IRCloggy #git 2012-09-04

Provider of IRC logs since 2005.

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

	2012-09-04
	← Chillance left	00:00
	→ btree_ joined	00:00
	→ Chillance joined	00:00
	← xxtjaxx left	00:01
	← lb_ left	00:02
	→ lb_ joined	00:02
	← lb_ left	00:02
	→ lb_ joined	00:02
	← cworth left	00:03
	→ xxtjaxx joined	00:03
	← xxtjaxx left	00:03
	→ xxtjaxx joined	00:03
	← didi left	00:05
	→ cworth joined	00:06
	← ironfroggy left	00:06
	→ brenden1 joined	00:08
	← Antisoche left	00:08
	→ Antisoche joined	00:08
	← caseymcg left	00:11
	← glennpratt left	00:11
	← kokainepizza left	00:12
	← DanyO83 left	00:12
	← iocor left	00:13
multiHYP	night all	00:14
	← multiHYP left	00:14
	→ billyoc joined	00:15
	→ Moussekateer joined	00:16
milki	lol	00:18
	i learned a lot from w3schools	00:18
	!	00:18
	mostly css stuff	00:18
	← mdw left	00:18
	→ jalama joined	00:19
	← archdori left	00:23
	→ mdw joined	00:24
	→ robottinosino joined	00:25
	← robottinosino left	00:25
	← guampa left	00:26
	→ robottinosino joined	00:26
	← robottinosino left	00:26
	→ gmagno joined	00:27
	→ etcetera joined	00:27
	← etcetera left	00:28
gmagno	erm... consider that I already have git installed (from source code). Should I unisntall it before cloning a new release of git and make install it? Or can I just make install it over the first one?	00:28
	← RaceCondition left	00:29
	→ maletor joined	00:29
	← Chillance left	00:30
	← AlbireoX left	00:31
	→ tmcmahon joined	00:31
	→ unreal joined	00:32
	← neurodrone left	00:33
	← hyperair left	00:34
milki	well, you need git to clone it first right?	00:35
	if it comes with an unintall script, i recommend using it though	00:35
offby1	gmagno: "make install" over the existing one should be fine; I do that all the time	00:37
	→ Petskull joined	00:38
	gitinfo set mode: +v	00:39
Petskull	so.. I ran 'git commit -m "first commit"' on my code folder.. Now how to I retrieve my backup?	00:39
cmn	what backup?	00:40
Petskull	well, my commit	00:40
cmn	that's not a backup	00:40
Petskull	I know	00:40
cmn	what do you want to retrieve? it's there, that's what you have	00:40
Petskull	but if I smashed up my code, can I revert to a previous version?	00:41
cmn	yes	00:41
	how depends on what you've done up to now	00:41
	the !book explains how you use git	00:41
gitinfo	There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable	00:41
Petskull	got a handle refernce for how?	00:41
	ah, yeah, I saw that one	00:41
cmn	first of all, ask git status what's going on, then !fixup tells you how to undo some things	00:42
gitinfo	So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!	00:42
	Petskull looks again	00:42
cmn	but this all depends on what you've done and what 'revert' means in your situation	00:42
Petskull	well, my main concern is using the old version if I mess up my code by writing a bunch of crap in it	00:43
	don't worry, I'll figure it out	00:43
	← dozymoe left	00:44
	← Akufen left	00:44
	← Vert left	00:44
gmagno	offby1, thank you o/	00:46
offby1	git is actually one of the easier Free Software thingies to build.	00:47
	→ CannedCorn joined	00:47
	→ cooloney joined	00:48
	Sargun_Screen → Sargun	00:49
	← chrisrr left	00:51
	→ pandeiro joined	00:52
	← xxthink left	00:52
	→ glennpratt joined	00:55
	→ iori joined	00:56
	ceradon-away → euphoria	00:58
	→ codeshah joined	01:00
	→ ecksit joined	01:02
ecksit	morning guys, here is my situation. I have a remote repo and a local repo. Essentially the same contents except for a few files. I need to get my local dir up to date from the remote master branch. what is the best way to achieve this without causing errors?	01:03
	← Guest79949 left	01:03
	→ jason joined	01:03
	jason → Guest77688	01:04
	← Yuuhi` left	01:04
	→ neurodrone joined	01:04
	← neurodrone left	01:04
	→ neurodrone joined	01:04
	euphoria → ceradon-away	01:04
milki	content doesnt matter so much as commit history	01:04
cmn	ecksit: fetch from the remote and see what you want to do, possibly merging	01:05
offby1	ecksit: I always do "git remote update", then "git sl HEAD..@{u}" to see what's new	01:05
	ceradon-away → euphoria	01:06
ecksit	cheers, will have a crack	01:06
	← sebastorama left	01:06
	→ rohan32 joined	01:06
	← w09x left	01:07
	→ hanasaki joined	01:08
	→ kokainepizza joined	01:08
	gitinfo set mode: +v	01:08
	→ intripoon joined	01:08
hanasaki	is there a way to undo a "git init" on a directory that it was run on by mistake?	01:08
bremner	rm -r .git	01:09
	→ SeoZ joined	01:09
hanasaki	what about the subdirectories on that directory?>	01:09
	← gigamonkey left	01:10
hanasaki	does each subdir have a .git like there is a .svn for svn?	01:10
cmn	no, and neither does svn now	01:11
ecksit	will git carry through chmod settings?	01:12
hanasaki	ah just the parent dir. I didn't know about the change in svn... what caused that in what ver?	01:12
	← intripoon_ left	01:12
	← agile left	01:12
cmn	ecksit: git doesn't care about chmod	01:12
	← diegoviola left	01:12
cmn	it only cares whether the file is executable for the user	01:12
Petskull	sigh git can't be this complicated...	01:13
	→ xiangfu joined	01:13
	→ etcetera joined	01:13
	← dhruvasagar left	01:13
chuckf	Petskull: the basic usage is not that complicated	01:14
	→ sebastorama joined	01:14
offby1	cmn: Huh! I didn't know that. Guess it's been a while since I paid any attention to Subversion.	01:15
	Sorry, Karl!	01:15
Petskull	yeah, I'm thinking it can't be that complicated	01:15
	must be it's not intended to be used between folders like this	01:15
offby1	Petskull: it is, and it isn't. The UI is horrible. But the underlying concepts are pretty simple	01:15
cmn	it's not a sync	01:15
Petskull	I know it's not rsync	01:15
offby1	Petskull: general-purpose git advice: 1) Read "Git For Computer Scientists" (http://eagain.net/articles/git-for-computer-scientists/); 2) Run "gitk --all"; 3) Learn about "git reflog"	01:15
Petskull	reading this now: http://git-scm.com/book/en/Git-Basics-Tagging	01:16
cmn	it's not rsync either, no	01:16
Petskull	already looked at gitk	01:16
	← Milossh left	01:16
	→ jdyin joined	01:16
Petskull	ok, how do I get the a specific version onto a folder?	01:17
ecksit	why is it i made a change but git isn't picking it up in the repo.	01:17
Petskull	that's what it's set up to do, isn't it?	01:17
	→ duckxx joined	01:17
cmn	that's one thing that it does	01:17
Petskull	from folder_a to folder_b?	01:17
cmn	use checkout	01:17
	→ hyperair joined	01:17
cmn	no, that's not what git is for	01:17
Petskull	not clone?	01:17
cmn	if you want to clone the repo, do that	01:18
	but getting a specific version isn't what clone is for, that's checkout	01:18
	→ diegoviola joined	01:18
cmn	clone is what you do before you can do anything	01:18
	→ acrocity joined	01:19
	← acrocity left	01:19
	→ gtuckerkellogg joined	01:19
Petskull	heh- nevermind.. I'ma go back to my 'backup-20120903' folder system	01:20
	← bizhanMona left	01:20
Petskull	that seemed to work better	01:20
	→ bizhanMona joined	01:23
	→ ironfroggy joined	01:23
	→ hwrd\|work joined	01:24
	← hwrd\|work left	01:24
	→ acrocity joined	01:25
	→ alberto56 joined	01:25
	← Eridius left	01:26
	→ hwrd\|work joined	01:26
	← cybersphinx left	01:31
	← gmagno left	01:33
	← frogonwheels left	01:33
	← noyb left	01:34
	→ frogonwheels joined	01:34
	← sebastorama left	01:35
	→ sebastorama joined	01:35
	← Vampire0 left	01:36
	→ Davey joined	01:37
	← acrocity left	01:37
	euphoria → ceradon-away	01:40
	← Ferris_ left	01:42
	→ Ferris_ joined	01:42
	← duckxx left	01:43
	← gtuckerkellogg left	01:46
	← alberto56 left	01:46
	→ aircan joined	01:47
	← sebastorama left	01:47
	→ redrocket joined	01:48
	← codeshah left	01:49
	→ nitrix joined	01:49
	← tjholowaychuk left	01:49
	→ caseymcg joined	01:50
	→ Vampire0 joined	01:50
	→ rgr joined	01:50
	← hwrd\|work left	01:50
	← iamtakingiteasy left	01:53
	← aircan left	01:55
	← Ferris_ left	01:57
	← hashset left	01:57
	→ iamtakingiteasy joined	01:57
	→ Ferris_ joined	01:58
	→ jim joined	01:59
jim	heya.	01:59
	→ hashset joined	02:00
artnez	Is anyone aware of a shortcut for HEAD that doesn't include caps?	02:00
	Typing out HEAD is no fun. Typing SHIFT+H+TAB is a little better, but it would be nice if there was one magic character.	02:02
	← ecksit left	02:02
	← Ferris_ left	02:03
cmn	you rarely have to type it	02:03
	make an alias for your most-used commands	02:03
artnez	cmn: Eh. I guess you're right. In ranges I just omit it. Only time I use it specifying refs in git-log.	02:04
	cmn: Aliases for things like HEAD^3 aren't very useful either though.	02:06
cmn	don't hard-code the modifiers	02:06
	→ Taichou joined	02:07
jim	I've been noticing that emacs has some kind of git module (saw it get loaded as I used emacs) and I had questions about it... is it considered any good? anyone have some experience with it? (artnez howbout caps lock, and another possibility, howbout running gtypist periodically so anything becomes second nature and -fast-)... and finally, are there other modules that go with emacs (text-only, as I like it in text but I consider emacs in x to be a disaster compare	02:08
	d to the simplest gui editors when they're done right)	02:08
	→ argoneus joined	02:09
argoneus	hello	02:09
	→ Ferris_ joined	02:09
	→ wolftankk joined	02:09
argoneus	if I have folder A, which has a git repo set up for github, and a folder B with different code files, and I want to replace all the files on github with those in folder B	02:09
	what's the best way to do that?	02:09
jim	I mean that, I -love- emacs in a terminal	02:09
	→ someprimetime joined	02:09
jim	hate it in x	02:09
bremner	jim: magit is the best choice for git in emacs afaik	02:10
jim	bremner, what's your favorite thing about it? where can I learn some more?	02:11
	← kukks left	02:11
someprimetime	is it a bad idea to rely on github for hosting my codebase?	02:11
	→ irqq joined	02:11
	← mattt_ left	02:12
	→ ricky joined	02:12
someprimetime	i just bought a paid subscription and am about to do it, but i want to be aware of any risks	02:12
argoneus	please can anyone help? :<	02:12
jim	someprimetime, well as you know git clone makes a complete copy	02:12
	→ harish joined	02:12
cmn	argoneus: delete your files, put the ones from B there, add, commit and push	02:12
	someprimetime: risks as in what?	02:13
	→ ipalaus joined	02:13
bremner	jim: http://magit.github.com/magit/	02:13
jim	thanks :)	02:13
someprimetime	cmn: just security… i know it'll be private since i'm paying but no one can access my stuff right?	02:14
cmn	if you don't make the repo public, github won't	02:14
argoneus	cmn: thing is	02:14
	do I just clone the repo and 'git rm .' ?	02:15
	and push back	02:15
	← mgorbach left	02:15
cmn	you already cloned	02:15
	put the files in that repo's workdir, commit and push, as usual	02:16
	this isn't any different from any other change you make to the files	02:16
argoneus	but the folder isn't empty	02:16
	and I don't want to lose the files there	02:16
	just pull them from github	02:17
	← FernandoBasso left	02:17
	→ tjholowaychuk joined	02:17
	← jinjonBoo left	02:17
jim	since you have the prev versions in the repo (you do, right?) then you should be able to do the git rm *	02:17
	then remove the files	02:17
	→ gavin_huang joined	02:17
jim	then copy the others	02:17
	then add/commit	02:17
argoneus	I see	02:17
	so I git rm everything	02:17
	move them somewhere else	02:17
	move new ones in	02:17
	and commit&push	02:17
jim	well one sec before you do that	02:17
cmn	which files are you talking about now?	02:18
jim	see I"m close to a beginner here	02:18
cmn	you said you wanted to replace one set of files with another, no?	02:18
jim	so before listening to me, consider other advice	02:18
argoneus	yes	02:18
	but I don't want to lose the one set	02:18
	just pull it from github	02:18
cmn	no	02:18
	→ sebastorama joined	02:18
cmn	that doesn't make sense	02:18
argoneus	why not?	02:18
	← someprimetime left	02:18
argoneus	ok here's the full story	02:19
	← gusnan left	02:19
	→ vitor-br joined	02:19
argoneus	basically I started working on a program	02:19
cmn	replace the files you have in the repo with the other ones	02:19
argoneus	periodically pushed it to github	02:19
cmn	then add, commit and push	02:19
	← Guest77688 left	02:19
argoneus	but then I realized my coding was terrible	02:19
	so I want to start over but use pieces from the original crappy one	02:19
	so I need to backup the current files and push new ones instead to a clear repo	02:19
	what would you do?	02:20
jim	you could make a dir that might be called attic	02:20
	and move the files to there	02:20
cmn	you don't push files, you push commits	02:20
	if you want to backup the files, do that	02:21
	← orafu left	02:21
cmn	but that's irrelevant for the git operations	02:21
	→ orafu joined	02:21
cmn	do you want to delete all of the history for the branch on the repo at github?	02:21
	→ acrocity joined	02:22
argoneus	yes	02:22
jim	isn't that a form of "rewriting history"?	02:22
cmn	then init a new repo	02:22
	and just start	02:22
	forget about the other files	02:22
	consider that it's !rewriting	02:23
gitinfo	[!rewriting_public_history] Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to `git pull --rebase` and even worse things if they have tagged or branched, so you must publish your humiliation so they know what to do. You will need to `git push -f` to force the push. The server may not allow this. See receive.denyNonFastForwards (git-config)	02:23
argoneus	wait	02:24
jim	don't you want to keep the whole history? how many devs on the project? also, anyone watching/forking/cloning your code?	02:24
argoneus	what happens if I init a new repo in another folder	02:24
	← hashset left	02:24
argoneus	and push to master	02:24
	also	02:24
	it's just me and no one gives a flying fuck about it except me	02:24
	:<	02:24
cmn	git will say that the history has diverged	02:25
	which is why you need to tell it to overwrite	02:25
	← irqq left	02:25
	→ xxthink joined	02:26
	→ Fingolfin joined	02:27
argoneus	I see	02:28
	thanks	02:28
jim	argoneus, if all that is the case... could you explain why it matters to you wheter you start a new repo or not?	02:28
argoneus	I never said it mattered :<	02:29
jim	in this question I'm being an advocate for keeping your history	02:29
	oh ok... so, in that case you are faced with two choices which in your mind are fairly equal	02:29
argoneus	for me it's either scrap it or scrap it	02:30
	→ jaapbadlands joined	02:31
argoneus	guess I'll just make a new repo	02:31
	→ hashset joined	02:31
argoneus	but, if I wasn't alone and wanted to do a full re-write	02:31
	how would I do that?	02:31
jim	ok... so, I'm going to stop, because I've said my peace and you've heard	02:31
	→ nothingbutlove joined	02:32
argoneus	oh	02:32
jaapbadlands	how do I remove git from a folder and start afresh?	02:32
	jim stands back and lets the more experienced answer that	02:32
argoneus	sorry if I insulted you in any way	02:32
jim	of course not; I'm absolutely neutral	02:32
argoneus	well, you took the 'do wtf you want i don't care' stance	02:33
jim	sorry if my irc-tone suggests otherwise :)	02:33
	← jalama left	02:33
jim	argoneus, well I'm sorry you heard it that way, and I take responsibility for that communication	02:34
	so I'll retry	02:34
	first, I'm glad you're interested enough to learn how to do the alternative	02:34
	→ herdingcat joined	02:35
jim	and second, it is yours and I would be reaching if I tried to -insist- you do any particular thing	02:35
cmn	argoneus: if you're working with other people, you don't do this	02:35
	if you did, everyone would be screwed	02:36
	jaapbadlands: rm -r .git && git init	02:36
jim	I consider it a win that you're at least willing to consider keeping your history or at least wanting to know how to do it	02:36
	← berserkr left	02:37
jim	the "do wtf you want" part, I wish that didn't come across that way	02:37
	→ jalama joined	02:39
jim	adding to "if you're working with other people, you don't do this", if you have additional clones of the same repo, any history rewrite will screw at least all except one	02:39
	← lb_ left	02:39
	→ lb_ joined	02:40
	← lb_ left	02:40
	→ lb_ joined	02:40
	← jalama left	02:41
	← pandeiro left	02:41
argoneus	so the idea is	02:41
	instead of just removing it	02:41
	I make a commit saying 'removed stuff'	02:41
	so people can go back to that point anytime?	02:41
jim	yes, and more than that,	02:42
	the commit should also include adding the new files	02:42
	→ JeremyWei joined	02:42
	← JeremyWei left	02:42
	→ angelsl joined	02:42
argoneus	so then people can choose between old/new and see what I actually did	02:42
	that sounds better	02:43
	though the con is	02:43
	it's basically saying "I wrote crappy code and had to re-write the whole thing"	02:43
jim	so you would git rm the files, then move the existing files out, then move the new files in,	02:43
jaapbadlands	thanks cmn, so I've removed that previous tracking of a wordpress site on my staging server. On my server is a full wordpress site, and in the new github repo - it is set to ignore everything except 2 folders. I would like to reinitialize this folder on my server to track files from the new github repository, only overwriting the tracked files. How do I go about this? Hopefully this...	02:43
jim	add them, then commit	02:43
jaapbadlands	...explanation makes sense, happy to clarify if not	02:43
argoneus	yeah I'll do that	02:44
	makes much more sense	02:44
jim	so that one commit represents the whole replacement	02:44
argoneus	I mean	02:44
	at this point I could just re-make the repo	02:44
	but while at it I could also go and work on some good practices	02:44
	thank you very much	02:45
	one last thing before I go sleep	02:45
	I think I already asked it but didn't really get an answer	02:45
	← lfaraone__ left	02:45
argoneus	the way I get it	02:45
	'master' branch represents the stable branch	02:45
	← thallada left	02:45
argoneus	but if you init a new repo and start a new project	02:45
	it's not a stable version in any way but it defaults to master	02:46
	should I develop the stable version on master or should I instantly switch to an experimental/development branch and then merge to master?	02:46
	is there a 'good practice' for this?	02:46
	→ lfaraone__ joined	02:46
	→ thallada joined	02:46
offby1	argoneus: your branches represent whatever you want them to.	02:47
	Many projects indeed use "master" to represent "the deployable state", but this is just a convention, and not all that widely honored, either.	02:48
argoneus	are there any people that instantly switch to another branch and use 'master' only for stable stuff?	02:48
offby1	probably.	02:48
argoneus	so I can basically do what I want	02:48
	and it's always correct	02:48
offby1	sure	02:48
	→ _huoxito joined	02:48
offby1	it's a social thing, not a technical thing.	02:49
	← rohan32 left	02:49
	→ cenuij joined	02:49
argoneus	I think I'll use 'master' for deployable stable versions then	02:49
	← cenuij left	02:49
	→ cenuij joined	02:49
	→ rohan32 joined	02:49
argoneus	should be cleaner	02:49
	← rohan32 left	02:49
offby1	If your collaborators understand certain branch names to make certain promises, then you'd be wise to honor them; but if not, go wild.	02:49
argoneus	you can remove the 'master' branch right?	02:49
	or, do branches only show if something is in them	02:50
offby1	you can remove any branch you want.	02:50
	not sure what your last question meant.	02:50
	sounds like you need to do some "Try It And See™"-ing.	02:50
argoneus	yup, doing that now	02:50
	sorry for the trouble	02:50
	thanks for the help	02:50
	have a nice night/evening :)	02:50
	→ noyb joined	02:51
	← argoneus left	02:52
offby1	why thank you	02:53
	← hashset left	02:56
	← etcetera left	02:57
	→ gigamonkey joined	02:58
	→ mgorbach joined	02:59
	→ cordoval joined	03:01
cordoval	hi guys	03:01
	bisect run failed:	03:01
	exit code 255 from 'build/bisect.sh' is < 0 or >= 128	03:01
	wonder why the exit code do not consider 255	03:01
	→ hashset joined	03:02
	← harish left	03:02
	→ Eridius joined	03:02
	← Fingolfin left	03:05
	→ dreiss joined	03:07
	← Esa_ left	03:08
	→ Anders_J joined	03:09
	← glennpratt left	03:09
	→ mh` joined	03:10
	← xplat left	03:10
	← lolfrenz left	03:11
jim	cordoval, my only suggestion is look at build/bisect.sh to see why it would throw that code	03:11
cordoval	jim is an exception of course	03:11
	but how to make it work even that exception	03:11
jim	find out what circumstances would have it thrown by reading the script (or is your goal something other than finding out what's wrong)	03:12
cordoval	yes when it started to give that exception	03:12
	← oogatta left	03:12
	ivan\\ → ivan\	03:13
	→ yawk joined	03:13
