IRCloggy #git 2012-07-25

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.

2012-07-25

→ kaen joined00:00
← kaen left00:00
→ towski joined00:02
← towski left00:03
← neurodrone left00:03
← Pupeno_W left00:04
← dvide left00:04
→ towski joined00:06
← btanaka left00:06
← thiago left00:06
← dreiss left00:08
→ gavin_huang joined00:08
→ bwwww joined00:09
→ iamjarvo joined00:09
← turtle_ left00:09
← bwwww left00:10
→ ehsan joined00:10
← lresende left00:11
← apok left00:13
← reuf left00:13
→ apok joined00:14
→ clifton joined00:15
← marhaban left00:15
→ Textmode joined00:16
clifton is there a way to stage a specific line of a file without going into interactive mode?00:17
← igotnolegs left00:19
cmn no, that's the point of interactive mode00:20
you can use git-gui, but that's still interactive00:20
frogonwheels clifton: also you can use something like fugitive00:20
clifton: (vim plugin)00:20
clifton: or git gui00:20
err yeah, cmn said that.00:20
clifton i have a pre-commit hook written that removes trailing whitespace, tabs -> spaces, and blank lines before eof00:21
→ btree joined00:21
cmn and why does that need this?00:21
clifton i just wanted to stage changes as part of the hook00:22
cmn stage the whole file00:22
→ brokenshakles joined00:22
cmn the changed lines will be updated00:22
→ duckxx joined00:23
frogonwheels clifton: and you're going to create a new commit for this? or are you going through and rewriting all the commits?00:23
clifton well, its very common we use interactive mode to add certain parts00:23
to begin with00:23
cmn but here you want to add the whole file00:23
clifton no, just the lines that are affected00:23
i want to keep our history and only change lines that were included in the commit00:24
many of the people on my team use TextMate or ST2 and I see trailing whitespace committed often00:24
cmn no, you want to stage the whole file00:24
frogonwheels ahh crap. pre-commit. just ignore me.00:24
cmn unchanged lines won't be changed00:24
frogonwheels cmn: I see what he means. He's talking about changing the file in the wd.00:25
cmn that won't work, though00:25
frogonwheels clifton: possibly a better way would be to change the file that's in the index - ie grab it out, process it, and put it back.00:25
cmn the only sensible thing is for the pre-commit hook to change the staged file and re-stage that00:25
anything else is going to be so full of edge cases it's not worth contemplating00:26
clifton hm00:27
← towski left00:27
clifton https://gist.github.com/317361100:27
thats what i have00:27
its so we can slowly but surely remove trailing whitespace without mucking up our history00:29
→ shawngo joined00:29
→ towski joined00:29
cmn this still has edge cases00:30
clifton any off the top of your head?00:31
cmn the file in the workdir doesn't have to be the same one that's being comitted00:31
frogonwheels clifton: looking at fugitive, it uses cat-file to read the file from the staging area, hash-object to create the object from stdin after processing it (from the vim buffer in this case) and update-index to update the file with a new blob00:31
← ipalaus left00:31
cmn so you're changing something that the user may or may not want to commit00:31
clifton ah frogonwheels: that's probably the best way to do it00:31
← btree left00:31
clifton cmn: so maybe its best to leave those changes unstaged00:32
→ w09x joined00:32
← Berglund left00:32
frogonwheels clifton: yeah - keep well away from the working-dir file00:32
cmn right, modify the file that's going to be committed00:32
← andrewhl left00:33
clifton k, any pointers on how to do that, im not very familiar with git objects00:33
frogonwheels clifton: I'm sure there's other code you could look at, but s:BufReadIndexFile and s:BufWriteIndexFile in fugitive.vim _might_ help00:33
clifton my OCD was just triggering all afternoon after seeing coworker commits00:33
cmn slap them00:33
you can't force them to use the pre-commit hook anyway00:33
→ redrocket joined00:34
clifton well, technically i could00:34
frogonwheels clifton: I put the whitespace correcting code in a Vim trigger :)00:34
cmn clifton: no00:34
→ alberto56 joined00:34
clifton they could always skip the validation if they needed to00:34
cmn right, that's not forcing anything00:34
reply to the commits mentioning the broken formatting00:34
whitespace damange is part of the review00:35
clifton yes but they won't skip it 99% of the time, they just need something they can without thinking about it much00:35
← chrisgeorge left00:35
clifton can use*00:35
← towski left00:35
clifton yes that's a very good point but id like that to be automated for the most part00:36
← htroyack left00:36
clifton thanks for the help so far guys and ill check out the fugitive source in my .vim dir00:37
← mtytel left00:37
← CannedCo_ left00:37
clifton https://gist.github.com/3171463 is a post-commit example00:38
but it didnt work for me00:39
frogonwheels clifton: git rev-parse :0:<filename> will give you the sha1 for the staged object. git cat-file blob <sha1> will give you the contents. git hash-object -w --stdin will write a hash object from stdin and give you the hash, git ls-files --stage <path> will give you the old 'index' entry, and git update-index --index-info will update the new entry (if you update the blob sh1 in the ls-files with the new blob)00:39
clifton: that should about cover it?00:40
clifton yep that looks fantastic00:40
awesome, thank you!00:40
frogonwheels clifton: np. I've delved into that code b4 trying to debug it for windoze, so I knew the rough idea00:40
clifton org00:41
mt00:41
ouch, windows00:41
← wilmoore left00:41
frogonwheels clifton: yeah. vim & git make it barable though.00:41
clifton: all my other computers/devices are linux based ;)00:42
← SkAzZ left00:42
→ nutron joined00:43
clifton i just use macs nowadays, except when im doing sys administration here-and-there00:43
→ SkAzZ joined00:43
← maxwellcorbin left00:43
clifton one of the dudes on my team worked at canonical for a few years and uses a mac now too :O00:43
→ igotnolegs joined00:43
→ maxwellcorbin joined00:44
clifton used linux exclusively through CS undergrad though00:44
→ towski joined00:44
→ defts joined00:45
← Guest31884 left00:46
← defts left00:46
← iwoj left00:46
clifton alright ill have this hacked up very fast00:47
→ lresende joined00:47
clifton thanks again for the help00:47
← SkAzZ left00:48
← jrnieder left00:53
→ enmand joined00:53
← towski left00:54
→ Facetious joined00:54
← lteo left00:55
→ dreiss joined00:56
→ towski joined00:56
← mcluskydodallas left00:58
→ xiangfu joined01:00
→ cooloney joined01:02
← lresende left01:02
→ thursday joined01:03
→ dc5ala joined01:03
→ weiyang joined01:03
→ jsoa joined01:04
→ tiglog joined01:05
← ivan\ left01:05
← harish left01:07
→ ivan\ joined01:07
← TheJH left01:07
← towski left01:08
← imrobert left01:08
← mocramis left01:10
→ mocramis joined01:10
← dermoth|home left01:10
← weiyang left01:11
← duckxx left01:11
← notfunk_ left01:11
← thursday left01:13
← Ryback_ left01:14
← Vampire0 left01:15
→ mcluskydodallas joined01:15
← linguini left01:15
→ duckxx joined01:17
→ w09x_ joined01:18
← mcluskydodallas left01:18
← w09x left01:19
← palm_vue left01:20
← bizhanMona left01:20
→ palm_vue joined01:21
→ bizhanMona joined01:21
→ imrobert joined01:21
← Maccer left01:23
← hackez left01:23
← segher left01:26
← alberto56 left01:26
← maletor left01:26
← artnez left01:27
→ weiyang joined01:27
→ rchavik joined01:28
← lonewulf` left01:28
→ segher joined01:28
→ lausi joined01:29
→ Vampire0 joined01:29
← lausi left01:30
← duckxx left01:30
→ dc5ala_ joined01:31
→ mcluskydodallas joined01:31
← hwrd|work left01:31
← WillMarshall left01:31
← Speed left01:32
← w09x_ left01:32
→ neurodrone joined01:32
← dc5ala left01:34
← mcluskydodallas left01:34
← adityamenon90 left01:34
→ lonewulf` joined01:35
→ adityamenon joined01:36
→ harish joined01:36
← iamjarvo left01:37
← segher left01:37
→ duckxx joined01:37
→ apok_ joined01:38
→ segher joined01:38
← diminoten left01:39
← segher left01:39
→ iamjarvo joined01:40
→ segher joined01:40
→ diminoten joined01:41
← AaronMT left01:41
← duckxx left01:42
← hachque left01:42
← apok left01:42
← apok_ left01:43
← aviraldg left01:44
→ intripoon_ joined01:45
→ GNUtoo joined01:46
← tinti left01:46
→ hackez joined01:46
← alexim left01:47
← SeoZ-work[AWAY] left01:48
← intripoon left01:48
→ seo joined01:49
seo → Guest127501:49
← Guest1275 left01:50
→ SeoZ joined01:50
→ Maccer joined01:51
→ chuckharmston joined01:52
← nadirvardar left01:53
← aruiz left01:54
← tiglog left01:56
→ Berglund joined01:56
→ WillMarshall joined01:58
← iamjarvo left02:02
GNUtoo sigh sigh sigh02:03
→ iamjarvo joined02:03
GNUtoo it's now blocked here:02:03
Bisecting: 8350 revisions left to test after this (roughly 13 steps)02:03
CareBear\,git checkout -f didn't work02:04
← arietis left02:04
GNUtoo cmn, ^^^02:05
SethRobertson define didn't work02:05
→ lresende joined02:06
GNUtoo I'm doing a bisect....02:06
and it often repeat the same message02:07
at the beginning the message change02:07
← iamjarvo left02:07
GNUtoo and at some point after using git bisect skip....02:07
← lresende left02:07
← orafu left02:08
→ orafu joined02:08
GNUtoo it keep saying the same message for differents commits02:08
after git bisect good or bad or skip02:08
Bisecting: 8350 revisions left to test after this (roughly 13 steps)02:08
← Araxia left02:08
→ mshroyer joined02:09
acidjazz so this happens every now and again .. i clearly have dpl/log/hook.log in my .gitignore and its not ignoring the file on production..02:09
SethRobertson If you git bisect visualize or git bisect log, what do you see?02:09
acidjazz: Is the file committed?02:10
acidjazz ah yes at 0 bytes02:10
SethRobertson acidjazz: !ignore_tracked02:10
gitinfo acidjazz: Git only applies ignore patterns to untracked files. You can't use ignore patterns to ignore changes to files that are already tracked by git. Still, see https://gist.github.com/1423106 for ways people have worked around the problem.02:10
← enmand left02:10
acidjazz cant i just remove the file from my tree?02:11
and itll start to ignore it?02:11
SethRobertson Yes02:11
→ iamjarvo joined02:11
← Stalafin left02:11
→ iwoj joined02:12
→ zdk joined02:12
← cooloney left02:13
acidjazz git filter-branch -f --tree-filter 'rm -rf dpl/log/hook.log' HEAD02:13
this hsould do it right02:13
git push w/ --force02:13
→ cilly joined02:14
GNUtoo I'll look02:14
I already restarted the bisect....02:15
← dinesh6654 left02:16
← Berglund left02:16
← Eridius left02:17
← iwoj left02:19
→ iwoj joined02:19
→ dhruvasagar joined02:20
← bgerber left02:20
CareBear\ GNUtoo : at this point it would be very helpful if you can pastebin a full log of your commands and the git output since git bisect start02:22
→ Eridius joined02:23
→ pdgreen joined02:23
gitinfo set mode: +v02:24
iwoj i just aborted a git rebase and it blew away the last two months of commits.02:24
← zdk left02:24
iwoj how to i bring back the last commit if I don't know it's hash code?02:24
→ fireh joined02:24
← duko left02:25
GNUtoo ok02:25
← chuckharmston left02:25
iwoj is there a way to undo a git rebase abort?02:26
thallada_ → thallada02:26
→ zdk joined02:26
GNUtoo I'll try to reproduce the issue and I'll do that02:27
iwoj, git reflog02:27
iwoj thanks gnutoo02:27
GNUtoo maybe it tries not to go in the skipped zone?02:28
← pdgreen left02:28
→ bgerber joined02:29
← eletuchy left02:29
iwoj git reflog show HEAD@{1} ?02:30
GNUtoo iwoj, you can even checkout it02:30
← nutron left02:30
← zdk left02:31
→ zdk joined02:32
→ cooloney joined02:33
← SiDz left02:34
infogulch → infogulch|afk02:35
→ SiDz joined02:36
← cakehero left02:37
→ shirajermilio joined02:39
← rehat left02:41
→ nadirvardar joined02:43
→ QaDeS joined02:44
← camdenx left02:45
← jsoa left02:45
→ eletuchy joined02:46
→ artnez joined02:48
← kerrick left02:48
→ cakehero joined02:50
→ pandeiro_ joined02:50
iwoj thanks02:50
← plaxx left02:51
← shirajermilio left02:52
→ rehat joined02:52
← cakehero left02:54
← kpreid left02:56
→ kpreid joined02:56
→ duckx joined02:58
iwoj how do I bring (no branch) back into a specific branch?02:59
… or how to I name (no branch)02:59
milki iwoj: git branch branchname03:00
iwoj thanks!03:00
← Textmode left03:01
→ kerrick joined03:03
→ CannedCorn joined03:04
→ hydrozen joined03:06
→ mankoff joined03:07
→ blorbx joined03:08
← iwoj left03:09
← Vert left03:10
→ Yuuhi``` joined03:11
→ williamcotton_ joined03:11
← Yuuhi`` left03:12
← Eridius left03:12
CareBear\ iwoj : you never bring things into branches03:14
iwoj : please stop for a moment and spend some time to learn the git data model, so that you can avoid further problems03:14
← dhruvasagar left03:15
→ Textmode joined03:15
← TomyLobo left03:15
← Anxi80 left03:17
CareBear\ iwoj : the commit is central. each commit refers to a tree and usually a parent commit. the tree is sortof a snapshot of all files on disk. all commits except the very first one in each repo has one or more parent commits.03:18
→ cakehero joined03:19
← Davey left03:19
CareBear\ iwoj : commits are immutable. once a commit has been created that commit will never change.03:19
iwoj : a branch points to the latest commit. git knows what the active branch is, and updates that branch if a new commit is created.03:20
iwoj : it's also possible to rewrite history of branches; that is, to replace commits on a branch with some other commits.03:21
→ dhruvasagar joined03:22
GNUtoo http://pastie.org/private/ubtrz6kjoxtakohmsfdfw03:24
is that normal?03:24
← adityamenon left03:24
← maxwellcorbin left03:24
← Textmode left03:25
CareBear\ GNUtoo : is what normal, exactly? it looks fine to me03:26
← goshawk left03:26
GNUtoo ok03:27
the steps remaining don't change03:27
CareBear\ the number of revisions does03:27
→ towski joined03:27
GNUtoo yes but very slowly....03:28
→ Textmode joined03:29
← tjholowaychuk left03:30
→ apok joined03:30
← Wulf4 left03:35
← kukks left03:37
→ gcw|msrv joined03:37
infogulch|afk → infogulch03:38
← osmosis left03:40
ivan` https://sites.google.com/a/chromium.org/dev/developers/how-tos/git-repo03:40
← segher left03:41
→ Wulf joined03:42
GNUtoo I'll go again03:42
bye03:42
← GNUtoo left03:42
→ nutron joined03:43
gitinfo set mode: +v03:43
→ segher joined03:43
gcw|msrv question - with git is it possible to have two branches checked out from the same (local) repository (without cloning to make a full copy) ? I've been using git for years but have not been able to come up with a solution for this...03:43
← segher left03:43
→ segher joined03:44
→ revagomes joined03:45
← cooloney left03:48
← segher left03:48
frogonwheels gcw|msrv: No, for most values of y.03:48
gcw|msrv: If you want a 'read-only' copy somewhere, you can always export it with git-archive03:49
gcw|msrv frogonwheels: hrm ok..03:49
thx03:49
→ duckxx joined03:49
frogonwheels gcw|msrv: in the git contrib there's a shell script that allows two working directories for one repository, but I do not know the consequencex03:50
s03:50
← revagomes left03:50
← duckx left03:50
→ revagomes joined03:50
→ segher joined03:50
frogonwheels gcw|msrv: (I only know of it 'cause I was fixing/hacking it to work for msysgit symlink patches.)03:51
→ cjs joined03:51
cjs How do I install git on a Linux machine that doesn't have git available in its repository (CentOS 5)?03:51
frogonwheels cjs: use another distro? wow, unbelievable .. ott - compile it from source03:52
cjs How do I get the source?03:52
(That was my first idea, but I've not been having much success with that.)03:52
→ Eridius joined03:52
frogonwheels cjs: use git to fetch it ;P03:52
gcw|msrv frogonwheels: ahh ok.. thanks I really appreciate the info..03:52
cjs (And yes, CentOS os totally OTT. The latest version, CentOS 6, uses gcc 4.4, which is no longer supported by the gcc project.)03:53
frogonwheels cjs: https://github.com/git/git [zip] possibly?03:55
cjs: ott is also other than that03:55
← weiyang left03:55
frogonwheels gcw|msrv: 'git-new-workdir' is the script.03:56
← duckxx left03:57
frogonwheels gcw|msrv: it basically soft-links everything that can be. but again, I do NOT know the consequences.03:57
→ stkrzysiak_ joined03:57
frogonwheels gcw|msrv: referenced here: http://www-cs-students.stanford.edu/~blynn//gitmagic/ch07.html actually :)03:58
cjs frogonwheels: Ah, that ZIP button was the trick! I hadn't noticed that.03:59
frogonwheels cjs: btw #github is a good place to go if you want to ask github interface questions :)03:59
→ adityamenon joined04:00
← stkrzysiak left04:01
gcw|msrv frogonwheels: hrmmm that looks pretty dangerous... I'm thinking I'll steer clear ;-)04:01
frogonwheels gcw|msrv: lol. yeah, indeed.04:01
← stkrzysiak_ left04:01
→ bddn joined04:05
← hyperair left04:06
← gavin_huang left04:10
→ ReekenX joined04:11
← iamjarvo left04:11
→ sangi joined04:11
← spyro left04:11
← cilly left04:12
→ robbyoconnor joined04:13
→ iwoj joined04:14
→ iamjarvo joined04:17
← robbyoconnor left04:17
→ jonwil joined04:18
→ cilly joined04:18
← cakehero left04:20
← apok left04:20
gitinfo set mode: +v04:21
jonwil Is there a way to display a nice graphical view of the commits and stuff in a git repository? I swear I did it before but now I cant remember how04:21
kpreid gitk04:22
opinions may differ on "nice", but it's standard and it's graphical04:22
→ beautifulmind joined04:22
jonwil thats what I wanted, thanks04:23
now I just gotta find where my git install put the binaries for it...04:23
(its windows so its non-standard)04:23
← cjs left04:24
→ mikepack joined04:24
← kurrataW left04:24
→ jnewt joined04:24
← blami left04:25
offby1 gitk actually (typically) looks better on Windows than on *nix04:25
true story04:25
→ Tegi_ joined04:26
jnewt i'm struggling with what it means to have a local branch "track" a remote branch. I have no idea what that means, unless it's just a name mapping, but it seems important, and i can't find an explanation04:26
→ blami joined04:27
→ subhojit777 joined04:27
ChanServ set mode: +v04:27
← williamcotton_ left04:27
→ Araxia joined04:27
jnewt is it just a name mapping, like "there's a branch on remote called x, but i want to call it y locally, so call it y but when i push, push it as x"?04:27
EugeneKay jnewt - it means that when you run "git pull" the remote branch will automagically be fetched and merged into your equivalent local banch, and when you push it will do the same.04:28
→ stkrzysiak joined04:28
← jonwil left04:28
→ spyro joined04:29
← spyro left04:29
→ spyro joined04:29
jnewt what do you mean by "equivalent"? if i clone a repo, all branches are equivalent. if i fetch, it will fetch all changes to my local. what is the point of tracking or not tracking, i'd say everything is tracked.04:29
EugeneKay s/equivalent/tracked04:29
