IRCloggy #git 2013-10-10

Provider of IRC logs since 2005.

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

	2013-10-10
	→ pr0ton joined	00:01
	→ LuxuryMode joined	00:01
	← pr0ton left	00:02
	← boombatower__ left	00:03
	→ bwreilly joined	00:03
	→ bitkiller joined	00:04
	→ _live joined	00:04
	→ cannonball joined	00:04
	→ FabTG_ joined	00:05
	← _live left	00:05
	← quazimodo left	00:05
	← YannH left	00:05
	→ _live joined	00:05
	→ pr0ton joined	00:05
	→ stamina joined	00:05
	← _iron left	00:05
	← _live left	00:05
	← reskp left	00:06
	→ Robarelli joined	00:06
	← Roots47 left	00:07
	← FabTG left	00:07
	← bakers left	00:08
	→ _live joined	00:09
	← _live left	00:09
	← hwrd\|work left	00:10
	← zjason` left	00:13
	← fowlduck left	00:13
	← alvaro_o left	00:13
	→ alvaro_o joined	00:14
	→ Dougie187 joined	00:15
	← stamina left	00:15
	→ zjason joined	00:15
	→ HereCas joined	00:18
	← vigs left	00:19
	→ maslowbeer joined	00:20
	← starfox21 left	00:20
	← brettnem left	00:20
	→ normanrichards joined	00:21
	← harish_ left	00:21
	← Baluse left	00:22
	→ petey joined	00:22
	→ Zushakon joined	00:22
	→ mFacenet joined	00:22
	→ dyng joined	00:23
	← CycoCid left	00:24
	← mFacenet left	00:25
	← petey left	00:26
	→ petey joined	00:27
	← dreiss_ left	00:29
	← mklappstuhl left	00:30
	← petey left	00:31
	→ mklappstuhl joined	00:32
	→ davidmturner joined	00:32
	← bwreilly left	00:32
	← PeterBooker left	00:33
	← acharlieh left	00:33
	→ petey joined	00:34
	← mattmcclure left	00:34
	← hyperair left	00:34
	← davidmturner left	00:34
	← bsaboia_ left	00:37
	← mklappstuhl left	00:37
	← rednaks left	00:37
	← Robarelli left	00:38
	→ Robarelli joined	00:39
	→ Swayz joined	00:39
	← bunbury left	00:41
	→ axl_ joined	00:43
	← Andrevan left	00:43
	← Zarthus left	00:44
	← rat_thing left	00:45
	← minim left	00:46
	← KillYourTV left	00:49
	→ fuzzybunny69y joined	00:50
	→ minim joined	00:52
	→ jaequery joined	00:53
	← cannonball left	00:54
	→ clreed87 joined	00:54
	← clreed87 left	00:55
	← minim left	00:55
	← HereCas left	00:56
	← exobyte left	00:56
	← axl_ left	00:57
	→ R1KRDO joined	00:57
perlstein	does anyone rememeber my problems from before? this is how i feel: http://devopsreactions.tumblr.com/post/61746167797/spent-all-day-fixing-merge-conflict-wrong-branch	00:58
	← freestyl3r left	00:58
	← tufflax left	00:59
	→ carbohydrate joined	00:59
	→ axl_ joined	00:59
	→ freestyl3r joined	00:59
	← Kabaka left	01:01
	→ Entomo joined	01:01
	→ luxgeek joined	01:01
	→ Kabaka joined	01:02
	→ Mikicacarica joined	01:03
	← kukks left	01:04
	← Mikicacarica left	01:04
	← shinnya left	01:06
	← R1KRDO left	01:06
	→ KillYourTV joined	01:07
	→ Vampire0 joined	01:08
	→ langhun joined	01:08
	→ hyperair joined	01:10
	← hahuang65 left	01:11
	← Vampire0_ left	01:12
	→ rchavik joined	01:13
	→ vinc joined	01:13
	← kmctown left	01:13
	→ foucist joined	01:14
	← normanrichards left	01:15
	→ rat_thing joined	01:16
	← Gooder left	01:16
	→ weiyang joined	01:17
	← thiago left	01:17
	→ gogobudyy joined	01:21
	→ delianides joined	01:22
	← gogobudyy left	01:23
	← maslowbeer left	01:23
	← rat_thing left	01:24
	→ introom23 joined	01:25
	← goshawk left	01:25
	← jhulten left	01:25
	gitinfo set mode: +v	01:25
	→ Goplat joined	01:25
introom23	hi. I have an idle linux box. and I'd like to set it as a git server for data backup use. Has anyone done that before? How's that?	01:25
	→ sayanee joined	01:26
	← jaequery left	01:26
	→ jon_r joined	01:27
	← axl_ left	01:27
	→ joshsmith joined	01:28
	→ jaequery joined	01:28
	→ intripoon_ joined	01:29
	← chrisss404 left	01:29
	← petey left	01:30
	→ petey joined	01:30
	→ chrisss404 joined	01:31
	← intripoon left	01:32
	→ aartist joined	01:32
	← petey left	01:32
Curtisbeef	don't use git for backing up binary files	01:32
	→ petey joined	01:32
Curtisbeef	use rsync and some scripting with a cron job	01:33
	← mribeirodantas left	01:34
	→ mribeiro1antas joined	01:34
	→ mattmcclure joined	01:34
	← batrick left	01:34
	← Swayz left	01:34
introom23	Curtisbeef: If I delete my file locally, I don't want the file will be deleted from the server. Can rsync fulfill that?	01:35
Curtisbeef	yup	01:35
	← shruggar left	01:36
Curtisbeef	look up "rsync backup incremental"	01:36
	→ jhulten joined	01:36
	→ daito joined	01:38
	← jaequery left	01:39
	← LuxuryMode left	01:39
	← Destos left	01:40
	→ batrick joined	01:40
	← pr0ton left	01:41
	→ byprdct joined	01:42
iateadonut	i've got a directory that i want to ignore: /home/me/project/www/ but i don't want to ignore /home/me/project/libs/dompdf/www/	01:42
	← byprdct left	01:42
iateadonut	the root of the project is /home/me/project	01:42
	→ SBoudria1 joined	01:42
iateadonut	i've got www/ in my .gitignore, but that is not correct - it's also preventing libs/dompdf/www from being inserted	01:43
	← Robarelli left	01:43
	← SBoudria1 left	01:43
	← petey left	01:43
	← yacks left	01:43
	→ petey joined	01:44
	← mattmcclure left	01:44
iateadonut	/usr/bin/rsync -ab --backup-dir=/home/back/backup/site_live/old_`date +%F` --exclude=/admin/templates_c/ site@site:/home/site/public_html/ /home/back/backup/site_live/	01:44
	introom23, that's my backup script - /home/back is the home directory of the backup user, while /home/site (user:site, /etc/hosts server name:site)	01:45
	i run that once a day	01:45
introom23	iateadonut: you backup it locally?	01:46
	← cmpera left	01:46
iateadonut	no	01:46
	locally?	01:47
	one server to another	01:47
	→ Newtonianb joined	01:47
iateadonut	go in #linux and ask more.	01:47
	← jhulten left	01:48
	← petey left	01:48
	→ ojacobson joined	01:50
	→ Robarelli joined	01:50
	→ resmo_ joined	01:50
	→ bwreilly joined	01:51
Hello71	iateadonut: because you want /www	01:53
	not www/	01:53
	man gitignore	01:53
gitinfo	the gitignore manpage is available at http://jk.gs/gitignore.html	01:53
	← tinti left	01:53
	← resmo left	01:54
	→ tmcmahon joined	02:00
	← adamwathan left	02:00
	→ synergy__ joined	02:02
	→ luc__ joined	02:03
	→ yacks joined	02:03
	→ refried_ joined	02:03
	→ bunbury joined	02:04
	← FernandoBasso left	02:06
	← GT left	02:06
	← gitpicker left	02:06
	→ kerrick joined	02:07
	← alvaro_o left	02:07
	→ LuxuryMode joined	02:08
	→ gitpicker joined	02:08
	→ rgr joined	02:10
	→ robbyoconnor joined	02:12
	← introom23 left	02:15
	← funfunctor left	02:16
	→ harish joined	02:16
	← LuxuryMode left	02:17
	← Squarism left	02:17
	← tmcmahon left	02:17
	← kerrick left	02:18
	← HenryG left	02:18
	→ MrKeuner joined	02:19
	→ mattmcclure joined	02:19
	← amcsi_work left	02:21
	← Robarelli left	02:23
	→ GT joined	02:25
	→ Robarelli joined	02:26
	← mattmcclure left	02:27
	→ angelsl joined	02:27
	→ vigs joined	02:28
	→ jaequery joined	02:29
	→ jamiis joined	02:32
	→ acharlieh joined	02:32
	← jaequery left	02:34
	← langhun left	02:35
	→ kerrick joined	02:37
	← kerrick left	02:38
	→ R1KRDO joined	02:39
	→ echo-area joined	02:41
	→ jaequery joined	02:42
	→ rurufufuss joined	02:43
	← bwreilly left	02:43
	→ mobileblue joined	02:45
	→ bwreilly joined	02:46
	→ hahuang65 joined	02:46
	← mobilebl- left	02:49
	→ maslowbeer joined	02:51
	→ dong joined	02:51
	← bwreilly left	02:52
	← AaronMT left	02:53
	→ dreiss_ joined	02:54
	→ minim joined	02:54
	← Goplat left	02:55
	← robbyoconnor left	02:55
	← simera left	02:56
	→ Goplat joined	02:56
	← Robarelli left	02:57
	← joki left	02:58
	→ robbyoconnor joined	02:59
	→ Swayz joined	03:00
	← robbyoconnor left	03:00
	← delianides left	03:00
	→ johnnyfuchs joined	03:00
	← johnnyfuchs left	03:00
	→ mattmcclure joined	03:02
	→ EdGrubemran joined	03:02
	← EdGruberman left	03:03
	→ joki joined	03:04
	← Dougie187 left	03:05
	← mattmcclure left	03:07
	← jaequery left	03:09
	← bandroidx left	03:10
	→ delianides joined	03:11
	→ jaequery joined	03:11
	← acharlieh left	03:12
	→ byprdct joined	03:12
	← byprdct left	03:13
	→ byprdct joined	03:13
	← Gamecubic left	03:15
	→ dsdeiz joined	03:16
	← bitkiller left	03:17
	← Swayz left	03:18
	← freestyl3r left	03:19
	→ chirayuk joined	03:20
	← hahuang65 left	03:21
	← dyng left	03:21
	→ dyng joined	03:21
	← MrKeuner left	03:24
	→ bwreilly joined	03:25
	← dyng left	03:26
	← rchavik left	03:28
	← willwh left	03:29
	← carbohydrate left	03:30
	→ redrocket joined	03:36
	← kermit left	03:36
	← echo-area left	03:37
	nitrix → cake	03:38
	cake → nitrix	03:38
	→ el3ctron_ joined	03:39
	→ relipse joined	03:39
	← PaulCapestany left	03:39
	← falconfoxx left	03:39
relipse	how do i get colors in git on my mac osx	03:39
	→ falconfox joined	03:40
	→ bandroidx joined	03:41
	← bandroidx left	03:41
	→ bandroidx joined	03:41
	→ woollyams joined	03:41
relipse	git config --global color.ui true	03:41
	got it	03:41
	how come when i type git commit -am foo, i get a fatal: Paths with -a does not make sense. on mac osx	03:42
	← delianides left	03:42
	→ Guest50535 joined	03:46
	→ functorfun joined	03:46
	functorfun → funfunctor	03:46
	← the_cybersphinx left	03:47
	→ rchavik joined	03:48
	gitinfo set mode: +v	03:48
	Guest50535 → fobius	03:49
	← jaequery left	03:50
fobius	I'm a git newb with a question:	03:50
karstensrage	relipse, doesnt -a mean all?	03:50
relipse	i restarted the terminal and it started working properly	03:50
fobius	Git uses a three way merge as the standard merge procedure I gather. Whenever users make edits to different parts of a file git has an auto merge procedure that it uses to create the new file	03:50
	i.e. if the changes don't conflict	03:51
	then keep both	03:51
	but what about code that has side effects, e.g. global variables?	03:51
	In that case, it seems that git might break the code if it merged in code which changed a global/side effect	03:51
	and code which depends on that global/side effect	03:51
	← chirayuk left	03:51
fobius	so, what are best practices for avoiding that type of situation?	03:52
	→ MrRoot joined	03:53
	→ preflex_ joined	03:53
foucist	fobius: if there's a commit that modifies the global variable, then presumably that same commit has modified the rest of the codebase to handle that change..	03:53
	or something like that	03:53
fobius	Well, I'm talking about two different commits	03:54
	for instance, say there's a method that's been add, addXYZ	03:54
	→ sqlnoob joined	03:54
fobius	and this method is a new feature	03:54
	originally it returns an array	03:54
	← preflex left	03:54
fobius	*not an array, a float	03:54
	preflex_ → preflex	03:54
	→ hahuang65 joined	03:54
fobius	anyway, one user writes some code that uses this method addXYZ	03:54
	but doesn't change the method	03:54
	now, another user comes along and changes the method	03:55
	to return an integer instead of a float	03:55
	These changes will be auto merged since they don't modify the same hunks	03:55
	but the method is now incompatible	03:55
foucist	uh, if another user changes the method, s/he should change the rest of the codebase that depends on that	03:56
	fobius: also, this is what CI (continuous integration) is for	03:56
	like travis CI	03:56
	it detects breakage like this	03:56
fobius	well, I mean, if the changes to the method	03:56
	are being made at the same time that the new code which uses the method is being written	03:56
	then when they merge together it will break, right?	03:57
	→ ingof joined	03:57
foucist	there's no same time though, someone pushes first, the other person that comes along has to pull & merge and make sure his changes don't break stuff (ideally)	03:57
	or if it's done through pull requests then sometimes the person needs to re-pull and redo his stuff and update the pull request	03:57
	if the merger encounters an issue	03:58
	i'm not really an expert here though	03:58
	→ dhruvasagar joined	03:58
foucist	just talking out my assk	03:58
	out of my ass*	03:58
fobius	lol. So then from what I gather here the solution is to basically just test after a merge and make sure everything still works proper?	03:58
	Because essentially the issue that I'm talking about is that the code base that you started with	03:59
	isn't the same one that you merge back into	03:59
	← bitkid left	03:59
foucist	fobius: i believe there's a way to have automatic tests that prevent merges if a merge is gonna break stuff	04:00
	← albel727 left	04:00
	← donmichelangelo left	04:00
	→ hagabaka joined	04:00
	→ donmichelangelo joined	04:01
	← druonysus left	04:01
	→ delianides joined	04:01
	← delianides left	04:01
	← ingof left	04:02
fobius	foucist, How would you know if it was gonna break stuff without knowing what the language is like? I mean, I think that in general this sort of thing shouldn't happen unless you have code with bad side effects anyway.	04:02
	→ kermit joined	04:03
	← kermit left	04:03
	→ kermit joined	04:03
	← mribeiro1antas left	04:05
foucist	fobius: well there'd need to be tests basically	04:05
	← funfunctor left	04:05
	→ YannH joined	04:06
fobius	foucist, Yeah, I suppose that makes sense. I think I'm looking for some kind of magical reassurance that auto merge always works out but it's probably not there. I was just wondering if this is a known problem or something that only people who are as OCD as myself are interested in.	04:06
foucist	fobius: i recommend you ask again when serious git people are around	04:07
	← hahuang65 left	04:07
	← shoshin left	04:07
foucist	ask in 12-14 hours from now :P	04:07
fobius	foucist, okay, thanks for your help.	04:07
	lol	04:07
	← fridim_ left	04:08
	→ albel727 joined	04:08
	← fobius left	04:09
	← YannH left	04:12
	→ Katafalkas joined	04:14
	→ hwrd\|work joined	04:14
	← hwrd\|work left	04:14
	← Newtonianb left	04:14
	← atSetKey left	04:16
	→ niggler joined	04:16
	→ axl_ joined	04:17
	→ atSetKey joined	04:17
niggler	hello! suppose i have 5 commits (initial,a,b,c,d) and I want to squash commits a,b,c into one commit. how can i do that with rebase?	04:17
	← natechan left	04:18
relipse	git rebase -i	04:18
	← atSetKey left	04:18
niggler	can you give it an initial and end commit, or do i have to rebase from initial	04:18
	→ breck7_ joined	04:19
	← breck7 left	04:19
	→ fragamus joined	04:20
SamB	niggler: see man git-rebase	04:22
gitinfo	niggler: the git-rebase manpage is available at http://jk.gs/git-rebase.html	04:22
	→ willwh joined	04:22
SamB	niggler: it has pictures and everything	04:22
	→ diegoviola joined	04:22
