IRCloggy #git 2013-04-16

Provider of IRC logs since 2005.

WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

	2013-04-16
	→ tmcmahon joined	00:00
	← ozette left	00:00
offby1	easiest way: "git push -u remotename branchname"	00:01
	← nigelb left	00:01
	→ nigelb joined	00:01
	→ knightshade3 joined	00:02
	← Noldorin left	00:03
	← tvw left	00:04
	← boombatower left	00:04
	← rvsjoen_ left	00:05
	← TheMoonMaster left	00:05
	← knightshade2 left	00:06
	← thasmo left	00:06
	← dreiss left	00:06
	→ Noldorin joined	00:06
	→ thiago joined	00:06
	→ chuckharmston joined	00:07
	→ boombatower joined	00:07
	→ kvda joined	00:08
	← rideh left	00:08
	→ TheMoonMaster joined	00:08
	← tmcmahon left	00:10
	→ ncd\|leeN joined	00:10
	← ncd\|leeN left	00:11
	← jacoluta left	00:11
	→ rickruby joined	00:12
	→ joaoh82 joined	00:13
	← leeN left	00:13
	← eletuchy left	00:14
	← AdrienBrault left	00:14
	→ pitzips joined	00:16
	← knightshade3 left	00:16
	← ehsan left	00:16
	→ Markvilla joined	00:17
	← Celelibi left	00:18
	← hwrd\|work left	00:20
	→ rvsjoen joined	00:20
	← rvsjoen left	00:20
	→ rvsjoen joined	00:20
	← wheezl left	00:21
	→ miso1337 joined	00:22
	→ infynyxx joined	00:22
	← pastime left	00:22
	← mloskot left	00:23
	← guns left	00:24
	← pitzips left	00:25
	→ rasto joined	00:26
	→ dideler joined	00:26
	→ private joined	00:28
	← m0ltar left	00:28
	← rurufufuss left	00:28
	← marc_kdab_ left	00:29
	→ bmac joined	00:29
	← daivdh left	00:31
	← joaoh82 left	00:31
	→ gerep joined	00:31
	← alansaul-laptop left	00:33
	← yshh left	00:33
	← vedant left	00:35
	→ simesy joined	00:36
	← ISF left	00:36
	← simesy left	00:36
	→ w09x joined	00:36
	← w09x left	00:37
	→ simesy joined	00:37
	→ lindenle_ joined	00:37
	← simesy left	00:37
	← hyperair left	00:38
	→ pitzips joined	00:38
	← poohtarou left	00:38
	← Bilge left	00:39
	← bytephilia left	00:39
	→ simesy joined	00:39
	→ maslowbeer joined	00:40
	→ poohtarou joined	00:40
	→ rideh joined	00:41
	→ Watabou joined	00:41
	→ kit joined	00:42
	→ joaoh82 joined	00:42
	← sluther left	00:44
	← chrisdev left	00:44
	→ Sosumi joined	00:44
	→ normanrichards joined	00:46
	← jaequery left	00:47
	← wrouesnel left	00:48
	← Alenah left	00:52
	← hellopat left	00:52
	→ hellopat joined	00:53
	← barneybook left	00:54
	→ Alenah joined	00:54
	← blami_orcl left	00:55
	→ blami_orcl joined	00:55
	← lindenle_ left	00:55
	← gerep left	00:56
	← bmac left	00:56
	← thiago left	00:57
	← hellopat left	00:57
	← benjings left	00:57
	← joaoh82 left	00:58
	→ ehsan joined	00:58
	← samholmes left	00:58
	→ kogent joined	00:58
	→ yshh joined	00:59
	→ Gentlecat__ joined	00:59
	← Gentlecat_ left	01:00
	→ agile joined	01:02
	← FabTG left	01:02
	→ FabTG_ joined	01:02
	← kogent left	01:03
	← electrog1ek left	01:03
	→ tinti joined	01:03
	→ MattStratton joined	01:04
	→ dharriga1 joined	01:04
	→ Joe9 joined	01:04
	← FatDarrel left	01:04
	← normmalno left	01:05
	Joe9 → Guest21424	01:05
	→ brocken_ joined	01:05
	← Joe8 left	01:05
	← brocken left	01:06
	← dharrigan left	01:06
	→ tjbiddle joined	01:06
	← scelis left	01:06
	→ R1KRDO joined	01:06
	→ flippo joined	01:07
	← natechan left	01:09
	← flippo left	01:09
	→ flippo joined	01:09
	→ nathanr_ joined	01:09
	→ jalama joined	01:11
	← trusktr left	01:12
	← nathanr left	01:13
	← aberrant left	01:13
	← kit left	01:13
	← R1KRDO left	01:13
	→ aberrant joined	01:14
	→ d3O joined	01:14
	→ electrogeek joined	01:14
	← electrogeek left	01:14
	→ electrogeek joined	01:14
	← aberrant left	01:15
	← d3O left	01:15
	→ d3O joined	01:15
	← MattStratton left	01:15
	← SaCu left	01:15
	→ security joined	01:16
	→ SaCu joined	01:16
	← megha left	01:17
	→ Bilge joined	01:18
	→ kit joined	01:18
	→ hyperair joined	01:19
	← esto left	01:19
	← ProLoser left	01:20
	→ jaequery joined	01:20
	→ r0bby_ joined	01:20
	→ eletuchy joined	01:20
	→ ProLoser joined	01:21
	→ rpg joined	01:21
	← rakl left	01:23
	← dmarkham_ left	01:23
	→ hellopat joined	01:23
	→ darinmorrison joined	01:24
	→ Goplat joined	01:24
	r0bby_ → robbyoconnor	01:25
	→ esto joined	01:25
	← eletuchy left	01:25
	← goshawk left	01:25
	← jaequery left	01:26
	→ nicksloan joined	01:27
nicksloan	Just curious, is there a way, and if not, would it be possible, to clone a repository as of a given commit?	01:27
	so if there are 100,000 commits, and I know I'll never nead to touch them, I can just start at commit 100k and work from there	01:28
bob2	not really	01:29
nicksloan	just seems silly. Who needs the inital import of a 10 year old project?	01:29
bob2	it's not silly	01:29
	→ BiggFREE joined	01:30
	← dkpi left	01:30
	→ Saby joined	01:30
	← Saby left	01:30
	→ Saby joined	01:30
nicksloan	I understand that someone may, but they could do a full clone. The full history should be maintained somewhere for sure	01:30
	→ erkules_ joined	01:30
	→ thegorn joined	01:30
	→ stoo_ joined	01:31
bob2	maybe you want man git-clone, but that'll cripple your clone	01:31
gitinfo	the git-clone manpage is available at http://jk.gs/git-clone.html	01:31
nicksloan	but I'm never going to need to look at the first commits from python. So why download all of that.	01:31
bob2	suggest just learning how git works	01:31
EugeneKay	nicksloan - man git-clone; there's a --depth option, but it really isn't a good idea. The resultant repo will eventually break.	01:31
gitinfo	nicksloan: the git-clone manpage is available at http://jk.gs/git-clone.html	01:31
	← simesy left	01:31
EugeneKay	If you want a tarball, just ask for a tarball.	01:32
	If you want to contribute, bite the bullet and wait for the history to download.	01:32
nicksloan	EugeneKay: what breaks it?	01:32
EugeneKay	The fact that it's a shitty hack which bypasses integrity checks on the git objects store	01:32
	← erkules left	01:32
	← yann\|work left	01:33
nicksloan	like I said, purely a thought exercise here. And curiousity about the workings of git and other vcs's	01:33
	← agu10^^ left	01:33
	→ joaoh82 joined	01:33
	→ simesy joined	01:33
	→ SuperSteff joined	01:34
	← amesha left	01:35
	→ amesha joined	01:36
	← boombatower left	01:36
	→ Vampire0_ joined	01:37
	→ barneybook joined	01:37
	← simesy left	01:37
	← tinti left	01:38
	← robbyoconnor left	01:40
	→ Davey joined	01:40
EugeneKay	What you can do with git is to "snip off" old history by creating an !orphan branch with the current state of the repo, and then work from there	01:40
gitinfo	To create an orphaned branch(unrelated to existing history): `git checkout --orphan newbranchname`. This will leave your index/worktree as-is(use `rm .git/index; git clean -dfx` to delete EVERYTHING). If you get 'error: unknown option `orphan`' see !orphan_old. For an empty/null commit see !orphan_null.	01:40
EugeneKay	The old history can be archived for historical purposes	01:40
bob2	but who cares	01:40
	for 99.9% of repos it doesn't matter	01:40
	← Vampire0 left	01:41
EugeneKay	In practice, this isn't done even with huge projects like the linux kernel, because disks space is cheap	01:41
	← subfauna left	01:41
	← maqr left	01:42
	→ hafeez3000 joined	01:42
	→ ergZay joined	01:42
	← Wulf left	01:42
	→ fredlwm joined	01:43
	→ Dougie187 joined	01:43
	← erkules_ left	01:45
	→ yann\|work joined	01:45
	← preflex left	01:45
	→ Dreamer3 joined	01:46
	→ preflex joined	01:46
	← tureba_ left	01:46
	→ erkules_ joined	01:46
	→ tureba joined	01:47
	← rickruby left	01:50
	→ simesy joined	01:50
	→ NiteRain joined	01:50
	→ subfauna joined	01:51
	← danishkhan left	01:51
	gitinfo set mode: +v	01:51
	→ simesy_ joined	01:52
fredlwm	Is git supposed to work with 128Mb of RAM ? I have a copy of the kernel tree, but 'git pull' doesn't do anything, except using the CPU until I kill it, or exiting by itself without printing anything, sometimes after hours. I'd also expect it to use the swap (almost 1Gb free), but it doesn't use it. It works fine with smaller repositories, so it looks like a memory issue ?	01:52
bob2	well, 128mb is a tiny amount for a huge repo like the linux kernel	01:53
	→ jdyin joined	01:54
	← simesy left	01:55
	→ teejmya joined	01:55
	Joefish → Joefish]off	01:56
	← kit left	01:57
fredlwm	Likely. But i'd expect git to print something. -v doesn't add any info. I wait, sometimes hours, but it never succeeds. I only see the CPU usage increasing and the timestamp in .git/FETCH_HEAD getting updated. I think it used to work with 384Mb. I'll see if it does with 256.	01:59
	← momomomomo left	01:59
	← simesy_ left	01:59
	→ alansaul-laptop joined	02:02
	← tiglionabbit left	02:02
	→ pastime joined	02:03
	→ R1KRDO joined	02:04
	→ BadQuanta joined	02:04
	→ stack_ joined	02:05
	→ brum joined	02:07
	← stack left	02:07
	→ robottinosino_ joined	02:10
	→ dreiss joined	02:10
	→ etcetera joined	02:10
	→ luc_ joined	02:12
	← tjbiddle left	02:13
	← harleypig left	02:15
	← d3O left	02:16
	→ d3O joined	02:17
	→ m0ltar joined	02:20
	→ rickruby joined	02:20
	→ tiglionabbit joined	02:21
	← tiglionabbit left	02:21
	← m0ltar left	02:24
	→ InvidFlower joined	02:24
	→ jaequery joined	02:25
InvidFlower	Any thoughts on merging vs rebasing? I guess it is a matter of taste...	02:25
	← jaequery left	02:25
	← joaoh82 left	02:26
madewokherd	they both have their uses	02:26
	→ jaequery joined	02:26
	← pitzips left	02:27
	→ avinashrbhat1 joined	02:27
InvidFlower	I like how it makes the graph simpler.. makes me feel a bit weird that it all comes in at the same time...	02:27
madewokherd	rebasing is good for cleaning up a set of changes before you add/send it to a project officially	02:28
	and keeping the history of that project tidy	02:29
	merging is good for long-lived forks	02:29
InvidFlower	That makes sense...	02:29
	← avinashrbhat left	02:30
	→ rchavik joined	02:32
	→ delinquentme joined	02:32
	→ engkur joined	02:33
dideler	InvidFlower: i rebase when my topic branch needs to be updated, and merge when the work on the topic branch is complete	02:33
madewokherd	I just remembered, merge --ff-only is pretty good too	02:34
	that's my default for updating things	02:34
	→ mstrcnvs joined	02:34
	→ xiukun joined	02:35
	Orbitrix → Huevo	02:36
InvidFlower	Yeah.. I've been working for a couple of days on a branch. Just rebased on the branch to get new changes from master..	02:36
dideler	if i'm getting updates from a remote repo, i pull --rebase. for merging the topic branch i use merge --no-ff	02:36
mstrcnvs	I have a script that postprocess my binary files and upload them IF THEY GOT CHANGED to a certain host. With rsync this was handled right, it uploaded only the changed. But with git it seems that every push that I do it changes the binaries and all of them get uploaded even if not changed...	02:36
	Huevo → Orbitrix	02:36
madewokherd	with a remote repo I use merge --ff-only too, so that I get interrupted and have to think about what to do if I have local changes	02:37
	← jalama left	02:38
	→ gumnos joined	02:38
InvidFlower	dideler: so forcing no merge when going to the branch and forcing the merge when bring back to master. That sounds like a good way to go..	02:38
SamB	madewokherd: I just wish you could then do a merge or a --rebase ...	02:38
	← Davey left	02:38
gumnos	When rebasing to split a commit, is there an easy way to specify that the current commit should retain the commit message from the source commit without retyping it?	02:38
	→ acharlieh joined	02:38
madewokherd	SamB: I don't follow; surely you can merge or rebase whatever you want..	02:38
	→ MattStratton joined	02:39
madewokherd	gumnos: how do you do that currently? I never really found a procedure that I liked	02:40
	← orafu left	02:41
	→ hwrd\|work joined	02:41
	→ orafu joined	02:41
gumnos	madewokherd: before rebasing the commit, dump the commit log to a temp file, do my rebase and read it into Vim. That's if I remember to dump the commit-log entry.	02:41
	that's why I'm asking if there's a better way :-)	02:41
	this evening's "solution" was simply to saw "aw, screw it" and copy over the first line of the commit-log, dropping the boring details that would have been in the body	02:42
bambams	gumnos: For starters you can use the -F option to read directly from the file. Assuming the file contains exactly the log you want.	02:42
madewokherd	I thought of doing a rebase -i and picking the commit twice (editing the first one), but I don't know if that's better..	02:42
	← ProLoser left	02:42
	→ preflex_ joined	02:43
madewokherd	wait, you can dump a series of commits to a file and then rebase those?	02:43
	← hwrd\|work left	02:43
bambams	gumnos: You can also read directly from inside of vim with :r!, possibly parsing the temp file.	02:43
	s/parsing/sparing/	02:44
	→ m0ltar joined	02:44
	← dvide left	02:44
gumnos	madewokherd: me? no. I'm just doing a split as described at http://git-scm.com/book/en/Git-Tools-Rewriting-History#Splitting-a-Commit but wanted to be lazy in keeping the commit message for one of the committed pieces.	02:45
	← preflex left	02:45
	preflex_ → preflex	02:45
	→ h2ero joined	02:45
gumnos	bambams: the trick would be finding the commit message in question. I'm guessing there's some way to coerce it out of git with minimal fuss for the commit that is currently being rebased. Just haven't banged on it long enough to find something I can remember easily enough to do next time.	02:46
	(I'm quite comfortable in Vim, so that part isn't the issue)	02:46
madewokherd	aha	02:46
	don't do git reset HEAD^	02:46
	well	02:47
	this'll be a procedure I haven't tried, but I think it'll work	02:47
gumnos	madewokherd: ? Is there a better way to get the effect of unstaging file2.txt but keeping file1.txt and the commit message, then commiting file2.txt with a new message?	02:47
bambams	gumnos: Well HEAD should point to the commit you're editing. So log HEAD~..HEAD should give you the log, I think.	02:47
	→ pitzips joined	02:48
	← engkur left	02:48
	→ maqr joined	02:48
bambams	I am quite tired though so I may be completely wrong.	02:48
madewokherd	except that you lost the commit by resetting	02:48
bambams	No, a soft reset won't change your working tree's head commit. Just basically resets the index and working directory.	02:49
	Again, unless I'm tired and thinking of this wrong...	02:49
madewokherd	that page doesn't say to do a soft reset	02:49
	← teejmya left	02:50
madewokherd	anyway, what I would do is copy the line that has your commit, so that it appears twice: making the first one an 'edit' and the second one a 'reword'	02:50
bambams	Wait, I'm thinking of mixed...	02:50
	← nathanr_ left	02:51
	→ nathanr joined	02:51
madewokherd	when it stops to edit your commit, don't reset, but do a git checkout -p HEAD^ and remove the parts of the commit you don't want (or otherwise add any changes you want to make to the first commit to the index)	02:52
	then do git rebase --continue	02:52
	← mstrcnvs left	02:52
	→ danishkhan joined	02:52
	← robottinosino_ left	02:53
madewokherd	(rebase --continue will amend your commit for you; I don't recommend amending it yourself or making commits yourself during a rebase, because doing that when resolving a conflict can screw it up)	02:53
	← etcetera left	02:53
	→ beakman joined	02:53
gumnos	That gives me some stuff to play with. Thanks, madewokherd & bambams	02:53
bambams	gumnos: Perhaps ORIG_HEAD is what you're after though.	02:53
	gumnos: Seems to do what I want after a reset --mixed HEAD~.	02:54
	← maqr left	02:54
	→ engkur joined	02:55
	→ LucasTizma joined	02:55
madewokherd	actually I'd probably revert the commit, pick the commit with -n, only commit the bits I want in my first commit, then delete the revert in the rebase, but I'm crazy	02:56
	→ ryez joined	02:56
madewokherd	I don't split commits very much; mostly I commit often enough that I'm combining them instead	02:57
	← danishkhan left	02:57
bambams	I usually end up adding edited patches instead. :-X	02:58
madewokherd	I never got the hang of hand-editing patches	02:58
	→ Garr255 joined	02:58
bambams	hack .. hack .. hack ... git add -p && git commit -m && git add -p && git commit -m ...	02:58
madewokherd	I'm just told they're malformed when I try to apply them	02:58
	→ tjbiddle joined	02:58
madewokherd	oh right add -p, I use that a lot	02:59
bambams	Yeah, it takes a bit of practice, but it's not too hard. My first experiences with hand editing patches was with Subversion... Probably resolving conflicts or something directly in a big patch instead of the files or something. I forget. It's been years.	03:00
Garr255	hey all, I need some help. I've pulled a master branch, modified it, then pushed back to master. I realized that I should have made changes to the "development" branch and not the master. is there any way to merge only MY modifications into the develop branch, but not the other ones committed to the master?	03:00
	← juanmabc left	03:00
bambams	I remember I lost a bunch of work and managed to recover it after hacking away at a patch file for an hour or something.	03:00
	← PaulCapestany left	03:00
nevyn	Garr255: yes.	03:01
bambams	I recently added some Vim mappings to make it a bit easier too.	03:01
	← engkur left	03:01
nevyn	Garr255: the simplest is probably to cherry pick it.	03:01
	Garr255: this doesn't remove it from master on the remote tho.	03:01
Garr255	nevyn, meaning what? and yes I'm okay with that	03:02
nevyn	git cherry-pick	03:02
bambams	Basically will duplicate the commits onto the develop branch.	03:02
	← dideler left	03:02
	← R1KRDO left	03:02
nevyn	well the one with your changes...	03:02
	if you specify it.	03:02
	→ rakl joined	03:02
	← AaronMT left	03:03
	bambams should really go to bed.	03:03
nevyn	git checkout develop && git cherry-pick <sha of desired change>	03:03
Garr255	any where do I find that?	03:03
nevyn	git log master	03:04
Garr255	should look like 028b3a6ecd047f9346e5b5105b49d9ba351b51ea ?	03:04
nevyn	yep.	03:04
Garr255	sweet	03:04
nevyn	Garr255: how long have you been using git.	03:04
Garr255	thanks	03:04
	one day	03:05
nevyn	Garr255: please read something...	03:05
	← hellopat left	03:05
nevyn	Garr255: like !book	03:05
gitinfo	Garr255: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable	03:05
	← MattStratton left	03:05
	→ simesy joined	03:05
Garr255	will do :P	03:05
bambams	I don't know how this bot works, but !backup	03:05
gitinfo	Taking a backup of a git repository is always a good idea, especially when taking advice over IRC. Usually, the best way to TACTICALLY back up a git repo is `git clone --mirror`. However, some unusual maintenance might require `tar cf repo-backup.tar repodir`. Testing in a clone is also an excellent idea. See also http://sethrobertson.github.com/GitBestPractices/#backups	03:05
Garr255	git is awesome :D	03:05
nevyn	I personally like !gcs	03:05
gitinfo	[!concepts] "Git Concepts Simplified" explains the basic structures used by git, which is very helpful for understanding its concepts. http://gitolite.com/gcs/	03:05
nevyn	!talks	03:06