	← bitkiller left	03:14
	← sebastorama left	03:14
	→ razdaz joined	03:14
	← noyb left	03:14
	→ oogatta joined	03:14
	→ AlbireoX joined	03:15
	→ goweol joined	03:20
	← goweol left	03:20
	← heytrav left	03:21
	→ Hasbro joined	03:21
	← brainbox1100 left	03:22
	→ Goplat joined	03:23
	← e66_ left	03:24
	← hyperair left	03:25
	← fisted left	03:25
	← rubious left	03:26
cordoval	solved it	03:27
jim	:)	03:27
cordoval	thansk	03:27
jim	well you did it :)	03:27
cordoval	git bisect bad	03:27
	and keep going	03:27
	→ g_bleezy joined	03:27
	← cordoval left	03:27
	→ hyperair joined	03:28
	→ fisted joined	03:31
	← Hasbro left	03:33
	← hashset left	03:33
	→ heytrav joined	03:34
	← g_bleezy left	03:34
	← gigamonkey left	03:35
	← juvenal left	03:37
	→ min\|dvir\|us joined	03:37
	← min\|dvir\|us left	03:37
	→ min\|dvir\|us joined	03:37
	← min\|dvir\|us left	03:37
	→ min\|dvir\|us joined	03:37
min\|dvir\|us	Hi. "fatal: Ambiguous object name: 'master'." How do I fix?	03:37
milki	maybe with a --	03:38
	→ etcetera joined	03:38
milki	or fulle path refs/heads/master	03:39
	→ jamescarr joined	03:40
	→ gtuckerkellogg joined	03:40
	← nothingbutlove left	03:40
min\|dvir\|us	I solved my problem by deleting the repo and cloning it again.	03:41
	But now I have a different problem, specific to Bitbucket.	03:41
	Apparently they use the HEAD branch to keep track of which commit is the latest.	03:42
	But that branch doesn't exist when cloning.	03:42
kpreid	HEAD is not a branch	03:42
min\|dvir\|us	I was able to do `git push origin :HEAD` but now I don't know how to make it point to the right branch on the remote.	03:42
	kpreid: I'm aware of that.	03:42
	But on Bitbucket it is.	03:43
	Or a tag of some sort.	03:43
milki	the repo sounds messed up -.-	03:43
	its supposed to be a symbolic-ref	03:43
min\|dvir\|us	milki: can you please help me fix it?	03:44
	milki doesnt even know what :HEAD would do	03:44
milki	i doubt it does what you think it does	03:44
	← neurodrone left	03:44
milki	i dont use bitbucket	03:44
min\|dvir\|us	Oh, :<ref> will delete the ref remotely.	03:44
milki	so no clue	03:44
	but since its a symbolic ref	03:44
	does it delete the HEAD ref or what it points to	03:45
	?	03:45
min\|dvir\|us	milki: the ref itself.	03:45
	← Ivoz left	03:45
milki	are you absolutely sure	03:45
min\|dvir\|us	This has worked on Bitbucket and Github.	03:45
milki	i would have to double check it locally	03:45
	cuz you should never need to worry about it	03:45
min\|dvir\|us	milki: yes, I'm sure. You can do the same with a tag.	03:45
	← tjholowaychuk left	03:46
min\|dvir\|us	Now, my problem is to restore the HEAD branch on the remote, and make it point to master.	03:46
milki	error: unable to delete 'HEAD': remote ref does not exist	03:47
	← wolftankk left	03:47
milki	hmmm	03:47
min\|dvir\|us	I fixed it.	03:47
milki	min\|dvir\|us: if this was github, you would be able to do it via the web interface	03:47
	→ zdevex joined	03:47
min\|dvir\|us	I made a local branch called HEAD, pushed it, and deleted the local branch.	03:48
	← Marbug left	03:48
milki	ya...that totally does not sound right	03:48
	→ Qiang joined	03:48
	← zdevex left	03:48
milki	well	03:48
	if it works	03:48
	good for you	03:48
min\|dvir\|us	I hope I don't run into any issues.	03:48
milki	you probably will	03:48
	:P	03:48
min\|dvir\|us	It fixed the commit log in Bitbucket.	03:48
milki	what does ls-remote say?	03:49
	← serj_ left	03:49
min\|dvir\|us	milki: HEAD, refs/heads/HEAD, and refs/heads/master all point to the same commit.	03:49
milki	refs/heads/HEAD should go away	03:50
	→ dv310p3r joined	03:50
milki	HEAD is fine though	03:50
	maybe bitbucket recreated it on your branch push	03:50
cehteh	oops .. is there something simpler to show the diff between 2 succesive commits than git diff $ID~..$ID ?	03:50
milki	dont use the ..?	03:50
min\|dvir\|us	milki: how would I get it away?	03:50
milki	min\|dvir\|us: try :refs/heads/HEAD	03:51
	→ wolftankk joined	03:51
cehteh	:]	03:51
min\|dvir\|us	milki: sweet. And the commit log in Bitbucket is fine.	03:51
	milki: how do I know that HEAD will always point to master?	03:51
milki	it doesnt	03:52
min\|dvir\|us	milki: I want it to refer to master.	03:52
	Not a specific commit.	03:52
jim	isn't each branchhas its ownmaster?	03:52
min\|dvir\|us	jim: master is a branch.	03:52
	← herdingcat left	03:52
milki	min\|dvir\|us: you said bitbucket is changing it themselves	03:52
	so you should talk to them	03:53
jim	so master has its own head, and other branches do too?	03:53
min\|dvir\|us	milki: no, I didn't say that.	03:53
	milki: I'm changing HEAD, clearly. Bitbucket uses HEAD to show the latest commit.	03:53
milki	a regular push should not change the symbolic-ref HEAD	03:53
	o	03:53
	wait what	03:53
	that doesnt make sense still	03:53
	so you are changing it because bitbucket uses it to show the latest commit...	03:53
min\|dvir\|us	Right. Without HEAD, there's no latest commit.	03:54
milki	why are you changing it?	03:54
min\|dvir\|us	I want HEAD to refer to master.	03:54
	Rather than a particular commit.	03:54
milki	its a symbolic-ref	03:54
	← Fandekasp left	03:54
min\|dvir\|us	milki: how do you know?	03:54
milki	it points to a branch which then resolves to a commit sha	03:54
min\|dvir\|us	Clearly HEAD on Bitbucket isn't HEAD locally.	03:54
milki	of course	03:54
	← dv310p3r left	03:54
min\|dvir\|us	milki: how can I see which branch the remote HEAD is pointing to?	03:54
	← razdaz left	03:55
milki	i dont know if you can actually verify what it is on the remote side	03:55
	ls-remote only shows what it resolves to	03:55
	→ serj joined	03:55
min\|dvir\|us	milki: are you sure about that?	03:55
	→ razdaz joined	03:55
	← AlbireoX left	03:56
	infogulch\|afk → infogulch	03:56
milki	unless you are messing with it yourself, it is by default a synbolic-ref	03:56
min\|dvir\|us	milki: well clearly I'm messing with it.	03:56
	← razdaz left	03:56
milki	as with any regular file on a *nix operating system, you can do whatever you want with the file	03:56
min\|dvir\|us	.git/HEAD shows a ref.	03:57
	(Not a commit.)	03:57
milki	your local HEAD doesnt matter	03:57
min\|dvir\|us	I'd like to confirm that's the case remotely.	03:57
milki	you said the problem is your remote HEAD	03:57
min\|dvir\|us	Right.	03:57
milki	and i dont think you can touch it	03:57
min\|dvir\|us	Well, when I pushed it, that's what was in .git/HEAD.	03:57
milki	unless bitbucket has a web interface for it	03:57
	your .git/HEAD is your local HEAD only	03:58
	it can either be a symbolic-ref or a commit in case of a detached head	03:58
	→ adiabatic joined	03:58
min\|dvir\|us	milki: are symbolic refs transferred to remotes when pushing?	03:58
milki	the remote head refers to the default branch that will be checked out on a clone	03:58
min\|dvir\|us	milki: can't it also be the default commit?	03:59
milki	min\|dvir\|us: the remote head is neighter a commit, branch, or tag	03:59
	so i dont think you can touch it	03:59
min\|dvir\|us	milki: but I did.	03:59
milki	ive never heard of a clone checking out a random commit	03:59
	milki tries	03:59
	→ Fingolfin joined	04:00
	← tmz left	04:00
	← Anders_J left	04:00
	→ nebulabox joined	04:00
	← gtuckerkellogg left	04:01
milki	lol, i cant even clone it now	04:01
	→ AlbireoX joined	04:01
	infogulch → infogulch\|afk	04:01
milki	so i dont think what you think is happening is happening	04:01
min\|dvir\|us	milki: all I know is that the remote HEAD either points to a ref or a commit.	04:01
	← nebulabox left	04:02
min\|dvir\|us	And I want to know which.	04:02
	→ dv310p3r joined	04:02
milki	and im telling you i dont think theres a way to tell	04:02
min\|dvir\|us	Fair enough.	04:02
	→ nebulabox joined	04:03
milki	if theres no bitbucket web thing for it	04:03
	← nebulabox left	04:03
milki	then i have no clue	04:03
	github has a web form for it	04:03
	pretty sure you need access to the remote server	04:03
artnez	min\|dvir\|us: What does git-ls-remote give you?	04:04
milki	and like i tried, you cant clone the repo if HEAD isnt a symbolic ref	04:04
min\|dvir\|us	artnez: shows HEAD and refs/heads/master pointing to the same commit.	04:04
	→ pppaul joined	04:05
	← dv310p3r left	04:05
artnez	min\|dvir\|us: You might look into the various options that git clone gives you.	04:06
	← Davey left	04:06
	← CannedCorn left	04:06
artnez	Perhaps you can clone some partial version of the repo so you could repair it.	04:06
	→ tmz joined	04:07
artnez	Like, maybe get a bare repo.	04:07
milki	clone should not mess with the remote repo	04:07
	→ dv310p3r joined	04:07
artnez	milki: Not saying it will. But if his clone is failing...	04:07
milki	there are few, if only 1 command that can change the remote side	04:07
	artnez: is it?	04:08
	the problem started with ambigious ref 'master'	04:08
	← pppaul left	04:08
min\|dvir\|us	artnez: I never said that.	04:08
	→ wwalker joined	04:09
milki	are you sure bitbucket doesnt have an interface for a default branch?	04:09
	→ Bosox20051 joined	04:09
min\|dvir\|us	milki: I am not sure.	04:10
	→ pppaul joined	04:10
	← pppaul left	04:10
artnez	min\|dvir\|us: In the admin area there is a "main branch" field.	04:11
min\|dvir\|us	artnez: oh, nice! It says "master".	04:11
	← dv310p3r left	04:12
milki	well then	04:12
	thats what HEAD should be	04:12
	next time you think its broken, check there first	04:12
min\|dvir\|us	OK, I'm pretty confident it's fixed.	04:12
	Thanks guys.	04:12
	← min\|dvir\|us left	04:13
milki	awesome	04:13
jim	wow, just finished a lot of separating and committing... took like 2 hours	04:13
	I wanted the commits -clean- and only having one specific change	04:14
	← simesy left	04:14
	→ simesy joined	04:14
artnez	I don't commit often, but when I do, I commit every line individually.	04:15
wwalker	https://gist.github.com/3616451 - I've researched this and i seems that the answers I found don't work.	04:16
	Anyone got a URL or man page that explains what I'm missing? Preferably not "read them all"	04:16
	→ hashset joined	04:16
milki	!doesnt	04:17
gitinfo	[!doesntwork] Sorry to hear it doesn't work. What happened? Did it fall asleep on your couch, with your sister? Eat your homework? What did you want it to do? What happened instead? Please be specific!	04:17
	← tmz left	04:17
nevyn	wwalker: git checkout -t origin/blah	04:18
milki	wwalker: ^	04:18
nevyn	forget the remotes bit.	04:18
milki	oo	04:18
	→ tmz joined	04:18
	← jamescarr left	04:18
milki	right, thats not even the full path anyways	04:18
	← tmz left	04:18
	← jaapbadlands left	04:18
	→ tmz joined	04:18
jim	hey!! I'm NOT a dog, and I DON'T eat my homework!	04:19
adiabatic	Do you eat anyone else's?	04:19
	← tmz left	04:19
	jim slaps gitinfo with a small bull mastief	04:19
	→ tmz joined	04:20
jim	adiabatic, so your temp increases as pressure decreases?	04:20
	← tmz left	04:20
jim	forgot what adiabats do	04:20
adiabatic	no heat goes in, no heat goes out	04:21
jim	but stuff can happen inside?	04:22
	← tmcmahon left	04:22
wwalker	nevyn: tried that, didn't work, exact response in the gist (https://gist.github.com/3616451)	04:23
	→ eedeep joined	04:24
adiabatic	jim: yeah, pretty much: http://en.wikipedia.org/wiki/Adiabatic_process	04:24
eedeep	hi there got a question which I'm wondering if anyone can clarify....scenario is you have a mainline style brnach, which the team is pushing to regularly, let's call it develop. then you have a long running feature branch which you created off develop, and multiple devs are working on that feature branch	04:25
	→ TheJH joined	04:25
nevyn	wwalker: git checkout -t refs/remotes/origin/contact_address ?	04:25
eedeep	ie, its a shared feature branch. if you rebase that branch from develop daily, to keep it fresh, and then push...is that going to cause headaches for the other devs working on that feature branch (because you have altered commit history and published...ie, I have heard that can cause headaches) ?	04:26
jim	eedeep, rebasing, isnt' that rewriting history?	04:27
wwalker	nevyn: same error. so does this mean the branch on the upstream repo is missing something?	04:27
milki	eedeep: yes it will cause headaches	04:27
jim	(any form of which is very bad berries in a multiperson project of any kind	04:27
milki	however	04:27
jim	)	04:27
milki	if all your devs rebase before pushing, it should alleviate the problem	04:28
eedeep	jim: milki : right, suspicions confirmed. but what is the solution then?	04:28
	ah k	04:28
milki	thea headache is if they dont	04:28
eedeep	so just prior to pushing, make sure they rebase	04:28
milki	yes	04:28
	→ gtuckerkellogg joined	04:28
eedeep	milki: jim : ok thanks a lot	04:28
	next phase of learning will be to understand properly why that is	04:29
jim	I'm doing single-person and I still do my best to never rewrite... since I might have multiple clones out there which will get screwed (as an actual person would) and have to reclone and you have to tell them all to reclone it	04:29
	milki always rewrites when single-person	04:29
milki	:P	04:29
	i choose a single repo to act as a master	04:29
	usually a gitolite server	04:30
jim	the best-practices recommendation is you should avoid any history rewriting	04:30
	→ Sigma joined	04:30
milki	only for public histories	04:30
	all bets are off for private histories	04:30
	its called !sausage	04:30
gitinfo	[!sausage_making] Some developers like to "hide the sausage making", transforming their commits before presenting them to the outside world. See http://sethrobertson.github.com/GitBestPractices/#sausage and !perfect	04:30
eedeep	right got ya	04:30
	!perfect	04:30
gitinfo	[!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.	04:30
eedeep	milki: hey that's rad	04:31
jim	well if it's private and still multidev (or multiclones are out there), you can still screw things up fairly badly by getting into the bad habit of rewriting history	04:31
	eedeep, also have you read "git from the bottom up"? cool 30 pg pdf	04:33
milki	jim: tis why theres a central server called gitolite	04:33
eedeep	jim: nah i'll check that out thanks	04:34
	how do I find out what all the ! keywords are in here? they look useful	04:34
jim	it's a really good technical intro	04:34
nevyn	jim: I'm not convince rewrites are useful	04:34
	!factoids	04:34
gitinfo	[!triggers] Please don't spam me! I'm just a poor bot! Here's everything I know: http://jk.gs/git/bot/trigger.php	04:34
nevyn	eedeep: ^^	04:35
eedeep	nevyn: sweet	04:35
	← TheJH left	04:35
jim	nevyn, agreed, that's why I try avoiding them and in the places where it's inconvenient, I still avoid them (knock on wood)	04:35
nevyn	huh	04:36
	I didn't mena to say that.	04:36
	I hate merges ;)	04:36
	I hate the unknownness of merges particularly	04:37
	that arnt' ff	04:37
	← codewrangler left	04:37
	← donri left	04:37
jim	at least if you have to merge you're better off with git than cvs	04:38
	→ toabctl joined	04:39
jim	merges are life... just ask a salmon why he swims upstream	04:39
	← Swathe left	04:39
	→ mindlogger joined	04:40
mindlogger	is it possible to use git and push it to mercurial?	04:41
	I hate branching in hg and I just can't stand its stupidity but at work I must deal with mercurial	04:41
nevyn	wwalker: git checkout -t remotes/origin/blah	04:42
jim	I think so... maybe you can build a push/pull situation	04:42
	so that every time the mercurial changes it triggers a push to git and when you commit to git it pushes to the merc	04:43
	that would be the ideal and most interactive	04:43
milki	ive seen an hg-git before	04:43
	im pretty sure theres a something like a git hg	04:44
	just like theres a git svn	04:44
	and bzr-git	04:44
eedeep	jim: so what is a good way to deal with this scenario where you have a long running shared feature that you need to keep freshening up from the main branch? would merging the main branch into it (rather than rebasing) avoid the problems associated with history rewriting?	04:44
	milki: ^	04:44
milki	no, you should rebase	04:44
	you will have lots of merge problems later if you dont rebase	04:45
artnez	eedeep: You don't have to rebase continuously. You can rebase when you're done with your feature.	04:45
milki	unless you want to deal with them when you merge the feature back	04:45
artnez	eedeep: I actually like the continuously-merge-then-rebase-when-done way because I can keep track of merge conflicts.	04:45
	→ mritz joined	04:45
eedeep	yeah I know....dealing with the catastrophic merge at the end (if you haven't been rebasing) is undesirable i think	04:45
	← gtuckerkellogg left	04:45
milki	artnez: o, will rebase work the same way as if there were no merges then?	04:46
	o..except for the merge conflicts...	04:46
eedeep	artnez: you mean merge the mainline branch into your feature regularly?	04:46
	instead of rebasing	04:46
artnez	milki: Yes if you `git rebase master` it'll nuke all the merge commits.	04:46
	→ irqq joined	04:47
artnez	eedeep: Yes. Then when you're ready to merge your feature back into the maineline, do a rebase first.	04:47
milki	artnez: oo, didnt know that	04:47
	then i guess merging is perfectly fine then	04:47
eedeep	artnez: oh, interesting	04:47
	so in that scenario, instead of rewriting history regularly, you're just doing it once, but in a big way	04:48
	← Sigma left	04:48
	← crocket left	04:50
	← c98_ left	04:51
	← Bosox20051 left	04:52
	← jjang left	04:55
	→ beautifulmind joined	04:57
	← vitor-br left	04:58
	→ jjang joined	05:00
	→ _iron joined	05:00
	→ boombatower joined	05:01
	→ Milossh joined	05:02
	→ c98_ joined	05:02
	→ crocket joined	05:02
	← nitrix left	05:02
	→ Swathe joined	05:03
	→ thiago joined	05:04
	→ binger\|away joined	05:05
	← binger\|away left	05:05
	→ hadi joined	05:05
	← hadi left	05:06
	pdelvo\|afk → pdelvo	05:06
	← binger left	05:07
	← quazimodo left	05:08
	← Petskull left	05:10
	← sawdisk_ left	05:14
	→ WillMarshall joined	05:14
	→ sawdisk_ joined	05:15
	→ und3f joined	05:15
	← etcetera left	05:16
	pdelvo → pdelvo\|afk	05:18
	← jcao219 left	05:21
	← adiabatic left	05:23
	← hyperair left	05:24
	← mritz left	05:24
	→ harish joined	05:25
	← _iron left	05:26
	← avpx left	05:26
	→ harshpb joined	05:27
	→ noyb joined	05:27
	← hashset left	05:29
	← Qiang left	05:32
	← irqq left	05:37
	→ quazimodo joined	05:39
	→ pantsman joined	05:40
	← pantsman left	05:40
	→ pantsman joined	05:40
	→ sangi joined	05:40
	→ Tegi joined	05:42
	← Milossh left	05:42
	→ patrick85 joined	05:43
	→ Milossh joined	05:45
	← AlbireoX left	05:46
	← kokainepizza left	05:48
	← dreiss left	05:49
	← harish left	05:50
	→ ThomasLocke joined	05:51
	→ tmz joined	05:51
	← caseymcg left	05:52
	→ AdrienBrault joined	05:52
	→ gtrak joined	05:52
	← angelsl left	05:53
	→ angelsl joined	05:53
	→ hyperair joined	05:54
	← Eridius left	05:54
	← madewokherd left	05:54
	← boombatower left	05:56
	← gtrak left	05:56
	← f0i left	05:57
	→ Qiang joined	05:57
	← d3vic3 left	05:59
	→ d3vic3 joined	06:01
	→ senny joined	06:04
	← AdrienBrault left	06:05
	← _jmg left	06:05
	→ duckxx joined	06:06
	← angelsl left	06:06
	→ MaximLacrima joined	06:07
	→ dvide joined	06:07
	→ soee joined	06:08
	← duckxx left	06:09
	→ kokainepizza joined	06:11
	← Fingolfin left	06:12
	← banghouse left	06:15
	→ cbreak-work joined	06:15
	→ j416 joined	06:17
	← mindlogger left	06:17
	→ mishok13 joined	06:17
	→ flijten joined	06:18
	← Milossh left	06:21
	→ gun-lpt joined	06:22
	← xxthink left	06:22
	← j416 left	06:23
	→ Marbug joined	06:24
	← mishok13 left	06:26
	← pdelvo\|afk left	06:28
	← Swathe left	06:28
	→ arti_t joined	06:28
	→ sebrock joined	06:29
	← unreal left	06:29
	→ unreal joined	06:30
	→ pdelvo\|afk joined	06:30
	← arti_t left	06:30
	→ JohnBat26 joined	06:37
	→ fisted_ joined	06:37
	→ Swathe joined	06:39
	← fisted left	06:40
	← noyb left	06:41
	→ blami_ joined	06:42
	← toabctl left	06:43
	→ mabrand joined	06:45
	→ giallu joined	06:46
	← giallu left	06:46
	→ giallu joined	06:46
	→ Advocation joined	06:46
	← Fuuzetsu left	06:47
	→ skorgon joined	06:48
	← skorgon left	06:48
	→ skorgon joined	06:48
	→ jonnu joined	06:48
	→ Fuuzetsu joined	06:49
	← diegoviola left	06:49
	← cenuij left	06:55
	→ RaceCondition joined	06:55
	← beautifulmind left	06:55
	← Flibberdy left	06:56
	→ Flibberdy joined	06:56
	→ dsantiago joined	06:56
Flibberdy	Anyone have a good link on setting up certificate based auth between a windows client and a linux server?	06:56
	← JohnBat26 left	06:57
	→ mishok13 joined	06:57
	← pdelvo\|afk left	06:57
	→ goweol joined	06:58
	→ pdelvo\|afk joined	06:58
	→ ph^ joined	06:59
	→ stepnem joined	07:00
IcePic	Flibberdy: any guide on using ssh with keys would work, I guess	07:00
	← pawdraig left	07:01
Flibberdy	IcePic: oh so if you set up ssh certs for interactive logins then git'll work too? interesting, thanks	07:01
