go wrong? | 19:50 |
| ← tr4656 left | 19:51 |
| → pppaul joined | 19:51 |
| ← pppaul left | 19:51 |
| → cakehero joined | 19:51 |
| → converge joined | 19:52 |
| → devinrhode2 joined | 19:52 |
|
gozes
| ???? | 19:53 |
| → daethorian joined | 19:53 |
| → zastaph joined | 19:53 |
| ← moguchiy left | 19:53 |
| ← thierryp left | 19:54 |
| → opus joined | 19:54 |
| → g3orge joined | 19:55 |
|
cmn
| use --init | 19:55 |
|
gozes
| cmn: did that fist got the error, so I try update | 19:55 |
| → skelterjohn joined | 19:56 |
| ← bartek left | 19:56 |
|
skelterjohn
| how do i move a branch to a particular commit? i want to roll back a merge i did to master | 19:56 |
|
cmn
| then whoever added the submdoule didn't do it properly | 19:56 |
|
gozes
| cmn: sync does the same thing | 19:56 |
| → bartek joined | 19:56 |
|
g3orge
| hi guys. when I'm doing a git commit --amend I get to have a problem. Due to the different commit locally and on the server I get lots of errors and I have to merge do get through them. Am i missing something? is there a command or a switch in push for that kind of fuctinality? | 19:56 |
|
skelterjohn
| and master back to its state from before the merge | 19:56 |
|
arietis
| what's best way to back up local git repo? | 19:56 |
|
g3orge
| arietis: clone it elsewhere... | 19:56 |
| → pppaul joined | 19:57 |
|
gozes
| cmn: I did and I did submodule add urltogitrepo | 19:57 |
| ← pppaul left | 19:57 |
|
gozes
| cmn: what did I miss | 19:57 |
|
adhawkins
| g3orge: !rewrite | 19:57 |
|
gitinfo
| g3orge: [!rewriting_public_history] Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to jump through hoops (and you have to tell them to), so it's infinitely better to just move on without rewriting. That said, you can use `git push -f` to force your new history through (may get rejected, though). See http://git-scm.com/docs/git-rebase.html#_recovering_from_upstream_rebase | 19:57 |
|
cmn
| gozes: a !repro | 19:57 |
|
gitinfo
| gozes: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session, or at least explain exactly what you did that led up to the problem. This will help immensely with troubleshooting. | 19:57 |
| → pppaul joined | 19:58 |
| ← pppaul left | 19:58 |
|
cmn
| is the merge you did important or not? does it happen with your submodule or with the one someone else did? | 19:58 |
|
| are you sure it's complaining about the submodule you're thinking about? | 19:58 |
| ← skelterjohn left | 19:58 |
| → pppaul joined | 19:58 |
| ← pppaul left | 19:58 |
| → LucasTizma joined | 19:59 |
| → pppaul joined | 19:59 |
| ← pppaul left | 19:59 |
|
gozes
| I added a few submodulse last night and made a commit and push after adding them | 19:59 |
| ← Davey left | 20:00 |
| → pppaul joined | 20:00 |
| ← pppaul left | 20:00 |
|
cmn
| show us | 20:00 |
| → Vile joined | 20:00 |
|
g3orge
| adhawkins: oh crap. yes. I'm always hearing about rewriting public history but I never actually thought about it. | 20:00 |
|
| thanks anyways | 20:00 |
|
gozes
| now In this computter did a pull and sub init follow but a update and thats when I got ther errroe No submodule mapping found in .gitmodules | 20:00 |
| ← babilen left | 20:00 |
| → pppaul joined | 20:00 |
| ← pppaul left | 20:00 |
|
arietis
| if i'll archive folder with repo using zip, will it contain git files? | 20:00 |
| → AlbireoX`Laptop joined | 20:01 |
|
cmn
| gozes: show us | 20:01 |
| → pppaul joined | 20:01 |
| ← pppaul left | 20:01 |
| ← PapaSierra left | 20:01 |
|
gozes
| wath do you wanat me to past git log? | 20:01 |
|
cmn
| no, !repro | 20:01 |
|
gitinfo
| Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session, or at least explain exactly what you did that led up to the problem. This will help immensely with troubleshooting. | 20:01 |
| ← AlbireoX left | 20:01 |
|
cmn
| commands, contents | 20:01 |
|
g3orge
| anyone can explain in plain english what the -u flag does in git push ?? thanks. | 20:02 |
| ← Advocation left | 20:02 |
| → pppaul joined | 20:02 |
| ← pppaul left | 20:02 |
| → pppaul joined | 20:02 |
| ← pppaul left | 20:02 |
| → pppaul joined | 20:03 |
| ← pppaul left | 20:03 |
|
gozes
| http://pastebin.com/n9afT3i5 | 20:03 |
|
adhawkins
| g3orge: It sets the local branch you've pushed to track the remote branch you've pushed to | 20:03 |
|
| git branch -vv will give more info | 20:04 |
| → pppaul joined | 20:04 |
| ← pppaul left | 20:04 |
| → pppaul joined | 20:04 |
| ← pppaul left | 20:04 |
|
g3orge
| isn't that the "default" state??? | 20:04 |
| → Davey joined | 20:04 |
|
g3orge
| what else does my local repo track adhawkins ?? | 20:04 |
|
cmn
| g3orge: your repo doesn't track anything | 20:05 |
|
gozes
| did git submodule sync and that was the oup put | 20:05 |
| → pppaul joined | 20:05 |
| ← pppaul left | 20:05 |
|
adhawkins
| No, if you create a branch locally, it has no link with a branch of the same name on the remote side | 20:05 |
|
| Unless you tell it to | 20:05 |
| ← bassliner left | 20:05 |
|
cmn
| gozes: so show us that the .gitmodules is right | 20:05 |
| → pppaul joined | 20:05 |
| ← pppaul left | 20:05 |
