IRCloggy #git 2012-06-21

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2012-06-21

→ chrisgeorge joined00:00
→ mcluskydodallas joined00:00
← Leemp2 left00:00
JoeCoolDesk Move the entire project, except the module, down to the module level, (git add), then move the original module back to where it was relative to the project, then move the entire git repo back to where the project was?00:00
cmn yes00:01
that way you might get a sensible history00:01
← harish left00:01
cmn alternatively, what I said above00:01
JoeCoolDesk OK this is going to be interesting...00:01
cmn which is doing it in the opposite direction00:01
but the changes of making sensible history are lower00:01
← twodollarz left00:02
→ PrinceAMD- joined00:02
→ mtx_init joined00:02
← fir31ight left00:05
→ fir31ight joined00:05
→ Leemp joined00:05
← anothervenue left00:06
← PrinceAMD- left00:06
Leemp I'm receiving a "non-fast-forward" error from a *new* project on github.. is that even possible?00:06
cmn yes00:08
rchavik gitignore stuff00:08
avpx Leemp: IIRC, new projects on github are often set up with a README00:08
This would be an initial commit that isn't an ancestor of the commit you are pushing00:09
Leemp Must be new, because i don't think i'm doing anything different00:09
cmn you can ask it to pre-fill it with some ignore stuff and other language-specific settings00:09
avpx: they don't have a README00:09
Leemp Normally i create a local project and then push00:09
avpx Leemp: fetch and see what the commit(s) is/are00:09
← juvenal left00:09
Leemp avpx: Goo idea00:10
good*00:10
→ k1t joined00:10
avpx git fetch && git log origin/master00:10
(Sorry, not sure how familiar you are with git)00:10
k1t hey folks, can i do a 'git reset FILE' on each file with a given extension?00:10
Leemp avpx: Don't be, my knowledge is very narrow and limited to my workflow. This is outside heh :)00:10
← twisted` left00:11
→ juvenal joined00:11
→ setmeaway joined00:11
→ twisted` joined00:12
avpx k1t: Wildcards?00:12
k1t: git reset *.txt00:12
k1t avpx: yep, that's the word.. and great, thx00:12
← Caius left00:12
avpx Are you just trying to get the files back to their state at HEAD?00:13
k1t i did a 'git add .' then 'git commit ...' and noticed it added a bunch of .png files it shouldn't have. i want to undo adding all those files before i push00:14
sorry, still catching onto the jargon00:14
git jargon ;p00:14
bremner k1t: !fixup00:15
gitinfo k1t: 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!00:15
→ setmeaway2 joined00:15
k1t bremner: gotcha, will check that out00:15
avpx k1t: Have you committed yet?00:15
Oh, sorry00:15
Reading comprehension fail00:15
k1t yeah, locally00:15
avpx Okay, so what you most likely want to do here is reset --soft HEAD^00:16
→ eicto joined00:16
avpx What this will do is keep your working tree the same but bring you back to before the commit happened00:16
So then you can use git add ... and just select what you want, before making a new commit.00:16
That, or git rm all of the .png files and use git commit --amend00:16
cmn unless it's the root commit; git rm --cached <whatever>; git commit --amend will always work00:17
← savid left00:17
→ joooooo joined00:17
avpx --amend is a very useful flag.00:17
← setmeaway left00:17
← joooooo left00:18
← setmeaway2 left00:18
Leemp avpx: fatal: ambiguous argument 'origin/master': unknown revision or path not in the working tree.00:18
→ QaDeS_ joined00:18
→ setmeaway joined00:18
Leemp avpx: There doesn't seem to be anything in the remote branch..00:18
cmn that's not what that means00:18
Leemp no?00:18
← wilmoore left00:18
→ notVert joined00:19
k1t avpx: keep in mind, i have a bunch of files that i do still want to commit00:19
cmn it means that origin/master isn't anything, did you set up the remote and fetch correctly? what commands did yo run?00:19
→ wilmoore joined00:19
← Vert left00:19
→ setmeaway2 joined00:19
k1t so just reset the local head, then re-do my add command00:19
and commit again, right?00:19
→ Caius joined00:19
Leemp git remote add origin path00:19
and git fetch00:19
avpx k1t: I'm actually thinking you should git rm and then git commit --amend00:19
→ PrinceAMD- joined00:19
Leemp then ofcourse git log origin/master00:20
avpx k1t: git rm <all of those .pngs you hate> && git commit --amend00:20
← gusnan left00:20
k1t git rm *.png && git commit -am '' --amend ?00:20
something like that, yeah?00:20
or i could skip '-am"" ' probably right00:21
avpx Err, -m '' will give the commit an empty message00:21
Yeah00:21
← dvide left00:21
← QaDeS left00:21
k1t yeh, just left it empty for IRC purposes.. heh00:21
ok cool i think i understand things a bit better now, ty00:21
one step closer to git domination.00:21
Leemp cmn: Is something wrong with that?00:21
k1t lol00:21
→ dvide joined00:21
avpx k1t: You're welcome00:21
cmn Leemp: not offhand, does fetch say something? does ls-remote origin say anything?00:22
← andrewhl left00:22
k1t avpx: this way definitely feels natural too00:22
i like it and doubt i would have came up with it lol00:22
Leemp fetch said nothing, ls-remote says.. (waits)00:22
k1t anyways back to teh codes00:22
← setmeaway left00:23
Leemp cmn: ls-remote just says "From path" and nothing00:23
cmn then there is something else going on00:23
← QaDeS_ left00:23
← wilmoore left00:23
avpx Leemp: I'm curious about the output of git show-ref00:24
Leemp: git show-ref | grep 'refs/remotes' | awk '{print $2}'00:24
← setmeaway2 left00:25
← PrinceAMD- left00:25
→ joooooo joined00:25
← mischat left00:25
→ mischat_ joined00:25
← rbarraud left00:25
← joooooo left00:25
→ joooooo joined00:25
Leemp avpx: Nothing00:26
← jjbohn left00:26
avpx Leemp: What command were you using to push?00:26
Leemp avpx: git push origin master, same as always00:27
avpx :-\00:27
Leemp (I also tried -u for the hell of it)00:27
avpx cmn: Thoughts?00:27
Leemp I'm thinking about blaming cloud9..00:27
But i have no idea what would cause it00:27
→ uman joined00:27
uman Is there a way to git to save my private key passphrase for a few minutes, a bit like sudo does with password by default00:28
cmn avpx: no idea; if ls-remote and fetch can't download anything, push can't bloody complain00:28
← udante left00:28
avpx Leemp: It is very strange. The remote isn't giving you any refs, but then when you go to push a ref it complains that it's not a fast-forward.00:28
Leemp I'm going to try pushing a false project00:29
My codebase is being developed on cloud9, as i am trying the ide out00:29
avpx I'd suggest physically destroying the remote.00:29
Leemp avpx: I have, actually lol. I deleted the project, and remade it00:29
avpx I meant with a suitable blunt instrument00:29
Leemp So.. i am going to try making a project from my local host (not cloud9 ide), and push it from that..00:30
avpx: .. yea, it worked.. i have no idea how, but it is somehow a problem with cloud9.. my local pushes to github just fine.00:31
avpx Unsurprising00:31
Leemp I've been using c9 for a bit, and haven't had problems.. but yea, i agree.00:32
→ trying2code joined00:33
Leemp avpx, cmn: Time to take my problem to them. Thanks for putting up with this odd scenario :)00:33
avpx Leemp: No problem, glad I could help00:35
→ bryan joined00:36
← akiress left00:36
→ macrover joined00:37
← btanaka left00:38
→ PrinceAMD- joined00:38
→ jhardinee joined00:40
← ph^ left00:40
← bryan left00:41
→ mobilegamelabs joined00:41
← intransit left00:41
k1t avpx: worked great btw00:42
k1t dances00:42
k1t not that kinda dance, perv!00:42
avpx Great! Glad to hear it00:42
k1t jk jk00:42
→ bglazer joined00:42
→ ph^_ joined00:42
← PrinceAMD- left00:43
mobilegamelabs → MGGGGG00:45
← MGGGGG left00:46
← ph^_ left00:47
→ JeremyWei joined00:47
→ PrinceAMD- joined00:47
← tvw left00:49
← jhardinee left00:49
→ tvw joined00:50
← PrinceAMD- left00:53
→ Targen joined00:53
→ PrinceAMD- joined00:55
→ enmand joined00:56
→ bricker88 joined00:57
→ maqr joined00:58
→ Goplat joined00:58
→ bulletmark joined00:59
← AaronMT left01:00
bricker88 Hello - I can't seem to get global git ignore to work. Here is how I have it setup. The files I'm trying to ignore are not being tracked in the repo. https://gist.github.com/296325701:02
→ barneybook joined01:02
→ rbarraud joined01:02
→ harish joined01:02
→ cooloney joined01:03
← PrinceAMD- left01:03
→ AaronMT joined01:03
→ xinming joined01:04
← etcetera left01:04
← notfunk left01:05
← bricker88 left01:05
bremner !next01:06
darn, we need a "another satisfied customer" factoid.01:06
← barneybook left01:06
infogulch → infogulch|afk01:06
→ etcetera joined01:07
← raykyri left01:07
← maqr left01:07
→ barneybook joined01:08
sitaram oh yesterday there was a guy who came on for about 20 seconds. Asked a question, said "oh I found it, nvm" or something like that, and left01:08
→ w09x joined01:08
avpx I think my favorite trigger is !refund01:08
gitinfo If you are not satisfied with git for whatever reason, you are entitled to a full refund of the purchase price, and are invited to use another VCS. Elsewhere.01:08
sitaram avpx: I like !doesnt -- before they changed it01:09
gitinfo avpx: [!doesntwork] 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!01:09
sitaram it used to say "is it sleeping with your sister?" ;-)01:09
avpx Ahaha01:09
sitaram or maybe "run away with your sister..." I forget now01:10
→ pdtpatrick_ joined01:10
avpx Now why would you go and do that?01:10
sitaram because I don't have a sister? ;-)01:10
cmn I think the running away was with money01:10
sitaram probably. What I'd like to know is who got cold feet (was it jast or someone else)?01:11
→ weiyang joined01:11
→ notfunk joined01:11
→ Vampire0_ joined01:12
← pdtpatrick left01:12
→ dhruvasagar joined01:12
→ mobilegamelabs joined01:12
← osmosis left01:13
← mtx_init left01:13
← barneybook left01:13
← etcetera left01:13
gerard0 → gerardo01:14
← pdtpatrick_ left01:14
← erkules left01:15
← Vampire0 left01:15
→ zewb joined01:15
← pfFred left01:16
→ erkules joined01:16
← zewb left01:16
→ tomek joined01:16
→ Zenith77 joined01:16
gitinfo set mode: +v01:17
→ etcetera joined01:17
← gerardo left01:17
→ gerard0 joined01:18
gerard0 → gerardo01:18
← chrisgeorge left01:18
← hwrd|work left01:19
← enmand left01:19
← Lgb left01:20
← palm_vue left01:20
→ mtx_init joined01:20
→ dpwright joined01:20
→ jjbohn joined01:21
← Mowsh left01:22
→ quazimodo joined01:22
→ palm_vue joined01:22
→ Lgb joined01:22
→ macmartine joined01:24
→ PrinceAMD- joined01:25
→ ISF joined01:25
← macmartine left01:26
tomek i've problem with git pull, i've my local repo with 2 remote added (one is repo which i forked and it's using master branch and secon one is another fork - but i use upstream branch from it), and i've problem with this second remote01:29
← PrinceAMD- left01:29
tomek when i do git fetch/git merge it's working fine - it use "upstream" branch, but when i just do - git pull - it is using master branch01:30
how can i set it to use "upstream" branch only?01:30
cmn man git branch --set-upstream01:30
gitinfo the git-branch manpage is available at http://git-scm.com/docs/git-branch.html01:30
cmn or just don't use pull01:30
← mocramis left01:30
→ PrinceAMD- joined01:31
→ soshdin joined01:31
→ mocramis joined01:31
← jjbohn left01:31
← yoklov left01:31
→ nakanara joined01:32
→ mdhopkins joined01:32
tomek thanks, i'll check it01:32
← soshdin left01:33
avpx Yeah, I think it's wise to just git fetch all of the changes from the remote and then git merge when you want your branches to point to them01:34
→ alexukf joined01:34
← PrinceAMD- left01:36
← macrover left01:36
→ macabre joined01:36
→ hwrd|work joined01:37
→ PrinceAMD- joined01:37
→ Zespre joined01:38
← Vinnie_mac left01:38
← williamcotton left01:41
→ apok_ joined01:41
← thierryp left01:41
← PrinceAMD- left01:42
→ intripoon joined01:42
← archdori left01:43
← Fandekasp left01:43
← apok left01:43
← adityamenon left01:43
→ jjbohn joined01:44
← mocramis left01:44
→ mocramis joined01:44
← apok_ left01:44
← intripoon_ left01:45
← asakura1 left01:46
→ PrinceAMD- joined01:46
← kenneth left01:47
→ asakura1 joined01:48
→ jackh joined01:48
← mcluskydodallas left01:49
→ zewb joined01:50
← zewb left01:50
→ chaowang joined01:50
← trying2code left01:50
← PrinceAMD- left01:51
← jjbohn left01:51
← Lulzon left01:51
→ zewb joined01:52
zewb DCC SEND startkeylogger 0 0 001:52
← jjasinski left01:52
← zewb left01:52
→ jjasinski joined01:52
← UukGoblin left01:52
→ UukGoblin joined01:53
← tabakhase left01:53
→ jjbohn joined01:53
← crabmanX left01:53
← Karmaon left01:54
← The-Compiler left01:54
→ crabmanX joined01:54
→ xiangfu joined01:55
← mattp_ left01:55
← JoshuaA left01:55
← fir31ight left01:55
gerardo → gerard001:56
← sriehl left01:56
→ mattp_ joined01:56
→ eregon joined01:56
→ Stravy joined01:56
→ pandeiro joined01:57
→ fir31ight joined01:57
→ Karmaon joined01:58
← Karmaon left01:59
← stkrzysiak left01:59
→ tabakhase joined01:59
→ Karmaon joined02:00
→ ferdna joined02:00
← cedr left02:00
→ The-Compiler joined02:00
← weiyang left02:00
→ adityamenon joined02:01
← HyperNerdV2 left02:02
→ JoshuaA joined02:03
→ sriehl joined02:03
→ PrinceAMD- joined02:03
← RaceCondition left02:04
→ ianawilson joined02:05
← ianawilson left02:05
→ jubei joined02:05
→ Fandekasp joined02:06
→ archdori joined02:06
→ yoklov joined02:07
→ zdk joined02:07
← mklappstuhl left02:08
← PrinceAMD- left02:08
→ savid joined02:09
→ PrinceAMD- joined02:11
milki thats not how it works zeom02:11
er02:11
zewb02:11
who isnt here02:12
jubei ^_^02:13
ChanServ set mode: +o02:14
EugeneKay set mode: +b02:15
EugeneKay set mode: -o02:15
← PrinceAMD- left02:16
← mdhopkins left02:16
Corey EugeneKay: Uh... it was klined. :-)02:17
EugeneKay shrugs02:18
→ PrinceAMD- joined02:18
← hwrd|work left02:19
← goshawk left02:20
→ ashphy joined02:21
← savid left02:23
← PrinceAMD- left02:23
← mobilegamelabs left02:24
→ dondikaio joined02:24
← dondikaio left02:25
→ weiyang joined02:26
← jjbohn left02:26
→ PrinceAMD- joined02:27
← subrosa left02:27
← orafu left02:27
→ orafu joined02:27
→ iamjarvo joined02:31
← PrinceAMD- left02:31
gerard0 → gerardo02:32
→ leilaM joined02:32
← Tinti_ left02:33
→ RaceCondition joined02:33
→ Tinti joined02:33
← RaceCondition left02:34
← Tinti left02:34
gerardo → gerard002:34
→ Keavon joined02:35
→ jhardinee joined02:35
→ RaceCondition joined02:36
← kpreid left02:37
← Destos left02:37
→ PrinceAMD- joined02:37
→ andrewhl joined02:39
gitinfo set mode: +v02:40
gitinfo set mode: +v02:40
Keavon Hello. I'm new to Git (or any version control) and I'm setting up version control for my game in Unity. I've installed Git Extensions and made a repository in BitBucket02:40
SethRobertson Read !book02:41
gitinfo 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 !designers !gitt !vcbe and !parable02:41
Keavon Now I have no idea how to link BitBucket to Git Extensions02:41
SethRobertson You clone your bitbucket URL you were provided02:41
→ ianawilson1 joined02:41
SethRobertson Or you git init and add bitbucket as a remote, whichever02:41
Keavon So I store my respository on my computer, plus on the server?02:42
← PrinceAMD- left02:42
SethRobertson Yes. The very basis of distributed version control02:42
Keavon ok02:43
→ macrover joined02:46
→ PrinceAMD- joined02:46
← philips left02:48
← k1t left02:49
→ philips joined02:49
← RaceCondition left02:50
Keavon I don't know how to clone a repository. Would you mind helping walk me though it, please? Sorry if I'm bugging you guys.02:50
ChanServ set mode: -v02:51
frogonwheels Keavon: git clone url-of-your-repository02:51
← hyperair left02:51
frogonwheels Keavon: possibly you should read some !doc02:51
gitinfo Keavon: 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.02:51
SethRobertson He is using git extensions. A GUI.02:51
→ hyperair joined02:51
SethRobertson We generally support the command line.02:51
However, if you look at the gitextensions home page, they have video tutorials02:52
Keavon is there a Git Extensions IRC to ask in then?02:52
← PrinceAMD- left02:52
Keavon okay, I'll try that02:52
← yaymukund left02:52
frogonwheels Keavon: explorer. right-click on folder Git Extensions|Clone02:52
→ michaelmartinez joined02:53
Keavon the folder I would use would be a new folder I make that I want to be the local respository files?02:53
frogonwheels Keavon: really - it's not that hard. I didn't know and it took me 10s to find out.02:53
Keavon: no. the parent folder02:53
Keavon: !sias02:53
Keavon: !tias02:53
gitinfo Keavon: Try it and see™. You learn much more by experimentation than by asking without having even tried. If in doubt, make backups before you experiment (see !backup). http://sitaramc.github.com/1-basic-usage/tias.html may help with git-specific TIAS.02:53
← ianawilson1 left02:54
frogonwheels Keavon: the dialog that comes up is pretty good02:54
Keavon I unchecked the contextual menu option though02:54
I'll try reinabling it02:54
frogonwheels Keavon: I'm not sure what the point of using git extensions is then :|02:54
Keavon re-enableing*02:54
I was told by somebody to use it who used Unity, and he left but I still have questions. I just did what he said, Extensions, when he was on02:55
frogonwheels Keavon: and really, if you ran the main git extensions program.. there's a Clone repository button.02:55
Keavon yes, I know02:56
← michaelmartinez left02:56
Keavon I don't know what to do with that though02:56
frogonwheels press it?02:56
→ michaelmartinez joined02:56
Keavon lol of course02:56
frogonwheels Keavon: just plonk the url into the Repository To Clone button02:56
*edit box02:56
→ PrinceAMD- joined02:56
Keavon URL? Okay02:57
That's where I was confused, since it was asking for a file path02:57
frogonwheels Keavon: it can be.02:57
Keavon so mine would be [email@hidden.address] >02:58
?02:58
← pandeiro left02:58
Keavon ( > meant to be ? )02:58
SethRobertson If that is what bitbucket told you02:59
→ burak575_ joined02:59
Keavon okay02:59
just making sure02:59
So "Destination:" would be a local path to where I want it to go on my hard drive, right?02:59
frogonwheels Keavon: that's an rsync style speficiation, and is valid02:59
Keavon: the parent directory03:00
Keavon of what?03:00
Git Exensions?03:00
frogonwheels the directory you want the clone to go into.03:00
Keavon ok03:00
→ jjbohn joined03:00
frogonwheels if you want the clone to go into c:\projects\repository-name-here then specify it as c:\projects and 'subdirectory to create' will probably default to repository-name-here anyway03:00
Keavon I see03:01
frogonwheels Keavon: but really, you can do just as well on the command line. and if you ask in here that's what people know03:01
→ ipoval joined03:01
frogonwheels Keavon: you will still be able to interact with it under git extensions.03:01
← PrinceAMD- left03:01
Keavon ok, well I've got this almost done now03:02
frogonwheels Keavon: git extensions doesn't add anything special to know that there's a git repo there.03:02
Keavon should I check Initialize all submodules? What that mean?03:02
frogonwheels Keavon: check it, it won't hurt. it depends on whether your repo has a !submodule03:02
gitinfo Keavon: 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.html03:02
← ben_alman left03:02
frogonwheels Keavon: if you look at man git-clone see the -r option.03:02
gitinfo Keavon: the git-clone manpage is available at http://git-scm.com/docs/git-clone.html03:02
← weiyang left03:03
frogonwheels Keavon: but as I said, checking it won't hurt.03:03
Keavon ok03:03
← eletuchy left03:03
Keavon hopefully this works03:03
← macrover left03:03
Keavon oh, dang03:03
it didn't xD03:03
Permission denied03:04
→ rurufufuss joined03:04
← Sidnicious left03:04
← jjbohn left03:04
Keavon Ok I'll do it by command line03:04
→ kenneth joined03:05
Keavon Would that be Git Bash?03:05
← kenneth left03:05
→ Yuuhi` joined03:05
→ PrinceAMD- joined03:06
← Yuuhi left03:07
→ glennpratt joined03:08
→ ben_alman joined03:08
← dreiss left03:08
← ipalaus left03:10
← PrinceAMD- left03:11
← simesy_work left03:11
→ PrinceAMD- joined03:13
→ simesy_work joined03:13
← ashphy left03:13
← alexukf left03:13
← leilaM left03:14
← duckxx left03:14
← thearthur left03:15
← andref_dias left03:16
← PrinceAMD- left03:17