IcePic	git rides on top of ssh, so any auth based on ssh will work for git also, of course.	07:02
	dont know if ssh-keys works as "certs" for you, I think it does, but thats just an opinion of mine.	07:03
Flibberdy	yeah that's pretty much what I meant	07:04
	← ph^ left	07:04
	← gavin_huang left	07:05
	→ gavin_huang joined	07:05
IcePic	some people have odd workplace requirements.	07:05
	← acl_ left	07:05
IcePic	wasnt trying to imply you had no clue or anything like that.	07:06
	← snappy left	07:06
	→ JohnBat26 joined	07:06
	→ ph^ joined	07:06
Flibberdy	IcePic: no that's fine. I'm partly responsible for setting the workplace requirements so it's free range at the moment	07:08
	→ pdelvol joined	07:08
	lb_ → lb	07:09
	← gun-lpt left	07:09
IcePic	could have been one of those "oh, ipsec isnt an 'ssl vpn', and the docs says we must run a ssl vpn so..."	07:09
	→ dpino joined	07:10
	→ hwrd\|work joined	07:10
Flibberdy	honestly we're trying to get this as passwordless but still secure as possible	07:10
	because users are lazy and complain a lot	07:11
	← dr_lepper left	07:11
	← kokainepizza left	07:11
IcePic	yes, and ssh-keys will be fine. That is the default mode for gitolite, so if you get them to use some kind of ssh-agent, you're all set. (putty has pageant or what its called)	07:11
	that <- meant 'ssh' in the above sentence.	07:11
	← pdelvo\|afk left	07:12
	← mabrand left	07:12
	→ mastro joined	07:12
	← mastro left	07:13
	→ mastro joined	07:13
	← White-- left	07:14
	→ cenuij joined	07:15
	← cenuij left	07:15
	→ cenuij joined	07:15
	← Fuuzetsu left	07:18
	→ stodan joined	07:18
	→ valtih joined	07:19
	→ Fuuzetsu joined	07:19
	→ diegoviola joined	07:19
	← bizhanMona left	07:20
	→ snappy joined	07:23
	← snappy left	07:23
	→ snappy joined	07:23
	→ bizhanMona joined	07:23
	→ dhruvasagar joined	07:27
	→ OOPMan joined	07:28
	→ mitchell joined	07:28
	→ Pitel_laptop joined	07:29
	→ herdingcat joined	07:30
Pitel_laptop	is there something like `git add --already-staged-files`	07:30
	→ arkiver joined	07:30
	→ chimay joined	07:31
	← chimay left	07:31
	→ chimay joined	07:31
	← cehteh left	07:32
	→ ngirard joined	07:33
	→ jceb joined	07:33
	← simesy left	07:33
	← WillMarshall left	07:33
	← blami_ left	07:35
artnez	Pitel_laptop: That doesn't make sense. If it's staged, it's already staged.	07:36
lb	Pitel_laptop: do you mean " already tracked files" ? then it's git add -u	07:36
	← quazimodo left	07:37
Pitel_laptop	I mean, I do `git add somefile`, then I do some changes to that file, and I want to automaticaly add the already added and chnaged files again.	07:37
lb	no	07:37
Pitel_laptop	maybe it's not called stage but index...	07:37
lb	you have to git add it again	07:38
	← foraks left	07:38
Pitel_laptop	ok	07:38
	→ f0i joined	07:38
	← maletor left	07:39
	→ klj613 joined	07:39
	← d3vic3 left	07:39
	← cenuij left	07:39
	→ blami_ joined	07:40
	→ jargon- joined	07:41
	← jargon- left	07:41
	→ jargon- joined	07:41
	← f0i left	07:42
	→ cenuij joined	07:43
	← cenuij left	07:43
	→ cenuij joined	07:43
	← Freeaqingme left	07:43
	→ Freeaqingme joined	07:43
	← rchavik left	07:43
crocket	git-gui and git-extensions fail to add files whose filenames are in either CP949 or UTF-8 encoding.	07:45
	Only "git add" adds them properly.	07:45
	Why?	07:45
	I use git on windows.	07:45
	← pantsman left	07:45
crocket	Windows XP SP3.	07:45
	→ amcsi_work joined	07:46
	← dogmatic69 left	07:46
	← Taichou left	07:47
	→ dfdsfds joined	07:48
lb	crocket: IMHO: gui's "suck" always. Windows and UTF8 "suck" always, and windows + git is something than can maybe under heavy circumstances work. but try to avoid it. you'll save a lot of brainpower and nerves and time and whatever.	07:48
	→ leokaito joined	07:48
	→ Taichou joined	07:48
crocket	lb : avoid what? Do I need to avoid git on windows at all costs?	07:49
	← ISF left	07:49
lb	crocket: IMHO :yes	07:49
	→ pantsman joined	07:49
	← pantsman left	07:49
	→ pantsman joined	07:49
lb	crocket: no, avoid windows at all costs, not git ;)	07:49
crocket	lb : I can't avoid windows.	07:50
	← dfdsfds left	07:50
crocket	lb : LabVIEW development happens on windows.	07:50
	← arkiver left	07:50
crocket	My boss wouldn't allow me and others to use linux at work.	07:50
	And my coworkers hate linux.	07:50
	→ g33k joined	07:50
crocket	More accurately, they fear linux.	07:50
klj613	lol	07:50
skorgon	crocket, use a linux VM	07:50
DeadZen	well thats fine	07:51
	use freebsd	07:51
	← g33k left	07:51
crocket	FreeBSD is more fearful.	07:51
	skorgon, That would only complicate matters for people.	07:51
DeadZen	FreeBSD IS DA DEBIL	07:51
crocket	And I like to visualize git with GUI.	07:51
skorgon	why people? only you would know. nobody else needs to know you're secretly using linux	07:51
	← Taichou left	07:52
	→ arkiver joined	07:52
crocket	skorgon, maybe	07:52
	← jceb left	07:53
crocket	Is it ok to compress git database?	07:54
	← diegoviola left	07:54
	→ gun-lpt joined	07:54
lb	crocket: yes	07:55
	← cenuij left	07:55
lb	if you mean "database" = ".git"	07:55
	→ teotwaki joined	07:55
	→ _Qman joined	07:55
	→ g33k joined	07:56
	← g33k left	07:56
crocket	damn	07:57
	→ phluks joined	07:57
crocket	Git takes very long for every operation on a repo with 6000~7000 objects.	07:57
	garbage collection is slow as hell.	07:57
lb	try git gc	07:57
	tu run it manually	07:57
	→ koltroll joined	07:58
crocket	lb : Too late, it's already running.	07:58
lb	;)	07:58
	← Swathe left	07:59
	← jh3 left	07:59
crocket	lb : hell	08:00
phluks	Yo! guys - it seems I need to learn something about refs: https://gist.github.com/3618183	08:00
frogonwheels	phluks: generally you don't need to deal with refs by themselves.	08:01
	← Goplat left	08:01
frogonwheels	phluks: what you have here is a branch (refs/heads/) and a remote tracking branch (refs/remotes//* )	08:01
	→ g33k joined	08:02
frogonwheels	phluks: remember also that refs can be packed into a file.	08:02
phluks	frogonwheels: yes, id expect to find it in remotes, but i didnt	08:02
	frogonwheels: oh, how and where?	08:02
	→ babilen joined	08:03
	← pantsman left	08:03
crocket	Does anyone use git GUIs?	08:03
phluks	frogonwheels: im trying to find out wether branches are in sync without checking them out - in a script	08:03
frogonwheels	phluks:it's in .git/packed-refs but you can do that without muckin' about like that.	08:04
Sajbar	crocket: I use 'git gui'	08:04
	and gitk	08:04
frogonwheels	phluks: look at man git-rev-parse	08:04
gitinfo	phluks: the git-rev-parse manpage is available at http://git-scm.com/docs/git-rev-parse.html	08:04
	→ jh3 joined	08:05
frogonwheels	phluks: you can see if they are the same with the output of git rev-parse unicode and git rev-parse origin/unicode	08:05
	sorry unicode was my test branch name. I meant git rev-parse $branchname ... &c	08:05
	← g33k left	08:06
	← arkiver left	08:06
phluks	frogonwheels: Ah, nice - i understand :-)	08:06
	frogonwheels: thanks	08:06
	← rgr left	08:07
	→ g33k joined	08:07
	← Fuuzetsu left	08:07
	→ rgr joined	08:07
	← eighty4 left	08:07
	→ arkiver joined	08:08
frogonwheels	phluks: if you want to see if A is a descendent of B then check that git rev-list -1 ^A B returns empty	08:08
	→ eighty4 joined	08:09
	← g33k left	08:09
	← arkiver left	08:09
frogonwheels	phluks: if you want to then see by how many.. i THINK that git rev-list A ^B \| wc - l should give it.	08:10
	wc -l	08:10
phluks	frogonwheels: well, in this case I can't know wich might be behind or in front. Its ok, if i know wether they are the same or not	08:10
	← Vile left	08:10
	→ angelsl joined	08:10
	→ w41__ joined	08:11
	→ LimeBlast joined	08:11
frogonwheels	phluks: you shouldn't really have to deal with .git dir .. except for perhaps .git/description and .git/info/exclude (and sometimes .git/config	08:12
	← _huoxito left	08:12
	→ mr6 joined	08:13
frogonwheels	phluks: man git-show-ref .. you can get at raw refs using this command :)	08:13
gitinfo	phluks: the git-show-ref manpage is available at http://git-scm.com/docs/git-show-ref.html	08:13
phluks	frogonwheels: nice to know. Sometimes the right way is not obvious but the wrong way is :-)	08:13
	← w41_ left	08:14
	→ gerryvdm_mbp joined	08:14
phluks	frogonwheels: yup	08:15
	← hwrd\|work left	08:16
	drarok → Drarok	08:17
	← __machine left	08:18
gerryvdm_mbp	i'm creating a programming course and i'd like to put the exercises in Git, does anyone have some great tips and ideas or resources i could look into?	08:18
	→ dfgdfgdfg joined	08:18
dfgdfgdfg	"git log origin.." shows 4 commits out of which 2 and 3 are merge branch 'master' into my local branch.. i want to combine commit 1 and 4 into a single one.. is that doable by git rebase -i HEAD^4	08:18
	→ Kingsy joined	08:20
Kingsy	if you have a git hook, like post-recieve or something, does that literally just execute its contents with bash? so ~/whatever/hello.sh would be a valid line in that file? and hello.sh would execute on each hook?	08:22
	→ dr_lepper joined	08:23
	SeoZ → SeoZ-work[AWAU]	08:23
	→ PigeonFriend joined	08:23
mr6	dfgdfgdfg: did you mean get log origin/master.. ?	08:24
	→ JENSU joined	08:25
	← dpino left	08:25
	← Moussekateer left	08:25
	→ gtuckerkellogg joined	08:25
	gitinfo set mode: +v	08:26
	→ dfr\|mac joined	08:26
JENSU	what's the difference between pull and fetch? 2 euro	08:26
	what's the difference between pull and fetch? 2 euro	08:26
	→ Moussekateer joined	08:26
	Kiall → zz_Kiall	08:26
jim	JENSU, one is part of the other (I think)	08:28
	pull is fetch+merge (I -think-)	08:29
	← Moussekateer left	08:29
	→ pdelvo\|afk joined	08:29
	→ tjegels joined	08:29
JENSU	thanks bro;)	08:29
	← JENSU left	08:30
	→ achadwick joined	08:30
tjegels	have any of you had lots of conflicts when regularly doing git pull --squash?	08:30
	seem to be getting conflicts way more than usual now	08:30
	← dfr\|mac left	08:30
jim	check the man page, I may have that wrong (but, I've been using pull, and it gets, followed by a fast-forward which I think is a merge)	08:31
lb	jim: !pull	08:31
gitinfo	jim: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus not the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/	08:31
	← pdelvol left	08:31
jim	oh I did remember right :) thanks :)	08:31
lb	:)	08:31
	→ jonnu[ghost] joined	08:31
	← jonnu left	08:31
	→ cebor joined	08:32
tjegels	thanks for the article link, will check it out	08:33
	zz_Kiall → Kiall	08:33
jim	!git from the bottom up	08:33
gitinfo	[!welcome] Welcome to #git, a place full of helpful gits. If you have a question, feel free to just go ahead and ask—somebody should answer shortly. For more info on this channel, see http://jk.gs/git/ Take backups (type !backup to learn how) before taking advice.	08:33
jim	lb, does it have a factoid for that?	08:34
lb	?	08:34
jim	lb, for the pdf "git from the bottom up"	08:34
lb	jim what are you refering to? i'm confused ^^	08:35
	you mean !bottomup	08:36
gitinfo	'Git from the bottom up' starts with explaining the building blocks of git and proceeds to tell you how they fit together. http://ftp.newartisans.com/pub/git.from.bottom.up.pdf	08:36
lb	?	08:36
jim	yes :) thanks	08:36
lb	then you maybe want too see !books	08:36
	← ciscon left	08:36
lb	then you maybe want too see !book	08:36
gitinfo	There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable	08:36
	→ dsirijus joined	08:36
	→ dfr\|mac joined	08:36
jim	great :)	08:37
	→ ciscon joined	08:37
	← artnez left	08:38
	→ dpino joined	08:39
	→ tatsuya_o joined	08:39
	← xiangfu left	08:39
	→ xxthink joined	08:40
	← herdingcat left	08:40
	← soee left	08:43
	adhawkins-away → adhawkins	08:44
	→ g33k joined	08:46
	← g33k left	08:46
	← dpino left	08:48
	← dfr\|mac left	08:49
	→ rmr joined	08:51
	← spec__ left	08:53
	→ xiangfu joined	08:54
	→ codeM0nK3Y joined	08:56
	→ Swathe joined	08:58
	→ Vile joined	08:59
	← legnaleurc left	09:00
	→ Taichou joined	09:02
	← kerrick left	09:03
	← oojacoboo left	09:03
	← ricky left	09:03
	→ jlg13 joined	09:04
	→ oojacoboo joined	09:05
	← Swathe left	09:05
	→ vervic joined	09:06
	← Taichou left	09:07
	→ legnaleurc joined	09:08
	→ Swathe joined	09:09
	← maqr left	09:09
	← robbyoconnor left	09:09
	→ spec__ joined	09:10
	→ d3vic3 joined	09:11
	← d3vic3 left	09:11
	→ d3vic3 joined	09:12
	← d3vic3 left	09:13
	→ d3vic3 joined	09:13
	← d3vic3 left	09:14
	← xiangfu left	09:14
	← ph^ left	09:15
	← jdyin left	09:15
	→ ph^ joined	09:16
	→ jdyin joined	09:16
	→ mklappstuhl joined	09:19
	← Advocation left	09:20
	→ ricky joined	09:21
	→ berserkr joined	09:21
	← mitchell left	09:21
phluks	whats !parable	09:22
	← Kingsy left	09:23
gitinfo	'The git parable' provides some good reasoning behind git. http://tom.preston-werner.com/2009/05/19/the-git-parable.html	09:23
	→ shruggar joined	09:23
	→ Kingsy joined	09:25
	← Kingsy left	09:25
	→ Kingsy joined	09:25
	→ tatsuya__ joined	09:25
	← vervic left	09:26
	→ Vert joined	09:27
	← tatsuya_o left	09:27
	→ irqq joined	09:28
	→ papegaaij joined	09:29
	→ Taichou joined	09:30
	← jonnu[ghost] left	09:32
	← sgambino left	09:35
	← gtuckerkellogg left	09:35
	→ sgambino joined	09:36
	→ xiangfu joined	09:39
	← sgambino left	09:39
	→ Geralt joined	09:39
	← Geralt left	09:39
	→ Geralt joined	09:39
	← legumbre left	09:40
	← thiago left	09:44
	→ Akufen joined	09:46
	← Akufen left	09:46
	→ Akufen joined	09:46
	→ pdelvol joined	09:46
klj613	is there a way to use `git checkout <treeish> -- filename.ext` and get it to output to console? easy to search for a specific version of the file?	09:47
skorgon	klj613, not sure, but are you probably looking for git show?	09:48
	→ arietis joined	09:49
	← pdelvo\|afk left	09:49
	← shruggar left	09:50
	→ atesc joined	09:51
	← atesc left	09:52
	→ shruggar joined	09:52
klj613	skorgon, not sure. at the moment im using `git blame HEAD~150 -- file.ext` etc.	09:53
	→ tmoskowite joined	09:53
skorgon	klj613, if you wanna look at a file at a specific ref you can use 'git show <ref>:<path>'	09:54
klj613	skorgon, thanks	09:55
	← patrick85 left	09:55
klj613	skorgon, the <path> got to be absolute to the root of the repo? lol. "file.ext doesnt exist but src/foo/bar/file.ext does"	09:56
	ah nevermind, works with ./	09:56
	← rgr left	09:57
	→ patrick85 joined	10:00
	← yatiohi left	10:00
JrCs	!cs	10:01
gitinfo	"Git for Computer Scientists" is a quick introduction to git internals for people who are not scared by phrases like Directed Acyclic Graph. http://eagain.net/articles/git-for-computer-scientists/ See also !concepts !bottomup	10:01
	→ ka2be joined	10:02
JrCs	!concepts	10:02
gitinfo	"Git Concepts Simplified" explains the basic structures used by git, which is very helpful for understanding its concepts. http://sitaramc.github.com/gcs/	10:02
	→ martinjlowm-lapt joined	10:03
	← Geralt left	10:07
	→ noisebleed joined	10:11
	← noisebleed left	10:11
	→ noisebleed joined	10:11
	← xiangfu left	10:15
	→ Dave^\| joined	10:15
	→ thiago joined	10:19
	→ phantomnomad joined	10:20
	← ipalaus left	10:20
	→ Element9 joined	10:20
	← martinjlowm-lapt left	10:21
	→ ipalaus joined	10:21
	→ d3vic3 joined	10:22
	← koltroll left	10:22
	→ snearch joined	10:25
	→ beautifulmind joined	10:27
	← SeySayux left	10:28
	← d3vic3 left	10:28
	← _W_ left	10:28
	→ _W_ joined	10:29
	← dsirijus left	10:29
	← tatsuya__ left	10:30
	→ tatsuya_o joined	10:31
klj613	how do i get version history of a single file? e.g. the commit SHA it started in, ... to the latest commit SHA it was modified etc?	10:31
frogonwheels	klj613: git log -- filename iirc	10:33
	← jjang left	10:34
	→ SeySayux joined	10:35
	← tatsuya_o left	10:35
	→ maqr joined	10:35
	← Cros79 left	10:37
	← gavin_huang left	10:38
	spyro → pink_panther	10:38
	→ g33k joined	10:40
	→ luc__ joined	10:41
	← hyperair left	10:41
	→ Roccivic joined	10:41
	← g33k left	10:42
	← OOPMan left	10:43
Roccivic	hi all	10:43
gitinfo	Roccivic: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.	10:43
	→ flaguy48 joined	10:43
	← maqr left	10:44
	← luc__ left	10:45
Roccivic	I'm doing a rebase on abranch and I'm not sure what has changed in a file where I have to resolve a conflict. So the question is: how can I list all commits that have touched a particular file? Many thanks :)	10:45
Sajbar	Roccivic: if you have gitk you can just do a gitk <file>	10:46
thiago	Roccivic: git log -- filename	10:46
Roccivic	thiago: thanks, exactly what I was looking for :)	10:47
	→ pfFredd joined	10:49
	← senny left	10:52
	→ senny joined	10:53
	← senny left	10:53
	→ kurrataW joined	10:53
	→ senny joined	10:53
	← Roccivic left	10:54
	← BombStrike left	10:55
	→ jceb joined	10:57
lb	thiago: what does the "--" stands for ? i use it always without them..?	10:58
bremner	lb: what happens if you have a branch and a file with the same name?	10:59
lb	ahha, never happend so far :)	10:59
	→ OOPMan joined	10:59
lb	ok now i see. thanks bremner	10:59
thiago	lb: there's another benefit. -- indicates that it is a path.	11:00
	including a path referring to a file that doesn't exist	11:00
	git log filename only works if filename is a file and it exists right now	11:00
lb	ok, will use it from now on. thanks	11:00
thiago	git log -- filename works even if the file has been deleted	11:00
	→ arvind_khadri joined	11:00
lb	hm, i'd like to flattr users for good answers....	11:01
	→ pdelvo\|afk joined	11:01
bremner	lb: you can use e.g. !thiago++ iirc	11:02
	hmm.	11:03
	thiago++	11:03
	bremner gives up	11:03
lb	what should happen?	11:03
bremner	some message from the bot about karma	11:03
lb	ahh :)	11:03
	gitinfo set mode: +v	11:04
	← Akufen left	11:04
	← pdelvol left	11:05
lb	bremner: !karma	11:05
gitinfo	bremner: This channel tracks karma based on who has gotten lots of thanks for being helpful. If you want to help someone reach karmic nirvana, please mention their name when thanking them. Try ".karma <nick>" or ".topkarma", but please do it via private message to avoid mass highlighting people.	11:05
	← wolftankk left	11:05
	→ ph^_ joined	11:05
	← ph^_ left	11:05
lb	Thank you thiago	11:05
	maybe that helped ;)	11:05
	← ph^ left	11:08
	→ QaDeS joined	11:09
	← Qiang left	11:10
_ikke_	lb: convention is: name: thanks, or something like that	11:10
lb	_ikke_: thanks	11:10
	thiago: thanks	11:11
	^^	11:11
	→ koltroll joined	11:11
	→ hyperair joined	11:11
_ikke_	:D	11:11
	→ akiress joined	11:14
	← spec__ left	11:14
	← cebor left	11:15
	← chimay left	11:16
	← arvind_khadri left	11:17
	→ wolftankk joined	11:19
	← ka2be left	11:19
	← jdyin left	11:21
	→ nbari joined	11:21
	→ xplat joined	11:22
	→ spaceonline joined	11:22
	→ zeppo joined	11:24
	→ rchavik joined	11:24
	← rchavik left	11:24
	→ rchavik joined	11:24
	→ Ivoz joined	11:24
	← nbari left	11:26
	→ spec__ joined	11:28
	← jlg13 left	11:29
	→ DanyO83 joined	11:31
	→ Chillance joined	11:32
	→ iocor joined	11:34
	→ Pupeno_W joined	11:34
	→ rwestphal joined	11:35
	→ pandeiro joined	11:35
	→ maqr joined	11:40
	→ gavin_huang joined	11:41
comps	does anybody know why the history printout of an interactive rebase was reversed some time after v1.5.x ? ... I seem to recall the latest commit on top, with older commits below it, now (v1.7.11.2) it's the other way around, oldest commit on top	11:42
	I tried searching (by commit message) in the 'git' history, nothing useful showed up	11:43
	← wolftankk left	11:43
selckin	1.5 is like years ago right?	11:44
comps	pretty much, yeah	11:44
selckin	i've never know it to be different, 'fixup/squash' etc is relative to the line above, which makes sense to me	11:45
	← maqr left	11:45
comps	well in my case, it's relative to the line below, .. perhaps fedora has some broken version of git-rebase	11:45
	→ sgambino joined	11:46
akheron	comps: I don't understand	11:47
	if you now have the oldest commit on top, doesn't a fixup/squash line get combined with the line above, i.e. the older commit?	11:47
comps	oh, it's relative to the commit above, nevermind ... though I still remember newer commits being on top and squashing older commits into newer ones	11:48
sitaram	comps: I don't remember it ever being that way, TBH, but then I wasn't on 1.5.x for more than a couple of weeks	11:49
	→ btree__ joined	11:49
	← harshpb left	11:50
akheron	I've used git since 1.5.4 and I don't remember it ever being the other way round	11:50
	→ aircan joined	11:50
	← aircan left	11:51
comps	then perhaps my memory has failed me again, I'll verify that in a moment on debian lenny	11:52
cbreak-work	squash is like commit --amend	11:52
	→ nitrix joined	11:52
	→ the_cybersphinx joined	11:53
	← hellopat left	11:53
	← btree_ left	11:53
jast	interactive rebase was introduced in 1.5.3	11:53
	→ cannonball joined	11:53
	→ v0n joined	11:53
	← v0n left	11:53
jast	the initial version in commit 1b1dce4bae760248a1fc3e29548a72c446e77270 already had --reverse in it	11:54