niggler	nvm my confusion stemmed from the fact that the rebase got rid of the merge commits (which, incidentally, was desired)	04:23
	← MrRoot left	04:23
	→ shoshin joined	04:24
	← daito left	04:24
	← bwreilly left	04:24
	→ daito joined	04:24
	← daito left	04:24
	→ dyng joined	04:24
	← R1KRDO left	04:25
niggler	actually SamB now I'm confused	04:27
	← rpg left	04:28
niggler	i think there was some timezone difference, and git is getting tripped up by it	04:28
SamB	why do you think that?	04:28
	I mean git doesn't usually do much with times	04:28
rjhunter	niggler: you can reset the commit date as part of rebase if you like	04:29
niggler	no the dates are correct, but git is getting confused by the timestamp	04:29
rjhunter	niggler: git itself doesn't care	04:29
SamB	what's it doing wrong?	04:29
	→ epicowl joined	04:30
	← foucist left	04:30
	← dyng left	04:30
niggler	so the history has a few commits	04:30
	in order	04:30
	→ dyng joined	04:31
niggler	Date: Sat May 18 13:55:18 2013 +0200	04:31
	Date: Sat May 18 08:19:17 2013 -0400	04:31
	Date: Sat May 18 15:02:31 2013 +0200	04:31
	Date: Sat May 18 06:08:49 2013 -0700	04:31
	← dyng left	04:31
	→ dyng joined	04:31
niggler	chronologically, they are the correct order	04:32
SamB	yeah?	04:32
niggler	then i rebased	04:32
SamB	and?	04:32
niggler	and the new history has the commits in the order	04:32
SamB	I'm guessing this is related to the merge	04:32
	← willwh left	04:32
niggler	so one of the merge commits was lost	04:33
rjhunter	if you want the commits in a particular order, use rebase --interactive	04:33
niggler	rjhunter i used rebase and tried to squash two unrelated commits	04:33
	→ eletuchy_ joined	04:34
niggler	but it appears that the commits were reordered based on the time without adjusting for time zones	04:34
	Sat May 18 08:19:17 2013 -0400 shows up before Sat May 18 13:55:18 2013 +0200	04:35
SamB	the commits would be in whatever order you left them in in the file you edited	04:35
	→ Smrtz joined	04:35
	← el3ctron_ left	04:36
	→ glosoli joined	04:36
	← eletuchy left	04:36
Smrtz	I'm trying to clone a repo into ~/<gitName>, I would do this with git clone git.github.com/<user>/<gitName> ~/<gitname> right?	04:36
	→ hahuang65 joined	04:37
rjhunter	niggler: i'll have to check to make sure, but i'm pretty sure git doesn't care about dates at all when choosing default order of commits for rebase.	04:37
glosoli	Hey I have a branch on which I did git rebase master, I had so conflicts, so I solved and pushed them (wanted to update it with the latest changes from master). Now history of my branch includes my commits twice (some sort of duplicates with different hashes) one before master commits I rebased and the other duplicates after	04:38
	any ideas ?	04:38
niggler	ok so it appears that the original view (from git log --oneline) is sorted correctly, but the file from git rebase -i is not sorted properly rjhunter	04:38
rjhunter	git doesn't care about dates much at all	04:38
niggler	and the sort is consistent with using the stated date but ignoring the timezone modifier	04:38
	← eletuchy_ left	04:38
rjhunter	niggler: you can move the lines into whatever order you like in the --interactive list	04:39
niggler	rjhunter is there a switch i can pass to rebase so that the sort is consistent with the git log order?	04:39
rjhunter	Smrtz: close -- you'll want a little bit more on your source URL. you can copy it from the Github web page	04:39
	→ bitkid joined	04:39
	→ Robarelli joined	04:40
	← AWinterman left	04:40
rjhunter	glosoli: i'll be with you in a second, but it sounds like your pre-rebase commits and your post-rebase commits both got pushed	04:41
	→ AWinterman joined	04:41
Smrtz	rjhunter: so https://www.github.com/<user>/<gitName> <~/<gitname>?	04:41
	← dyng left	04:41
	← nwertman_ left	04:41
	→ dyng joined	04:41
rjhunter	Smrtz: that looks pretty reasonable to me -- the last part is just the directory you want the results to go in -- if you don't provide one, git makes a new folder under the current directory	04:42
	← yacks left	04:42
Smrtz	rjhunter: thanks.	04:43
	→ longbeach joined	04:43
	→ robbyoconnor joined	04:43
	← joshsmith left	04:44
	← bitkid left	04:44
rjhunter	niggler: I don't see any control over the starting order in the manpage (I'll check the source shortly), but really if you want a specific order then just put them in that order. that's what --interactive is for.	04:44
	→ iamjarvo joined	04:45
	← AWinterman left	04:45
	← dyng left	04:46
	← glosoli left	04:48
	→ xjiujiu joined	04:49
rjhunter	niggler: the starting order for rebase is --topo-order (see man git-revlist for a description what that means)	04:49
	← GT left	04:50
rjhunter	niggler: in short, it's not the dates but the parentage that make the difference to the starting order (again, you can re-order them how you like in --interactive)	04:51
	→ dyng joined	04:51
	← Smrtz left	04:51
	← danillonunes left	04:52
	→ beautifulmind joined	04:52
	← beautifulmind left	04:53
	→ GT joined	04:53
	→ beautifulmind joined	04:53
	← hahuang65 left	04:54
	→ Tommy[D] joined	04:56
	← Tommy[D]_ left	04:56
	→ eletuchy joined	04:56
	← dsdeiz left	04:57
	→ LuxuryMode joined	04:58
	→ ebertx joined	05:01
	← eletuchy left	05:01
	→ hahuang65 joined	05:01
	→ thiago joined	05:02
	← breck7_ left	05:02
	→ hendry joined	05:03
hendry	what's the typical operation to try get the size of git repo down?	05:03
	← woollyams left	05:03
	→ woollyams joined	05:06
	← tango_ left	05:07
	← jwmann left	05:07
	→ rat_thing joined	05:08
	← akurilin left	05:09
	← iamjarvo left	05:09
	→ iamjarvo joined	05:10
	→ tango_ joined	05:11
rjhunter	hendry: du -sh .git	05:12
	→ zbcm joined	05:12
	→ momomomomo joined	05:12
rjhunter	hendry: oops, sorry, i misread your question	05:12
	hendry: `git gc` for garbage collection	05:13
	→ daito joined	05:14
	→ Caritade joined	05:15
rjhunter	hendry: you can use du -sh .git before and after to see the difference :-)	05:16
	← byprdct left	05:17
	← woollyams left	05:17
	→ maestrojed joined	05:18
	← Devastator left	05:19
	→ DonRichie joined	05:19
	→ Devastator joined	05:21
	→ JohnBat26 joined	05:21
	← angelsl left	05:22
	← awc737 left	05:24
	← NanoArrow left	05:24
	← iamjarvo left	05:25
	→ willwh joined	05:25
	← willwh left	05:25
	→ willwh joined	05:25
	← momomomomo left	05:25
	← madewokherd left	05:25
	→ iamjarvo joined	05:25
	→ atSetKey joined	05:27
	→ awc737 joined	05:27
	← rchavik left	05:28
	→ jwmann joined	05:28
	← iamjarvo left	05:30
	← willwh left	05:30
	→ v0n joined	05:30
	→ glosoli joined	05:30
	→ ej joined	05:31
	← atSetKey left	05:32
	← dreiss_ left	05:35
	→ rchavik joined	05:35
	→ functorfun joined	05:35
	functorfun → funfunctor	05:35
	→ chairabanta joined	05:40
	← drodger left	05:41
	→ drodger joined	05:42
	← Caritade left	05:42
	← Robarelli left	05:42
	→ ThomasLocke joined	05:42
	→ youlysse` joined	05:42
	← youlysses left	05:42
	→ eletuchy joined	05:43
	→ Dudalus joined	05:47
	← magu\|q left	05:50
	→ leeN joined	05:50
	← LuxuryMode left	05:50
	→ dreiss_ joined	05:50
	→ magu\|q joined	05:51
	→ AWinterman joined	05:51
	→ LuxuryMode joined	05:51
	← refried_ left	05:52
	→ rath joined	05:52
	→ TomyLobo joined	05:53
	→ lifechamp joined	05:55
	← tpollard left	05:55
	→ druonysus joined	05:56
	→ Zarthus joined	05:56
	← AWinterman left	05:56
	← druonysus left	05:59
	→ ciampix joined	05:59
	← LuxuryMode left	05:59
	→ druonysus joined	05:59
	← chairabanta left	05:59
	← lifechamp left	06:02
	→ fRAiLtY- joined	06:03
	→ lng joined	06:03
	→ momomomomo joined	06:05
	→ soee joined	06:06
	→ papegaaij joined	06:07
	← rjhunter left	06:08
	→ kcj joined	06:08
	← kwmiebach left	06:10
	→ rjhunter joined	06:10
	← ej left	06:10
	← Devastator left	06:10
	→ h0bbit_ joined	06:10
	→ ej joined	06:11
	→ kwmiebach joined	06:11
	← momomomomo left	06:11
	← Zushakon left	06:13
	← ebertx left	06:13
	← eletuchy left	06:13
	→ eletuchy joined	06:13
	→ a-l-e joined	06:14
	← ej left	06:15
	→ cbreak-work joined	06:15
	→ Devastator joined	06:16
	→ jargon joined	06:17
	→ willwh joined	06:17
	← eletuchy left	06:18
	→ beingbrown joined	06:19
	gitinfo set mode: +v	06:20
	→ h0bbit joined	06:20
beingbrown	hello.	06:20
gitinfo	beingbrown: 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.	06:20
	→ zarubin joined	06:22
beingbrown	Is there anyone around that might be able to answer a quick question?	06:22
	→ dsdeiz joined	06:23
	← GT left	06:24
	← beingbrown left	06:25
	← axl_ left	06:25
	← cbreak-work left	06:25
	← jon_r left	06:25
	← willwh left	06:28
	→ cbreak-work joined	06:28
	← dyng left	06:28
	→ dyng joined	06:29
	→ coredumb joined	06:31
	→ GT joined	06:31
	← coredumb left	06:31
	← reim left	06:32
iateadonut	depends on the question	06:32
	but now it's too late, because you left the room and will never see this.	06:33
	← epicowl left	06:33
	→ vurma joined	06:33
Woet	maybe that was his actual question	06:33
	you never know	06:33
	← dyng left	06:33
	→ dyng joined	06:33
	← youlysse` left	06:34
	→ youlysses joined	06:37
	→ angelsl joined	06:37
	→ quasimodo joined	06:37
	← relipse left	06:38
	→ threesome joined	06:40
	← TomyLobo left	06:46
	→ jnavila joined	06:47
	→ ej joined	06:48
	← JohnBat26 left	06:49
	→ rob_w joined	06:50
	← rjhunter left	06:51
	→ netiul joined	06:52
	→ JohnBat26 joined	06:52
	← ej left	06:53
	← threesome left	06:53
	← dyng left	06:53
	→ dyng joined	06:53
	← quasimodo left	06:54
	→ mishok13 joined	06:56
	← dyng left	06:57
	→ MrKeys88 joined	06:58
	→ dsdeiz_ joined	06:58
	← cbreak-work left	06:58
	← dsdeiz left	06:59
	→ kate_r joined	06:59
	→ lite_ joined	06:59
	→ dyng joined	06:59
	→ stamina joined	07:01
	→ marthinal joined	07:03
	→ hoschi joined	07:06
	← beautifulmind left	07:06
	← Devastator left	07:06
	→ h_kon joined	07:08
	← rat_thing left	07:09
	→ GeertV joined	07:10
	→ Xethron joined	07:10
	→ sm0ke joined	07:10
	→ ingof joined	07:11
	← ingof left	07:11
	→ lb1a joined	07:12
	← Goplat left	07:13
	→ Devastator joined	07:13
	→ gua_ould joined	07:13
	→ TomyWork joined	07:14
	→ sr78ger joined	07:14
	→ reim joined	07:15
	→ jon_r joined	07:16
	← jargon left	07:16
	→ dfanjul joined	07:18
	← dfanjul left	07:18
	→ dsdeiz joined	07:18
	→ staafl joined	07:18
	→ dfanjul joined	07:18
	→ Misan joined	07:19
	← dsdeiz_ left	07:19
	← dfanjul left	07:19
	← Norm_ left	07:19
	→ y_nk joined	07:21
	→ threesome joined	07:22
	← japaric left	07:22
	→ NanoArrow joined	07:23
	→ OOPMan joined	07:23
	← m0viefreak left	07:23
	→ kama_ joined	07:25
kama_	hi all	07:27
	why i am now on a nobranch, all changes are commits to nonbranch	07:27
	http://iyanwu.com/pastebin/showthread.php?tid=1828	07:27
	→ kewah joined	07:27
	← zequence_ left	07:29
	→ Robarelli joined	07:29
	→ Balliad joined	07:29
	→ zequence joined	07:30
	→ reskp joined	07:31
lb1a	kama_, because you've checked out a specifig ref	07:32
	specific*	07:32
_ikke_	detached	07:32
lb1a	with bang !	07:32
	^^	07:32
	!detached	07:32
gitinfo	A detached HEAD(aka "no branch") occurs when your HEAD does not point at a branch. New commits will NOT be added to any branch, and can easily be !lost. This can happen if you a) check out a tag, remote tracking branch, or SHA; or b) if you are in a submodule; or you are in the middle of a c) am or d) rebase that is stuck/conflicted. See !reattach	07:32
_ikke_	yeah, trying out my new external tabletkeyboard	07:33
lb1a	hardware keyboard for a tablet?	07:33
_ikke_	yeah	07:33
	← jfinn left	07:33
	← zarubin left	07:33
lb1a	very ambitious	07:33
_ikke_	lol	07:34
	but aparently shift+1 doesn't work	07:34
lb1a	_ikke_, you certainly heard about this new invention... it's called "Laptop"	07:34
	^^	07:34
	→ harobed_ joined	07:34
_ikke_	no?	07:34
lb1a	ok here's the deal: a computer, that is portable	07:35
	and it already has a keyboard attached to it	07:35
	← minim left	07:35
	← awc737 left	07:35
_ikke_	genious	07:35
	brilliiant	07:35
lb1a	and a monitor,too	07:35
_ikke_	no	07:35
lb1a	screen?	07:35
	→ matanya_ joined	07:36
lb1a	and the best: it's a flat screen :P	07:36
kama_	lb1a: seems i should name that non branch at first	07:36
	switch back to master branch, then merge that branch to master branch:)	07:37
lb1a	so throw away this shitty tablet, and run to your next walmart/whatever and buy a decent laptop for only 299$$$ ^^	07:37
	← bruce-zu left	07:37
lb1a	kama_, depending on what you did, that is a good solution	07:37
	→ awc737 joined	07:37
	→ arvind_khadri joined	07:39
	→ Guest95586 joined	07:41
	→ Dreadnaught joined	07:41
	← lite_ left	07:42
	← diegoviola left	07:42
	← Dreadnaught left	07:43
	← robstanley left	07:43
	→ peter5 joined	07:43
	← sm0ke left	07:43
	← maestrojed left	07:44
	← maslowbeer left	07:44
	→ sm0ke joined	07:45
	→ _iron joined	07:46
	← jon_r left	07:46
	← vigs left	07:47
	→ Asenar joined	07:47
	→ chirayuk joined	07:48
	→ jon_r joined	07:48
	← reskp left	07:49
	→ reskp_ joined	07:49
	← sm0ke left	07:53
	← reskp_ left	07:53
	→ starfox21 joined	07:55
	→ sm0ke joined	07:55
	→ lite_ joined	07:57
	→ marhaban joined	07:57
	← starfox21 left	07:58
	← dharriga1 left	07:58
	→ dharriga1 joined	07:58
	dharriga1 → dharrigan	07:58
	← s0undt3ch left	07:59
	→ jfinn joined	08:00
	← rainbyte left	08:00
	← Robarelli left	08:02
	→ kbart joined	08:02
	→ pantsman joined	08:04
	← pantsman left	08:04
	→ pantsman joined	08:04
	→ s0undt3ch joined	08:05
	← kate_r left	08:05
	← weiyang left	08:06
	→ tortal joined	08:07
	← dyng left	08:07
	← Dudalus left	08:08
	→ dyng joined	08:08
	← Katafalkas left	08:09
	→ Katafalkas joined	08:09
	→ Robarelli joined	08:10
	← Katafalkas left	08:10
	→ Katafalkas joined	08:11
	→ nuclearkitten joined	08:11
	← dyng left	08:13
	→ Dudalus joined	08:13
	→ ej joined	08:13
	→ yacks joined	08:13
	← Katafalkas left	08:14
	→ Katafalkas joined	08:14
	→ juril joined	08:15
	→ kenperkins_ joined	08:15
	← Katafalkas left	08:16
	→ Katafalk_ joined	08:16
	← Katafalk_ left	08:16
	→ Katafalkas joined	08:17
	← kenperkins left	08:17
	→ jargon joined	08:17
	→ darkmutt joined	08:20
	← sig left	08:20
	← Katafalkas left	08:20
darkmutt	would someone give me an example of how to test whether or not git push works on a server without any branches on it?	08:20
	→ Katafalkas joined	08:20
	← tortal left	08:21
	→ Sonderblade joined	08:22
	← dreiss_ left	08:22
	← Sonderblade left	08:22
	→ Katafalk_ joined	08:22