→ PrinceAMD- joined03:19
frogonwheels Keavon: you'll probably n eed the key in pageant03:20
Keavon: assuming you're using plink/putty03:20
→ HyperNerdV2 joined03:22
← bglazer left03:26
→ bglazer joined03:27
← AaronMT left03:27
→ enmand joined03:27
→ x0rs joined03:28
← PrinceAMD- left03:28
→ hoymkot joined03:29
→ igotnolegs joined03:29
Keavon I'm not03:32
I'm using... err, I forgot03:32
→ weiyang joined03:32
Keavon not PuTTY03:32
← micah left03:32
← mrtazz left03:32
Keavon OpenSSH03:32
← bglazer left03:32
← avpx left03:32
Keavon ugh I give up03:33
this is way too complicated03:33
→ leilaM joined03:33
frogonwheels Keavon: use the git:// url or http:// if it gives one03:35
Keavon: in order to use the ssh url, you need to have your key set up properly03:36
Keavon do I set it up in Extended?03:36
Extensions*03:36
← Eridius left03:37
frogonwheels Keavon: there should be a howto set up your ssh key. it's not really a git thing - it's just a ssh key. you can test it with ssh03:37
Keavon: did you generate a key for your bitbucket account?03:37
Keavon heh, I don't know about SSH, but I'll look it up03:38
not yet03:38
frogonwheels Keavon: ok. well without t hat, you can't use ssh protocol03:38
Keavon ok03:38
frogonwheels Keavon: bitbucket should give you a git or http: (hopefully a smart one) protocol for read-only acces03:39
Keavon it does03:39
so could anybody use that link to download my source code?03:40
← Internet13 left03:40
← joooooo left03:45
→ setmeaway joined03:45
→ Internet13 joined03:45
→ dreiss joined03:47
← thursday left03:47
← enmand left03:49
← ivan\ left03:49
→ ivan\ joined03:50
← dvide left03:50
→ subrosa joined03:52
→ rhl joined03:52
subrosa Hey guys what's the purpose of having a local/origin/live server setup. Can't I just do local/live ?03:53
offby1 idly wonders what a "local/origin/live server setup" is03:53
← bambams left03:54
← rchavik left03:54
subrosa offby1 that would be git on my local machine then a hosted server at hosting company X for the origin and then a different server with company X for the live site03:55
reboot03:55
opps sorry03:55
→ eletuchy joined03:56
← ISF left03:56
offby1 *shrug&03:58
← etcetera left03:59
→ dr_win joined03:59
subrosa ok different question then03:59
→ jlebar joined03:59
subrosa if I have a site locally do I create a database there to use or do I use the same database on the live server?04:00
jlebar For scripting purposes, how do I get the upstream branch of the current branch?04:00
subrosa or is the local database files pushed to the database?04:00
→ bearlulz joined04:00
→ mobilegamelabs joined04:00
← notVert left04:00
→ ashphy joined04:01
→ mdhopkins joined04:01
← eletuchy left04:02
← mischat_ left04:03
← mdhopkins left04:03
← glennpratt left04:05
← marcuy_ left04:06
offby1 subrosa: it's up to you, but I'd personally prefer to run a database on a machine that was professionally hosted, with a good internet connection.04:09
And backed up regularly etc.04:09
← cilly left04:09
offby1 Not sure what you mean by "have a site locally". Anyway, that question isn't pertinent to git, and I'm not a web programmer.04:09
→ joshteam joined04:10
→ lnostdal_ joined04:10
← Diego_g left04:11
→ hwrd|work joined04:12
← bearlulz left04:13
→ cilly joined04:13
← lindenle left04:14
← guampa left04:16
→ ISF joined04:16
← ivan\ left04:16
← ology left04:17
→ n2i joined04:17
→ beautifulmind joined04:18
← juvenal left04:18
→ duckxx joined04:20
→ ivan\ joined04:21
← subrosa left04:22
← uman left04:23
→ sangi joined04:24
← madewokherd left04:25
← neurodrone left04:25
← JeremyWei left04:29
← sangi left04:31
→ sangi joined04:32
← nakanara left04:33
← checkers left04:34
→ justgage joined04:34
← HyperNerdV2 left04:39
← philips left04:40
← wolfcore left04:40
← Keavon left04:43
→ akiress joined04:43
→ philips joined04:44
→ iwoj joined04:45
← joshteam left04:45
← justgage left04:45
→ osmosis joined04:46
→ mharris joined04:46
← macabre left04:47
→ ehsan joined04:48
→ macabre joined04:49
← bgerber left04:51
← d3vic3 left04:51
→ PRETTY_FUNCTION joined04:51
→ wilmoore joined04:51
← x0rs left04:52
→ d3vic3 joined04:53
← jlebar left04:55
→ yaymukund joined05:00
MiJyn bbl05:00
← Bass10 left05:00
→ bubbles10101 joined05:00
← _Vi left05:01
→ _Vi joined05:01
← tansell left05:02
→ macmartine joined05:02
→ cedr joined05:02
→ bgerber joined05:03
← macabre left05:04
→ h0bbit joined05:04
→ harshpb joined05:05
← iwoj left05:05
← Karmaon left05:06
r0bby_ → robbyoconnor05:06
← duckxx left05:07
→ nigelb joined05:07
← gavin_huang left05:08
← jason237 left05:08
→ gavin_huang joined05:11
→ JeremyWei joined05:11
→ thiago joined05:12
→ ColKurtz joined05:13
← kenperkins left05:14
→ tansell joined05:14
→ kenperkins joined05:15
→ brettgoulder joined05:17
← Targen left05:17
gitinfo set mode: +v05:18
brettgoulder how do you read a git diff header? (e.g @@ -1,5 +1,5 @@ )05:19
what's the 5 in it?05:19
thiago brettgoulder: it's just like a regular diff05:19
I assume you know it05:19
→ Garrick joined05:20
brettgoulder nope, i do not05:20
→ ToBeFree joined05:21
brettgoulder i'll figure it out though05:21
gitinfo set mode: +v05:21
Garrick hi everyone05:22
gitinfo Garrick: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.05:22
Garrick Is there a way to move/copy a file from one branch to another while including its commit history? I know git checkout <branch> <file> will do it, but without the files commit history.05:24
→ PrinceAMD- joined05:24
→ patrick99e99 joined05:25
DrNick files don't have history05:27
← bubbles10101 left05:28
← dhruvasagar left05:28
← _Vi left05:29
→ _Vi joined05:30
← patrick99e99 left05:30
→ pranjal710 joined05:30
← osmosis left05:31
→ mikepack joined05:31
← _Vi left05:31
→ imrobert joined05:32
→ _Vi joined05:32
imrobert Can I move local repositories around without messing anything up?05:32
e.g. I have a repo in /home/robert/repo and I want to move it to /home/robert/code/repo05:33
→ arti_t joined05:34
→ airborn joined05:35
← pranjal710 left05:35
← bartman`` left05:36
→ arvind_khadri joined05:36
← arvind_khadri left05:36
→ arvind_khadri joined05:36
wereHamster imrobert: mv /home/robert/repo /home/robert/code/05:36
→ angelsl joined05:36
wereHamster imrobert: and next time, !tias05:36
gitinfo imrobert: Try it and see™. You learn much more by experimentation than by asking without having even tried. If in doubt, make backups before you experiment (see !backup). http://sitaramc.github.com/1-basic-usage/tias.html may help with git-specific TIAS.05:36
imrobert wereHamster: Actually just tried it with cp -r :)05:37
→ pranjal710 joined05:37
← pranjal710 left05:38
imrobert thanks05:38
wereHamster++05:38
→ sq-one joined05:38
→ tigrang_ joined05:39
← tigrang left05:41
← imrobert left05:42
→ bartman`` joined05:43
wereHamster brettgoulder: google 'unified diff'05:45
brettgoulder cool, thanks wereHamster05:45
→ mishok13 joined05:48
→ siprbaum joined05:48
→ rchavik joined05:48
← rchavik left05:48
→ rchavik joined05:48
← dreiss left05:49
← Chaser_ left05:50
→ Eridius joined05:50
← bosspetta left05:50
← andrewhl left05:51
→ bosspetta joined05:51
→ bicranial joined05:51
→ jkitchen joined05:51
md_5|away → md_505:51
→ naught101 joined05:52
← tigrang_ left05:53
naught101 I have a file that I have some stub changes to, that will eventually need to be thrown out. I don't want to commit them, but git diff keeps telling em about them. Is there any way of telling git to ignore the changes to the file (considering the original version is already commited)?05:53
I thought a temporary branch might be good, but I don't really want to have to keep switching branches and rebasing05:54
gitinfo set mode: +v05:54
→ soee joined05:55
← yoklov left05:56
← n2i left05:56
→ Pupeno_W joined05:56
← sq-one left05:57
→ Axfy joined05:57
← ToBeFree left05:57
→ pantsman joined05:57
← pantsman left05:57
→ pantsman joined05:57
→ bosspetta_ joined05:57
→ dreiss joined05:57
→ flijten joined05:59
← madsy left05:59
← bosspetta left06:00
bosspetta_ → bosspetta06:00
→ marc_kdab joined06:00
← Sky[x] left06:05
← beautifulmind left06:05
→ afuentes joined06:07
→ Sitethief joined06:07
← macmartine left06:09
← tonyj left06:09
← mikepack left06:10
gitinfo set mode: +v06:10
dpwright I'm just taking a look through the git.git log to get a feel for their development structure06:12
← dreiss left06:12
dpwright and there's one or two things I'm not entirely clear on06:12
← mishok13 left06:12
dpwright as far as I can tell, patches are sent in series to the mailing list and then applied to a feature branch by the maintainer06:13
→ mikepack joined06:13
dpwright these branches seem to be local, but are merged into the "pu" branch which is public06:13
they're moved into "next" and eventually "master" as they are developed and tested and deemed ready for it06:14
so far I think I understand06:14
the thing I'm not entirely sure about is, if a patch series is discussed on the mailing list and it's determined that it should be done a different way, often a new patch series is created06:14
← airborn left06:15
dpwright what happens to the original branch? Does junio throw it away, apply the new patch series as a new branch, and then merge that new branch into "pu"? Doesn't that mean "pu" (a public branch) is constantly being rebased?06:15
→ hoschi joined06:16
dpwright similarly, as branches may be moved into "next" in a different order than they were applied to "pu", does this require a rebase?06:16
→ cbreak-work joined06:16
→ mischat joined06:16
→ ThomasLocke joined06:19
wereHamster dpwright: man gitworkflow06:19
dpwright: man gitworkflows06:19
gitinfo dpwright: the gitworkflows manpage is available at http://git-scm.com/docs/gitworkflows.html06:19
dpwright wereHamster, aha! Thanks, I'll give that a read06:19
→ beautifulmind joined06:19
← notmypudding left06:20
oo_miguel hi06:20
← Tabmow left06:20
← bosspetta left06:24
→ senny joined06:24
wereHamster !hi06:25
gitinfo [!welcome] Welcome to #git, a place full of helpful gits. If you have a question, feel free to just go ahead and ask—somebody should answer shortly. For more info on this channel, see http://jk.gs/git/ Take backups (type !backup to learn how) before taking advice.06:25
← hwrd|work left06:26
← thiago left06:26
← dcarper left06:28
← yaymukund left06:28
→ Jackneill joined06:28
→ Xizor joined06:29
← angelsl left06:31
← oo_miguel left06:32
→ Sitethief{afk} joined06:32
→ angelsl joined06:32
→ oo_miguel joined06:32
← arti_t left06:32
← bgerber left06:32
→ Rapp joined06:33
Rapp hi everybody06:34
gitinfo Rapp: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.06:34
Rapp i have a subdirectory in my repository, which i cannot remove with git: git rm -r somedir says "cannot remove somedir, it's a directory"06:34
← michaelmartinez left06:34
→ ph^ joined06:35
← Sitethief left06:35
Rapp but just doing a rmdir somedir doesn't help either, because later git will again fetch it from the remote06:35
how can i get rid of this directory?06:35
→ mishok13 joined06:36
← mikepack left06:36
naught101 Rapp: git doesn't track directories, just file (contents). So you need to delete everything in that directory06:36
sitaram git rm -r should work; there's something else wrong with this. Can you try this in a fresh clone of your repo?06:36
naught101 Rapp: if the directory is empty, git will ignore it06:36
Rapp the directory is empty, but it was another git repo, which i did not add as a submodule (but which i want to add as one now)06:37
sitaram aah submodules... I suddenly remembered an urgent meeting06:38
Rapp fresh clone, did ls -al on that dir -- it's empty06:38
so i think git half-added the submodule, and is confused now06:38
← mharris left06:39
Rapp hm, i think the problem is, that .gitmodules did not get the right entry after i accidentally added somedir06:40
→ bgerber joined06:41
← Sitethief{afk} left06:43
→ Sitethief joined06:44
← mobilegamelabs left06:44
→ JohnBat26 joined06:45
→ resmo_ joined06:45
→ kcj joined06:47
→ toabctl joined06:48
→ arti_t joined06:48
→ dc5ala joined06:49
← ipoval left06:50
→ marthinal joined06:51
← innrwrld left06:52
← Lemon|mbp left06:53
← yizumi left06:54
→ yizumi joined06:54
→ dreiss joined06:55
← bicranial left06:55
→ dsdeiz joined06:57
← tonesfrommars left06:58
← Xizor left06:59
← diegoviola left06:59
← yizumi left07:00
← arvind_khadri left07:00
→ yizumi joined07:00
→ caddoo joined07:00
dsdeiz hi! i renamed some files. and i think i did it wrong cause when i do 'git status' it says that the files were deleted an i need to add the renamed files. tried to do git mv [delete file] [new file] but god fatal: bad source. any suggestions? or is the output of 'git status' normal?07:00
→ macrover joined07:00
→ Innocentus joined07:03
← bgerber left07:04
→ frnkkpp joined07:06
→ robotmay joined07:06
→ Chib joined07:07
→ barmaglot joined07:07
dpwright that output sounds normal07:07
if you add the new files and delete the old ones, later, when you do blames/logs, git can track the movement07:08
cbreak-work dsdeiz: just add the new files07:09
dsdeiz os the log of the file name before gonna be recorded on the new one?07:09
is*07:09
cbreak-work and maybe you have to stage the removal too wight git add -u07:09
dsdeiz: no07:09
dpwright also, you'll notice when you do that the output of "git commit" will mark it as a file movement07:09
cbreak-work git does not recorde logs on any filename07:09
← toabctl left07:10
← ferdna left07:10
dpwright it will actually do this even if you don't rename a file, but if you just move large chunks of code from one file into a new one, which gives you the clue that it's working it out based on the content of the file, not its filename07:10
→ caseymcg joined07:10
→ goshawk joined07:10
→ Lemon|mbp joined07:11
dsdeiz ah cool. thx!07:11
→ purencool joined07:11
→ tsdh joined07:13
→ arvind_khadri joined07:14
← arvind_khadri left07:14
→ arvind_khadri joined07:14
→ toabctl joined07:14
purencool I have been working on a file in my git repo but I have wrecked it has not been added or commited to the repo. How do I get the last changes I made out of the git repo07:14
I am using the commandline07:14
← goshawk left07:15
→ bgerber joined07:15
→ jason237 joined07:15
barmaglot <purencool> git reset --hard commit_hash ?07:16
purencool what if I want to do it to one file not the rest07:16
cbreak-work purencool: just git checkout someoldcommit -- filename07:17
obviously the file has to be committed in some old version07:17
barmaglot <purencool> yes.. --filename.. i don`t right understand you07:17
purencool yes and thanks I will give a try thanks07:18
← naught101 left07:18
← kdub3 left07:18
← palm_vue left07:20
← Lgb left07:20
purencool that did it thanks07:20
← bgerber left07:21
→ dhruvasagar joined07:21
← cilly left07:21
→ palm_vue joined07:22
→ Lgb joined07:22
→ thierryp joined07:24
← ISF left07:25
← beneggett left07:25
→ thiago joined07:25
← robotmay left07:27
← toabctl left07:27
→ madsy joined07:29
← beautifulmind left07:30
quazimodo sitaram: recon gitolite can be extended to handle many servers, each copy taking its config from a central store?07:30
← marc_kdab left07:30
→ goshawk joined07:31
← dhruvasagar left07:31
→ dhruvasagar joined07:32
cbreak-work gitolite handles many servers07:33
I have one, and I bet I am not the only one :)07:34
(besides, you can push the admin repo to as many servers as you want)07:34
→ nhooey|mini__ joined07:35
← yizumi left07:36
→ bgerber joined07:36
→ yizumi joined07:37
← frnkkpp left07:37
→ hafos joined07:38
← macrover left07:38
← nhooey|mini_ left07:38
hafos Since my last commit I made some changes that I actually wanted to make in a new branch... what should I do to fix this?07:38
cbreak-work did you commit?07:38
hafos not yet07:38
cbreak-work then just switch to an other branch and commit there07:39
hafos ok thanks07:39
cbreak-work committing is what assigns state to a branch07:39
if you haven't committed, then your stuff is not on any branch07:39
hafos right, I had a hunch that would be the case :)07:39
just wanted to make sure I wouldnt lose the changes07:39
→ tylerwhitney joined07:40
tylerwhitney I'm a subversion guy, new to git, but going to use it for a few projects... if I already have a directory on a server that I want to make into a repository can I use the "git init" command on a full directory?07:41
← igotnolegs left07:42
← thierryp left07:43
→ klj613 joined07:44
← Lemon|mbp left07:44
← brettgoulder left07:44
thiago tylerwhitney: yes07:44
→ yaymukund joined07:45
← pantsman left07:45
→ needsch joined07:45
← simesy_work left07:46
tylerwhitney thiago: thank you, generally I think I've done "git init --bare" to create a new repo while playing around, but from what i'm reading of the docs its as simple as doing "git init" when I'm in the directory of existing files, right?07:46
← asakura1 left07:46
tylerwhitney with no other flags I mean07:46
sitaram quazimodo: I did not understand the question. Each gitolite instance can only control stuff on its own server. Unless you're talking about mirroring, which of course is supported07:46
← arvind_khadri left07:47
→ toabctl joined07:48
← dhruvasagar left07:48
← teukka left07:49
← toabctl left07:49
→ dhruvasagar joined07:49
→ thierryp joined07:51
← _huoxito left07:54
→ kdub3 joined07:54
← MissionCritical left07:54
→ arietis joined07:54
← spyro left07:55
→ spyro joined07:56
← Fudge left07:56
← nevyn left07:56
← maletor left07:56
← scientes left07:57
→ asakura1 joined07:59
→ francisfish joined07:59
← w41_ left07:59
→ maletor joined08:00
→ simesy_work joined08:01
← thiago left08:02
→ LimeBlast joined08:03
→ BarryCrayon joined08:04
→ bosspetta joined08:04
← goshawk left08:07
→ Fudge joined08:07
← dhruvasagar left08:09
→ dhruvasagar joined08:09
← wagle left08:10
afuentes I have a branch in a server so i can access from 2 computers that i would like to rebase with master from time to time... thing is... i am not supposed to rebase branches already in server... And i have to commit this branch in order to be able to access it from another computer :S08:14
→ monsieurlu joined08:14
afuentes how is the workflow supposed to go?08:14
→ dzonder joined08:15
FauxFaux You're not supposed to rebase branches that other people might care about. So long as you make sure the branch is PRIVATE-AFUENTES-master then you should be fine.08:15
afuentes yeah, well, im the only person using the repo... thing is i need to access from 2 computers, so i have to push it to the server08:16
FauxFaux Yeah, that's fine.08:17
afuentes but then i have to rebase the branch in the server :S08:17
from time to time08:17
how to do that? checkout branch, rebase it and and push it again overwritting branch in repo?08:17
is that how the flow is supposed to go?08:17
jast sitaram: there's a version history on the website, and feel free to change it and stuff08:18
fwiw the same person that added it removed it, too (or, rather, reworded it)08:19
← Gitzilla left08:19
→ lukosanthropos_ joined08:20
← lukosanthropos_ left08:20
← lukosanthropos left08:20
→ rtuin joined08:20
→ lukosanthropos joined08:20
sitaram jast: I only wanted to pull his leg/tease him about it, whoever it was. Nothing serious :)08:20
→ Desproges joined08:20
← dsdeiz left08:20
→ alexukf joined08:21
→ Gitzilla joined08:21
→ ziro` joined08:21
← monsieurlu left08:22
→ federic joined08:22
← alexukf left08:25
← Gitzilla left08:26
← mischat left08:26
← purencool left08:27
→ marcogmonteiro joined08:27
→ jkremser joined08:27
→ Gitzilla joined08:28
→ wagle joined08:28
→ dcarper joined08:28
→ OOPMan joined08:28
← MiJyn left08:28
jast sitaram: feel free to use the history for nothing serious, then. :)08:29
→ lnostdal__ joined08:29
→ nevyn joined08:30
→ MiJyn joined08:30
→ blacktulip joined08:30
← maletor left08:31
→ roofi joined08:32
→ dr_lepper joined08:32
← lnostdal_ left08:32
← roofi left08:32
→ whitman joined08:32
← dcarper left08:33
whitman I have a file which has been deleted in my develop branch but needs a hot fix in master. When I merge develop into master will there be a merge conflict even though the file is just being deleted?08:34
→ TheHaven joined08:34
→ und3f joined08:35
→ sebrock joined08:35
→ shruggar joined08:36
← Gitzilla left08:36
← GeorgeJ left08:36