	FauxFaux peers.	11:54
	→ speakman joined	11:54
	→ zarubin joined	11:55
	FauxFaux also believes he's seen it backwards.	11:55
speakman	hi folks! How can I look in the object store for a blob (file) containing a certain string? I've lost some changes in my working tree.	11:55
FauxFaux	git log -S / git grep	11:55
akheron	comps: I have access to a Ubuntu 6.06 which has git 1.5.4.5, and I just checked it gives the oldest commit first :)	11:56
FauxFaux	git fsck --unreachable \| fgrep blob \| awk '{print $NF}' \| xargs -1 git show	11:56
cbreak-work	git add -p needs --patience	11:56
akheron	so at least my memory didn't fail	11:56
	→ SkAzZ joined	11:56
cbreak-work	as does any other git command that works with diffs	11:57
	pdelvo\|afk → pdelvo	11:57
comps	akheron: you're right, 1.5.6.5 @ lenny does the same, I was wrong, sorry for the wasted time	12:00
	← acrocity left	12:01
akheron	comps: np, this was a fun excursion to the history :)	12:01
	→ Ryback_ joined	12:02
	← Ryback_ left	12:02
	→ Ryback_ joined	12:02
lb	there are more people alive in here than it seems to be :)	12:02
	→ _jmg joined	12:02
	→ wolftankk joined	12:02
	→ incognit0 joined	12:04
	← sangi left	12:04
incognit0	If I have a server at 'git://', how can I find/navigate to a repo? I don't know what's on the server.	12:04
jast	you can't	12:04
incognit0	I know there's one repo I need to work with, I simply don't know the address.	12:04
csmrfx	I have seem to have forgotten how to get one-line-per-commit history log?	12:04
jast	if the server doesn't have nother way of telling you, e.g. a web interface, you can only guess	12:04
csmrfx	or was that a plugin	12:05
jast	csmrfx: git log --oneline	12:05
	Drarok → drarok	12:05
csmrfx	hm, I seem to remember it showing the branches as well...	12:05
jast	csmrfx: git log --oneline --decorate	12:05
lb	csmrfx: --all --decorate	12:05
jast	you may be interested in !lol or !lg	12:05
incognit0	jast: Thanks for the help. I'll just have to wait to hear back from the repo owner.	12:05
gitinfo	git config --global alias.lol "log --oneline --graph --decorate"	12:05
	git config --global alias.lg "log --graph --pretty=format:'%Cred%h%Creset -%C(yellow)%d%Creset %s %Cgreen(%cr) %C(bold blue)<%an>%Creset' --abbrev-commit --date=relative"	12:05
csmrfx	oh yes!	12:05
	thats what it was, an alias	12:06
	thank you, sirs	12:06
	→ snadahalli joined	12:06
incognit0	Ah! I have a clone of the repo who will 'git pull' from that space. How can I determine that URL? Git status doesn't seem to show it.	12:07
	→ acrocity joined	12:08
	← snadahalli left	12:08
FauxFaux	incognit0: git remote -v	12:08
incognit0	FauxFaux: Thanks!	12:08
	← wolftankk left	12:09
	← gun-lpt left	12:11
	→ juvenal joined	12:11
	pdelvo → pdelvo\|afk	12:14
	→ neurodrone joined	12:15
	→ mikedanni joined	12:15
	gitinfo set mode: +v	12:16
mikedanni	Hi. I'm in a small team. I'm looking at having a central git repo on a file server. All the tutorial I find say that I should create a new git user to do this. What's wrong with users using their existing user accounts?	12:16
	→ pdelvol joined	12:16
mikedanni	It's over ssh I should say.	12:16
_ikke_	mikedanni: Using the same repo with different users	12:16
	mikedanni: file permissions and stuff	12:16
FauxFaux	Where did the gitolite !overkill document go?	12:16
mikedanni	mikedanni: so git might do funny stuff, creating files as private rather than group r/w and so on?	12:17
	should I really be running some kind of git server rather than using ssh?	12:17
FauxFaux	mikedanni: In general, group r/w stuff is much harder to handle than installing gitolite from packages.	12:17
	→ pfFred joined	12:17
	→ Stereo joined	12:17
Stereo	hi everyone	12:17
gitinfo	Stereo: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.	12:17
	← pdelvo\|afk left	12:19
Stereo	I have /etc in git on a machine I'm co-administering, and doing a commit is very slow. I tried gc and checking for huge files, with no success, and can't find any other things to try. What else could be making commits slow? This is git 1.7.2.5 on debian stable.	12:19
FauxFaux	git count-objects -v	12:20
	→ keller joined	12:20
sitaram	FauxFaux: went into http://sitaramc.github.com/gitolite/why.html	12:20
Stereo	FauxFaux: http://pastie.textmate.org/private/0ah55coo0tjtgwwfsedfg	12:20
	← pfFredd left	12:20
sitaram	FauxFaux: it was too big so I condensed it into the first part of the "alternatives" section in that link	12:20
FauxFaux	sitaram: Yeah, that does a good job.	12:20
sitaram	mikedanni: see http://sitaramc.github.com/gitolite/why.html for examples and reasoning of why unix permissions are not enough	12:21
FauxFaux	Stereo: That's bogglingly tiny. strace $(git --exec-dir)/git-commit # or wait for someone with a sensible guess.	12:22
	→ Davey joined	12:22
	→ afuentes joined	12:23
Stereo	hmm, it doesn't have --exec-dir	12:23
	→ gun-lpt joined	12:24
	→ monokrome joined	12:24
	← votz left	12:24
monokrome	Hello, all. Does anyone know if gitslave allows you to supply a specific branch for the repositories it clones?	12:24
Stereo	--exec-path seems to be what you meant?	12:24
	(sorry monokrome, I was replying to FauxFaux)	12:25
FauxFaux	Stereo: Yeah, that one.	12:25
	→ mabrand joined	12:25
monokrome	Stereo: I got it. Thanks :)	12:25
	→ yatiohi joined	12:26
sitaram	Stereo: per your nick, you should be able to talk to two people simultaneously anyway :)	12:26
Stereo	sitaram: but monokrome probably can't :)	12:27
sitaram	LOL!	12:27
	← DanyO83 left	12:30
	← JyZyXEL left	12:30
	→ geekbri joined	12:30
Stereo	ok, it hangs for a long time at line 8 in http://pastie.textmate.org/private/tkmxtcqdexp7nervqp7pfq	12:31
	→ SidGBF joined	12:31
Stereo	I'll see if removing the precommit hook makes it faster	12:32
	→ whitman joined	12:32
Stereo	heh, ok	12:32
	← nitrix left	12:32
Stereo	I just ran the pre-commit hook, and it's still running	12:32
	not git's fault then :)	12:32
SidGBF	http://pastebin.com/kBjFHDdT	12:33
	can anyone help me with this error?	12:33
	← acrocity left	12:34
	← Taichou left	12:34
	← phantomnomad left	12:34
	→ kokainepizza joined	12:34
thiago	SidGBF: hmm... I think "mensagem.git" is not a Git repository	12:34
SidGBF	thiago: it is, a bare repo, but it is...	12:34
	← dhruvasagar left	12:35
thiago	SidGBF: what git version are you using?	12:35
	→ dhruvasagar joined	12:35
SidGBF	1.7.4 on server	12:36
csmrfx	!ls .git/hooks	12:36
gitinfo	[!tools] See https://git.wiki.kernel.org/index.php/InterfacesFrontendsAndTools for information about known interfaces, frontends, and tools.	12:36
thiago	and locally?	12:36
csmrfx	oops, sorry	12:36
	← kokainepizza left	12:36
SidGBF	1.7.10	12:36
	→ nitrix joined	12:36
thiago	SidGBF: ok. Do you have shell access to that host?	12:36
SidGBF	yup	12:36
thiago	SidGBF: ssh in and then cd ~/mensagem.git	12:37
	SidGBF: does that exist?	12:37
SidGBF	yes	12:37
thiago	SidGBF: git log there. Does it work?	12:37
SidGBF	inside the folder?	12:38
	fatal: bad default revision 'HEAD'	12:38
	← Dave^\| left	12:38
	← juvenal left	12:40
	← GeorgeJ left	12:40
	→ GeorgeJ joined	12:40
thiago	SidGBF: looks like you have an empty repository	12:40
	cloning empty repositories is a recent feature	12:41
SidGBF	yes... a bare repo	12:41
csmrfx	When a post-commit script runs, what is the path context in the script? the git repo path?	12:41
	pink_panther → spyro	12:41
thiago	SidGBF: bare and empty	12:41
csmrfx	iow, as an idea, just having "cp ./* ../target/dir/foo" would work?	12:42
	← mikedanni left	12:42
	← monokrome left	12:43
cbreak-work	csmrfx: it has the same $PATH as the calling process	12:43
	→ ph^ joined	12:43
	← kcj left	12:46
	← snearch left	12:46
	→ Shanachan joined	12:46
	← Shanachan left	12:46
	→ Shanachan joined	12:46
	Shanachan → Fuuzetsu	12:47
csmrfx	well, tested that, called git commit from vim, no joy	12:47
SidGBF	thiago: and what should I do to be able to clone it?	12:48
	→ gusnan joined	12:49
thiago	SidGBF: make it non-empty	12:50
	SidGBF: or upgrade git	12:50
	← jceb left	12:50
csmrfx	hm, in a hook-shell script, how to block .files but not subdirs like './svg'?	12:50
_ikke_	csmrfx: When exactly do you want to block it?	12:51
csmrfx	post-commit -hook script	12:51
	I guess this is slightly off topic, actually	12:52
SidGBF	thiago: and how to make bare repo be a non empty?	12:52
_ikke_	You want to abort the commit?	12:52
csmrfx	no, just copy files post commit	12:52
	→ jceb joined	12:52
_ikke_	copying files is not blocking them	12:52
	csmrfx goes back to check again	12:53
	← afuentes left	12:53
	→ dendazen joined	12:53
csmrfx	ok, sorry, I mistook the message from cp as msg from git commit	12:53
dendazen	Does anyone know why this all of the sudden started giving me errors/	12:54
	csmrfx goes to add -r to cp	12:54
dendazen	git clone https://u:p@source.blanders.com/ep-compliance-ntp.pkg.git	12:54
	Cloning into 'ep-compliance-ntp.pkg'...	12:54
	fatal: I don't handle protocol 'https'	12:54
	Used to work for my account before.	12:55
thiago	SidGBF: push something there	12:55
SidGBF	thiago: how if I can't clone it?!	12:55
dendazen	And it is regardless on which machine i am.	12:55
	→ skyf joined	12:56
	← albel727 left	12:56
thiago	SidGBF: push without cloning	12:56
lb	SidGBF: create an new git repo, add your bare as a remote, add something, commit, push	12:56
	→ osxorgate joined	12:57
	→ Taichou joined	12:57
	gitinfo set mode: +v	12:57
SidGBF	I wanted something a bit more straighorward	12:57
	*forward	12:58
	← jargon- left	12:58
	← Taichou left	12:58
osxorgate	how can i ask git what the list of ignored files is? it seems my global ignore file is getting.. ignored	12:59
	→ freeed joined	12:59
lb	osxorgate: cat .gitignore	13:00
	→ codewrangler joined	13:01
lb	osxorgate: or try git status --ignored	13:01
SidGBF	Im used to work with svn, with is a code version tool anyway. I really like the idea of creating a repo on the server, then clone it and add my code in it. What's the git's flow for a new repo/first clone?	13:01
	→ albel727 joined	13:01
osxorgate	lb: uhm. i have a ~/.gitignore_global and told git to use it globally with excludesfile etc. yet the file types are still shown when i do git status	13:02
	→ specialGuest joined	13:02
	← specialGuest left	13:02
	→ specialGuest joined	13:02
lb	SidGBF: if there is a repo with stuff available, then you clone, otherwise you just start locally with git init, and add stuff. if you feel like sharing, then create a bare repo on a server and push your stuff up	13:02
	← Jay_Levitt left	13:02
	→ BadDesign joined	13:03
	→ chimay joined	13:03
BadDesign	Why it is called "pull request" instaead of "push request"?	13:03
	Pull it means you grab something from someone	13:03
	It should be push request not pull	13:03
charon	BadDesign: umm, because you ask the receiver of the request to pull from some location you specify?	13:03
	← chimay left	13:03
	→ chimay joined	13:03
lb	BadDesign: you send a request to someone. this person should pull from you	13:03
	← gun-lpt left	13:03
crocket	Has anyone had experiences with teaching git to coworkers?	13:04
	None of them has known version control system in their lifetime.	13:04
BadDesign	yeah but I'm not doing a pull request. he's doing	13:04
lb	osxorgate: sorry. just using .gitignore	13:04
BadDesign	I'm doing a push	13:04
lb	no, you request him to pull	13:04
	→ xiangfu joined	13:04
lb	you do nothin	13:04
crocket	My manager asked me to teach git to people.	13:04
charon	crocket: i always start with some graph-based concepts introduction similar to !cs but spanning all of git. but my advantage is that they're all computer scientists and not afraid of graphs and data structures	13:04
gitinfo	crocket: "Git for Computer Scientists" is a quick introduction to git internals for people who are not scared by phrases like Directed Acyclic Graph. http://eagain.net/articles/git-for-computer-scientists/ See also !concepts !bottomup	13:04
lb	crocket: then you should read a lot before teaching ;)	13:05
csmrfx	crocket: it is good they have no cvs or svn to overcome	13:05
SidGBF	lb: Ok, Im starting to figure this out... Now, what if I want tohave this cloud machine that is my code backup. How can I create first a repo there, and then clone my stuff?	13:05
crocket	lb : I've read 70% of progit.	13:05
csmrfx	crocket: I have the same job todo	13:05
	except these are devs with previous VC exp	13:06
crocket	csmrfx, shit	13:06
lb	SidGBF: like i said. prepare your repo local, then push it to the cloud box	13:06
crocket	csmrfx, They have no idea about VCS.	13:06
charon	crocket: but -- no offense -- my reading from the past few days was that you're not a very experienced user. perhaps try to get some deeper understanding before you start teaching it	13:06
csmrfx	I like how git is lucid in it's behaviour and directory structure -- so I am going for the 'lets install and make a git repo and look at all of .git's files'	13:06
	I call it Hackers Guide To Git	13:07
SidGBF	lb: is there any command that "magically" creates new repo remotelly?	13:07
	→ AaronMT joined	13:07
lb	SidGBF: nope	13:07
crocket	charon : I, myself, am a git beginner.	13:07
	→ erichynds joined	13:07
	← ngirard left	13:07
crocket	charon : It's been 4 days I started learning git.	13:07
csmrfx	crocket: well, if they're computer scientists, the reasons for using a VC or DVC will be easy to understand for them	13:07
crocket	4 days since I started.	13:07
	← setmeaway2 left	13:07
	→ tatsuya_o joined	13:08
	→ timini joined	13:08
csmrfx	crocket: see Schwartz Git video on youtube, also power your workflow with git has succinct overview of all the objs under the hood	13:08
crocket	csmrfx, The problem is I'll have to leave my work for 4 weeks next week.	13:08
	→ ddelrio1986 joined	13:09
lb	crocket: use those 4 weeks, to learn and practice git, then teach your buddies	13:09
SidGBF	lb: git thinks locally. It handles the versions locally (commits). It has the option to push out commits, in order to share/replicate the commits. Is this the idea?	13:09
	(be patient... 6 years of svn, 3 month of git)	13:09
csmrfx	git takes only 7 days to learn, or so	13:10
lb	SidGBF: that's correct	13:10
csmrfx	jmo	13:10
crocket	lb : I'll be conscripted for 4 weeks.	13:10
	→ tmcmahon joined	13:10
crocket	lb : no computer.	13:10
	damn south korea.	13:10
csmrfx	crocket's going to the hills! 8D	13:10
lb	crocket: omg ^^	13:10
	→ jargon- joined	13:11
	→ v0n joined	13:11
csmrfx	crocket: perhaps theres some good git vids in korean, too	13:11
timini	Is it possible to use a submodule with a particular branch?	13:11
SidGBF	lb: Ok, it's a mind changing process... but starts to make sense...	13:11
	Now I need to improve my understanding of bare and non bare	13:12
crocket	csmrfx : pro-git is fully translated to korean, but none of my company's people is willing to read it since they don't know any of VCS.	13:12
lb	SidGBF: bare means, no working copy	13:12
crocket	I had tried to explain how git works for 20 minutes after which my manager freaked out.	13:13
SidGBF	lb: yeah, but, I mean, how do I create a project.git without being bare?	13:13
BadDesign	crocket: switch to a workplace where people know Git already.	13:13
crocket	BadDesign, ha	13:13
csmrfx	crocket: 20 minutes? you can do it in 3 minutes!	13:13
cmn	what you need is to explain the problem	13:13
	→ hide joined	13:13
cmn	explaining a solution without a frame of reference isn't going to get you anywhere	13:13
jast	SidGBF: you might want to read !bare first	13:13
gitinfo	SidGBF: an explanation of bare and non-bare repositories (and why pushing to a non-bare one causes problems) can be found here: http://bare-vs-nonbare.gitrecipes.de/	13:13
csmrfx	Once it is cleas as water for you, then you can teach it	13:14
lb	SidGBF: mkdir newrepo && cd newrepo && git init	13:14
csmrfx	*clear	13:14
SidGBF	git 101 :) I like that	13:14
	lb: Thanks for your time and patience	13:14
lb	SidGBF: and by convention, only bare repos named XYZ.git	13:14
	SidGBF: you're welcome	13:15
crocket	cmn : ok	13:15
	I'll have to memorize examples in progit.	13:15
	→ Jax joined	13:15
	→ macabre joined	13:15
Jax	hello, just tried to push a commit and got this:	13:16
	error: insufficient permission for adding an object to repository database ./objects	13:16
BadDesign	Why you need to memorize? Just explain them those examples using a slideshow or live demo	13:16
	→ RWOverdijk joined	13:16
Jax	have committed and pushed 100 times at least on this repo without a problem	13:16
crocket	BadDesign, ok	13:16
Jax	what could this be?	13:16
jast	cmn: hey, no revealing secret consulting skills here ;)	13:16
speakman	Are there any way to "grep" through blobs/objects which is not in any branch/tag/is unreachable?	13:16
crocket	BadDesign, A slideshow would do.	13:16
cmn	jast: sorry, I'll keep the cabal safe from now on	13:16
	← cooloney left	13:16
charon	speakman: !xy but git fsck --unreachable gives you a list	13:16
gitinfo	speakman: This sounds like an "XY Problem" http://mywiki.wooledge.org/XyProblem So let's step back for a minute. What are you actually trying to achieve? Why are you doing it this way?	13:16
jast	Jax: my guess is someone pushed as a different user and the access rights got messed up because the repo wasn't configured properly for multi-user access	13:17
RWOverdijk	Does anyone know if I can fork public repos into github enterprise?	13:17
jast	RWOverdijk: sure, if you go the general git way of forking	13:17
Jax	jast: hm a chown -R git:git on the repo (on the server) did it.	13:17
	works again.	13:17
jast	you can bet that github won't show it in the list of forks then, though :)	13:17
cmn	RWOverdijk: that's very github-specific, they have their own support channels	13:17
speakman	charon: that's I mentioned earlier I was missing modifications which I know I've done previously but is probably lost due to "git branch -D".	13:17
RWOverdijk	jast, Which is?	13:17
	→ afuentes joined	13:18
RWOverdijk	(never forked git-style before)	13:18
jast	RWOverdijk: clone --mirror; push	13:18
charon	speakman: ah. but then it may be easier (if you recall the commit subject or so) to look in man git-reflog	13:18
gitinfo	speakman: the git-reflog manpage is available at http://git-scm.com/docs/git-reflog.html	13:18
Jax	thx!	13:18
	← Jax left	13:18
	→ gun-lpt joined	13:18
charon	speakman: you lose the branch's reflog by deleting it, but the HEAD reflog may still have the commits	13:18
csmrfx	Does Github for Windows work as a generic git client? (looking for git clients for Windows users)	13:18
jast	RWOverdijk: where the push goes to the new repository, of course :)	13:18
RWOverdijk	jast, I got that bit. Thanks :)	13:18
cmn	RWOverdijk: the gh forking process is to set up some references in the main repo, if you want it to be on a different computer, you'll probably have to transfer the data yourself	13:19
speakman	charon: I've no idea what was actually make me loose the changes, but I though I could just "grep" unreachable blobs.	13:19
BadDesign	csmrfx: No. Yes it will install Git Portable.	13:19
	-no	13:19
RWOverdijk	cmn, I think jast his answer did the trick for me.	13:19
csmrfx	Am I correct in thinking that Windows users only have msgit and cygwin git -clients?	13:19
charon	speakman: so go with git fsck --unreachable \| you write the magic that goes here	13:19
	← bizhanMona left	13:20
csmrfx	*msysgit, I mean	13:20
cmn	csmrfx: no, but they all need git	13:20
	→ ngirard joined	13:20
	← tatsuya_o left	13:21
	→ glennpratt joined	13:22
	→ bizhanMona joined	13:23
	← OOPMan left	13:23
	→ aircan joined	13:23
	→ monokrome joined	13:25
	→ hellopat joined	13:25
	→ pdelvo joined	13:30
	→ bartek joined	13:30
csmrfx	So, any of you windows users have a recommendation for (gui) client?	13:31
	← yatiohi left	13:31
crocket	csmrfx, "Git extensions", "git-cola", "git-gui"	13:31
BadDesign	csmrfx: TortoiseGit	13:32
csmrfx	crocket: which one do you think is the best?	13:32
BadDesign	or SmartGit	13:32
	← lb left	13:32
	← RWOverdijk left	13:32
crocket	csmrfx : I haven't tested them thoroughly, but git-gui doesn't seem to deal properly with UTF-8 file names.	13:32
	→ lb joined	13:33
	← lb left	13:33
	→ lb joined	13:33
crocket	csmrfx, If you have 7000 files to add or commit, git-gui is very slow.	13:33
	→ tplaner_ joined	13:33
	→ robbyoconnor joined	13:33
crocket	BadDesign, In your experience, what is the best?	13:34
	← pdelvol left	13:34
BadDesign	None.	13:34
	← BadDesign left	13:34
crocket	shit	13:34
csmrfx	hm, I wonder if that means: use command line git	13:35
lb	crocket: the shell is your friend	13:35
	→ cenuij joined	13:35
	← cenuij left	13:35
	→ cenuij joined	13:35
	→ tatsuya_o joined	13:35
	→ TeckniX joined	13:37
	→ jn_ joined	13:38