| ← dodo left | 20:06 |
| → tm1000 joined | 20:06 |
| → pppaul joined | 20:06 |
| ← pppaul left | 20:06 |
|
gozes
| cmn: http://pastebin.com/smYwddNE | 20:06 |
| → tworkin joined | 20:06 |
| → CollinPDX joined | 20:06 |
| ← bartek left | 20:06 |
| → pppaul joined | 20:07 |
| ← pppaul left | 20:07 |
| → bartek joined | 20:07 |
| → pppaul joined | 20:07 |
| ← pppaul left | 20:07 |
|
cmn
| don't use absolute pathnames | 20:07 |
|
| that's asking for trouble | 20:07 |
| → aruiz joined | 20:07 |
|
cmn
| which is why it doesn't fit | 20:07 |
| → pppaul joined | 20:08 |
| ← pppaul left | 20:08 |
|
gozes
| cmn: I dind't when I add it I use ~ | 20:08 |
|
cmn
| no | 20:08 |
| → dogarrhea2 joined | 20:08 |
| → pppaul joined | 20:08 |
|
gozes
| cmn: what is the proper way? | 20:08 |
|
cmn
| use a relative path | 20:08 |
|
cbreak
| relative to repository root | 20:08 |
|
| while you are in the repository root | 20:09 |
|
g3orge
| cmn: what's the practical difference.?? | 20:09 |
| ← Rhonda left | 20:09 |
| → Rhonda joined | 20:09 |
|
cmn
| between what and what? | 20:09 |
|
gozes
| ./dotfiels/bundelr/wahterver.git you mean? | 20:09 |
| → rohan32 joined | 20:09 |
|
dogarrhea2
| i think i figured it out. git push --force your-vagina my-penis will forcefully insert changes from my-penis into the remote, your-vagina. This doesn't actually merge anything right? it just shoves changes into the remote? | 20:09 |
|
cmn
| gozes: we don't know what the repo looks like | 20:09 |
|
| we can't say | 20:10 |
|
dogarrhea2
| and then on github, I do a pull request from the remote i shoved changes into, into the main repository. | 20:10 |
|
gozes
| https://bitbucket.org/gozes/dotfiles/ | 20:10 |
| ← opus left | 20:11 |
| → marius joined | 20:11 |
|
gozes
| cmn: will that help? | 20:11 |
|
cmn
| gozes: see what git is telling you that it's not finding | 20:11 |
|
| those are the paths it's looking for | 20:11 |
| ← mabrand left | 20:11 |
|
dogarrhea2
| wut. it appears -f can change the commit tree | 20:11 |
| → ISF joined | 20:12 |
| → metcalfc joined | 20:13 |
|
cbreak
| what's a commit tree? | 20:14 |
|
gozes
| cmn: aha but why is it not find them? if they are all there? | 20:14 |
|
cmn
| they're not | 20:14 |
|
| which is why it's telling you exactly what it's not finding | 20:14 |
|
dogarrhea2
| cbreak isn't this a history of commits? | 20:14 |
|
cbreak
| history is not a tree | 20:15 |
|
dogarrhea2
| hrm? seems like it's a tree. with branching and what not | 20:15 |
|
cbreak
| trees don't have merges | 20:15 |
|
| git has merges. | 20:15 |
|
| not a tree. | 20:15 |
|
| it's a DAG | 20:15 |
| → eletuchy joined | 20:16 |
|
dogarrhea2
| sounds like a tree to me | 20:16 |
|
cmn
| then read up on graph theory | 20:16 |
| → patrick99e99 joined | 20:16 |
|
gozes
| cmn: then can you tell me what can I dot to fix it and what not to do so it doesn't happen againg because I only beeing useing git for a week and I'm lost | 20:16 |
|
| do* | 20:16 |
| ← ReekenX left | 20:17 |
| → pfFred joined | 20:17 |
|
cmn
| set the paths to the correct ones, which git is helpfully complaining about | 20:17 |
| → Chillance joined | 20:17 |
|
patrick99e99
| so.. when I have a submodule.. I make a change to it.. I make a commit... And then I go out of the directory to the super project and git status shows the directory for the submodule modified (new-commits) | 20:17 |
|
cbreak
| dogarrhea2: if merges sound like trees to you, then you've never seen a tree. | 20:17 |
|
patrick99e99
| so do I really have to make commits on the super project saying "module updated" | 20:17 |
|
| every time? | 20:17 |
|
gozes
| cmn: TAHT is the corret path to my repo | 20:17 |
| ← irqq left | 20:18 |
|
cbreak
| patrick99e99: only if you want the parent to refer to the new child commit | 20:18 |
|
| patrick99e99: submodules are strong binding, they know exactly which commit they are developed for | 20:18 |
| → themalkolm joined | 20:18 |
|
cmn
| gozes: the path to your repo is irrelevant | 20:18 |
| ← intransit left | 20:18 |
| → pfFreddd joined | 20:18 |
|
patrick99e99
| cbreak: wouldn't I always want the parent to refer to the latest child ? | 20:18 |
|
cbreak
| no | 20:18 |
|
| of course not | 20:18 |
|
gozes
| or shuled I use $HOME/dontfles? | 20:19 |
|
patrick99e99
| of course not? | 20:19 |
|
| really? why | 20:19 |
|
cmn
| gozes: set the paths to the correct ones, which git is helpfully complaining about | 20:19 |
|
cbreak
| what if I make an experimental commit in some other branch that doesn't compile? | 20:19 |
|
| what if I make a new release with a new api that is not compatible? | 20:19 |
| ← pfFredd left | 20:19 |
|
gozes
| cmn: I don't know what you mean by that do a sub add againg? | 20:20 |
| ← kevwil left | 20:20 |
| → aselder joined | 20:20 |
|
cmn
| adding a new entry will leave the wrong ones in | 20:20 |
|
| change the entries so it fits what git is telling you it wants | 20:20 |
|
| use relative paths from the repo root path | 20:21 |
|
devinrhode2