→ Tabmow joined08:37
→ Gitzilla joined08:38
→ GeorgeJ joined08:38
← quazimodo left08:38
→ thiago joined08:40
→ MissionCritical joined08:40
← needsch left08:43
→ tjadc joined08:44
← thiago left08:46
toed when i do git remote -v i see that one source is there twice, for fetch and for push. but i'd like to remove the push one -- make it fetch only. is this possible?08:47
← tjadc left08:47
→ jatt joined08:49
→ Sitethief{afk} joined08:51
← yizumi left08:51
→ yizumi joined08:52
→ PigeonFriend joined08:52
← Sitethief left08:54
→ robotmay joined08:55
cbreak-work toed: what for?08:57
toed: man git-config08:58
gitinfo toed: the git-config manpage is available at http://git-scm.com/docs/git-config.html08:58
cbreak-work look for remote.<...>.url and pushurl08:58
toed i want to make sure i don't accidentally push into the wrong thing08:59
thanks09:00
← JoshuaA left09:00
→ Pentla joined09:00
→ arvind_khadri joined09:03
RichiH is there any way to suplement GIT_DIR with a path when using submodules?09:04
erm, GIT_WORK_TREE, i mean09:04
→ Element9 joined09:05
← hoschi left09:05
→ mischat joined09:06
→ timini joined09:08
→ berserkr joined09:08
→ Nedvajz joined09:09
→ stamina joined09:09
timini Hi I just switched to an ond branch hopeing to find some file there that wern't on the master branch and got this http://dpaste.com/761930/ The files I need are listed with the D at the beginning of the line, what does that mean?09:10
← tvw left09:10
→ jceb joined09:11
jast RichiH: meaning?09:12
timini: that refers to uncommitted changes you've made that have been 'floated over' into the branch09:13
remember that uncomitted changes don't belong to any specific branch09:13
→ stodan joined09:13
← spyro left09:13
timini so if i commit then checkout another branch things should change09:13
→ ipalaus_ joined09:14
← yaymukund left09:15
← harshpb left09:16
jast timini: yes. be sure to be on the right branch when committing, though. :)09:16
→ uraimo joined09:16
→ spyro joined09:16
← spyro left09:16
→ spyro joined09:16
← harish left09:16
→ harshpb joined09:17
timini so you can overrite the any branch with your current state?09:17
→ drizzd_ joined09:17
→ cissu joined09:18
RichiH jast: for reference: https://github.com/RichiH/vcsh/issues/27#issuecomment-535487009:20
timini Are the uncommited changes called HEAD?09:20
RichiH i am trying to figure out how to best solve this09:20
← drizzd left09:20
← Sitethief{afk} left09:21
RichiH that GIT_WORK_TREE overrides the full path in git submodule seems less than ideal (with limited understanding; i never used submodules)09:21
→ Sitethief joined09:21
→ cortana2 joined09:22
← arti_t left09:23
→ shreya joined09:24
→ hoschi joined09:26
← harshpb left09:26
→ harshpb joined09:26
→ barmaglot_ joined09:28
→ yaymukund joined09:29
← ipalaus_ left09:29
→ thiago joined09:30
→ andref_dias joined09:31
← _Vi left09:31
→ _Vi joined09:32
← OOPMan left09:32
← dhruvasagar left09:32
→ irqq joined09:32
→ dhruvasagar joined09:33
→ vano joined09:33
→ arti_t joined09:34
← spyro left09:34
→ Lord_DeathMatch_ joined09:34
← harshpb left09:35
→ tcurdt joined09:35
← tcurdt left09:35
→ harshpb joined09:35
→ spyro joined09:37
project2501b → project2501a09:38
boob00 hi09:39
→ srinivas07 joined09:40
boob00 git init <dir> I can put repo above it, but git init --bare, I cant, right?09:40
→ adhawkins joined09:41
boob00 git init --bare, is the upper level?09:41
vano I have fetched in from a wrong repository (after adding the wrong repo as remote) and have merged them. Thankfully the wrong repo is older than the new project, so I have the First commit of my project there and the rest of the stuff after it still intact. How can I fix this (remove all the old stuff previous to the first commit that has been included by mistake?) Thanks09:41
→ RaceCondition joined09:42
← dhruvasagar left09:44
← ehsan left09:44
← weiyang left09:45
→ dhruvasagar joined09:45
← shreya left09:46
→ mklappstuhl joined09:47
← vano left09:47
→ snearch joined09:47
→ beneggett joined09:50
← jceb left09:53
← jackh left09:54
→ pureluck_ joined09:54
jast timini: !HEAD09:57
→ pulse00 joined09:57
← Goplat left09:57
jast oh, come on, bot09:57
← dr_win left09:57
pulse00 hi all. can anyone recommend a self-hosted github alternative?09:57
gitinfo timini: HEAD is a 'pointer' in the repository that refers to whatever is currently checked out (or the default branch in bare repositories). It's not specific to any branch.09:57
jast there we go09:57
→ dr_win joined09:57
← pureluck_ left09:58
→ pureluck joined09:58
timini Thanks09:58
← dhruvasagar left09:58
→ dhruvasagar joined09:58
→ keex joined09:59
← mklappstuhl left10:00
dpwright pulse00, gitorious seems to be the done thing there10:00
http://gitorious.org/10:00
I have not tried it. It looks pretty good but I hear the setup is hard10:00
pulse00 dpwright: thanks. that's one of the 2 we're evaluting atm, the other one is http://gitlabhq.com/10:00
← Sitethief left10:00
pulse00 dpwright: yeah, that setup is what our admins are struggling with currently ;)10:01
← sebrock left10:01
→ ipalaus_ joined10:02
← irqq left10:02
→ raatiniemi joined10:02
→ irqq joined10:03
→ mobilegamelabs joined10:03
dpwright gitlab looks interesting -- very much github-inspired10:03
how is the setup for that? We are currently accessing repos on each others' computers directly using git-daemon, but it has its issues and the notorious gitorious setup (has a ring to it, that) is the only thing putting us off giving it a try10:04
→ gusnan joined10:04
← cissu left10:05
← mobilegamelabs left10:05
→ jceb joined10:06
← chaowang left10:07
← caseymcg left10:07
bremner !gitlab10:08
gitinfo gitlab is a git hosting package(similar to GitHub) built atop !gitolite(though it does not use the fancy ACL magic): http://www.gitlabhq.com/10:08
← dhruvasagar left10:09
→ mobilegamelabs joined10:09
rchavik dpwright, i had success using the vm they provided.10:09
→ dhruvasagar joined10:09
→ udante joined10:10
rchavik some issues during install process, but i managed to get it working after updating to latest stable10:10
→ beautifulmind joined10:11
jast dpwright: if all you need is repo access/management, gitolite is pretty easy to install10:12
and has little overhead10:12
→ marhaban joined10:13
← ipalaus_ left10:14
→ ipalaus_ joined10:14
dpwright jast, yeah, we're a small team, we don't need much, but some people like a bit of a visual interface and are familiar with github, so gitlab looked like a possibility10:14
ipalaus_ → ipalaus10:14
jast right10:14
dpwright anyway, thanks for the advice! time now to head to the pub10:14
← JeremyWei left10:15
← gavin_huang left10:15
← strobert left10:15
→ jn_ joined10:16
Lord_DeathMatch_ → Lord_DeathMatch10:17
← jkremser left10:18
← bgerber left10:19
← ph^ left10:20
→ bgerber joined10:20
→ ph^ joined10:20
boob00 any idea how I can put GIT build in CruiseControl.NET "Last build label" ?10:21
← mischat left10:23
← dhruvasagar left10:23
jast !GIT ;)10:24
gitinfo Git is not an acronym. Writing it in all caps is heresy and will be punished by Subversion or SCCS for more grievous offences. It's "git" or "Git". Thank you for your help in making the world a better place.10:24
← keex left10:24
→ dhruvasagar joined10:24
jast anyway, you can get the most recent commit ID on the current branch using this: git rev-parse HEAD10:24
maybe that helps10:24
→ harish joined10:24
→ n0xie joined10:25
← ph^ left10:25
gitinfo set mode: +v10:26
n0xie question: is there a simple way to merge part of a branch. I have a branch which i want to merge up until 2 commits ago. (so everything except the last 2)10:26
→ RaceCondishun joined10:27
← thiago left10:28
→ dcarper joined10:29
← RaceCondition left10:30
RaceCondishun → RaceCondition10:30
← noisebleed left10:32
← dr_win left10:33
← udante left10:33
← dcarper left10:33
→ udante_ joined10:33
→ udante joined10:34
← udante left10:35
→ mischat joined10:35
→ tjadc joined10:36
← caddoo left10:37
← udante_ left10:37
adhawkins +n0xie: I think you can merge against a specific commit can't you?10:38
md_5 → md_5|away10:38
adhawkins After all, a branch is just a pointer to the commit at the head of the branch.10:38
n0xie you can?10:39
you can!10:39
→ caddoo joined10:39
n0xie you learn something new every day10:39
thanks adhawkins10:39
adhawkins np10:39
→ RaceCondishun joined10:39
← RaceCondition left10:40
RaceCondishun → RaceCondition10:40
→ ph^ joined10:43
← jn_ left10:46
Jackneill can you help me? i have a repo with a branch 'master'. another person has the same repo with some + commits and i pulled it to 'sth'10:46
but it says CONFLICT errors10:46
FauxFaux And?10:47
Jackneill i want to merge my old repo to my friend's new commits10:47
FauxFaux Sounds like yuo tried to do that, and have conflicts. You have to resolve them by hand.10:47
Jackneill how can i do it? just because now i cant checkout master branch10:47
FauxFaux Read what "git status" says.10:48
Jackneill and then git add . and git commit -am ".."?10:48
after i can merge?10:48
FauxFaux You're already trying to do a merge. The above will commit the merge conflict markers, which exist /because you havea CONFLICT/.10:49
Jackneill then?10:49
FauxFaux 21/11:47:49 < FauxFaux> Sounds like yuo tried to do that, and have conflicts. You have to resolve them by hand.10:50
← gchristensen left10:51
← dreiss left10:54
← piotr left10:55
← wolog left10:56
heoa Is there some fast way to process conflicts after push?10:58
Suppose I clear the screen, how can I find the conflicts now?10:58
← pureluck left10:59
cbreak-work heoa: man git-diff10:59
gitinfo heoa: the git-diff manpage is available at http://git-scm.com/docs/git-diff.html10:59
cbreak-work heoa: also look at git status and the files themselves11:00
they should contain conflict markers11:00
→ piotr joined11:00
→ wolog joined11:01
← arti_t left11:01
→ w09x_ joined11:01
← dhruvasagar left11:01
→ neurodrone joined11:02
← neurodrone left11:02
→ neurodrone joined11:02
→ SkAzZ joined11:02
→ dhruvasagar joined11:02
← jatt left11:02
→ kpreid joined11:03
→ jkremser joined11:03
→ arti_t joined11:04
← w09x left11:04
← raatiniemi left11:04
← beneggett left11:05
← zdk left11:05
→ AdrienBrault joined11:06
← blacktulip left11:08
→ enmand joined11:08
← toed left11:09
→ mischat_ joined11:09
← harish left11:10
← mischat left11:11
mischat_ → mischat11:11
→ raatiniemi joined11:12
→ pureluck joined11:12
← arti_t left11:14
→ mklappstuhl joined11:18
← dhruvasagar left11:18
→ dhruvasagar joined11:19
→ jimmy1980 joined11:21
→ w09x joined11:21
→ sc joined11:21
→ pfFredd joined11:22
← w09x_ left11:22
← neurodrone left11:24
→ zinjacoder joined11:27
← dhruvasagar left11:28
→ ToBeFree joined11:28
→ dhruvasagar joined11:29
→ harish joined11:30
→ matessim joined11:31
→ chrisgeorge joined11:31
← PrinceAMD- left11:32
← rbarraud left11:33
→ fluxdude joined11:33
← sc left11:34
← mischat left11:34
← ph^ left11:34
fluxdude I'm getting more in to Git, can someone explain to me how storing files as blobs is a good idea, doesn't this mean that when making minor changes to txt files such as source code it has to store the whole thing again instead of just a small delta? Isn't this very space inefficient?11:34
→ mischat joined11:34
→ dabd joined11:35
← Cerise left11:35
→ ph^ joined11:35
→ Cerise joined11:35
cmn it's a tradeoff11:37
← Innocentus left11:37
cmn the loose objects mean faster access time11:37
and after a while, they'll be compressed in a packfile11:37
fluxdude so it will take up a lot more space? This isn't such good news for my SSD laptop with 6GB left...!11:38
sitaram fluxdude: http://keithp.com/blogs/Repository_Formats_Matter/ and search for the string "2.7"11:38
cmn huh?11:38
sitaram huh indeed11:39
cmn how is "they'll be compressed" -> "take up a lot more space"?11:39
← ph^ left11:39
sitaram cmn: he focused on the "after a while", which means right now they're uncompressed11:39
cmn ok, but a lot more space is then related to what?11:40
← iamjarvo left11:40
cmn he already said it'd take up the size of the file several times over11:40
gc won't increase the filesystem size11:40
← yizumi left11:41
→ jatt joined11:43
← dhruvasagar left11:43
→ BobMarley joined11:44
→ dhruvasagar joined11:44
BobMarley hi11:44
how do i remove all the tild files from the github?11:44
cmn tild files?11:45
is this github-specific?11:45
fluxdude sitaram: thanks for the link reading it now11:45
not sure how the size shrinks but reading the whole page11:45
→ erasmospunk joined11:46
→ jn_ joined11:46
→ hodgy joined11:46
BobMarley cmn, https://github.com/GreatMaster/edk2/tree/master/UefiCpuPkg/CpuDxe , all files that end with ~11:47
cmn same way you compress anything else, you identify common content and reference that instead of having it in the same object11:47
BobMarley: you delete them with git rm <whatever>~ just like any other file11:47
or use a glob, *~11:47
wereHamster BobMarley: git rm '*~' && git commit11:47
cmn it's just another file;11:47
erasmospunk does anybody knows how to change the type of an object? I have an error that the commit "is a tag, not a commit"11:48
← jceb left11:48
wereHamster erasmospunk: why. why change?11:48
BobMarley and how do i remove an entire folder?11:48
wereHamster BobMarley: git rm folder/11:48
FauxFaux erasmospunk: From what command? Most should automatically resolve. If not, append "^{}" or whatever it is.11:48
→ tvw joined11:48
wereHamster BobMarley: really, is it that not obvious?11:48
BobMarley wereHamster, i am begining to use the git, i know that i have really very easy qestions, but that is the reallity11:49
erasmospunk The repo corrupted, I get this error with git fsck error: Object 34c22466e6db485638f457d2337de577c0c8488c is a tag, not a commit and this error in commit 527abd718d58471b2ea3ac8b4046c7a5df549c29: parent objects missing11:49
→ blacktulip joined11:49
wereHamster BobMarley: http://git-scm.com/documentation11:49
erasmospunk the 527abd718d has the parent 34c22466e6db but it is a tag11:50
← fisted left11:50
wereHamster erasmospunk: restore the repo from backup11:50
cmn that output makes no sense11:50
← BobMarley left11:50
→ ph^ joined11:51
cmn so you've either cracked SHA-1 or been playing around with commit-tree11:51
erasmospunk wereHamster this is the last option11:51
← matessim left11:51
wereHamster erasmospunk: your only option, I'm afraid11:51
→ fisted joined11:52
erasmospunk the only backup is the distributed stuff we have11:52
cmn huh?11:52
→ neurodrone joined11:52
← neurodrone left11:52
→ neurodrone joined11:52
→ rearth_ joined11:53
cmn something created horribly a horribly broken commit, you need to restore11:53
← RaceCondition left11:53
erasmospunk I mean we have a central repo and the repos in our pcs11:53
cmn so you mean that you have no backups11:53
erasmospunk mine is not corrupted11:53
→ matessim joined11:54
cmn good for you11:54
erasmospunk not of the central repo11:54
cmn that's still not a backup11:54
erasmospunk you have a point11:54
← enmand left11:54
→ Vert joined11:55
← bartman`` left11:56
← hodgy left11:56
← fluxdude left11:57
← p3rror left11:58
← piotr left11:58
→ enmand joined11:59
← erasmospunk left11:59
→ piotr joined11:59
→ erasmospunk joined11:59
← piotr left12:00
← w09x left12:00
→ Stormbytes joined12:01
→ i42n joined12:01
→ JeremyWei joined12:02
→ Innocentus joined12:02
→ bartman`` joined12:03
← dhruvasagar left12:03
→ dhruvasagar joined12:04
→ piotr joined12:04
← bulletmark left12:04
← kcj left12:05
→ dvide joined12:05
→ marlun joined12:05
→ erichynds joined12:06
→ thiago joined12:07
← pfFredd left12:10
→ p3rror joined12:11
→ marc_kdab joined12:11
← beautifulmind left12:11
← madsy left12:12
← JoeCoolDesk left12:12
← dvide left12:12
→ dvide joined12:15
← Fudge left12:15
← rtuin left12:15
← berserkr left12:15
→ jajmon joined12:16
jajmon any idea why my latest commit does not show up in gitk? (tried restarting gitk)12:16
it shows in git log though12:16
← dhruvasagar left12:16
FauxFaux Running gitk with arguments?12:16
→ Stormbytes2 joined12:17
jajmon just &12:17
→ dhruvasagar joined12:17
→ Fudge joined12:17
FauxFaux shrugs.12:17
jajmon ok actually i had --all as well12:18
using an alias so forgot12:18
→ plasticle joined12:18
← thiago left12:18
jajmon also the commit does not get pushed when i push12:18
O_O12:19
→ glennpratt joined12:19
jajmon hmm12:20
if i stash my changes it appears12:20
FauxFaux I suspect that you're crazy.12:20
→ ianawilson joined12:20
jajmon ill show some screens12:20
→ thiago joined12:20
→ Tinti joined12:21
← Stormbytes2 left12:23
→ cannonball joined12:24
→ joshteam joined12:24
→ baazigar joined12:25
jajmon http://83.248.110.56/git-missing-commit-when-popped.png12:27
→ RaceCondition joined12:27
baazigar Hello, I am on errorbranch, and i want to commit the local changes to main branch , to that I have to change the branch to main branch by <git checkout main> but i cant because there are local uncommited changes. Any help?12:28
jajmon so i guess the commit is stashed somehow?12:28
baazigar jajamon : asking me ?12:28
jajmon no12:29
baazigar k12:29
jajmon FauxFaux, am i crazy?12:29
FauxFaux jajmon: That text doesn't look like bold, which implies you're not actually on the master branch, and are just confused by the --all output.12:29
← yaymukund left12:29
FauxFaux Or your shit operating system doesn't do bold. Which, y'know, is possible.12:29
→ d3c joined12:29
← Nedvajz left12:29
→ iamjarvo joined12:29
← dhruvasagar left12:30
← matessim left12:30
→ d0k joined12:31
→ beautifulmind joined12:31
← beautifulmind left12:32
jajmon FauxFaux, it does a ywllow dot12:32
→ sebrock joined12:32
→ w0m joined12:33
→ jcromartie joined12:34
→ TheJH joined12:34
→ matessim joined12:35
← matessim left12:35
→ beautifulmind joined12:36
→ Mowsh joined12:36
jajmon ok so i now my problem. i checked out the latest commit by the hash and not with "master"12:36
cbreak-work detached HEAD12:37
← thiago left12:37
→ tommyvyo joined12:37
FauxFaux So, not actually on the master branch.12:37
← glennpratt left12:37
→ RaceCondishun joined12:38
← blacktulip left12:38
← hyperair left12:38
← RaceCondition left12:39
RaceCondishun → RaceCondition12:39
→ resmo__ joined12:39
→ guampa joined12:40
← sangi left12:42
→ hyperair joined12:42
← resmo left12:43
→ Bass10 joined12:45
← jatt left12:45
→ glennpratt joined12:45
← p3rror left12:46
baazigar Hello, I am on errorbranch, and i want to commit the local changes to main branch , to that I have to change the branch to main branch by <git checkout main> but i cant because there are local uncommited changes.12:46
Please help12:46
cbreak-work baazigar: man git-stash12:46
gitinfo baazigar: the git-stash manpage is available at http://git-scm.com/docs/git-stash.html12:46
cbreak-work you'll have to resolve conflicts on stash pop with some likelyhood12:46
baazigar ok i will try that12:47
← Fandekasp left12:47
← archdori left12:47
→ rtuin joined12:48
← jMCg left12:48
baazigar cbreak-work : that helped, thanks12:50
→ jatt joined12:52
→ mastro__ joined12:53
← mastro left12:54
← jatt left12:56
→ achadwick joined12:56
→ mastro joined12:57
← plasticle left12:58
← piotr left12:59
→ piotr joined12:59
← harshpb left12:59
→ obiwahn joined13:00
← mklappstuhl left13:00
← PRETTY_FUNCTION left13:00
obiwahn i wirte a small install script that does for normal users a git fetch && reset --hard13:00
← i42n left13:01
obiwahn but should have the option to pull as well13:01
→ thiago joined13:01
cmn unattended merges are a bad idea13:01
← jimmy1980 left13:01
→ p3rror joined13:01
obiwahn git pull -no-commit13:01
← afuentes left13:02
obiwahn mh should i leave the option away?13:02
→ mklappstuhl joined13:02
cbreak-work if someone wants to pull then he can do that himself13:03
cmn not commiting wins you precisely zero; the problem with merging are conflicts13:03
baazigar cbreak-work : i am getting conflicts when i stash pop, how can i recover that file from stash area?13:04
← RaceCondition left13:04
→ RaceCondition joined13:04
cbreak-work baazigar: useless13:04
→ revagomes joined13:04
cbreak-work resolve the conflict13:04
(you can use git checkout --ours / --theirs but I recommend not doing that)13:04
← robbyoconnor left13:05
baazigar i checked out the new version of that file13:05