gitinfo	Some good video talks about Git: [yt] http://goo.gl/z72s (Linus Torvalds: History&Concepts); [yt] http://goo.gl/R9H2q (Scott Chacon: Git basics, live examples); http://vimeo.com/35778382 (Randal Schwartz: Git basics, descriptional); http://vimeo.com/46010208 (Jesica Kerr: Git basics, descriptional)	03:06
	bambams recommends a tarball.	03:06
	→ etcetera joined	03:06
	Garr255 is incompetant	03:06
	→ wesside joined	03:06
InvidFlower	Git isn't the most obvious software I have to say.. takes a while to get the hang of it	03:07
nevyn	InvidFlower: it's really very simple if you look at it from the inside out.	03:07
	InvidFlower: if you understand the plumbing it all makes sense	03:07
	if you try and understand it from the porcelin you'll never get it.	03:07
InvidFlower	nevyn: Yeah.. and I agree that stuff like Pro Git helps a lot for that.	03:08
	I just mean some of isn't as obvious as it could be perhaps...	03:09
	→ spyroboy_ joined	03:09
	← spyroboy_ left	03:09
	→ spyroboy_ joined	03:09
flippo	It would have been so simple to rename a few of the commands, and change a few of the default options	03:09
bambams	I prefer it this way.	03:10
flippo	Because Pain is Virtuous	03:10
	← pitzips left	03:10
nevyn	fetch and pull being semantic inversions from mercurial is infuriating (regardless of who's got it right and wrong)	03:11
bambams	I think that it makes better sense as it is. And so what if it's a slightly steeper hill to climb. You've only got to climb it once. Then you can drive the tractor back down. Or something like that.	03:11
	Mercurial is infuriating all by itself.	03:11
	← jaequery left	03:12
madewokherd	I was probably using git for weeks as simply a form of write-only code storage (literally commit and push to one branch and nothing else) before I did anything fancy	03:12
	← spyroboy left	03:12
	→ Davey joined	03:12
	→ saurabh joined	03:13
	→ R1KRDO joined	03:15
	→ danishkhan joined	03:15
	→ pppaul joined	03:16
	← brum left	03:16
	→ bimble joined	03:16
	← WinstonSmith left	03:19
	→ h0bbit joined	03:20
	← jdunck left	03:21
	← hafeez3000 left	03:23
	← SBoudrias left	03:24
	→ WinstonSmith joined	03:25
	← barneybook left	03:25
	→ barneybook joined	03:25
	→ mmlac joined	03:26
	← simesy left	03:27
	→ barneybook\|2 joined	03:27
	← m0viefreak left	03:28
	← barneybook\|2 left	03:28
	→ barneybook\|2 joined	03:28
	← barneybook\|2 left	03:29
	→ barneybook\|2 joined	03:29
	← barneybook\|2 left	03:29
	→ pitzips joined	03:30
	→ barneybook\|2 joined	03:30
	← barneybook left	03:30
	→ thiago joined	03:30
	← etcetera left	03:31
	→ simesy joined	03:31
	→ hellopat joined	03:31
	← flippo left	03:32
	→ leeN joined	03:32
	→ themill joined	03:33
	← dreiss left	03:33
	→ etcetera joined	03:34
	→ dreiss joined	03:34
	← the_cybersphinx left	03:35
	← m0ltar left	03:37
	← rakl left	03:37
	← Dougie187 left	03:38
	← beakman left	03:38
	← nathanr left	03:39
	→ momomomomo joined	03:39
	← dreiss left	03:41
	← d3O left	03:41
	← fredlwm left	03:43
	← avinashrbhat1 left	03:43
	← hellopat left	03:43
	→ d3O joined	03:44
	← danishkhan left	03:44
	→ hongymagic joined	03:46
	← hongymagic left	03:46
	→ undecim joined	03:46
	→ hongymagic joined	03:46
	→ stliu joined	03:47
	← hongymagic left	03:47
	→ oakdog8 joined	03:47
	→ hongymagic joined	03:48
	← hongymagic left	03:48
	← oakdog8 left	03:48
	→ hongymagic joined	03:48
	← hongymagic left	03:48
	→ robbyoconnor joined	03:48
	← subfauna left	03:51
	← h0bbit left	03:51
	→ jwsong joined	03:51
	→ dreiss joined	03:53
	← bimble left	03:54
	→ engkur joined	03:54
	← etcetera left	03:55
	→ etcetera joined	03:57
	← BiggFREE left	03:57
	← wesside left	03:58
	← scorphus left	03:58
	→ hafeez3000 joined	03:58
	← mh` left	03:58
	→ undecim_ joined	03:59
	→ macmartine joined	04:00
	→ chuckhar_ joined	04:00
	← uidzer0 left	04:01
	→ beautifulmind joined	04:02
	← DrHouse\|WKSTN1 left	04:03
	← chuckharmston left	04:03
	← undecim left	04:03
	← barneybook\|2 left	04:03
	→ scorphus joined	04:03
	→ barneybook\|2 joined	04:04
	→ subhojit777 joined	04:04
	→ ProLoser joined	04:04
	← barneybook\|2 left	04:04
	← ProLoser left	04:04
	→ barneybook\|2 joined	04:04
	← barneybook\|2 left	04:05
	→ barneybook\|2 joined	04:05
	← beautifulmind left	04:05
	→ kerrick joined	04:05
	← etcetera left	04:06
	→ beautifulmind joined	04:07
	→ Katafalkas joined	04:07
	← barneybook\|2 left	04:07
	→ barneybook\|2 joined	04:08
	← barneybook\|2 left	04:08
	→ barneybook\|2 joined	04:09
	→ hwrd\|work joined	04:09
	← madewokherd left	04:10
	← barneybook\|2 left	04:10
	→ barneybook\|2 joined	04:10
	→ etcetera joined	04:10
	← aartist left	04:10
	→ rakl joined	04:10
	← barneybook\|2 left	04:11
	← KBme left	04:11
	→ barneybook\|2 joined	04:11
	← Davey left	04:12
	← chuckf left	04:12
	← Sosumi left	04:14
	← Mez left	04:14
	← barneybook\|2 left	04:15
	→ barneybook\|2 joined	04:15
	← miso1337 left	04:17
	← Markvilla left	04:17
	→ miso1337 joined	04:17
	← xiukun left	04:18
	→ chuckf joined	04:19
	← rking left	04:20
	← R1KRDO left	04:20
	→ rking joined	04:21
	→ KBme joined	04:21
	← muddo left	04:22
	← amesha left	04:23
	→ amesha joined	04:24
	← InvidFlower left	04:24
	← paddymahoney left	04:24
	→ Davey joined	04:24
	← pppaul left	04:25
	← hafeez3000 left	04:25
	→ Mez joined	04:26
	← Mez left	04:26
	→ Mez joined	04:26
	← Noldorin left	04:26
	← ryez left	04:27
	← ehsan left	04:27
	→ ARMIGER1 joined	04:27
ARMIGER1	Hello, all! Is there a way to amend a commit that was two commits ago without rebase?	04:28
	→ Sosumi joined	04:28
	← mmlac left	04:28
	← hwrd\|work left	04:30
	← rideh left	04:30
	→ pppaul joined	04:31
ARMIGER1	I haven't pushed to remote yet, but I can't seem to figure it out. Would checking out the specific commit and then calling git commit --amend do it?	04:32
bob2	no	04:32
Adlai	ARMIGER1: try rebase --interactive	04:32
	and rebase onto the commit before the one you want to amend	04:32
bob2	you're asking how to check out a commit, apply a commit but not commit it, make a change, commit, then reapply and commit a few other commits	04:33
	which is what rebase does	04:33
Adlai	then you can put your changes in the line after the commit to amend, and make the status 'fixup'	04:33
	→ stack joined	04:33
Adlai	man git-rebase	04:33
gitinfo	the git-rebase manpage is available at http://jk.gs/git-rebase.html	04:33
	→ irqq joined	04:34
ARMIGER1	Thank you. Looks like I've got some reading to do.	04:34
	→ jceb joined	04:34
Adlai	np :)	04:35
	→ barneybook joined	04:35
	→ PaulCapestany joined	04:35
	← stack_ left	04:36
	→ Tommy[D]_ joined	04:36
	→ ProLoser joined	04:37
	→ barneybook\|3 joined	04:39
	← barneybook\|2 left	04:39
	← Tommy[D] left	04:40
	→ stack_ joined	04:40
ARMIGER1	Adlai: So in my case, I would want go use git rebase -i HEAD-3 and go from there?	04:41
	← barneybook left	04:41
	← Mangix left	04:42
	← stack left	04:43
	→ OOPMan joined	04:44
	→ avinashrbhat joined	04:44
	← stack_ left	04:45
SamB	ARMIGER1: if get rebase -i brings up the wrong list of commits in your editor, you can abort	04:46
	→ stack joined	04:46
SamB	*git	04:46
	← orafu left	04:46
	→ barneybook joined	04:46
	← dreiss left	04:46
	SamB wonders if ARMIGER1 is a relative of VGER	04:47
ARMIGER1	SamB: Oops. Didn't catch that until you pointed it out. :P	04:47
SamB	ARMIGER1: I'm not certain it's made clear in the manpage anyway	04:47
	I just remember seeing it in the file that comes up	04:48
	→ barneybook\|2 joined	04:48
	← stoo_ left	04:48
	← etcetera left	04:49
	→ barneybook\|4 joined	04:49
	← barneybook\|3 left	04:49
	← thegorn left	04:49
ARMIGER1	SamB: VGER? The thing from Star Trek?	04:50
Adlai	ARMIGER1: sounds about right	04:51
	← ProLoser left	04:51
	← macmartine left	04:51
	→ w09x joined	04:52
	← barneybook left	04:52
	← barneybook\|2 left	04:53
	→ barneybook joined	04:53
ARMIGER1	Adlai: The git command or the VGER thing?	04:54
	→ ProLoser joined	04:55
Adlai	the git command	04:55
	I'd first make a commit that makes the changes you want to amend in	04:55
ARMIGER1	As in, make the commit on HEAD and then apply it to HEAD-3?	04:56
	→ aberrant joined	04:56
	← barneybook\|4 left	04:56
	→ arvind_khadri joined	04:58
	← arvind_khadri left	04:59
	→ raging_swine joined	04:59
	→ arvind_khadri joined	04:59
Adlai	yeah	04:59
	well	05:00
	which commit is the one you want to amend? HEAD-3?	05:00
	← mossplix left	05:01
	→ subfauna joined	05:02
	← taylorlapeyre left	05:03
	→ mossplix joined	05:04
ARMIGER1	If that's the one that's two commits ago, then yes. Right now I'm on the current working copy, wanting to amend a file to the commit before the last one.	05:06
	→ jaequery joined	05:06
	→ pretty_function joined	05:06
	→ hoschi joined	05:06
	← raging_swine left	05:08
	→ longbeach joined	05:10
	← LucasTizma left	05:10
	→ hwrd\|work joined	05:12
	security → megha	05:12
	→ xiukun joined	05:15
	→ hafeez3000 joined	05:15
	← bgerber left	05:15
	← w09x left	05:18
	→ szx joined	05:18
	→ bgerber joined	05:19
	→ hellopat joined	05:20
	→ ozette joined	05:23
	← pppaul left	05:23
	← hellopat left	05:25
	← mossplix left	05:26
	← diegoviola left	05:28
	→ momomomomo_ joined	05:28
	← gnarf left	05:28
	← momomomomo left	05:28
	momomomomo_ → momomomomo	05:28
	← momomomomo left	05:28
	← kvda left	05:28
	← jceb left	05:29
	→ pantsman joined	05:30
	← pantsman left	05:30
	→ pantsman joined	05:30
	← szx left	05:31
	→ soee joined	05:32
	→ mossplix joined	05:32
	← flavius left	05:33
	→ joaoh82 joined	05:36
	→ barneybook\|2 joined	05:37
	← hwrd\|work left	05:38
	← barneybook left	05:40
	→ barneybook\|3 joined	05:40
	← joaoh82 left	05:41
	← Adlai left	05:42
	→ FatDarrel joined	05:42
	← engkur left	05:42
	→ Adlai joined	05:43
	→ barneybook joined	05:43
	← barneybook\|2 left	05:43
	→ gnarf joined	05:44
	→ dmarkham_ joined	05:45
	← barneybook\|3 left	05:46
	→ engkur joined	05:46
	← OOPMan left	05:46
	→ Wulf joined	05:47
	→ barneybook\|2 joined	05:47
ARMIGER1	Got it! Thank you!	05:48
	← barneybook left	05:48
	← barneybook\|2 left	05:50
	← jwsong left	05:51
	← stack left	05:51
	← Chillance_ left	05:52
	← Saby left	05:52
	→ hwrd\|work joined	05:53
	→ stack joined	05:55
	→ trusktr joined	05:56
	→ vedant joined	05:56
	→ fRAiLtY- joined	05:58
	← ARMIGER1 left	05:59
	← rakl left	06:01
	← hafeez3000 left	06:02
	→ stack_ joined	06:02
	← mihamina left	06:03
	← stack left	06:04
	← ozette left	06:07
	← longbeach left	06:08
	→ cbreak-work joined	06:10
	→ ThomasLocke joined	06:10
	← SuperSteff left	06:12
	→ papegaaij joined	06:12
	← heytrav left	06:13
	→ heytrav joined	06:14
	→ wrouesnel joined	06:15
	← mgedmin left	06:15
	→ mgedmin joined	06:15
	→ dreiss joined	06:16
	→ lb1a joined	06:17
	← lb1a left	06:17
	→ lb1a joined	06:17
	← darinmorrison left	06:18
	→ SuperSteff joined	06:18
	← wrouesnel left	06:18
	← cyphase left	06:19
	→ wrouesnel joined	06:19
	→ hahuang65 joined	06:19
	→ engkur_ joined	06:20
	← beautifulmind left	06:22
	← miso1337 left	06:22
	← engkur left	06:22
	← Goplat left	06:24
	← vedant left	06:25
	→ flijten joined	06:25
	→ tvw joined	06:25
	→ h0bbit joined	06:26
	→ rob_w joined	06:26
	→ jdunck joined	06:28
	← SuperSteff left	06:28
	← jaequery left	06:28
	← Vile left	06:28
	→ warthog9 joined	06:29
	→ miso1337 joined	06:30
	→ ph^ joined	06:30
	← alexwhitman left	06:31
	→ Noldorin joined	06:31
	→ netiul joined	06:31
	→ alexwhitman joined	06:32
	→ cyphase joined	06:32
	← rob_w left	06:33
	→ GeertV joined	06:33
	→ alunduil joined	06:34
	← undecim_ left	06:35
	→ undecim_ joined	06:35
	← Noldorin left	06:36
	← AlexanderS left	06:38
	vals_ → tango_	06:40
	→ wsebastian joined	06:40
	→ MestreLion joined	06:40
	→ fsvend joined	06:41
	→ AlexanderS joined	06:41
MestreLion	how can I list the files in a given commit? git ls-files SHA1 returns no files	06:41
_ikke_	ls-files is for the index	06:41
	use ls-tree	06:41
	→ hafeez3000 joined	06:41
	← fsvend left	06:42
	→ Raging_Hog joined	06:42
	← thiago left	06:42
	→ fsvend joined	06:42
MestreLion	thank you _ikke_	06:43
	→ HNF joined	06:45
	→ marthinal joined	06:48
	← dreiss left	06:49
	→ tbosua joined	06:49
	→ dominik_ joined	06:50
	← Davey left	06:50
	← LekeFly left	06:50
	→ wannado joined	06:52
	→ rob_w joined	06:52
	→ jargon joined	06:53
	← fsvend left	06:55
	← Chaser left	06:57
	→ rubikcube_work joined	06:57
benje	hi how to reget the remote after deleting some file in local ?	07:00
	← Bilge left	07:00
	→ whowantstolivefo joined	07:00
	→ lavong joined	07:00
bob2	do you mean 'git checkout -f'	07:00
	→ Chinorro joined	07:01
	→ psoo joined	07:02
	QshelTier → Bombe	07:02
	← subfauna left	07:02
	→ anderse joined	07:03
_ikke_	benje: git is decentralized, which means you already have everything local (unless you deleted something from the .git directory)	07:03
benje	bob2, don't know maybe i have deleted some file and want to reget them	07:03
_ikke_	benje: git checkout <file>	07:03
benje	not deleted in .git but in rep	07:03
	rep directory	07:03
	→ The-Compiler joined	07:04
	← hwrd\|work left	07:04
_ikke_	right	07:04
	git checkout <file> is enough	07:04
benje	ok thanks the -f get all ?	07:05
	← normanrichards left	07:06
	→ subfauna joined	07:06
_ikke_	git checkout -f overwrites anything not comitted in the working directory	07:06
	← bambams left	07:08
	→ timj1 joined	07:09
benje	good this is that ;)	07:09
	thanks	07:09
	→ ErKa joined	07:10
	← FatDarrel left	07:11
	← timj left	07:12
	→ Chaser joined	07:12
	← dmarkham_ left	07:13
	→ fsvend joined	07:13
	→ eatyourgreenery joined	07:13
	→ suy joined	07:13
	← PaulCapestany left	07:14
	→ vedant joined	07:14
	← eatyourgreenery left	07:15
	← wrouesnel left	07:15
	→ wrouesnel joined	07:15
	→ DaisyTheCow joined	07:15
	→ LekeFly joined	07:17
	← LekeFly left	07:17
	→ LekeFly joined	07:18
	← pantsman left	07:18
	← rickruby left	07:18
	→ KevinSjoberg joined	07:18
	← hahuang65 left	07:19
	→ dmarkham_ joined	07:21
	← dmarkham_ left	07:21
	→ alorence joined	07:22
	→ kvda joined	07:22
	→ rath joined	07:22
	← mossplix left	07:22
	← d3O left	07:23
	→ mossplix joined	07:23
	→ d3O joined	07:23
	→ mitz joined	07:26
	← tjbiddle left	07:26
	→ Yougi joined	07:26
	← nihilista left	07:27
	← mitz_ left	07:28
	→ juril joined	07:29
	→ szx joined	07:30
	→ ozette joined	07:30
	→ nihilista joined	07:30
	← nihilista left	07:30
	← simesy left	07:31
	→ maxmanders joined	07:31
	← Chaser left	07:31
	→ nihilista joined	07:31
	← spyro left	07:31
	→ spyro joined	07:32
	← heytrav left	07:33
	→ _Qman joined	07:34
	→ heytrav joined	07:34
	→ elixir joined	07:34
	→ chimay joined	07:35
	← Gentlecat__ left	07:35
	← huoxito left	07:36
	→ joaoh82 joined	07:37
	← subfauna left	07:41
	← joaoh82 left	07:41
	→ Noldorin joined	07:41
	← infynyxx left	07:41
	→ nodweber joined	07:42
	← ProLoser left	07:42
	← nodweber left	07:42
	← MestreLion left	07:42
	→ infynyxx joined	07:42
	→ nodweber joined	07:42
	erkules_ → erkules	07:43
	← xiukun left	07:43
	→ szx` joined	07:43
	Joefish]off → Joefish	07:45
	← ddv left	07:45
	→ ddv joined	07:46
	← szx left	07:46
	← Noldorin left	07:46
The-Compiler	Heya! I had two git repos and merged one into a subfolder of another, like described in the "subtree merge" strategy on http://stackoverflow.com/a/1426163 - now my tree looks like this: http://paste.the-compiler.org/view/6cd50adf - the commit on the top (626b0be) is the merge commit I just did. What I'd like is to prefix every commit message in the right tree-branch (9175e27..f0056be) with "bep: " so I can	07:46
	clearly differentiate what was in the old repo. How would I do that?	07:46
	← PerlJam left	07:47
_ikke_	The-Compiler: man git filter-branch	07:47
gitinfo	The-Compiler: the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html	07:47
	→ Bilge joined	07:47
	→ PerlJam joined	07:47
	→ LeMike joined	07:47
	→ xiukun joined	07:48
	← pen left	07:48
	→ penguin1 joined	07:48
	← vedant left	07:48
	→ vedant joined	07:48
benje	bye	07:50
	← benje left	07:51
	← tevio left	07:51
The-Compiler	_ikke_: thought so, but how could I refer to all the commits coming from one side of a merge?	07:51
jast	I would redo the merge...	07:52
	1) checkout the 'right side' of the merge into a new branch, 2) filter everything in there, 3) recreate the merge	07:52
	→ alkino joined	07:53
	← Raging_Hog left	07:54
	→ bambams joined	07:55
	← bambams left	07:55
	→ bambams joined	07:55
	← marcdel left	07:57
	→ Sadeukko joined	07:57
	szx` → szx	07:58
	← psoo left	08:00
	← Milossh left	08:00
	→ psoo joined	08:00
The-Compiler	okay, I did a reset --hard to the commit before the merge, then checked out bep/master into bep, changed the messages, and merged bep into master. But now I still have the original unchanged commits in my log, as refs/original/refs/heads/bep - how would I get rid of them?	08:02
jast	rm -rf .git/refs/original (check if there's anything else in there you want to keep -- probably not)	08:03
	→ Anders_J joined	08:04
	← amesha left	08:05
	→ LimeBlast joined	08:05
	→ amesha joined	08:05
	← wrouesnel left	08:06
	→ wrouesnel joined	08:07
	→ forlin joined	08:07
	← mnathani left	08:07
	→ mnathani joined	08:08
	← ozette left	08:10
	→ ozette joined	08:10
	dharriga1 → dharrigan	08:11
	← dharrigan left	08:11
	→ dharrigan joined	08:11
The-Compiler	I still seem to have the original commits in my log, see http://paste.the-compiler.org/view/61002835	08:11
_ikke_	You seem to have merged the old and the new commits together	08:12
The-Compiler	I just noticed	08:12
	→ ddssc joined	08:13
	→ jnavila joined	08:14
	→ stephan_ joined	08:14
	→ eletuchy joined	08:14
	← rubikcube_work left	08:14
	← dthorman left	08:16
	→ Macaveli joined	08:16
	→ dthorman joined	08:16
	→ schemula joined	08:19
	← schemula left	08:19
	→ schemula joined	08:19
	← wdxsnail left	08:21
	← madduck left	08:22
The-Compiler	okay, worked now :)	08:23
	→ wdxsnail joined	08:23
	← vedant left	08:24
	→ vedant joined	08:24
	← dominik_ left	08:24
	← wdxsnail left	08:26
	→ wdxsnail joined	08:27
	← kvda left	08:27
	→ madduck joined	08:28
	← wdxsnail left	08:28
	→ WormDrink joined	08:29
	← WormDrink left	08:29
	→ WormDrink joined	08:29
	→ Raging_Hog joined	08:29
	→ subfauna joined	08:30
	→ Yuuhi``` joined	08:30
	→ wdxsnail joined	08:30
	← schemula left	08:30
	→ edgeuplink joined	08:30
	→ schemula joined	08:31
	← saurabh left	08:31
	← wdxsnail left	08:31
	← yacks left	08:31
	← Yuuhi`` left	08:32
	← tirengarfio left	08:33
	← delinquentme left	08:33
	→ dominik_ joined	08:33
	→ maxandersen joined	08:33
	← maxandersen left	08:33
	→ maxandersen joined	08:33
	→ tevio joined	08:35
	← spyro left	08:36
	← hafeez3000 left	08:36
	← _Vi left	08:36
	→ joaoh82 joined	08:37
	→ spyro joined	08:38
	→ wdxsnail joined	08:38
	← wdxsnail left	08:40
	← Yougi left	08:40
	← alunduil left	08:40
	→ Yougi joined	08:41
	→ achadwick joined	08:41
	→ hafeez3000 joined	08:41
	← joaoh82 left	08:41
	→ Thra11 joined	08:42
	→ khushildep joined	08:44
	→ mattymo joined	08:46
	→ guns joined	08:48
	→ marhaban joined	08:48
	→ DoniRS joined	08:50
	→ wdxsnail joined	08:51
	← wdxsnail left	08:51
	← jdunck left	08:52
	← elixir left	08:52
	← rob_w left	08:52
	← private left	08:53
	→ eijk joined	08:53