It means that branch X is matched to remote/X04:29
jnewt isn't that always the case by default?04:29
EugeneKay If you checkout a branch from a remote branch, yes.04:30
→ apok joined04:30
EugeneKay YOu can change the tracked branch to a local branch, or a different remote branch, or to something involving space aliens. But you usually won't.04:30
jnewt what if i just clone a remote to local, then checkout the local branch, as is the usual case, wouldn't that local branch be tracking the remote branch?04:31
← tazle__ left04:32
← subhojit777 left04:32
EugeneKay Yup.04:32
→ subhojit777 joined04:33
← subhojit777 left04:33
→ subhojit777 joined04:34
jnewt and the only time i'd have an untracked branch is if i created one locally, committed, but didn't push correct?04:35
but then when i pushed, would i need to fetch to "enable" tracking?04:35
or use --track04:35
EugeneKay The latter, yes.04:38
← mikepack left04:38
jnewt so even if i push a local repo back to the origin (assuming a centralized model), i haven't done something that would make the branch tracked?04:40
it seems --track is default04:41
← iwoj left04:42
jnewt i'm reading on this page that is confusing the git out of me: http://git-scm.com/book/en/Git-Branching-Remote-Branches04:43
← CannedCorn left04:44
← revagomes left04:44
→ cooloney joined04:45
EugeneKay I advise you to stop worrying about it. If it does something you don't expect, use `git branch -avv` and `git branch --set-upstream` to make it do what you want. Then move on with your life.04:45
→ camdenx joined04:45
EugeneKay You have the basic idea, damn the details.04:46
← yoklov left04:46
jnewt the very bottom 2 sections are reading like "who the fuck came up with this crazy business"04:46
→ lresende joined04:46
jnewt then i remember...linus04:46
EugeneKay git is a proof of software evolution rather than intelligent design04:46
Linus envisioned git as an object model. The fact that the binary has become a full-fledged VCS rather than something which is wrapped in a user-friendly GUI is frankly terrifying04:47
So.... go with it. And pray it gets rewritten for v2.0 ;-)04:47
→ juvenal joined04:47
← bgerber left04:48
→ yoklov joined04:48
jnewt i've been using from cli, so far, and i personally would really appreciate someone writing a better gui / interface for windows than tortoiseSVN or some of the other's i've tried. they all seem confusing thus far, moreso than pecking away on git bash (win user here)04:49
← camdenx left04:49
→ wilmoore joined04:50
EugeneKay SmartGit04:52
chrisf|walrus jnewt: the github guys have done a 'friendly gui' for windows recently04:53
← avpx left04:54
chrisf|walrus jnewt: i dont know whether it's usable for repositories hosted elsewhere04:54
EugeneKay It's terrible and useless04:54
→ bgerber joined04:54
jnewt well, that's cleared up :)04:55
chrisf|walrus tbh, i dont get the appeal of a gui for version control at all, beyond a good history browser (for which tig wins)04:55
→ arvind_khadri joined04:58
← madewokherd left05:00
→ edude03 joined05:00
jnewt EugeneKay: just cloned a repo with smartgit, and the file list shows all untracked. smartgit looks good, but i'm right back to the misunderstanding of tracking. i just cloned it!05:01
EugeneKay Untracked or unchanged? :-p05:01
← bigkm left05:01
← Zenmaker__ left05:01
jnewt under the State Column, every single file shows untracked. granted i'm not on the tip (backed up using the log), but that seems to be at odds with what we discussed05:02
← rbarraud left05:02
→ Zenmaker__ joined05:02
→ bigkm joined05:03
← wilmoore left05:04
jnewt ok, just cloned another, tons of files are now in my local working tree. not a single file in the files list on smartgit. maybe i got a bad copy :)05:09
← yoklov left05:12
← Bass10 left05:14
→ iwoj joined05:14
← hydrozen left05:14
← towski left05:15
→ towski joined05:15
→ kleist joined05:16
→ tazle_ joined05:16
← lresende left05:18
→ lresende joined05:19
← project2501b left05:22
← lresende left05:24
→ lresende_ joined05:24
← dhruvasagar left05:24
→ Pupeno_W joined05:24
← adityamenon left05:25
→ samuel joined05:29
→ private joined05:31
private → Guest5817105:31
→ k1t joined05:34
→ andrewhl joined05:35
k1t hey folks, how might I squash the past 5 commits on origin/master ? if I do 'git rebase -i HEAD~5' it pulls up the past 7 commits, 2 of which are not mine. I just want to squash the past 5 commits from my forks remote05:35
← imrobert left05:36
→ JZA joined05:36
JZA hi anyone here know how can I recover from deleating my head files05:36
I am trying to upload my project to gitorious but I had issues pushing my code into it.05:37
so I delete the .git folder hoping it will re-generate but it seems it got worst now05:37
k1t https://github.com/whitmanc/spree_product_sort/commits/master id actually like to sqush commit 1, 3, 4 and 5... and leave 2 as its own commit, if I could05:38
id be happy to read a ref if someone could point me in the right direction05:38
← marc_kdab left05:38
_ikke_ JZA: Deleting the .git directory removes the actual repository05:38
JZA: You'd have to reclone it05:38
→ longbeach joined05:38
_ikke_ JZA: That's not a very smart thing you have done05:39
k1t: look at man git rebase -i05:39
gitinfo k1t: the git-rebase manpage is available at http://git-scm.com/docs/git-rebase.html05:39
← codewrangler left05:40
JZA _ikke_: I havent yet push my code to the gitorious website, so I don't really know how to re-clone it lo05:41
← iamjarvo left05:43
k1t thx05:43
→ iamjarvo joined05:43
_ikke_ JZA: Well, you're kind of out of luck then05:43
The only thing you have left is you working copy05:44
← edude03 left05:44
JZA _ikke_: I have folder1/<code> deleted folder1/.git/ then created folder2/ moved my code, want to generate a new repo.05:44
_ikke_: that's the point, there was always only one repo in my hd. I havent push it yet.05:45
_ikke_: I want to know how to create the master, and push it for the first time.05:46
jnewt clone the github repo, then put your code in05:46
_ikke_ jnewt: He's not using github05:46
JZA jnewt: is gitorious, not github.05:46
← glennpratt left05:47
JZA but I guess the process is the same05:47
jnewt re clone.05:47
_ikke_ JZA: Is there anything on gitorious?05:47
jnewt you lose your local history, you05:47
JZA _ikke_: no05:47
→ goshawk joined05:47
→ weiyang joined05:47
JZA first time I tried to push it, it didnt recognized my publickey05:47
_ikke_ JZA: git init then05:47
and the copy the code to that folder05:47
JZA ok05:48
_ikke_: ah I think that was it05:48
jnewt but you arent' going to push back to gitorwahtever that way05:48
_ikke_ jnewt: You can05:48
JZA: git remote add origin <url-to-gitorious-repo>05:48
jnewt fancy new trick for me!05:49
_ikke_ JZA: the first time you push master, you'd to: git push -u origin master05:49
← scientes left05:50
→ marc_kdab joined05:50
→ scientes joined05:51
JZA _ikke_: I get this.. git checkout master .. error: pathspec 'master' did not match any file(s) known to git.05:51
what does that mean?05:52
→ skorgon joined05:54
→ ThomasLocke joined05:55
_ikke_ JZA: When you create a repo (git init), there is no actual branch called master yet05:55
JZA: Just create a commit05:55
JZA ok05:55
← nadirvardar left05:56
← scientes left05:58
frogonwheels JZA: After init !HEAD contains a reference to master - the first commit will cause the branch to be created.05:59
gitinfo JZA: 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.05:59
JZA ok I am on the same error as before.05:59
I guess this is more on the gitorious end05:59
git push origin master05:59
Permission denied (publickey).05:59
fatal: The remote end hung up unexpectedly05:59
→ nadirvardar joined05:59
_ikke_ Your public key is not accepted05:59
frogonwheels JZA: yeah - you'll need to debug the ssh connection. ssh -V will help06:00
_ikke_ but that has nothing to do with your repository06:00
← dreiss left06:00
frogonwheels JZA: set up your pubkey with gitorious, and make sure your private key is available locally. using ssh will help you pinpoint the issue06:00
JZA frogonwheels: is there an attribute to select a different key, like ssh -i .ssh/otherkey.pub06:02
← EyesIsMine left06:02
frogonwheels JZA: /me shrugs06:02
argh06:02
you know what i mean.06:02
← iwoj left06:03
→ edude03 joined06:03
→ sq-one joined06:05
_ikke_ JZA: You'd have to use .ssh/config06:06
→ arti_t joined06:06
← lonewulf` left06:07
→ marthinal joined06:07
→ alexukf joined06:08
→ lonewulf` joined06:09
→ patrick85 joined06:09
← vlrk left06:10
← nadirvardar left06:11
← chrisward left06:12
← alexukf left06:13
→ michaelmartinez joined06:13
iamjarvo so i did a rebase and it rewrote the branches history. the branch was already pushed though. i am the only one working on the "feature" branch. Its probably not good practice but I am thinking i should clobber the branches on origin and push the new ones up06:14
→ cbreak-work joined06:14
frogonwheels iamjarvo: A nice way of handling this (an example I took from somebody else) kind of situation is to name your feature branch myfeature/v1 and then name it myfeature/v2 when you rebase it and push it06:15
→ sixteneighty joined06:15
frogonwheels iamjarvo: ott you can force push the branch up ( git push origin +featurebranch )06:15
→ leeN joined06:15
← sixteneighty left06:15
iamjarvo i like the first version better calling it a v206:15
frogonwheels iamjarvo: Just be aware that if you have myfeature you can't have myfeature/v206:16
→ maletor joined06:16
→ chrisward joined06:16
chrisward → Guest5776006:16
→ quazimodo joined06:16
frogonwheels iamjarvo: just cause in the first case 'myfeature' is effectively a file, and in the second 'myfeature' is effectively a diretory :) (even though when tags are compressed that may not be strictly true)06:16
iamjarvo frogonwheels i think i will make it easier and just tack on a -v206:17
frogonwheels iamjarvo: ah yes. that will work as well :)06:17
iamjarvo: you can then cull the old branches as necessary06:18
→ papegaaij joined06:18
← drl left06:19
iamjarvo thanks frogonwheels06:19
← skorgon left06:21
→ EyesIsMine joined06:21
← Dreamer3 left06:23
→ olaf_ joined06:24
→ bosspetta joined06:25
→ vlrk joined06:26
→ ottoshmidt joined06:28
→ nielsbusch joined06:28
← lresende_ left06:29
← rue_XIV left06:30
→ rue joined06:30
→ soee joined06:30
→ flijten joined06:30
← iamjarvo left06:31
← fisted_ left06:31
← edude03 left06:31
→ drl joined06:32
→ fisted joined06:32
← clifton left06:33
← Schwern left06:37
← Jay_Levitt left06:37
← michaelmartinez left06:39
← darren_ left06:41
← vlrk left06:46
→ Schwern joined06:46
→ tatsuya_o joined06:47
← _bart left06:47
→ dinesh6654 joined06:49
← sq-one left06:52
→ dreiss joined06:54
← tatsuya_o left06:55
← longbeach left06:55
← pandeiro_ left06:55
← ISF left06:56
→ dhruvasagar joined06:57
k1t is there a reason 1 commit wouldnt show up when i 'git rebase -i HEAD~2' ? It actually shows 4 commits, 3 from another individual from before i forked.06:57
any ideas heh06:57
milki maybe its merge related06:58
→ und3f joined06:59
→ bubbles10101 joined06:59
← txdv_ left07:00
→ skorgon joined07:01
→ jceb_ joined07:02
→ EricKamsky_ joined07:02
→ drizzd joined07:02
k1t https://github.com/whitmanc/spree_product_sort/commits/master considering theres a merge right before, definitely a possibility07:03
← EricKamsky left07:03
← nobdrais1ntone left07:03
→ txdv joined07:04
k1t so close, just need to squash those 2 top commits07:04
haha07:04
→ vlrk joined07:05
→ Chib joined07:05
← ottoshmidt left07:05
← imarcusthis left07:06
→ ottoshmidt joined07:06
→ clifton joined07:06
→ hoschi joined07:06
→ shruggar joined07:07
← towski left07:07
cbreak-work k1t: rewriting public history is bad07:07
→ imarcusthis joined07:07
k1t well its my fork, i was trying to clean up before i did a pull req... I should be working a branch, realy07:07
really*07:07
→ kurrataW joined07:08
k1t cbreak-work: either way, i agree with you..07:08
→ thiago joined07:08
← mishok13 left07:08
→ defts joined07:09
→ tatsuya_o joined07:09
→ ph^ joined07:09
k1t i dont understand why rebase wont show the 2nd commit though07:09
← kleist left07:11
← drizzd left07:11
← yuwang_ left07:12
cbreak-work rebase is not the correct solution07:13
not if you have a merge07:13
ottoshmidt is there a good guide on how to setup git server side somewhere?07:13
*anywhere07:13
cbreak-work just set up an ssh server and install git on it. done.07:13
if you want access control, take a look at gitolite07:14
EugeneKay Definitely gitolite.07:14
← Nibato left07:15
← jceb_ left07:16
→ jceb_ joined07:16
ottoshmidt thanks07:16
← Goplat left07:17
← kerrick left07:18
← stkrzysiak left07:18
→ tqmz joined07:19
← bizhanMona left07:20
← palm_vue left07:20
← NimeshNeema left07:20
← tatsuya_o left07:20
← wagle left07:20
→ bizhanMona joined07:21
→ palm_vue joined07:21
→ psoo joined07:22
→ _iron joined07:23
→ tatsuya_o joined07:25
← shruggar left07:25
← Chaser left07:26
→ wagle joined07:26
→ subhojit_ joined07:26
→ Chaser joined07:27
← bgerber left07:27
→ kwijibo joined07:28
→ mishok13 joined07:28
→ shruggar joined07:28
← andrewhl left07:29
← apok left07:30
← Eridius left07:30
→ solofight joined07:31
← solofight left07:31
→ giallu joined07:32
← dreiss left07:32
→ purencool joined07:33
← shruggar left07:33
← ph^ left07:33
← goshawk left07:33
→ shruggar joined07:34
→ goshawk joined07:34
→ tsdh joined07:34
← nyuszika7h left07:35
← maletor left07:38
→ bgerber joined07:39
← tatsuya_o left07:39
→ jn_ joined07:40
← shruggar left07:42
← beautifulmind left07:44
← eletuchy left07:45
→ ipalaus joined07:47
← WillMarshall left07:48
clifton i wrote a pre-commit hook to strip trailing whitespace and a few other changes from *only* files changed, looking for feedback (mostly did it to clean up coworker commits) http://git.io/pre-commit07:49
SeoZ hello07:51
gitinfo SeoZ: 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.07:51
← cooloney left07:52
SeoZ i have a question.07:52
_inte_ You don't.07:52
SeoZ hehe07:52
you knew it!07:52
i reset my git to previous revision to remove recent history.07:52
how can i push this?07:52
→ chris_dexter joined07:53
SeoZ i got07:53
! [remote rejected] HEAD -> refs/for/master (no new changes)07:53
error:07:53
gitinfo set mode: +v07:53
chris_dexter anyone can help me out with a phpStorm problem?07:54
→ polymar joined07:54
→ dpino joined07:54
← Cerise left07:55
chris_dexter just yesterday I have installed msysgit, but when I try to configure phpStorm to work with git, it says "Errors while executing git --version. exitCode=<some numbers>"07:55
any hints on that?07:56
→ tvw joined07:57
→ WillMarshall joined07:57
← dinesh6654 left07:57
→ Cerise joined07:58
← chris_dexter left07:58
→ _bart joined07:59
→ dr_lepper joined07:59
→ chris_dexter joined07:59
← Guest58171 left07:59
← tvw left08:00
← patrick85 left08:01
→ Lord_DeathMatch joined08:03
charon chris_dexter: what's the actual exit code?08:04
does it find the git executable?08:04
← WillMarshall left08:04
gitinfo set mode: +v08:05
chris_dexter -107374151508:05
yes it finds the executable08:05
← flijten left08:05
chris_dexter I have a hunch that it's because of some missing DLL08:06
→ LimeBlast joined08:06
chris_dexter oh yea, libiconv-2.dll08:06
I'll try to figure it out, I'll let you know if I fixed it08:07
thiago transform that number into hex and google for it08:07
c000013508:07
→ patrick85 joined08:08
← arti_t left08:08
chris_dexter thanks, I didn't know this worked08:08
→ cooloney joined08:08
→ arti_t joined08:08
→ OOPMan joined08:09
← igotnolegs left08:09
gitinfo set mode: +v08:10
ljbo question about git svn: in the SVN repo /trunk/boost@38326 was copied to /devel@38327 and then /devel@38328 was copied to /trunk@38329. git svn fetch logged it was feeding refs/remotes/svn/trunk (I used --prefix=svn/) but then it restarted from the beginning, this time feeding refs/remotes/svn/trunk@ 3832608:11
I would like to understand what is going on08:11
→ flijten joined08:11
thiago ljbo: do you mean /trunk was deleted between 38326 and 32328?08:12
chris_dexter fixed it :-) added mingw/bin to the %path%08:12
→ Eridius joined08:13
chris_dexter thanks guy, sorry for bothering with a silly issue like this08:13
← tabakhase left08:13
→ tabakhase joined08:13
charon chris_dexter: well, we're glad to help. i hope you learned not to elide "useless" information :)08:13
ljbo Changed paths:08:14
D /devel08:14
A /trunk (from /devel:38328)08:14
they surely did: svn copy devel trunk08:14
sorry: svn mv devel trunk08:14
← neurodrone left08:15
ljbo yeah, sorry, dunno why I wrote "copied" above08:15
need another coffee08:15
thiago ljbo: then you now understand what happened08:15
adhawkins-away → adhawkins08:15
ljbo svn-side yes, but I don't understand what git svn is doing08:17
thiago ljbo: it's trying to follow the SVN history08:17
ljbo: that doesn't mean it's succeeding08:17
ljbo so in the first phase, when it was feeding .../trunk, from where was it taking the history then?08:18
thiago moving trunk is a sure way to break it08:18
ljbo that was gonna be my next question…08:18
thiago did it finish importing?08:18
ljbo still running08:19
thiago when it finishes, inspect history and find out if it was correct08:19
ljbo it will take forever as there is a release branch to fetch next with the same issue08:19
I will do08:19
thiago that's because you're using git-svn, which is SLOW08:19
→ d0k joined08:19
thiago there's an alternative for one-off imports which is much quicker, up to 20x quicker08:19
svn2git08:19
ljbo I tried that and it is indeed blazingly fast08:20
but the imported repo has directories trunk, branches, etc08:20
thiago you need to edit the config file to tell it how to import08:21
ljbo well, of course, it's git we are talking about08:21
thiago unlike git-svn, it doesn't guess. It needs to be told of the layout.08:21
ljbo I think I had done that correctly08:21
I concluded that svn2git ignored the config file08:21
← kurrataW left08:21
ljbo or perhaps it has some special configuration of its own?08:21
← cooloney left08:21
thiago it doesn't run without a config file08:22
so it can't be ignoring it08:22
ljbo so, yeah, I was going to say that I could use git replace and git filter-branch to fix things afterward08:22
there was a config file08:22
thiago that's an option too08:22
ljbo but it did not honour it08:22
well, if you tell me that for sure it would honour it then I'll go back to that and check right now08:23
thiago more than like, it did honour it08:23
but your config file wasn't what you wanted it to do08:23
ljbo ok, great news then!08:23
thanks!08:23
→ angelsl joined08:25
→ shruggar joined08:25
→ fornext joined08:26
→ kurrataW joined08:27
← dhruvasagar left08:28
→ dhruvasagar joined08:29
→ tweedledumber joined08:30