the old version is lost?13:05
in the stash area?13:05
→ jimmy1980 joined13:07
← Lisimba3 left13:08
← sebrock left13:09
→ macabre joined13:09
← baazigar left13:10
→ baazigar joined13:12
← erasmospunk left13:13
baazigar someone help please : i need to recover a stashed file13:13
→ OOPMan joined13:14
→ oriba joined13:14
baazigar because i cant resolve a conflict13:14
← arietis left13:14
baazigar ...after stashing and then popping13:14
→ linusoleander joined13:15
→ Fandekasp joined13:15
→ archdori joined13:15
← arvind_khadri left13:16
→ dv310p3r joined13:16
→ maleknet joined13:17
cmn why can't you resolve it?13:18
→ nicesukhi joined13:18
cmn cbreak already gave you the command to put one of the versions in the working directory13:18
baazigar i dont know how, i dont have the previous file version13:19
git checkout --ours?13:19
← gerard0 left13:19
cmn you have a mixture of both versions13:19
← Lgb left13:20
← palm_vue left13:20
baazigar fatal: --ours/--theirs is incompatible with switching branches.13:20
cmn and you need to figure out what the conflicting parts should look like13:20
→ gerard0 joined13:20
→ AaronMT joined13:21
baazigar cmn : git checkout testing gives13:21
Scribus/scribus/ui/newfile.cpp: needs merge13:21
error: you need to resolve your current index first13:21
→ Lgb joined13:21
→ palm_vue joined13:21
cmn right13:21
baazigar how to do that merge?13:21
and resolve the index13:22
and i want to go back to the state before stashing13:22
← crs left13:22
baazigar because stashing created all problems13:22
cbreak-work text editor13:22
stashing didn't create any problems13:22
→ mkj_wk joined13:23
→ dcarper joined13:23
cmn you /are/ doing the merge13:23
→ hellopat joined13:23
baazigar my code is in the stash,13:23
cmn that's why you have conflicts; solve them13:23
← _Vi left13:23
baazigar i need the code from stash13:24
cbreak-work no13:24
git status/git diff13:24
cmn not really, and you already have it13:24
cbreak-work find the conflict13:24
open with text editor13:24
fix13:24
git add to resolve13:24
done13:24
baazigar i will check that again13:24
→ gyre007 joined13:24
← nicesukhi left13:24
→ nicesukhi joined13:26
→ nadavoid joined13:26
baazigar aha thanks, cbreak-work , resolved it finally, maybe i missed the git add step last time :P13:30
→ hwrd|work joined13:30
← mklappstuhl left13:30
← uraimo left13:30
← Innocentus left13:31
← hellopat left13:31
← cooloney left13:31
→ uraimo joined13:31
→ hellopat joined13:31
→ Innocentus joined13:32
→ PrinceAMD- joined13:32
→ gdmfsob joined13:32
← OOPMan left13:33
← The-Compiler left13:33
→ OOPMan joined13:34
← piotr left13:35
→ piotr joined13:35
← mishok13 left13:35
← piotr left13:36
→ Hounddog joined13:36
← TheJH left13:37
→ juanfer joined13:37
→ ehsan joined13:37
← maleknet left13:37
Hounddog is there a way to somehow prevent git from mergin a certain file through .gitattributes or something? i have seen that there is something with an sh file and config change but we have no ssh access or anything on that server to apply this...13:37
→ Dreamer3 joined13:38
→ stuntmachine joined13:38
← MiJyn left13:38
→ MiJyn joined13:38
← stuntmachine left13:39
→ piotr joined13:40
→ yoklov joined13:41
← beautifulmind left13:41
→ harshpb joined13:42
→ john2496 joined13:42
→ TeckniX joined13:44
← ehsan left13:44
cmn something with an sh file?13:44
Hounddog yeah13:44
i can't do that...13:44
our admins are securing everything so nicely...13:45
← jubei left13:45
cmn you call tell git to simply mark a file as conflicted and let you resolve it13:45
Hounddog cmn: this is on jenkins13:45
i cannot resolve everything13:45
ok... just for you to understand13:45
cmn I'm asking because "sh file" makes no sense in the case of a remote repo13:45
Hounddog yeah, let me just explain what i am doing13:45
cmn why on earth is jenkins even merging anything?13:45
FauxFaux "it wrong".13:46
→ Ecesis joined13:46
← tsdh left13:46
→ LittleIdiot joined13:47
Hounddog i have an autobuild.cfg in which i define which branches shall be autobuilt. this file i read out by a post receive on the remote. then i decide to do only a branch build or let jenkins merge it and build it.13:47
so it is wrong to let jenkins merge it?13:47
why?13:47
cmn if the merge isn't clean, there's no hope of doing anything sane13:47
Hounddog i let jenkins merge the branch on the master run the build and then only it get's pushed to master13:47
if the build with all the tests are successfull ;)13:48
← lnostdal__ left13:48
cmn but that can only happen on a clean merge13:48
← PrinceAMD- left13:48
Hounddog if there is a conflict it just aborts13:48
and you have to do it manually and push again to your branch13:48
this is to avoid people pushing on to the master13:48
cmn right, so what's the problem?13:49
← nicesukhi left13:49
FauxFaux Because it conflicts...13:49
Hounddog this autobuild.cfg shall not be merged but has to be pushed for the post-receive to evaluate if it shall be built on a personal branch or on to the master13:49
FauxFaux: it conflicts?13:49
have not had problems with that till now.13:50
← sitaram left13:50
Hounddog if we have a conflict it will simply not merge13:50
RandalSchwartz Hounddog !config13:50
gitinfo [!configfiles] the recommended way is to change your application so that you can (perhaps optionally) store site-specific configuration/data in separate files, then keep those files out of version control. Still, see https://gist.github.com/1423106 for ways people have worked around the problem.13:50
→ eijk joined13:50
→ sitaram joined13:50
cmn and you've already said how you deal with conflicts, you let a human solve it13:50
Hounddog RandalSchwartz: thx. as mentioned. i have no access to the config...13:50
← piotr left13:51
cmn you have no access to the file you're trying to merge? what?13:51
Hounddog .git/config i have no access to13:51
→ piotr joined13:51
cmn so?13:51
that file isn't tracked13:51
and has nothing to do with anything13:51
Hounddog hmm ok13:51
← dabd left13:52
→ jhegge joined13:52
→ Amicasa joined13:52
Hounddog ok, if it is not tracked... how to read it from the post-recieve hook?13:52
cmn however you want, it's at $GIT_DIR/config13:53
← Amicasa left13:53
cmn but that's repo-specific13:53
and you haven't mentioned why that file is relevant13:53
← kermit left13:53
Hounddog i havent?13:54
15:49 <Hounddog> this autobuild.cfg shall not be merged but has to be pushed for the post-receive to evaluate if it shall be built on a personal branch or on to the master13:54
15:47 <Hounddog> i have an autobuild.cfg in which i define which branches shall be autobuilt. this file i read out by a post receive on the remote. then i decide to do only a branch build or let jenkins merge it and build it.13:54
anyway13:54
← JohnBat26 left13:54
cmn since when is autobuild.cfg and .git/config the same file?13:54
Hounddog http://pastie.org/412645213:54
that is the file which i need to ignore13:54
from the merge13:54
→ erasmospunk joined13:54
RandalSchwartz Hounddog - did you not read what I sent you?13:54
I'm not talking about git config13:55
Hounddog am trying to understand it13:55
RandalSchwartz I'm talking about how to deal with files that should not be tracked because they are site-local13:55
I didn't read the whole scrollback, but that's usually the cause of things you're reporting as symptoms13:56
← gyre007 left13:56
Hounddog but this is just like creating autbuild.predefined.cfg which can be tracked and from there everyone shall only create their own autobuild.cfg13:56
i need it tracked. to read it on the remote13:56
RandalSchwartz Hounddog it should create autbuild.predefined.cfg.sample and check that in13:56
cmn so track it13:56
RandalSchwartz then everyone copies that to their own .cfg13:57
Hounddog i am tracking it13:57
RandalSchwartz NOT in the repo13:57
Hounddog ok...13:57
→ Destos joined13:57
RandalSchwartz then nobody has conflicts when you update it. :)13:57
← Stormbytes left13:57
cmn RandalSchwartz: that doesn't work when the build is being done on a different computer13:58
Hounddog yes nice but also my hook is obsolete cause it cannot read a non existant autobuild.cfg13:58
RandalSchwartz oh13:58
Hounddog i think i am not explaining myself correctly13:58
RandalSchwartz sounds like a mistake in design then13:58
cmn I was about to say the same; it sounds like you need a better way of describing what should happen13:58
because if you want to track it, it's going to get merged13:58
you can use merge=ours IIRC in the attributes, but that's bound to backfire13:59
Hounddog i tried that in the gitattributes but still merged it from the other branch14:00
http://pastie.org/412649014:00
← xiangfu left14:00
→ polymar joined14:00
← rearth_ left14:00
← soee left14:00
Hounddog this is what i have... now if i want my branch to be autobuilt i just add mybranch=autobuild and push it. the server receives it and runs the job to merge build the system14:00
but this should not get into master nor should the master file then get back into my branch...14:01
i saw the sh solution...14:01
LittleIdiot i've got a problem: "git clone https://android.googlesource.com/kernel/samsung.git" downloads 474M of data, but no files are in the directory... there's just a .git directory in it with two .pack files that seem to contain the actual contents, yet git doesn't unpack it - what am i doing wrong?14:03
← _sundar_ left14:03
FauxFaux LittleIdiot: What does git clone print?14:03
← thiago left14:05
→ JohnBat26 joined14:05
← baazigar left14:05
mkj_wk I have run into an line ending normalization problem that I haven't figured out how to resolve, after reading docs and googling... Some developers are using egit on windows, which seems to be not honoring text=auto in .gitattributes. When, running git on Linux, I check out commits in which egit one windows has failed to normalize CRNL->NL, I can't switch branches without first committing line ending normalization because my working directory is dirty, and I14:05
→ kukks joined14:05
← kukks left14:05
→ kukks joined14:05
LittleIdiot FauxFaux: just a sec, have to find the shell where i pulled it ;)14:05
← harshpb left14:05
→ andrewhl joined14:06
← archdori left14:06
← Fandekasp left14:06
← Ecesis left14:06
← Lord_DeathMatch left14:06
→ harshpb joined14:06
→ tyatpi joined14:07
→ ehsan joined14:07
→ eurbach joined14:07
PerlJam LittleIdiot: What makes you think *you're* doing something wrong? :)14:07
→ barq joined14:09
← Desproges left14:10
PerlJam LittleIdiot: when I browse around on android.googlesource.com, there's not much there. and nothing in kernel/samsung14:10
LittleIdiot PerlJam: there's the complete output: http://ix.io/2AB14:11
← cagedwisdom left14:11
← oo_miguel left14:12
mkj_wk is that repository intended to be used with the Android "repo" tool?14:12
FauxFaux It should have then printed more stuff.14:12
→ macmartine joined14:13
← ph^ left14:13
← codeM0nK3Y left14:13
← harshpb left14:13
→ codeM0nK3Y joined14:13
→ ph^ joined14:13
→ _sundar_ joined14:14
→ harshpb joined14:14
LittleIdiot mkj_wk: i thought so too, but this is the exact command that they provide in their official documentation (http://source.android.com/source/building-kernels.html)14:14
← zinjacoder left14:14
← und3f left14:14
→ und3f joined14:15
← h0bbit left14:15
← leilaM left14:16
← ShooterMG left14:16
← ToBeFree left14:16
mkj_wk LittleIdiot: oh, well then! :)14:17
→ baazigar joined14:17
← ph^ left14:18
← mobilegamelabs left14:19
→ jjbohn joined14:19
← ashphy left14:20
← marhaban left14:20
→ dhruvasagar joined14:20
→ ashphy joined14:20
LittleIdiot is there something i'm missing? some magical command or mechanism in git?14:20
← d3c left14:21
LittleIdiot or "may be missing" - i know you can't read minds... at least i hope ;)14:22
← axl_ left14:22
← PigeonFriend left14:23
→ axl_ joined14:23
← gdmfsob left14:23
← axl_ left14:23
→ Jay_Levitt joined14:24
haarg LittleIdiot: what do you get from running git log kernel?14:24
← ashphy left14:24
haarg it looks like those instructions want you to check out a specific commit after doing the clone14:25
← rhl left14:25
← piotr left14:25
→ piotr joined14:26
→ beatak joined14:26
→ axl_ joined14:27
→ zinjacoder joined14:28
cmn LittleIdiot: the default branch is emtpy, checkout one of the others14:28
LittleIdiot haarg: i've checked out the entire tree, so when doing that in the path device/samsung/crespo4g, "git clone" prints " unknown revision or path not in the working tree"14:28
i just found out that there are branches other than master, but i wonder why git branch --list did just show the master branch then14:29
→ leilaM joined14:29
cmn because that's the only local one14:29
← _sundar_ left14:30
LittleIdiot ahh ;)14:30
← baazigar left14:31
→ fly9 joined14:31
← zinjacoder left14:31
→ _sundar_ joined14:32
LittleIdiot okay, it works now when specifying the branch14:33
→ quazimodo joined14:33
← jcromartie left14:33
LittleIdiot just didn't they were only local - thank you guys! ;)14:33
← asakura1 left14:34
→ blacktulip joined14:34
← jason237 left14:35
← linusoleander left14:36
comps how can I push to remote repo's HEAD (bare)? .. is it possible or does it have to be set manually (locally)?14:37
→ lacrymology joined14:37
_ikke_ comps: What do you want to achieve?14:37
cbreak-work push to a branch14:37
← jargon- left14:37
→ airborn joined14:38
cbreak-work HEAD is a symref, you can only change it locally14:38
comps _ikke_: to set the default branch to be checkouted on clone14:38
cbreak-work: thanks14:38
cbreak-work (and normally that's not done)14:38
→ duckxx joined14:38
cbreak-work comps: man git-symbolic-ref for details14:38
gitinfo comps: the git-symbolic-ref manpage is available at http://git-scm.com/docs/git-symbolic-ref.html14:38
→ ryez joined14:39
comps cbreak-work: yes, but I hoped it can be done remotely14:39
→ troulouliou_dev joined14:39
_ikke_ nope14:39
← fayimora left14:40
→ PigeonFriend joined14:41
← axl_ left14:42
→ ipoval joined14:42
→ eletuchy joined14:43
→ axl_ joined14:43
troulouliou_dev hi i m trying to push to origin master on github and the command always freeze at " POST git-receive-pack (chunked)"14:43
any idea ?14:43
→ marhaban joined14:43
EugeneKay !ssh14:43
EugeneKay pokes gitinfo with a stick14:44
→ Vile joined14:44
gitinfo Please use SSH to talk to remote repos. http:// and git:// are completely unsecured. You can try to configure "smart" https:// access, but it is usually much easier to just stick with ssh://14:44
← andrewhl left14:44
← dhruvasagar left14:44
→ andrewhl joined14:44
← srinivas07 left14:45
← rchavik left14:45
← kenperkins left14:45
troulouliou_dev EugeneKay, ok will try that14:45
→ kermit joined14:46
← wilmoore left14:46
→ axl__ joined14:47
← axl__ left14:47
← Chib left14:47
→ jason237 joined14:48
← axl_ left14:48
→ asakura1 joined14:49
infogulch|afk → infogulch14:49
← jhardinee left14:51
→ axl_ joined14:51
← Araxia_ left14:52
← marhaban left14:52
→ dunpeal_ joined14:52
→ goshawk joined14:53
← flijten left14:53
← PigeonFriend left14:53
← dunpeal_ left14:53
→ dunpeal joined14:54
dunpeal Hey. If a developer sets `git config --global core.eol lf`, does that mean git will convert all CRLFs to LFs in all his code files?14:55
(Upon commit)14:55
→ PrinceAMD- joined14:55
← axl_ left14:56
→ berserkr joined14:56
→ axl_ joined14:57
FauxFaux dunpeal: !tias14:57
gitinfo dunpeal: Try it and see™. You learn much more by experimentation than by asking without having even tried. If in doubt, make backups before you experiment (see !backup). http://sitaramc.github.com/1-basic-usage/tias.html may help with git-specific TIAS.14:57
→ pranjal710 joined14:57
→ dfr|mac joined14:57
← axl_ left14:57
← tabakhase left14:58
→ tabakhase joined14:58
← ryez left15:00
← polymar left15:00
← PrinceAMD- left15:00
dunpeal FauxFaux: I tried it in Linux and it worked, but I don't have Windows XP, Windows 7, etc.?15:01
→ axl_ joined15:01
← macmartine left15:01
EugeneKay !crlf15:02
gitinfo to fix problems with line endings on different platforms, check out http://line-endings.gitrecipes.de/15:02
EugeneKay !read15:02
.trigger_edit read Don't expect everything to be spoon fed to you - the man pages and other documents do not bite, you need to spend some time and effort reading them.15:02
gitinfo EugeneKay: Okay.15:02
→ fluxdude joined15:02
← marcogmonteiro left15:03
→ Targen joined15:03
← axl_ left15:04
→ SegFaultAX joined15:04
dunpeal Uhmm, ok.15:04
← dunpeal left15:04
← ThomasLocke left15:04
← leilaM left15:05
→ PigeonFriend joined15:05
→ setmeaway2 joined15:05
→ axl_ joined15:05
← rtuin left15:06
← fluxdude left15:06
→ shtrb joined15:07
→ FreeArtMan joined15:07
→ sebrock joined15:08
← Innocentus left15:08
← axl_ left15:08
← setmeaway left15:09
← cbreak-work left15:09
← hafos left15:10
← FreeArtMan left15:10
→ axl_ joined15:10
shtrb [running on Win7] - Is there a way I could take a git svn clone and push it inside an svn directory without bringing havoc ? I have few gigabytes of source code under svn control - I wish to switch to git-svn what I clone with git svn download all in a diffrent dir but now I wish only to bring the git structure to where the SVN lives15:11
→ ph^ joined15:11
→ Twigg joined15:12
shtrb for example : I have c:\source I do git svn to c:\temp and wish to bring all the git stuff to c:\source (without the sources files them self) the question how to do that (I would still need to be able to do svn update/commit etc)15:13
FauxFaux shtrb: It should work if you just move the .git folder across.15:13
→ The-Compiler joined15:13
← harshpb left15:13
→ thiago joined15:14
shtrb hmm .. when I just moved all .git dirs I was getting the git client to crash15:14
The-Compiler Heya. Is there an easy way to get the diff of changes introduced with a commit-id? So, a more simple way to write git diff commitid~1 commitid ?15:14
FauxFaux The-Compiler: git show commitid, or git diff commitid^!15:14
→ harshpb joined15:14
barq How can I move my git repository from my local comp to my server?15:15
cmn move as in move all of it or transfer the history?15:16
EugeneKay barq - read chapter 4 of the !book15:16
gitinfo barq: 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 !designers !gitt !vcbe and !parable15:16
→ savid joined15:16
← thierryp left15:17
barq cmn: Move the whole thing and delete the local repository.15:18
← fission6 left15:18
cmn use your favourite file tranfer tool15:18
but you probably don't want to15:18
→ notmypudding joined15:18
barq I don't want to?15:18
cmn there is generally no good reason why you'd have a non-bare repo on anything considered a "server"15:19
barq Where else should it be?15:19
cmn wherever you work on it15:20
← PigeonFriend left15:20
cmn why else would you have files?15:20
← yoklov left15:20
cmn read that chapter of the book15:20
barq And how do the different users sync without a server?15:20
cmn 'cos it seems the bare/non-bare isn't clear to you15:20
again, that's what the docs are for;15:20
→ QaDeS joined15:21
cmn and I never said you wouldn't use a server15:21
→ davidjdagino joined15:21
← barmaglot left15:21
cmn the issue is the /non-bare/ repo15:21
→ fgro joined15:22
btree__ → btree15:22
← orbital left15:22
barq Aight, I'll read then15:23
fgro my fellow develper commmit some changes in the wrong branch. I'd like to undo the commits and move them to a new branch. How to do best?15:23
← harshpb left15:23
FauxFaux fgro: !fixup15:23
gitinfo fgro: 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!15:24
fgro it was actually one one commit15:24
thanx15:24
→ jaylevitt joined15:25
→ Mkaysi_ joined15:25
← erichynds left15:26
→ jbohn joined15:26
← jjbohn left15:27
← ianawilson left15:28
← barq left15:28
→ AeroNotix joined15:28
← ipoval left15:28
← Targen left15:28
→ mikepack joined15:29
← Jay_Levitt left15:29
→ wilmoore joined15:30
comps just out of curiosity: can I somehow do git checkout --orphan without creating a new local branch? .. like putting all zeros to .git/HEAD15:30
Mkaysi_ → Mkaysi15:30
← thiago left15:30
comps ie. HEAD would be moved to the newly created initial commit15:30
(still detached)15:30
← jbohn left15:31
← wilmoore left15:31
→ ianawilson joined15:31
→ wilmoore joined15:31
→ stkrzysiak joined15:31