| Hey, all little new to certain parts of git. I created a fork of a repo, cloned, made my changes, submitted a pull request, but I was asked to rebase my changes. I figured that out fine and squashed my commits all into one. But I can't sync these changes to my local repo to my remote fork on github. | 20:21 |
|
EugeneKay
| devinrhode2 - push -f | 20:21 |
| ← pfFred left | 20:21 |
| ← arietis left | 20:21 |
|
cbreak
| devinrhode2: push into a different branch | 20:21 |
| ← aselder left | 20:22 |
| → aselder joined | 20:22 |
| → banghouse joined | 20:22 |
|
gozes
| cmn: I just notices that the path are duplicatd for some reason | 20:22 |
| ← Ivo left | 20:23 |
| ← tworkin left | 20:23 |
| ← converge left | 20:23 |
| ← dfr|mac left | 20:23 |
|
cmn
| you mean because you're using an absolute path in the submdoule name? | 20:23 |
| ← guns left | 20:23 |
|
cmn
| the reason is that you put it in twice | 20:23 |
|
| once worng | 20:23 |
|
| the onther slightly less wrong | 20:23 |
|
gozes
| cmn: should I change both the the one insde [] and the on in path = or both? | 20:24 |
|
| nvm I see it now | 20:24 |
| ← marhaban left | 20:25 |
|
gozes
| but why did that happen what is the corret way? to make the submoduel | 20:25 |
| ← nitind left | 20:25 |
| → BiggFREE joined | 20:26 |
|
gozes
| giveing it the relative path in the scope of my repo? | 20:27 |
| ← anddam left | 20:27 |
| ← skyf left | 20:29 |
|
devinrhode2
| @EugeneKey, git push -f tells me everything is up to date, but the commits don't change on github | 20:29 |
| ← Sky[x] left | 20:30 |
|
canton7
| devinrhode2, github is having problems updating its ui atm | 20:30 |
| ← aruiz left | 20:30 |
|
canton7
| it's been going since the middle of yesterday | 20:30 |
| ← e66 left | 20:30 |
|
canton7
| loads of people have been complaining in #github | 20:30 |
|
devinrhode2
| oh.. so ok | 20:30 |
|
canton7
| git ls-remote origin will show the sha1's of the branches on github. if they match, github's up to date | 20:31 |
|
EugeneKay
| `git ls-remote origin`, see if the.... yeah, that. | 20:31 |
| ← CollinPDX left | 20:31 |
| ← dv310p3r left | 20:31 |
|
canton7
| heh, jinx! | 20:31 |
| ChanServ set mode: +o | 20:32 |
| EugeneKay kicked canton7 (jinx) | 20:32 |
| EugeneKay set mode: -o | 20:32 |
| → canton7 joined | 20:32 |
|
EugeneKay
| <_< | 20:32 |
|
devinrhode2
| is there a way to see the remote git log? | 20:32 |
| ChanServ set mode: +q | 20:32 |
|
EugeneKay
| devinrhode2 - log of what? Commits are commits. | 20:32 |
| ← jrajav left | 20:32 |
|
canton7
| REVENGE | 20:32 |
| ChanServ set mode: -q | 20:32 |
|
| EugeneKay flaunts his +v at canton7 | 20:32 |
|
devinrhode2
| the commit log `git log` | 20:32 |
|
| on the remote - since the ui isn't updating | 20:33 |
|
EugeneKay
| devinrhode2 - so long as you have the same SHA locally as remote(per git ls-remote), it's the same. | 20:33 |
|
| See also !lol | 20:33 |
|
gitinfo
| A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all | 20:33 |
|
gozes
| cmn: I fix it but now when I do a update or sync it does noting | 20:33 |
|
canton7
| EugeneKay, damn, you were modded before me. you win! | 20:33 |
|
gozes
| no even an error(wicht is good I guess) | 20:34 |
|
| EugeneKay plays for keeps | 20:34 |
| → dodo joined | 20:34 |
|
gozes
| cmn: nvm thenk you for your help an pasints with a new git user :-) | 20:35 |
|
| EugeneKay wanders off to find beer | 20:35 |
|
devinrhode2
| thanks though guys, you're awesome! | 20:35 |
|
| fyi verified the SHA's are the same | 20:36 |
| ← jjbohn left | 20:36 |
| ← erichynds left | 20:36 |
| → jjbohn joined | 20:36 |
| ← macrover left | 20:36 |
| ← devinrhode2 left | 20:37 |
| ← gozes left | 20:37 |
| → orated joined | 20:37 |
| ← les_sylvains left | 20:37 |
| → Macca- joined | 20:37 |
| → torjus_ joined | 20:38 |
|
Macca-
| git branch -r is listing branches that i deleted from another that accesses this repo. is there anything i need to do locally to make it aware of that? | 20:38 |
| ← ISF left | 20:38 |
|
m1sc
| Macca-: see man git-fetch --prune | 20:39 |
|
gitinfo
| Macca-: the git-fetch manpage is available at http://git-scm.com/docs/git-fetch.html | 20:39 |
|
Macca-
| dziękuje | 20:39 |
| → Chat8791 joined | 20:40 |
| ← _iron left | 20:40 |
| ← Chat8791 left | 20:40 |
| ← shruggar left | 20:40 |
| ← tplaner_ left | 20:40 |
| → jbohn joined | 20:40 |
| → torjus__ joined | 20:41 |
| → osse_ joined | 20:41 |
|
Macca-
| dowidzenia | 20:41 |
| ← Macca- left | 20:41 |
| → vdv joined | 20:41 |
| ← jbohn left | 20:41 |
| → jbohn joined | 20:42 |
| ← eletuchy left | 20:42 |
| ← torjus_ left | 20:42 |
| → Goplat joined | 20:42 |
|
ciaranm
| can i git-archive myself a tarball, excluding everything under a particular directory? | 20:42 |
| ← bartek left | 20:42 |
| ← jjbohn left | 20:42 |
| → bartek joined | 20:43 |
|
adhawkins
| ciaranm: man git-archive | 20:43 |
|
gitinfo
| ciaranm: the git-archive manpage is available at http://git-scm.com/docs/git-archive.html | 20:43 |
|
adhawkins
| Seems to suggest you can specify a list of paths to include | 20:43 |
|
| So that'd do it, although not exactly what you want. | 20:43 |
|
talexb