← arti_t left08:30
→ Speed joined08:30
← Speed left08:30
→ Speed joined08:30
→ arti_t joined08:30
← tweedledumber left08:31
← Null_Route left08:32
→ aviraldg joined08:32
Guest57760 → chrisward08:34
← bigkm left08:34
← Speed left08:35
→ JohnBat26 joined08:36
← ottoshmidt left08:36
→ ottoshmidt joined08:37
→ bigkm joined08:38
→ Speed joined08:39
← Speed left08:39
→ Speed joined08:39
← werdnativ left08:39
→ cooloney joined08:39
← maknz left08:39
→ d0k_ joined08:42
← d0k left08:43
d0k_ → d0k08:43
← ReekenX left08:43
→ tatsuya_o joined08:44
→ alexukf joined08:44
← kwijibo left08:44
→ whitman joined08:45
GeorgeJ Hello folks.08:45
Can I configure my repository so that doing a `git push` on my develop branch be equivalent to `git push heroku develop:master` ?08:45
→ maknz joined08:45
← telmich left08:46
→ camdenx joined08:47
→ ReekenX joined08:48
→ bullicon joined08:48
← camdenx left08:48
← aviraldg left08:49
← JohnBat26 left08:49
→ yaydin joined08:49
shruggar I don't think you can specify the default push remote on a per-branch basis, though you can specify the develop:master part08:49
→ JohnBat26 joined08:50
GeorgeJ shruggar: How could I do so?08:51
← yaydin left08:52
shruggar remote.origin.push=refs/heads/develop:refs/heads/master … but again, that's not per-branch, that just means that every time you "git push" anywhere without being more-specific, it will push "develop" to "master"08:52
that would be remote.heroku.push, in your case08:53
cmn you can set push.default to upstream and make heroku's master develop's upstream08:53
or just make an alais08:53
alias08:53
← thiago left08:54
→ RaceCondition joined08:54
shruggar btw, if you have a branch called "develop", you're probably doing something very wrong08:54
← RaceCondition left08:54
cmn why?08:55
shruggar because (especially as he's then just pushing directly back to master), it implies that you're not actually using branches08:56
→ rtuin joined08:56
jast this is heroku we're talking about08:56
→ kortti joined08:56
cmn for some people maybe08:57
jast normal rules of common sense don't apply ;)08:57
← alexukf left08:57
cmn but just because you're using a branch called develop it doesn't mean you're not using branches08:57
→ alexukf joined08:57
← oogatta left08:58
← ReekenX left08:58
← goshawk left08:58
→ oogatta joined08:58
→ ReekenX joined08:58
shruggar I'm guessing at least 99 out of 100 people who have a branch called "develop" either only work on one thing at a time, or have a lot of mixed history09:00
← arvind_khadri left09:00
→ RaceCondition joined09:02
→ caddoo joined09:03
← purencool left09:03
→ berserkr joined09:04
← LekeFly left09:04
bronsen git-flow encourages "develop" branch09:06
→ etcetera joined09:07
← k1t left09:07
← alexukf left09:07
charon bronsen: but it also uses a lot of topic branches09:08
GeorgeJ shruggar: Actually, the only reason I'm pushing the develop branch is to showcase the feature I'm currently working on.09:08
← RaceCondition left09:08
→ alexukf joined09:08
GeorgeJ I'm pushing to a 'daily' sub-domain. Which is only used for testing.09:09
cmn ok, so you've made shruggar's point09:09
GeorgeJ I am using the branches.09:09
I'm pushing master to the production repository.09:09
← artnez left09:10
→ jhllnd joined09:10
→ pranjal710 joined09:11
GeorgeJ I'm also using feature branches. Every time I merge a feature branch into develop, I also push develop to the testing sub-domain.09:11
← pranjal710 left09:11
← fornext left09:11
→ ph^ joined09:12
← Sky[x] left09:12
← albel727 left09:13
← alexukf left09:13
← cooloney left09:14
→ irqq joined09:14
→ TheJH joined09:15
→ lb joined09:15
→ vectorshelve joined09:16
gitinfo set mode: +v09:16
→ cooloney joined09:16
→ Davey joined09:16
← ph^ left09:16
→ nyuszika7h joined09:17
→ fsvend joined09:18
→ solofight_ joined09:18
→ LekeFly joined09:19
→ Vert joined09:23
← angelsl left09:26
← vectorshelve left09:26
← patrick85 left09:27
→ thiago joined09:27
→ fornext joined09:27
→ vervic joined09:31
← solofight_ left09:32
→ patrick85 joined09:34
← rapsys left09:35
← jjang left09:36
→ rbarraud joined09:36
→ jjang joined09:36
→ ph^ joined09:36
← ottoshmidt left09:37
→ ottoshmidt joined09:38
← TheJH left09:39
← und3f left09:40
→ Nibato joined09:43
← weiyang left09:44
← ljbo left09:44
→ Nibato` joined09:46
← harish left09:47
→ tvw joined09:48
← Nibato left09:48
← Lord_DeathMatch left09:51
→ Lord_DeathMatch joined09:51
← vervic left09:52
lb !beer09:52
gitinfo Beer! It's what's for dinner!09:52
← subhojit777 left09:53
← subhojit_ left09:53
← kurrataW left09:55
← matthiaskrgr left09:55
← nyuszika7h left09:55
← Sander^home left09:55
→ BassamJB joined09:57
← BassamJB left09:57
→ Camie joined09:57
→ BassamJB joined09:57
→ vervic joined09:58
→ danielpunt joined09:59
→ alexukf joined09:59
→ nyuszika7h joined10:00
← BassamJB left10:00
→ chowmain joined10:03
→ dvide joined10:03
← wagle left10:06
← giallu left10:07
← ottoshmidt left10:08
→ ottoshmidt joined10:09
→ wagle joined10:09
→ Fullmoon joined10:10
← Fullmoon left10:10
→ Sander^home joined10:11
→ subhojit777 joined10:11
→ subhojit_ joined10:11
← bosspetta left10:13
→ bosspetta joined10:14
← ottoshmidt left10:15
← subhojit_ left10:16
← subhojit777 left10:16
→ CaptainQuirk joined10:17
CaptainQuirk Hi !10:17
gitinfo CaptainQuirk: 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.10:17
← ph^ left10:17
CaptainQuirk I have a problem of synchronization between two git repos10:18
shruggar if you describe the problem, maybe someone will help you10:18
CaptainQuirk I had to revert from point B to point A on the local repo with reset --hard10:19
→ thm joined10:19
CaptainQuirk I then did git push -f origin master to synchronize both10:20
→ dsirijus joined10:20
thm how can I 'squash' multiple commits into one while cherry-picking? being on A, doing something like git cherry-pick B..C ?10:20
← bosspetta left10:21
CaptainQuirk but git log doesn't show what I would like it show10:21
which is, that the HEAD has been reset to the same commit as the local one10:21
git status on remote shows that I'm ahead with 9 commits10:22
shruggar sounds like !nonbare10:22
!bare ?10:22
gitinfo an explanation of bare and non-bare repositories (and why pushing to a non-bare one causes problems) can be found here: http://bare-vs-nonbare.gitrecipes.de/10:22
→ bosspetta joined10:22
CaptainQuirk more precisely that branch master is ahead of branch origin/master10:22
shruggar pushing to non-bare repositories does nothing to the work-tree and usually does nothing to the checked-out branch (Unless you have specifically overridden warnings and defaults or are using an ancient version of git)10:23
if you don't understand the implications, pushing to non-bare repositories is highly discouraged10:24
→ cissu joined10:24
shruggar how can I teach gitinfo about !nonbare ?10:24
← Schwern left10:25
→ ottoshmidt joined10:26
← jdav_gone left10:27
_ikke_ shruggar: .trigger_edit10:27
shruggar: !non-bare10:27
← bartman left10:28
CaptainQuirk shruggar, I don't fathom the whole implications. I know there's a hook on the remote server10:28
gitinfo shruggar: [!bare] an explanation of bare and non-bare repositories (and why pushing to a non-bare one causes problems) can be found here: http://bare-vs-nonbare.gitrecipes.de/10:28
CaptainQuirk but i don't know about its bare/non-bare property10:28
→ bartman joined10:28
_ikke_ CaptainQuirk: Basically, you should not push to non-bare repositories10:29
← cissu left10:29
CaptainQuirk _ikke_, well, there's a hook that worked just fine until now10:30
← rtuin left10:30
→ dfr|mac joined10:31
dfr|mac → dfr|moscow10:31
→ jdav_gone joined10:32
→ marhaban joined10:35
→ plazia joined10:36
plazia if a friend of mine (who is definitely not me) foolishly deleted a source file say "main.py" from his working folder ... by doing an "rm main.py" accidentally. could he use some sort of git command to restore that file?10:37
cmn plazia: ask status10:37
← alexukf left10:37
plazia cmn, you've lost me there matey10:37
← cthuluh left10:38
→ harish joined10:39
← berserkr left10:39
_ikke_ plazia: type in: git status10:39
plazia _ikke_, have done that10:39
← fireh left10:39
plazia i mean my friend has10:39
lb ^^10:39
plazia and it says main.py deleted10:39
adhawkins → adhawkins-away10:40
lb ? "git checkout -- main.py "10:40
_ikke_ plazia: read the rest10:40
→ stkrzysiak joined10:40
plazia ah10:40
cmn plazia: read what it says you should do to get it back10:40
plazia lb, that worked thanks... or at least that's what my friend says. :)10:40
cmn, _ikke_, my friend also thanks you for your help too10:40
lb plazia, not very convincing, your friend ;)10:40
plazia haha10:41
→ federic joined10:42
← irqq left10:42
← jhllnd left10:42
← chowmain left10:44
→ chowmain joined10:45
← Maccer left10:46
→ cthuluh joined10:46
← sweet_kid left10:46
→ rtuin joined10:46
→ changedNicks joined10:47
→ babilen joined10:47
→ camdenx joined10:47
changedNicks → sweet|kid10:49
← harish left10:49
→ alexukf joined10:50
← dhruvasagar left10:50
→ albel727 joined10:51
→ Nirvanko joined10:51
← camdenx left10:51
Nirvanko I've occasionally deleted a file now I have to restore it from Git repository. How to do it?10:52
← cilly left10:52
FauxFaux git checkout -- path/to/file10:52
← chowmain left10:52
Nirvanko error: pathspec 'app/config/parameters.ini' did not match any file(s) known to git.10:52
← dsirijus left10:52
canton7 how long ago did you delete it?10:52
lb Nirvanko, then your file was not tracked by git10:53
Nirvanko Not a long time ago10:53
lb what does git status tel you?10:53
canton7 as in, since the last commit?10:53
lb tell*10:53
Nirvanko This file is in .gitignore10:53
lb so it's not tracked10:53
so it's not part of your repo10:53
so you can't restore it from there10:53
← alexukf left10:54
Nirvanko But despite of that it was showed by Git as a file to be commited.10:54
→ alexukf joined10:54
canton7 Did you delete it before the last commit, or after it?10:54
→ beautifulmind joined10:55
lb did you added the file by typing "git add /path/to/file" ?10:55
cmn lb: just because it's in the ignore list, it doesn't mean it's not tracked10:55
← ottoshmidt left10:56
→ ottoshmidt joined10:56
cmn Nirvanko: you can use git log --diff-filter=D -- path/to/file to see what commit you deleted it in, and use commit^ in the checkout line10:56
lb cmn, "not tracked" is maybe unspecific.10:56
← OOPMan left10:56
cmn lb: was never tracked, then10:57
lb ;)10:57
cmn, i was just typing a long explanation ;)10:57
cmn, but of course you're right10:58
→ Pupeno_W_ joined11:01
← polymar left11:02
sweet|kid → sweet_kid11:02
← Pupeno_W left11:04
Pupeno_W_ → Pupeno_W11:04
→ nghbfgfgffggsdf joined11:05
← nghbfgfgffggsdf left11:06
→ drizzd joined11:07
→ cn28h joined11:09
fluxdude what is the equiv of svn up /path/to/dir_or_file when you only want to update to the latest revision of a part of your tree?11:09
→ Maccer joined11:09
fluxdude git fetch; git merge does everything I think11:09
← danielpunt left11:10
→ arietis joined11:10
fluxdude and I can't see from git merge --help if it's possible to only merge a subtree of the working copy11:10
← Textmode left11:11
→ himanshu_m786 joined11:11
← himanshu_m786 left11:12
→ dhruvasagar joined11:12
lb fluxdude, maybe git checkout SHA path/to/file ?11:12
(after git fetch of course)11:13
fluxdude lb: problem is I won't know what the SHA is without digging around and automating that is awkward11:13
← dfr|moscow left11:13
← ehsan left11:13
cmn merges work on whole trees11:13
→ landypro joined11:13
← zamba left11:13
cmn if you often need to grab some tree out of another commit, tagging it or pointing a branch to it can save you looking for it repeatedly11:14
← alexukf left11:15
← landypro left11:16
← wagle left11:16
→ landypro joined11:16
→ theCole joined11:16
→ angelsl joined11:17
← nyuszika7h left11:17
→ Stalafin joined11:17
← blacktulip left11:18
→ cissu joined11:18
← jjang left11:18
→ tgummerer joined11:19
jnewt cmn: that sounds like a good idea, but how would you point a branch to a portion of a tree at a specific commit?11:20
→ alberto56 joined11:20
→ jhoepken joined11:21
fluxdude '1811:21
opps wrong window11:21
shruggar well.. you could tag the subtree ;)11:21
→ dfr|moscow joined11:23
← bullicon left11:25
← ottoshmidt left11:26
→ the_cybersphinx joined11:26
→ ottoshmidt joined11:27
← cybersphinx left11:28
→ TomyLobo joined11:29
← Speed left11:30
← frogonwheels left11:30
← landypro left11:30
→ frogonwheels joined11:31
→ zackiv31 joined11:31
← arti_t left11:31
→ Seo9 joined11:32
← jhoepken left11:33
← zdk left11:33
→ TheJH joined11:34
cmn jnewt: you don't, you can tag a tree or make a branch to a commit11:35
→ tukja joined11:35
jnewt right, so the subtree approach he's talking about is impossible with git11:36
→ landypro joined11:37
← dfr|moscow left11:38
→ kleist joined11:39
← cooloney left11:40
← fsvend left11:41
→ roychri joined11:41
← tango_ left11:43
→ dfr|moscow joined11:43
→ aviraldg joined11:43
← dfr|moscow left11:44
roychri I am new to git (I used RCS (very long time ago), CVS (long time ago), SVN (until recently)). I created a topic branch (local) from my master (local) and I wonder if I can safely merge from origin in my topic branch, merge from origin in my master branch and then later merge my topic in my master safely?11:44
Or should I merge origin in my master first, then merge master in my topic so I can safely merge my topic branch in my master later?11:44
do I make any sense? :)11:44
canton7 what do you mean by 'origin'? A remote consists of a collection of branches, not a single branch11:45
selckin depending on your definition of safely, you can merge anything you want safely11:45
← Twey left11:45
roychri origin/master11:45
← dhruvasagar left11:46
roychri selckin: safely as in: nothing breaks like in SVN.11:46
canton7 both are fine. in the end, there's no real difference11:46
well, depends whether master has local changes11:46
→ dhruvasagar joined11:47
roychri canton7: logically, that is what I thought too. SVN 1.4 disagress :)11:47
disagrees11:47
selckin depending on your definition of breaks, you can merge anything you want11:47
roychri is glad git thinks more logically11:47
canton7 roychri, git won't disagree. if origin/master is at some commit, you can either merge that in, or update your local master to that commit and merge in the local master. git sees no difference11:47
well, the default commit message for the merge will differ, but that's ti11:48
selckin there are however strategies that will result in less conflicts or more logical history11:48
roychri selckin: oh? Such as?11:48
selckin but it should never "break"11:48
canton7 obviously, if master has local changes, then merging from origin/master will merge in different stuff than merging in the local master11:48
→ tango_ joined11:49
roychri canton7: I only merge with clean copies (no uncommitted changes)11:49
→ Twey joined11:49
canton7 roychri, sorry, I wasn't talking about uncommitted stuff. iirc git doesn't let you merge at all with uncommitted changes present11:49
roychri canton7: Oh, of course :) But I can always merge from my local/master as well then, right?11:49
← tukja left11:50
← rtuin left11:50
→ nyuszika7h joined11:50
canton7 roychri, sure, you can merge from either11:50
← Lord_DeathMatch left11:51
→ Lord_DeathMatch joined11:51
roychri canton7: at the same time I mean. Well, one after the other of course, but I mean, I don't have to choice one and stick with it. I can bring changes from both?11:51
← Milossh left11:53
← Davey left11:53
→ rtuin joined11:53
→ wagle joined11:54
→ Davey joined11:54
← cissu left11:55
← ottoshmidt left11:57
→ ottoshmidt joined11:57
→ neurodrone joined11:57
← quazimodo left12:00
→ iamjarvo joined12:01
→ matthiaskrgr joined12:02
→ tinti joined12:03
iamjarvo frogonwheels still around?12:04
→ ignas joined12:04
→ NimeshNeema joined12:05
lb roychri, maybe you should start using local feature / dev branches12:05
roychri lb: isn't it what my topic branch is?12:05
lb roychri, this way, you can always keep your local master up-to-date with origin/master12:05
meaning you can fastforward12:06
← kcj left12:06
lb and afterwards merge in your changes and push back to origin12:06
→ cannonball joined12:06
lb roychri, oh, i didn't see that12:06
← landypro left12:07
roychri lb: no problem :) Thanks for helping me understand :)12:07
SeoZ → SeoZ-work[AWAY]12:07
lb otherwise you can always work on feature X, keep master in sync with origin/master and merge in master from time to time (daily) in you feature branch12:07
there are some nice talks from schacon available on youtube12:08
should i give you the links?12:08
roychri lb: Right. I was just wondering if I can merge origin/master and local/master in my feature branch while merging origin/master into local/master anyway from time to time too.12:09
lb: oh yes, please :)12:09
→ cissu_ joined12:09
← samuel left12:10
← plazia left12:10
→ achadwick joined12:11
← Seo9 left12:11
→ doorp joined12:12
← juvenal left12:13
gitinfo set mode: +v12:14
doorp Hey, I'm trying to learn the very basics. I would like to setup GIT for a few users- is there a common approach for doing that while chrooting the users? (I tried to google, got lost very quickly)12:14
FauxFaux doorp: !gitolite12:14
gitinfo doorp: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html12:14
FauxFaux Unless you have some requirement for the chroot beyond mandatory access control.12:15
lb doorp, gitolite is the way to go12:15
→ blacktulip joined12:16
→ intransit joined12:17
← roychri left12:19
← simesy_work left12:19
→ alexukf joined12:21
→ Ryback_ joined12:22
← tqmz left12:23
→ cilly joined12:23
→ Zenmaker_ joined12:23
→ Milossh joined12:27
doorp FauxFaux: Thanks- it will take me a little time to grasp it, bbs :)12:27
← ottoshmidt left12:27
→ ottoshmidt joined12:28
← Zenmaker__ left12:28
→ bullicon joined12:28
→ skyf joined12:28
→ jsoa joined12:29
→ Athlon^ joined12:31
← sangi left12:32
← Lord_DeathMatch left12:32
→ arkiver joined12:32
→ Lord_DeathMatch joined12:33
→ Kredo joined12:35
adhawkins-away → adhawkins12:36
← etcetera left12:36
← bigkm left12:38
→ nickpresta joined12:38
→ etcetera joined12:38
→ Bass10 joined12:38
→ bigkm joined12:39
→ cakehero joined12:39
→ tommyvyo joined12:42
← Poapfel left12:42
→ Jay_Levitt joined12:43
← nielsbusch left12:45
← cakehero left12:45
→ plazia joined12:47
← unreal left12:47
→ camdenx joined12:48
→ unreal joined12:50
→ zamba joined12:51
← camdenx left12:51
→ adityamenon joined12:52
→ freeed joined12:54
→ chuckharmston joined12:56
← Chib left12:56
← fornext left12:57
← ottoshmidt left12:58
→ ottoshmidt joined12:58
← mshroyer left12:58
← etcetera left12:59
→ jkremser joined12:59
← Athlon^ left13:01
→ etcetera joined13:01