→ PigeonFriend joined15:32
← erasmospunk left15:32
SethRobertson comps: I think you can probably create a detached orphan commit if you try hard enough, but it surely will be easier to delete the branch afterward15:35
→ erasmospunk joined15:35
SethRobertson comps: This might do it: git checkout `echo "empty" | git commit-tree 4b825dc642cb6eb9a060e54bf8d69288fbee4904`15:36
→ mishok13 joined15:37
→ kenperkins joined15:38
→ thornofmight joined15:39
gitinfo set mode: +v15:39
→ cagedwisdom joined15:40
← eurbach left15:42
→ harshpb joined15:43
← harshpb left15:43
comps SethRobertson: I know it can be probably done with a custom set of commands that create an empty tree and use it for git-commit-tree (and update HEAD afterwards), but I was hoping for a --orphan revspec trick that would do the same thing15:44
→ harshpb joined15:44
← harshpb left15:44
comps SethRobertson: though there probably isn't any, nevermind then, thanks for the info15:45
→ harshpb joined15:45
→ macmartine joined15:45
← blacktulip left15:45
→ robbyoconnor joined15:47
→ aurum23 joined15:47
← mkj_wk left15:47
← uraimo left15:47
← adhawkins left15:48
→ mischat_ joined15:48
← totte left15:48
→ yoklov joined15:49
→ virunga joined15:49
→ barneybook joined15:49
← barneybook left15:49
virunga Hi, to download my repo on github should i clone it?15:49
→ barneybook joined15:49
← JeremyWei left15:50
virunga download the repo to my machine15:50
wereHamster maybe15:50
→ igotnolegs joined15:50
← kpreid left15:50
← mischat left15:50
mischat_ → mischat15:50
virunga wereHamster: are there other methods?15:51
→ needsch joined15:51
→ yaymukund joined15:51
← pureluck left15:51
→ dr_diamond joined15:52
← dr_diamond left15:52
→ dr_diamond joined15:52
thorbjornDX virunga: I think you can get just the source as a .zip or .tar.gz15:52
→ fission6 joined15:52
cmn depends on what you mean by "download"15:52
thorbjornDX virunga: but if you want to modify it and push, you definitely want to clone15:52
virunga thorbjornDX: yes, i want to do that, thank you :)15:53
← adityamenon left15:53
← aurum23 left15:54
→ suedoh joined15:54
→ avandenhoven joined15:55
suedoh hi all, i just removed a huge directory from my repo and pushed to HEAD. Still it seems a pull is trying to receive the objects in that large directory15:55
This causes a crash when doing a clone on a VM w/ small amount of memory15:55
cmn the dir is still in the history15:55
suedoh Any suggested way of permanently removing or removing from HEAD so that it doesn't bother grabbing anything about this directory on a clone15:56
→ lotus joined15:56
cmn man git filter-branch15:56
gitinfo the git-filter-branch manpage is available at http://git-scm.com/docs/git-filter-branch.html15:56
→ ToBeFree joined15:56
cmn but be aware of !rewrite15:56
gitinfo [!rewriting_public_history] 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)15:56
suedoh cmn, I've done: sudo git filter-branch --index-filter "git rm -rf --cached --ignore-unmatch ./incrementals/*" HEAD15:56
cmn upgrading git should also help deal with larger data15:56
suedoh I already tried this :-(15:56
lotus hey, I have a dev branch and a live branch. I have a .htaccess file which differs between the two branches. When I merge the two branches, I want it to consistently not merge the .htaccess file. How should I set this up?15:56
suedoh I am running version 1.7.915:56
I believe this is a good version15:56
cmn then that command didn't take or wasn't made correctly15:57
suedoh lotus, consider the use of symlinks?15:57
cmn other branches can still have that information15:57
and your branch might not have been rewritten15:57
lotus suedoh, doing the work in windows15:58
suedoh Rewrite b145bc6315e7aeba9b826f21e827ed635c0ee6e4 (32/32)error: duplicate parent a5b3f576b384c9ab0404d3e7c3cde8eb20026eb6 ignored15:58
should I worry about this?15:58
← troulouliou_dev left15:58
lotus suedoh, symlinks will be more trouble than they are worth. Also, I'm not sure how they'd help.15:58
obiwahn how can i ginore untracked files for the moment whithout deleteing them15:58
when i do a git diff --name-only15:58
← dfr|mac left15:59
→ davpoind joined16:00
suedoh cmn, could you recommend the command to run for removing an entire directory?16:01
← davpoind left16:01
← andrewhl left16:01
cmn that command looks fine, the manpage tells you what to do16:01
→ davpoind joined16:01
cmn but you may not have rewritten any branchesw16:01
suedoh I'm just confused by the following after that command:16:01
cmn or at least not all of them16:02
suedoh Rewrite b145bc6315e7aeba9b826f21e827ed635c0ee6e4 (33/34)error: duplicate parent a5b3f576b384c9ab0404d3e7c3cde8eb20026eb6 ignored16:02
Rewrite 7dc0f37cea91415039255623129c062a27ae1246 (34/34)error: duplicate parent 4718b410f54bf6751e73caccf33d726e0af87317 ignored16:02
Ref 'refs/heads/master' was rewritten16:02
]$ sudo git filter-branch -f --index-filter "git rm -rf --cached --ignore-unmatch ./incrementals/*"16:02
← hwrd|work left16:02
cmn no idea what that error means16:02
do you have other branches?16:02
suedoh After this command a 'git push' states the following:16:02
To ssh://gitro@108.166.212.51:26022/usr/local/git_root/wwwmotion.git16:02
! [rejected] master -> master (non-fast-forward)16:02
rror: failed to push some refs to 'ssh://gitro@108.166.212.51:26022/usr/local/git_root/wwwmotion.git'16:03
→ ferdna joined16:03
suedoh To prevent you from losing history, non-fast-forward updates were rejected16:03
confused16:03
_ikke_ suedoh: You have to pull first16:03
suedoh: The remote has changes you don't have locally16:03
suedoh _ikke_, I've done this before all this work16:04
so git pull claims 'Already up-to-date.'16:04
_ikke_ suedoh: Oh, I see you have altered history16:04
suedoh Then I run the remove16:04
_ikke_ !rewrite16:04
gitinfo [!rewriting_public_history] 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)16:04
cmn so the push was rejecteed16:05
thus the new history isn't there16:05
and pulling in that case is an extremely bad idea16:05
suedoh git push -f doesn't go through16:05
what should I do now?16:06
_ikke_ suedoh: What do you mean with, doesn't go through16:06
→ robotmay_ joined16:06
suedoh okay, let me paste16:06
should have done that from first :)16:06
← barneybook left16:07
_ikke_ yes16:07
← whitman left16:07
cmn doesn't go through?16:07
if your admin doesn't allow you to rewrite history, then that's that16:07
→ fl0w joined16:08
← robotmay left16:08
→ mobilegamelabs joined16:08
→ de_ joined16:08
→ mmc2 joined16:08
← oriba left16:08
→ pureluck joined16:09
→ jargon- joined16:09
→ sq-one joined16:09
→ FreeArtMan joined16:10
suedoh _ikke_, cmn, http://ideone.com/UjFVE16:10
→ egarcia joined16:10
suedoh This is the run through/break down of the commands I am running\16:11
and the responses I receive16:11
cmn and the force-push that you say doesn't work?16:11
suedoh no16:12
sudo git push -f, doesn't work16:12
← nadavoid left16:12
_ikke_ !doesn't16:12
suedoh it says rejected16:12
gitinfo [!doesntwork] 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!16:12
→ pdtpatrick joined16:12
suedoh Sorry16:12
_ikke_ Then the admin has disabled non-ff pushes16:13
→ aantix joined16:13
suedoh error: failed to push some refs to 'ssh://gituser@x.x.x.x/usr/local/git_root/project.git16:13
← senny left16:13
← john_anderson left16:13
← aantix left16:13
suedoh _ikke_, I'm not sure how I got into this fast forward stage16:14
→ senny joined16:14
→ john_anderson joined16:14
→ nadavoid joined16:14
cmn why on earth are you doing sudo?16:14
→ aantix joined16:14
← und3f left16:14
_ikke_ :-|16:14
cmn and what is the error message?16:14
→ psakrii joined16:14
suedoh Is there any way to enable fast forward, so I can get past this hurdle16:14
_ikke_ Do you have access to the repo in question?16:15
← joshteam left16:15
→ Vile1 joined16:15
suedoh what does this mean:16:16
Merge made by the 'recursive' strategy.16:16
?16:16
← Vile left16:16
← robotmay_ left16:17
cmn fast-forward is the normal case16:17
→ leilaM joined16:17
→ apok joined16:17
→ maletor joined16:17
cmn non-fast-forwards are when you need to force-push16:17
→ jackyyll joined16:17
← mikepack left16:17
cmn that message means it's merging; why are you merging if you're tring to overwrite history?16:17
→ hwrd|work joined16:18
suedoh i'm not, it's doing it auto-magically16:18
→ mwrosen joined16:18
← de_ left16:18
← jaylevitt left16:18
← senny left16:19
jackyyll hi i accidentally checked out a new branch, from another working branch that was ahead of my development branch.. so it had changes that aren't supposed to be in my new branch that i just created.. And then i made changes and commited them to the new branch.. Is there a way to remove the changes in the new branch that are from the branch i accidentally checked the new branch out from?16:19
cmn git doesn't merge automagically16:19
← nadavoid left16:19
→ jcromartie joined16:19
cmn and if you don't want to say what you're doing, there's no way to try to figure out what's happening16:19
jackyyll that way it matches development branch, but still has the changes i've made in the new branch16:19
→ alberto56 joined16:21
→ mikepack joined16:21
← axl_ left16:21
← psakrii left16:21
→ JoshuaA joined16:22
← project2501a left16:22
→ ottoshmidt joined16:22
← LimeBlast left16:22
JoshuaA Hey guys, is there a way to have git log show two repos, side by side? (preferably even as to branches, or chronologically sorted)16:23
jackyyll anyone? :(16:23
i need to merge my new branch into the development branch but i cna't have the changes in it that are from the accidental checkout16:24
charon JoshuaA: fetch one into the other and use a normal log?16:24
shtrb about the git-svn with svn in the same dir , I started to test modified and commited with svn but when I do git rebase I get (filename that I updated via svn: needs update \n update-index --refresh: command returned error: 1) how can I fix that ?16:24
charon jackyyll: !fixit16:24
gitinfo jackyyll: [!fixup] 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!16:24
→ nadavoid joined16:24
→ axl_ joined16:24
← berserkr left16:24
charon shtrb: tracking the same dir with both git-svn and svn sounds like a recipe for disaster. !repro16:25
gitinfo shtrb: 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.16:25
→ h0bbit joined16:25
← tjadc left16:25
shtrb charon , modifed a file localy and commited via svn , then I wish to bring git files to update with svn and did git rebase (now preparing a transcript to paste here)16:26
→ Shusshu joined16:27
← jackyyll left16:28
suedoh yea, I give up16:28
man, I don't know what I did to get here16:28
← Shusshu left16:28
suedoh Rewrite b145bc6315e7aeba9b826f21e827ed635c0ee6e4 (37/42)error: duplicate parent a5b3f576b384c9ab0404d3e7c3cde8eb20026eb6 ignored16:28
← harshpb left16:28
suedoh That's the problem ^16:28
← Hounddog left16:29
→ harshpb joined16:29
← harshpb left16:29
← klj613 left16:29
→ robottinosino joined16:29
→ harshpb joined16:30
JoshuaA Is there a way to merge the logs of the two branches after a subtree merge?16:31
As in, history will then appear chronologically?16:31
→ pandeiro joined16:32
FauxFaux JoshuaA: man git log --[something]-order16:33
gitinfo JoshuaA: the git-log manpage is available at http://git-scm.com/docs/git-log.html16:33
→ axl__ joined16:33
← axl__ left16:33
JoshuaA FauxFaux: ThanksQ16:34
→ axl__ joined16:34
← JohnBat26 left16:34
→ Einzig_ joined16:35
shtrb charon , my intention to track the dir with both svn and git (here is some transcript - http://pastebin.com/fdueaJZR )16:35
s/git/git-svn16:35
Einzig_ hello, I wonder if I can push only a specific commit, instead of all commits to a remote server.16:36
and which number can I use to identify it16:36
EugeneKay No, that's not how git's history model works.16:36
← BarryCrayon left16:36
← axl_ left16:36
axl__ → axl_16:36
charon shtrb: so umm. you change a file behind git's back, then you are surprised that it doesn't work?16:36
→ noisebleed joined16:36
← noisebleed left16:36
→ noisebleed joined16:36
suedoh so any ideas on duplicate parent error?16:37
FauxFaux Einzig_: You can push commits /up to/ a certain point, and not things after there, quite easily.16:37
charon shtrb: i think you can solve this particular problem simply by letting git redo the change: git reset --hard && git svn rebase16:37
shtrb: but i stick to my statement: managing it in both is a recipe for disaster16:37
suedoh only with the filter-branch?16:37
→ thiago joined16:37
→ williamcotton joined16:37
Einzig_ FauxFaux: I want to commit only some script, other code is broken and needs fixing first16:38
so ommit the code commits and push the documentation and scripts16:38
← mikepack left16:38
FauxFaux The easiest way is probably to branck from the remote tracking branch (@{upstream}) and cherry-pick what you want.16:39
Then rebase your old master atop the new branch. But, y'know, if taht doesn't make any sense you're pretty much screwed. !fixup16:39
gitinfo 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!16:39
→ arietis joined16:39
← ben_alman left16:39
shtrb charon , so baiscly when I would screw up the files behind git's back I could allways do hard reset ? what about local modification (not on the rep ) could I work around it ?16:39
← harshpb left16:40
→ dfr|mac joined16:40
→ etcetera joined16:41
→ tewecske joined16:41
← francisfish left16:41
→ Araxia_ joined16:41
← mischat left16:42
→ psakrii joined16:42
→ jrots joined16:43
← OOPMan left16:44
→ joshteam joined16:44
← stodan left16:45
→ lindenle joined16:45
→ Sky[x] joined16:45
→ tonesfrommars joined16:45
→ axl__ joined16:45
← axl__ left16:46
→ mxweas_ joined16:46
→ axl__ joined16:47
→ imrobert joined16:47
→ andrewhl joined16:47
→ axl___ joined16:47
→ bilel joined16:48
← achadwick left16:48
← axl_ left16:48
→ axl____ joined16:48
← axl____ left16:48
→ apok_ joined16:48
→ kmap2 joined16:49
→ admsyn joined16:49
gitinfo set mode: +v16:49
→ axl_ joined16:49
bilel hi, is there a simple way to do a branch that we don't want to track? something like a sandbox, but I don't want to track change history, it's only for tests things like that.16:49
→ cmullican joined16:50
→ ben_alman joined16:50
→ iwoj joined16:50
→ scientes joined16:51
→ mischat joined16:51
EugeneKay bilel - just make the changes, then `git reset --hard HEAD` when you wanna get back to normal.16:51
← axl__ left16:51
← apok left16:51
apok_ → apok16:51
← ziro` left16:51
← dr_lepper left16:52
→ pfFredd joined16:52
suedoh Rewrite 42b3bde9474fb428212c302db3691f2b11f9fcdc (48/48)error: duplicate parent 5de0fa27edae02d2fa0824ed9932c78fe34217a2 ignored16:52
Ref 'refs/remotes/origin/master' was rewritten16:52
Ref 'refs/heads/master' was rewritten16:52
cmullican i created a branch, and committed my working files, and it gave a bunch of create mode ... [path] and also some delete mode ... [path] output, and now, when I tried to switch back to master, it tells me "error: Untracked working tree file 'html/display_account_export.html' would be removed by merge.", only I see that file in the web UI for the repo, and it's in my directory. git status shows a LOT of changes I don't want, including dele16:52
← axl___ left16:52
suedoh Please help me get rid of this, I've tried every command set that I could find online, and nothing gets me past this stagte16:52
charon shtrb: dunno. i haven't thought about it for too long. i suppose it would work if, every time you switch vcs, you issue 'git reset --hard' or 'svn revert', depending16:53
shtrb: but what problem are you trying to solve?16:53
by using both, i mean16:53
suedoh is there something I am missing?16:53
← erasmospunk left16:54
suedoh I do a git pull, a git pull --rebase, git fsck, git fsck-full, git push origin master, git push, all these things16:54
← dc5ala left16:54
suedoh but when I run this command: sudo git filter-branch -f --index-filter "git rm -rf --cached --ignore-unmatch ./incrementals/*" origin master16:54
it gives me those errors16:54
and it doesn't allow me to push my changes16:54
→ brentratliff joined16:55
← fuzzybunny69y left16:55
→ codebeaker joined16:55
→ dreiss joined16:55
→ erasmospunk joined16:56
← apok left16:56
← Fudge left16:56
← shruggar left16:56
→ apok joined16:56
→ PrinceAMD- joined16:57
shtrb charon, currently I bound to have some scripts that will execute svn on a source tree (rarly called) , so until I fix them I wish to coninue to use git-svn (it would be a rare case to use both vcs but I need to know how to fix that if it happens)16:57
charon suedoh: i think duplicate parents are not an error16:57
suedoh What is your local repository name ? I know that my remote is origin master16:57
→ fuzzybunny69y joined16:57
suedoh charon, Yes, but I am unable to check in the changesd16:57
*changes16:57
→ OOPMan joined16:57
charon suedoh: that's highly surprising when you filter-branch. </sarcasm> see !rewriting16:57
gitinfo suedoh: [!rewriting_public_history] 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)16:57
shtrb charon, I had two diffrent source trees and I realized I could work in only one (I have one for git-svn and another for only svn)16:57
← marc_kdab left16:58
← jrots left16:58
→ alvaro_o joined16:58
bilel EugeneKay: in fact I want to keep the state somehow, it's just about the history I don't want to keep it as it's only for tests and a local use16:58
EugeneKay bilel - man git-stash may be of interest.16:58
gitinfo bilel: the git-stash manpage is available at http://git-scm.com/docs/git-stash.html16:58
suedoh charon, BTW, I've done all of what is said above by gitinfo bot16:58
← alvaro_o left16:58
← caddoo left16:58
→ Fudge joined16:58
EugeneKay You still keep the history, but it's not in a conventional branch16:58
→ osmosis joined16:59
EugeneKay I would just create a testing branch, but not configure it to be pushed16:59
charon suedoh: on the contrary, if you actually ran 'git pull --rebase' in the same repo as the one that did the filtering, you just blindly ran the commands from the quote16:59
← john_anderson left16:59
charon suedoh: at the very least read !gcs and some explanation of what filter-branch does, so that you will understand16:59
suedoh !gcs16:59
← aruiz left16:59
charon suedoh: then force the push. if the server denies forced pushes, you're out of luck16:59
hrm16:59
http://sitaramc.github.com/gcs/17:00
suedoh so if I am denied forced pushes, what option do I have?17:00
charon delete the ref remotely, and re-push17:00
← jargon- left17:00
FauxFaux .trigger_edit gcs http://sitaramc.github.com/gcs/17:00
gitinfo FauxFaux: Okay.17:00
charon but the admin may have wisened up to that trick, aand deny deletes17:00
suedoh delete the ref?17:00
charon in which case you're screwed17:00
← mischat left17:00
suedoh can't I just check or clone it again?17:00
charon yes, push --delete origin foo && git push origin foo17:00
← rurufufuss left17:01
charon huh17:01
suedoh shit, i might have to standby on this one17:01
← jkremser left17:01
suedoh the admin is me, that's why this is a mess :)17:01
the repo is on my VM17:01
→ spaceonline joined17:01
suedoh so i have full control over it17:01
i did do a lot of moving directories around17:01
so I believe this is what caused it17:01
bilel EugeneKay: I don't know, in that case I won't be able to use git stash for a normal use right? for example when there is an important thing to fix there is already a lot of changes17:02
EugeneKay bilel - you can have multiple stashes....17:02
JoshuaA Hey guys, I've been going through the process of figuring out how to work with git-svn (since at work I'm forced to use svn..) and I had a question related to svn:externals17:02
← TheHaven left17:02
suedoh charon, do you have a recommended way of cleaning up directories/files from the repository in a cached way? so that it is removed from the index17:02
JoshuaA I've found a little script that after the original git-svn clone let's you look for externals, git-svn clones them in .git_externals and symlinks them17:03
suedoh my objective was to make the repository size when doing a clone much smaaller17:03
bilel EugeneKay: oh didn't know about that, ok I will take a look at that, thanks17:03