| ciaranm, Alternatively, you could adjust the tarball after archiving. | 20:44 |
|
Plouj
| does git reflog work with tags? I'm not sure what it's showing me. I see one line of 6-char long hashes | 20:44 |
|
ciaranm
| adhawkins: that's a lot of specifying | 20:44 |
| → bfig joined | 20:44 |
|
ciaranm
| talexb: mm, not really. the directory i want to exclude is a couple of hundred megs. the rest is ten k... | 20:44 |
| ← maletor left | 20:44 |
|
bfig
| hello, i just made a mistake and `checkout -b origin/somebranch`. is it safe to do `git branch -d origin/somebranch` ? | 20:44 |
|
| (what i really wanted to do was download the branch from origin and work on it) | 20:45 |
|
ciaranm
| bfig: git checkout master | 20:45 |
|
| er, with a -b. | 20:45 |
| ← bartek left | 20:45 |
|
bfig
| what? | 20:46 |
|
| *why ? | 20:46 |
|
adhawkins
| ciaranm: How about something like git archive .... `ls | grep -v path-to-exclude` | 20:46 |
|
canton7
| no... | 20:46 |
|
adhawkins
| (I haven't tried it, but don't see why it shouldn't work. | 20:46 |
| → idletom joined | 20:47 |
|
ciaranm
| adhawkins: misses .files! | 20:47 |
|
adhawkins
| ls -a? | 20:47 |
|
ciaranm
| includes too much! | 20:47 |
| → serbang joined | 20:47 |
|
ciaranm
| alternatively i could just demand my supervisor finally gets himself a github account so i don't have to keep mailing him code | 20:47 |
|
| ciaranm mutters | 20:47 |
| ← kiffer84 left | 20:48 |
| → bartek joined | 20:48 |
| → diegoviola joined | 20:48 |
|
cmn
| there is no reason why you need that | 20:49 |
| ← g3orge left | 20:49 |
|
cmn
| just send the repo, a bundle, or put it in your server | 20:49 |
|
ciaranm
| oh, he doesn't use git either | 20:49 |
|
| or any form of version control other than lots of datestamp-named directories | 20:49 |
| → dfr|mac joined | 20:50 |
|
canton7
| bfig, !local_branch_from_remote | 20:50 |
|
gitinfo
| bfig: The following commands are all equivalent, assuming <branch> doesn't yet exist: 'git checkout -b <branch> <remote>/<branch>', 'git checkout -t <remote>/<branch>', 'git checkout <branch>'. The latter invokes some magic. | 20:50 |
| → maletor joined | 20:50 |
| ← zastaph left | 20:50 |
|
canton7
| (sorry, was distracted) | 20:50 |
|
cbreak
| ciaranm: just write yourself a script that mails him git format-patch output | 20:50 |
| ← jperry2_ left | 20:50 |
|
milki
| magic is awesome | 20:50 |
|
ciaranm
| cbreak: i was going to say "he doesn't believe in diff", but i've converted him on that at least... | 20:51 |
|
bfig
| canton7, thanks! | 20:51 |
|
milki
| lol | 20:51 |
| ← serbang left | 20:51 |
|
cbreak
| ciaranm: faith is irrelevant | 20:51 |
|
| the reality is unhampered by something as worthless as that | 20:52 |
|
ciaranm
| i was so happy the day he said he'd rm -fr'ed in the wrong directory. until he ruined my joy by remembering he had a week old copy on a memory stick. | 20:52 |
|
cbreak
| backups ftw | 20:52 |
|
| no, honestly | 20:53 |
| ← ThePendulum left | 20:53 |
|
cbreak
| just mail him the patches of all commits | 20:53 |
|
| that'll burry him in mails | 20:53 |
| → AdrienBrault joined | 20:53 |
| ← philips_ left | 20:53 |
| ← steveoliver left | 20:53 |
| ← dogarrhea2 left | 20:54 |
|
milki
| LOL | 20:54 |
| → philips_ joined | 20:55 |
| ← dvide left | 20:55 |
|
cbreak
| ciaranm: alternatively, send him git archive --format=tar versions of each commit | 20:55 |
|
| that should exhaust his mailbox quota rather quickly | 20:55 |
|
ciaranm
| i could mail him git archives of each commit, including the hundred plus megs of test data | 20:55 |
|
milki
| doooit | 20:55 |
|
cbreak
| yes. | 20:55 |
| ← sbodo left | 20:56 |
|
ciaranm
| he was talking about reintroducing floggings for students a few days back. i'd better not. | 20:56 |
|
milki
| haha | 20:56 |
| ← vdv left | 20:56 |
|
cbreak
| is that related with blogging? | 20:56 |
| → serbang joined | 20:57 |
| ← geekbri left | 20:57 |
| ← stewbydoo left | 20:58 |
|
esc
| ciaranm: i have tgz = "!bn=$(basename $(pwd)) && git archive --format=tar --prefix=$bn/ HEAD | gzip -c > $bn-$(git describe --tags --always).tgz" | 20:59 |
| ← serbang left | 20:59 |
| → tworkin joined | 21:00 |
| → SEToChan joined | 21:00 |
|
esc
| maybe you can write a shell-function to convert includes to excludes and throw it in there? | 21:00 |
| → eletuchy joined | 21:00 |
|
SEToChan
| i have repository where i develop my site and then i create another with "git init --bare" and then i want push from develop to this, named "hub" | 21:01 |
|
| and i do "git push hub maser" | 21:01 |
| ← kermit left | 21:01 |
|
SEToChan
| but got error | 21:01 |
|
canton7
| what error? | 21:01 |
| ← hyperboreean left | 21:01 |
|
esc
| ciaranm: maybe a $(date) too | 21:01 |
| ← whitman left | 21:01 |
|
SEToChan
| remote: error: refusing to update checked out branch: refs/heads/master | 21:01 |
|
| remote: error: By default, updating the current branch in a non-bare repository | 21:01 |
|
| remote: error: is denied, | 21:01 |
|
canton7
| SEToChan, well you didn't create it with --bare then, did you? | 21:01 |
|
SEToChan