← skyf left13:02
→ macabre joined13:04
← Lord_DeathMatch left13:05
→ Lord_DeathMatch joined13:05
→ aib joined13:06
→ enmand joined13:06
→ AaronMT joined13:06
← aib left13:06
→ aib joined13:06
aib how do I remove all SVN information (created by init and fetch) so I can retry everything from scratch?13:06
I also ran svn init -s inside .git by mistake once :/ anything to delete other than .git/.git ?13:07
jnewt i'm trying to figure out how to do something as a test case, and what i did was added a file.c, file_fixed.c, and a few others, then committed, and then made some changes to them and committed a few times. now i want to go back and make file_fixed.c into file.c in a commit and remove file_fixed from history.13:07
→ hyperair joined13:07
→ t0nic joined13:08
lb aib, all git information is inside the .git dir13:08
so deleting it, will delete your repo13:08
jnewt, if you didn't pushed your changes somewhere you can just git reset --soft STATEBEFORE, and commit your final state and your fine13:09
aib yeah. I created a git repo inside that, so I deleted it.13:09
anyway, I've lost my svn refs and svn fetch still thinks everything is up to date13:10
← diamonds left13:11
jnewt here's my plan. git reset --hard <SHA of initial commit> git -branch fix , make changes in working tree (rename file_fixed.c), then rebase the first commit of the "changes" onto that branch, rename the branch default. sound about right?13:11
aib ah, deleting .git/svn worked13:11
→ SkAzZ joined13:11
→ und3f joined13:11
← ianmcorvidae left13:12
jnewt lb: it has been pushed, i cloned it this way, and want to make the changes and push it back to the origin repo13:12
→ ianmcorvidae joined13:12
→ nielsbusch joined13:12
jnewt it's gonna take me forever to look up and decide on the proper commands and switches to do this though :)13:13
→ nwest joined13:13
lb jnewt, git reset hard on pushed history and republish is always a bad idea13:14
!rewriting13:14
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)13:14
jnewt lb: my pushed history is not public, it's just backup / shared source for me only so i can use from multiple locations.13:15
→ erichynds joined13:15
→ SirCmpwn_ joined13:15
canton7 the same issues can occur though -- be careful13:15
→ allquixotic_ joined13:15
→ Berglund joined13:15
jnewt canton7: if i destroy it, so be it, i'll try and learn from my mistake.13:16
→ jperry2_ joined13:16
jnewt i have backups13:16
→ tchan1 joined13:16
lb jnewt, still you could use git reset SHA1fromstart ( not --hard ) and rebuild your history13:16
→ aib|2 joined13:16
→ Adys_ joined13:16
canton7 jnewt, it's not destroying that's the issue, it's accidentally merging together two versions of history13:16
→ quazimodo joined13:17
← Berglund left13:17
selckin jnewt: google interactive rebase13:17
← jperry2 left13:18
← eightyeight left13:18
← chayin left13:18
→ martialcoder joined13:18
→ chayin joined13:18
← tchan left13:18
← blast_hardcheese left13:18
jnewt let me se if i can reset first,13:19
← SirCmpwn left13:19
← aib left13:19
← LmAt left13:19
← allquixotic left13:19
← Internet13 left13:19
← Adys left13:19
← kevlarman left13:19
← palm_vue left13:20
← bizhanMona left13:20
→ LmAt joined13:20
→ bizhanMona joined13:21
→ palm_vue joined13:21
→ dv310p3r joined13:21
← federic left13:21
→ bleen1 joined13:22
→ Internet13 joined13:22
→ eightyeight joined13:22
→ blast_hardcheese joined13:23
t0nic → diamonds13:24
→ jhoepken joined13:26
→ manolo joined13:27
jnewt how much of the SHA do i need for a reset. i put the first 9 but it gives me a failed to resolve13:27
CareBear\ jnewt : enough to be unique13:28
jnewt : put the command you ran into a pastebin please13:28
jnewt : along with the output you got13:28
← plazia left13:28
jnewt http://pastebin.com/de0cPM3j13:29
→ skyf joined13:29
← ottoshmidt left13:29
→ ottoshmidt joined13:29
→ kurrataW joined13:31
CareBear\ jnewt : why -- master on the end?13:31
→ glennpratt joined13:31
← patrick85 left13:31
→ hseg joined13:32
jnewt CareBear\, because i read somewhere that was how it was done.13:32
← kermit left13:32
canton7 !blog13:33
gitinfo Blog posts, while helpful and informative, are quite often outdated, give bad advice, or are just plain wrong. Please don't rely solely upon them, or treat them as authoritative.13:33
← LekeFly left13:33
CareBear\ jnewt : strongly suggest read man git reset instead to learn how the command actually works13:33
gitinfo the git-reset manpage is available at http://git-scm.com/docs/git-reset.html13:33
← jhoepken left13:33
jnewt they have that -- path there too13:34
→ neurodamage joined13:34
← aib|2 left13:34
→ aib joined13:34
canton7 that's a path to a file, not a branch13:35
← alexukf left13:35
canton7 and not in combination with --hard13:35
→ john2496 joined13:35
aib I'm trying to base master onto an empty branch (svn trunk with a single commit) but I'm getting merge conflicts where master has branched and then merged. what can I do?13:35
canton7 so you can say e.g. 'git resset HEAD [--] path/fo/file'13:35
aib rebase*13:36
→ cjdavies joined13:36
← tmz left13:36
→ axl_ joined13:36
← d0k left13:37
cjdavies I have a directory that I want to be under git version control, which contains a variety of different things (some code projects, some essays/papers, etc.) but I want the individual code projects within this repo to also be under their own individual git control13:37
is this where submodules come in?13:37
hseg Hey guys. I'm getting weird behavior from git-rebase: https://gist.github.com/317622813:38
_ikke_ cjdavies: !subprojects13:38
canton7 !subrepos13:38
_ikke_ gitinfo is slow13:38
gitinfo cjdavies: So, you want to add git repositories inside of other git repositories? Well, you have four main options. First is to just do it, add the repo to the outer project's .gitignore, and treat them entirely separately. Best if they are entirely seperate. Otherwise your best options are "!submodule" "!gitslave" and "!subtree" Try typing those commands into this IRC channel.13:38
[!subprojects] So, you want to add git repositories inside of other git repositories? Well, you have four main options. First is to just do it, add the repo to the outer project's .gitignore, and treat them entirely separately. Best if they are entirely seperate. Otherwise your best options are "!submodule" "!gitslave" and "!subtree" Try typing those commands into this IRC channel.13:38
canton7 yeah13:38
hseg For example, git-commit gives me # new file: FMS_5772/MainForm.h while git-status doesn't say a thing about the file. (It does exist after git-reset, so probably git-status believes it comes from a previous commit13:39
← vlrk left13:39
→ alexukf joined13:40
→ billy_ran_away joined13:40
cjdavies thanks guys, will read up on which is best for me13:40
billy_ran_away If I add a collaborator to my repo should they be able to push to my repo with their own ssh key?13:40
On github that is...13:41
canton7 billy_ran_away, yes13:41
billy_ran_away canton7: Hmm… got a friend that can't seem to push with his key even though I added him as a collaborator...13:42
cjdavies !submodule13:42
gitinfo git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://book.git-scm.com/5_submodules.html13:42
canton7 billy_ran_away, can he push to his own repos?13:43
cjdavies !gitslave13:43
gitinfo gitslave (http://gitslave.sf.net) is useful to add subsidiary git repositories to a git superproject when you control and develop on the subprojects at more or less the same time as the superproject, and furthermore when you typically want to tag, branch, push, pull, etc. all repositories at the same time.13:43
billy_ran_away canton7: Yes, here's the error he's getting.13:43
cjdavies !subtree13:43
gitinfo The git subtree merge method is ideal to incorporate a subsidiary git repositories directly in to single git repository with "unified" git history, where you only need to pull changes in from external sources not contribute your own changes back (which if technically possible is at least difficult). See http://progit.org/book/ch6-7.html Type "!subtree_alt" for more options13:43
canton7 cjdavies, !msg13:43
gitinfo cjdavies: [!privmsg] You can get gitinfo keyword expansion (e.g. !doc or this one) sent directly to you by sending a private message to me: /msg gitinfo !doc. This avoids spamming the channel. Also see See http://jk.gs/git/bot/trigger.php13:43
billy_ran_away http://cl.ly/IK1U13:43
It seems like it's associating his key with one of his specific repos...13:43
canton7 "denied to jekhokie/test-unit-completion". what's his git remote -v output?13:44
jnewt ok, i'm reset13:44
→ reuf joined13:44
← alexukf left13:44
→ alexukf joined13:45
← Vooloo left13:45
→ smw joined13:45
jnewt now i added a file to the working directory, and ran git reset HEAD <file to remove> then git add <new file>13:45
→ juvenal joined13:45
→ Vooloo joined13:45
← Vooloo left13:45
→ Vooloo joined13:45
← Err404NotFound left13:46
jnewt i think i'm ready to commit, is there anything else i need to do first?13:46
FauxFaux jnewt: Run "git commit" with no aruments and it'll help you.13:46
jnewt it doesn't show those files as removed, but shows the new file13:47
i did a reset on them13:47
← cjdavies left13:47
← reuf left13:48
billy_ran_away canton7: Sorry missed your reply, it's the ssh one, same as mine, that is [email@hidden.address]13:49
→ pandeiro_ joined13:50
← enmand left13:50
← mankoff left13:50
canton7 billy_ran_away, does the string "jekhokie/test-unit-completion." appear anywhere in his .git/config ? Anything on ssh_config ?13:50
s/on/in13:50
anyways, I had to head I'm afraid. Good luck!13:50
← canton7 left13:50
billy_ran_away canton7: Figured it out after screen sharing with him, he had the ssh key associated with the specific repo and not his account.13:51
← alexukf left13:51
→ hellopat joined13:51
lb jnewt, git reset doesnt deetes files13:52
→ hellopat_ joined13:52
lb jnewt, and it doesn't affect your working dir as long as you dont specify --hard13:52
jnewt right, and i've reset to the sha of the very first commit, so they just won't be committed right?13:53
with --hard13:53
lb jnewt, if you want to delete a file from the git repo ( in the next commit) you'll use git rm FILE13:53
← RichardBronosky left13:53
← hellopat left13:53
jnewt no, i just want the commits of those files split up13:53
→ mermeladas joined13:54
jnewt it's like i need to reset --hard to before the first commit, or do what i'm doing, on a different branch and then rebase that initial commit on top of this one13:54
→ andrewhl joined13:54
→ vlrk joined13:55
jnewt because if i reset to the initial, i'm actually working on the "next"13:55
→ alexukf joined13:56
lb jnewt, i'm not really sure about what your doing ;) i would prefer a git reset --mixed and then commiting the right things (which will create a new path in the history)13:56
but it sounds a little bit logical ^^13:56
→ RichardBronosky joined13:57
lb jnewt, git reset --hard INITIAL and then merge the commits you want to have is also possible.13:58
you dont need to rebase13:58
← und3f left13:59
→ Err404NotFound joined13:59
jnewt crap, it went after the first commit as a new one14:00
→ cjdavies joined14:00
→ LekeFly joined14:00
← martialcoder left14:00
← iamjarvo left14:00
cjdavies basic question - if I have a git repo & I copy another git repo into the first git repo, then do git add ./git commit on the first repo, does it automatically ignore the second?14:01
this is what seems to be happenning by my experiments at least14:01
→ marhaban1 joined14:01
← marhaban1 left14:01
→ Superlokkus joined14:01
fluxdude is it possible to have 2 checkouts from single clone repo at the same time?14:02
→ cakehero joined14:02
fluxdude I have 2 branches, 1 prod and 1 staging14:02
and I want both of them accessible on the file system at the same time from the same git clone14:02
← EvanDotPro left14:02
fluxdude I could have 2 separate clones and then git checkout staging on the second one but it seems like a waste to have to clone the whole repo twice14:02
← dhruvasagar left14:03
← hseg left14:03
← soee left14:03
→ zenchaos joined14:04
→ chrisgeorge joined14:04
→ cccaldas joined14:07
jnewt i need to see the results of what i've done, i reset to the initial sha, added a file, committed, and now i only have 2 log messages, whereas i had 5-6 before. i need to see at least one more to verify the results of what i've done. how do i see those other commits so i can get a SHA to switch to?14:07
→ EvanDotPro joined14:09
← alexukf left14:09
→ alexukf joined14:09
joshc jnewt: you're question is a bit confusing. what do you mean by `see the results of what i've done', `initial commit' (the root commit? some commit your branch used to point to?), and `verify the results'14:10
← rbarraud left14:10
← pandeiro_ left14:11
← beautifulmind left14:11
joshc if I squint my eyes a bit, it sounds like what your looking for is the git-reflog, but that makes a few specific assumptions about what you're really asking :)14:11
→ kermit joined14:11
lb fluxdude, you cant14:11
FauxFaux !gka14:11
gitinfo For a better way to view the reflog, try: gka() { gitk --all $(git log -g --format="%h" -50) "$@"; }; gka14:11
← Davey left14:12
→ tmz joined14:12
fluxdude lb: shame14:13
← chrisgeorge left14:13
joshc AFAIK, git makes allows (or makes by default?) some optimizations for clones that exist on the same filesystem14:13
jnewt joshc: well, i had 4-5 commits, then i reset to the initial / root / first one ever commit, added a file, and committed. then i typed git log and i now only show 2 commits, where as i have actually made 5-6, but it seems they've disappeared. maybe the reflog is what i need, i remember reading about it, i'll take a look14:14
lb joshc, fluxdude AFAIK git clone on the local fs use hardlinks14:14
so fluxdude clone the repo already existing in your fs should not waste too much space14:15
→ d0k joined14:15
← tango_ left14:16
→ giallu joined14:17
jnewt i expected to see an unnamed branch or something i guess14:17
← adityamenon left14:18
→ adityamenon joined14:19
fluxdude lb: good thinking14:19
jnewt i only have 3 things in my reflog, a clone, reset and my commit14:19
→ chrisgeorge joined14:20
→ nowhere_man joined14:21
→ KYle111111111111 joined14:21
← KYle111111111111 left14:21
← cissu_ left14:22
Nirvanko How that's possible http://pastebin.ca/217459414:23
← jn_ left14:23
→ jjbohn joined14:23
Nirvanko A file that marked as to be commited and changed but not updated.14:24
jnewt well, they're still there in smartgit, so they still live on. i think i do need a rebase here14:24
← axl_ left14:25
haarg Nirvanko: when you use git add, it stages the current file contents to be committed. further edits to the file aren't included in that automatically.14:26
→ axl_ joined14:26
← theCole left14:28
→ tango_ joined14:28
← stkrzysiak left14:29
joshc jnewt: could you elaborate a bit on what you are actually trying to do? its not common to reset to the root commit (essentially abandoning your 5-6 commits worth of history) and to start over.14:30
← jsoa left14:30
← alberto56 left14:30
← cccaldas left14:31
jnewt i was trying to split the root commit into two commits, from adding two files to adding 1 file and modifying that file14:31
→ tplaner_ joined14:31
← alexukf left14:32
→ alexukf joined14:32
→ jsoa joined14:33
cjdavies I've read up a bit on submodules/slave/etc. & can't quite work out which (if any) matches what I want to do, though I may be overcomplicating the situation14:33
I want something a bit like this http://paste2.org/p/208267114:33
where when I add/commit global_project it will track changes in all of the subprojects14:33
→ fly9 joined14:33
cjdavies but where I can add/commit/push the individual subprojects14:34
Adys_ → Adys14:34
← Lord_DeathMatch left14:34
← tango_ left14:34
jnewt joshc: i understand it's not common, but once my testing is done, and i learn how to do some of this stuff (i've just been reading up to this point), i'm going to import an svn repository and attempt to make it's history something logical and not huge. there have been piles of generated files, pdfs and other garbage committed to the svn repo that i'll need to clean up14:35
and commits that we're way to much at once to even make sense.14:36
→ tango_ joined14:36
jnewt it just so happened i started with a root commit, i didn't realize that would cause an issue14:37
← ottoshmidt left14:38
← ThomasLocke left14:38
← alexukf left14:39
→ polymar joined14:39
→ v0n joined14:41
→ lresende joined14:42
→ dzonder joined14:42
joshc jnewt: you could probably use git commit --amend to do what you are trying to do14:42
← papegaaij left14:42
→ ku joined14:42
joshc use it to amend the root commit14:42
→ avpx joined14:43
Kronuz-away → Kronuz14:43
→ alberto56 joined14:43
← Err404NotFound left14:44
← etcetera left14:45
← billy_ran_away left14:45
→ hseg joined14:46
→ ehsan joined14:46
→ stkrzysiak joined14:46
→ t4nkd joined14:46
jnewt joshc i think i need to cherry-pick a range of revisions from origin/master and rebase onto my new commit14:46
hseg Hey. I'm experiencing weird behavior from git-rebase: https://gist.github.com/317622814:47
jnewt then prune origin/master14:47
← cagedwisdom left14:47
hseg For example, git-commit shows that FMS_5772/MainForm.h was created, but git-status does not14:47
rvsjoen anyone aware of any neat tools to generate some stats of a repo ? I tried gitstats but i'm wondering if there are others14:48
← lb left14:48
→ williamcotton_ joined14:48
→ camdenx joined14:48
→ xissburg joined14:49
joshc jnewt: yes. 'cherry-picking a range' is also very similar to git-rebase --onto14:49
xissburg how to remove a path from .gitignore? I removed it from the .gitignore file but I still cannot add the file that was ignored14:50
axrfnu xissburg: prepend the file to be unignored with a !14:50
xissburg o_ O14:50
jnewt joshc: that was my first though, i tried to rebase the "old second commit" onto the "new second commit", but it just moved that commit, the rest of origin/master didn't go with it14:51
s/though/thought14:51
axrfnu or remove it from your global gitignore, if it is set there14:51
xissburg axrfnu: I added to the local .gitignore14:52
wait14:52
← psoo left14:52
xissburg where is the global gitignore?14:52
axrfnu under unix/linux typically in $HOME14:52
← erichynds left14:52
← camdenx left14:52
← Bass10 left14:53
→ iamjarvo joined14:53
xissburg no it's not in the global gitignore14:53
← xiangfu left14:53
→ urbanmonk joined14:53
xissburg What's with the ! ?14:53
jnewt ok, the cherry picking range worked, but prune completed successfully, without pruning the branch. used git prune origin/master, no changes14:54
← skyf left14:54