suedoh I run out of memory on many of my VM's17:03
← mxweas_ left17:03
→ vervein joined17:03
→ fromspain joined17:03
JoshuaA The problem is that git status now won't show me changes made in externals17:03
→ ces joined17:03
JoshuaA Would anyone know a solution to this?17:03
→ tonyj joined17:04
← axl_ left17:04
→ axl_ joined17:05
← raatiniemi left17:05
← axl_ left17:05
JoshuaA So tl;dr: how to get git log showing changes in svn:externals using git-svn?17:05
→ mxweas_ joined17:05
→ axl_ joined17:06
← SamB__ left17:07
← kmap2 left17:07
→ jhardinee joined17:07
→ kmap2 joined17:07
→ Superlokkus joined17:08
← federic left17:10
← codebeaker left17:10
← dvide left17:11
← sq-one left17:11
→ jjido joined17:11
→ intransit joined17:11
← jjido left17:12
→ Chillance joined17:13
→ jjido joined17:14
← erasmospunk left17:15
← cortana2 left17:17
→ RaceCondishun joined17:19
→ sq-one joined17:20
← sq-one left17:20
← fromspain left17:20
← bilel left17:21
← mobilegamelabs left17:22
→ beneggett joined17:22
← RaceCondition left17:22
← tonyj left17:22
← robottinosino left17:22
→ Lemon|mbp joined17:23
→ tonyj joined17:23
RandalSchwartz we've just about completely ruled out using fisheye and crucible with git now17:23
now it's time to figure out some other code-review tool for 650 git repos17:23
and not gerrit. :)17:23
cbreak perl17:23
← RaceCondishun left17:24
← mxweas_ left17:24
cbreak (joking of course... :)17:24
RandalSchwartz perl's not really a code review tool17:24
oh ok17:24
← yaymukund left17:24
shtrb cbreak, I once seen a perl module that was checking C code for coding standards17:25
and If you were using some forbidden compiler options or bad varibale names it would refuse your commit17:25
← iwoj left17:26
→ SamB joined17:26
→ shreya joined17:26
← pfFredd left17:26
shtrb [noob question ] if git clone crashes is there a way I can continue from the last point it was downloading ?17:27
→ mxweas_ joined17:27
← virunga left17:28
EugeneKay !resume17:28
gitinfo Git does not yet support resuming, torrenting, or parallel fetching of clones or other network traffic. Some people have successfully created bundles and then used torrents/resuming http download/etc to accomplish what they needed (possibly creating a reliable on-demand VM somewhere to do the initial download, bundle creation, and file hosting). See https://bundler.caurea.org17:28
→ whitman joined17:29
← pureluck left17:29
EugeneKay Oooh, that's an interesting service.17:29
cbreak you can also just torrent a whole bare repo directly17:30
← timini left17:30
← ipalaus left17:30
→ marcuy_ joined17:31
← beneggett left17:31
shtrb EugeneKay, thank you17:31
→ iwoj joined17:31
→ evilgeenius joined17:31
RandalSchwartz anyone using http://phabricator.org/ for git stuff here?17:32
→ pureluck joined17:32
→ beneggett joined17:32
← patr|ck left17:32
← robbyoconnor left17:32
→ codebeaker joined17:32
→ marc_kdab joined17:33
evilgeenius How can I diff the changes between my currently checkout out code and the code that I would see if I did a "git pull" ?17:33
FauxFaux evilgeenius: git fetch && git diff '@{u}' # man git rev-parse17:34
gitinfo evilgeenius: the git-rev-parse manpage is available at http://git-scm.com/docs/git-rev-parse.html17:34
evilgeenius FauxFaux: thanks. once I've done that, how do i go ahead and merge the code?17:34
FauxFaux git pull. :)17:35
← nadavoid left17:35
RandalSchwartz or git merge @{u}17:35
→ raykyri joined17:35
RandalSchwartz since you already fetched it17:35
that would avoid any phase error too17:35
JoshuaA Is it wise to use git-svn and then use submodules for the git-svn:externals?17:35
← tvw left17:36
← harish left17:36
→ nadavoid joined17:36
← pureluck left17:37
← PigeonFriend left17:37
→ pureluck joined17:37
← jjido left17:38
→ fayimora joined17:38
→ apok_ joined17:39
← ottoshmidt left17:39
→ dvide joined17:41
← shennyg_ left17:41
← apok left17:41
apok_ → apok17:41
← pureluck left17:41
→ Dave^| joined17:42
CareBear\ JoshuaA : it sounds awkward but may work fine17:43
→ Offlein joined17:43
JoshuaA CareBear\: :D17:43
I couldn't find anything on it online.17:43
→ pureluck joined17:43
← iwoj left17:44
CareBear\ RandalSchwartz : why not gerrit?17:44
JoshuaA The solution others have proposed is importing the externals into a (untracked) .git_externals/ subdir in the repo root dir and symlinking, but that doesn't work at all the way you'd want it to17:44
← jhardinee left17:44
Offlein Because it takes a few minutes for my git commits to go live on my website, I'm thinking about putting something into my site that displays the last commit hash -- so I can be 100% sure that it's running on the latest commit. Is it dangerous for people to know that commit hash?17:44
→ shwaiil joined17:45
shwaiil hi17:45
CareBear\ JoshuaA : I don't know what git-svn does with the tree links17:45
JoshuaA Nothing. ;) That's the problem17:45
→ Targen joined17:45
shwaiil I tought I was working in a given "branch", I tryed to push without any result. I checkout to branch development and I think I lost all my work. Can I go back ? :X17:45
Tks for looking17:45
→ TheJH joined17:45
CareBear\ Offlein : not dangerous, it's a good idea. use git describe --tags, or possibly git describe --all if you don't use annotated tags17:46
→ patr|ck joined17:46
CareBear\ shwaiil : please pastebin your terminal window17:46
Offlein CareBear\ Thanks! Is this a unique idea I've had, or is it standard practice?17:46
→ kpreid joined17:47
CareBear\ shwaiil : you left out important detail in your description of events past, and the easiest way is complete dump into a pastebin17:47
shwaiil this is the log http://pastie.org/private/kr25gwv3mwtzdp5b1lglha17:47
← Lemon|mbp left17:47
← LittleIdiot left17:47
→ jargon- joined17:47
shwaiil yes CareBear\ tks for looking, was doing that :D17:47
CareBear\ Offlein : git describe is there for a reason, not unique, not standard, some do it17:47
→ dr_lepper joined17:47
→ john_anderson joined17:47
shwaiil mentions git branch new_branch_name 77ddd4fb12aff7abd940ab8ead4ce3b1a2b0cbaa17:47
CareBear\ shwaiil : so actually git tells you exactly what has happened and exactly what you should do17:48
Offlein CareBear\: OK I'm going to pretend like it's a great idea I've had, then, and brag about it to all my friends.17:48
→ i42n joined17:48
← pureluck left17:48
shwaiil but didn't a switched ? I'll try the command17:48
was just making sure I didnt messed mored17:48
mored/more17:48
CareBear\ Offlein : it *is* your idea, just so happens that git allows you to do it really easily17:48
shtrb ok what can I do if when I do git svn clone git exists each time and the source files aren't in the dir (tried to fetch only 900 revitions and git still exits with any error indication other then warning empty dir)17:48
Offlein awesome! EVEN BETTER17:49
→ pureluck joined17:49
Offlein <3 thanks very much.17:49
CareBear\ shwaiil : good instinct, but as long as you have commits, you can always get contents back17:49
→ berserkr joined17:49
CareBear\ Offlein : I mean *doing it* is your idea ;)17:49
Offlein right right.17:49
← igotnolegs left17:49
→ mauiuku joined17:50
CareBear\ shwaiil : the git branch command is really good. you're working with some commits and you should really always work with commits on actual branches17:51
question: how to find subdirectories in the history (now deleted) which contain significantly more data than others?17:51
shwaiil CareBear\: Thanks for your time ;D A really appreciated it!17:51
CareBear\ shwaiil : you're welcome17:52
→ axl__ joined17:52
← axl__ left17:52
← dr_diamond left17:52
CareBear\ I will be using filter-branch to remove those subdirectories17:52
→ michaelmartinez joined17:52
CareBear\ but I don't know exactly where they are17:52
I'd prefer not searching manually17:52
→ mikepack joined17:53
← pureluck left17:53
← i42n left17:53
← enmand left17:53
→ i42n joined17:54
← mauiuku left17:54
→ mauiuku joined17:54
→ pureluck joined17:54
← axl_ left17:54
→ axl__ joined17:54
← shwaiil left17:54
← sebrock left17:55
← mxweas_ left17:55
← marc_kdab left17:55
← i42n left17:56
← tommyvyo left17:57
← leilaM left17:57
→ i42n joined17:57
← Targen left17:57
→ axl_ joined17:57
← fgro left17:57
→ fam_ joined17:59
← axl__ left17:59
← fam_ left17:59
← aantix left17:59
→ axl__ joined17:59
← codebeaker left18:00
← FreeArtMan left18:00
→ aantix joined18:00
→ FreeArtMan joined18:00
← axl__ left18:00
← kermit left18:00
→ RaceCondition joined18:01
→ kermit joined18:01
→ tigrang joined18:01
→ axl__ joined18:01
← axl__ left18:01
← axl_ left18:02
→ ISF joined18:03
→ axl_ joined18:03
→ imachuchu joined18:05
→ totte joined18:05
→ thearthur joined18:05
totte → Guest3309618:06
← ianawilson left18:09
→ marhaban joined18:10
→ yaymukund joined18:10
→ marc_kdab joined18:11
← fayimora left18:13
← macabre left18:13
→ madsy joined18:13
← i42n left18:14
→ i42n joined18:14
→ davidfetter_vmw joined18:15
davidfetter_vmw hello18:15
gitinfo davidfetter_vmw: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.18:15
← i42n left18:15
davidfetter_vmw !format-patch18:15
!format18:16
→ i42n joined18:16
davidfetter_vmw hrm18:16
!backup18:16
gitinfo Taking a backup of a git repository is always a good idea, especially when taking advice over IRC. Usually, the best way to TACTICALLY back up a git repo is `git clone --mirror`. However, some unusual maintenance might require `tar cf repo-backup.tar repodir`. Testing in a clone is also an excellent idea. See also http://sethrobertson.github.com/GitBestPractices/#backups18:16
← i42n left18:16
← michaelmartinez left18:17
← slakmagik left18:17
→ slakmagik joined18:17
← centipedefarmer left18:17
← slakmagik left18:17
→ slakmagik joined18:17
→ mkj_wk joined18:19
→ centipedefarmer joined18:19
← whitman left18:19
→ i42n joined18:19
mkj_wk how do I ask git to tell me whether it things a file is text or binary (i.e. whether it would modify the newlines in a file if so configured)?18:20
cmullican I created a branch, and committed my working files, and it gave a bunch of create mode ... [path] and also some delete mode ... [path] output.18:20
mkj_wk I'm guessing I'm missing some plumbing command...18:20
cmullican When I try to switch back to master, it tells me "error: Untracked working tree file 'html/display_account_export.html' would be removed by merge.", only I see that file in the web UI for the repo, and it's in my directory.18:20
git status shows a LOT of changes I don't want -- deleting a ton of files for no obvious reason. I tried 'git stash' then switching, but no luck, and I also can't pull from origin master (dirty index). Help?18:20
← btx94 left18:21
← yaymukund left18:21
→ adhawkins joined18:23
CareBear\ mkj_wk : let me know if you find out18:25
← john_anderson left18:25
→ harshpb joined18:25
→ Sigma joined18:25
← Vile1 left18:25
→ alexisabril joined18:25
CareBear\ cmullican : it sounds like your working copy has been tangled up a bit18:26
→ axl__ joined18:26
CareBear\ cmullican : it can certainly be untangled, but must find out how it is tangled and work targeted. what do you want to accomplish at this point?18:27
cmullican : are you still on your new branch? run git branch to find out18:27
cmullican : if yes, please pastebin git !lol18:27
gitinfo git config --global alias.lol "log --oneline --graph --decorate"18:27
cmullican CareBear\: A working branch, with the changes I had made which were the reason I created the branch, and without damaging master18:27
CareBear\ cmullican : that we should be able to do18:28
→ privman joined18:28
→ iwoj joined18:29
cmullican CareBear\: I am; I can't change to another. Pastebin forthcoming18:29
CareBear\: unrecognized argument: --oneline and it's very long without that. Do you want the whole thing anyway?18:30
← axl_ left18:30
axl__ → axl_18:30
← Mowsh left18:30
CareBear\ cmullican : eh?18:31
← cscribble left18:31
cmullican CareBear\: https://gist.github.com/296762518:31
CareBear\ cmullican : you added the alias, ran git lol, and got "unrecognized argument: --oneline" ?!18:31
wtf18:31
cmullican I used the expansion instead of adding the alias, but yes18:31
CareBear\ what version of git is this?18:32
→ agliodbs joined18:32
cmullican 1.6.2.5 according to --version18:32
CareBear\ hm, that's very old18:32
well, go without oneline then18:33
← tigrang left18:33
cmullican OK, but it's quite long.18:33
Just a moment.18:33
CareBear\ does it look "right" ?18:33
← jn_ left18:33
CareBear\ you could add -20 or so18:33
← psakrii left18:33
← noisebleed left18:33
raek or maybe --pretty=oneline18:34
CareBear\ good one yes18:34
← p3rror left18:35
← agliodbs left18:35
→ whitman joined18:35
cmullican CareBear\: https://gist.github.com/3b49a920b8f518222414 takes it back to last December18:35
looks OK to me, but I don't entirely know what I'm looking for18:36
→ tigrang joined18:36
CareBear\ so far so good18:36
please review git show 6be2d1e to check that it is the changes you want18:36
← admsyn left18:37
cmullican It's not; it's showing a bunch of extra files, not just the one I created and the one I eedited.18:37
somehow, when I created the branch, it decided to mark a bunch of files deleted, even though they're still there.18:39
← privman left18:40
→ macabre joined18:40
→ afuentes joined18:41
CareBear\ do you recall how you created the branch?18:41
cmullican git checkout -b fleet_blends18:42
CareBear\ and how did you create the commit?18:42
cmullican git commit -a18:42
← RaceCondition left18:42
CareBear\ ok, that's why18:42
→ nadirvardar joined18:42
→ Lemon|mbp joined18:42
cmullican *blink* It is?18:43
osmosis how do I checkout a 'tag' ?18:43
CareBear\ cmullican : commit -a is really greedy, often not desirable18:43
osmosis : git checkout tag18:43
osmosis : note that this checkout will be a detached HEAD18:43
cmullican CareBear\: OK, how do I fix it, and what should I have done instead?18:44
osmosis CareBear\, "You are in 'detached HEAD' state."18:44
CareBear\ osmosis : to play it safe, only check out branches18:44
right18:44
osmosis : by safe I mean simple18:44
osmosis CareBear\, i suppose its okay for code review18:44
CareBear\ yes, perfectly fine18:44
cbreak if you commit in a detached HEAD state18:44
you'll have to work around a bit to get them onto a branch18:44
which requires knowledge18:45
mkj_wk osmosis: if you want to branch off the tag, you can specify the tag as the starting point for the branch when you create the branch18:45
osmosis ok18:45
thx18:45
CareBear\ cmullican : first what you should have done instead; git add changedfile.html newfile.html && git commit18:45
mkj_wk osmosis: if you find you want to commit when you are in detached HEAD state, you can create a branch first and then you won't have to worry about fixing up after the fact :)18:46
← alexisabril left18:46
→ StuckMojo joined18:46
→ aristidesfl joined18:46
StuckMojo i need some help with subtree merges18:46
cmullican I had done a git add and a git commit -a before I created the branch, though, so I didn't realize I needed to do them again18:46
CareBear\ cmullican : the extra files that were included in the commit, are they few enough that you can process manually?18:46
StuckMojo http://git-scm.com/book/en/Git-Tools-Subtree-Merging18:47
CareBear\ cmullican : git commit -a is usually never helpful18:47
cmullican : sometimes, but usually not18:47
StuckMojo i can't figure this part out: "You can also do the opposite — make changes in the rack subdirectory of your master branch and then merge them into your rack_branch branch later to submit them to the maintainers or push them upstream."18:47
→ RaceCondition joined18:47
cmullican CareBear\: the ones it's trying to delete? I suppose so, if I must; it's enough to be annoying but not impossible.18:47
← intransit left18:47
→ fayimora joined18:47
→ intransit joined18:48
CareBear\ cmullican : also, I'm only talking about actions after you created the branch. I assume master is all right.18:48
← pranjal710 left18:48
cmullican CareBear\: as far as I can tell, it is.18:48
→ p3rror joined18:48
CareBear\ cmullican : can you pastebin git show 6be2d1e or at least git show --stat ?18:49
→ shruggar joined18:49
CareBear\ cmullican : sorry this is taking a while, but I like to fix things up slowly so that there are no further problems18:49
← evilgeenius left18:50
→ Targen joined18:50
cmullican The former is long; it's dumping the entirety of all the to-be-deleted files. https://gist.github.com/2481b66e9d66b5afc7b1 is the latter18:50
← berserkr left18:50
← lacrymology left18:50
cmullican I'd rather have it right and take more time than a quick "fix" that fails.18:51
→ saulus joined18:51
← kmap2 left18:51
CareBear\ cmullican : all right! :) good match then. ok, so do you want to change what this commit does in html or lib or both?18:52
cmullican right now? Neither.18:52
← snearch left18:52
→ Innocentus joined18:52
→ axl__ joined18:52
CareBear\ cmullican : oh - what should change?18:52
saulus hi, how can I execute some sed command before updating refs on the remote repository? Is it possible to change the diffs with hooks/pre-receive or can I only detect whether a ref shall be invoked?18:53
cmullican I had edited one file in lib, and created one file in html. I don't even see the new html file listed, and I've copied the new file and the edited version of the existing file elsewhere for safekeeping. None of these changes are real.18:53
← axl__ left18:54
cmullican I just want to get out of this mess.18:54
CareBear\ cmullican : oh - ok! so *everything* in this commit is bogus, and you have copies of the files you would actually like to commit in another place?18:54
cmullican CareBear\: yes18:55
← axl_ left18:55
← RaceCondition left18:55
CareBear\ cmullican : good! are there any other changes in your worktree that you want to keep?18:55
→ axl_ joined18:55
cmullican Nope!18:55
CareBear\ cmullican : perfect! git reset --hard master && copy file to html/ && copy file to lib/ && git add html/file lib/file && git commit18:56
← harshpb left18:56
→ axl__ joined18:56
mkj_wk CareBear\: since you asked me to tell you if I found out: reading convert.c it looks like the only result is a possible warning about line ending conversion18:56
→ harshpb joined18:56
mkj_wk CareBear\: so I think my solution will be .gitattributes setting text/binary explicitly in all cases and using git check-attr18:57
CareBear\ mkj_wk : hm, and can it be triggered without trying to do conversion?18:57
mkj_wk : as in, some kind of dry run18:57
mkj_wk doesn't matter for me since I want to know all cases :)18:57
I also found out that the auto-guessing won't work for me anyway18:58
→ dr_win joined18:58
CareBear\ mkj_wk : also consider crlf/-crlf in order to keep the text/binary detection18:58
oh, ok18:58
why?18:58
mkj_wk because it turns out that some windows editor (another committer; I'm on Linux) has edited files with only newlines in it and added crs ONLY to the edited lines (!)18:58
and if the number of cr and nl don't match in a file, git calls it a binary18:58
CareBear\ great18:59
← beneggett left18:59
mkj_wk and git won't normalize it even if .gitattributes calls the file "text"18:59
→ tjadc joined18:59
CareBear\ mkj_wk : if you can make $windeveloper do the right thing you can clean everything up in one commit and then tell the other person to start using autocrlf.18:59
mkj_wk so I'll script ls-tree, check-attr, and tr -d18:59
← axl_ left19:00
axl__ → axl_19:00
cmullican CareBear\: woot! Seems to be working now, thanks!19:00
CareBear\ cmullican : great!19:00
cmullican : you're welcome19:00
cbreak devs should use a non-shitty editor19:00
notepad++ can do it19:00
Visual Studio can do it19:00
CareBear\ cbreak : it's a people problem, not an editor problem19:00
mkj_wk it's possible that git did it, actually19:00
cbreak git doesn't muck around with newlines unless you tell it to19:01
mkj_wk doing a recursive merge with -X ignore-space-at-eol or something like that19:01
→ beneggett joined19:01
mkj_wk checks out a few more version nodes with od -xa to compare19:02
→ HolySinner joined19:05
HolySinner any ideas how these git guis are built ?19:06
→ fourq joined19:06
HolySinner i mean what do they use for the backend git processing ?19:06