	→ wdxsnail joined	08:54
	← wdxsnail left	08:54
	← tevio left	08:55
esc	is ther any way to somehow structure the gitolite config so that repos are grouped into section in gitolite?	08:55
	→ tevio joined	08:55
	← themill left	08:56
	→ mmlac joined	08:57
	← alp82 left	08:58
	→ eijk_ joined	08:58
	← heytrav left	08:58
	← marhaban left	08:58
	→ tirengarfio joined	08:58
	→ GarethAdams joined	08:58
	→ wdxsnail joined	08:58
	← hoschi left	08:59
	→ heytrav joined	08:59
	← khushildep left	08:59
	→ rudak joined	09:00
PaulePanter	Hi. Is there a command for `git svn` to get branches and tags from Subversion after having already done `git svn clone URL` without any options/switches?	09:00
	→ howlymowly joined	09:00
	← eijk left	09:02
	← CareBear\ left	09:02
	MatthewWilkes → mwilkes\|away	09:03
	→ LekeFly- joined	09:05
	← DaisyTheCow left	09:05
	← ErKa left	09:06
	← tirengarfio left	09:06
	← dominik_ left	09:06
	← hafeez3000 left	09:07
	← ferdev left	09:08
	← jnavila left	09:08
	→ yacks joined	09:08
	← LekeFly- left	09:08
	← LekeFly left	09:09
	→ jnavila joined	09:09
	← heytrav left	09:10
	→ heytrav joined	09:10
	→ DaisyTheCow joined	09:11
sitaram	esc: could you explain what you mean? maybe a pseudo example of some syntax you think should work?	09:12
	→ tirengarfio joined	09:13
	→ joaoh82 joined	09:15
esc	like: section = "git-dev"	09:15
	http://bp.zetatech.org/43f95ec	09:16
	for example like this	09:16
	→ security joined	09:16
	← megha left	09:17
	→ thierryp joined	09:18
	← thierryp left	09:18
	→ thierryp joined	09:18
	→ hafeez3000 joined	09:19
iveqy	PaulePanter: I didn't know it was possible to use svn AND git branches	09:20
	mwilkes\|away → MatthewWilkes	09:20
	← heytrav left	09:20
	→ heytrav joined	09:21
	→ BadDesign joined	09:21
	← BadDesign left	09:21
	→ BadDesign joined	09:21
	← Gaga left	09:22
	→ Juri_Agarin joined	09:22
	→ testrest joined	09:22
	→ diegoviola joined	09:24
	→ ferdev joined	09:27
	← TheJH left	09:27
	→ TheJH joined	09:28
	← alansaul-laptop left	09:28
	← HNF left	09:30
	→ ozcanesen joined	09:30
	← juvenal left	09:30
	← scorphus left	09:32
PaulePanter	iveqy: http://git-scm.com/docs/git-svn	09:32
	→ marhaban joined	09:33
PaulePanter	http://www.jukie.net/bart/blog/svn-branches-in-git	09:33
	→ mastro joined	09:36
	← luc_ left	09:36
	← mastro left	09:36
	→ luc_ joined	09:36
	← infynyxx left	09:37
	→ mastro joined	09:38
	← tvw left	09:38
	→ OOPMan joined	09:38
	← mmlac left	09:38
sitaram	esc: what is that supposed to achieve?	09:38
	→ irqq_ joined	09:39
esc	to put the repo in it's own section	09:39
	but i think I need config gitweb.section	09:39
	← osmosis left	09:39
esc	but I need to enable that in GL_GITCONFIG_KEYS	09:40
	but can't find the damn config file on this server	09:40
sitaram	esc: oh this is about gitweb...	09:40
	← BadDesign left	09:40
esc	well, no, cgit	09:41
	but apperently, cgit can read gitweb settings	09:41
	→ dhruvasagar joined	09:41
	← irqq left	09:42
	← KevinSjoberg left	09:42
esc	sitaram: http://git.zx2c4.com/cgit/tree/cgitrc.5.txt and grep for "enable-git-config"	09:44
	→ zoresvit joined	09:46
	→ saurabh joined	09:48
	→ _Vi joined	09:51
	→ elixir joined	09:51
	→ Houl joined	09:51
jast	repo infrastruktur/scriptsammlung category = Infrastruktur	09:53
	two lines, of course... but that works for me with gitweb	09:53
	with gitweb's $projects_list_group_categories = 1	09:54
esc	i have it now	09:54
	$GL_GITCONFIG_KEYS = "gitweb\..*";	09:54
	in gitolite.rc	09:54
	← vedant left	09:54
jast	I have '.*' for that :)	09:55
esc	http://bp.zetatech.org/54e12e6	09:55
	http://git.zetatech.org/	09:55
	and then i get the appropriate sections	09:55
	← elixir left	09:55
esc	and setting enable-git-config=1 in the cgitrc	09:55
	← ph^ left	09:56
	← HeN left	09:56
	→ HeN joined	09:56
	→ ph^ joined	09:56
esc	of course it would be nice to set these things in a more conveninet fashion in the gitolite conf	09:57
	← Thra11 left	09:57
	← joaoh82 left	09:58
	← thyseus left	09:58
esc	jast: do you do 'config gitweb.category', or just 'category'	09:58
jast	just 'category'	09:59
esc	jast: thanks, I'll try it	10:00
sitaram	esc: and your description can be just this line: desc = foo bar baz	10:00
	→ thyseus joined	10:00
	← eletuchy left	10:00
sitaram	you don't need to use the old syntax, which required the repo name on the same line (that was designed to keep all descriptions in one place, which still works of course)	10:00
	→ the_cybersphinx joined	10:00
	→ eletuchy joined	10:00
	→ dkpi joined	10:01
jast	though I'm having trouble finding the part of the docs that describes 'desc' and 'category'	10:02
	→ geekbri joined	10:02
jast	for some reason I pretty much always have trouble finding the relevant section :)	10:02
sitaram	external.html	10:03
jast	oh	10:04
sitaram	gitweb is an external tool, one of the few that we ship some quasi support for	10:04
jast	I suggest making the master TOC more detailed in that section	10:04
	← dropdrive left	10:04
esc	i'll experiment a bit	10:05
sitaram	jast: you mean not just mention the UMASK but this also... hmm	10:05
jast	yeah	10:05
esc	sitaram: is there a way to set category for a bunch of repos?	10:05
	→ elixir joined	10:06
sitaram	jast: will do; thanks	10:06
esc	I am using v2.0.3-45-ga7a2732 do I still need the old syntax?	10:08
	→ infynyxx joined	10:08
	← The-Compiler left	10:08
sitaram	esc: oh yeah you do; the new syntax is 3.x	10:09
jast	oh, you're from the past, why didn't you say so :P	10:09
esc	sitaram: i'll see if I can squeeze in an upgrade sometime soon	10:09
	→ marhaban1 joined	10:10
	← elixir left	10:10
	← marhaban1 left	10:10
	→ dropdrive joined	10:11
	→ rob_w joined	10:11
	← ph^ left	10:11
	→ ph^ joined	10:11
	→ Thra11 joined	10:12
	← j0ran1 left	10:15
	→ j0ran joined	10:15
	← infynyxx left	10:17
	→ spuz joined	10:18
	→ tinti joined	10:18
	← miso1337 left	10:21
	← bartman` left	10:21
	→ bartman` joined	10:22
	→ khushildep joined	10:22
	→ khushildep_ joined	10:23
	→ AdrienBrault joined	10:25
	← khushildep_ left	10:26
	← khushildep left	10:27
	← AdrienBrault left	10:27
	→ Milossh joined	10:27
	← Thra11 left	10:27
	→ alansaul-laptop joined	10:27
	→ AdrienBrault joined	10:28
	← chimay left	10:29
	← subfauna left	10:30
	← marhaban left	10:31
	→ KevinSjoberg joined	10:32
	← stephan_ left	10:32
	← d3O left	10:32
	← yann\|work left	10:32
	→ vi7000 joined	10:34
	← lyang0 left	10:34
	← alansaul-laptop left	10:36
	← penguin1 left	10:36
	← testrest left	10:37
	← engkur_ left	10:37
	→ tvw joined	10:37
	→ penguin1 joined	10:38
esc	is there already a tool that will mirror github repositories, given a github username	10:39
	← hyperair left	10:39
bob2	#github i guess	10:40
jast	probably not... shouldn't be too hard to make, though	10:40
	I suspect you can do it with shell script	10:40
	← geekbri left	10:41
esc	not it shouldn't	10:41
	→ dvide joined	10:41
esc	an initial prototype for sure not	10:41
	→ rideh joined	10:41
esc	but something that is nicely configurable	10:41
	← yshh left	10:43
	→ mikecmpbll joined	10:43
	← KevinSjoberg left	10:44
	→ yann\|work joined	10:44
jast	untested: USER=jast; for repo in $(curl https://api.github.com/users/$USER/repos \| grep '^ "name":' \| cut -d\" -f4); do git clone https://github.com/$USER/$repo; don	10:44
	e	10:44
	obviously not the most robust parser ever	10:44
	→ Thra11 joined	10:44
	← jdyin left	10:45
	← Yougi left	10:45
	→ KevinSjoberg joined	10:45
	→ Yougi joined	10:46
	← dropdrive left	10:46
	monty- → monty	10:47
	→ marc_kdab_ joined	10:51
	→ dropdrive joined	10:51
	← vi7000 left	10:53
	← monty left	10:54
	← rob_w left	10:54
	→ jnavila_ joined	10:54
	← h2ero left	10:56
	← jnavila left	10:58
	← dhruvasagar left	10:58
	→ monty joined	10:58
	← guns left	10:59
	→ knightshade joined	10:59
	← maxmanders left	10:59
	→ elixir joined	11:00
	→ dhruvasagar joined	11:00
	← Sadeukko left	11:01
	← luc_ left	11:03
	→ maxmanders joined	11:04
	← jnavila_ left	11:04
	→ jnavila__ joined	11:04
	→ mrwalker joined	11:05
	→ rurufufuss joined	11:06
	→ rob_w joined	11:06
	← rideh left	11:06
	← Bilge left	11:07
esc	jast: nice	11:09
	← BtbN left	11:10
	← sonOfRa left	11:10
	← mrwalker left	11:11
	← elixir left	11:12
	← stliu left	11:12
	← KevinSjoberg left	11:14
	→ dsiganos joined	11:14
	→ jnavila joined	11:14
	← jnavila__ left	11:14
	← stack_ left	11:15
	← rudak left	11:15
	→ Henoxek joined	11:15
	→ stack joined	11:17
	→ KevinSjoberg joined	11:18
	→ Bilge joined	11:22
	← eletuchy left	11:23
	→ nkr joined	11:24
	→ eletuchy joined	11:24
	← jnavila left	11:24
	→ jnavila_ joined	11:24
	→ simesy joined	11:25
	→ BtbN joined	11:25
	→ sonOfRa joined	11:26
	→ joaoh82 joined	11:27
	← simesy left	11:27
	← flaguy48 left	11:28
	← kcj left	11:28
	← eletuchy left	11:29
	→ hoschi joined	11:29
	← reikalus1kka left	11:29
	→ elixir joined	11:29
	← mikecmpbll left	11:30
	→ Katafalk_ joined	11:31
	→ flaguy48 joined	11:32
	→ gerep joined	11:33
	→ sgambino joined	11:34
	← jnavila_ left	11:34
	→ jnavila_ joined	11:34
	← Katafalkas left	11:35
	← rob_w left	11:36
	→ gusnan joined	11:37
	→ bmac joined	11:37
	← heytrav left	11:39
	← Thra11 left	11:40
	← bmac left	11:40
	← szx left	11:41
	→ mikecmpbll joined	11:41
	→ chimay joined	11:41
	→ heytrav joined	11:42
	→ reikalus1kka joined	11:43
	→ jnavila__ joined	11:44
	← jnavila_ left	11:44
	← sgambino left	11:45
	← hafeez3000 left	11:45
	← j0ran left	11:47
	→ j0ran joined	11:47
	← johnthebear left	11:48
	← LeMike left	11:48
	→ br07 joined	11:51
	→ lyang0 joined	11:51
	→ rob_w joined	11:52
	→ jnavila joined	11:53
	→ juvenal joined	11:53
	→ LeMike joined	11:54
	→ bmac joined	11:54
	← jnavila left	11:54
	→ jnavila joined	11:54
	← jnavila__ left	11:54
	→ Thra11 joined	11:55
	→ hafeez3000 joined	11:55
	← xiukun left	11:57
	→ aitiba joined	11:59
	← tinti left	11:59
	← KnightHawk3 left	12:01
	→ jalama joined	12:03
	← jalama left	12:03
	→ jnavila_ joined	12:04
	← jnavila left	12:04
	← joaoh82 left	12:05
	← bmac left	12:05
	→ rideh joined	12:06
	→ simesy joined	12:07
	→ juanmabc joined	12:08
	← dkpi left	12:10
	→ mobileblue joined	12:11
	← simesy left	12:12
	→ cannonball joined	12:12
	← nicksloan left	12:12
	← marc_kdab_ left	12:12
	→ dkpi joined	12:13
	→ freeed joined	12:13
	→ angelsl joined	12:14
	← jnavila_ left	12:14
	→ jnavila__ joined	12:14
	→ e66 joined	12:14
	← Katafalk_ left	12:16
	→ Katafalkas joined	12:16
	← Macaveli left	12:16
	→ marc_kdab_ joined	12:17
	← mossplix left	12:18
	→ Siraris joined	12:20
	← maxandersen left	12:20
	→ thelucid joined	12:21
	← mikecmpbll left	12:21
	← monty left	12:22
	← diegoviola left	12:23
	← kaictl left	12:23
	→ kristofers joined	12:23
	← juanmabc left	12:24
	→ tmcmahon joined	12:24
	← jnavila__ left	12:24
	→ jnavila joined	12:24
	← saurabh left	12:25
	→ mossplix joined	12:27
	→ mikecmpbll joined	12:28
	← mobileblue left	12:28
	→ dreiss joined	12:28
	gitinfo set mode: +v	12:30
	← mossplix left	12:32
	MatthewWilkes → mwilkes\|away	12:32
thelucid	Anyone know how to read a a binary blog e.g. jpg from a bare repo for writing directly to file or more specifically a MongoDB	12:32
	→ ZaherDirkey joined	12:33
	← TheJH left	12:33
_ikke_	thelucid: man git cat-file	12:33
gitinfo	thelucid: the git-cat-file manpage is available at http://jk.gs/git-cat-file.html	12:33
_ikke_	You can perhaps pipe that to mongo	12:34
thelucid	Oh cool, that'll work with binary files?	12:34
_ikke_	Yeah, I don't see why not	12:34
ZaherDirkey	Hi, I am in branch b1 I need to merge it to the master without switch or checkout the master, and continue developing with b1, "merge it to master"?	12:34
	Is that possible?	12:34
thelucid	And from a bare repo?	12:35
_ikke_	ZaherDirkey: Nope	12:35
	ZaherDirkey: Merge requires a working directory	12:35
	ZaherDirkey: Specifically of the branch being merged into	12:35
	thelucid: Shouldn't be a problem	12:35
	→ jnavila_ joined	12:35
_ikke_	thelucid: You just have to set something like GIT_DIR	12:35
	← jnavila left	12:35
thelucid	_ikke_: Nice, thanks	12:36
	→ TheJH joined	12:36
	→ mossplix joined	12:36
thelucid	I'm using the Rugged Ruby bindings, ideally I want to do it with that but the documentation is a bit thin on the ground.	12:37
	← br07 left	12:37
_ikke_	Can't help you with that	12:37
	→ br07 joined	12:37
thelucid	_ikke_: No probs, thanks anyway.	12:38
	← e66 left	12:38
	→ brocken joined	12:39
	← keller left	12:41
	← brocken_ left	12:42
	→ kdave joined	12:42
	← yacks left	12:42
	→ sgambino joined	12:42
	← _Qman left	12:42
ZaherDirkey	_ikke_: thanks	12:43
	→ jnavila__ joined	12:44
	← jnavila_ left	12:44
	← br07 left	12:45
	→ bmac joined	12:45
kdave	hi, is it possible to get a 'oneline' pretty format with custom fileds? My goal is to display author (%an) in interactive rebase	12:46
jast	I don't think the format in interactive rebases can be customized at all	12:47
	→ JaredMiller joined	12:47
	→ jnavila joined	12:47
	← jnavila__ left	12:47
kdave	git-rebase--interactive uses --pretty=online	12:47
jast	that's precisely my point	12:47
	← JaredMiller left	12:47
kdave	if I could tweak oneline in my local .git that would suffice	12:47
jast	the meaning of --pretty=oneline is static	12:47
	you can't define new pretty formats except by manually specifying a format string	12:48
	(which rebase -i gives you no way of passing to it)	12:48
	→ KaiSforza joined	12:48
kdave	all pretty formats force multiline output afaics (except oneline)	12:48
jast	if you use your own format string you can do whatever	12:49
	← thelucid left	12:49
	← dreiss left	12:49
	← Chinorro left	12:49
kdave	aha, --pretty and --format can be combined together	12:50
	→ Element9 joined	12:50
	← jnavila left	12:50
	→ jnavila_ joined	12:50
	← tmcmahon left	12:51
	→ aartist joined	12:51
	→ thelucid joined	12:51
	← thelucid left	12:51
	← NiteRain left	12:51
	← LeMike left	12:52
	→ m8 joined	12:52
	→ JaredMiller1 joined	12:52
	→ natechan joined	12:53
	← JaredMiller1 left	12:53
	← pastime left	12:53
	→ jelly1 joined	12:54
	← jnavila_ left	12:54
	→ robertodecurnex joined	12:54
	→ jnavila__ joined	12:54
jelly1	Hi, I have a current repo where I keep my dotfiles, now I have a new install and my dotfiles are changed, but I want to use the dotfiles git repo again. How could I easily git init and add the repo and merge the changes	12:55
	I don't really know what to google	12:55
	I have tried git init; git remote add gitrepo; git fetch -> works	12:55
	→ maxandersen joined	12:55
	← maxandersen left	12:55
	→ maxandersen joined	12:55
	→ scampbell joined	12:55
	→ hellopat joined	12:55
	→ engkur joined	12:56
	→ thehandler joined	12:57
	→ jnavila joined	12:57
	← jnavila__ left	12:57
	← trusktr left	12:58
	→ fightback joined	12:58
	→ cybersphinx joined	12:59
	→ Rulin joined	12:59
	→ uidzer0 joined	12:59
nevyn	yeah but it's ugly. (you have two initial commits in the repo	12:59
	← mossplix left	12:59
	→ trusktr joined	13:00
	← the_cybersphinx left	13:00
	→ LeMike joined	13:00
jelly1	nevyn: hmm	13:01
	nevyn: is there a "nice" solution	13:01
nevyn	clone?	13:01
jelly1	else I'll just need to clone the repo and then cp the files back	13:01
	hehe	13:01
nevyn	you could do it by rebasing the initial onto the old history if that's even possible...	13:02
	→ R1KRDO joined	13:02
jelly1	hrrm sounds Like I rather setup a new repo	13:03
	since i can't clone into a nonempty directory	13:03
	→ Vile joined	13:03
	← natechan left	13:04
	← ddssc left	13:04
jelly1	I think i have the solution, nevyn thanks	13:04
	→ jnavila_ joined	13:04
	← jnavila left	13:04
	→ ddssc joined	13:04
	→ axl_ joined	13:05
	→ mossplix joined	13:06
	← rob_w left	13:06
	→ Sigma joined	13:06
	→ dir joined	13:07
	→ rob_w joined	13:08
	maxandersen → maxwebinar	13:08
	→ EthanCooler1 joined	13:08
	← jnavila_ left	13:09
	→ jnavila__ joined	13:09
	→ RaceCondition joined	13:09
	← EthanCooler left	13:09
	→ anonymuse joined	13:10
	→ jceb joined	13:10
	→ jnavila joined	13:10
	← jnavila__ left	13:11
	→ breakingthings joined	13:11
	→ yacks joined	13:11
	← security left	13:13
	← jelly1 left	13:14
	← elixir left	13:14
	← acharlieh left	13:14
	← Raging_Hog left	13:14
	→ jnavila_ joined	13:14
	← jnavila left	13:15
	→ elixir joined	13:15
	→ normanrichards joined	13:15
	→ texasmynsted joined	13:15
	→ AaronMT joined	13:15
	→ p3lim joined	13:15
	→ eijk joined	13:16
texasmynsted	I have a submodule in one branch and want to add it to another branch by submodule add but keep getting a message "f you want to reuse this local git directory instead of cloning again from"	13:16
	← normanrichards left	13:17
	→ trungqtran joined	13:17
texasmynsted	Or "A git directory for 'foo' is found locally with remote(s):	13:17
	etc.	13:17
cmn	it also tells you what to do	13:17
	→ rcu joined	13:17
texasmynsted	How do I say, yes, that is correct, please clone it from the remote anyway	13:17
	it says to select another name with the —name option	13:18
	← jnavila_ left	13:18
	→ jnavila__ joined	13:18
	← AdrienBrault left	13:18
texasmynsted	that does not make sense. I know of no other name, and would not want a different name as I want the name to be the same in both branches	13:18
	I do not know why I have never seen this problem before.	13:19
	← rcu left	13:19
	← Vile left	13:19
	← eijk_ left	13:19
texasmynsted	is the right answer to remove my local clone?	13:19
	texasmynsted tries that	13:19
	← pretty_function left	13:20
	→ AdrienBrault joined	13:20
texasmynsted	hmm. nope	13:20
	→ kogent joined	13:21
	→ Ryback__ joined	13:22
texasmynsted	It is not in cache because I rm ed from cache	13:22
	it is not on the filesystem because I rm-ed it from the filesystem	13:22
	→ soee_ joined	13:23
	← soee left	13:23
texasmynsted	I am not clear where it thinks it is and what would happen if I forced it to reuse the "existing" version	13:23
	→ Bass10 joined	13:23
	→ bytephilia joined	13:24
	→ pretty_function joined	13:24
	→ jnavila joined	13:24
	← jnavila__ left	13:25
texasmynsted	it is not in cache, not in .gitmodules, or .git/config	13:25
	→ mrwalker joined	13:25
	← Ryback__ left	13:26
	→ Ryback__ joined	13:26
	← KevinSjoberg left	13:26
	→ p0wn3d joined	13:26
	→ MattStratton joined	13:27
	→ nathanr joined	13:27
	← Ryback__ left	13:27
charon	!repro	13:27
gitinfo	Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session. This will help immensely with troubleshooting.	13:27
	→ johnflux joined	13:28
	→ rabellamy joined	13:28
	→ m0ltar joined	13:29
	← aitiba left	13:29
	→ soee__ joined	13:30
	→ thl0 joined	13:30
	← soee_ left	13:30
	← wrouesnel left	13:31
	← pretty_function left	13:31
	→ mmitchell joined	13:31
	→ jnavila_ joined	13:31
	← jnavila left	13:31
	← engkur left	13:31
	→ hwrd\|work joined	13:31
amcsi_work	I don't know, but I would try googling it. There seem to be a few hits	13:32
	→ mezner joined	13:32
	← maxmanders left	13:33
	→ maxmanders joined	13:33
	→ Chinorro joined	13:35
	→ jnavila__ joined	13:35
	← jnavila_ left	13:35
texasmynsted	http://pastebin.com/96ipnLJU	13:36
	← kdave left	13:37
	→ KevinSjoberg joined	13:37