| why it says "non-bare rep is denied" when repo is bare | 21:01 |
|
| yes i did | 21:01 |
|
| i did | 21:01 |
|
canton7
| try this: cd bare_repo; ls | 21:01 |
| ← tomislater left | 21:02 |
|
canton7
| what do you see? | 21:02 |
|
ciaranm
| esc: mm. this is entering "less work for me to convince him to lrn2git" territory | 21:02 |
| → kermit joined | 21:02 |
|
SEToChan
| http://pastebin.com/83CFwD3k | 21:02 |
|
| it is indeed bare | 21:03 |
|
esc
| ciaranm: yeah, bummer it can't do an exclude | 21:03 |
|
canton7
| SEToChan, looks sensible. look in config -- what's the value of core.bare ? | 21:03 |
| → hyperboreean joined | 21:03 |
|
esc
| ciaranm: subtree split the directory, and then add it as a submodule (j/k ;)) | 21:03 |
|
ciaranm
| heh | 21:03 |
|
SEToChan
| where i can see this config? | 21:03 |
| → jjbohn joined | 21:03 |
|
canton7
| SEToChan, there's a file called 'config'. it's in your pastebin | 21:03 |
|
ciaranm
| esc: that might actually not be totally utterly crazy | 21:03 |
|
| the test data is "standard" | 21:04 |
| → jcao219 joined | 21:04 |
| ← bartek left | 21:04 |
|
SEToChan
| ah | 21:04 |
|
esc
| git subtree is part of mainline git now IIRC | 21:04 |
|
cmn
| it's in contrib/ | 21:04 |
| ← cannonball left | 21:05 |
| ← kerrick left | 21:05 |
|
SEToChan
| canton7: http://pastebin.com/vvLc0qNB | 21:05 |
| ← jbohn left | 21:05 |
|
SEToChan
| bare = true | 21:05 |
|
esc
| not sure what git archive will do to submodules though... | 21:05 |
|
canton7
| SEToChan, so it thinks it's a bare repo. in the remove you're pushing from, 'git remove -v' | 21:05 |
|
| * 'git remote -v' | 21:05 |
| → gusnan joined | 21:05 |
|
SEToChan
| you think i might push to another repo? | 21:06 |
|
esc
| i just did an online coding test | 21:06 |
|
| with codility | 21:06 |
|
canton7
| if the repo you should be pushing to is indeed bare, then the next option is that you're not pushing to that repo, yes | 21:06 |
|
esc
| one of the questions was: reverse a string, use python, you have 30 minutes to complete this | 21:07 |
|
SEToChan
| shit | 21:07 |
|
| you right | 21:07 |
|
canton7
| correct guess? :) | 21:07 |
|
esc
| i was like, WTF? | 21:07 |
|
canton7
| heh | 21:07 |
|
SEToChan
| yes | 21:07 |
|
| i should guess | 21:07 |
|
| so stupid i am | 21:07 |
|
| thanx | 21:07 |
|
canton7
| heh, it happens :) | 21:07 |
|
| np | 21:07 |
|
esc
| so i put s[::-1] and clicked submit after 1 minute | 21:07 |
| ← Mantic0re left | 21:08 |
|
cbreak
| and then it got rejected for lack of input handling? :) | 21:08 |
|
esc
| i dunno, did i miss something? | 21:08 |
| → serbang joined | 21:08 |
|
cbreak
| like, reading the string from somewhere? | 21:08 |
|
esc
| oh, right, it was return s[::-1], silly me :) | 21:09 |
|
cbreak
| or were you not supposed to write executable code? | 21:09 |
|
canton7
| I swear those things are sometimes written by people with no knowledge of the language in question | 21:09 |
|
esc
| no, just the function | 21:09 |
|
| the first was OK, some random array iteration | 21:09 |
|
| the second was real hard, a heavy integer question | 21:09 |
|
| and then that.. ??? | 21:09 |
|
| if you have to do it in c, ok | 21:10 |
| → snearch joined | 21:10 |
| ← GFLJay left | 21:10 |
| ← hinst left | 21:10 |
|
esc
| the requested runtime was O(n) | 21:10 |
|
| hmmm... | 21:11 |
| ← pandeiro left | 21:11 |
|
| esc scratches his beard | 21:11 |
| ← duckxx left | 21:11 |
| ← tworkin left | 21:12 |
| → Pupeno_W joined | 21:12 |
| ← griotspeak left | 21:12 |
| → duckxx joined | 21:12 |
|
cbreak
| esc: that's probably the best case | 21:13 |
|
| unless the representation intrinsically supports directions | 21:14 |
|
esc
| i guess it depends on if you know the length | 21:14 |
|
cbreak
| esc: for example, with the unicode right-to-left direction changing character | 21:14 |
|
esc
| lol | 21:14 |
|
| yeah, maybe it was something about hat | 21:14 |
|
cbreak
| not sure if python supports that | 21:15 |
| ← devth left | 21:15 |
| ← torjus__ left | 21:15 |
| ← eijk left | 21:15 |
| ← tjadc left | 21:16 |
|
cbreak
| seems it is | 21:17 |
|
| http://www.fileformat.info/info/unicode/char/202e/index.htm | 21:17 |
| → tjadc joined | 21:17 |
|
lectrick
| So today I learned about the -p option to "git add" which will let you stage changes to individual files in piecemeal fashion. I'm wondering if there is a similar facility to "git checkout" to *discard* individual changes to a file in piecemeal fashion? | 21:17 |
| ← rubious left | 21:17 |
|
canton7
| lectrick, checkout -p, yup | 21:17 |
|
lectrick
| NICE. | 21:17 |
|
| is that recent? I think I tried that once and it didn't work :/ | 21:18 |
|
canton7
| fairly I think | 21:18 |
|
lectrick
| git version 1.7.12 | 21:18 |
| ← snearch left | 21:18 |
| ← AaronMT left | 21:18 |
|
mgaunard
| lectrick: how does add -p work? | 21:19 |
| ← beneggett left | 21:19 |
| ← sq-one left | 21:19 |
| → lb joined | 21:19 |
|
lectrick
| mgaunard: It's fucking magic. It will bring up a menu driven interface to stage only parts of the file(s) | 21:19 |
| ← stopbit_ left | 21:19 |
| → psoo joined | 21:19 |