	→ Sonderblade joined	08:22
	← Katafalk_ left	08:23
	→ Katafalk_ joined	08:23
	← bunbury left	08:24
	← Katafalk_ left	08:24
	→ Katafalk_ joined	08:24
	→ dyng joined	08:25
	← dyng left	08:25
	← thiago left	08:25
	→ Katafal__ joined	08:26
	→ dyng joined	08:26
	← Katafalkas left	08:26
	← Katafalk_ left	08:26
	→ Katafalkas joined	08:27
	→ sig joined	08:28
	→ Anticom joined	08:28
	→ _Qman joined	08:28
	→ Davey joined	08:28
	← darkmutt left	08:29
	→ howlymowly joined	08:29
	← Katafalkas left	08:29
	→ Katafalkas joined	08:30
	→ stormwild joined	08:31
	→ ewnd9 joined	08:31
	← Katafal__ left	08:31
	← sylon left	08:33
	→ zarubin joined	08:33
	→ aChatir joined	08:34
	Davey → grumbles	08:35
	→ noname001__ joined	08:36
	→ pilif joined	08:38
	→ rat_thing joined	08:38
	→ atSetKey joined	08:41
	← pantsman left	08:41
	← Robarelli left	08:41
	→ el3ctron_ joined	08:41
	← chirayuk left	08:44
	→ amcsi_work joined	08:44
	→ renumi joined	08:44
	← terminalRepeat left	08:46
	→ rudak joined	08:46
	grumbles → Davey	08:49
	moritz can't even think of a way to have a server without any branches on it	08:53
	→ FernandoBasso joined	08:53
_ikke_	lol	08:54
	← Katafalkas left	08:54
	ChanServ set mode: +v	08:54
moritz	well, except a server without any git repos, of course	08:55
	→ tatsuya_o joined	08:55
	→ Katafalkas joined	08:58
	← leeN left	08:58
lb1a	moritz, but all the svn branches having a party :P	08:58
	→ mischief joined	08:59
mischief	hi	08:59
lb1a	mischief, !hi	08:59
gitinfo	mischief: [!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.	08:59
	→ ChrisAnn joined	08:59
mischief	i just made some commits i dont want any more because i was trying to test a pre-commit hook, so i comitted some garbage	09:00
	how can i 'rewind' 2 commits?	09:00
	i didn't push to remote yet	09:00
axrfnu	mischief: git reset --hard HEAD~21	09:00
	err HEAD~1	09:01
	or the like :(	09:01
_ikke_	HEAD~2	09:01
	← sayanee left	09:01
moritz	or just git reset --hard $last_good_commit	09:01
	→ starfox21 joined	09:01
	← noname001__ left	09:01
mischief	HEAD~2 means '2 commits behind the current one'?	09:01
_ikke_	yes	09:02
	the seccond ancestor of HEAD	09:02
mischief	moritz: are you moritz bartl?	09:02
_ikke_	so the parent of the parent of HEAD	09:02
moritz	you can preview with 'git log HEAD~2'	09:02
	mischief: no	09:02
mischief	haha	09:02
	i just deleted my pre-commit hook doing that	09:02
	is it gone, or can i go 'forward' again? :)	09:03
axrfnu	man git-reflog	09:03
gitinfo	the git-reflog manpage is available at http://jk.gs/git-reflog.html	09:03
_ikke_	git reset --hard @{1}	09:04
	was the hook comitted?	09:05
mischief	i did, but then it was gone with the reset	09:05
_ikke_	ok	09:05
mischief	i checked out the commit it was in (it said detached head mode), copied the shell script to tmp, then git checkout master	09:05
	so now i will commit it again since i figured out what was wrong before	09:05
	(the script wasn't marked +x)	09:05
_ikke_	you can also do git checkout <commit> path/to/file	09:06
	← GT left	09:06
	→ rainbyte joined	09:07
	← atSetKey left	09:08
	→ kate_r joined	09:10
mischief	hm	09:10
	is it a good idea to 'git stash' in my pre-commit hook?	09:10
	and pop at the end, obviously	09:10
	← rat_thing left	09:11
_ikke_	what does the hook do?	09:11
	→ Robarelli joined	09:11
mischief	http://sprunge.us/iBBc?sh	09:11
	when i do this, the first command fails	09:12
	but i end up with some junk in my file now from the stash commands	09:12
	git diff after running the hook manually is http://sprunge.us/HCJY	09:13
	→ rat_thing joined	09:13
	Guest95586 → diegoviola	09:14
	→ Norm_ joined	09:15
	← _Qman left	09:15
	→ jaseemabid joined	09:16
	→ rednaks joined	09:17
	→ intellix joined	09:17
	uslss → qwertz_	09:18
	→ lifechamp joined	09:18
	← rurufufuss left	09:18
	← ewnd9 left	09:18
amcsi_work	hi	09:19
mischief	_ikke_: any ideas? :3	09:19
	← zbcm left	09:19
amcsi_work	is it correct that it's into the HTTP protocol that to make an HTTP proxy, you are going to need to dedicate an entire IP and Port to it?	09:20
	hence most implementations that let you use proxies (such as curl) only ask for an IP and Port if you want to use a proxy?	09:20
	← mbear_nbr left	09:20
	→ matteo joined	09:21
	→ dbuch joined	09:21
amcsi_work	wrong channel	09:21
	← G0re left	09:22
	→ shuaib joined	09:22
	← daito left	09:23
	→ daito joined	09:24
	← sqlnoob left	09:25
	→ sqlnoob joined	09:26
	→ pantsman joined	09:26
	→ khushildep joined	09:27
	← daito left	09:28
	→ BiggFREE joined	09:29
	→ mikecmpbll joined	09:29
	← el3ctron_ left	09:29
	← MafooUK left	09:29
mischief	should i just not git stash in my pre-commit hook?	09:30
	← lifechamp left	09:30
	→ MafooUK joined	09:30
	← h_kon left	09:31
osse	is ./... a thing go understands?	09:31
	→ el3ctron joined	09:31
	→ rurufufuss joined	09:31
mischief	yea	09:32
	→ h_kon joined	09:33
	← h_kon left	09:34
osse	mischief: seems strange to supress go's error output and then write your own warning saying "please check the output"	09:34
	→ jkremser joined	09:34
	← jkremser left	09:34
	→ jkremser joined	09:34
mischief	it seems i had a git stash, but no outstanding changes on master	09:35
	→ Katafalk_ joined	09:35
mischief	so 'git stash' was a no-op	09:35
	but git stash pop was popping the old stash	09:35
	thats not what i wanted D:	09:35
	← pantsman left	09:35
	← Katafalkas left	09:35
	← Katafalk_ left	09:35
	→ Katafalkas joined	09:35
mischief	the idea is to only run the tests on code market for commit	09:36
	marked	09:36
osse	I don't see why your pre-commit hook shouldn't work.	09:37
	Except you shouldn't have an extension	09:37
mischief	the file is .git/hooks/pre-commit	09:37
	→ h_kon joined	09:37
mischief	it's a symlink to ../../pre-commit.sh	09:37
	and that file is being committed to the repo once it works	09:38
	→ atSetKey joined	09:38
	← y_nk left	09:39
mischief	osse: the problem is that if there is files staged, and no file is changed but unstaged, and something in the stash, the pre-commit hook will do 'git stash' but thats no-op, then 'git stash pop' which will pop the _old_ stash	09:39
	→ y_nk joined	09:39
	← y_nk left	09:39
	→ y_nk joined	09:39
mischief	so.. i need to only stash pop if stash actually stashed something	09:39
	→ Hetu joined	09:39
mischief	you see?	09:39
	← Davey left	09:41
	→ victorbjelkholm joined	09:41
	← Robarelli left	09:43
	→ glebihan_ joined	09:44
	→ Robarelli joined	09:44
	→ barq joined	09:44
	→ Davey joined	09:44
	→ daito joined	09:46
	→ rethus joined	09:46
	→ cbreak-work joined	09:47
	← atSetKey left	09:47
	← glebihan left	09:47
rethus	I have master, and Branch A, B and C.	09:48
	Say i have changed something in A and finish my work, i merge it to MASTER.	09:48
	Whats now the right way, if i work on B. Should i MERGE master into B, or have i to use REBASE in this case?	09:48
	→ babilen joined	09:48
	← tango_ left	09:48
	← glosoli left	09:49
	→ King_Hual joined	09:50
	← rudak left	09:50
	→ tango_ joined	09:51
	← BiggFREE left	09:51
	← albel727 left	09:52
	← NanoArrow left	09:52
	→ rudak joined	09:52
mischief	stupid git.	09:54
	→ PeterBooker joined	09:55
rethus	mischief ??	09:55
masak	well, that's why it's called "git", isn't it? because it's stupid.	09:56
	manpage even says: "git - the stupid content tracker"	09:56
	→ demiurgo joined	09:57
sonOfRa	git can mean anything!	09:58
	→ mklappstuhl joined	09:59
	→ tvw joined	10:01
	→ irqq joined	10:01
	← rethus left	10:02
	→ albel727 joined	10:02
	← sm0ke left	10:02
	← King_Hual left	10:02
	→ King_Hual joined	10:02
	→ sm0ke joined	10:03
	← rat_thing left	10:05
	← shoshin left	10:05
	→ shoshin joined	10:08
	→ dsdeiz_ joined	10:08
	← dsdeiz left	10:09
	← harish left	10:09
	→ onethfour3 joined	10:11
	← jaseemabid left	10:12
	→ Cornlius joined	10:13
	→ ryan_turner\|MTW joined	10:13
	← multiverse left	10:13
	Cornlius → multiverse	10:13
	← raijin left	10:13
	← wagle left	10:14
	← Davey left	10:14
	← dccc left	10:14
	→ raijin joined	10:14
	→ wagle joined	10:14
	→ dccc joined	10:14
	← cagedwisdom left	10:15
	← onethfour left	10:15
	← kate_r left	10:15
	← zarubin left	10:15
	→ cagedwisdom joined	10:15
	← hyperair left	10:16
	→ PigeonFriend joined	10:17
	→ noname001__ joined	10:20
	← ryan_turner\|MTW left	10:22
	← h0bbit_ left	10:23
	← h0bbit left	10:24
	← FernandoBasso left	10:24
	→ ryan_turner\|MTW joined	10:24
osse	mischief: use git diff to check if there is anything to stash	10:25
	and if there isn't, never pop	10:25
	→ carbohydrate joined	10:26
	→ kate_r joined	10:26
	→ rat_thing joined	10:26
	→ h0bbit_ joined	10:27
	→ gzg joined	10:27
	→ dsdeiz joined	10:28
	← dsdeiz_ left	10:29
	→ the_cybersphinx joined	10:30
	→ mithro-work joined	10:31
mithro-work	is there a way to find out why git thinks my .git directory is no longer a gitdir?	10:31
osse	mischief, beware, contains massive amouns of awesomeness within: http://sprunge.us/IAID	10:31
mischief	o.o	10:32
osse	heh, I forgot to actually git stash	10:32
mischief	yes that would help ^^	10:32
	→ chairabanta joined	10:32
osse	http://sprunge.us/LXVV	10:33
	← gzg left	10:33
Hetu	my local master is two commits ahead of its remote. Can I still make these two commits part of a new branch?	10:33
	← youlysses left	10:33
mischief	osse: so i put this at the top of the script	10:33
osse	mischief: I think that could work, yep. Seems to me you always want to pop the stash right before exiting	10:33
mischief	yes	10:34
	but only if a stash was pushed :P	10:34
	→ DoniRS joined	10:34
osse	which this achieves	10:34
	← mishok13 left	10:34
osse	you might have to change #!/bin/sh to #!/bin/bash	10:34
mischief	oh.. :s	10:35
mithro-work	I've never seen git fail to recognize a .git directory before :/	10:35
	→ mishok13 joined	10:35
osse	mithro-work: maybe some of the GIT_ environment stuff is set	10:36
cmn	the .git directory isn't the issue, it's the contents	10:36
osse	mithro-work: try echo $GIT<tab><tab>	10:36
	← lng left	10:36
mithro-work	"export \| grep GIT" returns nothing	10:36
osse	mischief: but proably not, if sh is a symlink to bash anyway	10:36
mischief	no this is real sh	10:37
cmn	what's in that .git dir then?	10:37
	← starfox21 left	10:37
	→ slassh joined	10:37
mischief	osse: i use git on several systems. openbsd and linux, mainly	10:37
	← rednaks left	10:37
mischief	but this seems to work on openbsd sh	10:37
	→ Guest83628 joined	10:38
osse	mischief: isn't sh a symlink to zsh there? :P	10:38
mithro-work	It looks like a valid .git repository to me. I was happily working away on this repository and my machine power cycled. Now git doesn't think its a git repository and won't tell me why	10:38
mischief	osse: no ._.	10:38
cmn	strace would tell you	10:38
osse	mischief: good news: trapping the "signal" EXIT is POSIX sh blabla. So it'll probably work everywhere	10:39
mithro-work	Last thing it opens is .git/HEAD then it says it's not a git repository	10:40
	→ maslowbeer joined	10:40
	→ AzizLight joined	10:41
cmn	so what did you find out about it?	10:41
	← AzizLight left	10:42
mithro-work	ahh ha, I tried replacing HEAD's content with a value from refs/heads/ and it seems to recognise it now	10:42
mischief	osse: do you think i should add --keep-index to stash?	10:42
osse	mischief: oh, definitely. I forgot about that	10:42
mischief	ok	10:42
osse	mischief: use the same git stash commands you already use	10:42
mischief	well, the idea is to test staged files	10:43
	without --keep-index staged files would be stahed	10:43
	→ atSetKey joined	10:43
	→ AzizLight joined	10:44
	← kewah left	10:44
AzizLight	Hello everybody	10:44
	→ _nova joined	10:45
AzizLight	is there a command to add all files to the staging area and remove the ones that have been deleted please? (Essentially a combination of git add and git rm)	10:45
	→ jaseemabid joined	10:45
	← marhaban left	10:45
mischief	git do-all-the-things	10:46
	← chairabanta left	10:46
osse	AzizLight: add -A	10:46
	← maslowbeer left	10:46
	← Robarelli left	10:47
	← daito left	10:47
osse	mischief: http://sprunge.us/FGCh	10:47
	→ dsdeiz_ joined	10:47
	→ daito joined	10:47
AzizLight	osse: cool, thanks a lot ;)	10:48
	← dsdeiz left	10:48
	← atSetKey left	10:48
osse	mischief: but I would reconsider silencing stderr	10:48
	← kc8qvp_ left	10:48
	← rat_thing left	10:48
osse	ie. dropping 2>&1	10:48
	← dyng left	10:49
	→ dyng joined	10:50
	← mklappstuhl left	10:51
	← aChatir left	10:51
mischief	osse: i'll see how this goes	10:52
	← daito left	10:52
mischief	https://github.com/PacketFire/go-ircd	10:52
	thanks again, now we wont have fail code committed (hopefully)	10:52
	← _nova left	10:53
	→ _nova joined	10:54
	← grilix left	10:54
	→ dyng_ joined	10:54
	→ imarcusthis- joined	10:54
	← dyng left	10:55
	→ lng joined	10:56
	← dyng_ left	10:56
	→ dyng joined	10:56
	→ grilix joined	10:58
osse	no credit to me? :( Hee hee	10:59
	→ starfox21 joined	10:59
	← donmichelangelo left	11:00
	← dong left	11:00
mischief	osse: sorry	11:00
	osse: file an issue >:)	11:01
	→ donmichelangelo joined	11:01
	← dyng left	11:01
	→ ewnd9 joined	11:02
	← Xethron left	11:02
	← kcj left	11:02
	← rudak left	11:02
	← slassh left	11:05
	← dbuch left	11:06
	← diegoviola left	11:06
	→ Katafalk_ joined	11:06
	← Katafalkas left	11:07
	→ dsdeiz joined	11:08
	← dsdeiz_ left	11:09
	← cbreak-work left	11:12
	← stormwild left	11:12
	→ stormwild joined	11:12
	← frgtn left	11:12
	→ haukur joined	11:12
	→ frgtn joined	11:13
haukur	I have a master and a dev branch. I wanted to merge dev into master, as I've usually done, but after doing so git diff master (from dev) shows that there's still a difference between the two branches (one simple line).	11:14
	How can I fix this?	11:15
arand__	was that line added/modified in dev or in master?	11:16
	→ kewah joined	11:17
	→ CheckDavid joined	11:17
	← frgtn left	11:17
haukur	I'm pretty sure it was modified in dev, at least it should've been. One second.	11:17
	→ vilesoft joined	11:18
	← vilesoft left	11:18
	→ asQuirreL joined	11:20
	→ Katafalkas joined	11:21
	→ D-Boy joined	11:21
	→ kirsty joined	11:22
	→ cmpera joined	11:24
	← Katafalk_ left	11:24
	→ hpm joined	11:27
	← TomyWork left	11:28
	← mikecmpbll left	11:29
	→ raonyguimaraes joined	11:29
	→ twoface88 joined	11:31
	→ Ropt joined	11:31
Ropt	Is this the appropriate channel to ask questions about GitLab?	11:32