	→ jtri joined	13:38
crocket	lb : "git bash" is very limited compared to cygwin.	13:38
jtri	hi, i've been using git on bitbucket, with http	13:39
	→ agile joined	13:39
crocket	"git bash" also misses man pages.	13:39
lb	i recommended linux yesterday, didn't i? ;)	13:39
jtri	some files are not tracked in the local directory	13:39
	→ adityamenon joined	13:39
jtri	i'd like to switch over to ssl	13:39
	→ wiremaster joined	13:39
crocket	lb : I'd recommend linux to engineers.	13:39
	← adityamenon left	13:39
csmrfx	crocket: what do you mean by "git bash"	13:39
	← babilen left	13:39
jtri	can i do so without doing a clone and trying to salvage the un-tracked files	13:40
	← pfFred left	13:40
crocket	csmrfx, After you install git on windows, git creates msysgit which is called "git bash".	13:40
xiangfu	jtri, git clean -xdf	13:41
csmrfx	oh, you have to do 'git bash add .' or 'git bash commit' in windows command line?	13:41
crocket	csmrfx, no	13:41
	← tmcmahon left	13:41
crocket	csmrfx, If you launch "git bash", a MinGW shell is launched that includes git.	13:41
csmrfx	heh, just checking	13:41
skorgon	jtri, look at man git remote	13:41
gitinfo	jtri: the git-remote manpage is available at http://git-scm.com/docs/git-remote.html	13:41
	→ maqr joined	13:42
	← LekeFly left	13:42
skorgon	you should be able to change the current remote url or add the https url as another remote and then pull/push to that one	13:42
	→ skyf_ joined	13:42
	→ gerard0 joined	13:43
	→ votz joined	13:43
	← Anubisss left	13:43
Flibberdy	I have a private git server (runnign on linux) and a bunch of windows clients. How do I specify that git use a private key when doing push/pull etc?	13:45
	← maqr left	13:46
	← jspiros left	13:46
	→ Ke_ joined	13:46
	→ jspiros joined	13:46
	← glennpratt left	13:47
	← arietis left	13:47
canton7	Flibberdy, you're serving the repos over ssh? Using gitolite, or just a user? If the latter, deny password login for that user (PasswordAuthentication no in sshd_config)	13:47
cmn	Flibberdy: you don't tell git, you tell ssh	13:48
	if you only have one key under ~/.ssh, it should work by itself, but still do the PasswordAuthentication thing	13:49
	← Ke_ left	13:49
	→ hwrd\|work joined	13:49
	→ bitkiller joined	13:49
	← cannonball left	13:50
	← jtri left	13:50
Flibberdy	canton7, cmn, thanks. And it's not gitolite, just standard users	13:51
	this would be so much easier if we all ran linux. Ah welll	13:51
	← votz left	13:53
	→ Fingolfin joined	13:53
	→ kukks joined	13:54
	← kukks left	13:54
	→ kukks joined	13:54
	→ cannonball joined	13:54
	← timini left	13:55
	→ juvenal joined	13:55
	→ dozymoe joined	13:56
	→ glennpratt joined	13:56
	← jn_ left	13:57
	← Fingolfin left	13:57
	← iori left	13:57
	→ mewm joined	13:58
	→ rgsteele joined	13:58
	← warthog9 left	13:58
	→ BadDesign joined	13:59
	← phluks left	14:00
	← kermit left	14:01
	← hellopat left	14:01
	→ les_sylvains joined	14:01
	→ hellopat joined	14:01
	drarok → Drarok	14:03
	→ gdmfsob joined	14:03
	→ ehsan joined	14:04
	→ d3vic3 joined	14:04
	→ alexisabril joined	14:04
	→ AlbireoX joined	14:05
	← tatsuya_o left	14:05
	→ kermit joined	14:05
	→ acrocity joined	14:06
	→ congera joined	14:06
	← mishok13 left	14:06
	← kermit left	14:07
	← acrocity left	14:07
	→ maestrojed joined	14:07
	→ Destos joined	14:08
	→ kermit joined	14:09
	→ bertani joined	14:09
	→ quazimodo joined	14:09
	→ chrisgeorge joined	14:11
	← bertani left	14:11
	→ leeN joined	14:12
	← Zeeshan_M left	14:12
	→ fly9 joined	14:12
	← maestrojed left	14:13
	← mabrand left	14:13
	→ adityamenon joined	14:14
	→ pandeiro_ joined	14:14
	← dozymoe left	14:15
	← dhruvasagar left	14:15
	→ anish[1] joined	14:15
	→ dozymoe joined	14:15
	← wiremaster left	14:16
	→ acrocity joined	14:16
	← pandeiro left	14:16
	← ph^ left	14:17
	→ goombas joined	14:17
	← acrocity left	14:17
	→ jlg13 joined	14:17
	→ wziPc joined	14:17
	← beautifulmind left	14:18
	→ chuckharmston joined	14:18
	→ stewbydoo joined	14:18
	gitinfo set mode: +v	14:19
	→ Guest29525 joined	14:19
wziPc	hello	14:19
gitinfo	wziPc: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.	14:19
	gitinfo set mode: +v	14:19
stewbydoo	hey all	14:19
wziPc	hello	14:20
	← Guest29525 left	14:20
	← congera left	14:20
	← gdelfino left	14:20
	gitinfo set mode: +v	14:20
stewbydoo	so I keep getting this issue when I am trying to use the git help command where it says it cannot find the 'man' directory	14:20
	→ tatsuya_o joined	14:20
	→ Jay_Levitt joined	14:20
stewbydoo	anyone have any suggestions?	14:20
	I am on Windows FYI	14:20
	→ DanyO83 joined	14:21
	← tatsuya_o left	14:21
goombas	if you're developing a web platform and using git for versioning, how would you deal with multiple people editing the files? i know that you can merge changes to the repo, but what testing changes before committing them? should each person be running a local web server on their machine or can everyone develop on a dev server at the same time?	14:21
	→ Synthead joined	14:21
	← Jay_Levitt left	14:21
	← SidGBF left	14:22
	→ Jay_Levitt joined	14:22
mjt	i don't see a prob giving each dev its own webserver "partition" on a dev server.	14:22
canton7	stewbydoo, how did you install git?	14:22
	→ wiremaster joined	14:22
	← wziPc left	14:22
stewbydoo	canton, what specific part are you referring to?	14:22
	what setup wizard options did I choose?	14:23
goombas	mjt: so each has their own partition then commits to the actual merged source?	14:23
mjt	s/commits/pushes/	14:23
	→ wolftankk joined	14:23
canton7	stewbydoo, there are lots of different versions of git for windows. msysgit, the github client, etc etc	14:23
stewbydoo	I went to git-scm.com	14:24
	used the install package there	14:24
	← adityamenon left	14:24
goombas	mjt: thanks :P	14:24
canton7	stewbydoo, aha so that'll be msysgit	14:24
	← sebrock left	14:25
jast	I don't think msysgit has proper manpage viewing set up	14:25
	at least it never worked for me	14:25
canton7	the version i'm running points the browser at <install dir>/doc/git/html/manpage.html	14:25
	← gavin_huang left	14:25
canton7	but other versions have been borked	14:25
	← goombas left	14:25
jast	in any case, the manpages are also available at, for example, http://git-scm.com/docs/git-checkout	14:25
stewbydoo	so do I need to manually point it at the correct location	14:25
canton7	yeah, I normally just use the online docs too	14:26
stewbydoo	right but I would like to be able to access them through bash	14:26
	← d3vic3 left	14:26
	→ goombas joined	14:26
	→ faraway joined	14:26
jast	the best thing you can expect is that it'll launch a web browser that will show you the manpage	14:26
stewbydoo	damn	14:26
jast	so you might as well use the web browser straight away	14:26
	msysgit doesn't include a manpage viewer for the console	14:26
stewbydoo	I can't set it up through the confi	14:26
	config	14:26
	gitinfo set mode: +v	14:26
stewbydoo	damn okay	14:26
	← wolftankk left	14:27
	infogulch\|afk → infogulch	14:27
stewbydoo	so should I use a different git?	14:27
	or just use the online man	14:27
canton7	I doubt any of the others are any better	14:27
	check whether you have the manpages in program files/git/doc/git/html	14:27
	→ Yuuhi joined	14:28
canton7	if not, it's probably a msysgit packaging error, and you could report it to their list	14:28
	(manpages in html format, that is)	14:28
stewbydoo	it is there, its called user-manual in html format and text format	14:28
	It won't bother me if I have to look up the manual online just as long as my install works properly except for that little issue	14:29
speakman	/join #qt-labs	14:29
monokrome	So, anyone know if I can tell gitslave to use specific branches>	14:30
faraway	hi i have a question about git-flow. i currently have a repos with a master branch and some branches that are for the different parts of the projects i'm working on. do i need to finish my currently open feature branches before i switch to git-flow or can i migrate to this ? is there some tutorial how to do this the best way ?	14:30
goombas	mjt: would each dev's repo be a fork of the main repo?	14:30
mjt	s/fork/clone/ ;)	14:30
	→ ishi joined	14:30
ishi	hi	14:30
goombas	:P i'm new to this as you see	14:30
	i've only used git as a single user	14:31
ishi	I do git checkout otherbranch and it's aborting claiming that the checkout would overwrite my local changes. However, when I do status, there are no local changes.	14:31
goombas	both push and pull operations ask to merge changes right?	14:31
canton7	but 'git help checkout' fails with 'cannot find man directory'? Check the config key help.format. also browser.<tool>.path, web.browser, and it all else fails poke the msysgit mailing list	14:32
	stewbydoo, ^	14:32
goombas	if there's a difference between the local and repo files	14:32
cmn	goombas: pull is a merge, push isn't	14:32
_ikke_	faraway: Well, git-flow has two parts. The workflow part, and a git plugin	14:32
cmn	goombas: !pull	14:32
gitinfo	goombas: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus not the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/	14:32
faraway	_ikke_: the git plugin i mean	14:33
	goombas reads	14:33
	← xiangfu left	14:33
canton7	faraway, I'd suggest renaming master to develop, and creating a new master branch. that way your feature branches will be based on develop (as per git-flow)	14:33
_ikke_	faraway: iirc, it doesn't really matter	14:33
	← kermit left	14:33
_ikke_	faraway: You just have to merge it in manually	14:33
stewbydoo	yea fails with that saying there is no man viewer	14:34
canton7	stewbydoo, ah that's a different error to what you said before? what exactly does it say?	14:34
	also, what's the output of 'git config help.format' ?	14:34
	← blami_ left	14:35
stewbydoo	warning: failed to exec 'man': no such file or directory fatal: no man viewer handled the request	14:35
	→ btx94 joined	14:35
	← btx94 left	14:35
cmn	ishi: !repro	14:35
gitinfo	ishi: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session, or at least explain exactly what you did that led up to the problem. This will help immensely with troubleshooting.	14:35
	← BadDesign left	14:35
lb	canton7: stewbydoo is using windows	14:35
jast	yeah, it tries to use 'man' but that's obviously not shipped with msysgit	14:35
stewbydoo	so it's just missing the actual manual	14:36
_ikke_	git help command should open the brwoser	14:36
jast	'man' is the program that displays manpages on a terminal	14:36
stewbydoo	iikke it results in the error above	14:36
canton7	lb, I know. read the scrollback	14:36
_ikke_	stewbydoo: man git foo, or git help foo?	14:36
gitinfo	stewbydoo: the git manpage is available at http://git-scm.com/docs/git.html	14:36
canton7	stewbydoo, what is the output of 'git config help.format'?	14:37
	if it's blank or 'man', change it to 'html'	14:37
stewbydoo	ouputs html	14:37
canton7	hrm	14:37
stewbydoo	ikke git help foo	14:37
	← dfgdfgdfg left	14:37
	→ Zenmaker_ joined	14:38
_ikke_	Well, then do what canton7 says	14:38
canton7	_ikke_, it's already html :S	14:38
faraway	_ikke_, canton7: so in the worst case i need to merge my current "feature" branches manually to the develop branch when they are finished. but the rest should work if i rename master to develop? (i'll try it anyway but i just wondered if there are known/common pitfalls)	14:38
canton7	stewbydoo, try setting web.browser?	14:38
_ikke_	faraway: Not many people here use the plugin	14:39
canton7	(might also need browser.<tool>.path\|cmd . I haven't played with those options before)	14:39
	faraway, nah the tool will merge into the branch called 'develop'. if the branch you want to merge them into is called 'develop', the tool will work fine	14:40
cmn	you shouldn't need to, my git for win installation works with just help.format	14:40
	→ torjus_ joined	14:40
stewbydoo	set web.browser firefox.exe?	14:40
	← hwrd\|work left	14:40
faraway	_ikke_, canton7: ok thx both of you	14:40
canton7	hang on. I need to update msysgit anyway, so i'll grab the latest and play	14:40
stewbydoo	alright	14:40
cmn	it will start the default browser... maybe you don't have one set?	14:41
stewbydoo	yea how do you set it?	14:41
	set web.browser firefox.exe	14:41
	?	14:41
	im very new to git	14:41
cmn	that'd be a git setting, I mean default	14:41
	← papegaaij left	14:42
cmn	the browsers generally ask you when you start them	14:42
canton7	i'm not sure what sort of path it'll want. /c/Program Files/ or C:\\Program Files\\	14:42
stewbydoo	right well my default is firefox	14:42
cmn	but try setting it, if it looks like it's not picking it up	14:42
	← faraway left	14:42
stewbydoo	okay	14:42
cmn	what Windows version are you using?	14:42
	→ gtrak joined	14:42
stewbydoo	win 7 pro	14:42
cmn	hm, ok, same here	14:43
canton7	stewbydoo, https://github.com/msysgit/msysgit/issues/40	14:43
cmn	thought maybe Win8 had messed it up	14:43
	← specialGuest left	14:43
canton7	hrm, Git-1.7.11-preview20120704 works fine for me out of the box	14:44
	→ pdelvo_ joined	14:44
canton7	which installer did you download stewbydoo?	14:45
stewbydoo	damn it	14:45
	umm one sec	14:45
	← kpreid left	14:45
	→ fluxdude joined	14:45
stewbydoo	Git-1.7.11-preview20120620	14:45
	→ kpreid joined	14:45
canton7	..which is the broken one. grab the latest one	14:45
	← oojacoboo left	14:45
canton7	http://code.google.com/p/msysgit/downloads/detail?name=Git-1.7.11-preview20120704.exe	14:46
jast	so, git-scm.com links to the broken one? nice :)	14:46
	someone might want to let scott know	14:46
stewbydoo	right...so the main site sucks	14:47
cmn	google code sucks harder	14:47
	← pandeiro_ left	14:47
cmn	it makes it very hard to figure out what's going on with the releases	14:47
	→ pandeiro joined	14:47
stewbydoo	well if google code has the working verson than it does not suck	14:47
	→ linusoleander joined	14:48
cmn	what?	14:48
stewbydoo	the main git-scm site links to the wrong install...might wanna change that	14:48
	← pdelvo left	14:48
cmn	right	14:48
	→ wolftankk joined	14:48
linusoleander	Git is asking for a password for host A when trying to push to host B	14:48
	git push origin master	14:48
	[email@hidden.address] password:	14:48
	Why is that?	14:48
jast	git remote -v to see what is configured	14:48
cmn	because that's how you log in to computers	14:49
linusoleander	91.123.195.252 isn't the host I'm trying to push to	14:49
thiago	linusoleander: because you're trying to push to A	14:49
	linusoleander: yes, it is	14:49
	← patrick85 left	14:49
	← wolftankk left	14:49
linusoleander	thiago: Not according to the .git/config file	14:49
	origin is set to host B	14:50
thiago	linusoleander: what branch are you on?	14:50
linusoleander	master	14:50
stewbydoo	well hopefully that does the trick	14:50
	thanks to all	14:50
thiago	linusoleander: what is master's upstream branch?	14:50
jast	doesn't matter	14:50
thiago	oh, right	14:50
	it's git push origin master	14:50
jast	what does 'git config remote.origin.url' say? what does 'git config remote.origin.pushurl' say?	14:50
thiago	linusoleander: git remote -v. What does it say?	14:50
jast	or, yeah, what I said before :)	14:51
linusoleander	Origin points (according to git remote -v) to host B	14:51
	← stewbydoo left	14:51
jast	both url and pushurl?	14:51
	does it contain an IP address or a hostname?	14:52
Silex	hi, to completely remove a folder from the repository, should I use tree-filter?	14:52
linusoleander	Nope	14:52
	← zeppo left	14:52
rudi_s	Hi. Is there support to authenticate HTTPS fingerprints like hg's hostfingerprints option?	14:52
canton7	stewbydoo, the problem is that git-scm, although "official" insomuch as it's recognised as the main git website, isn't maintained by git's maintainer, but rather by a contributer. so stuff isn't always in sync	14:52
_ikke_	Silex: Does it need to be removed from history?	14:52
jast	Silex: index-filter is much faster. there should be an example somewhere in the manpage.	14:52
linusoleander	jast: They are all pointing to github	14:52
	Which is the correct host	14:52
	But when I try to push to origin it uses my VPS for some reason	14:53
	→ kermit joined	14:53
jast	linusoleander: and ~/.ssh/config doesn't override the hostname for github.com?	14:53
	*the host	14:53
Silex	_ikke_: yes, it's for a friend he added his whole img/ folder	14:53
	→ zeppo joined	14:53
Silex	which is not needed in the repo	14:53
	→ BSaboia__ joined	14:53
	→ TheJH joined	14:53
	→ pppaul joined	14:53
_ikke_	Silex: https://help.github.com/articles/remove-sensitive-data	14:54
thiago	linusoleander: ssh -v <hostnamehere>. Where does it connect to?	14:54
lb	Silex: !rewriting_public_history	14:54
gitinfo	Silex: Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to `git pull --rebase` and even worse things if they have tagged or branched, so you must publish your humiliation so they know what to do. You will need to `git push -f` to force the push. The server may not allow this. See receive.denyNonFastForwards (git-config)	14:54
linusoleander	jast: It looks like this problem is my ssh config file	14:54
Silex	lb: I know.	14:54
lb	good	14:54
jast	... another problem pinpointed, yay ;)	14:54
Silex	_ikke_: the example is about one file only. Can I simply add -r and change the filename to a directory?	14:55
	→ guampa joined	14:55
_ikke_	yes	14:55
	infogulch → infogulch\|afk	14:56
	→ shreya joined	14:56
Silex	thanks	14:56
jast	'add -r'? since when do we have that?	14:56
linusoleander	jast: Changing the config file worked, thanks	14:56
	→ w41_ joined	14:57
jast	oh, never mind	14:57
	← BSaboia left	14:57
jast	linusoleander: you're welcome	14:57
	→ bodom joined	14:58
	→ shreya_ joined	14:58
	← shreya left	14:59
	← john_anderson left	14:59
	← w41 left	15:00
	→ tatsuya_o joined	15:00
	← und3f left	15:00
	→ martinjlowm joined	15:01
	→ wolftankk joined	15:01
	← Pitel_laptop left	15:01
	← stodan left	15:02
	→ trungqtran joined	15:02
	← w41_ left	15:02
	← w41__ left	15:02
	→ mritz joined	15:03
	← iocor left	15:03
	← dvide left	15:03
	← iamtakingiteasy left	15:04
	← crocket left	15:04
	→ legumbre joined	15:05
	→ dfgdfgdfg joined	15:05
	← Jay_Levitt left	15:05
	← wolftankk left	15:05
	→ Jay_Levitt joined	15:06
	→ crocket joined	15:06
	→ procrastubator joined	15:07
procrastubator	Guys how do I find the name of the branch that my repo usually pushes to?	15:07
	git branch only returns master	15:07
	which is the branch I am nin	15:07
cmn	usually pushes to? that depends on the configuration, and there is no "usually" there	15:08
jast	by default, 'push' updates all branches that exist on both sides	15:08
cmn	git branch returns the local branches	15:08
	→ axl_ joined	15:08
canton7	git branch -r returns the remote branches, and git branch -a shows everything	15:09
procrastubator	I just added it again and linked it with the URL	15:09
	→ erdnaxeli joined	15:10
	← erdnaxeli left	15:10
canton7	"linked it with the URL"?	15:10
dendazen	Why do i get this error when doing git clone https://u:p@whatever	15:10
	write(2, "fatal: I don't handle protocol '"..., 42fatal: I don't handle protocol 'https'	15:10
	) = 42	15:10
	← zeppo left	15:11
cmn	because you build git or curl without https support	15:11
	built*	15:11
	→ nwest joined	15:11
	→ SirCmpwn joined	15:12
	← MaximLacrima left	15:12
	← rchavik left	15:13
	→ ph^ joined	15:13
	← tjegels left	15:14
	← cbreak-work left	15:14
	→ dvide joined	15:14
	← cagedwisdom left	15:15
	← Destos left	15:15
	→ specialGuest joined	15:16
	→ Destos joined	15:17
	→ someprimetime joined	15:17
	→ t4nkd joined	15:18
	→ jjbohn joined	15:18
	→ brainbox1100 joined	15:18
	← linusoleander left	15:19
procrastubator	canton: git remote add <branch> < .git URL>	15:21
	← flijten left	15:21
	→ banghouse joined	15:22
	← osxorgate left	15:22
	← skorgon left	15:22
	→ wolftankk joined	15:23
	Werring → W\|Food	15:23
	← goombas left	15:24
	← wolftankk left	15:24
	→ osxorgate joined	15:24
	← ThomasLocke left	15:24
	→ tjholowaychuk joined	15:25
	← mritz left	15:26
	← Merwin left	15:26
	→ Merwin joined	15:26
	→ toabctl joined	15:28
dendazen	cmn	15:28
	but it worked before.	15:28
	→ adityamenon joined	15:28
dendazen	what should i do?	15:29
	→ kenperkins joined	15:29
	→ blorbx joined	15:29
dendazen	re-install git?	15:29
	Or there are settings?	15:29
	← carbolymer left	15:29
cmn	before what?	15:30
	← wiremaster left	15:30
	→ _iron joined	15:30
	Merwin → ThibautD	15:31
	← amcsi_work left	15:32
	→ gigamonkey joined	15:32
	→ wiremaster joined	15:32
dendazen	It worked before i updated my workstation from F16 to F17.	15:33
	→ anabelle joined	15:33
	→ etcetera joined	15:34
	← dfgdfgdfg left	15:34
canton7	looks like a problem with the git package in the F17 repos? (I'm assuming F16 and F17 have different repos)	15:34