| → jbohn joined | 21:19 |
|
lectrick
| Great when you have things like debug harness code around actual changed code | 21:19 |
|
mgaunard
| seems text-driven to me | 21:20 |
|
| no graphical menu | 21:20 |
| → stopbit_ joined | 21:20 |
| ← duckxx left | 21:20 |
| → devth joined | 21:20 |
|
lectrick
| menu driven does not mean GUI :) | 21:20 |
| ← stopbit_ left | 21:20 |
| → griotspeak joined | 21:20 |
| → dr_lepper joined | 21:20 |
| ← dr_lepper left | 21:20 |
| → dr_lepper joined | 21:20 |
| ← skorgon left | 21:21 |
|
cbreak
| This is a test | 21:22 |
|
mgaunard
| cbreak: #test | 21:22 |
| ← jjbohn left | 21:22 |
|
cbreak
| hmm. Wonder if that came accross | 21:22 |
| ← Jippi left | 21:22 |
|
DeadZen
| this is only a test... if this were an actual emergency this would be followed by frantic screaming | 21:23 |
| ← d3vic3 left | 21:23 |
| ← TeckniX left | 21:24 |
|
cbreak
| lectrick: add -p and checkout -p are fairly old | 21:24 |
| ← whaley left | 21:25 |
| ← mmc1 left | 21:26 |
| ← lb left | 21:29 |
| → ipalaus joined | 21:33 |
| → lresende joined | 21:33 |
| ← drizzd_ left | 21:34 |
| ← j416 left | 21:34 |
| ← gfontenot left | 21:34 |
| → opus joined | 21:34 |
| → rr3 joined | 21:35 |
|
rr3
| hi newbie needs help. i have project with just one branch 'master'. I made some changes. i don't want to commit these changes to the 'master' branch, rather I want to commit them to a new branch 'javascript' and commit the changes to them. | 21:37 |
| ← etcetera left | 21:37 |
|
canton7
| rr3, 'git checkout -b javascript', then add/commit them | 21:38 |
| ← maletor left | 21:39 |
| ← john_anderson left | 21:39 |
|
rr3
| :canton7, ty | 21:40 |
| → maletor joined | 21:41 |
| → dsirijus joined | 21:43 |
| → babilen joined | 21:44 |
| → aspotashev joined | 21:44 |
| → giallu joined | 21:44 |
| babilen → Guest65475 | 21:45 |
| ← Guest65475 left | 21:45 |
| ← themalkolm left | 21:45 |
| → BCook joined | 21:45 |
| → babilen_ joined | 21:45 |
| ← rr3 left | 21:45 |
| → moscownights joined | 21:45 |
| babilen_ → babilen | 21:46 |
| ← derpops left | 21:46 |
| → iron_houzi joined | 21:47 |
| ← moscownights left | 21:47 |
| ← mwolff6 left | 21:47 |
| → Tohuw joined | 21:47 |
| gitinfo set mode: +v | 21:48 |
|
iron_houzi
| OK - why is it that two different files that contain the same data (i vimdiff'ed them) will compile differently? One is inside a git repo on a new branch I'm working on, the other file is just in my home/ ..Any clues? | 21:48 |
|
BCook
| ahah! | 21:48 |
| ← maletor left | 21:48 |
|
BCook
| is anyone around that can offer a quick bit of help with repository setup? i installed Git on both my laptop and the server, and created a repository on the server... what i need help with now is accessing the server's repository, but i'm not sure how to add that from my laptop... | 21:48 |
|
canton7
| BCook, is the server running ssh? | 21:49 |
|
iron_houzi
| BCook: Are you using linux or SSH? | 21:49 |
|
BCook
| they are both windows machines on the same company network | 21:49 |
| → thierryp joined | 21:49 |
|
iron_houzi
| Never used git with Windows - sorry.. | 21:49 |
|
canton7
| BCook, to serve a git repo, you need either ssh, or a configured http server | 21:49 |
|
BCook
| i see... the server it is installed is running SQL Server 2008 and IIS which hosts some ASP.net pages i created | 21:50 |
|
nevyn
| BCook: so a repository in git is every checkout. | 21:50 |
|
Tohuw
| For whatever reason, I'm having trouble wrapping my head around the "don't pull, merge" philosophy. I always pull (sometimes pull --rebase), and I don't typically want merge commits in the history. Can you help me understand a general workflow without using pull? I manage my own GIT repo, and test locally from my system by pulling in the latest changes. If a pull is just a fetch and merge, why is not using pull superior? | 21:50 |
| → maletor joined | 21:50 |
|
BCook
| can i configure IIS somehow to host the repository on the server? | 21:51 |
|
nevyn
| BCook: most git use cases involve multiple repositories. | 21:51 |
| ← brasom left | 21:51 |
| → brasom joined | 21:51 |
|
nevyn
| BCook: fighting windows platform issues for hosting git is probably lose. can you get user access to a linux host? | 21:52 |
| ← ShadeTornado left | 21:52 |
|
cbreak
| no need for a web server for hosting git repos | 21:53 |
|
| just use an sshd | 21:53 |
| → newbie999|3 joined | 21:53 |
| → acl_ joined | 21:53 |
|
Tohuw
| BCook: What is your use case for presenting the repo over the web? | 21:53 |
|
BCook
| we're a windows based shop so nothing else available | 21:53 |
|
nevyn
| BCook: this might help http://therightstuff.de/CommentView,guid,b969ea4d-8d2c-42af-9806-de3631f4df68.aspx | 21:53 |
| → PatF joined | 21:54 |
| → etcetera joined | 21:54 |
| ← thierryp left | 21:54 |
| ← upper2bits left | 21:54 |
|
BCook
| i don't really need it over the web, i'm just trying to get away from a fileshare based development environment | 21:54 |
|
nevyn
| it involves cygwin and stuff tho. | 21:54 |
|
| right. | 21:54 |
|
canton7
| Tohuw, !pull | 21:54 |
|
gitinfo