	← haukur left	11:32
Ropt	Oops I found the #gitlab channel, sorry ignore my question please	11:32
	→ Droolio joined	11:32
	← hpm left	11:35
	← sdx23 left	11:35
	→ tmcmahon joined	11:35
	← nemesit\|znc left	11:35
	→ Roots47 joined	11:35
	← luc__ left	11:36
	← Roots47 left	11:37
	← Happzz left	11:37
	← King_Hual left	11:38
	← farn left	11:38
	→ Roots47 joined	11:38
	→ King_Hual joined	11:38
	← Cerise left	11:40
	← carbohydrate left	11:41
	→ HenryG joined	11:42
	→ mklappstuhl joined	11:43
	→ atSetKey joined	11:43
	→ LuxuryMode joined	11:44
	← ojacobson left	11:45
	→ cannonball joined	11:46
	→ VictorCL joined	11:46
	→ bunbury joined	11:47
	→ dbuch joined	11:47
	← tmcmahon left	11:47
	← irqq left	11:47
	→ ThomasCle joined	11:48
	→ vipzrx joined	11:48
	← atSetKey left	11:48
	← twoface88 left	11:49
	→ NanoArrow joined	11:49
	← dhruvasagar left	11:50
	← jaseemabid left	11:50
	→ kcj joined	11:51
	← merlin83_ left	11:51
	→ irqq joined	11:52
Hetu	hi, can I enforce a three way merge when my feature branch is just ahead of master? I don't want the feature commits to get rebased into master.	11:52
_ikke_	git merge --no-ff	11:53
	a fast-forward merge doesnt rebase anything	11:53
	→ dsdeiz_ joined	11:53
	← dsdeiz left	11:54
Hetu	thanks a lot	11:54
	right, by "rebase" I probably meant them each and every commit being merged into master	11:55
	← NanoArrow left	11:55
	← vipzrx left	11:55
	→ Happzz joined	11:57
	← LuxuryMode left	11:57
	→ nemesit\|znc joined	11:57
	→ farn joined	11:57
	→ Cerise joined	11:58
	→ beautifulmind joined	11:58
	→ dhruvasa1ar joined	11:58
_ikke_	The only thing that happens is that a branch (a pointer) gets moved	11:58
	→ frgtn joined	11:59
	← stamina left	11:59
	→ sdx23 joined	11:59
	← beautifulmind left	11:59
	→ beautifulmind joined	12:00
	→ gzg joined	12:00
	← Katafalkas left	12:01
	← h_kon left	12:01
	→ Katafalkas joined	12:01
	→ h_kon joined	12:02
	→ vipzrx joined	12:02
	← dhruvasa1ar left	12:02
	← Katafalkas left	12:02
	← jon_r left	12:03
	→ dhruvasagar joined	12:03
	→ blackjack joined	12:04
	← blackjack left	12:04
	→ Kulrak joined	12:05
	→ blackjack joined	12:05
	← blackjack left	12:05
	[emmajane] → emmajane	12:05
	→ blackjack joined	12:06
	← blackjack left	12:06
	← dsdeiz_ left	12:06
	→ dsdeiz joined	12:06
	→ diegoviola joined	12:08
	← asQuirreL left	12:09
	← nuclearkitten left	12:10
	← h_kon left	12:10
	→ ozette joined	12:11
	→ h_kon joined	12:11
	→ ojacobson joined	12:11
	→ jon_r joined	12:12
	← Entomo left	12:12
	← mklappstuhl left	12:12
	→ mklappstuhl joined	12:12
	→ Man joined	12:13
	→ panpainter joined	12:13
	← v0n left	12:14
	← stormwild left	12:14
	← CodeShark left	12:16
	→ mklappst_ joined	12:16
	← mklappstuhl left	12:17
	→ CodeShark joined	12:17
	← CodeShark left	12:18
	← mklappst_ left	12:20
	→ Jackneill joined	12:20
	→ rethus joined	12:20
osse	mischief: no worries. it probably doesn't work anyway	12:20
rethus	have this error:	12:20
	→ mklappstuhl joined	12:21
rethus	git show	12:21
	error: object file .git/objects/6b/fecc157959895b732e237f276dc7a84142a27f is empty	12:21
	fatal: loose object 6bfecc157959895b732e237f276dc7a84142a27f (stored in .git/objects/6b/fecc157959895b732e237f276dc7a84142a27f) is cor	12:21
	rupt	12:21
	what can i do to fix it	12:21
	← grilix left	12:22
	→ breakingthings joined	12:22
	→ grilix joined	12:23
	→ reskp joined	12:24
	→ iveqy joined	12:25
	→ bitkiller joined	12:25
	← mklappstuhl left	12:25
	← jon_r left	12:25
	← rfoust left	12:26
hakril	have you pushed this object ?	12:26
	← grilix left	12:27
	← Kulrak left	12:27
	→ g0bl1n joined	12:28
	← kenperkins_ left	12:28
hakril	(if you have no idea, you can verif this by cloning again your repos and using 'git rev-parse 6bfecc157959895b732e237f276dc7a84142a27f')	12:28
	← aspiers left	12:28
	→ mklappstuhl joined	12:28
	→ aspiers joined	12:29
	→ rat_thing joined	12:29
	← Devastator left	12:29
	→ Devastator joined	12:29
	→ Saurabh_ joined	12:31
	← vipzrx left	12:32
rethus	hakril: the horrible true is, that this is only a local repo	12:32
	→ iamjarvo joined	12:32
	→ freeed joined	12:32
rethus	no copies at all	12:32
	→ m0viefreak joined	12:32
	→ Squarism joined	12:33
rethus	before i had a OS freeze and to restart (without shutdown)	12:33
	← sdx23 left	12:35
	→ kenperkins joined	12:36
iveqy	rethus: you do know that your OS doesn't write things to disc instantly and when you save something it's only i RAM and will be lost forever if you force a reboot or if the system freeze	12:36
	→ carbohydrate joined	12:36
rethus	iveqy yes. but this answer not help in any direction.	12:37
	← Saurabh_ left	12:38
rethus	I didn't ask why this happend, i ask how to fix it to got git back to work	12:38
	← nemesit\|znc left	12:38
	← freeed left	12:38
iveqy	rethus: I was disconnected so I couldn't read your question. However a freeze can be fatal	12:38
	at least git let you know about it	12:38
	→ nemesit\|znc joined	12:40
	← Milossh left	12:41
rethus	error: object file ./objects/6a/b1aeda549cdd998b4069f6d40f6ad5063164c9 is empty	12:42
	fatal: loose object 6ab1aeda549cdd998b4069f6d40f6ad5063164c9 (stored in ./objects/6a/b1aeda549cdd998b4069f6d40f6ad5063164c9) is corrupt	12:42
	what if i only remove this file	12:42
	→ freeed joined	12:42
rethus	or fill it with content of another one?	12:42
	← ciampix left	12:42
	→ sdx23 joined	12:42
iveqy	rethus: have you runned git fsck and git gc?	12:42
rethus	thats the result of git fsck	12:43
	→ EdGruberman joined	12:44
	→ atSetKey joined	12:44
	→ nnyk_ joined	12:44
	← EdGrubemran left	12:45
	→ rnsanchez joined	12:46
	→ Milossh joined	12:46
	→ nova_ joined	12:46
	→ freestyl3r joined	12:47
	← sony left	12:47
	← rethus left	12:48
	→ glosoli joined	12:48
	← atSetKey left	12:49
	← _nova left	12:49
	→ vipzrx joined	12:49
	← vipzrx left	12:49
	→ Quanteek joined	12:50
	← dsdeiz left	12:50
	→ rethus joined	12:50
	← Quanteek left	12:50
rethus	I have now removed 4 corrupted objects and run git fsck --full	12:50
	than i got:	12:51
	→ vipzrx joined	12:51
rethus	Checking object directories: 100% (256/256), done.	12:51
	error: HEAD: invalid sha1 pointer 6bfecc157959895b732e237f276dc7a84142a27f	12:51
	error: refs/heads/master does not point to a valid object!	12:51
	error: 6ab1aeda549cdd998b4069f6d40f6ad5063164c9: invalid sha1 pointer in cache-tree	12:51
	← vipzrx left	12:51
	← sm0ke left	12:51
	→ Robarelli joined	12:51
	← tvw left	12:51
	← iamjarvo left	12:52
	← freeed left	12:52
	→ iamjarvo joined	12:52
	→ tvw joined	12:53
	→ TheBrainWork joined	12:54
	→ dsdeiz joined	12:54
	← haaja left	12:54
	→ haaja joined	12:55
	→ HumanSky joined	12:55
	→ kmctown joined	12:56
	→ freeed joined	12:56
iveqy	rethus: I would say, move it manually	12:56
rethus	to?	12:56
	from?	12:56
iveqy	rethus: to the latest commit you've that's still there	12:56
	seems like you just removed your latest commit	12:57
	→ nuclearkitten joined	12:57
	→ Burer joined	12:57
	← Burer left	12:57
	→ Burer joined	12:57
	→ mikecmpbll joined	12:57
rethus	don't see what you mean. What should i move	12:57
	→ rfoust joined	12:58
rethus	if i remember right, last thing i've done was a rebase	12:58
	→ DanielKarp joined	12:58
iveqy	a branch is a pointer to a commit	12:59
	your branch master points to a commit you've removed	12:59
	you need to point it to acommit that exists	12:59
	→ grilix joined	13:00
rethus	thats the last entry in my log:	13:01
	1d5dcf311b96fcbfd27d0ad017854d9e2876a0e8 6bfecc157959895b732e237f276dc7a84142a27f Samuel Suther [email@hidden.address] 1381399067 +0200 merge refs/heads/#66Buttonbeschriftung: Merge made by recursive.	13:01
	→ psynor joined	13:01
rethus	exactly my broken one ... 6bfecc....	13:02
	so how can i now set a new pointer	13:02
	← freeed left	13:02
	→ rjhunter joined	13:03
	→ delianides joined	13:03
iveqy	rethus: man git reset but in this case I would have just edited the file .git/refs/heads/master	13:03
gitinfo	rethus: the git-reset manpage is available at http://jk.gs/git-reset.html	13:03
	← shoshin left	13:04
rethus	.git/refs/heads is emty	13:04
	← hoschi left	13:04
rethus	no file in there	13:04
	→ freeed joined	13:05
	← sr78ger left	13:05
	← raijin left	13:06
	→ jerius joined	13:06
rethus	k, got thin now: http://pastebin.com/KJ0xHc7f	13:06
	← nessib left	13:07
	→ Xethron joined	13:07
iveqy	no files at all? What's in your .git/refs?	13:08
rethus	folders: heads, stash tags	13:08
	→ jaseemabid joined	13:09
rethus	whats with my pastebin, isnt it a step forward?	13:09
jast	if the branch doesn't exist in .git/refs/heads/, it probably lives in .git/packed-refs instead	13:09
	(which is a single file)	13:10
	→ iderik joined	13:10
rethus	http://pastebin.com/sVrFF5mS	13:11
	thats the content of the file	13:11
	→ decotii joined	13:11
iveqy	jast: I thought the current branch never should be packed?	13:12
	rethus: there you can see refs/heads/master point to 6bfecc157959895b732e237f276dc7a84142a27f	13:12
	← nuclearkitten left	13:12
rethus	so should i replace it by 1d5dcf311b96fcbfd27d0ad017854d9e2876a0e8	13:13
iveqy	rethus: I'm not sure with that you should replace it with, do a backup and try	13:13
barq	I changed .gitignore to ignore *.class files, but git status still shows them as modified. How can I refresh?	13:14
	→ nuclearkitten joined	13:14
	→ buibex joined	13:14
	→ geekbri joined	13:15
axrfnu	barq: git rm --cached on the *.class files	13:15
	barq: that'll remove them from the indexd	13:16
	*index	13:16
jast	barq: !ignoretracked	13:16
gitinfo	barq: [!ignore_tracked] Git only applies ignore patterns to untracked files. You can't use ignore patterns to ignore changes to files that are already tracked by git. Still, see https://gist.github.com/1423106 for ways people have worked around the problem.	13:16
jast	if you want to stop tracking the files, run 'git rm --cached' on them	13:16
	← iderik left	13:16
rethus	iveqy: thanks a lot. You save my day.	13:17
	seems to work now	13:18
	← ThomasLocke left	13:18
	→ GODDOG joined	13:18
rethus	so what's a good way to got even local git repos saved?	13:18
	→ shoshin joined	13:19
	→ YannH joined	13:19
rethus	I have a shop-project local, but no git on the Live-Server. So should i anyway copy the .git-Folder to the Server to be save?	13:19
	→ v0n joined	13:20
	← iamjarvo left	13:20
	→ iamjarvo joined	13:21
	← grilix left	13:21
	→ DolphinDream joined	13:21
	← matanya_ left	13:21
	→ Guest26260 joined	13:22
	← Fandekasp left	13:22
	← sdx23 left	13:23
	→ sdx32 joined	13:23
	→ Wilduck joined	13:24
	← Wilduck left	13:24
	→ rethus1 joined	13:24
	→ momomomomo joined	13:24
	← threesome left	13:24
	← rat_thing left	13:24
	← diegoviola left	13:24
barq	How can I rm --cached everything that's listed? rm --cached * had rather undesirable results.	13:24
	← Robarelli left	13:25
	← momomomomo left	13:25
	← iamjarvo left	13:25
sonOfRa	listed?	13:26
	← panpainter left	13:26
barq	listed did not match any files	13:26
	→ Robarelli joined	13:26
	→ tbmiller joined	13:27
	→ m0ltar joined	13:27
	← rethus left	13:27
	→ rat_thing joined	13:27
	← nemesit\|znc left	13:27
	→ nemesit\|znc joined	13:27
iveqy	rethus1: nice to hear =). I suggest you clone your local repo and work there instead. A corrupt git repo won't care any corruption over a clone	13:28
	← tbmiller left	13:29
iveqy	rethus1: the best way for a backup IMHO is to have a remote repo to push to somewhere (bitbucket gives you one for free) however there's always git clone --mirror	13:29
	→ tbmiller joined	13:29
	← tbmiller left	13:30
	→ \|Freeman\| joined	13:30
	→ tbmiller joined	13:30
	→ Dougie187 joined	13:30
	← rethus1 left	13:31
	→ rethus1 joined	13:31
	← rob_w left	13:31
	→ axl_ joined	13:32
	→ LuxuryMode joined	13:33
	→ AaronMT joined	13:33
	← rethus1 left	13:34
	→ raijin joined	13:34
	← lb1a left	13:34
	→ rethus joined	13:34
rethus	ok. good idea. for first short solution, i could clone the repo to another hdd, and use this one as "kind of" remote repo?	13:34
	→ tdussa joined	13:34
	→ apipkin joined	13:34
	← Arokh left	13:35
	→ cbreak-work joined	13:35
	→ fridim_ joined	13:35
	← h_kon left	13:35
	gitinfo set mode: +v	13:35
	Guest26260 → diegoviola	13:35
	← Guest83628 left	13:35
tdussa	Does anyone have any experience with git bundles?	13:36
	I seem to be unable to figure out how to prune a bundle by date properly.	13:36
moritz	99	13:37
	→ LuxuryMo_ joined	13:37
	← Xiti left	13:37
moritz	sorry, keyboard hung	13:37
	← LuxuryMode left	13:37
	→ Xiti joined	13:38
	→ lumberg joined	13:38
	→ kukks joined	13:38
	← kate_r left	13:38
barq	I am trying to git rm --cached on a .metadata/ directory, but git says: did not match any files. Is there a problem with .dirs?	13:39
	← lumberg left	13:39
	→ Arokh joined	13:39
	← AaronMT left	13:39
	→ frio80 joined	13:39
jast	barq: use -r	13:40
barq	I have	13:40
jast	(with --cached)	13:40
	→ yahoololman joined	13:40
jast	well, you didn't say :}	13:40
	← HumanSky left	13:40
barq	git rm -r --cached path/.metadata/	13:40
	← v0n left	13:40
jast	and that complains about not matching anything?	13:41
	← apipkin left	13:41
barq	.metadata did not match any files	13:41
jast	is this still the same question? i.e. does that dir contain files that currently show up as 'modified'?	13:41
	← jaseemabid left	13:41
barq	it shows as untracked files	13:41
	→ Guest32557 joined	13:42
jast	oh. well, you can't untrack files if they're already untracked ;)	13:42
moritz	'git rm' is only for tracked files	13:42
barq	oh, so manually add to .gitignore then -.-	13:42
jast	'git rm --cached' and .gitignore are related but different	13:42
moritz	use regular 'rm' to delete untracked files	13:42
tdussa	Rephrasing my problem: Am I mistaken to believe that a "git bundle create" command should only include the commits that, say, a "git log" command would show if the same limitations are applied?	13:42
	← Xiti left	13:42
	→ Voting joined	13:42
	← lite_ left	13:42
	← sdx32 left	13:43
	← el3ctron left	13:43
	→ lite_ joined	13:43
tdussa	So if I say "git log --since 2013-10-01", and it shows, say, one revision, then shouldn't that be all that is included in a "git bundle create Foo --all --since 2013-10-01"?	13:43
	→ sdx23 joined	13:43
jast	.gitignore is for stopping untracked files from being listed/added; 'git rm --cached' is for getting rid of the 'tracked' status.	13:43