cmn	then make sure you have a curl version that understands https	15:35
	← mklappstuhl left	15:35
	← etcetera left	15:35
	← adityamenon left	15:35
	→ etcetera joined	15:36
dendazen	I do.	15:37
cmn	then it's your git installation that doesn't know this	15:38
dendazen	Then the thing is that i have build machines in my company and if i do the same thing under my user from those build machines. i get the same error, and those are F16	15:38
	Before i was able to do that.	15:38
canton7	what is the exact error, character for character?	15:38
dendazen	So i log on to those machines.	15:38
canton7	what you pasted before was somewhat distorted, but there was some space between 'h' and 'ttps'. I'm suspecting some weird character	15:40
	← Tegi left	15:40
	← osxorgate left	15:40
	← Soliton left	15:40
	→ mritz joined	15:40
dendazen	(sshing) with my ldap credentials, do kinit with mine credentials and do: git clone https://u:p@source.blenders.com/ep-compliance-ntp.pkg.git	15:40
	← irqq left	15:40
	→ fxn joined	15:41
dendazen	and it would clone before. and get this: https://u:p@source.eladian.com/ep-compliance-ntp.pkg	15:41
	Cloning into 'ep-compliance-ntp.pkg'...	15:41
	fatal: I don't handle protocol 'https'	15:41
canton7	complete with that space in the middle of the https?	15:42
	→ harshpb joined	15:42
dendazen	what do you mean?	15:42
cmn	I can't see any spaces...	15:42
dendazen	with the space in the middle?	15:43
	← bodom left	15:43
cmn	are you copy-pasting it or typing it out?	15:43
canton7	ooh this is fun. xchat is gettin very confused, and moving the 'h' around when I highlight the https	15:43
	canton7 suspects weird unicode/unprintable stuff even more	15:43
fxn	hey, I have a repo I cannot rename and would like to collapse all the history into one single commit and force push, is git rebase -i the way to go?	15:43
	→ irqq joined	15:43
	→ Soliton joined	15:44
canton7	fxn, probs easier to nuke the .git repo and start over. that or create a new orphan branch	15:44
fxn	canton7: so I delete .git, git init etc and forche push	15:45
dendazen	And the thing is that other people on my group can clone with no problems.	15:45
cmn	are you copy-pasting it or typing it out?	15:45
dendazen	Can it be related to my kerberos ticket?	15:45
	← JohnBat26 left	15:45
canton7	fxn, that approach will nuke all config actually. I'd prefer 'git checkout --orphan newbranch && git commit -m "New commit" && git branch -D original_branch && git branch -m original_branch && git push -f"	15:46
cmn	canton7: can you see anything odd in the web log?	15:46
	dendazen: are you copy-pasting it or typing it out?	15:46
dendazen	As my keytab has been changed, since i've gotten a new F17 OS	15:46
	i am copying it out from bash	15:46
canton7	cmn, nah the weblog is fine	15:46
cmn	then type it out	15:46
	the likely cause right now is some unvisible space, so don't copy it	15:47
	← rwestphal left	15:47
cmn	invisible, even	15:47
	← unnamedwill left	15:47
thorbjornDX	cmn: non-uninvisible, perhaps?	15:48
	← procrastubator left	15:48
	← specialGuest left	15:48
	→ aLeSD joined	15:48
	← gerryvdm_mbp left	15:48
aLeSD	hi all	15:48
cmn	that'll be why it's confusing git	15:48
	→ dzonder joined	15:48
aLeSD	is it possible to setup a remote repository without ssh ? I mean open and without password ?	15:48
	← _Qman left	15:49
	← monokrome left	15:49
	← Kingsy left	15:49
cmn	use the raw git transport	15:49
	→ Milossh joined	15:49
dendazen	git clone https://u:p@source.blenders.com/ep-compliance-ntp.pkg	15:50
	← Vert left	15:50
dendazen	cmn, it worked	15:50
	how the space can be hidden?	15:50
cmn	because it's un unprintable character	15:50
canton7	cmn, dendazen, there's the byte sequence E2 80 8B between the ' and h	15:50
	→ rwestphal joined	15:50
	← thiago left	15:50
	← lb left	15:50
dendazen	yeah it worked, when i just typed by hands.	15:50
	→ Desproges joined	15:51
dendazen	Thanks, weird.	15:51
bigmeow	anybody here use both hg and git	15:51
	?	15:51
	W\|Food → Werring	15:51
	→ unnamedwill joined	15:51
bigmeow	how to use git to clone hg repo?	15:51
canton7	dendazen, and that sequence is a utf-8 zero width space	15:52
cmn	you use hg-git	15:52
	← t4nkd left	15:52
	← hyperair left	15:52
	← resmo left	15:52
	← Desproges left	15:52
dendazen	yeah.	15:52
	← jlg13 left	15:53
	→ argoneus joined	15:53
dendazen	the thing is i used to copy that git clone thing from our trac wiki	15:53
	← tatsuya_o left	15:53
cmn	right, and as soon as the space was mentioned, you should have typed it out	15:53
	→ specialGuest joined	15:53
argoneus	is it possible to 'move' a git repo from one folder to another?	15:53
dendazen	when i used firefox and would copy it str8 to the bash, it worked	15:53
	← shreya_ left	15:54
dendazen	but turned out did not work when i used chrome.	15:54
	as a browser	15:54
	→ gregm_ joined	15:54
bigmeow	how to use git to clone hg repo?	15:54
	→ tatsuya_o joined	15:55
	← angelsl left	15:55
	The24z_away → The24z	15:55
cmn	bigmeow: you use hg-git	15:55
	→ angelsl joined	15:55
bigmeow	cmn: dude, it should be git-hg	15:56
	← tatsuya_o left	15:56
	→ tatsuya_o joined	15:56
cmn	if you know the tool to use, then use that instead of asking us	15:56
bigmeow	cmn: but this maybe error prone:(	15:56
cmn	what is?	15:56
	→ pdelvo joined	15:56
argoneus	is it possible to 'transfer' a git repo to another directory and add files from there?	15:57
	because I can't move the files to the folder	15:57
	only the other way around	15:58
cmn	you can clone	15:58
bigmeow	what is the hg command corresponding to git clean?	15:58
cmn	or init, fetch etc if you already have files there	15:58
	bigmeow: you should ask in the mercurial channel	15:59
	← brainbox1100 left	15:59
argoneus	cmn: as in	15:59
	clone the repo from the other dir	15:59
	rm the unneeded stuff	15:59
	and commit push?	15:59
	← lausi left	16:00
	→ lausi joined	16:00
	← fxn left	16:00
	← pdelvo_ left	16:00
	← lausi left	16:00
cmn	why do you need a new clone just to delete some stuff?	16:00
argoneus	I need to replace some of the files with others	16:00
	← variousnefarious left	16:00
argoneus	and the others	16:00
	are in a different dir	16:00
	from which I can't move them	16:01
cmn	then read what I said above	16:01
argoneus	yeah I'll do that	16:01
	thank you	16:01
	← argoneus left	16:02
	→ ISF joined	16:02
	→ variousnefarious joined	16:03
	→ boombatower joined	16:03
	← gusnan left	16:05
	→ shwaiil joined	16:05
shwaiil	hi	16:05
	The24z → The24z_away	16:06
	← TheJH left	16:06
	← leokaito left	16:06
adhawkins	What happened to the bot?	16:06
	!ask	16:06
gitinfo	Yes, it's okay to ask questions here.... in fact, you just asked one! ;-) Pretty much any question is fine. We're not terribly picky, but we might be asleep. Please be patient and you should get an answer soon.	16:06
	→ Protux joined	16:06
	← senny left	16:07
	→ iocor joined	16:07
shwaiil	I'm finding more and more people using GIT to do deployment. Because I also use it as well to do deployment and because I've been trying to learn Capistrano etc (but didn't had time), I'm having a little bit of dificulties finding if I really should stop using GIT to deploy or not, because seems that everyones doing it. Any comments on that ? Thanks!	16:07
adhawkins	!deploy	16:07
gitinfo	Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html	16:07
shwaiil	- I know GIT is not ment for deployment	16:07
	great..	16:07
cmn	ti's also not called GIT	16:07
adhawkins	I don't know much about using it for deployment, but presume that URL contains useful advice	16:07
	→ andrewstewart joined	16:08
cmn	shwaiil: basically, you can get away with it if you have very minimal needs, but anything other than putting a bunch of files somewhere needs a real deployment	16:08
shwaiil	cmn: tks for looking. I know it's not for deployment, that's the reason I've been trying to find the right tool for it	16:09
	I found capistrano, but, let's say I may just create a project using JS	16:09
	sometimes PHP	16:09
	sometimes other	16:09
	Is Capistrano good for me ?	16:09
cmn	the right tool usually depends more on the framework you're using than the VCS	16:09
shwaiil	I just don't have time and I'm lacking energy to learn a deployment tool and I just want to make the right choice	16:10
	→ oojacoboo joined	16:10
csmrfx	rake	16:10
shwaiil	cmn: I'm sorry about asking here, but I really can't find a better place	16:10
csmrfx	or what else is there - ant and its descendant	16:10
	and so on.	16:10
shwaiil	cmn: isn't there a deployment tools that let me deploy anything ?! :T	16:10
csmrfx	I'd say it depends on the language and also the frameworks	16:10
cmn	shwaiil: yes, but those involve a lot of writing stuff out	16:11
csmrfx	I'm the kinda person that would write build scripts in the same language as deployment or app, but thats not always possible	16:11
shwaiil	cmn: what do u use ?	16:11
cmn	AFAIK you can use capistrano for anything, but you'll end up writing a bunch of scripts	16:11
	I don't do deployments myself; if I need to test something out, I outsource that to heroku or openshift	16:12
	→ hwrd\|work joined	16:12
	→ diegoviola joined	16:12
csmrfx	hehe, deploy app or deploy the whole service? anyway, pardon my off topic	16:12
	← robbyoconnor left	16:13
	→ fly9_ joined	16:13
rudi_s	Hi. Is there support to authenticate HTTPS fingerprints like hg's hostfingerprints option?	16:16
	← fly9 left	16:16
	fly9_ → fly9	16:16
	adhawkins → adhawkins-away	16:16
	← jjbohn left	16:16
	← chimay left	16:17
cmn	rudi_s: git uses curl	16:20
	→ ncd\|leeN joined	16:21
	→ chimay joined	16:21
	← chimay left	16:21
	→ chimay joined	16:21
rudi_s	cmn: Thanks. But looks like curl can't do it.	16:24
cmn	then that's the answer	16:24
rudi_s	:-/ Ok.	16:24
	← leeN left	16:24
bremner	ssh is your friend.	16:24
rudi_s	Can I pass options to curl?	16:25
	→ gtuckerkellogg joined	16:25
rudi_s	bremner: Only helps if I have SSH access.	16:26
	→ maletor joined	16:26
	← youngblood left	16:26
	← ping-pong left	16:28
	→ wolftankk joined	16:28
	← wolftankk left	16:28
	→ thiago joined	16:28
	→ fornext joined	16:29
	← giallu left	16:29
	→ vmassuchetto joined	16:30
vmassuchetto	I have a global ~/.gitconfig file with my user name and email, but I can't seem to override this on a project's .git/config. Is it possible?	16:31
cmn	yes, that's how you do it	16:31
	→ dfgdfgdfg joined	16:32
	→ Shusshu joined	16:32
	← Shusshu left	16:32
	← achadwick left	16:32
	← pdelvo left	16:33
	→ pdelvo joined	16:33
vmassuchetto	cmn, Well, it's not working. I get commits with the ~/.gitconfig user.email as author	16:33
cmn	get commits meaning what?	16:33
	what does git config user.email say?	16:33
vmassuchetto	git config user.email get the project's .git/config user.email	16:34
	→ dreiss joined	16:34
	← LimeBlast left	16:34
	→ cehteh joined	16:34
vmassuchetto	'git config --list' will show both user.email options, with project's one at the bottom.	16:35
	→ pandeiro_ joined	16:35
	← pandeiro left	16:35
	→ Turicas joined	16:35
cmn	yes	16:35
	then it's set up	16:35
	← klj613 left	16:36
	← hwrd\|work left	16:36
	Drarok → drarok	16:37
	→ wolftankk joined	16:37
	← dreiss left	16:39
vmassuchetto	cmn, it just came to me that, a long long time ago I've put GIT_AUTHOR_NAME\|EMAIL in my ~/.bashrc	16:39
	cmn, thanks anyway =D	16:39
milki	lol	16:39
	← vmassuchetto left	16:39
cmn	there you go, then	16:39
	← wolftankk left	16:41
	→ ping-pong joined	16:41
	← QaDeS left	16:41
	← ISF left	16:41
	→ flo` joined	16:46
	← harshpb left	16:46
flo`	hi	16:46
gitinfo	flo`: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.	16:46
	→ brainbox1100 joined	16:46
	→ harshpb joined	16:47
	→ cakehero joined	16:47
flo`	i have a git-repo created with git-svn. however, at some point, the master branch ends (there are just no new commits) and a 'trunk' branch has been created.	16:47
	→ hyperair joined	16:48
	← harshpb left	16:48
flo`	how can i put that 'trunk' branch into master, deleting trunk and having master contain it all? my repo is not shared with anyone, so i can cleanly rewrite history	16:48
	→ harshpb joined	16:49
	← fly9 left	16:49
	→ fly9 joined	16:49
milki	flo`: merge trunk to master?	16:50
	→ tomislater joined	16:50
	→ Sym3tri joined	16:50
	← nightcrawler786 left	16:50
flo`	milki: can i then delete trunk without losing any history?	16:51
	← gdmfsob left	16:51
	← dr_lepper left	16:51
milki	after youve merged, yes	16:51
flo`	and how can i make new stuff which gets svn-committed to svn_'s trunk (by other developer) be put into master instead into git's trunk-branch/	16:52
	← angelsl left	16:52
flo`	*?	16:52
esc	flo`: git merge trunk?	16:52
	← tatsuya_o left	16:52
milki	magic	16:52
	im sure theres some branch mapping thing somewhere...	16:53
	i dont know much about git svn	16:53
	→ flavius joined	16:53
	upasna → sweet_kid	16:53
flo`	esc: that went fine so far. but will that also make git-svn store stuff coming from svn's trunk into git's master?	16:53
milki	what does git branch -v say for trunk?	16:53
esc	flo`: if you need to still retrieve stuff from svn, don't delete the trunk	16:54
	← harshpb left	16:54
esc	flo`: usually, that isn't the case	16:54
flo`	esc: but i want to delete the trunk, and have git-svn put it all into master.	16:54
	and i do want to still retrieve from svn	16:54
	(unfortunately)	16:55
esc	the workflow is to git svn fetch	16:55
milki	flo`: what does git branch -v say for trunk?	16:55
esc	the master is a local branch	16:55
	→ nightcrawler786 joined	16:55
cmn	you can give git-svn a mapping, but making it overwrite your local branches is not a good idea	16:55
esc	trunk tracks svn	16:55
milki	make master track svn then?	16:56
flo`	when on branch master: ..., trunk 1e76122 nothing	16:56
esc	i don't well, it's not quite the same	16:56
	trunk is a special branch	16:56
milki	o	16:56
flo`	when i then do git checkout trunk, "warning: refname 'trunk' is ambigous", git branch -v tells me the same as before then.	16:56
esc	kind of like remote-tracking branches, e.g. origin/master	16:56
	← iocor left	16:56
flo`	esc: but trunk got created some day by git-svn.	16:57
milki	aah ok, so git-svn stuff	16:57
flo`	for a whole bunch of revisions, it correctly put it into master	16:57
	but some day, it started to put it into trunk. and that's what i want to revert.	16:57
	→ kerrick joined	16:58
esc	well, usually, what happens, is the first time you git svn clone, trunk and master point to the same commit	16:58
flo`	btw, the gitg frontend displays as first entry in its list: (master) (trunk) (trunk) nothing	16:58
	why does it display trunk twice?	16:59
	→ oakdog8 joined	16:59
esc	because you have two trunk branches, apparently	16:59
	→ Fullmoon joined	16:59
	← AlbireoX left	16:59
	→ AlbireoX`Laptop joined	16:59
	← Sym3tri left	16:59
esc	that's also the bit about being ambigouus you posted earlier	16:59
flo`	esc: i see. git checkout <TAB> <TAB> displays a list, containing heads/trunk and remotes/trunk	16:59
esc	yes	17:00
	remote branch and the svn-tracking-branch (for lack of a better word)	17:00
	← Vile left	17:00
esc	eh, local branch, i meant	17:00
	heads/trunk is your local branch, i suppose	17:00
	and remotes/trunk is the one that tracks the trunk from svn	17:00
	→ Advocation joined	17:01
flo`	okay, so i somehow need to tell git that i want remotes/trunk to write its stuff to master, and not to heads/trunk?	17:01
	is that true?	17:02
esc	donsn't sound right	17:02
	i can only tell you, what my workflow is	17:02
	banghouse → banghouseAFK	17:02
	← irqq left	17:03
	→ argoneus joined	17:03
argoneus	does each branch have its own 'remote'?	17:03
esc	flo`: i would recommend to delete the heads/trunk	17:04
flo`	nope, just trunk	17:04
esc	flo`: and get master to track trunk	17:04
	→ devth joined	17:04
flo`	there are also strange branches: branchname and branchname@1234	17:05
	(1234 is apparently a svn revision number)	17:05
esc	git branch master --set-upstream trunk	17:05
	yeah, its from when branches were moved around in the svn repository	17:05
	← RaceCondition left	17:05
flo`	esc: so these are two different branches which share the same name, e.g. by creating, deleting and re-creating a svn branch?	17:06
	← shruggar left	17:06
esc	something like that	17:06
	→ RaceCondition joined	17:07
esc	but i find it comes from layout changes in the repository and other unorthodox nonsense people do with svn repsoitories	17:07
	← argoneus left	17:07
esc	i have seen commits similar to those you mentioned, but i could never work out exactly what caused them	17:07
flo`	esc: unorthodox nonsense has happened to the repo.	17:08
esc	yes, this is usually the case, amazingly enough...	17:08
	flo`: http://andy.delcambre.com/2008/03/04/git-svn-workflow.html	17:08
	here is a good workflow	17:08
flo`	after git branch --set-upstream remotes/trunk, git branch -v says master 1e... [remotes/trunk] nothing	17:08
	which is good i suppose?	17:08
	→ BSaboia joined	17:09
	→ aruiz joined	17:09
esc	it's git branch master --set-upstream remotes/trunk	17:10
	→ mwrosen joined	17:10
esc	basically this allows you to 'git pull' after doing 'git svn fetch'	17:10
	to merge the changes from remotes/trunk into master	17:10
	which kind of mimics a git-workflow	17:10
flo`	git svn rebase does both things for me, right?	17:10
	← neurodrone left	17:11
esc	no, git-svn rebase, does a git svn fetch and then rebases master onto remotes/trunk	17:11
	→ opus joined	17:11
esc	it's mainly used, when someone else has commited to svn, and you need to put the stuff you have in your local master on top of that	17:11
flo`	ah right.	17:12
esc	you get the latest changes with git svn fetch	17:12
	and then you rebase you master onto remotes/trunk	17:12
	← BSaboia__ left	17:12
esc	then you would normally, git svn dcommit to commit your local changes to the remote svn	17:12
	basically, as a rule of thumb, your need to linearise your history, before dcommitting	17:13
	← PigeonFriend left	17:13
esc	otherwise git svn becomes confused	17:13
	anyway, i need to leave now	17:13
	← xplat left	17:14
	→ goshawk joined	17:14
esc	flo`: good luck with git svn, if you read a little bit more about it, things should clear up	17:14
flo`	thank you	17:14
	uhm, linearize means what?	17:14
	→ lacrymology joined	17:15
	→ harshpb joined	17:15
	→ patrick85 joined	17:19
	← harshpb left	17:21
	→ cklein joined	17:21
	→ k0nichiwa joined	17:21
k0nichiwa	what is the quickest (smallest bandwidth) way i could get the source code for version 1.9.0 of the phonegap repository ?	17:22
	← cklein left	17:22
	→ iocor joined	17:22
k0nichiwa	https://github.com/phonegap/phonegap	17:22
	i'd rather not clone the whole repository	17:22
FauxFaux	k0nichiwa: a) #github. b) Hit the big "zip" button.	17:22
	→ apok joined	17:23
k0nichiwa	zip would get me 2.0	17:23
	i want an older version, 1.9	17:23
FauxFaux	So you want to get the source of a /tag/? I wonder what could possibly show you zips of tags.	17:23
k0nichiwa	i want source for 1.9	17:24
	i didn't say a tag	17:24
FauxFaux	1.9 is a tag.	17:24
k0nichiwa	tag is some stupid git concept, i don't give a fck ab out it	17:24
	i want the source for version 1.9	17:24
muep	there is no other simple way than tags to specify an user-visible version of source stored in git	17:24
canton7	k0nichiwa, stop being rude, click the tags button, then click on '1.9.0.zip'	17:24
FauxFaux	You're trying to use github to work around a vendor problem without talking to the vendor. We're trying to support git, which is used by github, but github actually has nothing to do with git..	17:25
	← koltroll left	17:25
k0nichiwa	thanks i think that will work	17:26
esc	flo`: linearise, means to adjust the directed acyclic graph, that is the git history, into a graph where each node has only a single parent	17:29
	no bifurcations or joins	17:29
	← specialGuest left	17:30
	← shwaiil left	17:32
	← jceb left	17:32
flo`	esc: ok, and how would i do this?	17:32
	→ paul_andrew joined	17:33
esc	well, by using git svn rebase for example	17:33
cbreak	you should use git rebase...	17:33
	git svn rebase is much more limitted	17:34
	← fornext left	17:34
	→ tigris_ joined	17:34
EugeneKay	git svn screwitallup	17:35
	← someprimetime left	17:35
	→ eli joined	17:36
	→ sfrank joined	17:36
	→ thirdknife joined	17:36
esc	EugeneKay: +1	17:36
tigris_	hi all... i have a script that creates and writes directly to a bare repo... i'm seeing a really weird issue where all the commits are visible in the log, but when i clone the repo, not all the files are there	17:36
cbreak	files?	17:37
	← jh3 left	17:37
cbreak	why would there be any files?	17:37
EugeneKay	Are you updating a ref to point to your new commits?	17:37
cbreak	does the last commit contain them?	17:37
	(the HEAD commit)	17:37