axrfnu well, if you have, say, !*.o in your local .gitignore *.o files will be added to a repository, despite they've been ignored in a global ignore file14:55
xissburg oh14:55
axrfnu xissburg: see man gitinore14:55
err, man gitignore14:55
gitinfo the gitignore manpage is available at http://git-scm.com/docs/gitignore.html14:55
hseg Hey. I'm experiencing weird behavior from git-rebase: https://gist.github.com/317622814:55
→ Davey joined14:55
← ipalaus left14:56
→ SirCmpwn joined14:56
→ fireh joined14:56
xissburg axrfnu: thanks, it works C:14:56
fluxdude when cloning a repo it doesn't automatically track all the branches14:58
I can fix this on my local clone easily14:58
but when other ppl clone the repo they are also missing the branches14:58
FauxFaux fluxdude: They're not missing the branches. They ahve them. Use git branch -avv.14:58
fluxdude is there a way I can just get them to auto-track the branches without them having to fiddle their local clones?14:58
← bleen1 left14:58
FauxFaux Having all the branches as local branches is a bad thing which has no advantages at all, stop doing it.14:59
→ neil_ joined14:59
→ CannedCorn joined14:59
jnewt double crap, i tried to delete origin/master, and it did, now i just have headless branch that i can't prune. fsck me15:00
← thm left15:00
FauxFaux If you've messed it up, just delete and recreate the remote. It has no useful information in it.15:00
fluxdude FauxFaux: ah, I see. It does track the remote branch on first checkout15:00
when I created that branch I had to go around doing git branch --set-upstream staging origin/staging15:01
→ EricKamsky joined15:01
← cbreak-work left15:01
fluxdude since otherwise git checkout staging was failing15:01
but I'm not sure why now15:01
← LekeFly left15:02
← polymar left15:02
→ lekefly_ joined15:02
← EricKamsky_ left15:02
← flijten left15:02
→ skyf joined15:03
→ psoo joined15:03
→ airborn joined15:04
→ gioele joined15:04
→ axl__ joined15:05
jnewt ok, i managed to delete the origin/master from my local copy, but only the last commit in that branch. now it is not tagged. how can i delete the commits in that lineage?15:05
hseg Hey. I'm experiencing weird behavior from git-rebase: https://gist.github.com/3176228 - git-commit is displaying files as changed while git-status does not report them15:06
jnewt no, i guess i deleted it from both. i think15:06
← neil_ left15:06
FauxFaux jnewt: You are one very confused puppy.15:06
jnewt FauxFaux: and the more i read on git-scm, the worse it gets15:07
gioele hello, I would like to make the local `master` branch track the branch `extra` on the remote `server`, so I can avoid `git push server master:extra`.15:07
I tried `git branch --set-upstream master server/extra` but git complains with `error: Not tracking: ambiguous information for ref refs/remotes/server/extra`15:07
FauxFaux jnewt: You never, ever want to fiddle with anything named "origin/". If you have a remote based problem, delete all your remotes, re-add them, run git remote update, /then/ see what your problem is.15:08
← fisted left15:08
→ pandeiro joined15:08
→ weiguan joined15:08
FauxFaux gioele: What does ls -lR .git/refs/remotes # look like?15:08
thiago hseg: I don't see what the problem is15:09
← axl_ left15:09
axl__ → axl_15:09
hseg For example, git-commit shows that FMS_5772/MainForm.h was created, but git-status does not15:09
gioele FauxFaux: .git/refs/remotes/: server/ .git/refs/remotes/server: master extra15:09
→ chris__ joined15:10
jnewt FauxFaux: maybe i should start over15:10
thiago hseg: it's there15:10
chris__ i installed git with the xcode developer tools and can't move the git-credential-osxkeychain to /usr/loca/git/bin because it doesn't exist15:10
thiago # Untracked files:15:10
# FMS_5772/15:10
chris__ does anybody know where this folder is when installing git with the developer tools15:10
→ diegoviola joined15:11
→ fisted joined15:11
← Tegi_ left15:11
hseg So it means it's not tracking the directory and all files in it (yet):?15:11
← tchan1 left15:11
thiago chris__: run this: git --exec-path15:11
→ tchan joined15:11
thiago hseg: correct15:11
hseg Oh. That makes my job simpler. I was trying to commit all changes to that directory separately15:12
chris__ thiago thx a lot15:12
hseg thx thiago15:13
→ john_anderson joined15:14
→ OOPMan joined15:14
← e-ndy left15:15
← andrewhl left15:15
→ jbohn joined15:15
← Superlokkus left15:15
→ phishy joined15:16
hseg How do I stage a deletion?15:18
← shigekimorimoto left15:18
← QaDeS left15:18
→ cbreak-work joined15:18
hseg https://gist.github.com/317673915:19
→ beneggett joined15:19
→ Zenmaker__ joined15:19
← jjbohn left15:19
← cbreak-work left15:20
→ Harley joined15:20
← chris__ left15:20
← defts left15:20
crispus git rm <filename>15:20
→ sanguisdex1 joined15:20
→ defts joined15:20
← arkiver left15:20
← Harley left15:21
hseg thx15:21
→ ipalaus joined15:21
← Zenmaker_ left15:21
← sanguisdex1 left15:22
← felher left15:22
→ sanguisdex1 joined15:23
sanguisdex1 is there a quick way to get around the EOL errors when trying to add files made on nother platform?15:24
FauxFaux dos2unix the files.15:24
Or, if you're referring to some hook you have in place, man git commit --no-verify.15:25
gitinfo the git-commit manpage is available at http://git-scm.com/docs/git-commit.html15:25
← bgerber left15:26
← Milossh left15:27
← thiago left15:27
← phishy left15:27
→ phishy joined15:28
← williamcotton_ left15:30
→ bgerber joined15:30
→ felher joined15:30
jnewt what happens if you rebase and create a situation where the same named file was created twice?15:31
→ camdenx joined15:31
joshc you'll hit a conflict during the rebase15:32
MacGyver I'd assume that it simply creates a conflict within the file.15:32
← hseg left15:32
→ Zenmaker_ joined15:32
← fireh left15:33
→ arkiver joined15:33
← stkrzysiak left15:34
→ stkrzysiak joined15:34
→ tewecske joined15:35
← marhaban left15:36
→ hellopat joined15:36
→ Kaisyu joined15:36
← hellopat left15:36
← rtuin left15:37
← hellopat_ left15:37
← Zenmaker__ left15:37
→ hellopat joined15:37
→ dreiss joined15:38
jnewt ok, i've got two branches now, origin/master and master, but they are not "the same" or not "tracked" or whatever it's called, they're separate. how do i make them tracked / rewrite my new master history in to my origin history?15:38
→ erichynds joined15:39
shruggar you'll need to explain your situation a bit more, you're using some weird terminology so I'm not quite sure what you mean. Your situation could be absurdly easy to rectify, or it could just be absurd15:40
← beneggett left15:40
← tvw left15:40
jnewt i created a new branch, and committed it, then rebased master onto it. so that i could get a new initial commit. it worked. but the old line of commits is still hanging around. i want to get rid of it or overwrite it when i push.15:41
milki what do you mean by "old line of commits still hanging around"15:42
shruggar if you push to the same place, it will kill them. Problem solved?15:42
← cakehero left15:42
→ maletor joined15:42
→ beneggett joined15:42
milki unreferenced commits will go away with a git gc15:42
jnewt i mean my local copy, when viewed through any of my gui tools shows two branches, origin/master and master, and unlike normally when they are the same, they are completely unconnected now.15:43
← tsdh left15:43
shruggar when you push, they will be the same again15:43
jnewt ok15:43
→ etcetera joined15:43
jnewt i just didn't want to push a mess15:44
milki you dont push remote-tracking branches, only local ones15:44
← sedavand left15:45
→ thelazydeveloper joined15:46
← brentdanley_ left15:46
← thelazydeveloper left15:46
← diamonds left15:47
jnewt git push origin master15:47
→ andrewhl joined15:47
jnewt update were rejected because the tip of your current branch is behind it's remote counter part. Merge the remote changes before pushing again15:48
→ nikuyoshi joined15:48
← psoo left15:48
jnewt nothing has been changed on the remote end. how did this happen15:48
← nikuyoshi left15:49
← rchavik left15:49
→ cakehero joined15:49
jnewt not sure i want to pull, seems like it might either overwrite my commits, or try to do some merging, which i don't need/want15:49
joshc since you're !rewriting history, you'll need to force push15:50
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)15:50
jnewt ah, well, that makes sense15:51
→ thelazydeveloper joined15:52
← JohnBat26 left15:52
← jbohn left15:53
← sanguisdex1 left15:55
→ EdwardIII joined15:55
← nickpresta left15:55
EdwardIII hrm, topic is up to date, no changes have been made to master since it was started - should i just merge them?15:56
← hellopat left15:56
→ pppaul joined15:57
→ nickpresta joined15:57
← bullicon left15:57
→ systemclient joined15:59
systemclient is there a git way to push every repo into my projects/ folder to it's “github” remote? Or should I just use a bash-for-loop?15:59
→ jjbohn joined16:00
cjdavies I'm using duply/duplicity to do backups & I want to add/commit changes in git repos before the backup runs - the backup script runs as root, is it okay for root to do the add/commit even if the repos are owned by a normal user?16:01
← whitman left16:01
→ btanaka joined16:01
→ towski joined16:01
systemclient cjdavies: then normal users cannot use the repo any more16:01
cjdavies: “git gc“ will not work it root created any new files in there16:02
← arkiver left16:02
← dpino left16:02
← dougsland left16:02
← dreiss left16:02
cjdavies systemclient: right, thanks16:03
← stkrzysiak left16:03
cjdavies I guess I can just set a cron job as the normal user to add/commit before the cron job that runs the backup as root16:03
adhawkins → adhawkins-away16:04
← vervic left16:04
← skyf left16:06
← Davey left16:06
→ bwwrd joined16:06
→ kidsalsa joined16:06
← kidsalsa left16:07
→ stkrzysiak joined16:07
→ theswimmingfrog joined16:08
← pigeon left16:10
→ Davey joined16:10
→ Dragooon joined16:10
← fisted left16:11
← neurodrone left16:11
← tatsuya_o left16:11
← bubbles10101 left16:12
→ bubbles10101 joined16:13
→ thiago joined16:14
→ bullicon joined16:15
→ boombatower joined16:16
→ dinesh6654 joined16:17
→ jbohn joined16:17
← theswimmingfrog left16:18
→ wilmoore joined16:18
→ theswimmingfrog joined16:19
← theswimmingfrog left16:19
← jjbohn left16:20
→ theswimmingfrog joined16:20
← theswimmingfrog left16:21
← LimeBlast left16:21
→ theswimmingfrog joined16:21
→ rtjure joined16:21
← theswimmingfrog left16:22
← beneggett left16:22
← chris_dexter left16:23
→ theswimmingfrog joined16:24
← Kyzz left16:24
← theswimmingfrog left16:24
→ harish joined16:25
← _bart left16:25
→ theswimmingfrog joined16:25
← theswimmingfrog left16:25
← LmAt left16:25
→ Kyzz joined16:26
→ theswimmingfrog joined16:26
fluxdude I want to rename my master branch to prod, how will this affecf16:26
← theswimmingfrog left16:26
fluxdude affect the upsteam central repo we use and other users?16:26
I want this change to propagate rather than just my local branch name being difference16:26
different16:26
→ beneggett joined16:27
→ theswimmingfrog joined16:27
← theswimmingfrog left16:27
→ imachuchu joined16:27
→ theswimmingfrog joined16:28
← theswimmingfrog left16:28
← lonewulf` left16:29
← john_anderson left16:29
→ LeMike joined16:30
→ theswimmingfrog joined16:31
← lresende left16:31
→ john_anderson joined16:31
← theswimmingfrog left16:31
→ goshawk joined16:32
→ theswimmingfrog joined16:32
→ tvw joined16:32
← theswimmingfrog left16:32
→ enmand joined16:32
← jargon- left16:33
→ eletuchy joined16:33
← TheJH left16:33
← eletuchy left16:33
→ nadirvardar joined16:34
→ theswimmingfrog joined16:34
→ apok joined16:35
← theswimmingfrog left16:35
rking (Asking for a buddy, so I can't test it on my own easily): If he's working on a feature branch and does a "git pull", he can then simply do "git merge master" to get the updated remote's master's HEAD, right?16:36
→ theswimmingfrog joined16:36
← bosspetta left16:36
← theswimmingfrog left16:37
selckin pull = fetch+merge16:37
→ hobodave joined16:37
← bubbles10101 left16:37
selckin (sometimes fetch+rebase (configuration etc))16:38
→ theswimmingfrog joined16:38
← theswimmingfrog left16:38
rking selckin: Right, but that's merging only the currently selected branch, no?16:38
selckin sure16:39
imachuchu rking: yeah, he'll get the remote master's head through fetch/pull (really that's it's job, to pull down a branch (by default master) and any commits that you don't already have)16:39
→ theCole joined16:39
fluxdude on my git server I did git branch -m master prod16:39
selckin so you'll need origin/master, since master is not merged or rebased16:39
fluxdude but on the clients I still see refs to master, how do I get rid of them?16:39
→ lresende joined16:39
selckin fluxdude: don't think you can16:40
→ theswimmingfrog joined16:40
rking OK. Because the only place where you could possibly have non-committed changes is the working tree, so that's the only one where the mergewould make sense to not update HEAD all the way to the remote's HEAD. (right?)16:40
selckin you can change the default "master" branch tho, for new clones16:40
rking selckin: I still really don't understand rebase. I've not manually used it even once. =\16:40
← theswimmingfrog left16:40
selckin rking: just another way to merge basicly16:40
rking: but no, if you type "git merge master" you won't get the freshly fetched stuff, those are in origin/master16:41
→ theswimmingfrog joined16:41
← mishok13 left16:41
→ artnez joined16:41
→ dpino joined16:41
← theswimmingfrog left16:41
→ kukks joined16:41
← kukks left16:41
→ kukks joined16:41
→ theswimmingfrog joined16:42
← theswimmingfrog left16:42
→ nhooey|mini joined16:42
imachuchu umm, I wouldn't say that rebase is akin to merge, really at all. It's a way of grafting one subtree onto another (alternatively it reapplies the changes introduced by patches, in order, at a specified point)16:43
→ lonewulf` joined16:44
← caddoo left16:44
→ theswimmingfrog joined16:44
← OOPMan left16:45
← theswimmingfrog left16:45
selckin they both make the history of multiple tree visible from one of them16:45
FauxFaux It's akin to merge in that it's a way to automaticially reconcile divergent histories.16:45
(semi-automatically)16:45
← CaptainQuirk left16:46
→ theswimmingfrog joined16:46
→ Goplat joined16:47
← theswimmingfrog left16:47
imachuchu FauxFaux: true16:47
→ theswimmingfrog joined16:47
→ alvaro_o joined16:47
FauxFaux Does the documentation actually use a consistent term for a "history" somewhere? i.e. a branch of history, but not necessarily a git branch.16:47
← theswimmingfrog left16:48
→ neurodrone joined16:48
← alvaro_o left16:48
→ theswimmingfrog joined16:48
← theswimmingfrog left16:49
→ dreiss joined16:49
← jnewt left16:49
→ jnewt joined16:49
→ fireh joined16:49
selckin plumming commands use tree no?16:50
→ hellopat joined16:50
→ mwrosen joined16:50
→ theswimmingfrog joined16:51
← Nirvanko left16:51
→ cmullican joined16:51
→ sedavand joined16:51
← Kaisyu left16:51
← theswimmingfrog left16:51
FauxFaux selckin: Tree is an entirely different thing; it's a list of files, instead of a list of commits.16:52
← iamjarvo left16:53
← ipalaus left16:53
→ theswimmingfrog joined16:53
→ pranjal710 joined16:53
selckin hmm right16:53
→ dangerboy joined16:53
→ adityamenon90 joined16:54
← theswimmingfrog left16:54
← tango_ left16:54
→ ipalaus joined16:54
← pandeiro left16:54
← teotwaki left16:55
← adityamenon left16:55
→ tango_ joined16:56
→ tatsuya_o joined16:56
← beneggett left16:56
← rns-laptop left16:56
→ skyf joined16:56
← btanaka left16:57
→ horar79 joined16:58
→ Learath2 joined16:58
→ Speed joined16:59
← Speed left16:59
→ Speed joined16:59
→ apok_ joined16:59
← lresende left17:00
selckin all the manuals i checked in boredom assume you already understand the word branch17:00
→ raykyri joined17:00
gitinfo set mode: +v17:00
← ignas left17:00
← gioele left17:00
← shruggar left17:00
FauxFaux Which means two different things.17:00
→ beneggett joined17:01
← babilen left17:01
← lonewulf` left17:03
← apok left17:03
apok_ → apok17:03
← tchan left17:04
→ eletuchy joined17:04
→ vervic joined17:05
talexb Has anyone tried changing the diff options in gitweb? I'd like to add '-b' to get around checkins that merely chang line-endings.17:06
→ kevwil joined17:06
← kevwil left17:06
← airborn left17:06
talexb I know where to change it, I'm just curious to know if anyone has good/bad stories about it.17:06
→ rns-laptop joined17:08
→ mrwalker joined17:08
← bullicon left17:08
← horar79 left17:10
→ horar79 joined17:10
← systemclient left17:11
talexb Patch worked fine. \o/17:12
← apok left17:12
→ apok joined17:12
Marcux83|away → Marcux8317:13
→ kenperkins joined17:13
mocramis hello.17:15
is there a way to find out why a patch is corpted (or to get more details ?17:15
← skyf left17:16
← dangerboy left17:16
→ dhruvasagar joined17:17
talexb mocramis, Corrupted? Is your hard drive having problems?17:17
→ pigeon joined17:19
→ jargon- joined17:19
mocramis no, i did edi the patch a little bit, but i just found out that --recount solved my problems17:20
→ airborn joined17:20
← t4nkd left17:20
← horar79 left17:21
← giallu left17:21
→ tchan joined17:22
← tchan left17:22
→ tchan joined17:22
→ shavus joined17:23
→ lresende joined17:24
← warthog9 left17:24
← jceb_ left17:25
← jason237 left17:25
← wilmoore left17:26
→ wilmoore joined17:27
← john_anderson left17:28
← ReekenX left17:28
← weiguan left17:29
← axl_ left17:29
→ john_anderson joined17:29
← wilmoore left17:30
← rns-laptop left17:30
→ darren_ joined17:31
→ abstrusenick joined17:31
→ skyf joined17:32
→ ISF joined17:33
→ _huoxito joined17:34
← rehat left17:35
→ martin__ joined17:36
← Stalafin left17:36
Mpenzi → mpenz17:36
mpenz → Mpenzi17:36
← pppaul left17:36
→ pppaul joined17:37
→ axl_ joined17:37
← martin__ left17:38
→ martin__ joined17:38
→ shruggar joined17:38
← martin__ left17:39
→ mklappstuhl joined17:39
→ lonewulf` joined17:39
mklappstuhl hey17:39
I'm looking for a way to measure "developer contribution" starting from a given commit... any ideas how I could do that?17:40
→ ph^ joined17:41
← ctaloi left17:41
→ maestrojed joined17:41
selckin define contribution17:42
→ Bass10 joined17:42
→ ctaloi joined17:42
→ theswimmingfrog joined17:43
→ andrew_knowles joined17:43
maestrojed Is there such thing as adding an exception in gitignore? I want to track files/foo/ but block everything else in files/. The structure of files/ is somewhat dynamic so its not ideal to block each subdirectory individually (files/bar/, files/example/, etc.)17:43
mklappstuhl selckin, something like "# of changed lines/characters"17:44
← theswimmingfrog left17:44
selckin maestrojed: anything tracked is not ignored17:44
← shruggar left17:44
→ theswimmingfrog joined17:44