PerlJam	tdussa: sounds right.	13:44
barq	git rm -r doesn't find the dir either	13:44
	→ stamina joined	13:44
Voting	So, after I fetch from something that is NOT my origin, how do I do the merge? I want to something like "git merge -Xtheirs dev" but it was dev from the repo that I just did the fetch from that I want to do the merge. Is the question clear?	13:44
jast	barq: if it's all untracked, that makes sense	13:44
tdussa	PerlJam: That's what I thought, but the bundle includes way more.	13:44
PerlJam	barq: git also doesn't track dirs per se	13:44
jast	tdussa: compare to 'git log --all --since ...', not 'git log --since ...'	13:44
tdussa	Ah.	13:44
	jast: THX.	13:44
jast	and if you don't want --all, don't use --all ;)	13:44
tdussa	Good point, of course.	13:44
	→ HumanSky joined	13:44
	→ AaronMT joined	13:44
	→ atSetKey joined	13:44
	→ el3ctron_ joined	13:44
tdussa	jast: Same effect, though.	13:45
frio80	A colleague asked my to rebase my commits into a single commit and open a pull request so he can merge my changes. Of my 3 commits, the 1st one is simply merged a pull request from his repo to 'stay up to date'. Do I need to squash all those commit as well?	13:45
jast	Voting: have you added a remote for the other repo? (i.e. git remote add foo ...)	13:45
	← kermit left	13:45
	→ Destos joined	13:45
Voting	jast: do I need to? I did the fetch already.	13:45
barq	So how can i get rid of those untracked files in the status?	13:45
PerlJam	barq: .gitignore :)	13:45
	← cwillu_at_work left	13:46
jast	Voting: so you did a 'git fetch <url>'?	13:46
	← diegoviola left	13:46
Voting	PerlJam: any other options beside .gitignore ?	13:46
	jast: yes	13:46
jast	Voting: he wasn't talking to you, actually :)	13:46
barq	Is it good practice to gitignore .gitignore?	13:46
Voting	jast: I know but I have junk in my git status also...	13:47
jast	barq: not really. .gitignore is for shared ignore patterns. if you want extra ignore patterns for yourself, put them in .git/info/exclude	13:47
barq	ok, thanks	13:47
	→ Xiti joined	13:47
jast	Voting: the problem with fetching 'anonymously' is that it fetches all branches and stores them in FETCH_HEAD; you can't really merge a specific of those branches then	13:47
PerlJam	Voting: why do you want something other than .gitignore ? IS there something wrong with that solution?	13:47
	← yahoololman left	13:48
jast	if you're okay with a bit of manual work, do a 'cat .git/FETCH_HEAD', find the right branch, and merge by commit ID	13:48
	← lite_ left	13:48
Voting	PerlJam: .git/info/exclude does look like the right thing for just junk...	13:48
jast	in general I recommend creating a remote, fetching that, and then 'git merge theremote/thebranch'	13:48
Voting	jast: I can find the ID by looking at the source repo that I pulled from, right?	13:49
jast	sure	13:49
PerlJam	Voting: well, "junk to you, for this repo". If it's "just junk" and you think everyone using your repo will have the same idea of "junk" .gitignore is a good option	13:49
	← harobed_ left	13:49
jast	and FETCH_HEAD knows it, too, until you fetch something else :)	13:49
	← atSetKey left	13:49
PerlJam	Voting: adding another remote like jast says is way easier to deal with. (fwiw)	13:50
	→ harobed_ joined	13:50
jast	for one-off merges there's nothing wrong with doing it the less easy way if it saves a bit of time	13:50
	but the next time you do something like this, creating a remote will probably be easier	13:50
Voting	jast: so, assuming my comit is 4321, what do I type to kick off the merge?	13:50
jast	'git merge 4321'	13:50
Voting	jast: git merge 4321 -Xtheirs ?	13:51
	→ panpainter joined	13:52
Voting	jast: do I put in the branch somewhere?	13:52
PerlJam	one-off merges are like one-off code; you almost always need to use it more than once. (IMHO)	13:52
	→ bamj0rb joined	13:52
jast	no need. merge doesn't care about the source branch, it simply resolves the branch to its topmost commit	13:52
	gitinfo set mode: +v	13:53
	← AaronMT left	13:53
jast	and the target is always the branch you have checked out, so no name needed for that, either	13:53
Voting	jast: I'm grateful for your support!	13:53
jast	you may want to mention the source in your merge commit message, though	13:53
	for posterity	13:53
	you're welcome :)	13:54
	→ NanoArrow joined	13:54
	← JohnBat26 left	13:55
	→ grilix joined	13:55
	→ marhaban joined	13:55
	← Xethron left	13:55
Voting	jast: you rule! So much easier than I was GOING to do.	13:56
	← marhaban left	13:56
jast	's all a matter of experience	13:57
	and you get a lot of experience if you hang around here :)	13:57
	(and actually do stuff yourself from time to time, I suppose)	13:57
	→ kate_r joined	13:57
	→ Wilduck joined	13:58
	← soee left	13:58
	→ saurabh joined	13:58
	← Robarelli left	13:59
	→ mizu_no_oto joined	13:59
	← donmichelangelo left	14:00
	← VictorCL left	14:00
	→ juvenal_ joined	14:00
	→ donmichelangelo joined	14:00
	→ mFacenet joined	14:02
	← juvenal left	14:02
	juvenal_ → juvenal	14:02
	← buibex left	14:04
	→ aChatir joined	14:04
	→ buibex joined	14:04
	→ herdingcat joined	14:05
	← ThomasCle left	14:05
Voting	this is a great room! Now, what is easier when I set up more "remotes" than just the one I cloned off of?	14:05
babilen	Voting: It's called "channel" on IRC, but yeah :)	14:05
	→ Guest54977 joined	14:06
	← sqlnoob left	14:06
	← fridim_ left	14:07
	→ buibex_ joined	14:08
	← jnavila left	14:08
	→ jnavila_ joined	14:08
	← irqq left	14:08
	← kewah left	14:08
	← buibex left	14:09
	← nnyk_ left	14:09
	→ ThomasCle joined	14:09
	→ marhaban joined	14:09
	← mizu_no_oto left	14:09
	← comjf left	14:10
	→ rob_w joined	14:10
	← robbyoconnor left	14:10
	→ iamjarvo joined	14:10
	→ rpg joined	14:11
	← glosoli left	14:11
	→ mizu_no_oto joined	14:12
	← beautifulmind left	14:13
	← fragamus left	14:14
	→ asQuirreL joined	14:14
Voting	babilen: lol - I knew that! I'm just not that careful with my words... What are the best books and websites to really understand git so I'll be more efficient? I LOVE http://marklodato.github.io/visual-git-guide/index-en.html	14:14
	What else should I be reading?	14:14
	→ cwillu_at_work joined	14:15
jast	depends on what kind of approach you like	14:15
	!doc	14:15
gitinfo	A list of useful documentation of all kinds is here: http://git-scm.com/documentation -- or try the wiki at http://git.wiki.kernel.org/. Also try typing "!book" "!cs" "!bottomup" "!parable" "!best_practices" or "!vcbe" or "!designers" here in IRC. !book is probably the most helpful.	14:15
jast	I like !gcs for an explanation of all the data structures and implications from that	14:15
gitinfo	[!concepts] "Git Concepts Simplified" explains the basic structures used by git, which is very helpful for understanding its concepts. http://gitolite.com/gcs/	14:15
Voting	jast: I am not clueful enough to have an "approach" that I am aware of. (Beside "keep it simple")	14:15
	← Guest54977 left	14:16
Voting	jast: what are the possible "approaches" one could choose among?	14:17
	← funfunctor left	14:18
	← NanoArrow left	14:18
jast	well, I mean... are you a computer scientist or mathy person? we have something for that. are you not much of a developer at all? we have something for that. do you think you might benefit from seeing the building blocks of git? we have something for that. etc.	14:18
	← frio80 left	14:18
	→ Robarelli joined	14:19
	← Roots47 left	14:19
	← saurabh left	14:20
	← asQuirreL left	14:20
	→ phile joined	14:21
	→ wildfire joined	14:21
	← druonysus left	14:21
	← juril left	14:21
	→ ehmatthes joined	14:21
wildfire	Hi, is there a way to mark a local branch as something that should not be pushed to a remote location?	14:21
	→ v0n joined	14:22
	→ tang^ joined	14:22
lov	no, I don't believe so.	14:23
	but, wait for the rest of the brain trust to wake up :)	14:23
	→ bklane joined	14:23
	← ezakimak_work left	14:23
	← papegaaij left	14:24
	→ merlin83_ joined	14:24
	← merlin83_ left	14:24
	→ merlin83_ joined	14:24
	← carbohydrate left	14:26
	→ hholtmann joined	14:26
	→ kewah joined	14:26
	→ Swayz joined	14:28
canton7	wildfire, no, there isn't. if you look at push.default in man git-config, there are options that stop all branches being pushed automatically, so a branch won't be pushed unless you explicitely push it	14:28
gitinfo	wildfire: the git-config manpage is available at http://jk.gs/git-config.html	14:28
iveqy	canton7: only tracking branches will be pushed anyway	14:28
	so yes, a normal branch won't be pushed. or do I miss something big here?	14:29
canton7	wildfire, ooh, no I remember something. You can set branch.<name>.pushremote to "Don't push here you numpty" or something I seem to recall? That'll cause pushes to fail	14:29
	← marhaban left	14:29
	← OOPMan left	14:29
canton7	iveqy, I was assuming that this might be a branch which does exist on the remote, but which he doesn't want to push	14:29
babilen	Voting: I liked http://git-scm.com/book	14:29
	← barq left	14:29
canton7	iveqy, well, by default (on older git versions) matching branches are pushed - not tracking branches.	14:30
Voting	babilen: Tx! Anything else?	14:30
canton7	!doc !book	14:30
gitinfo	A list of useful documentation of all kinds is here: http://git-scm.com/documentation -- or try the wiki at http://git.wiki.kernel.org/. Also try typing "!book" "!cs" "!bottomup" "!parable" "!best_practices" or "!vcbe" or "!designers" here in IRC. !book is probably the most helpful.	14:30
	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	14:30
canton7	I especially liked bottomup	14:30
	→ rainbyte16 joined	14:30
	→ aChatir1 joined	14:31
iveqy	!talks (I like Jessicas)	14:31
gitinfo	Some good video talks about Git: [yt] http://goo.gl/z72s (Linus Torvalds: History&Concepts); [yt] http://goo.gl/R9H2q (Scott Chacon: Git basics, live examples); http://vimeo.com/35778382 (Randal Schwartz: Git basics, descriptional); http://vimeo.com/46010208 (Jesica Kerr: Git basics, descriptional)	14:31
	→ boboc joined	14:31
	← Man left	14:31
Voting	thanks everyone!!!	14:31
	→ asQuirreL joined	14:31
	→ wrouesnel joined	14:32
	→ petey joined	14:32
canton7	(!bottomup)	14:32
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 (http://ikke.info/git.from.bottom.up.pdf)	14:32
	← aChatir left	14:32
	← xjiujiu left	14:32
	← herdingcat left	14:33
Voting	perhaps I know enough now to read that gbo thing!	14:33
	→ AaronMT joined	14:34
	← onethfour3 left	14:34
babilen	Hmm, why isn't the bot answering to private requests? I tried "/msg gitinfo parable" and got nothing. Do I have to use it in the channel and spam everybody else?	14:34
	→ onethfour joined	14:35
canton7	babilen, you forgot the !	14:35
lov	10:35:12 -!- Irssi: Starting query in freenode with gitinfo	14:35
	10:35:12 <lov> !parable	14:35
	10:35:12 -gitinfo(jast@infobot.gitrecipes.de)- 'The git parable' provides some good reasoning behind git. http://tom.preston-werner.com/2009/05/19/the-git-parable.html	14:35
gitinfo	'The git parable' provides some good reasoning behind git. http://tom.preston-werner.com/2009/05/19/the-git-parable.html	14:35
onethfour	I have a bunch of modified files. How do i return my branch to the state of the last commit	14:35
lov	yeah, don't forget the !	14:35
	← Dudalus left	14:35
	qsusOFFcs → qsuscs	14:35
canton7	onethfour, 'git status' will tell you	14:35
	← kcj left	14:35
babilen	canton7: That's the address character in the channel .. Why would you include that in the factoid names? But thanks, "/msg gitinfo !parable" works. Strange decision though ...	14:35
gitinfo	canton7: 'The git parable' provides some good reasoning behind git. http://tom.preston-werner.com/2009/05/19/the-git-parable.html	14:35
	babilen: you're welcome, but please note that I'm a bot. I'm not programmed to care.	14:35
canton7	remember you can use directory names, . and * as well as filenames	14:35
iveqy	onethfour: git reset --hard HEAD	14:36
onethfour	thanks canton7 and iveqy	14:36
canton7	babilen, the bot responds to a range of stuff, some of it starting with !	14:36
onethfour	what if i dont put --hard?	14:36
canton7	then nothing happens	14:36
	→ byprdct joined	14:36
canton7	(in this case)	14:36
wildfire	canton7: excellent - thanks	14:36
canton7	reset --hard = move the branch and the working copy and the index to the named commit (in this case HEAD) - so the moving the branch bit is a no-op, but the rest takes effect	14:37
iveqy	babilen: !privmsg	14:37
gitinfo	babilen: You can get gitinfo keyword expansion (e.g. !doc or this one) sent directly to you by sending a private message to me: /msg gitinfo !doc. This avoids spamming the channel. Also see http://jk.gs/git/bot/trigger.php	14:37
onethfour	Now, how do I copy 1 file from another branch over to my current branch?	14:37
babilen	canton7: yeah, fair enough. Seems to be a strange design decision, but I won't argue the point. :)	14:37
canton7	reset [--mixed] = move the branch and the index to the named commit, leave the working copy alone	14:37
iveqy	onethfour: man git checkout	14:37
gitinfo	onethfour: the git-checkout manpage is available at http://jk.gs/git-checkout.html	14:37
canton7	again, the moving the branch bit is a no-op, so it'll just reset the index	14:37
	reset --soft = move only the branch, leaving the working copy and index alone. that's a complete no-op in this case	14:38
	← kewah left	14:38
babilen	iveqy: yeah, I was simply surprised by the fact that I have to include the "!" as that should simply be necessary to address the bot in the channel (a lot of bots allow you to configure this "address character" and I wouldn't expect it to be necessary in a query. But thanks.	14:38
canton7	onethfour, my preferred approach, and the one recommended by 'git status' is 'git checkout -- .'	14:38
iveqy	canton7: you know about !resets?	14:38
gitinfo	canton7: tl;dr of man git-reset: --soft moves HEAD, --mixed moves HEAD+index, --hard moves HEAD+index+work-tree, -- foo.txt will !unstage	14:38
canton7	iveqy, I did not know that trigger, thanks	14:38
	!reset also	14:38
gitinfo	A good resource explaining git-reset is http://git-scm.com/2011/07/11/reset.html	14:38
iveqy	canton7: I learned it yesterday, I like it =)	14:39