tigris_	cbreak: when i clone the bare repo into a normal one i mean, only the latest file is there	17:37
esc	tigris_: did you forget to check the files in?	17:37
	→ duckxx joined	17:38
tigris_	https://gist.github.com/3623894	17:38
	→ harshpb joined	17:38
	→ jh3 joined	17:38
	→ shruggar joined	17:39
cbreak	tigris_: that's normal	17:39
	only the latest file is supposed to be there	17:39
	a commit is a snapshot after all	17:39
tigris_	cbreak: i'm doing something wrong with :parents aren't i	17:39
cbreak	do a git log --stat	17:39
	pastebin the result if you want us to interprete it :)	17:40
tigris_	cbreak: oh i think i see what you mean... when i create each commit, that tree should contain all existing blobs, even unchanged ones	17:40
cbreak	unless you want them to be deleted	17:40
	-> snapshot	17:40
	you're better off using git's own facilities for this	17:40
	i.e. a non-bare repo with index for staging future commits	17:40
	← fluxdude left	17:40
	← Ivoz left	17:41
	← etcetera left	17:41
tigris_	cbreak: i was doing that originally, but this code can create 500,000+ repos in about 10 minutes, shelling out to git was taking about 3 hours	17:41
	→ mabrand joined	17:42
	→ SegFaultAX\|work2 joined	17:42
	→ sebastorama joined	17:42
cbreak	you know that you can copy existing repositories, right?	17:42
	a.k.a. clone	17:42
bremner	that's a lotta repos	17:42
	→ etcetera joined	17:42
	← gun-lpt left	17:43
	→ diptherial joined	17:43
	→ argoneus joined	17:43
argoneus	hello	17:43
	what is HEAD^^?	17:43
	HEAD - 2 commits?	17:43
	← sebastorama left	17:43
	→ gun-lpt joined	17:43
cbreak	parent of parent of HEAD	17:43
	(first parent each time)	17:43
	→ sebastorama joined	17:44
tigris_	cbreak: you are correct, git log --stat shows previous files getting deleted, thanks!	17:44
cbreak	np.	17:44
	← sebastorama left	17:44
flo`	`.	17:44
	`.	17:44
	~	17:45
argoneus	parent of HEAD is the last commit before HEAD?	17:45
	e.g.	17:45
	if I do git rebase -i HEAD^	17:45
	I revert 1 commit?	17:45
flo`	meh, sorry >_<	17:45
	→ stratwine joined	17:45
	← stratwine left	17:45
tigris_	i'm moving a cms revision based system from mysql to git... currently 525,000 customers, total of 2.3 million files in total, and about 6 million revisions over all... i am going with 1 repo per client, hence creating new ones and not cloning	17:46
	→ Akufen joined	17:46
	← Akufen left	17:46
	→ Akufen joined	17:46
	→ sebastorama joined	17:46
	← noisebleed left	17:46
	← jgoss left	17:46
	← sanmarcos left	17:46
	→ sanmarcos joined	17:46
	← harshpb left	17:46
	← spaceonline left	17:46
cbreak	argoneus: rebase -i doesn't revert anything	17:46
	→ foraks joined	17:47
	→ harshpb joined	17:47
	→ spaceonline joined	17:47
argoneus	if I accidently commited bad or accidently merged the wrong way	17:47
tigris_	and also why you see that script deleting any tiny amount of space per repo, even the 200 byte config file, since any small amount becomes a big amount when multiplied by 525,000	17:47
argoneus	what do I do?	17:47
cbreak	HEAD^ refers to the parent of HEAD, to the commit that was the newest commit in the branch before the newest after that was made	17:47
	argoneus: you can reset it away if you haven't pushed yet	17:47
	git reset HEAD~1	17:47
argoneus	~1?	17:47
	or HEAD^	17:47
cbreak	git reset HEAD^	17:47
	same thing	17:47
argoneus	and if I pushed?	17:48
cbreak	argoneus: man git-rev-parse	17:48
gitinfo	argoneus: the git-rev-parse manpage is available at http://git-scm.com/docs/git-rev-parse.html	17:48
	← gun-lpt left	17:48
cbreak	then you can use git revert	17:48
argoneus	thank you	17:48
cbreak	this will make a new commit undoing the past's effects	17:48
	but with merges, that has all kinds of associated troubles	17:48
	→ vervic joined	17:48
argoneus	how so?	17:48
	← shruggar left	17:48
	← Advocation left	17:49
	→ TheJH joined	17:49
sfrank	hi all - I am trying to rebase my master branch into my feature branch, and git has auto-merged changes by excluding mine on a file that was moved to another folder in master - any ideas why it has ignored my changes?	17:50
	← diptherial left	17:50
cbreak	argoneus: if you merged something then that's part of history	17:50
	reverting the merge won't change that	17:50
	→ srayner joined	17:50
cbreak	so you can never again merge it	17:50
	and of course it will show up in git log	17:50
argoneus	so if I have a 'stable' and 'dev' branch	17:51
cbreak	sfrank: bad idea	17:51
argoneus	and I accidentally merge them the wrong way	17:51
sfrank	oh?	17:51
argoneus	then it's game over?	17:51
cbreak	do not rebase master into anything	17:51
	← nwest left	17:51
milki	ever	17:51
sfrank	oh i meant i was rebasing the changes from master into my branch	17:51
	and i squashed all the commits into my branch into one	17:51
cbreak	that's usually not a good idea	17:51
sfrank	so basically just trying to get caught up to master	17:51
	→ Sigma joined	17:51
cbreak	then rebase your own branch onto master	17:52
	that's the way to get up-to-date to master	17:52
	if you rebase master, then you CHANGE MASTER HISTORY	17:52
	-> bad	17:52
	argoneus: merge the wrong way?	17:52
sfrank	oh i see	17:52
cbreak	you can rewrite history if you want, with reset or so	17:52
argoneus	cbreak: as in merge stable into dev and push it	17:52
cbreak	but it's tricky	17:53
	← pandeiro_ left	17:53
	← rmr left	17:53
	→ pandeiro joined	17:53
	← Jay_Levitt left	17:54
cbreak	sfrank: basically, if you rebase master on your feature, then you make git change master so that it looks as if it was based on your feature. :)	17:54
sfrank	ah - so what is the best way to get my branch caught up to master so that I can put in a pull request now that i'm done with my feature	17:55
	← aruiz left	17:55
	→ shruggar joined	17:55
EugeneKay	man git-merge	17:56
gitinfo	the git-merge manpage is available at http://git-scm.com/docs/git-merge.html	17:56
EugeneKay	git supports creating merge commits; use them. Having two parents is not heresy, it is a biological mandate.	17:56
argoneus	cbreak: how do you mean tricky?	17:57
	← axl_ left	17:57
	→ dr_lepper joined	17:58
cbreak	sfrank: git checkout yourbranch	17:58
	git rebase master	17:58
	argoneus: it's public history rewriting	17:58
	you have to do it on every single repository in existance	17:58
	so yours, the one you pushed to	17:58
	→ gun-lpt joined	17:58
cbreak	and everyone who fetched	17:58
argoneus	if I do it 3 minutes later	17:59
	no one should be messed up, no?	17:59
cbreak	depends	17:59
	you have to force push for once	17:59
	and then you have to check everyone elses repository	18:00
	if you work alone then that's easy	18:00
argoneus	yeah I do	18:00
cbreak	if you work in a team with dozens of people	18:00
	then you have some work to do	18:00
argoneus	so tl;dr	18:00
	check before I push	18:00
sfrank	cbreak: I attempted to do that.. however it decided to leave out some of my changes - which is the original problem i was having. i can manually go back through and resolve the issues by pulling back in the changes of mine that it discarded but i fear its error prone	18:00
argoneus	to avoid headaches	18:00
	→ Jay_Levitt joined	18:00
cbreak	sfrank: what?!	18:00
	banghouseAFK → banghouse	18:01
cbreak	you didn't use -i, did you?	18:01
sfrank	no just "git checkout mybranch" then "git rebase master"	18:01
	and it found one conflict… automagically resolved most of the differences by accepting my changes but oddly discarded my changes on a file that had moved	18:02
cbreak	hmm...	18:02
	it didn't conflict?	18:02
	← opus left	18:03
sfrank	so it found one conflict.. somebody had added a new function to that file… i had to manually resolve that one… but it discarded the other changes i made to the file without a conflict on those line numbers	18:03
cbreak	then you did something wrong.	18:04
	← Turicas left	18:04
cbreak	when you manually resolved the conflict	18:04
	→ specialGuest joined	18:04
cbreak	try again :)	18:04
	← paul_andrew left	18:04
	← patrick85 left	18:05
cbreak	sfrank: if renames make problems, check out man git-rebase	18:05
gitinfo	sfrank: the git-rebase manpage is available at http://git-scm.com/docs/git-rebase.html	18:05
cbreak	look for -m and rebase-threshold	18:06
	which is an -X option	18:06
	→ tehgeekmeister joined	18:06
tehgeekmeister	can a .gitconfig source another .gitconfig?	18:06
	i have stuff i want to share between machines, and stuff i don't	18:07
cbreak	I think that's a new feature	18:07
cmn	from 1.7.10 onwards it ca	18:07
	→ opus joined	18:07
	← vervic left	18:07
cmn	can	18:07
sfrank	thanks - i'll try that right now	18:07
argoneus	alright thanks for help	18:07
	← argoneus left	18:07
	→ rcu joined	18:07
tehgeekmeister	hmm, i can afford to upgrade	18:08
	→ neurodrone joined	18:08
tehgeekmeister	is it just like you'd do in bash?	18:08
	→ axl_ joined	18:08
cbreak	I doubt it	18:08
	→ haYabuSa joined	18:09
cmn	not even a little	18:09
tehgeekmeister	hmm. is there a better way to override just the name/email bits on a per machine basis?	18:10
	← haYabuSa left	18:10
tehgeekmeister	that's the only thing i don't want to share between all boxes.	18:10
	← srayner left	18:10
	→ Advocation joined	18:11
	→ paul_andrew joined	18:11
	← tomislater left	18:11
	← etcetera left	18:11
cmn	use the system config	18:11
tehgeekmeister	oh, cool	18:11
	didn't know that existed	18:11
	thanks	18:11
	→ mrOrangeChuck joined	18:11
	← thirdknife left	18:12
	← paul_andrew left	18:12
	→ paul_andrew joined	18:13
	infogulch\|afk → infogulch	18:14
	← Advocation left	18:15
	← paul_andrew left	18:15
	→ etcetera joined	18:16
	→ Advocation joined	18:17
	← kerrick left	18:17
	→ pantsman joined	18:17
	← pantsman left	18:17
	→ pantsman joined	18:17
	→ kerrick_ joined	18:17
	→ caseymcg joined	18:18
	→ paul_andrew joined	18:18
	← fread2281 left	18:19
	← paul_andrew left	18:19
	→ zeppo joined	18:19
	← dfgdfgdfg left	18:19
	← _jmg left	18:21
	← harshpb left	18:21
	→ harshpb joined	18:22
	← zeppo left	18:22
	→ zeppo joined	18:23
	→ imachuchu joined	18:23
	← mabrand left	18:23
	→ Austin__ joined	18:23
	→ paul_andrew joined	18:23
	← bartek left	18:25
	← berserkr left	18:25
	← paul_andrew left	18:25
	← sirdancealot left	18:25
	TheFuzzb_ → TheFuzzball	18:25
	→ iamtakingiteasy joined	18:26
	→ ShadeTornado joined	18:26
	← sfrank left	18:27
	→ koltroll joined	18:27
	← pandeiro left	18:27
	→ sfrank joined	18:27
	→ irqq joined	18:28
	→ paul_andrew joined	18:28
	← paul_andrew left	18:29
	← shruggar left	18:31
	→ converge joined	18:31
	→ bartek joined	18:32
	→ apipkin joined	18:32
	→ doherty joined	18:32
	→ sirdancealot joined	18:32
apipkin	i have an upstream repository. is it possible to set the default `pull` to a specific branch? if yes, how can i see what that branch is	18:32
	← etcetera left	18:33
cbreak	no	18:33
	← specialGuest left	18:33
cbreak	pull pulls to the current branch	18:33
	no exceptions.	18:33
thiago	pull pulls the current branch's upstream	18:33
	so just change the upstream	18:33
_ikke_	It needs a working tree to merge	18:33
thiago	oh, pull to a branch?	18:33
	it always merges with the current HEAD	18:34
	← Advocation left	18:34
cbreak	did you mean source or destination?	18:34
apipkin	so something like `git reset upstream/master —hard` would always be the master branch of upstream?	18:34
	→ gigamonkey_ joined	18:34
_ikke_	yes	18:34
thiago	upstream/master is just one branch	18:34
cbreak	if you have an upstream remote	18:34
_ikke_	Since you're last fetch anyway	18:34
	→ fread2281 joined	18:35
thiago	if that's refs/remotes/upstream/master, it's usually the upstream remote's master branch	18:35
apipkin	yeah refs/remotes/upstream/*	18:35
	→ etcetera joined	18:35
	→ Annika_Doe joined	18:35
	← gigamonkey left	18:36
	gigamonkey_ → gigamonkey	18:36
thiago	when you fetch upstream, it saves upstream's refs/heads/* into refs/remotes/upstream/*	18:36
	← caseymcg left	18:36
	← Annika_Doe left	18:37
cbreak	unless you fetch it somewhere else... :/	18:37
	→ Vortex34 joined	18:37
	ameoba__ → ameoba	18:37
	← DanyO83 left	18:38
apipkin	the issue i'm running into is with issuing a pull request on github (may or may not be in scope of the channel). explaining in here feels like writing a book haha one sec	18:38
thiago	pull requests are not in scope	18:38
	→ patrick85 joined	18:38
thiago	a pull request reduces to notifying someone that your work is ready in a given branch, at a given remote	18:39
	← patrick85 left	18:39
	← Vortex35 left	18:39
thiago	is your problem with making your changes available in that repository?	18:39
	→ paul_andrew joined	18:39
apipkin	ah okay :)	18:39
	← guampa left	18:40
	→ maqr joined	18:40
	→ guampa joined	18:40
	← guampa left	18:40
	← paul_andrew left	18:41
apipkin	this was my process if that helps any https://gist.github.com/4e51054c0efd00f74d35	18:42
	and please let me know if i did it all wrong :)	18:42
	← anish[1] left	18:42
	← koltroll left	18:43
apipkin	and though pull requests are out of scope, the issue is that the pull request says there were 180 files changed and 80 commits	18:43
	so i'm not sure if there is a cleaner way to do this, or if that's as good as it gets	18:43
_ikke_	apipkin: Then you made the pull request against the wrong branch	18:44
	→ pandeiro joined	18:44
	→ patrick85 joined	18:44
	→ specialGuest joined	18:44
	→ paul_andrew joined	18:45
	← paul_andrew left	18:45
	← fly9 left	18:46
	→ fly9 joined	18:46
	← Sigma left	18:46
	← juvenal left	18:47
	→ revagomes joined	18:48
	← ipalaus left	18:49
apipkin	ah okay. i'll take another look at it and see where i went wrong. thanks guys/gals	18:49
	→ ipalaus joined	18:49
	→ juvenal joined	18:50
EugeneKay	Women are not gits. They're gitesses	18:50
_ikke_	lol	18:51
	Where thit that came from?	18:51
	← revagomes left	18:52
	→ artnez joined	18:52
EugeneKay	I just made it up	18:52
	→ jgoss joined	18:54
	→ guampa joined	18:55
	→ paul_andrew joined	18:56
	→ Sigma joined	18:56
_ikke_	Obviously	18:56
	← tigris_ left	18:57
	← paul_andrew left	18:57
	→ AdrienBrault joined	18:57
	← fisted_ left	18:59
	→ Advocation joined	19:00
mjt	that's inspired by the !hello message i guess	19:00
gitinfo	[!welcome] Welcome to #git, a place full of helpful gits. If you have a question, feel free to just go ahead and ask—somebody should answer shortly. For more info on this channel, see http://jk.gs/git/ Take backups (type !backup to learn how) before taking advice.	19:00
	← rcu left	19:00
	→ fisted joined	19:00
EugeneKay	I think I wrote that	19:01
	→ uri111 joined	19:01
	→ paul_andrew joined	19:01
	← paul_andrew left	19:01
	← chimay left	19:01
mjt	so to be politically correct you have to add "and gitesses" there :)	19:01
_ikke_	EugeneKay: You wrote the seccond version. V3 was written by SethRobbinson	19:01
	→ Hasbro joined	19:01
	← uri111 left	19:02
_ikke_	SethRobertson*	19:02
	← Hasbro left	19:02
	→ Hasbro joined	19:02
	← apipkin left	19:03
EugeneKay	Sounds right.	19:03
_ikke_	ftr, I wrote V1	19:03
	But that was different	19:03
EugeneKay	But that one didn't mention "a place full of helpful gits"	19:03
_ikke_	nope	19:04
	→ axiom92 joined	19:05
	← brenden1 left	19:06
	→ hwrd\|work joined	19:07
	← devth left	19:07
	← rgsteele left	19:07
	← axiom92 left	19:08
	← zeppo left	19:10
	→ paul_andrew joined	19:12
	← valtih left	19:12
	← paul_andrew left	19:13
	→ doorp joined	19:14
	→ zeppo joined	19:14
	← doorp left	19:15
	→ dorp joined	19:15
	← keltvek left	19:15
	← tjholowaychuk left	19:15
dorp	When I "git add -A ." ... it actually reads each and every file? Or only the files properies, timestamps/length comparison?	19:16
	I've read the manual, but I can't find an option to force by-content mode?	19:16
cmn	it tries to avoid reading in the file	19:16
	→ donri joined	19:16
thiago	it does a timestamp comparison first	19:16
	if the timestamp differs from the index, it hashes the file to check if it's the same	19:17
dorp	I think I have a scenario where git add wouldn't add anything because the file has the same length and modified-timestamp	19:17
	Is it optional to make git add to read all the files contents for comparisons?	19:17
mmattice	why wouldn't the timestamp be getting updated?	19:18
	→ IrcBotBundle joined	19:18
	→ zastaph joined	19:18
thiago	exactly: why is the timestamp the same?	19:18
dorp	I'm trying an automated script for building a branch, while the export process is very very fast on a SSD, which produces many files with the exact same timestamp	19:19
	So either I add a "sleep", which is nasty... or start putting random last-mod stamps, which is nasty never-the-less :\	19:20
	← bizhanMona left	19:20
	→ neunon_ joined	19:20
dorp	So I guess git-add can't evaluate the file contents?	19:20
	→ keltvek joined	19:20
mmattice	why would git need to add anything for a build?	19:20
	→ devth joined	19:21
	→ Kramer_ joined	19:21
	→ shruggar joined	19:21
dorp	mmattice: Think of something like a conversion from non-git branch, to a git-branch, so I'm trying to re-build revisions as a git branch from an external non-git source	19:21
	← neunon_ left	19:21
	← Davey left	19:22
	← Kramer_ left	19:22
dorp	mmattice: It's not a "build" as-in compile/make, but actually "creating" the branch with preexisting revisions	19:22
	→ Shusshu joined	19:22
	→ uri111 joined	19:22
	← Shusshu left	19:22
thiago	dorp: file timestamps record nanosecond resolution	19:22
	← uri111 left	19:22
	→ bizhanMona joined	19:23
	← converge left	19:23
	← Sigma left	19:24
	→ Davey joined	19:24
	← IrcBotBundle left	19:24
	→ Eridius joined	19:25
	→ dangerousdave joined	19:25
dorp	thiago: nanosecond? it's not depended on the util / filesystem?	19:25
	→ converge joined	19:25
	→ someoh43 joined	19:27
	← dr_lepper left	19:27
doherty	I've been doing `git merge -r --no-merge \| grep stuff` to find relevant and un-merged branches. How do I see the changes that remain to be merged between one such branch and master?	19:27
	← iocor left	19:27
thiago	dorp: sure, it depends on the fs	19:28
	dorp: but fast systems compiling is the reason why milliseconds are stored since the 1990s	19:28
	→ brenden1 joined	19:28
	← keltvek left	19:28
thiago	if your system is fast enough that you can modify-commit-modify within 1 second, you should really have a modern fs	19:29
OMGOMG	2 Modern view	19:29
	3 See also	19:29
	← erichynds left	19:30
dorp	Or a very small repo, high-performance ssd, etc etc	19:30
	← axl_ left	19:30
	→ iocor joined	19:31
	← someoh43 left	19:31
	→ avpx joined	19:31
	→ chuckhar_ joined	19:32
	→ srayner joined	19:32
	← mewm left	19:32
	→ keltvek joined	19:32
cmn	dorp: use update-index instead	19:32
	that one is for scripts and lets you control parts of it	19:33
	→ Kramer_ joined	19:33
	→ dr_lepper joined	19:33
	← dr_lepper left	19:33
	→ dr_lepper joined	19:33
	← Kramer_ left	19:34
	← harshpb left	19:34
	→ Kramer2718 joined	19:34
dorp	cmn: Thanks, I'm reading about it	19:35
	→ harshpb joined	19:35
	← chuckharmston left	19:35
	← Kramer2718 left	19:36
	← patrick85 left	19:38
	→ kevwil joined	19:38
	← harshpb left	19:39
	← hwrd\|work left	19:39
	→ harshpb joined	19:39
linduxed	i've got a submodule placed under .foobar, and i want to move it to .barfoo. When i try to do "git mv" i get "fatal: source directory is empty, source=.oh-my-zsh, destination=.zprezto"	19:41
	how would i best go about doing this?	19:41
	← chrisgeorge left	19:42
	← specialGuest left	19:42
NfNitLoop	why are you putting submodules in hidden directories? That seems dubious...	19:42
	→ dardevelin joined	19:43
	← sanmarcos left	19:43
NfNitLoop	Have you tried adding the submodule in the new location and removing it from the old one?	19:44
	(I don't see a git submodule mv, unfortunately.)	19:44
	→ tjholowaychuk joined	19:44
	→ SidGBF joined	19:44
	→ specialGuest joined	19:44
	→ paul_andrew joined	19:45
	← rue left	19:45
	← harshpb left	19:46
	← paul_andrew left	19:46
	→ rue joined	19:46
	→ harshpb joined	19:46
SidGBF	http://pastebin.com/M02wQgMn guys, Im trying to send some stuff to a new (bare) repo. But Im missing something, can someone help me?	19:47
EugeneKay	linduxed - I would rm the old location and add the new one	19:47
	Given as that's all that `git mv` does, it is pretty git-y.	19:47
	← cworth left	19:47
	→ kvanderw joined	19:48
	← specialGuest left	19:49
	→ specialGuest joined	19:49
	→ someoh43 joined	19:50
	→ paul_andrew joined	19:50
	← paul_andrew left	19:50
bigmeow	Git push: fatal: Out of memory, malloc failed, why?	19:51
	http://stackoverflow.com/questions/11135764/git-push-fatal-out-of-memory-malloc-failed	19:51
linduxed	EugeneKay: ok	19:52
thiago	bigmeow: because you're out of memory	19:52
	bigmeow: malloc failed, so it can't continue	19:52
EugeneKay	bigmeow - because you don't have enough RAM. How big is your repo / the RAM of machines involved?	19:52