texasmynsted	Yes there are two hits in google one in german one in english. Neither make a difference.	13:37
	← jnavila__ left	13:37
	→ jnavila joined	13:37
texasmynsted	The suggestion to edit .gitmodules files does not make sense because the entry is not there (as shown in my above paste)	13:37
	← trusktr left	13:38
texasmynsted	and 'submodule sync' would not and does not do anything since the entry is not there	13:38
	→ Voziv joined	13:38
	→ jnavila_ joined	13:39
	← jnavila left	13:39
texasmynsted	Sense the files are not in the index nor on the filesystem nor referenced in the .gitmodles, nor in the .git/config, I am clearly missing something so early this morning.	13:39
	hehe	13:39
	→ phantomtiger joined	13:40
texasmynsted	I guess I will try force	13:40
	← fightback left	13:40
	← subhojit777 left	13:40
	← amesha left	13:41
	← mossplix left	13:41
	mwilkes\|away → MatthewWilkes	13:41
	→ amesha joined	13:42
	→ aitiba joined	13:42
texasmynsted	hmm. I guess I need to re-read how git works internally	13:42
	← JSharpe_ left	13:43
	→ fightback joined	13:43
	→ jnavila__ joined	13:43
	← jnavila_ left	13:43
	→ thiago joined	13:44
	→ JSharpe joined	13:44
	← RaceCondition left	13:47
	← hwrd\|work left	13:47
	← hellopat left	13:48
	→ pmdarrow joined	13:48
	→ leehambley joined	13:48
leehambley	is it possible to completely ignore a certain path during rebasing? Perhaps not, I guess, because of all the hashing for tree integrity ?	13:48
	→ RaceCondition joined	13:48
leehambley	I'm in a Rails project where the `Gemfile.lock` conflitcts /every/ time, and the solutino is just to delete it, and call `bundle`	13:48
	→ jaequery joined	13:49
GarethAdams	if it conflicts every time then you're not committing enough	13:49
	→ mossplix joined	13:49
GarethAdams	;)	13:49
_ikke_	leehambley: You might want to use filter-branch then	13:49
leehambley	:-) we're a lot of people, on cutting edge Rails 4, so the head ref from the Rails dependendy changes every two hours :-)	13:49
	→ lacrymology joined	13:49
	← dhruvasagar left	13:49
	→ freestyler7 joined	13:50
	→ _Qman joined	13:50
leehambley	_ikke_: filter-branch looks perfect, but perhaps too much learning for the next half an hour :-)	13:50
	← Siraris left	13:50
GarethAdams	leehambley: deleting and calling `bundle` is not the same as fixing the conflicts however. You may end up upgrading gems that no one had planned to upgrade	13:50
leehambley	GarethAdams: I know, but it's close enough :-)	13:50
GarethAdams	which shouldn't be a problem if you're bleeding edge anyway but it's a consideration	13:51
leehambley	GarethAdams: right, certianly agree "all the newest things" is a terrible way to fix confifcts	13:52
GarethAdams	leehambley: so you're happy that different devs should all be working with different edge Rails commits?	13:52
leehambley	especially as it guarantees that this file will conflicts on /every/ other change	13:52
	← AdrienBrault left	13:52
	← dkpi left	13:53
GarethAdams	leehambley: sounds like you're saying that you don't want your (local) version of Rails to be affected by an update that someone else has made	13:53
	← DrShoggoth left	13:54
	← rabellamy left	13:54
	→ SuperDMP joined	13:54
leehambley	GarethAdams: no - but if I "bundle install", then the Gemfile in a commit from a week ago will be brouht upto date	13:54
	so it'll conflict with the 6.5 days old change, and the 6 day old change, and the next, etc	13:54
	→ yshh joined	13:55
	→ AdrienBrault joined	13:55
	← johnnyfuchs left	13:55
	← johnflux left	13:56
	→ JohnBat26 joined	13:56
	→ johnflux joined	13:56
	← johnflux left	13:56
	→ johnflux joined	13:56
	← soee__ left	13:57
	← Voziv left	13:57
	→ jxpx777 joined	13:59
	← ferdev left	14:01
	← rob_w left	14:01
	→ ferdev joined	14:02
	← ph^ left	14:02
	→ ehsan joined	14:03
	→ erichynds joined	14:03
	← rpg left	14:03
	→ aitiba_ joined	14:04
	← mossplix left	14:05
	→ hellopat joined	14:05
	→ jordilopezamat joined	14:06
	→ daniele91 joined	14:06
	← bytephilia left	14:06
	← juvenal left	14:06
	← aitiba left	14:06
	← mklappstuhl left	14:07
jordilopezamat	hi there! Nice to be here with you... i have a newbie question on the github web interface, is anyone willing to give me a hand ?	14:07
	:-)	14:07
lb1a	jordilopezamat: !hi	14:07
gitinfo	jordilopezamat: [!welcome] Welcome to #git, a place full of helpful gits. If you have a question, feel free to just go ahead and ask—somebody should answer shortly. For more info on this channel, see http://jk.gs/git/ Take backups (type !backup to learn how) before taking advice.	14:07
	→ cbot_ joined	14:08
lb1a	jordilopezamat: there is also #github	14:08
	→ Raging_Hog joined	14:08
jordilopezamat	great! thanks for, lb1a ... moving to	14:08
	→ dkpi joined	14:09
	→ YS-L joined	14:10
	← cagedwisdom left	14:10
	→ natechan joined	14:12
	← fightback left	14:12
	→ rob_w joined	14:13
	← mrwalker left	14:14
	→ acharlieh joined	14:14
	→ mossplix joined	14:14
	→ juvenal joined	14:16
	→ infynyxx_ joined	14:17
	← juvenal left	14:17
	→ spaceonline joined	14:17
	← phantomtiger left	14:17
	→ ojacobson joined	14:18
	← KevinSjoberg left	14:18
	→ ustunozgur joined	14:18
	→ m0viefreak joined	14:18
	→ juvenal joined	14:18
ustunozgur	Hi there, sometimes when I make some commits, and push and then when a coworker pulls the latest stuff, my newly added files are shown as untracked, and my changes are reverted with no commit that shows the revert. I have given a more detailed explanation of the problem here: http://stackoverflow.com/questions/16039320/my-coworkers-changes-remove-my-commits-in-git?noredirect=1#comment22883433_16039320	14:20
	so basically git log -p a does not show my changes in the file, while --follow does. Any ideas as to what might be the problem?	14:20
	← aberrant left	14:21
	→ oakdog8 joined	14:21
ojacobson	ustunozgur: I would look really closely at what 'git pull', 'git fetch', and 'git push' commands you're both running	14:22
	→ mklappstuhl joined	14:22
	→ geekbri joined	14:22
ojacobson	this looks a lot like someone using git push --force	14:23
ustunozgur	I suspected that he was running something like git pull --ours, but that doesn't seem to be the case.	14:23
	→ vedant joined	14:23
	← rchavik left	14:23
ustunozgur	hmm is there a way to find out of a forced git without looking at command history?	14:24
ojacobson	If your shared repo (the one you're both pushing to and fetching from) has reflogs turned on, yes	14:24
	otherwise, not really.	14:24
	← lb1a left	14:24
ojacobson	You'll see "(forced update)" during git fetch and git push	14:24
ustunozgur	OK, we are using github, I'll ask them to see if they have a clue.	14:25
	Thank you.	14:25
	→ ickyb0d joined	14:25
	← jast left	14:25
	← daniele91 left	14:25
	→ aitiba__ joined	14:25
	→ ndimitrij joined	14:26
	→ wesside joined	14:26
	← uidzer0 left	14:26
	→ aberrant joined	14:26
	→ jast joined	14:26
	← jast left	14:28
	← mikecmpbll left	14:28
	← aitiba_ left	14:28
	→ fightback joined	14:29
	← vedant left	14:29
	→ vedant joined	14:30
	→ KevinSjoberg joined	14:30
	→ Artpicre joined	14:30
	→ johnnyfuchs joined	14:31
	← jceb left	14:31
	→ jast joined	14:31
	← MattStratton left	14:32
	← pitzips left	14:32
	← Rulin left	14:32
	← wesside left	14:32
	← aitiba__ left	14:33
	← yshh left	14:33
	→ Rulin joined	14:33
	→ yshh joined	14:33
	→ MattStratton joined	14:33
	← Katafalkas left	14:34
	→ TheFox joined	14:34
	→ wesside joined	14:34
	→ juanmabc joined	14:35
	← jordilopezamat left	14:35
	← marc_kdab_ left	14:36
	→ marc_kdab_ joined	14:36
	KaiSforza → kaictl	14:36
	← Rulin left	14:37
	→ Rulin joined	14:38
astor	does pushing a tag automatically push the commit and all dependant objects?	14:38
vlt	Hello. what command could show me that I’m behind the remote branch?	14:38
ojacobson	vlt: 'git status' will tell you	14:39
	← rob_w left	14:39
ojacobson	'git fetch' will update the cached copy of the remote branch that 'git status' (and friends) uses	14:39
	as will git branch -vv	14:39
	← avinashrbhat left	14:40
	→ delinquentme joined	14:40
	→ rabellamy joined	14:40
	→ rpg joined	14:41
	→ trusktr joined	14:41
	→ jonatasnona joined	14:41
vlt	ojacobson: Hmmm … no, it doesn’t. git is version 1.7.9.5, `git log` locally shows a different last commit than remote. I did `git fetch`. Any idea?	14:42
	`git fetch` showed me that the remote branch has "updates".	14:43
ojacobson	!lol	14:43
gitinfo	A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all	14:43
ojacobson	that'll help you visualize what's going on	14:43
mgedmin	or gitk --all	14:43
ojacobson	'git log @{u}..HEAD' and 'git log HEAD..@{u}' will tell you about the commits on either side of the split	14:44
mgedmin	(for a GUI version)	14:44
ojacobson	(even if the split is one of the two being strictly ahead of the other)	14:44
lov	Ah, the joys of having a coworker that clones a repo to make a branch because he doesn't understand branching.	14:44
vlt	And I get output from `git diff origin/topicbranch` but not from `git status`	14:44
alansaul	Hey guys, is there anyway to rename a file to one which already exists, merging the two files, and maintaining their history?	14:46
vlt	ojacobson: lol shows me (what I already knew). I just don’t know why `git status` doesn’t tell me (anymore).	14:46
	→ aitiba__ joined	14:46
ojacobson	Does your tracking config say what you think it says?	14:46
	'git branch -vv' will tell you what branch each local branch is tracking	14:46
alansaul	i.e. I have A_a, and A, and I want to rename A_a to A and have A have the contents of A before and also merge in the functions from A_a	14:46
ojacobson	is your current branch tracking the upstream you were expecting to compare with?	14:46
	alansaul: cat A_a >> A; git rm A_a; git add A;	14:47
	← schemula left	14:47
ojacobson	i.e., do it the obvious way	14:47
	git will decide whether it thinks that's a rename of A_a to A or a change to A and the removal of A_a itself; either outcome is a merge hazard if others are working on either file	14:47
	← jaequery left	14:48
	← Yakko left	14:48
	→ _gmc joined	14:49
vlt	ojacobson: You were right: for that topicbranch there’s no origin branch in -vv.	14:49
	→ anderse_ joined	14:49
ojacobson	upgrade to 1.7.10 or later and use git branch -u to change the tracking info, or read man git-branch closely and use --set-upstream (the argument order is surprising)	14:50
gitinfo	the git-branch manpage is available at http://jk.gs/git-branch.html	14:50
	→ hierbabuena joined	14:50
ojacobson	or use git-config to manually set up tracking, but that's no fun :)	14:50
	maxwebinar → maxandersen	14:51
	← anderse left	14:51
	anderse_ → anderse	14:51
	→ mauke joined	14:51
	→ cagedwisdom joined	14:52
hierbabuena	Hi guys. One quick question: I used ''git update-index --assume-unchanged a_file'' on a file with changes. Then I did ''git stash''. The changes were staged. I would have expected the changes to be ignored, because of the first command. Is this by design or a bug?	14:53
	→ Siraris joined	14:54
	← RaceCondition left	14:54
	ChanServ set mode: +v	14:55
jast	--assume-unchanged is for telling git that it's not necessary to check the file for changes	14:55
	it's not meant to forbid git from checking the file	14:55
	← m0viefreak left	14:55
	← thiago left	14:55
	← leehambley left	14:55
	→ tga joined	14:55
jast	there are several operations that discard/ignore that flag in the index	14:55
tga	greetings	14:55
	<rant>does this kind of error message actually make sense for anyone?	14:56
	error: src refspec master does not match any.	14:56
_ikke_	yes	14:56
mauke	no	14:56
EugeneKay	Of course it does; we have the man pages tattooed on our buttocks.	14:56
tga	I don't mean recognizing it	14:56
EugeneKay	!repro your problem though	14:56
gitinfo	Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session. This will help immensely with troubleshooting.	14:56
	→ mrwalker joined	14:56
tga	I mean understanding what that things says	14:56
mauke	the answer to my question I haven't even asked yet: git show <commit1> <commit2> ...	14:56
cmn	tga: spell the branch properly	14:56
jast	yes. I know what it means.	14:56
tga	you get that when you try to push a new repo before commiting	14:57
_ikke_	yes	14:57
jast	is that so?	14:57
_ikke_	it says master doesn't exist yet	14:57
tga	it's the actual error text that I find horrible	14:57
	← OOPMan left	14:57
alansaul	ojacobson, Oh, so theres no real say of doing it I guess?	14:57
mauke	that's because it is horrible	14:57
cmn	it tells you exactly what's wrong, the source in the refspec dosen't match anything git knows about	14:57
alansaul	I mean, thats just a copy and paste effectively	14:57
_ikke_	mauke: </troll>	14:57
mauke	I'm not sure it's even english	14:58
ojacobson	alansaul: I'm not sure what you're looking for beyond that	14:58
EugeneKay	tga - feel free to raise your concern on the mailing list; git's porcelain is historically unfriendly to new users.	14:58
ojacobson	how would the solution you have in mind (no matter how fuzzy it is) differ from the obvious one, in observable ways?	14:58
EugeneKay	It has been getting improvements, but things like `git checkout` are just bad.	14:58
	← MattStratton left	14:58
alansaul	ojacobson, Doesn't matter that much, but it kind of means you can't revert "part of a file"	14:58
EugeneKay	Note that just whining isn't useful. Provide a suggestion to make it better	14:58
jast	even better, provide a patch	14:59
tga	hah alright, fair enough	14:59
	→ stliu joined	14:59
tga	git needs a humanization team	15:00
_ikke_	Hmm, maybe even add a hint when master isn't created yet	15:00
ojacobson	tga: you volunteering? :)	15:00
tga	I would if I would understand it better	15:00
ojacobson	(please please PLEASE make any new, more verbose "hints" disableable via git-config the way the existing advice.* entries are. :)	15:00
_ikke_	yes, that's why I meant	15:01
ojacobson	tga: you ran 'git push origin master', correct?	15:01
	→ _iron joined	15:01
	→ MattStratton joined	15:01
	→ ph^ joined	15:01
tga	yeah, with -u	15:01
	spyroboy_ → spyroboy	15:01
ojacobson	Is this a brand-new repository, or a clone of a brand-new repository?	15:01
	(I'm going to try to get you up to speed :)	15:02
	← netiul left	15:02
tga	init, add, (forget to commit), add remote, push	15:02
	← Raging_Hog left	15:03
tga	so that would make it a brand new one, no empty master	15:03
ojacobson	yeah, ok	15:03
	→ jaequery joined	15:03
	→ huoxito joined	15:03
ojacobson	So, "git push" takes a thing called a refspec, because you can push more than just branches	15:03
	you can push tags, groups of branches, or even non-branch "refs" (git's underlying naming mechanism, on which branches are built)	15:04
	A newly-created repository, until you run 'git commit' at least once, has no branches	15:04
	not even master	15:04
jast	afaics this error message comes from plumbing	15:04
ojacobson	(git "cheats" and allows HEAD to point to a nonexistant branch in that case, so that it can tell what branch the initial commit should create)	15:04
tga	right, so I told it to push that and it went huh, there is no master	15:04
	← ozette left	15:05
jast	so, replacing it would be a rather involved project	15:05
	← YS-L left	15:05
ojacobson	git push is complaining that the "refspec" master didn't match any local refs	15:05
	given that, what would a better message look like :)	15:05
	→ thiago joined	15:05
jast	(remote.c is used by plumbing, too, right?)	15:05
	→ rpg_ joined	15:06
	→ ZoeB joined	15:06
	← rpg left	15:06
tga	heh at least something like "Source refspec master doesn't exist locally"	15:06
_ikke_	Unknown local branch master?	15:06
tga	that sounds better actually	15:07
jast	neither would be correct	15:07
	example: 'Unknown local branch refs/remotes/*'	15:07
	← psoo left	15:07
jast	doesn't make sense	15:07
	in any case, can we change the message without breaking scripts?	15:07
	I sort of doubt it	15:07
_ikke_	Are those messages part of the plumbing 'api'?	15:08
	← cbreak-work left	15:08
_ikke_	I hope not actually	15:08
tga	"Unknown source master"	15:08
jast	I'm not sure	15:08
cmn	push is porcelain, you shouldn't be relying on the text	15:08
ojacobson	it'd have to use the branch name if and only if the resulting refspec's source was (A) not a wildcard refspec and (b) started with either refs/heads/ or +refs/heads/	15:08
	→ houser joined	15:08
jast	"Source refspec 'master' does not match any local refs" would be more readable and still accurate	15:08
_ikke_	rigt	15:08
	← ThomasLocke left	15:08
ojacobson	it still presupposes you care about refs rather than about the SCM-like abstraction built on them	15:09
	I propose that most users are in the latter set	15:09
	← ZoeB left	15:09
tga	Source 'master' does not match any local refs	15:09
	up to you to know what master is, you requested it	15:09
	← KevinSjoberg left	15:09
tga	oh, wait, I still said ref	15:09
jast	ojacobson: okay, but then you'd still get messages like 'Branch refs/remotes/origin/* does not exist' which make even less sense	15:09
cmn	'mater' can never be an actual ref, though; that's the short name for a branch	15:09
	→ Cromulent joined	15:10
ojacobson	it's a short name, yes	15:10
jast	and you can push notes, too, can you not?	15:10
cmn	you know when you only have the lhs, so you can tailor it to that	15:10
jast	notes are not branches :)	15:10
ojacobson	it's only a short name for a branch if the branch exists, given the way branch names are resolved :)	15:10
	→ aitiba_ joined	15:10
cmn	right	15:10
	which is why the word "match" is used	15:10
ojacobson	tga: "up to you to know what 'master' is" is, I think, sane	15:10
	← rath left	15:10
	→ madewokherd joined	15:10
ojacobson	Some people still manage to blow their feet off by creating local branches that shadow remote-tracking branches, but it's rarer than I would've expected	15:11
jast	how about something like: "Source 'master' not found"	15:11
ojacobson	(or branches with the same name as a tag)	15:11
	← MattStratton left	15:11
	← jude left	15:11
tga	jast: that is more informative	15:12
	→ alunduil joined	15:12
	→ leehambley joined	15:12
	← aitiba__ left	15:12
	← texasmynsted left	15:12
	→ jude joined	15:13
	→ MattStratton joined	15:13
	← houser left	15:13
	← juril left	15:13
	← Element9 left	15:13
	→ aitiba__ joined	15:14
	→ diverdude joined	15:14
	← infynyxx_ left	15:14
diverdude	I have a git repo on my clientmachine cloned from mydomainA.com. I now have a new server mydomainB.com where i made a --bare repo. My question is, is it possible to change my client repo to this new repo and upload files there?	15:14
	→ security joined	15:14
ojacobson	diverdude: push commits and refs to it as normal	15:15
	you can add it as a second remote, or re-point existing remotes to it; see man git-remote	15:15
gitinfo	the git-remote manpage is available at http://jk.gs/git-remote.html	15:15
	← JEEB left	15:15
diverdude	ojacobson, but how do i instruct the repo to push to the new server instead?	15:15