canton7	babilen, the bot also responds to 'man something', some wiki stuff, other docs, dot-prefixed commands to edit triggers, generate graphs, etc	14:39
	→ HereCas joined	14:40
canton7	the stuff prefixed with '!' is a subset of the stuff it responds to, so keeping the prefixes keeps everything nice and unambiguous	14:40
	← merlin83_ left	14:40
canton7	(it also responds to new joiners, people saying things, and some other odd stuff - all in response to non-prefixed keywords)	14:40
onethfour	iveqy, so this: git checkout master~2 Makefile	14:40
	i would do git checkout <otherbranch> path/to/file	14:41
wildfire	canton7: unfortunately I can't find .pushremote as an option	14:41
iveqy	onethfour: what's stopping you?	14:41
wildfire	iveqy: git push origin -- seems to push all local branches	14:41
onethfour	ok i'll try that	14:41
	thansk	14:41
canton7	wildfire, it's new I think? it's in man git-config	14:41
gitinfo	wildfire: the git-config manpage is available at http://jk.gs/git-config.html	14:41
canton7	wildfire, set push.default to something sensible	14:41
	like I said at first	14:41
wildfire	whereas I want to configure which ones go, or require explicit branch names -- so that local things are not inadvertedly pushed up	14:41
iveqy	wildfire: no it won't. See canton7s answer to me earlier. It just push matching branches	14:42
	wildfire: if you create a branch with a different name than any branch that's on your remote it won't be pushed	14:42
babilen	canton7: I don't think it would be appropriate to continue this discussion here, but I still think it is a weird design decision. Either way I understand how gitinfo works now and will keep future "conversations" with it in private.	14:42
canton7	wildfire, yes, that's what push.default configures. new git versions use 'simple' as the default value, but old git versions use the less intuitive 'matching'	14:42
	babilen, cool. it has its reasons, that's all. jast's the creator if you want to question him personally :P	14:43
	it might be worth making the '!' optional in privmsg to avoid confused people like yourself	14:43
iveqy	canton7: actually, git uses just a warning right now but will switch to simple when version 2.0 is released	14:43
	← rpg left	14:43
canton7	iveqy, oops, I mis-remembered	14:43
babilen	canton7: Nah, it works and that's what counts. I wouldn't have designed it that way but it is by no means horrible enough to warrant changing it now. :)	14:43
iveqy	babilen: gitinfo is on github and I'm sure jast likes pull requests =)	14:44
	← grilix left	14:44
canton7	github.com/jast/gitinfo iirc	14:44
	→ herdingcat joined	14:44
	→ merlin83_ joined	14:44
	← merlin83_ left	14:44
	→ merlin83_ joined	14:44
babilen	canton7: fun, line after line of undocumented perl	14:44
	→ ThomasClem joined	14:45
babilen	(almost undocumented)	14:45
canton7	isn't that par for the course in perl? :P	14:45
	→ dsdeiz_ joined	14:45
	→ atSetKey joined	14:45
	← Norm_ left	14:45
	← dsdeiz left	14:45
iveqy	babilen: good code doesn't need documentation they say ;)	14:45
onethfour	iveqy, it didn't work. I did git checkout mybranch path/to/file && git status and it says nothing changed	14:45
canton7	maybe the file didn't change in mybranch?	14:46
	← kbart left	14:46
iveqy	onethfour: !repro	14:46
gitinfo	onethfour: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.	14:46
onethfour	oh maybe	14:46
iveqy	onethfour: you can check that with git diff	14:46
	→ reskp__ joined	14:46
	→ fragamus joined	14:47
	← p3lim left	14:47
	← sebastianmarkow left	14:47
	→ p3lim joined	14:47
	→ teejmya joined	14:47
wildfire	canton7: ahh, my problem is my local git is a bit old in the teeth and doesn't have know 'simple'	14:48
	← ThomasCle left	14:48
	→ sebastianmarkow joined	14:48
wildfire	canton7: thank you again though	14:48
	← reskp left	14:48
canton7	wildfire, set it to 'current' or 'upstream'	14:48
	(depending on your preferences)	14:48
	← rat_thing left	14:48
canton7	'simple' is like a combination of those two, that's even safer for newbies	14:48
	← fragamus left	14:49
	← raijin left	14:49
	→ p3lim_ joined	14:49
wildfire	canton7: cool, I'll read up on those (but upgrade in progress anyway)	14:50
	← Robarelli left	14:50
	← atSetKey left	14:50
	← teejmya left	14:50
canton7	wildfire, cool :)	14:50
	← mFacenet left	14:50
	← cmpera left	14:51
	← p3lim left	14:51
	→ willwh joined	14:52
	→ d2dchat joined	14:52
	← boboc left	14:53
	← aChatir1 left	14:53
	→ normanrichards joined	14:53
	→ eddyb joined	14:54
	→ nnyk_ joined	14:54
	← merlin83_ left	14:55
eddyb	is there a way to force submodule update to clone only the HEAD of a branch?	14:55
	setting the branch in .gitsubmodules and running git submodule update --init --depth=1 doesn't work :(	14:56
	→ glosoli joined	14:56
	← v0n left	14:57
eddyb	it seems to fetch master	14:57
	→ v0n joined	14:57
	← mikecmpbll left	14:57
	← bunbury left	14:57
	← buibex_ left	14:57
	← willwh left	14:58
cannonball	Somehow, my local master has diverged in some small way from the origin/master. As a result, when I 'git push --tags', it generates a large (5MB) upload to the server, and then 20 or so tags (previous years of work) get rejected because it's non-fast-forward.	14:58
	→ bwreilly joined	14:58
	← ej left	14:58
	→ buibex joined	14:58
	← harobed_ left	14:58
eddyb	s/gitsubmodules/gitmodules	14:58
cannonball	I want to delete my local master, and then pull a fresh copy of master from origin. I'm trying to wrap my head around it, but I can't figure out how to get rid of whatever is fubar'd to cause this difference.	14:58
	→ fobius joined	14:59
	← juvenal left	14:59
	gitinfo set mode: +v	14:59
iveqy	eddyb: !broken	14:59
gitinfo	eddyb: [!doesntwork] Sorry to hear it doesn't work, but how about some details? At this point we don't even know whether it did nothing, spit out an error message, or abducted your pet unicorn in a flying saucer. That makes it hard to troubleshoot – unicorn abduction requires a completely different solution than an error message, for instance. So, please be specific...	14:59
	→ Robarelli joined	14:59
ToxicFrog	cannonball: so, for just "delete local master and get a fresh copy" - git fetch; git checkout master; git reset --hard origin/master	14:59
eddyb	iveqy: git submodule update --init --depth=1 fetches the last commit from master, not the branch in .gitmodules	15:00
	← lng left	15:00
ToxicFrog	cannonball: that said, it sounds like it has diverged in more than "a small way" if it's trying to push that many non-ff tags and the like.	15:00
	cannonball: are you sure you haven't rebased or filter-branched master?	15:00
	← bklane left	15:00
canton7	I'm surprised by the non-ff tags. Are you sure it isn't trying to push some other branches as well, which are being rejected?	15:01
	→ bklane joined	15:01
	← rob_w left	15:01
iveqy	eddyb: the --depth thing is quite bad. Since git only allows fetching refs (and their parents) the --depth=2 thing to a submodule says "give me master and 1 parent". However if the superproject point to a submodule commit that's currently a grandparent to master the correct commit won't be fetched. So that's actually quite bad and git submodule --depth won't be usefull until git allows you to fetch commits directly and not just refs	15:01
	→ matthiaskrgr joined	15:01
	gitinfo set mode: +v	15:01
	← Streward left	15:01
	→ mikecmpbll joined	15:02
cannonball	ToxicFrog: filter-branched, definitely not. rebased...I don't think so. The tags that it's complaining about are tags between 2005 and 2009. Everything since then is just fine.	15:02
	← bklane left	15:02
eddyb	iveqy: the commit is the HEAD of the branch I set in .gitmodules. I want it to fetch from that branch, not master	15:02
matthiaskrgr	does someone know a regex to extract git commit hashes (long and short) from text ?	15:02
	← v0n left	15:02
	→ bklane joined	15:02
iveqy	eddyb: --depth is a clone only command. So if your submodule already is clone, it won't effect the fetch	15:02
ToxicFrog	cannonball: honestly, I'd recommend firing up gitk and looking at the history and seeing exactly what's happened.	15:02
cannonball	I'm gonna try the reset --hard, but I don't expect it to fix the issue, because the issue appears to be back years ago, not with current commits.	15:02
	← buibex left	15:03
cannonball	ToxicFrog: Agreed.	15:03
	→ v0n joined	15:03
eddyb	iveqy: the submodule isn't initialized	15:03
ToxicFrog	cannonball: are you sure someone else hasn't rebased or filtered master and then force pushed it?	15:03
iveqy	eddyb: good.	15:03
	← v0n left	15:03
cannonball	No, I'm not sure of that. Am digging now.	15:03
	→ v0n joined	15:03
	→ raijin joined	15:03
	← bklane left	15:03
	← v0n left	15:03
cannonball	git reflog just shows my actions, right? I can't see what others would have done to the remote repo.	15:03
iveqy	eddyb: so it should clone the branch and not master =(	15:04
	→ v0n joined	15:04
iveqy	eddyb: have you played with reference?	15:04
	→ zarubin joined	15:04
	→ bklane joined	15:04
	→ acharlieh joined	15:04
eddyb	iveqy: why is it not fetching again, after deinit?	15:04
	← iamjarvo left	15:05
ToxicFrog	cannonball: as far as I know, yes	15:05
	→ dsdeiz joined	15:05
	→ buibex joined	15:05
eddyb	iveqy: oh, there we go, rm -rf .git/modules	15:05
	→ Cromulent joined	15:05
	→ natechan joined	15:06
eddyb	iveqy: # On branch master	15:06
	→ iamjarvo joined	15:06
	→ willwh joined	15:07
	← willwh left	15:07
	→ willwh joined	15:07
	jnavila_ → jnavila	15:07
	← dsdeiz_ left	15:07
eddyb	iveqy: I have branch = rust-2013-08-28-do-not-delete in .gitmodules, it does a shallow clone (the commit is the HEAD of that branch), but the resulting clone only has master's HEAD	15:08
	→ BSaboia joined	15:08
	← glosoli left	15:08
	→ cmpera joined	15:08
	← bklane left	15:08
cannonball	ToxicFrog: Ok, so it was easier than I thought. The problem was with my repo. A simple "git pull --tags" fixed it (which appears to be equivalent to git fetch --tags). My local repo didn't know of those tags (they were plain not present in gitk). After the pull --tags, they were visible and now push works as expected.	15:09
ojacobson	cannonball: not quite equivalent	15:09
	→ PaulCapestany joined	15:09
ojacobson	'git pull --tags' still does the merge	15:09
	If you want to download changes or refs, use git-fetch	15:09
	← shuaib left	15:10
cannonball	would that be equivalent to git pull --no-merge --tags ? (fetch and merge minus merge == fetch) ?	15:10
ojacobson	no	15:10
	← Cromulent left	15:10
eddyb	iveqy: this works as expected: git clone --depth=1 --branch rust-2013-08-28-do-not-delete https://github.com/alexcrichton/libuv.git	15:10
cannonball	Ok, thanks all for the feedback. I have a better understanding now.	15:10
	← ozette left	15:11
	← willwh left	15:11
	→ willwh joined	15:12
	← willwh left	15:12
	→ willwh joined	15:12
	← matthiaskrgr left	15:12
	← cbreak-work left	15:12
	← chuckharmston left	15:13
	→ harobed joined	15:13
eddyb	iveqy: git-submodule is broken :(. it checks out the branch after cloning	15:14
	which means I have no chance of fixing that today. bloody thing	15:15
	← kama_ left	15:15
iveqy	eddyb: nah that's by design. The --branch thing is for people that need the submodule to be on a branch instead of beeing on the tracked commit-id	15:15
	→ kama_ joined	15:15
	qsuscs → qsusOFFcs	15:15
iveqy	however I'm a bit surprised about the clone issue you're having	15:15
	is it a public repo?	15:16
	(I'm btw. the one that added the --depth option to git submodule, but I did not write the clone part)	15:16
eddyb	iveqy: clone --branch works, but branch in .gitmodules doesn't, because it does checkout after clone. you just confirmed that's the intended behavior, what's surprising?	15:17
	→ japaric joined	15:17
fobius	Newb Question: What are best practices for making sure that auto merges work as expected? For example, say that we have two users, user1 and user2. user1 introduces a new method addXYZ which returns a float. User2 now goes and writes some code which uses addXYZ but does not change the method. User2 pushes to the server. Now, in the meantime user1 rewrites addXYZ to return an integer. User1 pulls down changes made by user2 and merges them. The chang	15:17
	es made by user2 are auto merged because they do not try to modify the same lines since user2's code simply uses addXYZ. Yet, the changes user1 made to addXYZ break user2's new code.	15:17
eddyb	iveqy: this repo: https://github.com/mozilla/rust/blob/master/.gitmodules (first two submodules are really big)	15:17
	→ sttau joined	15:18
iveqy	eddyb: it's surprising that git submodule update doesn't care about branches.	15:18
eddyb	it made no difference before --depth	15:19
	← v0n left	15:19
	← lavaman1 left	15:19
iveqy	eddyb: the -b branch option isn't about cloning	15:19
	→ tuxx joined	15:19
iveqy	eddyb: hmm that's true	15:19
eddyb	(whether it was a clone argument or a checkout afterward)	15:19
	← tuxx left	15:19
	→ ngladitz joined	15:20
	← kama_ left	15:20
iveqy	eddyb: well, the correct solution is to make --depth fetch a commit and its history, not a branch. So making git submodule update respect branches when cloning is not a final solution I'm afraid	15:20
	← ngladitz left	15:21
	→ teejmya joined	15:21
	→ m8 joined	15:21
iveqy	fobius: User1 did the merge and is responsible that it works. He should run the test-suite and compile before pushing that	15:22
	← willwh left	15:22
	← iamjarvo left	15:22
	→ pepijndevos joined	15:23
iveqy	eddyb: look at man git clone at -b. It looks like it's not possible to clone from a refs as it is to fetch from one	15:23
gitinfo	eddyb: the git-clone manpage is available at http://jk.gs/git-clone.html	15:23
	→ iamjarvo joined	15:23
iveqy	eddyb: or wait, maybe with adding --single-branch	15:23
eddyb	git clone --branch works fine	15:23
pepijndevos	I just did "git reset --hard HEAD" but "git status" still shows a lot of modified files. What's up with that?	15:24
	→ rpg joined	15:24
iveqy	eddyb: doesn't that give you all the branches?	15:24
eddyb	iveqy: with --depth=1 it does exactly what I want	15:24
	← twisted` left	15:24
iveqy	eddyb: :(, not according to the docs. Too bad	15:24
	← kirsty left	15:24
eddyb	and it's fast (which is the important thing here. I don't want half an hour in the build process)	15:25