← jason237 left19:06
cmn which ones? the ones that come with git use git19:06
HolySinner no i meant other ones19:07
like github for mac19:07
cmn most call and parse git19:07
HolySinner source tree19:07
octopus19:07
RandalSchwartz where's the git wiki link on the git-scm.com website?19:07
cmn some use libgit219:07
HolySinner u mean the porcelain commands ?19:07
cmn yes19:07
HolySinner but u can't pull and push with libgit can you ?19:07
well how often does the format of porcelain commands change ?19:07
RandalSchwartz it has19:07
that's all that matters :)19:08
"git status" has changed quite a bit over teh years19:08
HolySinner i mean how reliable is the porcelain19:08
i want to make a gui19:08
so it has got to work over all the machines19:08
RandalSchwartz you should be using plumbing for that19:08
not porceilein19:09
mkj_wk that's what the plumbing was designed for, after all :)19:09
HolySinner so you mean I should develop my own19:09
→ badfish129 joined19:09
HolySinner porcelain19:09
RandalSchwartz if you need some, yes19:09
but there's a lot already there19:09
HolySinner ok19:09
that seems like a lot of job19:09
cmn yes, it's a big job19:09
kevlarman HolySinner: git status --porcelain isn't supposed to change19:10
HolySinner could you give me some references19:10
kevlarman even though it's missnamed19:10
badfish129 can someone help me with some branch stuff19:10
mkj_wk kevlarman: well, it's named right if you think of it as "git status for porcelain to read"...19:10
cmn references to what? Documentation/ in git.git has all the documentation you should need on using git19:10
→ mcluskydodallas joined19:10
HolySinner no i meant19:11
cmn if you want to use libigit2 to make your job easier, see libgit2.github.com19:11
HolySinner references to some porcelain projects19:11
does libgit allow pulling/fetching etc19:11
i have not been able to find19:11
badfish129 I forked a project on github, cloned it locally, and made all my changes on master. Now I want master to remain true to the original project, move all those changes to a new branch a, and create a new branch to make completely unrelated changes called b, how do I do this?19:11
← RobertLaptop left19:11
→ rotham joined19:12
rotham hey19:12
→ RobertLaptop joined19:12
→ maqr joined19:12
cmn what on earth is a porcelain project?19:12
rotham how do i delete files from git cache, but leave them in the folder (I added them to .gitignore)19:12
HolySinner cmn: I meant a porcelain that parses plumbing commands and returns stuff like structs19:13
adhawkins badfish129: How many commits have you made?19:13
→ martinjlowm joined19:13
HolySinner to represent the underlying objects19:13
cmn there are many, but most lack in many places19:13
← dr_lepper left19:13
→ madsy_ joined19:13
HolySinner hmmm so u think i should go with libgit ?19:14
cmn but that's not what a porcelain is19:14
HolySinner cmn: then ?19:14
cmn I'm biased, but libgit2 should be the best bet right now19:14
badfish129 adhawkins: quite a few19:14
how do I check that?19:14
adhawkins git log19:14
← madsy left19:15
→ tommyvyo joined19:15
HolySinner cmn: just one thing - does it allow stuff like cloning/pull/fetch etc19:15
i have tried to find but couldn't find any concrete proof ?19:15
cmn why "allow"?19:15
HolySinner as far as i know it does not19:15
cmn it doesn't support merging yet19:15
but it shouldn't be far off19:15
← marc_kdab left19:16
cmn cloning is being implemented right now, but with a bit of legwork you can implement it in a way that fits you19:16
and be done sooner19:16
HolySinner cmn: and if i parse the git porcelain - what are the chances it will fail on some machine due to diff versions19:16
cmn pushing is being implemented right now as well19:16
porcelain doesn't guarantee anything19:16
the output is free to change at any time and for any reason19:17
→ erasmospunk joined19:17
← i42n left19:17
← harshpb left19:18
badfish129 adhawkins: when I do a git log, I get a bunch of commits made by the project owners and not by me (I did pulls along the way)19:18
cmn pumbling is guranteed to be stable, but it's a PITA, which is why libgit2 and the wrappers exist19:18
adhawkins Ah ok, that makes things harder.19:18
kpreid rotham: git rm --cached19:18
mkj_wk badfish129: hmm, doesn't seem that hard19:19
badfish129: create a new branch from where you are19:19
and check out that branch19:19
adhawkins git log --author=....19:19
mkj_wk then delete your master branch19:19
→ SirCmpwn joined19:19
mkj_wk and then re-create it as a tracking branch of origin/master19:19
now you'll have master tracking origing/master and a different branch that you created locally with the work you've done19:19
← palm_vue left19:20
← Lgb left19:20
adhawkins mkj_wk He'll end up with the upstream's commits in both master and his created branch19:20
Will it handle that?19:20
mkj_wk didn't he want that?19:20
I thought he wanted his master to point at where origin/master is now19:20
badfish129 yeah19:20
adhawkins I assume he wants a seperate branch to that he can send a pull request or something.19:20
mkj_wk well, after doing this, could always rebase the newly-named branch in order to send a pull request19:21
assuming the usual warnings about rebase :)19:21
→ palm_vue joined19:21
→ Lgb joined19:21
adhawkins I was going to suggest he created a branch from where he originally cloned, then cherry pick his commits across.19:21
mkj_wk well, to move master it's still easy to create a new branch to hold the ref, then delete master, then recreate master as a tracking branch of origin/master19:23
← Tabmow left19:23
adhawkins Yeah19:23
mkj_wk then after that, it's possible to cherry pick commits to yet another branch off of master19:23
HolySinner cmn: Thanks so much - I think I can use libgit and call git commands for stuff like cloning and all since parsing them is not required19:23
thx man19:23
← HolySinner left19:23
mkj_wk once the branch has been picked clean, it can be deleted and gc'ed, or closed with an -s ours merge, as appropriate :)19:24
→ avpx joined19:24
← Sigma left19:25
→ dmsuperman joined19:26
dmsuperman When I make a branch of a branch19:26
→ Vile joined19:26
dmsuperman say master -> branchA -> branchA.119:26
what's the command to refresh branchA.1 with all the changes from branchA?19:26
without pushing back19:26
cmn refresh? you mean merge?19:27
→ Corey_ joined19:28
dmsuperman Preferably using rebase (I think?)19:28
→ Tabmow joined19:28
dmsuperman so it doesn't create a merge commit19:28
cmn then use that; as long as the branches haven't been published19:28
dmsuperman git rebase?19:28
cmn otherwise it's !rewriting19:28
gitinfo [!rewriting_public_history] 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)19:28
dmsuperman branchA is published19:28
branchA.1 is not19:28
it's my local modification branch19:28
avpx dmsuperman: Then rebase, sure19:29
dmsuperman thanks19:29
→ marc_kdab joined19:29
dmsuperman What if I push branchA.1 and nobody else makes any changes to it?19:30
would a rebase still be appropriate?19:30
avpx No, because git is distributed and you can't know for sure who is basing their work off of your commits19:30
dmsuperman I do know for sure though, it's for work and the branch is in my name19:30
convention dictates that people don't make changes to other people's named branches19:30
→ project2501a joined19:31
dmsuperman if nobody makes any changes will it force them to rebase ?19:31
→ nadavoid_ joined19:31
→ teweWORK joined19:31
← Corey left19:31
avpx Of people don't base their work off of the commits which you are replacing, then there is no problem19:31
The problem is when you rebase commits that other people have built / will build work on.19:32
← nadavoid left19:32
cmn as long as it's clear that nobody should be basing work on top of a branch, you can rewrite it19:32
→ dr_lepper joined19:32
← dr_lepper left19:32
→ dr_lepper joined19:32
dmsuperman Awesome19:32
thanks guys :)19:32
← dr_win left19:32
mkj_wk ok, looking into my crnl issue: git cat-file blob .. | od -xa shows me that the blog for a file (hash from git ls-tree -r master) has only NL, no CR19:32
avpx Of course19:32
mkj_wk I'm on Linux19:33
and in attempts to normalize I set core.eol=lf (should be a no-op) and core.autocrlf=input19:33
avpx mkj_wk: What is this issue?19:33
mkj_wk in .gitattributes I have text=true for the file19:34
when I check it out, I get CRNL19:34
avpx When you say NL, do you mean \n (LF)?19:34
mkj_wk right, \r\n if you prefer19:34
avpx Okay, I see19:34
mkj_wk so, it's \n only in the blob19:34
but \r\n on the filesystem19:34
avpx And is this the blob that's pointed to by the commit at HEAD?19:35
mkj_wk yup19:35
avpx And you've run git reset --hard?19:35
mkj_wk yup -- and just re-checked19:35
avpx Okay19:35
mkj_wk oh, but rm .git/index and *then* git reset --hard fixed it19:35
avpx Ahhh19:36
mkj_wk I thought I had done that since my last configuration change19:36
→ tjadc_ joined19:36
→ jhardinee joined19:36
mkj_wk (been trying to figure out how to not introduce newline changes between Linux/mac and Windows developers using both Git and EGit. It's a challenge. :)19:37
← aristidesfl left19:39
← tjadc_ left19:39
→ d3c joined19:42
← fly9 left19:42
→ ivan12348 joined19:42
→ ivan12349 joined19:42
← rotham left19:42
→ ivan12350 joined19:43
→ matsebc joined19:43
→ fly9 joined19:43
← tjadc left19:43
→ rotham joined19:43
matsebc Hello guys, newbie here, and not even sure this is the proper place to ask, I followed setup instructions for gitolite but when I try to clone the admin repo it says that it does not appear to be a valid repository19:43
any hints on that?19:44
EugeneKay !repro19:44
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.19:44
← joshteam left19:44
Corey_ → Corey19:45
← ivan12347 left19:45
← rotham left19:46
← ivan12345 left19:46
← ivan12346 left19:46
← davpoind left19:46
pdtpatrick Question - i've done git svn clone --stdlayout <remote> .. and i see all the tags with git branch -r19:46
→ rotham joined19:46
pdtpatrick how can i pull all the tags?19:46
← dmsuperman left19:46
→ john_anderson joined19:47
cmn you already have them19:47
avpx pdtpatrick: git branch -r doesn't show the tags19:47
cmn git-svn doesn't use the git ref layout19:48
→ sebrock joined19:48
pdtpatrick http://pastie.org/private/j2spr862ru2nvhzhhfyw19:48
← nadavoid_ left19:48
avpx Oh.19:48
matsebc If followed instructions on this link http://sitaramc.github.com/gitolite/install.html and when I got to the cloning point I get this https://gist.github.com/296810519:48
← gerard0 left19:48
matsebc S/if/I19:49
← Jackneill left19:49
→ mischat joined19:49
pdtpatrick Looks like this might do the trick: https://github.com/nirvdrum/svn2git19:50
CareBear\ pdtpatrick : see also http://john.albin.net/git/convert-subversion-to-git19:51
→ pfFredd joined19:53
→ apetrescu joined19:55
→ francisfish joined19:55
← codeM0nK3Y left19:55
→ blacktulip joined19:56
→ aristidesfl joined19:56
← StuckMojo left20:00
→ BHSPiMonkey joined20:01
← BHSPiMonkey left20:01
→ BHSPiMonkey joined20:01
BHSPiMonkey Hi all20:01
gitinfo BHSPiMonkey: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.20:01
→ nadavoid joined20:01
→ talntid joined20:02
← pandeiro left20:02
talntid when adding a remote path, what is it looking for?20:04
→ klj613 joined20:04
_ikke_ sitaram: Are you there?20:04
talntid I have a git repository on my website, but when I put in my website URL, it's not pulling20:04
← barmaglot_ left20:04
→ btree_ joined20:04
→ robbyoconnor joined20:05
← marc_kdab left20:05
→ cdash004 joined20:05
← marthinal left20:07
← nadavoid left20:07
← john_anderson left20:07
cmn a repo on the website? are you trying to use the dumb http protocol?20:08
← btree left20:08
← fuzzybunny69y left20:08
→ nadavoid joined20:08
talntid cmn, no, cmn20:08
← mmc2 left20:09
talntid good tip though, I didn't use ssh:// to do uit20:09
← krosenbl left20:09
← matsebc left20:09
cmn what?20:09
← JoshuaA left20:09
cmn this doesn't make any sense. what does the website have to do with ssh?20:10
talntid i was adding the remote repo using http://www.thewebsitename.com20:11
← adhawkins left20:12
← apetrescu left20:12
cmn if there isn't a repo there, there's not much git can do20:12
← resmo__ left20:13
→ apetrescu joined20:13
← madsy_ left20:13
→ alexukf joined20:15
talntid needed to be ssh://20:16
_ikke_ Somehow, my gitolite setup seems broken20:16
→ john_anderson joined20:18
avpx talntid: So it wasn't really a "website"20:18
→ marc_kdab joined20:19
← h0bbit left20:20
← sebrock left20:21
→ gerard0 joined20:21
→ AlbireoX`Laptop joined20:22
← FreeArtMan left20:23
← brentratliff left20:24
→ fsvend joined20:24
← matthiaskrgr left20:24
← needsch left20:25
→ resmo__ joined20:25
← Fullmoon left20:26
← jhardinee left20:28
→ Mowsh joined20:29
← angelsl left20:29
← ISF left20:31
→ burak575__ joined20:31
← iamjarvo left20:31
← ToBeFree left20:33
← warthog9 left20:34
← burak575_ left20:34
shtrb is there any thing I could do to go around :r60900 = 013234d913dd40224c60b266b651c1b728207791 (refs/remotes/git-svn)20:34
error: git-svn died of signal 1320:34
2 hours of clone20:35
cmn signal 13 is EPIPE, no?20:36
shtrb 13 is SIGSEGV afaik20:36
→ jhardinee joined20:37
avpx According to the man page on signal(7), 13 is SIGPIPE20:37
shtrb ok what can I do ?20:37
CareBear\ shtrb : move closer to the svn server and start over20:38
(network wise)20:38
cmn pretty much, yeah20:38
that means that the connection was closed20:38
IIRC git svn fetch should continue20:38
shtrb CareBear\ : it happen even one a machine connected directed to server20:38
→ joshteam joined20:39
CareBear\ try it on the server too20:39
cmn but if you're going to move all to git, importing locally would be a lot better20:39
CareBear\ cmn++20:39
cmn or use svn-dump-fast-import (-export?) which should be at least as fast, maybe faster20:40
→ btree__ joined20:40
→ eletuchy_ joined20:41
← eletuchy left20:41
→ fourlastor joined20:41
shtrb moment I can take exported svn and then some how use git-svn to use against original rep ?20:42
→ btree joined20:42
shtrb i.e. I have all the sources files allready (did svn checkout in a diffrent directory and all went ok)20:42
← dr_lepper left20:42
cmn a svn checkout is completely different from a git-svn clone20:43
shtrb :-(20:43
talntid having a hell of a time getting a git repo working20:43
on my remote server via ssh, i did a "git init"20:43
avpx You should do git init --bare20:44
← btree_ left20:44
talntid on my local machine, i have a folder, which is setup as a git repo, and when I push to it, it tells me "not all refs pushed"20:44
← john_anderson left20:44
shtrb cmn , is there a way I could take exported files and push it into git svn clone dir ?20:44
→ vdv_ joined20:44
shtrb I don't need the history20:44
← vdv_ left20:44
cmn what are "exported files"?20:44
← vdv left20:44
avpx talntid: We'll need to know more specifically what commands you are using20:44
cmn if you need the history, why are you cloning?20:44
just do git init somewhere20:45
← btree__ left20:45
→ vdv joined20:45
shtrb cmn, I don't need the history (can live without it) I need to be able to do git svn commit git dcommit20:45
→ berserkr joined20:45
cmn then limit what you're cloning20:45
shtrb cmn, to some svn rep20:45
← hoschi left20:45
← savid left20:45
shtrb I did for 90 revitions and even for one20:45
all fail20:45
cmn fail how?20:46
← iwoj left20:46
cmn what did you actually do?20:46
→ john_anderson joined20:46
← tigrang left20:46
shtrb git svn clone -r60900:HEAD https://ip/svn/name/trunk (one revition back) and I get rXXX: hash and git20:47
exits20:47
← andrewhl left20:47
shtrb in a diffrent event I got the signal 1320:47
→ matthiaskrgr joined20:48
← d3c left20:48
shtrb the clone is for a sevral GBs20:48
← mkj_wk left20:48
← edude03 left20:49
→ eurbach joined20:50
← eurbach left20:50
→ edude03 joined20:50
← lotus left20:50
→ adamm joined20:53
→ savid joined20:54
cmn different event?20:55
→ oops joined20:55
cmn does git say anything on that rev? did trunk get modified in that rev?20:55
oops I want to set upstream without using the git push -u command. How can I do it using git-config?20:56
→ trying2code joined20:56
← edude03 left20:56
cmn why do you want to use config? it's annoying, man git branch --set-upstream20:56
gitinfo the git-branch manpage is available at http://git-scm.com/docs/git-branch.html20:56
shtrb cmn, first event run on time second event time + 2 hours (no commits in between)20:56
← mishok13 left20:57
oops cmn, thanks20:57
→ toabctl joined20:57
← Araxia_ left20:58
shtrb cmn , If I would do a local svn rep after doing an svn export and do git svn clone will be able set git to commit to the original source ?20:58
cmn probably20:58
shtrb cmn, any ideas how ?20:59
cmn it depends on how the GUIDs change20:59
shtrb have no idea20:59
cmn me neither20:59
but the problem is grabbing data from the repo, not importing the repo20:59
← guampa left20:59
cmn you still haven't answered my questions about what git did21:00
shtrb I mean I have now all the files in a folder the files are on the exact version as in the rep so if I could only tell git use them instead of downloading them21:00
badfish129 I have a forked repo cloned locally, I've made changes on master, and two branches a and b. How do I make master reflect the original repository (that I forked from) while keeping my changes on a and b?21:00
cmn if you can recreate the git-svn metadata by hand21:00
← imrobert left21:01
shtrb is searching how he could recreate the metadata21:01
shtrb cmn, I think it could actually work21:01
→ imrobert joined21:01
→ edude03 joined21:01
← imrobert left21:01
← eicto left21:01
→ imrobert joined21:01
← imrobert left21:01
shtrb cmn, for preious attempts git was getting into some file writing revition number : hash and then just exiting21:02
→ needsch joined21:02
shtrb no error code , no messages just exit21:02
→ nadavoid_ joined21:03
← comps left21:03
cmn and the question that might shed some light? the last one in that sentence?21:04
← tommyvyo left21:04
← nadavoid left21:05
badfish129 anyone?21:05
← Tinti left21:05
imachuchu badfish129: essentually it is "git checkout master" then "git reset --hard origin/master"21:06
badfish129: but read up on reset first and I'm not sure the "origin/master" part is right as it depends on what you called your remote (or if you cloned) and what the remote's branch that you are tracking is called21:07
badfish129: lastly that won't "make master reflect the original repository" but "make master reflect the original branch" which I think is what you mean21:08
badfish129 yup21:09
so if I were to an add a remote to the original repo21:09
and then do a reset hard with that remote21:09
it should work?21:09
← fission6 left21:09
← alberto56 left21:12
← intransit left21:12
→ venom00 joined21:14
→ eicto joined21:14
venom00 hello, I've had to rewrite part of the history of a git repo (wrong commiter e-mail address)21:14
→ Goplat joined21:14
imachuchu badfish129: yes, but you most likely just want one branch for each remote so the reset may not be needed/what you actually want21:15
venom00 then I `git push --force`d, now the other clones of the repo have two entries for the rewritten commits21:15
is there a way to repair the history/log on the already existsing clones?21:15
← shreya left21:16
imachuchu venom00: first off !rewrite secondly when you say "two entries" you mean the before and after right?21:17
gitinfo venom00: [!rewriting_public_history] 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)21:17
← toabctl left21:18
badfish129 imachuchu: things are not working21:19
I did a git add remote base https://therepothatIforkedfrom.git21:20
and then I did git reset --hard base/master21:20
→ madsy joined21:20
← madsy left21:20
→ madsy joined21:20
← egarcia left21:20
← klj613 left21:20
badfish129 but then it gives me git checkout -b newdesign origin/newdesign21:20
whoops21:20
← john_anderson left21:21
badfish129 it gives me fatal: ambiguous argument 'base/master': unknown revision or path not in working tree21:21
← davidfetter_vmw left21:21
avpx Did you run git fetch?21:21
→ john_anderson joined21:21
← eletuchy_ left21:22