	← aitiba_ left	15:16
_ikke_	diverdude: git remote --set-url origin <new_url>	15:16
cmn	it's set-url, no dashes	15:16
	← vedant left	15:16
	← batrick left	15:16
diverdude	_ikke_, ok, and how do i see the current?	15:17
_ikke_	d'oh	15:17
	→ aitiba_ joined	15:17
_ikke_	diverdude: git remote show -n origin	15:17
	(-n so that it doesn't probe the remote)	15:17
cmn	it's easier to just do git remote -v	15:17
	but maybe that's just me	15:18
	→ Katafalkas joined	15:18
	← aitiba__ left	15:19
	← GeertV left	15:19
	→ v0n joined	15:20
	← stliu left	15:21
	← rabellamy left	15:21
	→ barneybook joined	15:21
	← _iron left	15:22
	← johnflux left	15:22
	← papegaaij left	15:22
	→ johnflux joined	15:22
	← johnflux left	15:22
	→ johnflux joined	15:22
	← fsvend left	15:23
	→ n00dle joined	15:24
	→ batrick joined	15:25
	→ infynyxx joined	15:26
	→ b1rkh0ff joined	15:26
	→ Dougie187 joined	15:27
	← h0bbit left	15:27
	← infynyxx left	15:27
	→ lavaman joined	15:27
	← leehambley left	15:28
	← Siraris left	15:29
	→ NiteRain joined	15:29
	→ nioncode joined	15:30
	← thl0 left	15:31
	← Cromulent left	15:32
	→ thl0 joined	15:32
	→ ThePawnBreak joined	15:32
	→ rath joined	15:32
	SeoZ → SeoZ-work[AWAY]	15:33
	← cagedwisdom left	15:33
	← rasto left	15:33
	← ustunozgur left	15:33
	→ br07 joined	15:33
	← delinquentme left	15:33
	← fightback left	15:33
	→ edude03 joined	15:34
	→ t-mart joined	15:34
	← aitiba_ left	15:35
	→ vedant joined	15:35
	→ thor7070 joined	15:35
	← ilteris_ left	15:36
	← spaceonline left	15:36
	→ aitiba_ joined	15:36
	← bmac left	15:36
	← aartist left	15:36
	← ddssc left	15:36
	← thierryp left	15:37
	← br07 left	15:37
	← thor7070 left	15:37
	→ JEEB joined	15:38
	← tbosua left	15:39
	← TheFox left	15:39
	← jdav_gone left	15:40
	← vedant left	15:41
	→ ilteris_ joined	15:42
	→ Donckers joined	15:42
	→ LekeFly joined	15:42
	→ thierryp joined	15:42
	← n00dle left	15:43
	→ aitiba__ joined	15:43
	→ mikecmpbll joined	15:43
	← aberrant left	15:43
	← soulshake left	15:45
	→ phantomtiger joined	15:45
	← aitiba__ left	15:46
	← aitiba_ left	15:46
	→ rasto joined	15:46
	→ aitiba__ joined	15:46
	→ mg983 joined	15:46
	→ aberrant joined	15:46
mg983	hi guys	15:46
	security → megha	15:46
	← gerep left	15:47
	← ilteris_ left	15:47
mg983	for some reason whenever i checkout from my repo it is setting permissions on my files to 665…. and i have to manually set them to 755 to use them properly… how can I resolve this	15:47
ojacobson	mg983: what does 'git ls-tree HEAD -- path/to/directory/containing/affected/files' print? gist/pastebin it	15:47
	I suspect the solution is 'chmod foo +x; git add foo; git commit' but the ls-tree output will help decide	15:48
	← ergZay left	15:48
mg983	http://pastebin.com/TsQfmi2v	15:48
	→ soulshake joined	15:48
ojacobson	Figured. They're checked in with the +x bit unset	15:48
mg983	there are also files underneath lib	15:48
ojacobson	so git-checkout will not set the +x bits	15:48
mg983	oy	15:49
	how do i fix that?	15:49
ojacobson	(php scripts don't need to be +x unless you're invoking them via exec(), though)	15:49
	(mod_php, in particular, will happily interpret them)	15:49
thiago	mg983: what was the permission on the file when you added/committed it?	15:49
	→ avinashrbhat joined	15:49
thiago	anyway, to fix: chmod +x filename && git add filename && git commit	15:49
	→ habva2013 joined	15:49
mg983	ok cool	15:50
	for each one?	15:50
	→ ilteris_ joined	15:50
	→ delinquentme joined	15:50
ojacobson	chmod the files you think should be +x, git add them all, and git commit them all at once at the end	15:50
	← Katafalkas left	15:50
thiago	well, chmod and add per file	15:50
	commit once	15:50
mg983	yea	15:50
ojacobson	no need for a separate commit per file	15:50
spyroboy	wait	15:50
mg983	gracias	15:50
spyroboy	git stores file permissions?	15:50
thiago	spyroboy: no, just one +x bit	15:50
spyroboy	oh I see	15:51
	so does it return a+x or u+x?	15:51
thiago	it's either 644 or 755	15:51
	→ Katafalkas joined	15:51
spyroboy	ah.	15:51
	I see.	15:51
thiago	it tests 0100 and then sets to 755 if that is set	15:51
spyroboy	I see.	15:51
	thanks for that.	15:51
	very useful	15:51
	← asedeno left	15:51
ojacobson	spyroboy: all of the +x bits get set, except for the ones masked out by your umask	15:51
mg983	you guys rule	15:52
thiago	right, files are created 0666 or 0777	15:52
	the umask removes the extra bits	15:52
	← ThePawnBreak left	15:52
	→ PaulCapestany joined	15:52
thiago	but if you do git ls-tree, you'll see git says 100644 or 100755	15:52
	← thierryp left	15:53
	→ starsinmypockets joined	15:54
	← zoresvit left	15:54
	← ilteris_ left	15:55
starsinmypockets	I'm starting development work on an existing webservice... does it make sense to add a dev branch to the existing live repo, then merge with the live branch, or is it smarter to use separate repos?	15:55
cmn	separate repos isn't going to bring you advantages unless you have some odd configurations	15:56
	← mitz left	15:57
	← MattStratton left	15:57
	→ dmarkham_ joined	15:57
	→ spaceonline joined	15:57
	→ rudak joined	15:58
	← Chinorro left	15:58
	← trungqtran left	15:58
	→ ilteris_ joined	15:58
	← Adlai left	15:58
	→ leehambley joined	15:59
PerlJam	cmn: it might if he doesn't have direct access to the webservice repo	15:59
cmn	but in that case it wouldn't even come up	15:59
	→ uidzer0 joined	16:00
starsinmypockets	Right, I do have access... it's common practice to have dev and production branches in the same repo, no?	16:00
	→ asedeno joined	16:00
PerlJam	starsinmypockets: sure	16:00
cmn	yes	16:00
	← thiago left	16:00
PerlJam	starsinmypockets: or, alternatively, it's weird to have them in different repos :)	16:00
	← hierbabuena left	16:01
cmn	especially if it's the branch you use for testing or similar	16:01
starsinmypockets	Right - ok	16:01
	Yeah exactly	16:01
cmn	you'd have topic branches in your own repo if there's too many people for everyone to push to the main repo and keep an overview (or if access is a problem)	16:01
	→ ThePawnBreak joined	16:01
mg983	thanks	16:02
	← acharlieh left	16:02
	← asedeno left	16:02
	→ asedeno joined	16:03
	← pmdarrow left	16:03
	← jxpx777 left	16:03
	← tureba left	16:04
	← kogent left	16:04
	← flijten left	16:05
	← GarethAdams left	16:05
	→ Spydemon joined	16:05
	← alorence left	16:06
	→ Siraris joined	16:06
	→ acharlieh joined	16:06
	→ tureba joined	16:06
Spydemon	Hi. We have some wrong commint in a remote branch. Is it a good idea to use `git revert` for undoing the wrong one?	16:06
	← alunduil left	16:06
	→ pmdarrow joined	16:07
	← jargon left	16:07
	← Remram left	16:07
diverdude	_ikke_, hmmm it seems to not work....it suggests this syntax: git remote set-url --add <name> <newurl>	16:07
	_ikke_, should i then delete the old url first? and how do i know the name of the old url?	16:08
cmn	Spydemon: that's what it's for	16:08
	→ Lucha joined	16:08
	← jaequery left	16:08
	← jnavila__ left	16:08
	← leehambley left	16:08
Spydemon	Ok cmn thanks. I was thinking that it's maybe not a good idea to do it on the remote one. ;-)	16:09
	→ MattStratton joined	16:09
	→ pastime joined	16:09
cmn	you won't do it on the remote one, the change happens in whatever branch you are	16:10
	→ kurkale6ka joined	16:10
	hpa → hpa_sick	16:10
	← rath left	16:11
	← erichynds left	16:12
	← delinquentme left	16:13
	← JohnBat26 left	16:13
	← SuperDMP left	16:14
	→ Remram joined	16:14
	← v0n left	16:14
	← LeMike left	16:14
	→ thiago joined	16:14
	→ marhaban joined	16:15
	← elixir left	16:15
	→ ozette joined	16:16
	j416_ → j416	16:16
	→ LeMike joined	16:16
	→ alainus joined	16:17
alainus	how do I set the shell prompt to show the current git branch ?	16:17
	→ hyperair joined	16:19
	← amesha left	16:19
	← _Qman left	16:19
	→ BadDesign joined	16:19
	← BadDesign left	16:19
	→ BadDesign joined	16:19
	→ rakl joined	16:19
	← Yougi left	16:19
	← mmitchell left	16:19
	→ amesha joined	16:19
	→ john2496 joined	16:20
	← LeMike left	16:20
	→ ergZay joined	16:21
	→ mathstuf joined	16:21
	← Watabou left	16:21
	→ boombatower joined	16:21
	← LekeFly left	16:22
mathstuf	hi, it seems as though with the latest git, MERGE_HEAD no longer exists after a merge; was this intentional?	16:24
	→ jordilopezamat joined	16:24
	→ mitz joined	16:24
cmn	it sounds just right	16:25
	MERGE_HEAD is there to assist git in doing a merge with several steps	16:25
	after the merge, it's meaningless	16:25
	← hyperair left	16:25
cmn	even if it did exist, you shouldn't be reading from it	16:25
	→ jaequery joined	16:26
	← rpg_ left	16:26
mathstuf	we were just testing for existence	16:26
	so the proper way to detect a merge commit in commit-msg hooks is to count parents?	16:26
	or is there some other way?	16:26
cmn	that doesn't happen after the merge	16:26
	← jaequery left	16:26
	→ a-l-e joined	16:27
	→ jaequery joined	16:27
mathstuf	hmm	16:27
	→ LeMike joined	16:28
	← LeMike left	16:28
	← nathanr left	16:28
	→ jdav_gone joined	16:30
	→ bytephilia joined	16:30
	→ infynyxx joined	16:30
	← angelsl left	16:31
	→ ultimateboy joined	16:31
cmn	are you checking during or after the merge?	16:31
	→ mmitchell joined	16:31
	← mikecmpbll left	16:31
	→ erichynds joined	16:31
mathstuf	in the commit-msg hook	16:32
ultimateboy	Question, is there an opposite of "git apply --no-add" to apply a patch without removing anything?	16:32
mathstuf	though theres some weirdness going on; ill see where it thinks it is	16:32
	Guest59049 → mf_	16:33
mathstuf	(weirdness == customg merge script with gerrit integration)	16:33
	← mitz left	16:33
	← yshh left	16:33
	→ yshh_ joined	16:34
	→ Davey joined	16:34
	← Sigma left	16:34
	→ JaredMiller joined	16:34
mathstuf	cmn: looks like we merge --no-commit followed by a --commit	16:35
	err commit -e	16:35
	← kurkale6ka left	16:35
	← fRAiLtY- left	16:35
	→ dankest joined	16:36
cmn	then it should be there, it's how git-commit knows to create a merge	16:36
	→ tjbiddle joined	16:38
	→ yshh joined	16:38
	← yshh_ left	16:38
	← Katafalkas left	16:39
mathstuf	seems it disappears before the commit-msg hook is called	16:39
	→ khushildep joined	16:40
	→ Shusshu joined	16:41
cmn	well, commit-msg is only meant for checking the contents of the message, maybe there's an assumption in the code that it won't need it	16:41
mathstuf	well, as a stopgap, ive edited the hook to let me work, but there probably should be a way to detect a merge commit in commit-msg (other than the first line matching '^Merge ')	16:42
	← spuz left	16:43
	→ lolmaus joined	16:43
	← john2496 left	16:43
cmn	are you sure you need that hook?	16:43
	← mtk left	16:44
lolmaus	I've filed an incomplete commit. Instead of doing another commit, how do i add to the previous one?	16:44
mathstuf	it checks the first line < 78 wide	16:44
bremner	lolmaus: man git commit --amend	16:44
gitinfo	lolmaus: the git-commit manpage is available at http://jk.gs/git-commit.html	16:44
	→ john2496 joined	16:44
cmn	so why do you need to know if it's a merge?	16:44
mathstuf	merge commits are exempt from that condition	16:44
	as are reverts	16:44
	← kristofers left	16:44
	← nodweber left	16:45
lolmaus	bremner, could you please give a hint which option will help me?	16:45
	→ etcetera joined	16:45
bremner	I could and I did.	16:45
lolmaus	Oh missed it, sorry	16:46
	bremner, thank you a lot	16:46
bremner	no problem	16:46
	← etcetera left	16:46
	← elnn left	16:46
	← suy left	16:47
	→ jdunck joined	16:47
	← ergZay left	16:47
	→ kristofers joined	16:47
	→ TeckniX joined	16:48
	→ danman joined	16:48
	← ZaherDirkey left	16:49
	→ hyperair joined	16:49
	← lolmaus left	16:49
	→ mikef joined	16:49
cmn	checking the message seems to make sense, as you want an exception for a particular type of message	16:49
	→ Noldorin joined	16:49
	← pastime left	16:49
ojacobson	I don't think there's a good mechanistic way to detect reverts that's morally superior to checking the message	16:50
	← mikef left	16:50
mathstuf	ojacobson: thats all it does right now	16:50
	but the merge detection was trying to be smart	16:50
ojacobson	you could, in principle, compare the diffs, but you'd have to examine all of history, and even then it'd be tricky	16:50
	merges at least leave footprints	16:50
mathstuf	'Merge files A and B together' is a plausible message	16:50
ojacobson	are you checking this on push, too, or only on commit?	16:50
mathstuf	commit	16:50
cmn	would you want to lift the restriction for a merge commit that doesn't have the canonical format?	16:51
	← acharlieh left	16:51
mathstuf	cmn: i doubt it	16:51
	→ mikef joined	16:51
cmn	then the message seems the more correct option	16:51
mathstuf	k	16:51
	← Shusshu left	16:52
	← irqq_ left	16:53
	← Noldorin left	16:54
	→ hwrd\|work joined	16:54
	→ JaredMiller1 joined	16:54
	← JaredMiller left	16:54
	→ Paul92 joined	16:56
	← kermit left	16:57
	→ kermit joined	16:57
	← hwrd\|work left	16:58
	→ etcetera joined	16:58
	→ wheezl joined	16:59
	→ hwrd\|work joined	16:59
	→ Gentlecat__ joined	16:59
	← danman left	17:00
	← anderse left	17:00
mg983	thiago: are you around?	17:01
	← hwrd\|work left	17:01
	← dkpi left	17:01
	→ mitz joined	17:02
	→ maqr joined	17:02
	← maxmanders left	17:02
	→ acharlieh joined	17:03
	← penguin1 left	17:03
mg983	ojacobson: are you around?	17:03
	← MattStratton left	17:03
mg983	ojacobson: thiago that solution doesn't seem to have helped :-\	17:03
	for some reason whenever i checkout from my repo it is setting permissions on my files to 665…. and i have to manually set them to 755 to use them properly… how can I resolve this	17:03
	→ Saby joined	17:04
	← Saby left	17:04
	→ Saby joined	17:04
	← yshh left	17:04
	→ leehambley joined	17:04
	← hyperair left	17:04
	← JaredMiller1 left	17:05
thiago	mg983: occupied	17:05
	→ dreiss joined	17:05
	→ hyperair joined	17:06
	→ trungqtran joined	17:06
	← tureba left	17:06
EugeneKay	mg983 - git tracks the +x bit; chmod +x && git add Foo.txt && git commit	17:06
mg983	I did that	17:06
	it doesn't work	17:06
EugeneKay	git update-index --chmod=+x -- Foo.txt	17:06
	Is this Windows? ;-)	17:06
mg983	no	17:06
	osx	17:06
EugeneKay	Odd. What's `git config core.filemode` say?	17:07
	← [1]JPeterson left	17:07
mg983	true	17:07
	← mikef left	17:07
mg983	i ran that update-index	17:07
	to no avail	17:07
EugeneKay	Really odd. I blame Steve Jobs.	17:07
	What does `git diff` say?	17:07
mg983	blank	17:07
	this is frustrating :\	17:08
EugeneKay	Hm. So the file is +x in the repo, but not when checking out	17:08
	← amesha left	17:08
	→ aksnot joined	17:08
EugeneKay	No clue on that - I blame per above ;-)	17:08
mg983	when files are created in linux do they automatically have a specified permission assigned to them based on the user?	17:08
	→ amesha joined	17:09
EugeneKay	umask is applied, or whatever the app creates it as	17:09
	← leehambley left	17:09
mathstuf	mg983: are you checking out on NTFS or FAT?	17:09
mg983	osx	17:09
	FAT	17:09
	→ tinti joined	17:09
EugeneKay	Oh, FAT, not HGFS	17:09
	THERES your problem.	17:09
	← mitz left	17:09
mg983	wait...	17:09
	i'm checking out on a linux box	17:09
	i'm sorry	17:09
	long day so far	17:09
EugeneKay	-_-	17:09
mg983	you have no idea how frustrating this is	17:10
	:-\	17:10
EugeneKay	ext3/4 then?	17:10
	→ kmels joined	17:10
mg983	how can I tell for certain	17:10
mathstuf	FAT doesnt support *nix permissions	17:10
EugeneKay	`mount`	17:10
	← ferdev left	17:10
	→ bmac joined	17:10
	→ tjbiddle_ joined	17:10
	→ tureba joined	17:10
	← gwty left	17:10
mg983	http://pastebin.com/0sBgZsM9	17:10
	that's the result on my checking out box	17:10
mathstuf	mg983: are you checking out on /dev/vzfs ?	17:11
EugeneKay	Ahhhh reiserfs.	17:11
	The official filesystem of wife-murdering.	17:11
mg983	mathstuf: i have no idea	17:12
	→ ferdev joined	17:12
mg983	i'm just a developer trying to get this to work	17:12
	→ anderse joined	17:12
mg983	so please excuse my ignorance	17:12
EugeneKay	I've never used git on reiser, so I dunno really	17:12
mathstuf	`pwd`	17:12
EugeneKay	It may not be handling it correctly	17:12
mg983	my pwd is /home/mg983/public_html/angular	17:13
EugeneKay	I would suggest that you post something to the mailing list - the important details are the fact that it's +x in the repo, reiserfs filesystem, and +x is not being set.	17:13
	← hyperair left	17:13
mg983	is there any way I can make 100% sure it is +x on the repo? maybe via github?	17:13
EugeneKay	I would not be surprised to leearn that this is a corner-case - reiser is not a popular filesystem nowadays.,	17:13
	→ LucasTizma joined	17:13
	→ madh888 joined	17:14
	← tjbiddle left	17:14
	tjbiddle_ → tjbiddle	17:14
	→ yshh joined	17:14
mg983	ugh	17:15
	→ nathanr joined	17:15
EugeneKay	Talk to your sysadmin about using a more-common FS, eg ext3/4	17:15
jast	I never ran into any problems like that while I used reiserfs, though admittedly that was quite some time ago	17:15
mg983	how can I be 100% certain it's properly set as +x in the repo?	17:15
	← bytephilia left	17:15
jast	git ls-tree HEAD	17:16
	normal files show as 100644, executable files show as 100755	17:16
mg983	i want these to be 755...	17:16
	they are 755	17:16
jast	(it doesn't actually mean '755' literally, just that the file is marked as executable)	17:17
mg983	all of the files when i ls -l	17:17
	are green with an asterisk	17:17
	what does that mean?	17:17
	→ marcdel joined	17:17
mg983	again excuse the ignorance	17:17
jast	that normally means that all the files are executable	17:17
	depends on your colour config, of course	17:17
mg983	ok	17:17
EugeneKay	Smells like reiser problems to me	17:17
	It'd be a shame if it.... MURDERED YOUR REPO	17:17
mg983	ok	17:18
EugeneKay	(sorry, I'll stop)	17:18
jast	that would be a ridiculous problem for reiser to have	17:18
	← chamunks left	17:18
jast	I've run a freaking _server_ on reiserfs (though not for long)	17:18
	→ chamunks joined	17:18
EugeneKay	Doesn't git do fs capability checking somewhere?	17:18
jast	all the common linux filesystems do all this correctly	17:19
mg983	i'd be happy to give someone root access to this box if they can investigate?	17:19
	→ elixir joined	17:19
mg983	this is so frustrating	17:19
jast	be careful with stuff like that	17:19
mg983	i know	17:19
	i'm at a loss	17:19
jast	I suppose I could spare a few minutes...	17:19
EugeneKay	I'm guessing this isn't a public repo	17:19
mg983	yeah it is	17:20
	← etcetera left	17:20
mg983	actually	17:20
	it's on github	17:20
EugeneKay	Linky?	17:20
mg983	1 sce	17:20
	sec	17:20
	https://github.com/matstars/angular_mat	17:20
jast	why the heck do you have mode=755 for /dev	17:20
	mm, I have that too	17:20
	makes no sense to me, but whatever :)	17:20
	→ etcetera joined	17:20
	→ ergZay joined	17:21
EugeneKay	What file is giving you the issues?	17:21
jast	well, if you want me to look, let me know	17:21
EugeneKay	I see add-todos.js changed to +x on the last commit	17:22
mg983	all of them	17:22
jast	why is a .js file supposed to be +x anyway? :}	17:22
	→ DonRichie joined	17:22