	← chriswk left	15:25
	→ todd_dsm joined	15:25
iveqy	eddyb: well, it would be quite easy to modify git-submodule.sh to allow it to pass -b to git clome	15:25
	clone*	15:25
	pepijndevos: are those tracked files?	15:26
eddyb	fobius: I would use rebase in that context, so the commit user1 made (changing return type) could be amended to include updated usage for addXYZ, and it will be directly on top of user2's last commit	15:26
	← asQuirreL left	15:26
pepijndevos	iveqy, yes	15:26
	← skelterjohn left	15:26
iveqy	pepijndevos: what does git status tell you?	15:26
pepijndevos	they are listed under changes not staged for commit	15:26
iveqy	and what does git reset --hard HEAD tell you?	15:27
eddyb	iveqy: let me try without --depth=1	15:27
pepijndevos	iveqy, HEAD is now at blabla	15:27
iveqy	pepijndevos: !repro	15:27
gitinfo	pepijndevos: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.	15:27
eddyb	iveqy: same thing, only the branch specified in clone's --branch argument gets fetched	15:27
	→ maestrojed joined	15:28
pepijndevos	iveqy: http://www.fpaste.org/45794/14189081/	15:28
fobius	eddyb, thanks. I'm not really sure how rebase works so I'll go look it up.	15:28
	→ LukeKetley joined	15:29
iveqy	eddyb: that's a good thing... I'm afraid you found a bug :)	15:29
eddyb	iveqy: oh c'mon...	15:29
fobius	iveqy, It just seems strange to me that the choice in git was to auto merge code that could be potentially harmful.	15:29
iveqy	eddyb: it's easy to fix if you really need it though...	15:29
eddyb	fobius: that's why you have a choice :P. git pull --rebase	15:30
iveqy	fobius: tell me one other VCS that doesn't do this	15:30
	→ FernandoBasso joined	15:30
	→ skelterjohn joined	15:30
	← ThomasClem left	15:30
iveqy	pepijndevos: what does git diff tell you? What's changed in those files?	15:31
	← Burer left	15:31
	← ELLIOTTCABLE left	15:31
iveqy	eddyb: open git-submodule.sh and edit line 278 to add -b to clone	15:31
eddyb	fobius: merge combines changes which are in the wrong order. rebase moves local changes on top, as if all local changes were made after the latest remote ones	15:31
fobius	iveqy, well, I don't mean git specifically. I'm kind of a newb to VCS in general. I'm really just trying to understand the workflow better. I.e. what is expected of me if I need to push code that includes other people's changes.	15:31
pepijndevos	iveqy, you want to see the full diff? It's quite big. But it has actual changes...	15:31
eddyb	iveqy: -b and the branch, which isn't provided to module_clone AFAIK	15:32
iveqy	fobius: it's expected that you make sure you changes work...	15:32
	← howlymowly left	15:32
	→ lavaman joined	15:32
	← dhruvasagar left	15:32
iveqy	pepijndevos: very strange, I'm afraid I can't give you a straigth answer on why. Usually there's a good explanation for this	15:33
	← primitive left	15:33
iveqy	eddyb: no it isn't. But you could manually provide it first to test if it works as expected	15:33
pepijndevos	ok :(	15:33
	← HereCas left	15:33
	→ irqq joined	15:33
	→ primitive joined	15:33
fobius	iveqy, Well, sure, but does that mean writing tests for other people's code if they're not provided? If git auto merges other people's changes then how should I know what to test even?	15:34
	← Ropt left	15:34
	← mishok13 left	15:34
	→ chuckharmston joined	15:34
iveqy	fobius: it doesn't. It merges your changes	15:34
	← steme_jomps_ left	15:35
eddyb	iveqy: I was ready to make a pullreq to rust... no way I'm getting all of this setup today. I just wanted faster build times :(	15:35
iveqy	fobius: in your example you used a function in your code that had been changed upstream. That make your code errornous	15:35
	→ v0n joined	15:36
iveqy	eddyb: that's the glory of open source. You find a bug and then you can actually fix it. It's much faster than trying to get a company to give you a fix	15:36
fobius	iveqy, sure, but how would I know that it's erroneous? It wasn't changed when I started working on it. When I merged the changes into my code before I pushed git didn't even tell me that it had changed.	15:36
	← rchavik left	15:36
iveqy	eddyb: and you got direct help from one of the submodule developers...	15:36
	→ jasonroyle joined	15:37
eddyb	iveqy: oh I know but there's like 3 places where I need to make changes, and the entire thing is a hack because there's no clone-by-commit-id :(	15:37
	→ mribeirodantas joined	15:37
iveqy	fobius: of course git did told you that. Didn't you read the merge diff?	15:37
	→ ngladitz joined	15:37
iveqy	eddyb: yes. And that's a security feature... By design we're not allowed to fetch commnits that's not a parent to a reference. So to implement a fetch by commit_id, it will be quite expensive on the server-side	15:38
	← tdussa left	15:38
iveqy	and even when we do, most hosting services like github/bitbucket etc. will take some time to get the changes into their codebases	15:38
	so it's not completely trivial	15:38
fobius	iveqy, Aha, okay. That's pretty much what I'm looking for. Again, I'm a git newb. How do you get ahold of the merge diff to show what was automatically merged in?	15:38
	← ngladitz left	15:39
iveqy	fobius: man git diff	15:39
gitinfo	fobius: the git-diff manpage is available at http://jk.gs/git-diff.html	15:39
	iveqy is afk	15:39
	← renumi left	15:39
fobius	iveqy, Okay, thanks.	15:39
eddyb	iveqy: still, in the case of rust, all of those submodules are pointing to branch HEADs, so it would work, but only maybe in 1.8.5 :(	15:40
	← d2dchat left	15:40
	→ boboc joined	15:40
	→ worblehat joined	15:41
	→ joshsmith joined	15:41
	← cmpera left	15:41
	→ Cromulent joined	15:41
psynor	ok, i know this question isnt "right" but how do i switch the active branch of a bare repo?	15:43
	← jnavila left	15:43
	← jargon left	15:44
	← Balliad left	15:44
	← iamjarvo left	15:44
	→ iamjarvo joined	15:45
PerlJam	psynor: do you have access to the filesystem that the bare repo lives on?	15:45
	→ atSetKey joined	15:45
	→ Roots47 joined	15:46
	← Roots47 left	15:46
PerlJam	psynor: if so, you should be able to run "git symbolic-ref HEAD refs/heads/some_branch" to change it.	15:46
	→ Roots47 joined	15:46
PerlJam	psynor: if you don't have access to the filesystem, then I have no clue how to do it. Google or stackoverflow might be your friend.	15:47
	→ lifechamp joined	15:47
	→ druonysus joined	15:49
	← druonysus left	15:49
	→ druonysus joined	15:49
	→ jargon joined	15:49
	← iamjarvo left	15:49
	→ cmpera joined	15:50
	→ dhruvasagar joined	15:50
	← atSetKey left	15:50
	→ atrigent joined	15:50
	→ willwh joined	15:50
	← willwh left	15:50
	→ willwh joined	15:50
	→ shinnya joined	15:50
	→ grilix joined	15:51
atrigent	hey guys, I've got a blob in the history of my repo that I would like to compare with a file that is outside my repo - is there a way to do this with git diff?	15:51
	← jargon left	15:51
	→ glosoli joined	15:51
	→ asQuirreL joined	15:51
	← zarubin left	15:52
	← Misan left	15:52
milki	atrigent: hm, you could replace that blob with your new file in a new commit and hten use git diff?	15:52
	→ infynyxx joined	15:52
	→ jargon joined	15:52
milki	or maybe just add the file so that its a blob and compare two blobs	15:53
	→ jaequery joined	15:54
	→ leeN joined	15:55
	← axl_ left	15:55
atrigent	hmm well if it came to that then I'd rather just check out the blob somewhere and use regular diff	15:56
osse	atrigent: maybe something like diff <(git show HEAD:path/to/file) path/to/other/file	15:56
atrigent	osse: ah yeah, that'll work	15:57
	I guess that's good enough	15:57
	thanks	15:57
osse	atrigent: according to the man page you can do git diff -- some/path ../outside/repo/...	15:58
atrigent	osse: right, you can compare two files, but not a file and a blob it seems	15:58
	→ Amandil joined	15:59
	→ rat_thin_ joined	15:59
	← ehmatthes left	16:00
	← jasonroyle left	16:01
	← epta left	16:01
	→ axl_ joined	16:02
	→ hyperair joined	16:02
	← boboc left	16:02
	← bwreilly left	16:03
	← Amandil left	16:03
	→ Amandil joined	16:04
	→ Norm_ joined	16:04
	← acharlieh left	16:04
	← Anticom left	16:04
	→ AWinterman joined	16:05
	→ rednaks joined	16:06
	← MafooUK left	16:07
	→ acharlieh joined	16:07
	← dsdeiz left	16:07
	← AWinterman left	16:07
	← freestyl3r left	16:08
	← petey left	16:08
	→ freestyl3r joined	16:08
	← jargon left	16:08
	→ epicowl joined	16:08
	← willwh left	16:10
	→ ngladitz joined	16:10
	← ngladitz left	16:11
	← mizu_no_oto left	16:12
	→ ehmatthes joined	16:12
	← Roots47 left	16:12
	→ Exodist joined	16:13
	→ Roots47 joined	16:13
	→ NanoArrow joined	16:15
	← nathanbbb left	16:15
	→ nathanbz joined	16:15
	← v0n left	16:17
	→ chirayuk joined	16:17
	→ HereCas joined	16:18
	← cmpera left	16:19
	← HumanSky left	16:19
	← \|Freeman\| left	16:19
	→ ebertx joined	16:21
	→ MafooUK joined	16:22
	→ zhost joined	16:22
	← MafooUK left	16:23
	→ huli_ joined	16:23
	← DanielKarp left	16:23
	→ zxvf joined	16:24
	→ harish_ joined	16:24
	gitinfo set mode: +v	16:24
	→ rfree joined	16:24
	← herdingcat left	16:25
	← multiverse left	16:25
	← Voting left	16:25
	→ bwreilly joined	16:25
	← huli_ left	16:25
	← harish_ left	16:26
	→ harish_ joined	16:26
	← glosoli left	16:26
	← harish_ left	16:26
	← mklappstuhl left	16:27
zxvf	hello. a question a bit off-topic: Where can I find info about percentage of use of different DVCS systems?	16:27
	→ ngladitz joined	16:27
	→ epta joined	16:28
ojacobson	probably nowhere, if you want numbers with any kind of statistical relevance	16:28
	why do you want them?	16:28
	← arvind_khadri left	16:29
	← chirayuk left	16:29
LuxuryMo_	I had a conflict when rebasing. I resolved the conflict and did "git add . && git rebase --continue", but git tells me "No changes - did you forget to use 'git add'?"	16:30
	← donmichelangelo left	16:30
	→ v0n joined	16:30
	← mikecmpbll left	16:30
	→ Xethron joined	16:30
	← Xethron left	16:30
	→ Xethron joined	16:30
	→ donmichelangelo joined	16:30
LuxuryMo_	Should I just --skip?	16:30
	← noname001__ left	16:31
rfree	consider that I forked a repository and then made 10 changes (git commits) to it	16:31
	how to select 3 of thoes changes and make a nice poll-request for the upstream consisting only of them?	16:31
	consider that I forked a repository and then made 10 changes (git commits) to it (that is my question also from #github)	16:31
	← reim left	16:31
	→ xpe joined	16:31
	← ngladitz left	16:32
rfree	uh. the git question is, what would be the git command to make such a branch and cherry pick into it	16:32
osse	oojacoboo: wild guess: to convince his boss to switch	16:33
	oops	16:33
	wrong person	16:33
	← aarwine left	16:33
rfree	actually, I guess http://nathanhoad.net/how-to-cherry-pick-changes-with-git displays it enough	16:33
zxvf	ojacobson: I am curious	16:33
	→ willwh joined	16:34
	→ aarwine joined	16:34
	→ Peter_Occ joined	16:34
	← bamj0rb left	16:35
	← sttau left	16:36
	gitinfo set mode: +v	16:36
	← HereCas left	16:36
	→ HereCas joined	16:37
	← YannH left	16:38
	→ lazers joined	16:38
	→ YannH joined	16:39
	→ zarubin joined	16:39
	← willwh left	16:40
Peter_Occ	What does this mean? Your branch and 'origin/master' have diverged, and have 1 and 1 different commit each, respectively.	16:40
	→ aard_ joined	16:40
osse	Peter_Occ: you have one commit on your computer that's not present on 'origin', and there is one commit on origin you don't have on your computer	16:41
	← Sonderblade left	16:41
ojacobson	Peter_Occ: have a visualization: 'git log --graph --oneline --decorate --color HEAD origin/master'	16:41
LuxuryMo_	Looks like git will view resolving a conflict when rebasing by essentially discarding your commit as not "adding" anything…so --skip seems to be the way to go...?	16:41
ojacobson	LuxuryMo_: the "this" part of your "if this then that" logic doesn't hold	16:42
	git doesn't view resolution that way, except in specific circumstances	16:42
	→ dsdeiz joined	16:42
ojacobson	if the resolution ends up being identical to HEAD, then it's viewed as no change, because it _is_ no change from HEAD	16:42
	← HereCas left	16:43
LuxuryMo_	right	16:43
	thanks ojacobson	16:43
	is that sort of what git means when it says ""If there is nothing left to stage, chances are that something else already introduced the same changes; you might want to skip this patch.""?	16:43
	→ glosoli joined	16:43
	→ harish_ joined	16:43
	← rednaks left	16:44
	← harish_ left	16:44
	← harobed left	16:44
	← jamiis left	16:44
	→ gnrlbzik joined	16:45
	→ atSetKey joined	16:46
	→ sguselnikov joined	16:46
	← sguselnikov left	16:47
	→ nickpresta joined	16:48
	→ lng joined	16:49
	→ arietis joined	16:50
	← atSetKey left	16:51
	→ dreiss_ joined	16:51
	→ maslowbeer joined	16:52
	← GeertV left	16:52
	→ GeertV joined	16:53
	→ thiago joined	16:53
	→ Baluse joined	16:53
	→ hambone joined	16:54
	← buibex left	16:54
Peter_Occ	Thanks for the help.	16:54
	glosoli → cowsoli	16:54
	cowsoli → glosoli	16:55
	← zhost left	16:55
	← acharlieh left	16:55
	→ JoeQuery joined	16:55
	← babilen left	16:56
JoeQuery	Hi! I'm using git svn, and I recieve this error when running 'git svn dcommit': "Permission denied: Can't open file '/var/svn/myproject/db/txn-current-lock': Permission denied at /usr/lib/git-core/git-svn line 922"	16:56
	← hambone left	16:56
JoeQuery	That /var/svn directory doesn't even exist though. I tried making it just in case, and that doesn't help.	16:57
	← GeertV left	16:57
JoeQuery	(Also made it writable after making it)	16:57
	← DoniRS left	16:58
	→ rvince joined	16:58
rvince	I just changed every jsp file in a java web project. Now I have to commit them. Since htey are in all different directories, do I have to do a git add on each individual file? How can i tell it to just add every changed cjsp from a certain directory and lower?	16:58
	→ iamjarvo joined	16:58
lov	git add /path/to/root/of/changes/	16:59
	→ twisted` joined	16:59
rvince	lov : and it will add all files from the root on down?	16:59
	→ blackstar257 joined	16:59
	→ acharlieh joined	16:59
lov	yes, it's recursive	16:59
	← xpe left	16:59
rvince	ah! Thank you!	16:59
	← rath left	17:00
rfree	how to remove one commit abcd1234 from my history?	17:00
	or apply this commit in reverse	17:00
	← rjhunter left	17:00