← nasso_ left21:22
→ eletuchy joined21:22
→ nasso joined21:23
→ thiago_ joined21:23
← thiago left21:23
← bgerber left21:24
badfish129 avpx: ah thats what i was missing, thanks21:24
infogulch → infogulch|afk21:24
avpx badfish129: Yeah, git remote add just sets up the URL. git fetch is what gets you the refs / objects from the remote21:24
← trying2code left21:25
→ rcu joined21:25
→ warthog9 joined21:25
→ andrewhl joined21:26
← shtrb left21:26
← avandenhoven left21:26
← hellopat left21:26
→ kcj joined21:27
← wanders left21:27
→ marcogmonteiro joined21:27
← d0k left21:27
← aristidesfl left21:27
← irqq left21:27
← mwrosen left21:28
→ wbednarski joined21:28
→ bgerber joined21:28
gitinfo set mode: +v21:28
→ aristidesfl joined21:29
← andrewhl left21:30
→ shtrb joined21:30
venom00 imachuchu, yep, before and after21:31
→ pantsman joined21:32
← pantsman left21:32
→ pantsman joined21:32
venom00 imachuchu, I'm aware of the problems of rewriting, but what are the alternatives in my case?21:32
← martinjlowm left21:32
→ wanders joined21:33
← mpenz left21:33
→ alberto56 joined21:34
tuxx_ → Guest6810121:34
venom00 it'd be nice to have a list of committers/authors identified by some ID so one can change their info easily without rewriting, but I don't know if it's doable21:34
cmn mailmap21:34
FauxFaux For what tool? In git log? Just sed it.21:34
← whitman left21:34
→ pfFred joined21:35
shtrb cnm , I think I found a solution - after a failed (signal 13) git svn clone doing git fetch from a sub dir and it continued :D21:36
why didn't I think about that few hours ago :-(21:36
imachuchu venom00: it's a catch-22. Git won't change history behind someones back (and shouldn't!) yet you really want to change history. Your options (as far as I know are): 1. Accept the problem and ignore. 2. Revert the changes and recommit (most likely very messy). 3. Fixup the history, then have all other developers rebase/reset their branches to use the new history21:37
← nadavoid_ left21:37
cmn shtrb: I told you about that earlier21:37
imachuchu venom00: the difference between 2 and 3 is that 3 changes history (so all developers must acknowledge and shift their stuff around) while 2 just moves on from where we already are21:38
shtrb cmn, I missed that I guess (to try to do a fetch in the cloned dir)21:38
gitinfo set mode: +v21:38
thiago_ → thiago21:38
← pfFredd left21:38
→ nadavoid joined21:39
venom00 imachuchu, in truth I'm the only developer, but now others are going to join21:39
luckily21:39
however I'm afraid this will happen again :P21:39
→ diegoviola joined21:39
→ madewokherd joined21:40
cmn venom00: man git config mailmap.file21:40
gitinfo venom00: the git-config manpage is available at http://git-scm.com/docs/git-config.html21:40
venom00 I think I'll opt for third solution21:40
← Innocentus left21:40
← Superlokkus left21:41
venom00 cmn, thanks I'll investigate mailmap, whatever it is21:41
← etcetera left21:42
cmn it's what you were asking for21:42
a waya to map one identity to another21:42
→ etcetera joined21:44
← ferdna left21:44
← p3rror left21:44
← edude03 left21:45
→ p3rror joined21:46
venom00 cmn, looks like a patchwork but could solve my problem21:46
thanks a lot21:46
← alexukf left21:46
cmn there is no other way to have such a mapping21:46
→ HyperNerdV2 joined21:47
→ edude03 joined21:47
→ endzYme joined21:48
endzYme So I went to push a new commit to my remote server via ssh and I am getting a fatal: object corrupt21:48
SJr How do I get my git repo to always store files as a certain group, and with a certain directory and file permission.21:48
endzYme FS and OS are fine, how can I resolve this?21:48
FauxFaux SJr: !deploy21:49
gitinfo SJr: Git is not a deployment tool, but you can build one around it for simple environments. Here is an example hook to get you started: http://goo.gl/l68hq21:49
SJr ?21:49
← robbyoconnor left21:49
SJr how is this a deployment issue?21:49
imachuchu endzYme: git fsck21:49
SJr One problem we keep having is every time we check files in, the person needs to go to the repo, and then do chmod -R 775 *, and chown -R *21:49
FauxFaux SJr: The fact that multiple people have access implies that either you're using a shared filesytem (VERY BAD) or that you're editing on a deployed copy of the application (EVEN WORSE)?21:50
→ deltree_ joined21:50
endzYme imachuchu: just shows two "dangling tag"s21:51
SJr shared file system :)21:52
← AaronMT left21:52
FauxFaux Yeah, don't do that.21:52
cmn SJr: see man git init --shared if you must21:52
gitinfo SJr: the git-init manpage is available at http://git-scm.com/docs/git-init.html21:52
→ comps joined21:52
SJr Yeah I think it is initialized as --shared21:52
FauxFaux cmn: Is that even going to help with multiple users /committing/ inside the same repo?21:53
imachuchu sorry but I have to go, can someone else help endzYme21:53
← imachuchu left21:53
→ jason237 joined21:53
cmn FauxFaux: not against that, but I assumed that was way too silly to be what's hapenning21:53
FauxFaux I'm suddenly too scared to switch jobs again, there's too much stupid in the world.21:54
SJr Because of this?21:54
→ raatiniemi joined21:54
→ psoo joined21:54
FauxFaux Well, the idea that shared filesystems could seem more attractive than provisioning a linux box to run a git server.21:55
cmn SJr: that depends on what you're actually doing21:55
endzYme anyone have a good link of how to determine why a repo has been corrupted and how to resolve?21:55
FauxFaux There's various writups under titles like "dealing with repo corruption".21:56
← airborn left21:56
endzYme FauxFaux: as in google it?21:56
SJr Basically on a Shared File System we have a --bare --shared repo, that we clone and use as an origin/master21:57
to our own local working copies.21:57
So when we push to upstream we have problems with permissions.21:57
← john2496 left21:57
← fly9 left21:58
cmn are the people in the same group? do all OSs support groups sanely?21:59
SJr Yes, but some have different primary groups22:00
I believe it's all Linux machines22:00
Is this just a umask problem, and unrelated to git?22:00
→ ISF joined22:00
→ dfrey joined22:00
→ imrobert joined22:00
→ matsebc joined22:00
matsebc is there a gitolite support irc channel?22:00
← afuentes left22:00
← Mowsh left22:01
cmn doing all those chmod probably doesn't help22:01
→ alanp joined22:01
← alanp left22:01
→ alanp joined22:01
← Maccer left22:01
← dv310p3r left22:01
← imrobert left22:02
→ imrobert joined22:02
← alanp_ left22:02
endzYme matsebc: yes #gitolite22:02
← Destos left22:02
matsebc ops sorry, thanks22:02
← OOPMan left22:03
→ Tinti joined22:03
← john_anderson left22:03
← imrobert left22:03
→ imrobert joined22:03
→ guampa joined22:03
← imrobert left22:03
→ OOPMan joined22:03
→ imrobert joined22:03
→ john_anderson joined22:04
→ bicranial joined22:05
→ Fidelix joined22:05
← imrobert left22:05
← glennpratt left22:05
→ imrobert joined22:05
Fidelix What is a good git workflow with automatic "deploy" for stage/live?22:05
I tried using gitlab, but it does not provide that.22:05
Only clumsy "web-hooks"22:06
That are worthless for the average users22:06
← imrobert left22:06
→ imrobert joined22:06
← ces left22:06
← francisfish left22:06
cmn generally to push to whatever branch is live and trigger a deployment22:07
Fidelix cmn, using what, standard hooks?22:08
cmn and to make sure that branch only gets updated with tested code22:08
using whatever fits your needs best22:08
maybe normal git hooks, or you can signal via one of those web hooks22:08
SJr cmn I think it's probably just setting the umask to be 0007, and then setting it g+s22:08
cmn deployment is not something that git does; so you need to identify how you want to deploy and make sure pushing to that branch triggers it22:09
Fidelix cmn, this sounds too much hassle for something so simple. I currently do standard hooks, with no checks for staging, and to live, only manually.22:09
→ fuzzybunny69y joined22:09
cmn gitlab uses gitolite underneath, so it should be possible to use "normal" hooks22:09
Fidelix cmn, I know that. I was hoping one of you could recommend a tool that could help with that22:09
← dzonder left22:09
cmn simple? deployment isn't that simple22:09
← fayimora left22:09
← maletor left22:10
cmn there are many deployment tools, your framework or environment would probably make you choose one over another22:10
Fidelix cmn, gitlab overwrites your hooks22:10
cmn this is not something that git does22:10
Fidelix It already implements a post-receive hook, and if you edit it, it will overwrite it22:10
cmn, OK. Do you have some tool to recommend?22:10
cmn gitolite has secondary hooks22:10
there's on in !deploy22:10
gitinfo Git is not a deployment tool, but you can build one around it for simple environments. Here is an example hook to get you started: http://goo.gl/l68hq22:10
← juanfer left22:11
cmn capistrano is popular among ruby people22:11
Fidelix Now that's a start. Capistrano. Thanks. Anything else?22:11
← imrobert left22:11
→ imrobert joined22:11
cmn no idea, I either outsource my deployment or let jekyll do a crappy one22:11
→ maletor joined22:12
cmn and ignore the race condition22:12
← nadavoid left22:12
← ISF left22:12
← TheJH left22:13
→ ro0 joined22:13
← raatiniemi left22:13
← AeroNotix left22:17
→ ferdna joined22:18
← marhaban left22:19
SirCmpwn → behind22:21
behind → SirCmpwn22:21
→ airborn joined22:21
→ fayimora joined22:21
← duckxx left22:22
→ duckxx joined22:23
→ cybersphinx joined22:23
→ imran-uk2 joined22:24
← jhegge left22:24
← davidjdagino left22:26
← the_cybersphinx left22:26
← tr-808 left22:27
← kermit left22:27
→ tr-808 joined22:27
→ tr-808_ joined22:28
← eicto left22:28
→ nadavoid joined22:28
→ kermit joined22:28
← arietis left22:29
← burak575__ left22:29
→ Diego_g joined22:29
← duckxx left22:29
← Einzig_ left22:29
← apetrescu left22:30
← fsvend left22:30
← Tabmow left22:31
← Tommy[D] left22:32
← eijk left22:32
→ apetrescu joined22:32
→ Tommy[D] joined22:33
← tr-808 left22:33
→ tr-808 joined22:33
→ bulletmark joined22:34
→ Corey_ joined22:36
→ Tabmow joined22:36
← airborn left22:37
← venom00 left22:37
← wanders left22:38
→ sshres15 joined22:39
← Corey left22:40
→ lemonade` joined22:41
← bulletmark left22:41
← ehsan left22:41
→ eicto joined22:41
→ adityamenon joined22:42
→ alexukf joined22:42
→ AlexMax joined22:42
→ ehsan joined22:42
← ehsan left22:43
← shruggar left22:43
← AdrienBrault left22:43
→ wanders joined22:43
← marcuy_ left22:44
AlexMax I have a repository on github that is actually a git-svn repository. I notice that when i git clone it, none of the SVN metadata is set up. Is it safe to git svn fetch on this repository?22:44
← mikepack left22:44
← endzYme left22:44
← aristidesfl left22:45
Corey_ → Corey22:45
kevlarman AlexMax: you'll need to set up the svn metadata manually22:45
AlexMax But is it safe to do so?22:45
kevlarman AlexMax: the git-svn man page has an example of how to do that22:45
← jajmon left22:45
→ aristidesfl joined22:46
kevlarman AlexMax: it's the third one in the examples section22:46
← nadavoid left22:46
← eicto left22:46
← HyperNerdV2 left22:46
→ zinjacoder joined22:47
→ Maccer joined22:48
→ eicto joined22:48
← macabre left22:48
← revagomes left22:49
→ pranjal710 joined22:51
← pranjal710 left22:51
→ Fandekasp joined22:51
→ archdori joined22:52
← alanp left22:53
← fourlastor left22:56
→ fayimora_ joined22:57
→ marcuy_ joined22:58
← AlbireoX`Laptop left22:58
← psoo left22:59
← zinjacoder left22:59
← mtytel left22:59
← fayimora left23:00
fayimora_ → fayimora23:00
→ w09x joined23:01
← Dave^| left23:02
→ bulletmark joined23:02
← marcogmonteiro left23:02
← erasmospunk left23:02
← beatak left23:02
→ marcogmonteiro joined23:03
← Vampire0_ left23:04
← unreal left23:05
diegoviola http://russellwarner.deviantart.com/art/F-you-30906451023:06
← tr-808 left23:06
→ dv310p3r joined23:07
→ kcm1700_ joined23:08
→ cilly joined23:08
→ hwrd|wor_ joined23:09
→ unreal joined23:09
← hwrd|work left23:09
→ HyperNerdV2 joined23:10
→ ISF joined23:10
← kcm1700 left23:11
← dfr|mac left23:17
← cmullican left23:17
← bulletmark left23:18
← shtrb left23:19
← imran-uk2 left23:19
→ egarcia joined23:20
→ intransit joined23:23
← cagedwisdom left23:23
→ avinashhm joined23:23
← fayimora left23:24
← wbednarski left23:25
← blacktulip left23:25
→ delinquentme joined23:25
delinquentme hey all I just forked a project and made some code edits23:25
now im getting a 403 when I try to push to my forked branch23:25
badfish129 is your remote set to the original project or your fork?23:26
delinquentme in my config file it shows my fork as the origin23:26
url = https://github.com/delinquentme/bioruby.git23:26
← apetrescu left23:26
badfish129 wrong username/password?23:27
i was just getting a 403 with that23:27
→ ehsan joined23:27
← bgerber left23:27
← rcu left23:27
→ fayimora joined23:27
→ apetrescu joined23:28
→ cagedwisdom joined23:29
→ bgerber joined23:29
delinquentme hmmm badfish129 how do I check that23:29
whats the normal login command anyways23:29
← tyatpi left23:30
← pantsman left23:30
badfish129 I would just follow the github instructions and regenerate your ssh key23:30
→ glennpratt joined23:31
← oogatta left23:31
suedoh Hi, could someone recommend a method to remove all tracking history for a certain directory which doesn't exist on the local filesystem anymore and has been git rm'ed as well? When I try to do a checkout on a virtual machine, it ends up crashing because the repo is too large. So I'm trying to shrink it down.23:31
cbreak suedoh: that's unlikely23:32
← PrinceAMD- left23:32
cbreak but if you want to remove history, you have to rewrite history23:32
suedoh: man git filter-branch23:32
gitinfo suedoh: the git-filter-branch manpage is available at http://git-scm.com/docs/git-filter-branch.html23:32
cbreak look at index filter23:32
suedoh Oh, that sounds scary23:32
Rewrite history23:32
heh23:32
thiago it is scary23:32
← TeckniX left23:32
cbreak not really23:32
thiago everyone would need to re-clone23:32
(to put it in simple terms)23:32
cbreak you'll just end up on a separate timeline23:32
suedoh I'm not very worried about everyone, because only two people exist in this project23:32
cbreak you do know time travel movies, right? :)23:33
← ehsan left23:33
avpx I'm surprised nobody has used !rewrite yet23:33
gitinfo [!rewriting_public_history] 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)23:33
suedoh Yea of course! The Doc. is right here23:33
thiago only in a few sci-fi movies & shows do they stick to the new timeline23:33
suedoh (From Back to the Future) :)23:33
thiago in most, they go back to the original later23:33
cbreak if you change anything in the past, be it a single byte, everything that follows will change as well :)23:33
thiago: or one that looks very similar23:34
← delinquentme left23:34
thiago in Eureka they stayed in the new timeline23:34
I don't like the new Star Trek timeline23:34
← intransit left23:34
cbreak in back to the future repeatedly :)23:34
avpx filter-branch is a good way to accomplish this, though. Basically it's a command that goes over a sequence of commits and allows you to make some modification to them.23:34
→ oogatta joined23:34
cbreak but really, if your VM crashes on clone23:35
avpx So for instance, if you want to go over HEAD and its ancestors and modify commits so that you remove a file called "foo", you can run: git filter-branch --index-filter 'rm --cached --ignore-unmatch foo' HEAD23:35
cbreak then you likely have too big files23:35
not too big repo23:35
suedoh Rewrite b145bc6315e7aeba9b826f21e827ed635c0ee6e4 (5/5)error: duplicate parent 664901326e4a57065e4def99fb2bdfc9f3c529b1 ignored23:35
What does this mean?23:35
My command was23:35
$ git filter-branch --subdirectory-filter new26 -- --all23:35
I want to remove the directory new26/ and all objects history below it23:36
cbreak well23:36
→ xnm joined23:36
cbreak your command removes everything BUT new2623:36
suedoh I'm worry that this duplicate parent is something that's alerting me my repository is screwed23:37
I just now did a brand new clone, and that's the one I'm working with23:37
So to me that sounds scary23:37
avpx Did you make a !backup before proceeding?23:37
gitinfo Taking a backup of a git repository is always a good idea, especially when taking advice over IRC. Usually, the best way to TACTICALLY back up a git repo is `git clone --mirror`. However, some unusual maintenance might require `tar cf repo-backup.tar repodir`. Testing in a clone is also an excellent idea. See also http://sethrobertson.github.com/GitBestPractices/#backups23:37
cbreak would have to look at history graph23:37
but going to bed, so don't have time23:37
suedoh avpx, Yea I made backups23:37
cbreak but what you want is probably index filter, not subdirectory filter23:37
do something like git rm --cached as written above23:38
suedoh cbreak, Thanks much, appreciate your support. Have good rest23:38
cbreak, I would do rm --cached but these directories have already been hit w/ that command23:38
→ _huoxito joined23:38
← rotham left23:38
cbreak not in the past23:38
suedoh I'm trying to remove all history, because my checkout crashes on virtual machines with smaller memory23:38
fatal: Out of memory, malloc failed (tried to allocate 734276974 bytes)23:38
fatal: index-pack failed23:38
→ tigrang joined23:38
suedoh that is what I'm trying to resolve :)23:39
cbreak filter-branch will execute that command in the past, in every commit23:39
→ rotham joined23:39
cbreak suedoh: that happens if you have big files23:39
suedoh Well, the problem with filter-branch is that I keep receiving this:23:39
cbreak not a big repo23:39
suedoh Yea, well the large files are in that directory23:39
cbreak -> filter branch23:39
suedoh error: failed to push some refs to23:39
← etcetera left23:39
suedoh To prevent you from losing history, non-fast-forward updates were rejected23:39
→ mikepack joined23:39
cbreak that's normal23:39
as we said: history rewriting23:39
suedoh Funny part is, that in my config on remote server, fast forward is enabled23:40
cbreak the result is a different timeline23:40
incompatible23:40
avpx suedoh: Yeah, that's not really funny23:40
It's totally expected23:40
suedoh Okay23:40
cbreak you have to force push or just make a new repo23:40
suedoh So how do I get this pushed in?23:40
I've tried git push --force, that doesn't work23:40
cbreak it does23:40
avpx suedoh: First we should explain what you are doing23:40
cbreak maybe your server disallows it23:40
suedoh ! [remote rejected] master -> master (non-fast-forward)23:40
error: failed to push some refs to23:40
avpx Ding ding ding!23:40
We have a winner23:40
cbreak you made your server disallow it23:41
avpx suedoh: The server will *only* accept fast-forward commits23:41
Do you know what that means?23:41
suedoh cbreak, I have:23:41
[receive]23:41
denyNonFastforwards = false23:41
cbreak just make a new repository, that's easiest23:41
suedoh cbreak, would it be safe to just go on my repo server rename the directory, and re-initialize this project23:41
← williamcotton left23:41
cbreak sure.23:41
suedoh or should I entirely change the name of the project?23:41
→ etcetera joined23:41
avpx suedoh: It's up to you, really23:42
cbreak either way would work23:42
suedoh okay23:42
thanks guys23:42
cbreak make sure to git init --bare23:42
← needsch left23:42
suedoh I wish I knew what I did23:42
cbreak and don't throw your old repo away23:42
suedoh to get here23:42
cbreak you probably will need it eventually :)23:42
avpx suedoh: So yeah, basically you can't just git push a rebased commit because it's not a descendent of the remote branch23:43
← bgerber left23:43
← thiago left23:43
cmn .version23:45
→ bgerber joined23:45
cmn where does the bot look for the version?23:45
→ phpD3vLuhv joined23:46
→ hellopat joined23:46
← pfFred left23:46
← savid left23:49
← phpD3vLuhv left23:49
← yoklov left23:50
← notmypudding left23:52
← p3rror left23:52
← warthog9 left23:53
→ ology joined23:53
→ tyatpi joined23:55
← SirCmpwn left23:56
→ acl_ joined23:58
← scientes left23:58
← Element9 left23:58
← wilmoore left23:59

Logs Search ←Prev date Next date→ Channels Documentation