	← maslowbeer left	17:22
EugeneKay	Well, it looks right to me.	17:22
	What does `umask` say?	17:23
jast	not to me... it's insanity for all those files to be marked as +xc	17:23
grawity	jast: Because it's a program?	17:23
jast	*+x	17:23
	grawity: it's a web app	17:23
	the JS files are delivered to clients	17:23
EugeneKay	grawity - in this case it isn't; it's a webapp. Only the .php should be +x	17:23
jast	not even that	17:23
	few PHP installations are plain CGI	17:24
EugeneKay	Depends upon your php config, but usually yes.	17:24
	I require them to be +x anyway	17:24
jast	most use mod_php or type-/extension-based handlers	17:24
	→ Noldorin joined	17:24
EugeneKay	mg983 - <EugeneKay> What does `umask` say?	17:24
mg983	0002	17:24
EugeneKay	Hm. Now I'm stumped.	17:25
mg983	nuuu	17:25
	:)	17:25
EugeneKay	Lemme make up a test fs	17:25
	→ MattStratton joined	17:26
jast	I'm 99.9% certain that it's not because of reiser	17:26
	← lavong left	17:27
	→ kevwil joined	17:27
	→ daniel_- joined	17:27
	← tvw left	17:28
	→ jxpx777 joined	17:28
	← jxpx777 left	17:28
	→ jxpx777 joined	17:28
	← etcetera left	17:29
EugeneKay	Bah. And my kernel doesn't speak reiser.	17:31
	→ gwty joined	17:32
mg983	does it speak Bocce?	17:32
EugeneKay	No, but it does interface with dehumibifiers	17:33
	← mathstuf left	17:33
	→ ProLoser joined	17:33
	→ rpg joined	17:33
	→ motto joined	17:33
jast	so, it's not because of reiser :)	17:33
	← m8 left	17:35
	← forlin left	17:35
	← tinti left	17:35
	→ hyperair joined	17:36
	→ breck7 joined	17:36
	← starsinmypockets left	17:37
	← spaceonline left	17:37
mg983	jast = the man!	17:37
	→ Raging_Hog joined	17:37
	→ spaceonline joined	17:38
	→ tinti joined	17:38
	→ mitz joined	17:39
	→ maslowbeer joined	17:39
	→ mtk joined	17:40
	→ JPeterson joined	17:40
	→ cbreak-work joined	17:40
	← elixir left	17:41
	→ elixir_ joined	17:41
	← Davey left	17:41
	← Milossh left	17:42
	→ jwmann joined	17:45
	→ hwrd\|work joined	17:45
	→ goshawk joined	17:46
	← mitz left	17:47
	→ swiety joined	17:48
	→ JaredMiller joined	17:48
	→ vedant joined	17:48
	← johnflux left	17:49
	→ johnflux joined	17:49
	← johnflux left	17:49
	→ johnflux joined	17:49
	→ flavius joined	17:49
	← lacrymology left	17:49
	← hwrd\|work left	17:51
	MatthewWilkes → mwilkes\|away	17:53
	→ hwrd\|work joined	17:53
	← marc_kdab_ left	17:54
	← soulshake left	17:54
	→ treaki__ joined	17:54
	→ marc_kdab_ joined	17:55
	← mishok13 left	17:55
	→ soulshake joined	17:56
	→ maxmanders joined	17:56
	← nkr left	17:57
	→ m8 joined	17:57
	→ danishkhan joined	17:57
	← motto left	17:58
	← jordilopezamat left	17:59
	← treaki__ left	17:59
	→ jordilopezamat joined	17:59
	← JaredMiller left	17:59
	← m8 left	17:59
	← ferdev left	17:59
	→ treaki joined	18:00
	← oakdog8 left	18:00
	→ InvidFlower joined	18:00
	→ m8 joined	18:01
	→ tenmilestereo joined	18:02
	← hyperair left	18:03
	→ dhruvasagar joined	18:03
	→ hyperair joined	18:04
	→ motto joined	18:04
	← tevio left	18:04
	→ rpg_ joined	18:04
	→ opus joined	18:04
	← rpg_ left	18:04
	→ rpg_ joined	18:05
	← rpg left	18:05
	→ leehambley joined	18:05
	← alkino left	18:05
	→ Watabou joined	18:05
	← aksnot left	18:06
	← chimay left	18:06
	← Lucha left	18:06
	← Noldorin left	18:06
	← m8 left	18:06
	→ normanrichards joined	18:07
	→ oakdog8 joined	18:08
	← thiago left	18:08
	← mrwalker left	18:08
	← marc_kdab_ left	18:08
	→ SuperDMP joined	18:09
	→ marc_kdab_ joined	18:09
	← marc_kdab_ left	18:09
	→ marc_kdab_ joined	18:10
	← vedant left	18:10
	← habva2013 left	18:11
	→ nkr joined	18:11
	→ m8 joined	18:12
	← gumnos left	18:13
	← motto left	18:13
	→ exoon joined	18:14
	← plasticboy left	18:16
	← arvind_khadri left	18:17
	← ergZay left	18:17
	→ rednaks joined	18:17
	← GoshaZa left	18:19
	→ lacrymology joined	18:19
	→ jpwhiting joined	18:19
	→ plasticboy joined	18:19
	← plasticboy left	18:20
	→ alvaro_o joined	18:21
	← leehambley left	18:21
	← opus left	18:21
	→ plasticboy joined	18:21
	→ motto joined	18:21
	→ macabre joined	18:21
	→ GoshaZa joined	18:21
jpwhiting	hey all, I have two git repos kde-dev-scripts and kde-cmake-scripts	18:21
	← plasticboy left	18:21
	→ opus joined	18:21
jpwhiting	kde-cmake-scripts contains 61 commits all under cmake-utils prefix	18:21
	← opus left	18:22
	→ lifechamp joined	18:22
	→ plasticboy joined	18:22
jpwhiting	I want to merge the kde-cmake-scripts commits into kde-dev-scripts and keep file history (author, committer, etc.)	18:22
lifechamp	are there user permissions for r/w within individual repos, or how is that done typically using git?	18:22
	← m8 left	18:22
	← ProLoser left	18:22
lifechamp	well, write i suppose you don't have to accept the push, but what about read restriction?	18:22
	← juvenal left	18:23
jpwhiting	how can I do this? my first thought was to use git format-patch to create patches of the 61 commits, but I hear that will keep author, but not committer if I do git am each patch	18:23
_Vi	lifechamp, Reading permission is better to be managed with per-repo granularity.	18:23
	→ scorphus joined	18:23
bremner	jpwhiting: fetch one into the other and merge?	18:23
	→ mitz joined	18:23
	→ marni_ joined	18:23
lifechamp	_Vi you mean, have a separate repo for the group of users which should be allowed to read the entire repo? or?	18:24
_Vi	lifechamp, You should split repository into public and secret parts (possibly one being submodule of the other)	18:24
jpwhiting	bremner: just git remote add cmake kde-cmake-scripts, then git merge cmake/master master or something?	18:24
	from within a local clone of kde-dev-scripts	18:24
	← jbub left	18:24
	→ mikef joined	18:25
_Vi	lifechamp, Like in "Number of user groups (considering read access) == number of repositories".	18:25
bremner	jpwhiting: something along those lines.	18:25
	← przemoc left	18:25
_Vi	Is it possible to configure "git submodule update --init" to first check local collection of Git repositories instead of downloading (and preferrably put downloaded subrepos to the collection as well)?	18:26
	← macabre left	18:26
	→ macabre joined	18:26
	→ thasmo joined	18:27
	→ jbub joined	18:27
	← marni_ left	18:27
	→ przemoc joined	18:28
lifechamp	_Vi perfect explanation. Just curious tho, is it possible to restrict read access per file in a repo? was this a bad/diff design to implement if not? why?	18:28
	→ LekeFly joined	18:28
Paul92	if i have 2 branches: master and A, how could I replace the master branch with the branch A and remove master branch's content? basically, I want to delete the current master branch and make A become the new master branch.	18:28
_Vi	lifechamp, In one repository, Git can pack together different files (even ones that are not expected to share anything common).	18:29
	lifechamp, And fetching works using packs.	18:29
	← trungqtran left	18:29
_Vi	lifechamp, Maybe you have hack by restricting access per-ref, but I don't expect this to be reliable.	18:29
	s/you have/you can/	18:30
	← BrianR left	18:30
grawity	_Vi: Doesn't the git daemon create a new pack containing just the objects to be downloaded?	18:30
	← mitz left	18:30
_Vi	grawity, Maybe. But I would not rely on this for security.	18:30
	grawity, For example, some implementation may choose to give cached version with a little of extra objects...	18:31
	→ juvenal joined	18:31
	← mmitchell left	18:31
	→ arietis joined	18:31
	→ wrouesnel joined	18:32
_Vi	"Secret" part should be kept small (and users able to access the secret part should not be numerous).	18:32
	← juvenal left	18:32
grawity	the answer "it's impossible" is correct anyway, since the protocol cannot modify commits in any way, so no skipping files	18:32
_Vi	Why do you want to restrict read access to some files? Is it a config files with database password or something like that?	18:32
grawity	but I think there was something in the specs that forbade sending unrelated objects over the smart protocol	18:33
	← breck7 left	18:34
	→ thiago joined	18:34
milki	Paul92: delete the master branch and rename A as master	18:34
	→ ferdev joined	18:34
	← ehsan left	18:34
grawity	Paul92: alternatively, reset master to where A is, then delete A	18:35
_Vi	grawity, Maybe yes, maybe no... In any case fine grained read access is not a Git feature, as far as I understand. It should be done by a bit other means.	18:35
	→ hellz joined	18:36
Paul92	milki, grawity: ok, thank you	18:36
grawity	if you're on master right now, `git reset --hard A`	18:36
	→ bauruine joined	18:36
_Vi	So, is there some thing like "central local storage of submodules" (as a contrib script, for example)?	18:37
	← exoon left	18:37
	→ mmitchell joined	18:38
	← jbub left	18:38
	→ m8 joined	18:38
	← lausi left	18:38
	→ lausi joined	18:39
	← jordilopezamat left	18:39
	gitinfo set mode: +v	18:39
hellz	quetion from someone looking to switch to git....No team, its just me. Can I create a local repo and leave it at that? Do I need to run one on a server? Any things I'll be missing doing it that way? Figure I'll just back up (copy/tar) the entire project folder including .git stuff to dropbox or S3 or something. Any problems with this or suggestions for a better plan?	18:39
	← motto left	18:40
	← danishkhan left	18:40
imMute	hellz: you can run it completely locally. you lose nothing	18:40
	→ jbub joined	18:40
hellz	does that backup plan make sense?	18:40
imMute	could get by just synching the .git folder (unless you want to backup uncommitted things too)	18:41
	← Skaag left	18:41
	→ Skaag joined	18:41
imMute	there's probably scripts to pack the data more efficiently (or into a single file, etc)	18:41
xrq	hellz: it's easier to push to a server than make backups... unless you don't have any servers to push to	18:41
	← avinashrbhat left	18:42
xrq	hellz: from what you said though you could push to your dropbox or S3 account	18:42
bremner	hellz: tarballs of the whole thing is fine. Synching repos via dropbox is a bad idea	18:42
	or of course pushing to a git server is fine too, but not mandatory.	18:43
xrq	hellz: I would recommend pushing as often as possible unless you have a continuous backup system	18:43
hellz	right ok. i have a cron job to push to S3 nightly. I can add a tar easily to it.	18:43
	← mossplix left	18:43
	← normanrichards left	18:44
hellz	i eventually will setup a server for git, but dont want to stop development right now if I dont need to. if I can wait a month or so....thats what ill do.	18:44
	← St_Marx left	18:45
xrq	hellz: you can also push to dropbox on your own machine in that case	18:45
	→ trungqtran joined	18:45
robertodecurnex	hellz: yea, you can just have your local repo	18:46
	← mmitchell left	18:46
xrq	not that I recommend putting source code on dropbox :/	18:46
robertodecurnex	hellz: and push it in a year :P	18:46
	→ mmitchell joined	18:46
	← erichynds left	18:46
xrq	robertodecurnex: spoken like someone who's never lost a hard drive	18:46
robertodecurnex	hellz: better plan, there are some free, private, git services out there	18:47
	xrq: haha, yea, i was going to explain that	18:47
hellz	robertodecurnex: rather not put source on there.	18:48
robertodecurnex	hellz: so you can pick one of those, start there, and migrate to another one (yours or something like a paid github repo) whenever you can	18:48
	→ CareBear\ joined	18:48
	→ sq-one joined	18:48
	→ tvw joined	18:49
	→ St_Marx joined	18:49
hellz	ill figure out the serverside plan when i NEED to....just wanted to make sure that I can do everything I need to locally for now.	18:49
	→ RalfR joined	18:49
robertodecurnex	hellz: sure you can, you can lost your hardrive too (just trying to keep xrq happy :P)	18:50
	→ mossplix joined	18:50
xrq	just trying to avoid giving out bad advice :)	18:50
	← Raging_Hog left	18:50
	← mmitchell left	18:51
	→ system64 joined	18:51
	→ Chillance joined	18:51
	→ mmitchell joined	18:51
	← alansaul left	18:52
	→ snearch joined	18:52
	← ddv left	18:52
	→ ddv joined	18:52
	← phantomtiger left	18:52
	← edgeuplink left	18:54
	→ jdolitsky joined	18:54
	← amesha left	18:55
	→ diegoviola joined	18:55
	→ amesha joined	18:55
	→ tevio joined	18:56
	→ v0n joined	18:57
	→ mitz joined	18:58
	← jdolitsky left	18:59
	← megha left	19:00
	→ [1]habva2013 joined	19:00
	← _Vi left	19:01
	→ jdolitsky joined	19:02
	← Villadelfia left	19:03
	← SeySayux left	19:03
	← MattStratton left	19:03
	→ chisler joined	19:03
	→ alansaul joined	19:04
	← achadwick left	19:04
	← lacrymology left	19:04
	← hoschi left	19:04
	→ MattStratton joined	19:04
	→ private joined	19:04
	→ lacrymology joined	19:05
	← hwrd\|work left	19:05
	→ e66 joined	19:06
	→ cristi_ joined	19:06
	cristi_ → Guest31120	19:06
	← ThePawnBreak left	19:08
	← diegoviola left	19:08
	→ marn joined	19:08
	→ megha joined	19:09
	← marn left	19:09
	← Guest31120 left	19:09
	→ breck7 joined	19:10
	→ maksim_ joined	19:10
	→ ThePawnBreak joined	19:10
	→ danishkhan joined	19:10
	← megha left	19:10
	← snearch left	19:10
maksim_	Hi, is there a way to configure git to send an email everytime my repo gets cloned?	19:11
	→ brenelz joined	19:11
	→ nicklasbo joined	19:11
	← brenelz left	19:11
	→ thierryp joined	19:11
cehteh	maksim_: in some way for sure	19:12
	→ hwrd\|work joined	19:12
cehteh	not git directly, but with some tricks	19:12
	→ Gabri joined	19:13
	← kristofers left	19:13
maksim_	cehteh: cool :) can you guide to some resource about that?	19:13
	→ megha joined	19:13
	→ Chaser joined	19:13
cehteh	how do you serve the repo? git-daemon or http?	19:13
	→ franck joined	19:13
	franck → Guest89812	19:14
	← madh888 left	19:14
maksim_	i have one repo under git-daemon, and i have one using bitbucket.. if the later can do that job i would use it too	19:14
	← lacrymology left	19:14
cehteh	and its prolly not that trivial to do, as you need to figure out how to differentiate a update(fetch) from a clone	19:14
	i'd track the logs of git-daemon	19:15
	well looking into the git hooks might also be worthwhile, but i dont know if there are any hooks run by git-daemon	19:15
maksim_	mmm	19:16
	→ MistuhKurtz joined	19:16
cehteh	non-trivial but doable :)	19:17
	→ rcu joined	19:18
maksim_	cehteh: with try checking the doc, looks challenging :D	19:18
	← system64 left	19:18
	← thiago left	19:19
	← danishkhan left	19:19
	← mitz left	19:20
cehteh	why do you want that anyways .. i dont see much real use for it	19:21
cbreak	maksim_: you can't easily tell appart a clone and a fetch	19:21
	→ exoon joined	19:21
	← maksim_ left	19:22
cbreak	gitolite could probably allow you to hook read access	19:22
	but it's not straight forward	19:22
	→ juvenal joined	19:22
	→ pantsman joined	19:22
	← tevio left	19:22
	→ goodwin_i joined	19:23
	→ system64 joined	19:24
	← hafeez3000 left	19:24
Seveas	I don't think there is a ready-made point for this. Clone and fetch both simply connect and call upload-pack. You'll need to patch that command and somehow check whether it's a clone or fetch. Not sure it's as trivial as you think :)	19:24
	← jdolitsky left	19:26
cbreak	it's not possible in general	19:26
	since a clone is an init and a fetch	19:26
	if you could somehow listen in on the communication between the git instances you could try to detect if the client already has some blobs	19:27
Seveas	a clone is a fetch that has nothing for any branch, this may be detectable in upload-pack	19:27
cbreak	but that's definitely non-trivial	19:27
	← thierryp left	19:27
	← WormDrink left	19:28
	← axl_ left	19:28
Seveas	I don't know the details of the git protocol, but iirc it goes something like 'I have branch X at sha Y, give me what's newer' for all branches. X will be "nothing" for a clone.	19:28
	Y will be nothing*	19:28
cbreak	it's not about branches	19:29
	it's about hashes	19:29
	I think you try to establish a common front of known history	19:29
	(by sending tips of what you have)	19:29
Seveas	same difference, you will send "I have nothing"	19:30
	but yeah, looking at upload-pack.c, this still doesn't seem trivial :)	19:30
	→ ehsan joined	19:30
	← sniffells left	19:30
	← SuperDMP left	19:31
	→ sniffells joined	19:31
cbreak	I overread the protocol once.	19:32
	→ rpg joined	19:32
	→ axl_ joined	19:32
cbreak	the idea is simple. But I don't think it's easy to hook	19:32
	← rpg left	19:32
	← rpg_ left	19:32
	→ rpg joined	19:33
	← rcu left	19:33
	← marhaban left	19:34
	← jonatasnona left	19:34
	← marthinal left	19:34
cehteh	cbreak: yeah prolly not worth the trouble, needs some heuristics and some work	19:34
	→ h0bbit joined	19:35
	→ lacrymology joined	19:35
	→ rcu joined	19:36
	← a-l-e left	19:37
	← ehsan left	19:38
	→ jdolitsky joined	19:38
	→ ehsan joined	19:38
	→ dkpi joined	19:40
	→ SuperDMP joined	19:41
	→ SeySayux joined	19:41
	→ Villadelfia joined	19:42
	← Gentlecat__ left	19:43
	← Gabri left	19:43
	← lacrymology left	19:44
	← trusktr left	19:44
	→ kiumars joined	19:44
	← hwrd\|work left	19:45
	→ phantomtiger joined	19:45
	← tinti left	19:45
	← amesha left	19:45
	→ kiumars1234 joined	19:46
	← chisler left	19:46
	← kiumars1234 left	19:46
	→ amesha joined	19:46
	← undecim_ left	19:46
	→ DaisyThe1ow joined	19:46
	→ undecim_ joined	19:46
	→ Gentlecat joined	19:47
	→ chisler joined	19:47
	← troyt left	19:48
	← goodwin_i left	19:48
	← bmac left	19:48
	→ AdrienBr_ joined	19:49
	← v0n left	19:49
	← jpwhiting left	19:49
	→ v0n joined	19:49
	← kiumars left	19:50
	← AdrienBrault left	19:50
	→ troyt joined	19:50
	→ vedant joined	19:53
rking	How can one get the same list of commits that will be merged that the GitHub UI shows on a pull request screen?	19:54
	I'm trying to get better about not mashing the Big Green Button, since I think it's a bad practice to not actually run the code locally before merging.	19:54
	→ tinti joined	19:55
rking	But I made a mess by accidentally merging some junky commits that I would've seen if I had've gone to the UI.	19:55
	→ etcetera joined	19:56
	← Paul92 left	19:56
	← vedant left	19:56
	→ mitz joined	19:56
	→ vedant joined	19:56
	← jdolitsky left	19:57
	← rcu left	19:58
	← freestyler7 left	19:58
bremner	rking: I know nothing about github, but why can't you look at the proposed branch to merge with gitk or git log	19:58
	→ LeMike joined	19:59
	→ Parhs8 joined	20:00
	← InvidFlower left	20:00
	← Tommy[D]_ left	20:01
	← khushildep left	20:02
	← Parhs left	20:02
	← aitiba__ left	20:03
rking	bremner: I can try gitk, sure… but I just want a plain CLI way of doing it. What would the 'git log' invocation be?	20:03
	← mitz left	20:03
bremner	roughly, git log master..pr-branch	20:04
rking	I tried 'git log ..master' and 'master..' plus the three-dot versions, a few other things. Each one showed a diff that I don't see in the Pull Request UI.	20:04
	← DaisyThe1ow left	20:05
cbreak	rking: git log ..master means HEAD..master	20:05
	do you merge master?	20:05
	(your local master branch?)	20:05
	← MattStratton left	20:05
hellz	is there a different between 'git add -A' and 'git add -a' ? help shows only -A but talks about -a in other options.	20:06
	mwilkes\|away → MatthewWilkes	20:06