| Tohuw: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/ | 21:54 |
|
Tohuw
| BCook: Can't users just use the Git client to obtain the files? | 21:54 |
| ← dr_lepper left | 21:54 |
|
Tohuw
| canton7: Thanks | 21:54 |
| → john_anderson joined | 21:54 |
|
canton7
| Tohuw, so the "don't pull; merge" means "don't just run 'git pull'. run 'git fetch', inspect origin/branch_of_interest to see if you need to merge, and if you do, run an explicit 'git merge origin/branch_of_interest' | 21:55 |
|
nevyn
| Tohuw: he wants to use git to manage asp code in a windows shotp | 21:55 |
|
BCook
| that is correct nevyn | 21:55 |
|
iron_houzi
| Sorry for nagging, but is there any reason why two different files with the same content should compile differently if one is inside a git-repo and the one that works is not in a git repo ?? | 21:55 |
| ← newbie999 left | 21:55 |
|
nevyn
| BCook: if you want to use a central repo workflow then I'd strongly reccomend using gitolite | 21:55 |
|
| (note that this is a workflow choice not a tool imperative) | 21:56 |
| → guampa joined | 21:56 |
|
canton7
| iron_houzi, are they the same? md5 the | 21:56 |
|
| *them | 21:56 |
| → Uriziel joined | 21:57 |
|
iron_houzi
| no - one contains some commented lines .. | 21:57 |
| ← serbang left | 21:57 |
|
BCook
| i'm really the only one developing so this isn't crucial, just a desire to get in place so that i can have version control | 21:57 |
| ← Destos left | 21:57 |
|
iron_houzi
| should I try to make them exactly the same? | 21:57 |
|
Uriziel
| Is there an easy way to "unfork" the project? | 21:57 |
|
BCook
| without just making a copy of the project, and updating the copy | 21:57 |
|
canton7
| iron_houzi, well it sounds like they're not exactly the same tbh. run a diff on them | 21:57 |
|
Uriziel
| As in, taking down "fork" status" | 21:57 |
|
nevyn
| BCook: also you're being a tad imprecise wiht terminology in that with git.. a repository on the server is no different to a repository on your workstation | 21:57 |
|
canton7
| Uriziel, on github? don't think so | 21:57 |
|
| oh, I think you might be able to email them and get them to do it | 21:58 |
|
| or delete/re-create | 21:58 |
|
Uriziel
| Oh, yeah, forgot, on github indeed | 21:58 |
|
nevyn
| BCook: so you could just use git on your workstation (if you're the only dev) and copy stuff to a share as a publish step | 21:58 |
|
Uriziel
| kk | 21:58 |
|
iron_houzi
| canton7: OK, I'll take out the commented stuff and do a new compare.. | 21:58 |
| ← PatF left | 21:58 |
|
BCook
| sorry, i'm not familiar with all the terminology... i've only used SVN in the past and it's a bit quirky | 21:58 |
| pdelvo → pdelvo|afk | 21:58 |
| ← cakehero left | 21:58 |
| ← resmo__ left | 21:58 |
|
nevyn
| BCook: you really really should read !book | 21:59 |
|
gitinfo
| BCook: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable | 21:59 |
|
nevyn
| BCook: if you just want to get version control going for you for code you're the only one changing. | 21:59 |
| ← walters left | 22:00 |
|
nevyn
| then dealing with multiple repos and push and blah is possibly more effort than it's worth. | 22:00 |
|
| you do need version control in this case. | 22:00 |
| ← gusnan left | 22:01 |
|
BCook
| you're probably right | 22:01 |
| ← jefferai left | 22:01 |
|
nevyn
| but you don't need to have any more than one repo | 22:01 |
|
BCook
| the odds of our small shop hiring another dev is unlikely | 22:01 |
|
nevyn
| so when that happens you can always just put a central repo server in to be a co-ordination point | 22:01 |
| → jefferai joined | 22:01 |
|
nevyn
| and keep all the history and stuff | 22:01 |
|
BCook
| yeah | 22:02 |
|
nevyn
| but you don't need to do it now | 22:02 |
| → gusnan joined | 22:02 |
| ← dzonder left | 22:02 |
|
nevyn
| BCook: so a git clone. is not like a checkout | 22:02 |
| ← allquixotic_ left | 22:02 |
|
nevyn
| gtg. | 22:02 |
|
| meeting in 80mins and have to get to the office still. | 22:02 |
| → allquixotic joined | 22:02 |
|
BCook
| if i put a local repo and work from that, i'm guessing i would have two folders... a dev folder that i work out of with my MS tools, and then commit those changes to my repo (which is my production directory then) | 22:03 |
| ← TTilus left | 22:03 |
|
BCook
| and then i can copy the files from my repo to the fileshare for others as needed | 22:03 |
|
| am i piecing that together correctly? | 22:04 |
| ← hobodave left | 22:04 |
| → ShaunR joined | 22:04 |
| → aruiz joined | 22:04 |
|
iron_houzi
| canton7: Sorry - getting late and I'm getting blind.. thanks for your help though | 22:05 |
|
BCook
| d'oh, i missed that nevyn left... oh well, i think i get the general idea | 22:05 |
|
| i gotta run too | 22:05 |
|
canton7
| heh, no worries. i'm almost certain that git has nothing to do with it, and i'm fairly sure there's a little change somewhere (not a comment) that you've missed. a normal diff will spot it, of course | 22:05 |
|
pantsman
| BCook, keep your repository in your working directory, not wherever you deploy it... use a deployment tool to do that | 22:06 |