mklappstuhl selckin, Problem with changed lines will be that we probably hat a lot of whitespace fixes which will add up when checking for changed lines17:44
← goshawk left17:44
← theswimmingfrog left17:45
selckin counting lines is pritty uselss for pritty much anything you want to do17:45
→ EricKamsky_ joined17:46
← EricKamsky left17:46
← sirdancealot7 left17:46
mklappstuhl selckin, I'm open to any other suggestions17:46
→ arkiver joined17:46
mklappstuhl I'd just like to have some measurement for "done work" for myself17:47
← tvw left17:47
mklappstuhl (I don't want to judge anybody based on that or similar17:47
← arkiver left17:47
← dinesh6654 left17:48
→ TheShagg joined17:48
→ goshawk joined17:48
selckin google gitstats i guess17:48
gitinfo set mode: +v17:48
→ dinesh6654 joined17:48
→ arkiver joined17:48
TheShagg I setup a remote tracking branch, checked it out, and fetched, and I didn't get the most recent changes as far as I can tell17:48
→ iamjarvo joined17:48
TheShagg anyone understand how this is supposed to work?17:48
selckin what are the exact commands you ran17:49
iamjarvo say i merge in changes before doing a pull from a remote branch. and github hangs up. is it wise to do git pull?17:49
→ n1x joined17:50
TheShagg selckin: git branch --track central_master central/master; git checkout central_master; git fetch central; gitk -1017:50
→ aantix joined17:50
← raykyri left17:50
← defts left17:50
selckin iamjarvo: sure, if you want to avoid the extra merge maybe pull --rebase17:50
← SirCmpwn_ left17:50
→ raykyri joined17:50
selckin TheShagg: git fetch will will update central/* only, you need merge or rebase (pull = fetch+merge)17:51
iamjarvo selckin thanks17:51
← clifton left17:51
← n1x left17:52
adhawkins-away → adhawkins17:52
→ shruggar joined17:52
TheShagg selckin: so what are my options then, can I just use central/master like any other branch? I am guessing that if my local copy does not have any changes that the merge will more or less get what I am expecting?17:52
selckin: let me clarify17:52
selckin: can i just merge with central/master as if it were any other branch?17:53
selckin TheShagg: yes17:53
maestrojed selckin so you are saying track what I want, then add the parent folder to gitignore? Should work. Thx17:53
selckin maestrojed: yeah, just ignore the folder, and explicity 'git add' stuff you want git to know about17:54
→ andrew_k1owles joined17:54
← ctaloi left17:54
maestrojed selckin ok thanks17:54
mklappstuhl selckin, Gitstats unfortunately provides only stats based on # of commits and this is not saying a lot when people often do 1-line commits17:55
→ surripere joined17:55
← Vampire0 left17:55
selckin lines don't say anything either17:55
talexb This sounds like a PHB statistic to me ..17:55
selckin but one of the stats one does it i'm sure, they are multiple with the same name17:55
← bgerber left17:55
← angelsl left17:55
talexb A ten line change may take ten minutes and little thought.17:56
A one line change might require tons of research and days of testing.17:56
→ berserkr joined17:56
← PerlJam left17:56
→ surriper1 joined17:56
← achadwick left17:56
selckin turning knob 5$, knowing what knob to turn priceless17:56
talexb It's probably better to look at the estimate of the complexity of the originating issues ..17:57
→ Vampire0 joined17:57
EugeneKay My experience is that reducing SLOC is a better program practice than increasing.17:57
maestrojed if your site has a blog, would you track the images added to blog posts?17:57
talexb Nicely put.17:57
selckin images good for reddit and driving traffic!17:57
EugeneKay +ming17:57
→ PerlJam joined17:57
→ tjadc joined17:57
→ bgerber joined17:58
→ fusion27 joined17:58
talexb And a high number of commits could indicate that the developer made a change that included a change, fixed the change and introduced another error, fixed the error and documented, then fixed the documentation ..17:58
Rhetorical: Is that better or worse then someone who Got It Right The First Time?17:58
selckin or fixing up the errors of everyone else17:58
should really look way higher level17:59
fusion27 Alright a couple of Git noobies here17:59
← abstrusenick left17:59
fusion27 We're working remotely on a Git repository hosted on an Ubuntu 10.04 box17:59
talexb Isn't everyone? ;)17:59
TheShagg selchkin; so lets say I am on master, and I do a git fetch, if I want to manually fastforward would I do something like "git rebase central/master" or "git merge central/master", is this the expected way to work if you fetch instead of pull?17:59
fusion27 Developer A is working on the master, Developer B cloned the master and branched it18:00
selckin TheShagg: those will do the same thing, unless you have made commits yourself18:00
talexb TheShagg, Why wouldn't you just git rebase master?18:00
→ BD- joined18:00
selckin TheShagg: ie a fast-forward merge18:00
talexb fusion27, Why is developer A working on master?18:00
fusion27 We want to merge Developer B's branch in with Developer A's master18:00
→ werdnativ joined18:01
→ HAL_9001 joined18:01
← HAL_9001 left18:01
→ HAL_9001 joined18:01
fusion27 I think it has more to do with being a noobie than making a deliberate choice to do so18:01
TheShagg selckin: I understand, I just have never used a remote/branch pair by name in my commit merging process - I've been doing it all on my local machine18:01
talexb fusion27, So Dev A pushes, Dev B pulls and does the necessary merging .. and pushes the result.18:01
Kronuz → Kronuz-away18:01
TheShagg talexb: wouldn't doing git rebase master pluck master and rebase it onto master (i.e. do nothing)18:01
← adityamenon90 left18:01
← ph^ left18:01
talexb fusion27, Typically no one actually does any work on master. You always create a separate branch, complete your work there, then merge back into master.18:02
← shruggar left18:02
selckin TheShagg: well it looks right to me18:02
fusion27 Thank you18:02
→ adityamenon joined18:02
← arkiver left18:02
← Samual left18:02
← axl_ left18:02
TheShagg so, basically you use master as your remote tracker? and you git pull it from the remote and you never have to merge it, and then you do all the merging in your local branch, and then push it up to the remote?18:02
→ axl_ joined18:03
talexb TheShagg, If you're on master, you may be out of step with the server .. then you just do a pull, and you're up to date.18:03
→ bradrice joined18:03
selckin TheShagg: well if you just did pull, it would fetch and merge from the remote branch it is tracking18:03
talexb You may have mis-understood what 'fast-forward' means.18:03
← bradrice left18:03
TheShagg selckin: I hate using pull, i would rather peek around first before merging18:03
selckin TheShagg: if you do it manually, it doesn't really mather you are tracking, except for when you push18:04
← fusion27 left18:04
← juvenal left18:04
→ bradrice joined18:04
TheShagg but my understanding is you keep master "virgin" so you don't have to do any merging when you pull, and then you can do all your rebasing/merging in your topic branch using master as your source for remote commits18:04
does this sound correct?18:04
talexb If you pull from master, do a bunch of commits, then push back to origin, that counts as a fast-forward merge. Although I prefer a non-ff merge, since it lumps the commits together into a recognizable blob.18:04
→ fusion27 joined18:04
selckin sounds like more work, depends on how you like to work i guess18:04
← resmo left18:04
fusion27 TheShagg: were you just helping me? My chat client wasn't showing the ID of respondents18:05
I just reset and can see now18:05
selckin like the linux kernel has a policy of only merging to master18:05
TheShagg fusion27: i think a few people are talking about the same thing, but I wasn't meaning to ;)18:05
selckin: as opposed to fast forwarding?18:06
← tatsuya_o left18:06
fusion27 thanks TheShagg18:07
TheShagg selckin: my main problem is that I do not like to pull, I like to fetch, and that is where my confusion arises18:07
selckin TheShagg: like some people like to keep mostly linear histry and rebase branches on top of master, or if you're a small team, other people prefer to like always work in branches, and then merge them in master, some ppl just git pull randomly and create a clusterfuck of a history and like that18:07
bradrice I cloned a site, did a branch, but now I can't push using git push branchname18:07
→ Synthead joined18:07
selckin TheShagg: yeah that sounds like a good thing, not sure what part is confusing you18:08
TheShagg selckin: but I think I have an idea of how this works now, a workflow that will work for me18:08
fusion27 talexb: thanks for your help, we're taking a swing right now18:08
→ juvenal joined18:08
TheShagg selckin: my original confusion was that I couldn't figure out how to get the fetched commits, because my branch was unchanged after fetch, but it looks like I can get them by referencing remote/branch or creating another branch that I can just pull (master could work for this), either would work I think18:09
fusion27 talexb: I'm Dev A, when you say push, are you saying to push to Gitbub?18:09
← grandy left18:09
fusion27 I think the box I'm working on is properly set up as a Git server18:09
and I'm (Dev A) working on the master right on that box18:10
→ axl__ joined18:10
selckin TheShagg: by default it's named origin, so you clone a repo, and have origin/master which is the state of the branch on the remote (update with fetch), then your own master branch is stuff you want to end up on the remote master (origin/master) when you push, and the remote had changes you didn't see yet, you have to first merge or rebase into your own master, before you can push your new stuff out18:11
TheShagg: if that makes sense18:11
EugeneKay !workflow18:11
gitinfo Finding the right workflow for you is critical for the success of any SCM project. Git is very flexible with respect to workflow. See http://sethrobertson.github.com/GitBestPractices/#workflow for a list of references about choosing branching and distributed workflows.18:11
EugeneKay Grrr.... that's not the one.18:11
← skyf left18:11
EugeneKay !isgitbetter18:11
gitinfo We think so, but if you don't, please go away and use $YourFavoriteSystem instead. For a comparison to other VCSes, see http://thkoch2001.github.com/whygitisbetter/18:11
xissburg :)18:11
EugeneKay fusion27 - See the workflow section of that page for some diagrams which might help bring sense to your situation18:12
xissburg git is the world standard18:12
fusion27 thanks EugeneKay18:12
→ resmo__ joined18:12
← axl_ left18:12
axl__ → axl_18:12
TheShagg selckin: I get all that, I was just confused on how to get those commits from the remote to do a local merge, (assuming a fetch, not pull), and the answer is "origin/master" - I just do a git merge origin/master if I am happy with what I see, or I can do some rearranging on my side or whatnot18:12
← beneggett left18:13
selckin TheShagg: cool18:13
TheShagg: git branch -a18:13
TheShagg selckin: although it sounds like people frown on using master as a place to work18:13
all depending18:13
→ shruggar joined18:14
TheShagg but I can always do a git push HEAD:origin/master if I am on a topic branch18:14
selckin yeah, new people to git tend to use git pull a lot and create a mess, thinking its the same to svn update, but that's more like a rebase18:14
TheShagg I am using gerrit code review, which has some built in merging crap that makes things more strange, so I am trying to get a lot of commits in sync with it18:15
rebasing and such18:15
→ beneggett joined18:15
EugeneKay .trigger_edit flexible git's status as a Distributed VCS really means that it is flexible - you can build any layout of repos you like to suit your !workflow. Here are some diagrams of common ones: http://git-scm.com/about#distributed18:15
gitinfo EugeneKay: Okay.18:15
EugeneKay .trigger_edit layout @!flexible18:15
gitinfo EugeneKay: Okay.18:15
← Bass10 left18:16
← resmo_ left18:16
← tchan left18:16
→ jeffisabelle joined18:16
selckin EugeneKay: link is wrong s,#,/,18:17
EugeneKay No it isn't?18:17
selckin unless you didn't intent to link disrubuted18:17
← lresende left18:18
EugeneKay Oh, lovely. git-scm has one of those "smart" interfaces that relies upon JS18:18
← phishy left18:18
EugeneKay The # works with JS off, but the / with it on -_-18:18
→ phishy joined18:18
gitinfo set mode: +v18:19
selckin oh it does those # links in chrome, but / in firefox18:19
← meLon left18:19
selckin only / works from a fresh link for me, bad site18:19
EugeneKay This is where I begin to scream at the html5 nuts18:19
→ skyf joined18:19
selckin yeah thats pritty bad18:19
TheShagg scream in a bad way?18:19
← vervic left18:20
EugeneKay Murderously.18:20
TheShagg dunno anything about html5 except it does "flash" like things ;)18:20
EugeneKay The problem is people forget to make it degrade gracefully18:20
→ tatsuya_o joined18:20
EugeneKay I browse with NoScript because I beleive that if your website needs to eat 100MB of my RAM to show me a fancy flash animation then I probably don't care what you have to say18:21
← dv310p3r left18:21
selckin i wouldn't go that far18:21
→ dfr|moscow joined18:21
selckin but getting urls right is the minimal you can do18:21
can't really do much useful without js these days18:21
EugeneKay Sure you can.18:21
← BD- left18:22
EugeneKay It's called STOP BOUNCING AROUND LIKE A CRACKED UP CHIHUAHUA18:22
selckin well i 'm glad i have gmail and not 1999 horde18:22
← beneggett left18:23
selckin if its for "websites" where you like go to read news and documentation its all good without js, but if you need interaction, ie app stuff, like for configuration & shit, you really need js18:23
→ marhaban joined18:24
← ipalaus left18:24
→ dv310p3r joined18:25
→ BD- joined18:25
selckin ie here, git-scm would be fine without it, maybe bit slower to swap between the pages #care18:25
← tatsuya_o left18:25
← juvenal left18:26
← jeffisabelle left18:26
← raykyri left18:27
← BD- left18:27
→ BD- joined18:27
→ juvenal joined18:29
→ williamcotton_ joined18:29
← shruggar left18:29
← williamcotton left18:30
williamcotton_ → williamcotton18:30
→ ShooterMG joined18:30
→ lresende joined18:31
← manolo left18:32
→ ipalaus joined18:32
→ manolo joined18:32
→ clifton joined18:32
→ tchan joined18:33
← tchan left18:33
→ tchan joined18:33
← Dragooon left18:34
→ raykyri joined18:36
→ Xizor joined18:36
→ devth joined18:36
TheShagg LOL, is there a way to rebase multiple branches at once18:36
→ scientes joined18:37
← clifton left18:37
FauxFaux Nope.18:37
TheShagg awww, i have like 10 commits with common ancestry that need to rebase far back18:38
10 branches18:38
i mean18:38
→ p3rror joined18:38
← bgerber left18:38
→ rehat joined18:38
cbreak TheShagg: that'd make no sense at all18:38
SirFunk i'm getting this problem "error: Your local changes to the following files would be overwritten by checkout: config/application.yml Please, commit your changes or stash them before you can switch branches." but when i git stash it says "no local changes to save"18:38
TheShagg cbreak: why not?18:38
cbreak because rebase rebases the current branch18:39
only one current branch18:39
only one working directory, which rebase needs to resolve conflicts18:39
FauxFaux SirFunk: Is that a tracked file? If not, then you need to deal with it in other ways.18:39
TheShagg cbreak: it could be done iteratively18:39
SirFunk FauxFaux: it was tracked then i assume unchanged it18:39
TheShagg which is what I am going to have to do by hand18:39
→ fisted joined18:39
→ bgerber joined18:40
→ SirCmpwn_ joined18:40
← SirCmpwn_ left18:40
FauxFaux SirFunk: If you assume-unchanged, all bets are off, and anyone's ability to help you goes down the drain, especially if you don't tell us. Have fun fixing the problem yourself.18:40
SirFunk i just told you :-P i can un-assume-unchanged it but i'm not sure if that will get the desired results. I need to the file to have a certain version in the repo (blank config file) and another version locally (for developing my app)18:41
→ sq-one joined18:41
selckin SirFunk: welcome to windows and line endings18:42
→ tqmz___ joined18:42
← eicto left18:42
SirFunk selckin: huh? I'm on linux18:42
cbreak SirFunk: delete the file.18:42
selckin then i dunno18:42
oh yeah the file isn't tracked and was added18:42
cbreak that'll make the changes gone18:42
SirFunk cbreak: how does that make it exist in both the repo and locally?18:43
cbreak of course it's tracked18:43
← doorp left18:43
cbreak SirFunk: what?18:43
you have a file with changes18:43
git was told to assume there are no changes18:43
what do you want git to do?18:44
SirFunk: the sane solution would be to just kill the local changes since they are obviously worthless18:44
since they are not even committed18:44
SirFunk I want a copy in my git repo (that has default configs) and then i want git to ignore it after the initial commit18:44
cbreak sounds stupid.18:44
!config18:44
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.18:44
→ SirCmpwn_ joined18:46
→ towski_ joined18:47
← dfr|moscow left18:47
→ freimatz joined18:48
gitinfo set mode: +v18:48
← cjdavies left18:48
→ eicto joined18:48
→ Schwern joined18:48
← towski left18:50
freimatz Hi, with git status i get wrong file i.E. # Untracked files:/# "BMP/Symbole/Aktuell/A_Gr\303\266\303\237er.bmp" - the files containes german umlauts18:50
→ e-ndy joined18:50
← SirCmpwn_ left18:50
cbreak how's that wrong?18:50
thiago looks like UTF-8 to me18:50
freimatz hm, i expted18:51
→ Chillance joined18:51
→ schultza joined18:51
← dv310p3r left18:51
thiago that's not a verb18:51
freimatz BMP/Symbole/Aktuell/A_Größer.bmp18:51
← towski_ left18:52
freimatz (Windows 7)18:52
what can i do?18:53
selckin working as intented18:53
thiago what's the problem?18:53
→ beneggett joined18:53
← adityamenon left18:54
← _huoxito left18:54
schultza warning: remote HEAD refers to nonexistent ref, unable to checkout... im doing a local bare clone of a git server (which apparently is a clone of a svn server) how do i fix this?18:54
fusion27 anyone ever seen this one before? http://pastebin.com/RaeLPbJT 2 noobie git developers were able to generate this. Dev A started a project all on his own, never understanding that anyone else would ever join him. Committed all of his work to Git, working in a local repo, committing regularly, but only on the master branch. Dev A got a different job, then was re-hired by former employer to continue project. Dev B joined in. He cloned Dev A's re18:54
thiago schultza: check out the branch that you want18:55
selckin !faq nonbare18:55
fusion27 how's that look bradrice18:55
?18:55
selckin !faq non-bare18:55
gitinfo Pushing to non-bare repositories is discouraged; more details available at http://jk.gs/gitfaq#non-bare18:55