thiago	also: are you on 32-bit? And do you have large files inside the repo?	19:52
	→ cworth joined	19:53
	→ chuckharmston joined	19:54
	→ centipedefarmer joined	19:54
	← srayner left	19:55
	→ Sym3tri joined	19:55
	← Advocation left	19:55
	← shruggar left	19:55
	← etcetera left	19:55
	← chuckhar_ left	19:56
	→ Advocation joined	19:57
	→ mewm joined	19:57
	← someoh43 left	19:57
	← mdw left	20:00
bigmeow	EugeneKay: linux kernel, 1.2G	20:00
EugeneKay	Haha yeah	20:01
	Get more RAM. ;-)	20:01
	yardenac → torquemada	20:01
bigmeow	EugeneKay: git is kind of hell, why so much memory is needed?	20:01
EugeneKay	Because git is checking the checksums of all files in the repo at clone time	20:01
	The fact that it mallocs the entire thing is inefficient, but blame the git who wrote it. Patches welcome ;-)	20:02
	← RaceCondition left	20:02
cmn	it needs to hash and delta a lot of objects	20:02
	which means that it needs to keep them in memory	20:02
	→ axl_ joined	20:02
	→ john_anderson joined	20:03
	← Milossh left	20:03
	→ _adnasa_ joined	20:03
	← _adnasa_ left	20:03
	← Advocation left	20:04
	← dsantiago left	20:04
	→ _adnasa_ joined	20:04
	→ etcetera joined	20:04
	← harshpb left	20:04
EugeneKay	For "most" repos it isn't a problem because no sane person would have a 1GB repo. But linux.git is.... popular.	20:04
	→ Milossh joined	20:05
	→ RaceCondition joined	20:05
	← Milossh left	20:05
	→ mdw joined	20:05
Nugget	"Here's a nickel, kid, go buy yourself a better computer."	20:05
abackstrom	can i replay a series of commits, kinda like a rebase that stops after every commit is applied?	20:06
	← Hasbro left	20:06
	→ Milossh joined	20:06
canton7	what do you want to do when it stops?	20:06
thiago	abackstrom: git cherry-pick each commit	20:06
	← gun-lpt left	20:06
thiago	abackstrom: just write a script	20:06
	← Milossh left	20:06
abackstrom	canton7: run my integration tests, i want to see which commit is breaking the tests	20:06
canton7	or you can rebase -i and s/pick/edit?	20:06
	abackstrom, man git-bisect	20:07
gitinfo	abackstrom: the git-bisect manpage is available at http://git-scm.com/docs/git-bisect.html	20:07
canton7	we have tools for that :)	20:07
SidGBF	:-/	20:07
	→ Milossh joined	20:07
abackstrom	canton7: thank you	20:08
	→ mishok13 joined	20:08
	← Milossh left	20:08
	→ Milossh joined	20:09
	→ srayner joined	20:09
	→ Advocation joined	20:10
SidGBF	YAY!!!	20:10
	← Advocation left	20:11
	← RaceCondition left	20:11
canton7	SidGBF, got it?	20:11
	otherwise, pastie the output of 'git branch'	20:12
	→ lb joined	20:12
	← axl_ left	20:12
	→ koosha joined	20:12
	→ patrick85 joined	20:13
koosha	Do the Git community guys have a bug tracking system to see open bugs and write patch for them?	20:13
	→ RaceCondition joined	20:13
canton7	koosha, we use the mailing list. [email@hidden.address]	20:14
koosha	canton7: mailing list is a bit weird for bug traking!	20:14
canton7	koosha, there's a very convincing mail from the maintainer somewhere on the subject. anyone got a link?	20:14
	← fly9 left	20:14
	→ fly9 joined	20:15
	← koosha left	20:15
	→ hwit joined	20:15
lb	btw, who's maintaining git ?	20:16
milki	how do you bug trake?	20:16
canton7	lb, Junio C Hamano	20:17
lb	thanks	20:17
	canton7: thank you	20:17
	→ gfontenot joined	20:17
	← Samual left	20:18
	← mdw left	20:18
	→ Samual joined	20:18
	← Samual left	20:18
	→ Samual joined	20:18
	← Fullmoon left	20:18
	← skyf_ left	20:18
	→ Fullmoon joined	20:19
	← doherty left	20:20
bc	does setting core.excludesfile to ~/.gitignore make /project/.gitignore no longer looked at?	20:21
	← skyf left	20:22
	→ Sigma joined	20:23
	→ bluj joined	20:24
	→ beneggett joined	20:25
bluj	help? i have a second remote added in my repo. i have one branch already which tracks a branch of this second remote. i'm trying to do a new "git checkout -b newbranch -t secondremote/newbranch" but it isn't aware of this branch at the second remote? if i do a "git branch -a" it's not aware of any branches on this second remote except the one i'm already somehow tracking	20:26
canton7	bluj, git fetch secondremote ?	20:27
	also, that -t isn't needed: it's implied	20:27
	← centipedefarmer left	20:27
mr6	bluj: if after git fetch secondremote it still doesn't show up in git branch -a you haven't pushed the branch yet	20:27
bluj	canton7: tried, "git branch -a" still just shows me the branch on the secondremote i'm currently tracking (but i dont want that one)	20:27
	pushed the branch? hmm...	20:28
	← aircan left	20:28
mr6	bluj: git push secondremote newbranch:newbranch	20:28
canton7	bluj, git config remote.secondremote.fetch ?	20:28
	→ tatsuya_o joined	20:28
bluj	canton7: +refs/heads/branchidontcareabout:refs/remotes/brett/branchidontcareabout	20:29
	← bloony- left	20:29
canton7	bluj, aha, it's stange how often this problem crops up	20:29
bluj	mr6: i don't get why i need to push anything (and where shuld that be executed?) the branch exists on the secondremote, and it currently 1 commit ahead of the upstream branch IT tracks	20:29
canton7	as you can see, git's configured to only fetch that branch	20:29
	bluj, change that to ~refs/heads/:refs/remotes/brett/	20:29
bluj	canton7: thanks, sec	20:30
	← hwit left	20:30
canton7	oops +refs/heads/:refs/remotes/brett/	20:30
	← Ryback_ left	20:30
mr6	bluj: fair enough, but you can see how that could happen if you didn't have the branch on secondremote	20:30
bluj	canton7: thanks that worked. how did this happen? i don't use secondary remotes very often, did i create the remote incorrectly or something?	20:31
	→ mdw joined	20:31
canton7	bluj, how did you add that remote? git remote add -t ?	20:31
	← mdw left	20:31
bluj	canton7: git remote add brett ssh://ip/home/brett/git/repo	20:32
	if i recall	20:32
	wrong, i did use -t actually, oops!	20:32
canton7	I bet it was git remote add -t branchyoudontcareabout brett url :P	20:32
bluj	you bet right.	20:32
canton7	that's why ;)	20:32
bluj	thank you	20:32
	→ maestrojed joined	20:32
mr6	hah so why use -t if you don't know what it does?	20:33
	← maestrojed left	20:33
	→ aircan joined	20:34
bluj	mr6: i use it on "checkout -b" all the time because i know what it does, even though as mentioned above it may not be required. looks like i just made a mistake here	20:34
	→ savr joined	20:35
savr	hi	20:35
	→ axl_ joined	20:35
mr6	bluj: np just curious as i hadn't ever used that flag to add remote	20:36
	← Sigma left	20:36
	← _iron left	20:36
mr6	bluj: wondered if I was missing out on some extra benefit heh	20:36
savr	I previously merged from a development branch from an upstream remote	20:37
	← zeppo left	20:37
savr	I now want to revert back down to the upstream master	20:37
	→ zeppo joined	20:37
savr	is it safe to git merge upstream/master ?	20:37
	← zeppo left	20:38
	← converge left	20:38
	→ chuckhar_ joined	20:39
	→ cagedwisdom joined	20:39
	← geekbri left	20:40
	← TheJH left	20:41
	← chuckharmston left	20:41
	→ jn_ joined	20:41
	← bartek left	20:42
	→ converge joined	20:42
savr	?	20:44
lb	savr: you want to get back to the state of upstream/master?	20:44
savr	yeah	20:44
	← SidGBF left	20:44
	← axl_ left	20:44
	← mishok13 left	20:44
savr	and keep my one commit	20:44
lb	then merge is not the way to go	20:44
savr	but I can commit it again	20:45
lb	so you want to rebase your commit on top of upstream/master?	20:45
savr	I guess so	20:47
	→ chrisgeorge joined	20:47
	← s0ber left	20:48
lb	man git rebase	20:49
gitinfo	the git-rebase manpage is available at http://git-scm.com/docs/git-rebase.html	20:49
	→ s0ber joined	20:49
	← Vortex34 left	20:50
	← ShooterMG left	20:50
	→ zastern joined	20:50
	→ mklappstuhl joined	20:53
	→ ShooterMG joined	20:53
	← etcetera left	20:53
	← incognit0 left	20:53
	→ xiaotian joined	20:54
	→ bartek joined	20:54
	← imachuchu left	20:54
	← kevwil left	20:55
	→ towski joined	20:56
savr	thanks lb	20:57
	← irqq left	20:57
berndj	is there a way to get a script i wrote to answer y/n to git add -p?	20:57
	→ jcao219 joined	20:57
	→ rubbish joined	20:57
	← freeed left	20:58
	← rubbish left	20:58
_ikke_	berndj: Maybe expect	20:59
	→ maestrojed joined	21:00
_ikke_	http://linux.die.net/man/1/expect	21:00
	→ shruggar joined	21:00
	← patrick85 left	21:00
berndj	i hadn't thought of expect(1), thanks	21:00
	← shruggar left	21:00
	→ axl_ joined	21:01
berndj	it's (should be!) an order-only diff, in case you're wondering, but beyond what sort(1) on its own could decide	21:02
	→ etcetera joined	21:02
rking	What kind of Twilight Zone am I in if I see "M"s for "git status", but "git diff --cached" and "git diff" show nothing?	21:03
	→ xplat joined	21:03
	← rwestphal left	21:03
	← maestrojed left	21:03
	→ ddelrio1986_ joined	21:03
rking	Aww, crud. Has something to do with my ~/.gitconfig	21:03
FauxFaux	Almost certainly autocrlf fucking you.	21:03
	← Zenmaker_ left	21:03
	→ Zenmaker_ joined	21:04
	← ShadeTornado left	21:04
rking	I didn't have -X in pager = less -FR	21:04
	← tatsuya_o left	21:04
rking	What are those flags supposed to be? Gentoo fubars it by setting $LESS	21:04
	→ Zenmaker__ joined	21:04
	← mritz left	21:05
rking	$LESS is set to -R -M --shift 5, so I have to set pager = less -_something_ to make git work as it wants.	21:05
	← ddelrio1986 left	21:05
	← etcetera left	21:05
rking	I know -R has to be in there to get colors, and -F makes sense as quit-if-one-screen, and apparently -F without -X is pathological in that it shows nothing.	21:06
	← sebastorama left	21:06
	→ etcetera joined	21:06
	← bartek left	21:06
	← etcetera left	21:06
	← Sym3tri left	21:07
	← spec__ left	21:08
	→ etcetera joined	21:08
	→ YoMomma joined	21:08
	← Zenmaker_ left	21:09
	← Element9 left	21:09
	← tplaner_ left	21:09
	→ maestrojed joined	21:09
	← tinti left	21:10
	← toabctl left	21:10
	← axl_ left	21:11
	← etcetera left	21:12
	← xiaotian left	21:12
	→ xiaotian joined	21:12
	← xiaotian left	21:13
	← torjus_ left	21:13
	← converge left	21:13
	→ xiaotian joined	21:13
	← ncd\|leeN left	21:14
	← cannonball left	21:14
	→ etcetera joined	21:14
	← ddelrio1986_ left	21:15
	→ axl_ joined	21:15
	→ aguynamedben joined	21:17
	→ gusnan joined	21:18
	← aguynamedben left	21:19
	→ jjbohn joined	21:19
	← Austin__ left	21:21
	→ aguynamedben joined	21:21
	← ipalaus left	21:24
	→ _huoxito joined	21:24
BSaboia	hello everyone. how do I completly removed a file from being tracked when i push? ie.: in my local repository, i created a file, then i deleted it in a posterior version. i do not want this file to go to the repository when i push. what should i do?	21:26
	← les_sylvains left	21:26
thiago	BSaboia: have you pushed the commit that added it in the first place?	21:27
BSaboia	no	21:27
	← hellopat left	21:27
thiago	have you done any merges since you added it/	21:27
BSaboia	thiago, no	21:28
	everything is only local since them	21:29
	but i want to stop tracking the file, and delete it from my local history	21:29
	← srayner left	21:29
	ceradon-away → euphoria	21:30
	→ drizzd joined	21:31
	← etcetera left	21:31
thiago	BSaboia: git rebase -i firstcommit~	21:34
	including the ~	21:34
	then edit that commit and remove the file	21:34
	← martinjlowm left	21:34
	← iocor left	21:34
BSaboia	thiago, you mean by fisrt commit the hash of the commit that i added the file?	21:37
thiago	yes	21:37
	or any commit before it	21:37
BSaboia	before? but if i do it before, the file will appear nevertheless?	21:38
	→ kcj joined	21:38
cmn	with any commit before it, you could be rewriting published history	21:38
	← savr left	21:38
BSaboia	i'm a little bit confused, let me see if i understand everything. i am in commit N, lets say. i added the file in some previous commit I can't recall, like N-y. I deleted it already in N-1. my expected behaviour is that when i push, that file should be, to git, as if i never added it... it never existed	21:40
	so, i should rebase to a commit where this file does not exist yet	21:40
	but then, how would i delete it?	21:40
thiago	BSaboia: that's not the expected behaviour	21:40
	if you push now, without rewriting history, the file will be pushed	21:40
	if you want to remove it from history, use rebase	21:41
BSaboia	by expected behaviour i mean what i want to do	21:41
	i know that if i push	21:41
thiago	edit the commit that added the file so it will never have been added	21:41
BSaboia	it will be there	21:41
thiago	if you modified the file in any commits, they'll conflict	21:41
BSaboia	ok	21:41
thiago	the deletion at the end will disappear because you'd be deleting a file that doesn't exist	21:41
BSaboia	i never modified it, so no worries here	21:41
	← flavius left	21:43
	← opus left	21:43
	← zastaph left	21:43
	← ngirard left	21:43
	microcolonel → kiken	21:44
	→ youngblood joined	21:44
	← lfaraone__ left	21:46
	→ lustucru joined	21:46
	← mewm left	21:46
	→ lfaraone__ joined	21:46
	← spaceonline left	21:47
	→ chrisgeorge_ joined	21:47
	← chrisgeorge_ left	21:47
	→ chrisgeorge_ joined	21:47
	→ opus joined	21:49
	→ alberto56 joined	21:49
	← chrisgeorge left	21:51
	← lfaraone__ left	21:52
	← mklappstuhl left	21:53
	← dvide left	21:53
	← sgambino left	21:54
	← YoMomma left	21:54
	← whitman left	21:56
	← alexisabril left	21:57
	→ BiggFREE joined	21:58
	→ lfaraone__ joined	21:59
	euphoria → ceradon-away	21:59
	← bapa left	21:59
	← kermit left	22:01
	← dr_lepper left	22:01
	← youngblood left	22:02
	→ dr_lepper joined	22:02
	← dr_lepper left	22:02
	→ dr_lepper joined	22:02
	← wanders_ left	22:04
	→ bapa joined	22:04
	← Chillance left	22:05
	→ kermit joined	22:05
	→ mklappstuhl joined	22:06
	→ robbyoconnor joined	22:06
	→ sfrank1 joined	22:07
	← sfrank1 left	22:07
	← Destos left	22:07
	← lavaman left	22:09
	← sfrank left	22:09
BSaboia	thiago, using this git log --diff-filter=D --summary i was able to find the hash for the commit i've deleted the file	22:09
	→ YoMomma joined	22:10
	← dangerousdave left	22:10
	pdelvo → pdelvo\|afk	22:10
	oakdog8 → _oakdog8	22:10
	← YoMomma left	22:10
thiago	BSaboia: use filter A and find the commit where you added the file	22:10
	→ YoMomma joined	22:11
	→ wanders joined	22:12
	← ehsan left	22:12
	→ madewokherd joined	22:12
	← ph^ left	22:12
	← stepnem left	22:13
	→ Chillance joined	22:13
BSaboia	thiago, done	22:14
	pdelvo\|afk → pdelvo	22:14
thiago	now rebase -i from one commit before	22:15
BSaboia	using ~1?	22:15
thiago	yes	22:16
	→ skyf joined	22:16
	← drizzd left	22:16
	← ketas left	22:17
BSaboia	ok, and then?	22:17
	pdelvo → pdelvo\|afk	22:18
	← chrisgeorge_ left	22:18
	← juvenal left	22:18
	→ chrisgeorge joined	22:18
	pdelvo\|afk → pdelvo	22:19
	← YoMomma left	22:19
	← fly9 left	22:20
	← mklappstuhl left	22:21
	→ eintw1ck joined	22:21
eintw1ck	can I get help on how to use git here	22:22
thiago	no, we just talk about svn and hg :-)	22:22
eintw1ck	Just checking	22:22
	→ sebastorama joined	22:22
	→ Vert joined	22:23
eintw1ck	I got a repo at github.com/eintw1ck/We-Browser-Code-Only.git	22:23
	← nomadmonad left	22:23
eintw1ck	I've done the git init on my local machine what next	22:23
	→ nomadmonad joined	22:24
	← Swathe left	22:24
diegoviola	is there a way to check what commits i haven't pushed yet to my remote repo?	22:24
cbreak	you should git clone the original eintw1ck	22:24
	if you want to stick with your inited one, git remote add, then fetch, then checkout, and so on	22:24
diegoviola	compare what commits i have in my local repo with my remote one, etc	22:24
cbreak	but clone does all that in one command	22:24
	diegoviola: git remote update	22:24
eintw1ck	I've done git remote add thingy	22:25
cbreak	diegoviola: and then git log @{u}..	22:25
eintw1ck	But then it says invalid repository	22:25
diegoviola	thanks	22:25
cbreak	eintw1ck: maybe it's invalid :)	22:25
	eintw1ck: do a git ls-remote on the URL	22:25
eintw1ck	Then I cd into my local git repo and it kinda works	22:26
	It was working on my mac but now im on my iPhone it doesn't really work	22:26
cmn	then? you need to do git operations in git repos	22:26
	← TeckniX left	22:27
cbreak	eintw1ck: as I said above, you're better off cloning	22:27
	→ ketas joined	22:27
eintw1ck	Ok	22:27
	← eintw1ck left	22:27
	→ t0rc joined	22:28
	→ iocor joined	22:29
	← dr_lepper left	22:32
	→ eintw1ck joined	22:33
	← tehgeekmeister left	22:33
	← glennpratt left	22:34
	← jbub left	22:36
eintw1ck	cbreak: I did it like you said and this is what I got	22:39
	error: Cannot access URL https://github.com/eintw1ck/We-Browser-Code-O nly.git/, return code 77 error: failed to push some refs to 'https://github.com/eintw1ck/We-Bro wser-Code-Only.git'	22:39
	← pantsman left	22:39
cbreak	push?	22:39
	you did not do what I said.	22:39
	also, you have random spaces in your URL	22:39
eintw1ck	I did	22:39
	Weird	22:40
	→ markatto joined	22:40
eintw1ck	I'll fix that	22:40
	← neurodrone left	22:40
	← s0ber left	22:40
	← Protux left	22:40
	← Olipro left	22:40
cbreak	> git ls-remote https://github.com/eintw1ck/We-Browser-Code-Only.git	22:40
	works for me	22:40
	→ s0ber joined	22:41
	← chuckhar_ left	22:41
	→ neurodrone joined	22:41
	← Chillance left	22:41
	← eintw1ck left	22:41
	← meLon left	22:42
	→ eintw1ck joined	22:42
eintw1ck	cbreak: It works for me too I don't know why I cant push	22:43
	→ DanyO83 joined	22:43
cbreak	pushing is a different issue	22:43
	ideally you'd use the ssh URL for that	22:43
eintw1ck	Oh	22:43
	Why?	22:43
cbreak	although I think github uses smart HTTP with push support	22:43
	← aircan left	22:44
	← axl_ left	22:44
eintw1ck	I did it 4 hours ago on my mac maybe it's the version of git?	22:44
	→ bloony- joined	22:44
eintw1ck	Could it be?	22:45
cbreak	is yours really ancient?	22:45
	git had push support since many years.	22:45
eintw1ck	No I got the latest iPhone compiled git	22:45
	→ Olipro joined	22:45
	← cakehero left	22:45
	← ShooterMG left	22:46
cbreak	missing proper libcurl?	22:46
eintw1ck	Got it	22:46
	Well the iPhone version	22:46
	→ meLon joined	22:46
	→ jbub joined	22:46
	← meLon left	22:46
	→ meLon joined	22:46
cbreak	well, try ssh.	22:47
eintw1ck	It doesn't seem to work	22:47
	← pandeiro left	22:47
eintw1ck	I've tried	22:47
	← afuentes left	22:47
	→ lavaman joined	22:47
	← skyf left	22:47
eintw1ck	ssh (the ssh URL they give) and it returns an invalid URL	22:47
cbreak	sounds like your libraries or git is broken.	22:48
	it should not classify URLs as invalid if they are valid	22:48
eintw1ck	most likely git I'll speak to iPhone jailbreak people as this is more iPhone talk but I when it get git working I'll come back	22:49
	→ dsantiago joined	22:51
	→ skyf joined	22:51
	← Synthead left	22:51
	← sebastorama left	22:53
	→ dr_lepper joined	22:53
	← dr_lepper left	22:53
	→ dr_lepper joined	22:53
	→ metcalfc joined	22:53
	→ billyoc` joined	22:54
	← billyoc` left	22:55
	← billyoc left	22:55
	← gfontenot left	22:56
	← skyf left	22:56
	→ skyf joined	22:57
	← opus left	22:57
	← macabre left	22:57
	→ jjang joined	23:00
	← _adnasa_ left	23:00
	← lb left	23:01
	← Olipro left	23:03
	← xiaotian left	23:03
	→ opus joined	23:03
	← jargon- left	23:05
	pdelvo → pdelvo\|afk	23:05
	← jn_ left	23:06
	→ simesy joined	23:07
	← zastern left	23:08
	← thiago left	23:09
	→ Olipro joined	23:10
	→ adamm joined	23:10
	← mastro left	23:10
	→ ehsan joined	23:10
	← dzonder left	23:11
	← simesy left	23:11
	→ simesy joined	23:12
	→ cakehero joined	23:12
	← lacrymology left	23:12
	← Olipro left	23:14
	→ ChristopheT joined	23:15
	← gigamonkey left	23:16
	→ gigamonkey joined	23:18
	← jjbohn left	23:18
	← SirCmpwn left	23:19
	→ tinti joined	23:23
	→ ph^ joined	23:23
	← maestrojed left	23:24
	→ Olipro joined	23:24
	← Davey left	23:25
	→ axl_ joined	23:27
	← ph^ left	23:27
	→ nowhere_man joined	23:29
	← akiress left	23:29
	← ChristopheT left	23:32
	→ dfr\|mac joined	23:32
	← boombatower left	23:32
	→ Swathe joined	23:33
	→ osmosis joined	23:35
	← chrisgeorge left	23:36
	← eintw1ck left	23:36
	← cakehero left	23:37
	← Swathe left	23:37
	← specialGuest left	23:38
	→ Esa_ joined	23:40
	→ cakehero joined	23:42
	→ klj613 joined	23:42
	→ acl_ joined	23:44
	← opus left	23:44
	← dsantiago left	23:46
	← v0n left	23:46
	→ etcetera joined	23:47
	← AdrienBrault left	23:47
	← BiggFREE left	23:47
	← axl_ left	23:48
	→ glennpratt joined	23:49
	← hyperair left	23:50
	← robbyoconnor left	23:50
	→ Goplat joined	23:51
	← kvanderw left	23:52
	→ opus joined	23:54
	→ cybersphinx joined	23:55
	← BSaboia left	23:56
	← AlbireoX`Laptop left	23:56
	→ dfr\|mac_ joined	23:57
	← lavaman left	23:57
	← dfr\|mac left	23:57
	→ jdav_gone joined	23:57
	→ dsantiago joined	23:57
	← jjang left	23:57
	← opus left	23:57
	→ rchavik joined	23:58
	← rchavik left	23:58
	→ rchavik joined	23:58
	← the_cybersphinx left	23:58
	← jdav_gon` left	23:59

Logs Search ←Prev date Next date→

Channels Documentation