| ← BCook left | 22:06 |
|
pantsman
| noooooooooo, he will do it wrong now | 22:06 |
| ← alunduil left | 22:07 |
| → TTilus joined | 22:07 |
| → kiffer84 joined | 22:08 |
| → poseidon joined | 22:10 |
| ← martinjlowm left | 22:10 |
| ← tfitts left | 22:16 |
| ← lresende left | 22:16 |
| ← bapa left | 22:18 |
| ← BiggFREE left | 22:18 |
| → adamm joined | 22:20 |
| → apok_ joined | 22:21 |
| ← apok left | 22:24 |
| apok_ → apok | 22:24 |
| → beneggett joined | 22:24 |
| → BiggFREE joined | 22:24 |
| ← robde left | 22:24 |
| → yoklov joined | 22:25 |
| ← macabre left | 22:26 |
| → mischief joined | 22:27 |
| ← anth0ny_ left | 22:28 |
| → madewokherd joined | 22:28 |
| ← trungqtran left | 22:29 |
| ← avtobiff left | 22:29 |
| ← grey left | 22:29 |
| ← yoklov left | 22:29 |
| → grey joined | 22:30 |
| → ppppaul joined | 22:30 |
| → yoklov joined | 22:30 |
| ← quazimodo left | 22:31 |
|
Plouj
| how can I push multiple refs under a namespace to a given remote? | 22:31 |
| ← pppaul left | 22:31 |
| ppppaul → pppaul | 22:31 |
| → rubious joined | 22:31 |
| ← psoo left | 22:32 |
| ← AdrienBrault left | 22:34 |
| ← grey left | 22:34 |
| → kcj joined | 22:35 |
| → grey joined | 22:35 |
| ← chuckharmston left | 22:36 |
|
mischief
| does anyone know of a good resource for learning about using git with multiple developers? | 22:36 |
| ← scarlisle left | 22:37 |
| → dcs_ joined | 22:37 |
| → jason237 joined | 22:38 |
| ← hellopat left | 22:38 |
| → erichynds joined | 22:39 |
| ← sgambino left | 22:40 |
| ← aruiz left | 22:41 |
| ← jason237 left | 22:42 |
| → orated_ joined | 22:43 |
| ← orated left | 22:43 |
| ← etcetera left | 22:44 |
| → derpops joined | 22:45 |
| ← scelis left | 22:45 |
| ← Mpenzi left | 22:46 |
| FrenkyNet → FrenkyNet|away | 22:46 |
| → angelsl joined | 22:46 |
| ← glennpratt left | 22:47 |
| → serbang joined | 22:47 |
| ← Synthead left | 22:48 |
| ← griotspeak left | 22:49 |
| ← tdowg1 left | 22:49 |
| → etcetera joined | 22:50 |
| → trungqtran joined | 22:50 |
| ← TheJH left | 22:53 |
| ← angelsl left | 22:53 |
| ← beneggett left | 22:54 |
|
nevyn
| !workflow | 22:56 |
|
gitinfo
| Finding the right workflow for you is critical for the success of any SCM project. Git is very flexible with respect to workflow. See http://sethrobertson.github.com/GitBestPractices/#workflow for a list of references about choosing branching and distributed workflows. | 22:56 |
| ← pppaul left | 22:57 |
| ← mtlife left | 22:58 |
| → jperry2_ joined | 23:00 |
| → _Vi joined | 23:03 |
| ← SirCmpwn left | 23:04 |
| → glennpratt joined | 23:04 |
| SeoZ-work[AWAY] → SeoZ | 23:04 |
| ← serbang left | 23:06 |
| ← derpops left | 23:07 |
| ← Davey left | 23:08 |
| → jason237 joined | 23:09 |
| ← dfr|mac left | 23:09 |
| → derpops joined | 23:10 |
| ← leen^| left | 23:10 |
| ← juvenal left | 23:11 |
| ← apok left | 23:11 |
| → apok joined | 23:12 |
| SirCmpwn_ → SirCmpwn | 23:12 |
| ← Anders_J left | 23:12 |
| ← eletuchy left | 23:13 |
| ← derpops left | 23:14 |
| → kvanderw_ joined | 23:16 |
| → wereHamster joined | 23:18 |
| → Leonidas joined | 23:18 |
| ← kvanderw_ left | 23:18 |
| ← emdete left | 23:18 |
| → qknight joined | 23:18 |
| → emdete joined | 23:19 |
| → derpops joined | 23:20 |
| ← guide left | 23:20 |
| ← kvanderw left | 23:20 |
| → guide joined | 23:20 |
| ← erichynds left | 23:21 |
| → avjpl joined | 23:22 |
| ← avjpl left | 23:22 |
| ← SEToChan left | 23:23 |
| → macrover joined | 23:23 |
| → serbang joined | 23:24 |
| ← zastern left | 23:25 |
| ← iocor left | 23:25 |
| ← techhelp left | 23:26 |
| → Bert joined | 23:27 |
| ← giallu left | 23:28 |
| ← john_anderson left | 23:28 |
| ← jbohn left | 23:28 |
| ← pantsman left | 23:29 |
| gitinfo set mode: +v | 23:29 |
| → iocor joined | 23:29 |
|
milki
| workflows are hard | 23:30 |
|
| issues as a to do list is kind of interesting | 23:31 |
| ← pfFreddd left | 23:32 |
| → steveoliver joined | 23:32 |
| ← macrover left | 23:33 |
| ← fisted left | 23:33 |
| → lresende joined | 23:33 |
| → yaroslav joined | 23:34 |
| ← derpops left | 23:36 |
| ← chrisgeorge left | 23:37 |
| ← friskd left | 23:38 |
| → ISF joined | 23:39 |
|
SaCu
| Is there a special reason, why 'git tag -l "foo/*"' invokes fnmatch() without the FMN_PATHNAME option, essentially disabling the "**" match? | 23:42 |
|
milki
| thats a better question for the mailing list | 23:42 |
|
SaCu
| probably :) | 23:43 |
| ← jason237 left | 23:43 |
| ← panpainter left | 23:43 |
| → eletuchy joined | 23:43 |
| ← cbreak left | 23:44 |
| ← yaroslav left | 23:44 |
| → alunduil joined | 23:44 |
| ← eletuchy left | 23:45 |
| → fisted joined | 23:47 |
| → cooloney joined | 23:47 |
| → warthog9 joined | 23:47 |
| → ahferroin7 joined | 23:48 |
| ← thiago left | 23:48 |
| ← diegoviola left | 23:48 |
| → gavin_huang joined | 23:49 |
| → gitinfo joined | 23:50 |
| ChanServ set mode: +o | 23:50 |
| → thierryp joined | 23:50 |
| ← thierryp left | 23:55 |
| ← glennpratt left | 23:55 |
| → rchavik joined | 23:56 |
| ← rchavik left | 23:56 |
| → rchavik joined | 23:56 |
| ← gitinfo left | 23:58 |