← dpino left18:55
bradrice ok18:55
schultza it's a bare copy ... do i still have to run the checkout command?18:55
bradrice I don't quite understand why I couldn't just push my branch18:56
← BD- left18:56
bradrice but it is merged in master now18:56
→ BD- joined18:56
← rtjure left18:56
← dys left18:56
schultza git checkout trunk # returns # fatal: this operation must be run in a work tree18:56
→ jeffisabelle joined18:57
→ towski joined18:57
adhawkins schultza: You can't checkout into a bare repository18:57
You need to clone it18:57
schultza yes... and im having that other error cloning from it18:58
thiago schultza: I meant in the repository you've just cloned18:58
schultza did i not set up the bare clone right?18:58
thiago schultza: if you want to change the branch that the bare repository points to, you need to use git symbolic-ref to change HEAD18:58
fusion27 should we do a git reset —hard here?http://pastebin.com/RaeLPbJT18:58
→ AdrienBrault joined18:58
← enmand left18:59
→ Shusshu joined18:59
← BD- left18:59
mklappstuhl EugeneKay, talexb: I'm completely on your site regarding the statistics but knowing the project I'd like to have them anyway :) Actually the situation is that our CTO (only devs are him and me) apparently stopped working "real" on the project. He is only doing css stuff for weeks and when there is some "real work" he just sends me an email with the error. He is doing commits like changing one number in a css file (which is generally ok but h18:59
ard to measure). I would just like to know if I'm just thinking that or if its like that for real...18:59
selckin fusion27: http://jk.gs/gitfaq#non-bare18:59
← Shusshu left18:59
→ pdtpatrick joined18:59
rking I've got a (git-n00b) coworker that said he ended up with like 5 siblings of a repo file, something about foo.LOCAL.ext and foo.REMOTE.ext ? What might've happened?18:59
bremner some mergetool left those19:00
← tango_ left19:00
EugeneKay mklappstuhl - beat the boss. I recommend an aluminum bat19:00
← ameoba left19:00
mklappstuhl EugeneKay, I'd like to know if I'm right with my thoughts before doing so19:01
→ tango_ joined19:01
selckin mklappstuhl: xyproblem, git log -p --author=boss | diffstat19:01
← beneggett left19:01
→ BD- joined19:01
→ enmand joined19:01
→ dys joined19:01
← kleist left19:02
→ captainfurious joined19:02
→ sungnyc joined19:02
→ _bart joined19:02
→ elux joined19:02
mklappstuhl selckin, whats xyproblem? ;)19:03
selckin google it19:03
mklappstuhl selckin, what if he is commiting from different machines with 5 different users? :D (git noob)19:04
→ dv310p3r joined19:04
mklappstuhl (he actually didnt use version control until I joined him)19:04
(because he's been working alone, you know...)19:04
fusion27 so selckin we're confident of what both developers have done19:04
adhawkins !xyproblem19:05
fusion27 If I correctly understand the documentation, we're calling for git reset --hard19:05
?19:05
← nielsbusch left19:05
→ beneggett joined19:05
selckin mklappstuhl: add more --authors19:05
fusion27 the pusher did a git pull from the master, then merges his branch in to it19:05
→ wilmoore joined19:05
mklappstuhl selckin, Oh cool, didnt know there is a name for that. Felt back for blaming someone so I started with Y19:05
fusion27 merged19:05
→ TheJH joined19:06
selckin fusion27: isn't that output from a push?19:06
fusion27 selckin: yes it is19:06
← skyf left19:06
selckin fusion27: then read the link i gave you twice19:06
← d0k left19:07
selckin or say you didn't understand what that has to do with it, but just repeating the same question is kinda futile19:07
← captainfurious left19:07
→ d0k joined19:07
schultza i got it working... i cant just run git clone on a bare init... had to add remote and fetch...19:08
darn weekend getting in the way.. .:)19:08
talexb mklappstuhl, XY Problem: http://www.perlmonks.org/?node_id=54234119:08
selckin you need at least 1 commit i think19:08
adhawkins schultza: clone creates the repository for you, you don't need to init it19:08
schultza clone also allows a working area... i was only setting up a bare repos for a server over ssh19:09
talexb mklappstuhl, Also, http://perl.plover.com/Questions3.html19:09
mklappstuhl talexb, found that alread19:10
freimatz yeah: commit -a was the solution for me19:10
selckin talexb: mklappstuhl i like this one http://mywiki.wooledge.org/XyProblem19:10
adhawkins → adhawkins-away19:11
← beneggett left19:11
schultza now im getting it again...19:12
← dhruvasagar left19:12
← fisted left19:12
→ skyf joined19:12
← jeffisabelle left19:12
→ fisted_ joined19:12
→ clifton joined19:12
→ psoo joined19:12
→ tatsuya_o joined19:12
→ ameoba joined19:13
schultza i have a server running over ssh... trying to clone for work area on local machine.. have server on 2... local on 3... git clone ssh://git@2/dir/project.git responds with warning: remote head refers to nonexistent ref, uanble to checkout19:13
selckin default branch doesn't exist maybe19:13
mklappstuhl selckin, second example is great19:13
schultza again... the top originating server for this project is svn...19:14
master on svn is "trunk"19:14
selckin "git svn" repos should not be shared19:14
→ Bass10 joined19:15
← eicto left19:15
selckin unless you know what you're doing (and even then)19:15
schultza im not planning on sending the data back up the ranks... but keeping company devopments inside19:15
be back...19:16
selckin then just init a bare repo ont he server, add it as a remote in the git svn repo and push it up19:16
cmullican I'm sure this is simple, but I'm relatively new to git. I accidentally made some changes on a project branch instead on on the main branch, The project is not ready for prime time. I need to get the last three commits from that branch onto master, but not any of the previous commits.19:16
selckin git checkout main (master?); git cherry-pick sha-1 sha-2 sha-319:17
← rehat left19:17
selckin go back to branch, and remove the commits (git reset --hard sha-before-the-3) (warning will remove all uncommited changes from workdir)19:18
← theCole left19:19
← etcetera left19:19
← palm_vue left19:20
← bizhanMona left19:20
→ eicto joined19:20
imachuchu take the best of both worlds, create a new branch off of master, cherry-pick the commits you want to it, then merge it into master. You still have your old branch and you can muck around with the commits in your new merge branch19:20
← Wooga left19:20
selckin sounds harder for someone new to git19:21
← goshawk left19:21
→ bizhanMona joined19:21
→ palm_vue joined19:21
→ nielsbusch joined19:21
← stkrzysiak left19:21
→ nadavoid joined19:22
← nadavoid left19:22
← tqmz___ left19:23
schultza doesnt git push upload back to the server rather than download?19:23
selckin yes,19:24
→ jason237 joined19:24
schultza why would i push a bare init up to a remote server? (it has nothing)19:24
→ na3r joined19:25
← shavus left19:25
← hyperair left19:26
selckin how did you make the repo on the server19:26
→ goshawk joined19:26
→ moltar_net joined19:27
→ hyperair joined19:28
← p3rror left19:28
← camdenx left19:28
moltar_net Hi! I want to push changes to the origin, and someone has previously already commited to the branch on the origin. When I do "git pull" to sync, it creates a merge commit message. I don't think this is right, is it? Should I be doing rebase instead? When I tried rebase, it tried to rebase everything from a long time ago for some reason. Maybe I have some weird deviation from the origin or something? What is the appropriate course of action here? Thanks!19:29
schultza git --bare init ; git remote add origin git:[apache]/ofbiz.git ; git fetch (where apache is the location for apache's git server for ofbiz.git)19:29
→ beneggett joined19:30
schultza do i need to restart this whole thing over again?19:30
← fusion27 left19:30
selckin schultza: probably need like: git symbolic-ref HEAD refs/heads/trunk19:30
→ _huoxito joined19:31
→ etcetera joined19:31
schultza that's better19:31
← LeMike left19:31
→ anothervenue joined19:31
cbreak moltar_net: if you can not fast-forward push, then you have to make your local history fast-forwardable19:31
either by merging or by rebasing19:31
both will work, in different ways19:32
schultza nope ..still getting an error19:32
selckin moltar_net: what rebase command did you try for that? does git pull --rebase do want you want?19:32
moltar_net hmm... i did "git rebase origin"19:32
not "git pull --rebase"19:32
I'll try "git pull --rebase" next time.19:32
thanks!19:32
cmullican Huh? https://gist.github.com/1dc0bbc5efc76f283a3819:32
cbreak origin is bad19:33
← tatsuya_o left19:33
selckin cmullican: you're probably in a conflict and need to resolve it19:33
cbreak don't use remotes as branches or you'll mess up your understanding19:33
← bradrice left19:33
cbreak since I bet you don't know which branch origin actually refers to19:33
cmullican selckin: HOW? I thought I already had!19:33
← manolo left19:35
selckin cmullican: what does just git status say? if you resolved the conflict you may still need to like 'git commit' to tell git you resolved it, or like rebase --continue or merge --continue, but i think this was from a cherry pick? just commit should work19:35
← jkremser left19:35
cbreak no19:35
to tell git that you resolved a conflict, git add is sufficient19:35
← dc5ala_ left19:35
cbreak commit is only needed if you want to conclude a merge that stopped due to conflict19:35
← urbanmonk left19:36
selckin or rebase that stopped, ro cherry pick that stopped19:36
cmullican selckin: git status is in the gist19:36
selckin cmullican: without the path19:36
← alberto56 left19:36
cmullican I already committed, which is why I'm confused.19:36
→ urbanmonk joined19:36
→ millerti joined19:36
selckin dunno, hard to say without knwonig what exactly you did before19:36
millerti This should be a simple question, but I'm coming up dry on Google. If I "git commit -a", I get a list of files that are going to be committed. How can I just get that list printed to the console?19:37
selckin millerti: git status19:37
millerti Thanks!19:37
selckin: Actually lists a ton of untracked files. How can I restrict it to tracked files?19:38
FauxFaux You probably don't actually want to do that. Ignore the untracked files?19:38
selckin millerti: -u=no, or gitingore them19:38
millerti Well, I'll work on git ignore later. And thanks for the -u=no19:39
Apparently, -uno also works.19:39
selckin can do git add -u; to stage all changes to already tracked files, then just commit that19:39
← AdrienBrault left19:40
→ patrick85 joined19:40
→ imrobert joined19:40
← imrobert left19:40
→ imrobert joined19:40
millerti selckin: I have some tracked files that I don't want to commit right now.19:41
selckin its just a really nice timesaver (and way nicer then -a)19:41
so just stage those files19:42
millerti I'm not sure I understand. I'll look this up.19:42
→ branduren joined19:42
← raykyri left19:42
→ martinjlowm joined19:42
imachuchu millerti: git uses an opt-in system, just "git add" the files you want, then "git commit" to commit just them19:43
millerti Ah, I see. I'll have to remember thiat.19:43
selckin git add file-i-want-to-commit, check git status, it'll be at the top in green, if you then do commit it'll only commit the staged files19:43
FauxFaux Or maintain your ignore lists properly and use git add -A most of the time.19:43
millerti Ah, I see. I'll have to remember that. And yes, I was "git add"ing some files individually.19:43
imachuchu millerti: and git -p can add just sections of some files (so that you don't have to commit everything in the file)19:43
→ fayimora joined19:44
millerti My problem is that I was stupid and added some binaries to the repo. It's nice for when I want to make a release, but for the rest of the time, they're baggage, and the repo is getting bloated.19:44
selckin yeah, never a good idea19:44
imachuchu millerti: ewww19:44
selckin why don't these apparemnt listings have the address ¬|@¼ŧ^¬|#¼19:45
schultza working with java projects.. you might want to add a "*/build/*" in your .gitignore... you may have several directories with a build subdir19:45
selckin schultza: blasphemy /target/19:45
→ steffo joined19:45
schultza really?19:45
selckin mvn > ant :(19:45
schultza want all *.o files?19:45
selckin holy wars19:45
schultza ok ok... whats the better way to handle those bloated files?19:46
selckin i making a bad joke, ignore it19:47
→ raykyri joined19:47
thiago don't add *.o to the repository19:47
← enmand left19:47
← mklappstuhl left19:47
schultza thats why they are in the .gitignore right?19:47
_ikke_ yes19:48
← iamjarvo left19:48
← raykyri left19:50
→ enmand joined19:50
→ linoj joined19:53
← avpx left19:53
linoj hiya, looking for help how to change the color of "! [rejected] master -> master (non-fast-forward)" when i do a git push origin master. I have some console coloring but that is in white and i frequently don't notice that the push failed.19:54
→ tatsuya_o joined19:55
_ikke_ linoj: afaik, that's not supported by default19:55
selckin the output is like 5 lines longer then normal19:55
_ikke_ maybe you can make your terminal color stderr different or something19:56
linoj hmm19:56
← mechanyancat left19:56
← s0ber left19:56
selckin alias mypush="git push 2>&1 | grep -15 --color rejected" super duper lame19:57
→ s0ber joined19:57
→ allan_hi_ joined19:57
← Freeaqingme left19:58
← hobodave left19:58
← jonmasters left19:58
← mermeladas left19:58
← Schwern left19:58
← olaf_ left19:58
← daniel-s__ left19:58
← sente left19:58
← qwertz_ left19:58
← ku left19:58
→ olaf_ joined19:58
→ sente joined19:58
fisted_ → fisted19:58
→ qwertz_ joined19:59
→ robonyankitty joined19:59
→ Guest72152 joined19:59
selckin git config --global alias.mypush '!git push || echo "$(tput setaf 1)I FAILED YOU MY MASTER$(tput sgr0)"'19:59
git mypush19:59
← netj left19:59
→ strtok_ joined19:59
← Tabmow left19:59
→ Schwern joined19:59
← strtok left20:00
← unreal left20:00
← Bucciarati left20:00
Guest72152 → Freeaqingme20:00
→ Tabmow joined20:00
→ ku joined20:00
← ku left20:00
→ ku joined20:00
selckin add some "${@}" etc, yadada20:00
← strtok_ left20:00
← john_anderson left20:00
← beneggett left20:00
→ LeMike joined20:00
→ _daniel-s__ joined20:00
← lae left20:00
→ strtok joined20:00
strtok is there a nice way to keep a repository's history truncated? I want to build something that automatically truncates history at 30 days20:01
→ unreal joined20:01
→ pandeiro joined20:01
selckin no that goes against the core of git20:01
_ikke_ strtok: Git is built for keeping all history20:01
selckin why do you want this?20:01
_ikke_ While it's doable, it's not very simple20:01
→ lae joined20:01
strtok i understand that, but if i'm able to remove a single commit and then prune?20:01
→ beneggett joined20:01
_ikke_ strtok: The problem is that all new history is dependend on old history20:02
selckin its doable, but the repo won't be easy to use, are you trying to make backups?20:02
→ netj joined20:02
strtok selckin: using git for something other than source management20:02
selckin: no20:02
FauxFaux You can use replace/filter branch, git checkout --orphan and filter-branch, but it makes the entire repo unusable. strtok: !annex20:02
gitinfo Storing lots of binary files in git causes repo balloon. git-annex and git-media are two solutions to work around this by keeping the blobs out of the repo. http://git-annex.branchable.com/ and https://github.com/schacon/git-media20:02
_ikke_ !UGFWIINI20:02
gitinfo You appear to be Using Git For What It Is Not Intended - https://www.google.com/search?q=UGFWIINI Are you sure git is the right tool for the job? See also !wrong20:02
strtok we're playing with using git for replication for something20:02
selckin uh oh20:02
try rsync20:02
← aviraldg left20:03
FauxFaux _ikke_: I bet that took longer to type than the whole factoid. ¬_¬20:03
→ shawngo_ joined20:03
strtok rsync won't work bceause i need to calculate diffs quickly between 2 points in time, which git is good at :)20:03
_ikke_ FauxFaux: It doesn't help that I'm on a very slow connection right now20:03
← na3r left20:03
_ikke_ FauxFaux: But I had it right the first time20:03
← _huoxito left20:03
→ Jippi joined20:04
← moltar_net left20:04
← tommyvyo left20:04
← enmand left20:05
← shawngo left20:05
shawngo_ → shawngo20:05
→ Bucciarati joined20:06
→ Seo9 joined20:06
_ikke_ Weird20:07
selckin science20:07
← Seo9 left20:08
_ikke_ I did a git commit --amend -CHEAD, but that caused some files to be removed from git20:08
← Seveas left20:08
→ jonmasters joined20:08
← anothervenue left20:08
_ikke_ while that was not staged20:08
→ adek05 joined20:09
selckin shouldn't20:09
_ikke_ git status showedd the files being new and staged (after the commit), but git reset made them go away :S20:09
(ie, not modified or something)20:10
→ Seveas joined20:10
→ s0ber_ joined20:10
_ikke_ oh wait, they're in a directory20:10
weirdness20:10
selckin you didn't mention runnning reset20:10
→ mfonda joined20:10
→ sirdancealot7 joined20:10
← s0ber left20:10
_ikke_ No20:10
s0ber_ → s0ber20:10
→ Seo9 joined20:11
selckin it won't list files in untracked directory, so if you amended a delete that resulted in no files being tracked in the dir20:11
or something20:11
_ikke_ git add <file>; git commit --amend -CHEAD; (files suddenly get deleted from git)20:11
← steffo left20:11
→ k1t joined20:11
→ fly9_ joined20:11
← fly9_ left20:11
← fly9 left20:12
→ fly9_ joined20:12
_ikke_ Well, another git ammend fixes it20:12
amend*20:12
← nielsbusch left20:12
k1t hey folks, i have several cloned repos in my repo (they are extensions) and now on github it shows green folders next to the folders where the extensions should be: https://github.com/whitmanc/sisbro/tree/master/vendor/extensions i was wondering if anyone could point me in the right direction20:12
selckin head spins20:12
k1t im assuming those green folders are references but id like to just have the files/folders there, rather than references20:12
milki points k1t to #github20:13
← bwwrd left20:13
→ enmand joined20:13
selckin maybe you want submodules, but probably not20:13
→ aviraldg joined20:14
mfonda i'm tasked with converting from svn to git and my experience with git is somewhat limited--i'm wondering what the "git way" of doing the following is: we have a script that does `svn export /path/to/repo/specific/file-or-folder`, but from what i understand there isn't really an equivalent in git, or at least it's not the correct way to go. any advice on best practices for this situation?20:16
selckin what does export do20:17
→ steffo joined20:17
selckin mfonda: man git-archive20:18
gitinfo mfonda: the git-archive manpage is available at http://git-scm.com/docs/git-archive.html20:18