cbreak	hellz: -a is an option for git commit	20:07
	git commit -a is similar to git add -u && git commit	20:07
rking	cbreak: So you're asking if my local 'master' is the same as the one on GitHub?	20:07
	Like if I've made the local up to date with the remote?	20:08
hellz	ok got it	20:08
	→ Desproges joined	20:08
	→ jdolitsky joined	20:09
cbreak	rking: I am asking:	20:09
	if you want to know what you get by merging	20:09
	← chadi left	20:09
rking	Yes, that's exactly what I want to know. =)	20:09
cbreak	then why not git log ..thingyouwanttomerge	20:09
	like bremner said	20:09
	→ Milossh joined	20:10
rking	Well, that was my first guess, but it differs from what I see in the Pull Request web UI, which I'm trying to get away from.	20:10
	← nicklasbo left	20:10
	→ jaseemabid joined	20:10
	→ moondoggy joined	20:10
	→ penguin1 joined	20:10
	→ danishkhan joined	20:11
	→ lacrymology joined	20:11
cbreak	maybe that pull request thing is broken? :)	20:11
	because what I said will tell you the absolute truth.	20:11
	→ drizzd joined	20:12
jaseemabid	How do I get the manual on a config variable? I'm looking for something like 'git --help config pull.rebase'	20:12
	→ chadi joined	20:13
	→ hwrd\|work joined	20:13
cbreak	jaseemabid: man git-config	20:13
gitinfo	jaseemabid: the git-config manpage is available at http://jk.gs/git-config.html	20:13
cbreak	just search in there	20:13
	→ franksNbeans joined	20:13
	→ ProLoser joined	20:13
rking	cbreak: OK, thanks. I'll just start comparing what I see there to what I see on the CLI, to what I see as a result.	20:14
	← mmitchell left	20:14
	← arietis left	20:14
rking	bremner: Thanks also. ☺	20:14
	→ DaisyThe1ow joined	20:14
	← hwrd\|work left	20:14
	→ hwrd\|work joined	20:15
hellz	say i have a bunch of binary resources in a project (images, audio, etc). when i make changes and commit them, it replaces the file. But does the original still live somewhere in the .git? if so, anyway to permanently get rid of them so I don't cause huge bloat to my repo?	20:16
cbreak	hellz: it will be in the repository for ever	20:17
	and it is impossible to rip it out without destroying the history	20:17
hellz	well damn	20:17
	← AlexanderS left	20:17
cbreak	that's why the usual recommendation is to not store big binary files or fast changing binary files in a git repository	20:17
	store them in a submodule, a slave repository, with git annex, ...	20:18
	→ AlexanderS joined	20:18
cbreak	naturally, git is fully capable to rewrite history so that it appears as if those files never existed	20:18
	← etcetera left	20:18
	→ wrouesnel1 joined	20:19
cbreak	but this will create new history, and people who already have the old will have to work to get to the new	20:19
drizzd	cbreak: does it work gracefully if the submodule history is rewritten?	20:19
	← danishkhan left	20:19
drizzd	I mean will git submodule update complain or just do it?	20:19
hellz	cbreak: when you say 'rewrite history' do you mean delete that history and those binary files?	20:20
	→ mmitchell joined	20:20
	→ __david__ joined	20:20
	← Watabou left	20:20
	gitinfo set mode: +v	20:20
	← elixir_ left	20:21
__david__	hey folks, I created a feature branch A from master, and then a feature branch B based on A	20:21
	now I would like to merge both A and B into master as squashed commits	20:21
bremner	hellz: man git-filter-branch	20:21
gitinfo	hellz: the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html	20:21
	→ Clint_Beancake joined	20:22
	→ Katafalkas joined	20:22
dr_lepper	__david__: have B and A diverged?	20:22
	→ lb_ joined	20:22
dr_lepper	or is B a direct descendant of A?	20:22
cbreak	drizzd: submodules don't really care about history, they just bind to a commit	20:22
	← LekeFly left	20:22
__david__	dr_lepper, A is fully merged into B	20:23
	so I guess that makes it a direct descendent?	20:23
	→ kit joined	20:23
	lb_ → lb1a	20:23
dr_lepper	but you want to have them as two separate squashed commits in master?	20:23
__david__	exactly	20:23
hellz	this contradicts ->it will be in the repository for ever	20:23
__david__	essentially I'd like to have a `git diff A..B` as a squashed commit in master	20:24
	→ opus joined	20:24
	← BtbN left	20:24
	← sonOfRa left	20:24
bremner	hellz: no, git-filter-branch destroys history because it changes the hash of every commit	20:24
hellz	so git-filter-branch allows actually deleting some commit	20:24
	so if i keep my binary commits all together and put no src in there, i can remove that commit?	20:24
dr_lepper	__david__: you can try using git rebase -i to squash commits inside the feature branches, and then merging them into master	20:25
hellz	oh oh, it destroys the hash but not the file data?	20:25
dr_lepper	or just merge --squash them into master individually	20:25
bremner	hellz: see !rewrite	20:25
gitinfo	hellz: [!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to). If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://git-scm.com/docs/git-rebase.html#_recovering_from_upstream_rebase	20:25
hellz	i get it. ty.	20:26
	→ Davey joined	20:26
SuperDMP	does Django have anything built-in that will have a go at turning rubbish into date objects? at present I have http://dpaste.org/Ouo4u/, but if something nicer exists already I'd like to know	20:26
	→ BtbN joined	20:26
SuperDMP	sorry, wrong #	20:26
	→ Watabou joined	20:26
	→ sonOfRa joined	20:26
SamB	there are occasionally problems that make rewriting public history the least bad thing to do	20:26
	→ etcetera joined	20:26
__david__	dr_lepper, the latter is what I tried, and I got merge conflicts. Seems like git wants to merge things "twice" because it doesn't realize the squashed commit contains stuff it also has in its branch	20:27
SamB	like if you accidentally pushed huge binaries or something	20:27
__david__	I'll look at the rebase	20:27
	→ thirtythreeforty joined	20:27
thirtythreeforty	Hi. What is the appropriate way to handle a 'next' branch in a public project?	20:27
	Is it occasionally rebased?	20:27
SamB	thirtythreeforty: rebasing is probably a bad idea if "next" means anything like I think it does	20:28
thirtythreeforty	Err, I guess I meant 'reset', SamB.	20:28
	Actually, maybe it would be better if you told me how next is supposed to be used.	20:29
SamB	well, what were you thinking it would be meant to hold?	20:29
	← kit left	20:30
thirtythreeforty	I think it is a place that starts off pointing at master, then additional branches are merged into it.	20:31
	→ mitz joined	20:32
	← etcetera left	20:32
SamB	oh, that sounds like that one branch of git ... was it called "pu"?	20:32
	→ cansf joined	20:32
cbreak	next is just a name	20:32
	→ MattStratton joined	20:33
cbreak	branch names carry no inherent notion of stability	20:33
SamB	yes	20:33
thirtythreeforty	Yes. pu basically points to whatever, and you're not supposed to base work off it.	20:33
SamB	yeah	20:33
__david__	how can I get an "Automatic cherry-pick" failed in rebase -i ? I'm just squashing all the commits (except the first)	20:33
thirtythreeforty	I know, but "traditionally" how is it done, cbreak?	20:33
cbreak	depends on who you ask...	20:33
	I don't really mind the occasional history rewriting :)	20:34
thirtythreeforty	Maybe can you explain "Large-Merging Workflows" from http://git-scm.com/book/en/Distributed-Git-Maintaining-a-Project to me?	20:34
SamB	there's something that happens when you clone git.git, isn't there, that informs git to expect this?	20:35
	→ tsp joined	20:35
	← tsp left	20:35
	→ tsp joined	20:35
cbreak	SamB: downloading the readme file? :)	20:35
SamB	cbreak: no, that's for users	20:35
	← ph^ left	20:35
cbreak	I am not aware of anything else.	20:36
	→ diamonds joined	20:36
	→ ph^ joined	20:36
diamonds	I stashed & rebased then tried to git stash pop	20:36
	← cansf left	20:36
diamonds	got a couple issues:	20:36
thirtythreeforty	cbreak, I don't understand what happens to next (in that page's example) when some topics are merged to master. Does next now point somewhere entirely different that may/may not be a descendent of the old next?	20:36
tsp	How can I explicitly set git's editor? I tried setting both EDITOR and VISUAL to ed, but still get error: cannot run vi: No such file or directory	20:36
thirtythreeforty	What does he mean by "next is rebased occasionally"?	20:37
	tsp, set core.editor	20:37
	http://stackoverflow.com/questions/2596805/how-do-i-make-git-use-the-editor-of-my-choice-for-commits	20:37
cbreak	thirtythreeforty: no idea. but in general, the only branch affected by a merge is the current branch	20:37
SamB	cbreak: I don't even see 'pu' mentioned in README	20:37
diamonds	http://pastie.org/pastes/7613366/text	20:37
	then it has "unmerged paths" in git status..	20:38
	I didn't attempt a merge, per se	20:38
	so I can unstage or add .. ??	20:38
tsp	Thanks, that fixed it	20:38
diamonds	why is it staged, I just wanted to apply stashed changes, not stage anything	20:38
thirtythreeforty	cbreak, oops... I was unclear. When the old next is not wanted, do we just git branch -D it and then push up a new branch? Or something else?	20:38
SamB	gitworkflows(7) seems to talk about this	20:38
offby1	diamonds: that actually does a merge behind the scenes	20:39
cbreak	thirtythreeforty: you can do what ever you want. force push, delete, ...	20:39
	← BadDesign left	20:39
arand_	diamonds: well, it does merge the stashed changes with the working tree.	20:39
	← troyt left	20:39
drizzd	tsp: maybe you did not export the environment variable?	20:39
diamonds	yes	20:39
thirtythreeforty	cbreak, I know you "can" but I want to know what is considered standard or good practice	20:39
	→ P4C0 joined	20:39
diamonds	arand_ offby1 but then it stages it (?)	20:39
thirtythreeforty	SamB, thanks for that link	20:39
tsp	drizzd: I'm using cygwin's git, it's doing some odd things	20:39
cbreak	as said, no idea :)	20:39
P4C0	Hello, how can I merge changes from a branch but interactive? I mean, I only need part of a commit	20:40
diamonds	I am now forced to commit or reset HEAD the file?	20:40
SamB	I found it by grepping for "'pu'" (outer quotes were for the shell) in git's tree	20:40
diamonds	AAAAAAAAAAAARTGGGHHh	20:40
arand_	diamonds: it stages things that was staged when you stashed it.	20:40
	→ mrwalker joined	20:41
arand_	afaik	20:41
diamonds	no matter how many times I git stash pop, fix, git add . && git commit I still have 2 things in my stash stack	20:41
SamB	diamonds: you can unstage, or stash again, or whatever	20:41
diamonds	SamB: but staging means resetting back to head, no?	20:41
	← rpetre left	20:41
SamB	diamonds: oh, try git stash drop	20:41
diamonds	which is not my stashed version	20:41
	well I don't know if the changes were applied	20:41
	← mitz left	20:41
SamB	I meant once you've decided you're finished with that stash	20:42
	← chadi left	20:42
diamonds	the most irritating thing is the fact that it's a single blank line causing this :/	20:42
	SamB: how do I know if the stash was applied?	20:42
SamB	I usually look at things in magit	20:42
	but I guess you can also use gitk	20:42
	or git gui	20:42
	→ troyt joined	20:43
	← rpg left	20:43
diamonds	hmm...	20:43
	how view stash in gitk?	20:43
	→ chadi joined	20:43
	← ThePawnBreak left	20:43
SamB	oh, to see what the stash was I use "git stash show" or whatever it's called	20:43
cbreak	P4C0: merging doesn't merge a commit	20:44
	it merges history	20:44
	merging only part of history makes zero sense	20:44
	maybe you want cherry-pick?	20:44
P4C0	cbreak: maybe sherry pick without commit or similar?	20:44
	→ slacx joined	20:44
	SamB wonders why it's called "cherry" pick	20:44
SamB	(why cherries?)	20:44
	← MattStratton left	20:44
	← hellz left	20:45
P4C0	cbreak: then thing is that i have everything on a temporal branch, and i need to split it in several branches, I don't really need the history just the changes	20:45
cbreak	P4C0: yes... you can use -n if you want	20:45
	← infynyxx left	20:45
cbreak	or amend the commit afterwards	20:45
SamB	darn, it's not in the dictionary	20:45
P4C0	cbreak: thanks	20:45
	→ thiago joined	20:45
	← mossplix left	20:45
thirtythreeforty	cbreak, "the integration branch next may optionally be rewound and rebuilt from the tip of master using the surviving topics on next."	20:46
P4C0	The thing is that I started coding and I have everything in tmp branch, tmp branch implements fix1, fix2 and fix3, what I want is to have a fix1 branch, fix2 branch and fix3 branch (and delete tmp)	20:46
thirtythreeforty	--gitworkflows(7)	20:46
	So that answers it :)	20:46
cbreak	P4C0: sounds dumb.	20:47
arand_	SamB: https://en.wiktionary.org/wiki/cherry-pick at least a few notes ;)	20:47
cbreak	P4C0: you should make a new topic branch for each :)	20:47
	← jaseemabid left	20:47
P4C0	cbreak: yes, I know, I usually don't do it like that	20:47
diamonds	SamB: that shows the files changed & number of lines but not the contents	20:48
SamB	I think P4C0 knows it's kind of dumb and wants to salvage things	20:48
	→ hahuang65 joined	20:48
SamB	diamonds: oh	20:48
	in truth, I usually use magit to look at them	20:48
arand_	git show stash{n} ?	20:48
SamB	but it must use some underlying functionality	20:49
	→ momomomomo joined	20:49
SamB	looks like you can alsu use "git stash show -p stash@{$N}"	20:49
	or skip the last word for the top one	20:50
	← anderse left	20:50
diamonds	arand_: http://pastie.org/pastes/7614179/text	20:50
	← Saby left	20:50
cbreak	salvaging is easy	20:50
	← Bass10 left	20:51
cbreak	just squash merge the whole branch, then reset it away and selectively commit the debries	20:51
SamB	cbreak: I think it's usually possible to do a bit better than that	20:51
arand_	diamonds: word order.	20:51
	→ mossplix joined	20:52
arand_	git show stash@{n}	20:52
diamonds	ah	20:52
	→ prc3webs joined	20:52
SamB	diamonds: or add a -p	20:52
	← rideh left	20:52
diamonds	SamB: I don't get that one	20:52
	← amesha left	20:52
	→ Bass10 joined	20:52
diamonds	"With --patch, you can interactively select hunks from the diff between HEAD and the working tree to be stashed. The stash entry is constructed such that its index state is the same as the index state of your repository, and its worktree contains only the changes you selected interactively. The selected changes are then rolled back from your worktree. See the `Interactive Mode'' section of linkgit:git-add[	20:52
	1] to learn how to operate the `--patch mode."	20:52
	I don't see how... I just don't get how this means "show diff"	20:52
	→ amesha joined	20:52
__david__	dr_lepper, ugh, rebase doesn't like it that I merged master into A and resolved conflicts...	20:53
diamonds	I don't want to interactively select hunks	20:53
	→ m0viefreak joined	20:53
	← lb1a left	20:53
SamB	diamonds: I'm going from the paragraph under "show [<stash>]"	20:53
arand_	diamonds: -p acts differently when used with log or stash than it does with add.	20:53
	← exoon left	20:53
dr_lepper	__david__: you didn't mention merging master into the feature branches	20:53
SamB	which says you can use git-diff(1) options to select patch formats, and gives -p as an example	20:53
	→ khushildep joined	20:54
diamonds	ah "patch" == "diff"	20:54
	→ aitiba__ joined	20:54
	→ karstensrage joined	20:54
SamB	no, patch and diff are near-inverses ;-P	20:54
diamonds	so git show stash@{n} == diff	20:54
__david__	dr_lepper, oh, ups. I did that in an attempt to prepare the feature branch for merging (maybe wrongly so?)	20:54
	← prc3webs left	20:54
SamB	(though when used to denote a file format, they mean the same thing)	20:55
karstensrage	hi, were using gitolite and i was wondering if there were a smart way to itemize the current users and push their .pub keys to servers for ssh login?	20:55
diamonds	git stash show stash@{n} -p == .... inverse diff?	20:55
	← drizzd left	20:55
	← pmdarrow left	20:55
SamB	karstensrage: you mean, you want to graba /home//.ssh/.pub and add those?	20:55
	→ john_anderson joined	20:56
__david__	dr_lepper, I think in this case I achieve exactly what I want with `git log A..B` and `git diff A..B` - the former lists all commits, the latter all changes	20:57
	→ mloskot joined	20:57
__david__	then manually apply that after squash merging A into master	20:57
arand_	diamonds: git stash show -p stash@{n} looks like it's showing the diff as it was stored in the stash, and as it would be applied if pop/applied.	20:57
	← wsebastian left	20:57
karstensrage	SamB, gitolite-admin/keys/*.pub and add those	20:57
	is that what you mean?	20:57
diamonds	arand_: as opposed to what	20:57
	← tenmilestereo left	20:58
	← Bass10 left	20:58
	← wrouesnel1 left	20:59
	← mg983 left	20:59
	→ Bass10 joined	20:59
SamB	I think "git show stash@{n}" doesn't show everything, actually, because a stash is really two commits isn't it?	20:59
	← juvenal left	20:59
arand_	yeah... git show stash@{n} does not show the diff of new files...	20:59
	← robertodecurnex left	21:00
	← thiago left	21:00
	← p0wn3d left	21:00
dr_lepper	__david__: were there too many conflicts?	21:00
	← momomomomo left	21:00
arand_	And it seems to treat it as a merge diff instead...	21:01
dr_lepper	__david__: 'cause I'd recommend you to remove the merges, if you still want the squashed commits	21:01
	having _both_ the merges and squashed branches doesn't make any sense	21:01
	it's like having the same history twice	21:02
SamB	arand_: yeah, I believe it is a merge commit	21:02
__david__	dr_lepper, how would I remove the merges?	21:03
	← DaisyThe1ow left	21:03
	← wheezl left	21:03
dr_lepper	__david__: reset the branches to a commit before the merge with "git reset"	21:03
	→ _Vi joined	21:03
	→ thiago joined	21:04
	→ axl__ joined	21:04
	← vedant left	21:04
dr_lepper	e.g. "git checkout A ; git reset --hard A^"	21:04
	← Bass10 left	21:04
dr_lepper	make sure "A^" is the parent commit you want, and that you don't have any unstaged changes	21:04
__david__	dr_lepper, but there are changes after the commit that I also need	21:04
	→ Bass10 joined	21:04
	← geekbri left	21:04
__david__	s/commit/merge/	21:04
	← freeed left	21:04
	→ bmac joined	21:05
	→ wheezl joined	21:05
dr_lepper	oh	21:05
SamB	why are we trying to squash again?	21:05
	← axl_ left	21:05
	axl__ → axl_	21:05
dr_lepper	__david__: how many?	21:05
__david__	a couple	21:05
	SamB can see the point when you're basically preparing a patch series, not so much in a main development repository	21:06
__david__	SamB, just to clean things up	21:06
	← Bass10 left	21:06
	→ ShooterMG joined	21:06
__david__	SamB, it is a feature branch, but it's stable now and should go into production (master)	21:07
	← suspekt left	21:07
	← metze left	21:07
	← cannonball left	21:08
	→ metze joined	21:08
	← goshawk left	21:08
__david__	hm, but the merge without squashing just works... maybe I should save myself some time and just do it this way	21:08
dr_lepper	__david__: good thought	21:08
	← DoniRS left	21:08