_ikke_ copy all files except the .svn direcotry20:18
selckin: https://gist.github.com/bf6033b82ac6241c9a0f20:18
selckin _ikke_: that looks strange20:19
adek05 I wanted to ask you guys a few questions about git diff. First of all I have file with lines like {...} {..} ... {...} (all the same number of mathchin parenthesis}, now I would like to generate HTML diffs from it taking few interesting bits of content of those parenthesis. I found git diff --word-diff option, which seems reasonable for that purpose and I would most likely use it, but the new lines may appear, can I somehow force git diff to show only word chang20:19
← airborn left20:19
adek05 And one more question, diff always shows first file then second file line, when some change occured, right?20:19
→ giallu joined20:20
← chuckharmston left20:20
→ mikeshultz joined20:20
→ simesy_work joined20:20
← cn28h left20:20
mikeshultz So using git flow, why would it prevent me from having multiple hotfix branches? What if a high priority bug comes through that trumps whatever I'm currently working on?20:21
← Seo9 left20:21
← aviraldg left20:21
→ airborn joined20:21
selckin see no reason it should prevent it20:22
_ikke_ mikeshultz: git flow is a convention20:22
← erichynds left20:22
mikeshultz _ikke_: Fair enough.20:22
selckin local branches are free, can like stash your current work, or make some tmp brach to store it, or create a new branch for the new stuff20:22
← v0n left20:23
mikeshultz I'm more curious about the reason they chose to do so. Seems strange to me.20:23
selckin ocd20:24
mikeshultz heh20:24
_ikke_ mikeshultz: I don't really get the problem, could you elaborate a bit about what you mean?20:24
← enmand left20:25
mikeshultz _ikke_: I could give you shell output to tell the story.20:26
mfonda selckin: to reword by question--my current procedure is: git clone repo.git; cp repo/some-file some-file; rm -rf repo; -- i feel like i'm fighting againt git in doing this, and wondering what the "git way" is for just getting one specific file/folder from a repo is, or if i should be taking a different approach?20:26
← acidjazz left20:26
← pandeiro left20:26
selckin mfonda: you already have the repo locally?20:26
mfonda selckin: it will be remote20:27
← marthinal left20:27
selckin mfonda: don't think you can get any files without cloning then, not sure, why don't you keep a local clone around?20:27
→ tubbo joined20:28
mikeshultz _ikke_: http://pastebin.com/QwY0tr9920:28
tubbo i believe i have written the longest commit message20:29
it is 176 lines20:29
_ikke_ woa20:29
selckin mikeshultz: ask git flow people? maybe they have a mailing list20:29
_ikke_ selckin: git flow is created by one person20:30
and it's not really a project20:30
selckin: It's mostly based on a blog post20:30
selckin a comment on his uber blog then20:30
still doubt many ppl use it20:30
← clifton left20:30
→ GNUtoo joined20:30
_ikke_ mikeshultz: Oh, you're talking about the git plugin20:31
tubbo yeah there's a plugin that helps you along20:31
→ acidjazz joined20:31
_ikke_ selckin: Well, mostly they use some variant of git flow20:31
selckin helps or makes it even more complicated20:31
→ mklappstuhl joined20:31
_ikke_ mikeshultz: I think you should just pick the things from it that fit your project20:31
→ kaen joined20:32
mikeshultz Not sure I get to make that call, but I guess some things just have to be done without it.20:32
_ikke_ fit flow is a model, it's not to be copied exactly, because it doesn't fit every project20:32
→ enmand joined20:32
_ikke_ mikeshultz: If you need a second hotfix branch, just create it20:33
→ tommyvyo_ joined20:33
GNUtoo hmmm, it's still blocked on 12 steps: http://pastie.org/private/ruozignklpw2uqitpjhf3g20:33
does somebody knows why?20:34
selckin is there a question?20:35
→ TeckniX joined20:35
→ aviraldg joined20:35
→ paissad joined20:36
← skyf left20:37
→ lacrymology joined20:37
GNUtoo yes I wonder if someone knows the cause of why it's blocked on 12 steps20:37
because some people told me it's normal20:38
but I'm not sure20:38
I'm bisecting since too long20:38
← pranjal710 left20:38
_ikke_ GNUtoo: Are there branches in between?20:38
GNUtoo it's the linux kernel....20:39
master/origin branch between v2.6.39 and v3.520:39
← kaen left20:39
GNUtoo so there are merge branches20:39
I guess20:39
I don't know the vocabulary well....20:39
_ikke_ GNUtoo: Merges confuse git bisect a bit20:39
GNUtoo ah ok20:39
_inte_ merges don't really confuse git-bisect20:40
paissad hello guys, i would like to know the real interest of the -bare option of "git init" ... is it "only" useful when we want that related repository to be the "central" repo where everyone makes the pull and push ?20:40
thanks in advance20:40
_inte_ It might confuse *you* because its jumps are too logical for you to handle ;)20:40
→ cn28h joined20:40
→ pandeiro joined20:40
SaCu paissad, it also helps with backups, mirrors or if you want local proxy repositories20:40
_ikke_ paissad: If you want to push to a branch, you want that branch to be bare20:40
pasky: !bare20:40
gitinfo pasky: an explanation of bare and non-bare repositories (and why pushing to a non-bare one causes problems) can be found here: http://bare-vs-nonbare.gitrecipes.de/20:40
_ikke_ paissad: *20:41
paissad ok20:41
GNUtoo basically that behaviour(keeping the same number of steps) seem to be introduced after the first git bisect skip(not in the pastebin, from yesterday)20:42
← chrisf|walrus left20:43
← sungnyc left20:44
← jsoa left20:44
← beneggett left20:44
GNUtoo maybe I should to it manually?20:45
like for instance git checkout commit20:46
then git bisect good|bad|skip20:46
← Kyzz left20:47
→ Kyzz joined20:47
← pandeiro left20:48
GNUtoo I'll do that20:48
← _iron left20:49
← mklappstuhl left20:50
→ sungnyc joined20:50
→ pantsman joined20:50
← pantsman left20:50
→ pantsman joined20:50
→ camdenx joined20:51
← jperry2_ left20:51
← enmand left20:52
→ Caritade joined20:52
← zenchaos left20:53
→ ameoba__ joined20:53
← artnez left20:54
→ jol1naras joined20:55
→ kaen joined20:55
← Bass10 left20:55
← freeed left20:56
→ michas joined20:57
→ raykyri joined20:57
← ameoba left20:57
michas hi, when showing patches/diffs -w ignores all whitespace. is there a way to ignore line breaks, too?20:58
→ dfr|moscow joined20:59
→ lavaman joined21:00
← ameoba__ left21:00
→ theCole joined21:00
→ AdrienBrault joined21:01
← freimatz left21:02
← kaen left21:03
← tommyvyo_ left21:03
← Jay_Levitt left21:04
← berserkr left21:05
→ cagedwisdom joined21:05
← tinti left21:05
→ john_anderson joined21:06
→ rehat joined21:06
← lacrymology left21:06
→ alanroemen joined21:07
→ shruggar joined21:07
gitinfo set mode: +v21:07
← Caritade left21:08
→ teweWORK joined21:08
← dfr|moscow left21:08
→ mtytel joined21:08
← psoo left21:09
← alanroemen left21:10
→ alanroemen joined21:11
→ psoo joined21:11
→ kcj joined21:12
← spyro left21:12
← phishy left21:12
→ ameoba__ joined21:12
→ spyro joined21:13
→ skorgon_ joined21:13
← tplaner_ left21:13
alanroemen if I have a .gitmodules file already defined from another similar repo, is it possible to initialize them from the data in that file?21:13
→ shavus joined21:14
_ikke_ alanroemen: afaik not21:14
← hoschi left21:14
→ j416 joined21:15
alanroemen _ikke_: bummer, that's what I thought but wanted to ask. thanks121:15
← olaf_ left21:16
→ mcluskydodallas joined21:16
← etcetera left21:16
← gtrak left21:17
← dsantiago left21:17
← paissad left21:18
→ Bass10 joined21:21
→ enmand joined21:22
← dzonder left21:24
→ alberto56 joined21:24
← hellopat left21:25
← jjasinski left21:27
→ artnez joined21:28
← glennpratt left21:29
← AaronMT left21:29
→ madewokherd joined21:29
→ glennpratt joined21:30
→ AaronMT joined21:31
← martinjlowm left21:34
← d0k left21:34
→ Wooga joined21:34
← adek05 left21:35
← lonewulf` left21:35
← macabre left21:35
→ dfr|moscow joined21:35
← blacktulip left21:36
GNUtoo I think I mistaken a skip for a bad21:37
I said bad instead of skip....21:37
← theCole left21:37
→ lonewulf` joined21:37
GNUtoo so I'll restart with tested good and bad from last bisect....21:37
← j416 left21:38
→ jjasinski joined21:40
← aselder left21:41
← BD- left21:41
→ edude03 joined21:42
→ yoklov joined21:42
← linoj left21:44
← drizzd left21:44
← jol1naras left21:45
← elux left21:46
→ Shagane joined21:47
← Shagane left21:47
← nwest left21:48
← Xizor left21:49
→ w09x joined21:50
→ tty011 joined21:51
← cannonball left21:52
gitinfo set mode: +v21:52
tty011 anyone know how i can add a ssh pub key via the gl-auth-command?21:52
→ hobodave joined21:53
→ cannonball joined21:53
← EricKamsky_ left21:54
schultza i have have created a bare repos, added remoted to outside repos and fetched it.. how do i get it serving up the repos to the other local computers (one download the the entire company) and in-house checkouts from there21:54
← bigkm left21:55
→ osmosis joined21:56
← nickpresta left21:57
← urbanmonk left21:57
→ bigkm joined21:57
← dfr|moscow left21:57
← fly9_ left21:58
← kermit left21:59
schultza how do i set up git to auto track all remotes/origin branches?22:00
← eletuchy left22:01
← michas left22:01
← ehsan left22:03
→ jeffisabelle joined22:03
→ Textmode joined22:03
← shruggar left22:04
→ Milossh joined22:04
← intransit left22:04
← Ryback_ left22:06
→ Dreamer3 joined22:06
← TeckniX left22:06
← AaronMT left22:07
→ davidjdagino_ joined22:09
← axl_ left22:11
← dv310p3r left22:11
← CannedCorn left22:11
← psoo left22:12
← davidjdagino left22:12
davidjdagino_ → davidjdagino22:12
← john2496 left22:12
← kenperkins left22:12
→ macIvy joined22:13
→ kenperkins joined22:13
→ Anxi80 joined22:13
imachuchu schultza: "git clone --mirror <sourceServer>"?22:14
→ TeckniX joined22:15
← Speed left22:15
→ toofar joined22:15
← raykyri left22:16
→ zastaph joined22:18
← Facetious left22:19
→ toddr joined22:20
toddr I've noticed that the git client is significantly slower making remote connections on CentOS 6 vs CentOS 5. I think the reason is that git is trying IPv6 first before trying IPv4.22:22
← Vert left22:22
toddr Does anyone know if there's a way to disable this?22:22
← tty011 left22:23
← GNUtoo left22:23
→ tinti joined22:24
→ tweedledumber joined22:25
← skorgon_ left22:25
→ gtrak joined22:26
→ AaronMT joined22:26
→ caseymcg joined22:27
← giallu left22:28
← tatsuya_o left22:28
→ xeon123 joined22:28
← shavus left22:29
xeon123 I've a local repository, and I want to put a copy of my repository in a backup disk without losing the history. so, sometimes I want to push the local repository to the backup. How do i do that?22:30
→ hwrd|work joined22:30
→ kermit joined22:31
← thiago left22:32
→ Speed joined22:33
← Speed left22:33
→ Speed joined22:33
← pppaul left22:33
→ thiago_ joined22:33
imachuchu xeon123: on remote "git clone --mirror <workRepo>", then "git push --all --tags", I think22:34
→ mishok13 joined22:34
imachuchu xeon123: sorry "git push --all --tags <remote>"22:34
← pantsman left22:34
← lekefly_ left22:35
→ LekeFly joined22:35
← rehat left22:36
→ ph^ joined22:36
← glennpratt left22:37
← sq-one left22:39
← steffo left22:40
xeon123 imachuchu: can you give me an example of <remote>? My repository locally calls test. After I did clone, it created test.git repository.22:41
what I put in git push?22:41
is it test or /backupdrive/repo/test.git?22:42
← fireh left22:43
← nutron left22:43
imachuchu xeon123: well we can think of two repos, local (the one we are working on) and remote (the one we want to push/pull from). It sounds like in your situation that "test.git" is your remote, so it would be push "git push --all --tags test.git" (assuming that it knows of test.git as a remote)22:44
xeon123: if not, man git-remote22:44
gitinfo xeon123: the git-remote manpage is available at http://git-scm.com/docs/git-remote.html22:44
→ bubbles10101 joined22:45
← Samos123 left22:46
frogonwheels xeon123: did you work it out?22:46
xeon123 frogonwheels: yes.22:47
← chrisgeorge left22:47
frogonwheels xeon123: you can also do a 'fetch' from the mirror clone to get the latest22:48
xeon123: (you might want to follow up with a git gc ocassionally as well)22:48
→ LucasTizma joined22:49
xeon123 frogonwheels: what is git gc?22:49
frogonwheels xeon123: (which reminds me. I need to 'garbage collect' my own backup clones)22:49
← bubbles10101 left22:50
← edude03 left22:50
frogonwheels xeon123: it gets rid of stale objects and creates pack-files. should be done every now and again.22:50
→ Berglund joined22:52
SeoZ-work[AWAY] → SeoZ22:52
← patrick85 left22:53
→ jjang joined22:53
xeon123 I did "git fetch /backupdrive/repo/test.git" in another dir, and I've got "Stopping at filesystem boundary (GIT_DISCOVERY_ACROSS_FILESYSTEM not set)."22:53
what iswrong?22:54
← Bass10 left22:54
→ EricKamsky joined22:55
→ rehat joined22:56
gitinfo set mode: +v22:56
imachuchu xeon123: as a silly sanity check, you didn't copy this repo from your hard drive to the backup drive did you? (like through "cp")?22:56
xeon123 imachuchu: no. I did it with git clone and git push22:57
← hobodave left22:58
SaCu xeon123, "cd /backupdrive/repo && git clone --mirror /path/to/repo" for setup and "cd /backupdrive/repo && git pull" for update - ought to be enough to set it up22:58
← schultza left22:58
← jjang left22:58
→ jjang joined22:59
xeon123 that's what I did, and it's ok.23:00
imachuchu: now, I'm in another dir: ~/temp/ and I want to fetch the test.git23:00
← SirCmpwn left23:00
imachuchu xeon123: you want to create a new repo in ~/temp, or just pull the tip of backups master?23:00
→ Samos123 joined23:01
→ CannedCorn joined23:01
xeon123 just pull from the backups master23:01
→ cybersphinx joined23:03
→ axl_ joined23:03
imachuchu xeon123: "git pull /backupdrive/repo master", !pull4 but it's ok if it's just a test, or you really only want it once, else properly set up a remote23:04
gitinfo xeon123: [!fetchfour] Never use the four-word version of git-fetch or git-pull (e.g. git fetch remote refspec). It always ends in tears. Yes, if you understand the implications of FETCH_HEAD it can technically be done, but really it is easier to just fetch the whole remote (or perhaps edit the fetchspec if you never want other bits).23:04
EugeneKay !last23:04
gitinfo git-last is a script which will give you a GitHub-like view of the current directory(showing the last commiter of each dir/file): http://goo.gl/zb6xi23:04
imachuchu xeon123: but you will still need to create a repo at temp. or just "git clone /backup/repo"23:05
← CannedCorn left23:05
→ macabre joined23:05
← leeN left23:06
imachuchu EugeneKay: so kind of a blame but on a folder level instead of a file?23:06
← the_cybersphinx left23:06
← resmo__ left23:06
EugeneKay imachuchu - sort of. I just needed the link. :-p23:06
It has it's issues.23:07
← macIvy left23:07
imachuchu !privmsg but 1. I don't think anyone here minds, and 2. I think you maintain the bot so there really isn't any reason to tell *you* this ;)23:07
gitinfo You can get gitinfo keyword expansion (e.g. !doc or this one) sent directly to you by sending a private message to me: /msg gitinfo !doc. This avoids spamming the channel. Also see See http://jk.gs/git/bot/trigger.php23:07
EugeneKay jast's bot. I'm just the chief abuser.23:08
xeon123 imachuchu: it was missing the git clone. now it works. It's preferable to use git fetch or pull?23:09
← seungers left23:10
imachuchu xeon123: not preferable, just different. pull = fetch + merge23:11
← john_anderson left23:12
← Samos123 left23:13
← zackiv31 left23:13
Marcux83 → Marcux83|away23:15
xeon123 how do I merge conflicts in git?23:15
← AdrienBrault left23:15
imachuchu xeon123: carefully? !eekaconflict23:16
gitinfo xeon123: Merge conflicts are a natural part of collaboration. When facing one, *stay calm*. Read "How to resolve conflicts" in man git-merge. Carefully go through the conflicts. Picking one side verbatim is not always the right choice!23:16
→ nutron joined23:16
← thelazydeveloper left23:16
← lonewulf` left23:16
xeon123 is there a tool to do it, or it must be in text mode?23:17
imachuchu could we change "stay calm" to "don't panic" (as a reference to Hitchhikers Guide)?23:17
EugeneKay Go forit.23:17
→ raykyri joined23:18
→ etcetera joined23:18
→ lonewulf` joined23:18
← Zyrkon left23:18
imachuchu xeon123: there are plenty of diff tools (I use vim), heres a list of some: http://www.handlewithlinux.com/Linux-GUI-diff-utilities-a-visual-tour23:19
EugeneKay: I have no idea how/wouldn't want to mess up anything23:19
EugeneKay !bot23:19
gitinfo [!gitinfo] I am an IRC bot which responds to certain keywords to provide helpful(?) information to humans. Please see http://jk.gs/git/bot for more information about how to use me.23:19
← devth left23:19
→ resmo joined23:21
← zastaph left23:21
← Jippi left23:22
imachuchu EugeneKay: while I have that page bookmarked (and use it often) it doesn't say how to change anything23:22
why? !cuz23:22
gitinfo [!why] Why? Because screw you, that's why.23:22
EugeneKay ianmcorvidae - yes it does ;-)23:22
→ Disreali joined23:23
imachuchu *sigh*, go go grep23:23
EugeneKay !fish23:23
gitinfo In the long run, it is much better to give pointers to answers than the answers themselves.23:23
EugeneKay ;-)23:23
← Chillance left23:24
imachuchu ahh, I bookmarked the trigger page, forgot there was the page before it23:24
→ SirCmpwn joined23:24
← SirCmpwn left23:24
← alanroemen left23:24
← segher left23:26
→ SirCmpwn joined23:27
→ macIvy joined23:27
← AaronMT left23:27
← teweWORK left23:27
← caseymcg left23:28
→ segher joined23:28
→ AaronMT joined23:28
← airborn left23:29
← segher left23:29
← gtrak left23:30
imachuchu .trigger_edit eekaconflict Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge. Carefully go through the conflicts. Picking one side verbatim is not always the right choice!23:32
gitinfo imachuchu: Okay.23:32
imachuchu !conflict23:32
gitinfo [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge. Carefully go through the conflicts. Picking one side verbatim is not always the right choice!23:32
→ segher joined23:32
← Zenmaker_ left23:32
→ Zenmaker_ joined23:33
← boombatower left23:33
→ Samos123 joined23:34
→ pdtpatrick_ joined23:35
← pdtpatrick left23:35
pdtpatrick_ → pdtpatrick23:35
← segher left23:36
→ segher joined23:36
← macIvy left23:38
→ tty01 joined23:39
← maestrojed left23:39
← AaronMT left23:39
← nutron left23:40
← tty01 left23:43
← toddr left23:45
→ david1010 joined23:45
← goshawk left23:48
→ zdk joined23:48
← LeMike left23:53
infogulch is there a way to exclude specific branches when i want to `git push --all` ?23:53
→ Zenmaker__ joined23:55
← yoklov left23:58
← TeckniX left23:58
← xeon123 left23:59
← Zenmaker_ left23:59

Logs Search ←Prev date Next date→ Channels Documentation