__david__	it just always bugs me when something that seems to simple turns out to be so difficult ;-)	21:09
dr_lepper	well, you have created a certain structure in the history	21:10
	by merging branches	21:10
__david__	I think I need to use --ff more often	21:10
	though I'm not sure that would have fixed everything	21:10
dr_lepper	no, it wouldn't	21:10
	← [1]habva2013 left	21:10
dr_lepper	what you probably want (or even mean) is git rebase	21:11
__david__	yeah, I tried to avoid rebase because sometimes other devs fetch from my repo	21:11
dr_lepper	do they fetch your feature branches as well?	21:11
	→ danishkhan joined	21:11
__david__	or maybe I just shouldn't worry about those "forced update" messages so much	21:11
	yeah	21:11
	← ToothlessGear left	21:11
__david__	to check on its status	21:11
dr_lepper	just to check?	21:12
	not to develop on them, or use them in their own work?	21:12
__david__	usually, yes, though occasionally when something stalls they add to it	21:13
	→ MattStratton joined	21:13
	← scampbell left	21:14
	← sq-one left	21:14
	← rudak left	21:15
	→ hobodave joined	21:15
	→ InvidFlower joined	21:16
	→ Jojomofoman joined	21:17
	→ mitz joined	21:17
__david__	dr_lepper, seems like 'rerere' may have been the solution to the rebase conflicts - but it seems strange to me that it's even necessary	21:17
	← e66 left	21:17
	→ e66 joined	21:17
	→ rudak joined	21:17
__david__	actually, the intro to the rerere man page looks exactly like my situation	21:18
	→ schemula joined	21:18
	→ caspingus joined	21:19
	← caspingus left	21:19
	← sirdancealo2 left	21:19
	→ caspingus joined	21:19
	→ Bass10 joined	21:19
	← danishkhan left	21:20
	→ DaveLib joined	21:20
	gitinfo set mode: +v	21:22
	→ goshawk joined	21:22
	← h0bbit left	21:23
	← maxmanders left	21:23
	← mitz left	21:23
	← system64 left	21:24
	← nkr left	21:24
DaveLib	hello i am a git newbie i am trying to use git as a backup of a directory tree i will mirror on 2 linux computer in my home LAN over SSH i created a git repo for directory common/ on 1 computer my user account is dave on both computers and i previously got SSH working between the dave accounts on the 2 machines	21:24
	→ mmlac-bv joined	21:25
	← v0n left	21:25
	← Thra11 left	21:25
	→ Gentlecat__ joined	21:25
DaveLib	i went to the second machine and in the parent directory of the intended git mirror common/ i did	21:25
karstensrage	SamB, any suggestions other than just giving read access to the repo and lettting them download the keys to be pushed?	21:26
	clone the keys?	21:26
DaveLib	git clone ssh://192.168.1.102/path-to-other-common/	21:26
SamB	karstensrage: what is this about downloading keys ?	21:26
DaveLib	both accounts are dave so i left dave@ off	21:27
	→ mishok13 joined	21:27
	← amesha left	21:27
SamB	DaveLib: so, you know git isn't really good for backups right?	21:27
DaveLib	i got back error message ....	21:28
	→ amesha joined	21:28
SamB	I mean, it works pretty well for backing up the source history of open source projects, but that's about it	21:28
DaveLib	no i didn't i had thought of that on my own last couple days hey why not use git for backups	21:28
SamB	and, I mean, it works okay for private sources sometimes	21:29
DaveLib	i had previosly been using dar for backups to external and unison to mirror directories	21:29
	← hobodave left	21:29
	← Gentlecat left	21:29
	→ eijk_ joined	21:29
	← eijk left	21:29
DaveLib	yes it is just for me on my own computers	21:29
	to duplicate accross computers agaisnt hardware fair;lires	21:30
	← Artpicre left	21:30
DaveLib	failures	21:30
	also against accidental rm files	21:30
	i thought git compression and delta's in packfiles the revision controls and so easy to push and pull and spread out the copies	21:31
SamB	but I suppose you'll probably be wanting to solve the error message anyway	21:31
	→ stephCoue joined	21:31
	← e66 left	21:31
DaveLib	yes	21:31
	← kmels left	21:31
DaveLib	i have 3 copmputers and an external	21:31
	→ sirdancealo2 joined	21:31
DaveLib	so i thought mkae 4 copies and use git to send them around	21:31
	why is git bad for backups in general SamB ?	21:31
	these would all be private just my own files for myself on my own computwers	21:32
	i have ssh on the lan with generated keys	21:33
SamB	well, I'm not sure it's always bad, but the circumstances where is sufficient are certainly quite limited.	21:33
	← khushildep left	21:33
	← Bass10 left	21:33
	→ Bass10 joined	21:34
DaveLib	i thought my unison mirroring adn if i mistakenly delete a file or change my mind anout a deletion then unison would have deleted all copies	21:34
	with git i can get them back	21:34
SamB	yeah	21:34
	← Davey left	21:34
	← uidzer0 left	21:34
DaveLib	also if the repos are getting too big i cane rebase -s to squash some old cimmits away	21:35
	SamB doesn't actually have any media to do backups on, personally :-(	21:35
DaveLib	i guess same squash on ech repo copy but i don't mind	21:35
__david__	DaveLib, unless you really want versioning anyway, there are probably better solutions	21:35
	→ Noldorin joined	21:35
DaveLib	ok what are good ? i had been usiong dar as external hd backups and unison to mirror important files	21:36
	→ butblack joined	21:36
	← Watabou left	21:36
__david__	like rdiff-backup or "back in time"	21:36
butblack	does deleting a repository delete all of the commit and history of that repo?	21:36
	→ astroo- joined	21:36
DaveLib	i thought versioning is just a safety net	21:36
butblack	essentially wipe it clean	21:36
	← penguin1 left	21:36
	→ etcetera joined	21:37
__david__	DaveLib, you may need to combine them with unison or so to get the network synchronization though	21:37
butblack	this is for a case where no one forked, starred, followed the repo	21:37
	← mossplix left	21:37
__david__	but they solve the accidental rm problem	21:37
	butblack, followed? are you talking about github or something?	21:38
butblack	__david__:	21:38
	yeah	21:38
	i thought this was the channel?	21:38
	lol i guess not	21:38
	haha	21:38
	thanks (Y)	21:38
DaveLib	there is a #github channel somewhere	21:38
__david__	github uses git, but what you're asking is not really about git	21:38
butblack	DaveLib: yeah, found it, thanks	21:39
	← mikef left	21:39
DaveLib	ahh #github here at freenode	21:39
butblack	and yeah definitely github specific	21:39
	← butblack left	21:39
	← cbreak-work left	21:40
	← b1rkh0ff left	21:40
DaveLib	anyway i have 3 computers + the external so mkae 4 cooies and i thought git would make that easier than unison also i thought git would be more intuitive to recover stuff than dar	21:40
	← P4C0 left	21:40
DaveLib	a couple years ago i researched backup and at the time dar sounded better to me than rdiff or back in time	21:41
	→ amcsi joined	21:41
	→ osmosis joined	21:41
	→ jjido joined	21:41
	← DonRichie left	21:41
	→ marhaban joined	21:42
	← ojacobson left	21:42
DaveLib	anyway my error message was ssh: cannot resolve host name name or service unknown	21:42
	← mmitchell left	21:42
	← lacrymology left	21:42
	← Dougie187 left	21:42
DaveLib	but from the same machine when i do ssh 192.168.1.102 its ok	21:43
	← barneybook left	21:43
__david__	try using 192.168.1.102:/my/path	21:43
	→ barneybook joined	21:43
DaveLib	thanks _david_	21:43
	→ b1rkh0ff joined	21:43
DaveLib	so no ssh: prefix at all ?	21:43
__david__	hm though I guess ssh:// is more or less an alias for that	21:44
DaveLib	ok	21:44
	i will try	21:44
	← thasmo left	21:44
__david__	DaveLib, in any case I'd recommend for you to either setup /etc/hosts or to teach your router to resolve the names	21:44
cbreak	ssh:// is a URL prefix	21:44
	→ redhotvengeance joined	21:44
	← thiago left	21:44
__david__	imho that's much better than working with IPs	21:44
	→ infynyxx joined	21:45
cbreak	no prefix means local path, or maybe scp syntax	21:45
	← tvw left	21:46
DaveLib	thanks _david_ that worked :)	21:46
	its processing now	21:46
	← kadoban left	21:46
	← thirtythreeforty left	21:46
	→ donmichelangelo joined	21:46
	← Anders_J left	21:47
	← TeckniX left	21:47
	→ mossplix joined	21:47
	← rednaks left	21:48
__david__	cbreak, no prefix, no : means local path, with host.name: it is scp-like	21:48
	→ josh_ joined	21:48
	→ w09x joined	21:48
cbreak	yes, as I said	21:49
	← tinti left	21:49
cbreak	it's a stupid system	21:49
	← josh_ left	21:49
DaveLib	i got that prefix version that didn't work from a help page on the web	21:49
cbreak	can't even handle paths that contain :	21:49
__david__	I like scp for its brevity	21:49
cbreak	I recommend using url syntax always	21:49
	apart from local clones	21:50
__david__	DaveLib: my guess is ssh:// urls don't start in your homedir, but at /	21:50
	→ josh_ joined	21:50
	← josh_ left	21:50
__david__	if you do host.name:foo it resolves to ~/foo	21:50
	→ kadoban joined	21:50
	← b1rkh0ff left	21:51
DaveLib	oh yes that web page gave ssh://user@host:/~/rest-of-path	21:51
	→ lando23 joined	21:51
	← lando23 left	21:51
SamB	awwwwwwwww	21:51
	→ lando23 joined	21:51
SamB	--cherry-mark doesn't work in gitk :-(	21:52
DaveLib	it is best to only push to bare repos and not to a wroking repo with working directory ?	21:52
milki	yes	21:52
DaveLib	ok	21:52
SamB	doesn't git tell you this?	21:52
DaveLib	i have a working repo on ocmputer A	21:52
	← R1KRDO left	21:52
	→ mitz joined	21:53
DaveLib	yes i had read that in documentation and was following it but just confirming	21:53
	→ R1KRDO joined	21:53
	← jjido left	21:54
	→ Conic joined	21:55
	→ jjido joined	21:55
	→ kcj joined	21:55
	← infynyxx left	21:55
	← Houl left	21:56
	← chisler left	21:56
	→ infynyxx joined	21:56
	← goshawk left	21:57
	← rudak left	21:58
	→ kit joined	21:58
	← jast left	21:58
	← oakdog8 left	21:58
	→ pastime joined	21:59
	← SuperDMP left	21:59
	→ wrouesnel1 joined	21:59
	← breakingthings left	21:59
	→ thiago joined	21:59
	← gitinfo left	21:59
	← pantsman left	22:01
	← __david__ left	22:01
	← john2496 left	22:02
	← etcetera left	22:02
	← R1KRDO left	22:02
	← Clint_Beancake left	22:02
	← marhaban left	22:02
	→ jordanm joined	22:04
	← wannado left	22:04
	← JPeterson left	22:04
	← jjido left	22:04
	→ johnflux_ joined	22:05
	→ astroo-_m joined	22:05
	→ danishkhan joined	22:05
	← johnflux left	22:05
	← Alenah left	22:05
	← wrouesnel1 left	22:06
	→ ivan_on_trac joined	22:07
	← CodeShark left	22:07
	→ CodeShark joined	22:07
	← jcromartie left	22:07
	← infynyxx left	22:07
	← astroo- left	22:07
	← mossplix left	22:07
	astroo-_m → astroo-	22:07
	← danishkhan left	22:07
	← natechan left	22:07
ivan_on_trac	In my company has 25 staff developers. I need suggestions on how to choose a workflow. Tutorial suggestion?!	22:08
	→ Alenah joined	22:08
	← Milossh left	22:08
jordanm	http://en.wikipedia.org/wiki/Agile_software_development	22:08
	→ jcromartie joined	22:08
	→ goshawk joined	22:09
	← Raptors left	22:09
	→ diegoviola joined	22:09
	← MattStratton left	22:09
	→ Sebboh joined	22:10
	→ etcetera joined	22:10
	← Sebboh left	22:10
	→ Sebboh joined	22:10
	→ JPeterson joined	22:11
	← mrwalker left	22:11
	← Sebboh left	22:11
	← trungqtran left	22:12
	← johnnyfuchs left	22:12
	→ mossplix joined	22:12
	← redhotvengeance left	22:13
	← mitz left	22:15
	← jwmann left	22:16
	← Rulin left	22:16
	← kit left	22:17
	← bmac left	22:18
	← lavaman left	22:18
	→ ToothlessGear joined	22:18
	← mezner left	22:18
	← cbot_ left	22:19
	→ bytephilia joined	22:19
	→ _iron joined	22:19
	← macabre left	22:20
	← Sosumi left	22:21
	← w09x left	22:21
	← Siraris left	22:22
	← ivan_on_trac left	22:22
	→ redhotvengeance joined	22:22
	← jude left	22:23
	← comps left	22:23
	→ jude joined	22:24
	→ natechan joined	22:24
	→ uidzer0 joined	22:25
	← etcetera left	22:25
	← maslowbeer left	22:26
	← tjbiddle left	22:26
	← Desproges left	22:27
	→ gitinfo joined	22:27
	ChanServ set mode: +o	22:27
	← kermit left	22:27
	→ kermit joined	22:29
	→ LuxuryMode joined	22:30
	→ thasmo joined	22:30
	→ Milossh joined	22:30
	← kevwil left	22:30
	→ Pupeno joined	22:31
	← goshawk left	22:32
	← Pupeno left	22:32
	← diamonds left	22:32
	← gitinfo left	22:32
	→ calamari joined	22:32
	→ danishkhan joined	22:33
	← arnej left	22:34
	→ Tommy[D] joined	22:34
	→ arnej joined	22:34
	← wrouesnel left	22:35
	→ johnflux__ joined	22:36
	← johnflux__ left	22:36
	→ johnflux__ joined	22:36
	← johnflux_ left	22:37
	← LuxuryMode left	22:37
	→ Davey joined	22:40
	← private left	22:40
	→ w09x joined	22:42
	← uidzer0 left	22:42
	← ozcanesen left	22:43
	← nioncode left	22:44
	→ hendry joined	22:44
	→ tinti joined	22:45
	← axl_ left	22:45
	← bytephilia left	22:47
	← Kicer86 left	22:47
	→ infynyxx joined	22:47
	← Milossh left	22:48
	→ Sosumi joined	22:49
	← acharlieh left	22:49
	→ Kicer86 joined	22:49
	→ wrouesnel joined	22:49
	← hwrd\|work left	22:50
	→ mitz joined	22:50
sanmarcos	is there a way to make code reindents without it screwing up the entire history of git blame?	22:50
	or to have git blame skip certain commits?	22:51
thiago	git blame -w	22:51
	← thl0 left	22:52
	→ uidzer0 joined	22:55
	→ Adlai joined	22:56
	→ comps joined	22:57
	→ kevwil joined	22:58
	← mitz left	22:58
	← m8 left	22:59
	→ khushildep joined	23:00
	← ToothlessGear left	23:01
	← opus left	23:01
	→ ToothlessGear joined	23:02
	← thiago left	23:03
Caelum	so I'm in project/trunk in git-svn, and we recently started using branches/ but I didn't do the git-svn clone with --std-layout, how do I tell it I have a branches/ path now so I can make branches?	23:04
	→ maslowbeer joined	23:04
	← imMute left	23:04
	→ opus joined	23:05
	→ eletuchy joined	23:05
	← w09x left	23:05
	← kevwil left	23:06
	→ hwrd\|work joined	23:06
	→ Xethron joined	23:08
	← john_anderson left	23:08
	→ werewolf2 joined	23:08
	← werewolf2 left	23:09
	← infynyxx left	23:09
	→ kvda joined	23:10
	→ TeckniX joined	23:10
Caelum	what's the difference between refs/remotes/git-svn and refs/remotes/trunk	23:10
	← TeckniX left	23:10
	→ goshawk joined	23:10
	→ simesy joined	23:11
	→ imMute joined	23:13
	← imMute left	23:13
	→ imMute joined	23:13
	← astor left	23:15
	← InvidFlower left	23:16
	← goshawk left	23:16
	→ SkAzZ joined	23:16
	→ etcetera joined	23:17
	← mossplix left	23:18
	→ markalanevans joined	23:18
	← ehsan left	23:19
	← markalanevans left	23:20
	→ markalanevans joined	23:20
	← opus left	23:23
	→ gitinfo joined	23:24
	ChanServ set mode: +o	23:24
	→ opus joined	23:24
	→ mossplix joined	23:24
	← eijk_ left	23:26
	← Internet13 left	23:27
	→ mitz joined	23:27
	← m0ltar left	23:28
	← gitinfo left	23:28
	← pastime left	23:28
EugeneKay	!botsnack	23:29
	Oh, he died again.	23:29
	→ Dougie187 joined	23:30
	← Davey left	23:30
	HedgeMage blames EugeneKay	23:30
EugeneKay	It's because I'm not wearing pants, isn't it.	23:31
HedgeMage	Probably	23:31
	I thought you were on a trip with your dad, so the lack of pants thing is more concerning than usual.	23:31
	← DaveLib left	23:31
EugeneKay	I am; we have a 2-bedroom suite.	23:32
	← simesy left	23:32
	→ thiago joined	23:32
	→ Internet13 joined	23:32
	← danishkhan left	23:32
HedgeMage	ahh	23:32
	← schemula left	23:33
	← uidzer0 left	23:33
	← NiteRain left	23:33
	← phantomtiger left	23:33
	← mossplix left	23:34
	← natechan left	23:34
calamari	I've set up git-daemon, and specified --enable=receive-pack, and it's still insanely slow. any tips on debugging this?	23:35
	← boombatower left	23:35
	← etcetera left	23:35
calamari	(pushes are slow)	23:35
HedgeMage	Sorry, I've never used git-daemon. I use gitolite for both my git servers.	23:36
	HedgeMage == lazy	23:36
	→ uidzer0 joined	23:36
calamari	gitolite uses ssh and shared keysm	23:36
	?	23:36
HedgeMage	no shared keys	23:36
	everyone uses their own key(s)	23:37
	← mitz left	23:37
calamari	I'm hoping for a solution where everyone has access with a minimum of hassles	23:37
HedgeMage	that would be gitolite	23:37
	→ simesy joined	23:38
milki	or no authentation	23:38
	auth*	23:38
calamari	yeah I don't care about auth, we are behind a firewall	23:38
HedgeMage	calamari: gitolite took me <10 minutes to set up the first time, <5 the second	23:38
milki	calamari: then just use ssh?	23:38
	→ jfinn joined	23:39
calamari	milki: empty password?	23:39
	→ bitkiller joined	23:40
milki	sure?	23:40
	i wouldnt recommend it, but that would certainly be hassle free	23:40
	← thasmo left	23:41
	→ boombatower joined	23:41
milki	assumming users understand ssh >.<	23:41
	which might be a high bar for your organization o.O	23:41
EugeneKay	calamari - I strongly encourage you to require users to auth to access repos in an organizational context	23:41
	If gitolite looks like too much work, there are other solutions, such as Atlassian Stash, that are cheap-and-easy.	23:42
	← aitiba__ left	23:42
	← anonymuse left	23:42
	Joefish → Joefish]off	23:42
	← Noldorin left	23:42
	← simesy left	23:42
calamari	EugeneKay: if people have to deal with ssh keys their eyes will gloss over and they'll say "too hard"	23:43
EugeneKay	Get less-stupid users? :-p	23:43
calamari	lol	23:43
HedgeMage	calamari: If people think ssh key auth is "too hard" they are not competent devs and you should fire them.	23:43
calamari	not my call	23:44
HedgeMage	calamari: that is unfortunate	23:44
	→ mitz joined	23:44
EugeneKay	Look at Stash. It's basically bitbucket in a box	23:44
	← sgambino left	23:44
EugeneKay	And for <=10 users it's $10	23:44
calamari	thanks for the ideas. it seems I am on the wrong path. I will check what I cfan do with ssh	23:44
EugeneKay	You can also try setting up gitolite with http(s)	23:45
	That'll take user+pass auth	23:45
calamari	maybe a single account with password everyone knows	23:45
	I heard http is slow	23:45
	← opus left	23:46
EugeneKay	It is a terrible protocol to wrap git in	23:46
	← AdrienBr_ left	23:46
calamari	I must have something wrong.. that push was running for 30 minutes and never finished	23:47
	(git-daemon).. but that's ok	23:47
	→ goshawk joined	23:47
	← breck7 left	23:47
MatthewWilkes	calamari: How big are your repos!?	23:47
	→ Watabou joined	23:47
	→ breck7 joined	23:48
calamari	8 megabytes, the commit was 1 kb	23:48
	← goshawk left	23:48
MatthewWilkes	yeah… that's certainly in the 'did something wrong' bracket	23:48
calamari	(commented out a line to test the push)	23:48
	→ simesy joined	23:50
	← simesy left	23:50
	← breck7 left	23:50
	→ simesy joined	23:51
	→ jast joined	23:51
	ChanServ set mode: +v	23:51
	→ subfauna joined	23:52
	→ gitinfo joined	23:52
	ChanServ set mode: +o	23:52
	→ SBoudrias joined	23:52
	→ caspingu_ joined	23:52
	→ normanrichards joined	23:52
	← TheMoonMaster left	23:52
	→ rideh joined	23:53
	← Dougie187 left	23:53
	→ TheMoonMaster joined	23:53
	← wesside left	23:53
EugeneKay	!botsnack	23:53
gitinfo	Om nom nom	23:53
EugeneKay	Yay	23:53
	→ trusktr joined	23:54
	← RalfR left	23:54
	→ mr_mustard joined	23:55
	← ozette left	23:55
	← caspingus left	23:55
	← Watabou left	23:57
	← boombatower left	23:58
	← Caelum left	23:58
	→ Caelum joined	23:59
	← Xethron left	23:59
	→ opus joined	23:59

Logs Search ←Prev date Next date→

Channels Documentation