ToxicFrog	rfree: "git revert" to reverse a commit without rewriting history.	17:01
	→ safinaskar joined	17:01
	→ rjhunter joined	17:01
ToxicFrog	"git rebase -i" to remove a commit from history entirely (warning: impolite)	17:01
	→ k4rtik joined	17:01
	← ChrisAnn left	17:01
	→ Sigma joined	17:01
	← iamjarvo left	17:02
	→ iamjarvo joined	17:02
	← arietis left	17:03
	← YannH left	17:03
	← pepijndevos left	17:04
	← kulp left	17:04
	→ jnavila joined	17:05
	→ Sigma[Mtp] joined	17:05
	← rat_thin_ left	17:05
LuxuryMo_	What does a conflict with (add/add) mean?	17:05
	→ arietis joined	17:05
	← rjhunter left	17:05
	→ rat_thing joined	17:06
	← Vampire0 left	17:06
	→ y_nk_ joined	17:06
	→ HumanSky joined	17:06
	← Sigma left	17:07
	← hahuang65 left	17:07
	← lng left	17:08
	← khushildep left	17:08
	→ kulp joined	17:08
LuxuryMo_	just means it was added anew in both branches?	17:08
	→ whunt joined	17:09
	→ cmpera joined	17:09
	→ HereCas joined	17:09
	← y_nk left	17:09
	← rat_thing left	17:10
	← JoeQuery left	17:10
	→ chirayuk joined	17:11
	← y_nk_ left	17:11
	← Squarism left	17:11
	→ Vampire0 joined	17:11
	→ chairabanta joined	17:11
	← chairabanta left	17:12
	→ babilen joined	17:12
	→ guillaum joined	17:12
	← dhruvasagar left	17:12
	→ Sigma joined	17:12
	← guillaum left	17:13
	← Wilduck left	17:13
	← Sigma[Mtp] left	17:14
	→ guillaum joined	17:14
	← guillaum left	17:15
	← todd_dsm left	17:15
	← nnyk_ left	17:15
atrigent	I'd like to do a git log starting a ways back in my projects history, and I'd like to get the log for a directory that no longer exists	17:15
	is there a way to do that?	17:15
	→ guillaum joined	17:15
	→ minim joined	17:16
	→ gghoffman joined	17:16
atrigent	oh wait nevermind	17:16
	just needed the --	17:16
	← minim left	17:16
	→ Burer joined	17:17
	→ multiverse joined	17:17
	← Burer left	17:18
	← multiverse left	17:18
	← gghoffman left	17:18
	→ boombatower joined	17:18
	→ multiverse joined	17:18
	→ Burer joined	17:18
	← NanoArrow left	17:19
	← Norm_ left	17:20
	← fobius left	17:20
	← Xethron left	17:20
	→ atc3030 joined	17:20
	gitinfo set mode: +v	17:21
	← ToxicFrog left	17:22
	→ OOPMan joined	17:22
	→ minim joined	17:22
atc3030	Hello everyone, as I am still gitting used to git ;) I had a question. If i was wanting to cherry-pick this commit from github (http://goo.gl/mERv0K) is there any way that I can find what has followed after, or perhaps what would have it as a dependency. Gerrit makes this process much much easier but their gerrit is shady at best.	17:23
	thank you for any help	17:23
	← rurufufuss left	17:23
atc3030	I will also need some help in figuring out how to actually cherry-pick said commit, but any commit in general will work as I will be doing it alot	17:23
	← Roots47 left	17:24
	→ davividal joined	17:24
Peter_Occ	If I delete a repository and then create a new repository and don't add or commit anything, git status says :	17:25
	On branch master	17:25
	Initial commit	17:25
	What does that mean?	17:25
davividal	hi. I have this one (legacy) SVN project. I want to import the SVN branch "foo" into a new git repository. Is that possible using git svn? I found this http://ivanz.com/2009/01/15/selective-import-of-svn-branches-into-a-gitgit-svn-repository/ but I'm not sure if it is what I'm looking for (SVN newbie :( )	17:25
	→ bwreilly_ joined	17:26
	→ b0rsuk joined	17:26
	→ bamj0rb joined	17:26
	← Cromulent left	17:26
	gitinfo set mode: +v	17:26
	← bwreilly_ left	17:27
	← nuclearkitten left	17:27
	← zxvf left	17:27
	→ btree joined	17:28
	← btree left	17:28
	→ btree joined	17:28
	← boblee left	17:28
	← HenryG left	17:28
	→ ToxicFrog joined	17:29
	→ DoniRS joined	17:29
	← bwreilly left	17:29
	→ ELLIOTTCABLE joined	17:29
	→ ngladitz joined	17:29
	← gnrlbzik left	17:29
	→ grampajoe joined	17:30
	→ bwreilly joined	17:30
	← ngladitz left	17:30
	← amcsi_work left	17:30
	← angelsl left	17:30
Vampire0	davividal, I personally would not recomment git-svn for a one-time conversion	17:31
	For using Git as frontend client to a SVN repository with two-way communication it is great	17:31
davividal	Vampire0: what about migration? I was going to follow http://git-scm.com/book/en/Git-and-Other-Systems-Migrating-to-Git	17:32
Vampire0	But for doing a real switch over I greatly prefer svn2git from the KDE project and probably svneverever before to help analyze the history of the repository	17:32
	http://blog.hartwork.org/?p=763	17:32
	This is a blogpost about svneverever that also has a link to the right svn2git tool	17:32
	← jkremser left	17:33
	← intellix left	17:34
	→ chriswk joined	17:34
davividal	Vampire0: it says something about 15,000 revs. I have 'only' 4k. Would you still use svneverever + svn2git ?	17:34
	→ steme_jomps_ joined	17:34
	← GODDOG left	17:35
	← panpainter left	17:35
	→ mizu_no_oto joined	17:35
	← Burer left	17:35
osse	atc3030: you can do something like git branch --contains foo to see which branches contain foo (in this case 518c329)	17:35
	atc3030: as for cherry-picking it: that might be difficult since it is a merge commit	17:35
	← kpreid left	17:36
atc3030	ok. I am going to take the time to read up more on git actually to hopefully be able to understand it a bit more rather than just copy past commands and praying they work. lol.	17:37
	THank you osse	17:37
	thank you #git	17:37
	← OOPMan left	17:37
	← guillaum left	17:37
osse	atc3030: to see a log of everything that follows, try git log --all ^518c329 (meaning the log of everything leading up to all refs, but exclude everything leading up to 518c329)	17:37
Vampire0	davividal, I didn't read that page. Where does it say something with 15000 revs?	17:37
	← dreiss_ left	17:37
	← Droolio left	17:37
	→ Wilduck joined	17:37
	→ panpainter joined	17:37
Vampire0	I personally would always use svn2git for the Migration I used it multiple times on projects of varying sizes and you have superior control over what goes where and so on	17:38
	→ MoHaX joined	17:38
osse	atc3030: git is hard to just "start using" if you don't know anything about it. Reading a bit would be a good idea.	17:38
	!book	17:38
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	17:38
osse	!talk	17:38
gitinfo	[!talks] Some good video talks about Git: [yt] http://goo.gl/z72s (Linus Torvalds: History&Concepts); [yt] http://goo.gl/R9H2q (Scott Chacon: Git basics, live examples); http://vimeo.com/35778382 (Randal Schwartz: Git basics, descriptional); http://vimeo.com/46010208 (Jesica Kerr: Git basics, descriptional)	17:38
	→ dhruvasagar joined	17:38
Vampire0	I even splitted a big fat SVN repository with it into multiple small Git repositories per module with one run	17:38
davividal	Vampire0: "So youre about to convert a larger SVN repository to Git, say 15,000 revisions." but I'll use it anyway	17:38
	it's a 8 year project...	17:38
	← cagedwisdom left	17:38
atc3030	cool. thank you osse	17:39
Vampire0	Ah ok davividal, on my link, I thought on your link	17:39
	That's just an example with the 15.000, I would even use svneverever and svn2git for a project with 50 commits	17:40
	← eddyb left	17:40
	← OMGOMG left	17:40
	→ eddyb joined	17:40
osse	atc3030: you're welcome	17:41
	→ hahuang65 joined	17:41
	→ OOPMan joined	17:42
	→ Voting joined	17:43
	→ schemanic joined	17:43
	← victorbjelkholm left	17:44
	→ mdentinho joined	17:44
	→ victorbjelkholm joined	17:44
	← dhruvasagar left	17:46
	→ ngladitz joined	17:46
	← ngladitz left	17:46
	→ atSetKey joined	17:46
	← b0rsuk left	17:47
	← tatsuya_o left	17:47
	→ jhulten joined	17:47
	← Guest32557 left	17:47
	← Voting left	17:47
	→ ngladitz joined	17:48
	→ tatsuya_o joined	17:48
	← ngladitz left	17:48
	→ Roots47 joined	17:48
	→ guillaum joined	17:48
	← guillaum left	17:48
	→ heylouiz joined	17:48
	→ str joined	17:49
	→ mishok13 joined	17:49
	← MoHaX left	17:49
	← wildfire left	17:49
	→ matanya_ joined	17:50
	→ TomyLobo joined	17:51
str	hi everyone. I have a git repository with the 2 last commits wrongly commited to master. How do I tell git to create a branch from the HEAD-2 revision and move those changes to that branch?	17:51
	← atSetKey left	17:51
iveqy	str: create a new branch from master and then reset master	17:52
	← BSaboia left	17:52
str	iveqy, create a new branch from mast <- done,	17:52
	→ danillonunes joined	17:52
	← tatsuya_o left	17:52
str	iveqy, and then reset master <-- how do I do that?	17:52
	→ Andrevan joined	17:53
	gitinfo set mode: +v	17:53
heylouiz	it's not better to rebase to the another branch?	17:54
	this way he'll keep those two commits	17:54
	← chirayuk left	17:54
	→ tufflax joined	17:56
jasom	iveqy: git checkout master; git reset HEAD~2	17:56
	← infynyxx left	17:56
Vampire0	ivegy: git reset HEAD~2 master	17:56
	str: git reset HEAD~2 master	17:56
	→ dhruvasagar joined	17:57
	← btree left	17:57
Vampire0	Peter_Occ, it just means that you are preparing the first (initial) commit and that there is no commit currently in the repository but that it is empty and you are on the branch called "master" that is created by default when you create a new Git repository.	17:58
jasom	Oh, for anyone interested, I was talking yesterday about storing .tgz in unzipped form for better packs; I (ab)used smudge/clean filters to do that and my pack went from 7G to 2.3G so I will call that success	17:58
jrnieder	jasom: nice :)	17:59
	jasom: any ideas about how this could be documented better for the next person?	17:59
	← tvw left	17:59
	→ tvw joined	18:00
jrnieder	jasom: e.g., a more complete example (including the filter-branch step, etc) for the EXAMPLES section of gitattributes(5)?	18:00
	→ breck7 joined	18:00
	← multiverse left	18:00
	→ str_ joined	18:01
	→ multiverse joined	18:01
jasom	jrnieder: I might be able to write it up this weekend; should I just send it to the mailing list?	18:01
jrnieder	yeah, that would be excellent	18:01
	→ apipkin joined	18:03
	← PaulCapestany left	18:03
	← str left	18:04
	← multiverse left	18:04
	→ ngladitz joined	18:04
	← asQuirreL left	18:05
	← ngladitz left	18:05
	→ pantsman joined	18:06
	← pantsman left	18:06
	→ pantsman joined	18:06
jasom	jrnieder: okay, it's on my todo list, but below about a half-dozen pull requests for another project I'm working on, so no promises :)	18:06
	← str_ left	18:06
	jasom thinks you can tell when his kids are with the grandparents by the flurry of his activities on FOSS projects	18:07
	← dhruvasagar left	18:08
	→ nnyk_ joined	18:09
	← byprdct left	18:11
	← mdentinho left	18:14
	← epicowl left	18:15
	→ byprdct joined	18:16
	← schemanic left	18:16
	→ Norm_ joined	18:16
	→ PaulCapestany joined	18:17
	← maslowbeer left	18:18
	→ dreiss_ joined	18:18
	→ tavoe joined	18:20
	→ d2dchat joined	18:20
	gitinfo set mode: +v	18:21
tavoe	Looking for help. I have "node_modules/" in my root .gitignore. I expect this to ignore every folder named "node_modules/" and all their contents. It does not.	18:22
	← heylouiz left	18:23
	→ ngladitz joined	18:24
	← jsec left	18:24
	→ alvaro_o joined	18:25
cbreak	tavoe: why do you think it does not?	18:25
onethfour	How do i grab a file from another branch and pull it into my current directory? (without checking out cp to temp dir checkout out other branch and cp from temp)	18:25
cbreak	tavoe: does such a directory show up as untracked?	18:26
atrigent	can git log --follow and --full-diff be used together? I'm getting "usage: git logs can only follow renames on one pathname at a time", but I'm only supplying one filename...	18:26
cbreak	onethfour: man git checkout	18:26
gitinfo	onethfour: the git-checkout manpage is available at http://jk.gs/git-checkout.html	18:26
onethfour	cbreak what do i use for the path?	18:26
	← ngladitz left	18:26
	← MrKeys88 left	18:26
cbreak	atrigent: pastebin	18:26
onethfour	git checkout OTHERBRANCH path/to/file ?	18:26
cbreak	onethfour: yes.	18:26
	onethfour: be aware that you will not get any history	18:26
	so it is NOT advised to do that	18:26
	onethfour: you should make a proper merge instead	18:27
onethfour	i dont need history	18:27
	← lifechamp left	18:27
	→ infynyxx joined	18:27
atrigent	cbreak: pastebin what? "git log --follow <path>" works, but "git log --follow --full-diff <path>" gives the message I gave above	18:27
	← synergy__ left	18:28
	→ Sendoushi joined	18:28
cbreak	atrigent: put a -- in as separator	18:28
	← kwmiebach left	18:28
atrigent	cbreak: that doesn't work either, I already tried that	18:28
	→ atSetKey_ joined	18:29
cbreak	hmm...	18:29
	← Roots47 left	18:29
cbreak	you are using -p?	18:29
	maybe --full-diff implies -p	18:29
	→ kwmiebach joined	18:29
onethfour	cbreak: i tried and it gave me a	18:30
	fatal: Could not switch to .... error	18:30
	← frgtn left	18:30
	→ iamjarvo_ joined	18:30
	← shoshin left	18:30
cbreak	onethfour: you don't have to switch.	18:30
	← iamjarvo left	18:30
onethfour	i didnt	18:30
cbreak	onethfour: as said above, put in a -- as separator as well	18:30
atrigent	cbreak: why would -p be a problem?	18:30
	→ frgtn joined	18:30
cbreak	that helps git tell appart the path and the argument part	18:30
	atrigent: not having -p would be one	18:30
	→ pr0ton joined	18:31
onethfour	cbreak: i tried: git checkout MYBRANCH -- /path/to/file.php and it gave me that fatal Could not switch error	18:31
jasom	isn't -p the default?	18:31
atrigent	ok well "git log -p --follow --full-diff <path>" doesn't work either	18:31
	← pr0ton left	18:31
atrigent	still not sure why -p is relevant though	18:31
	just --follow by itself works, as I said above	18:32
cbreak	onethfour: well, clearly, / is outside the repository	18:32
	→ snuffeluffegus joined	18:32
onethfour	ok i took out the preceding /	18:32
cbreak	atrigent: because --full-diff doesn't make sense if you don't show a diff	18:32
	← lazers left	18:32
onethfour	and now i'm getting '' did not match any file(s) known to git	18:32
cbreak	onethfour: then that commit doesn't have that file	18:32
	→ ezakimak_work joined	18:33
	← infynyxx left	18:33
	→ pr0ton joined	18:33
onethfour	its not a commit it is a branch	18:33
cbreak	onethfour: it's a commit	18:33
	→ infynyxx joined	18:33
	← arietis left	18:33
	← tavoe left	18:34
	→ maroloccio joined	18:34
onethfour	ok i switched it to git checkout MYBRANCH -- path/to/file.php and typed ls and the file doesn't exist?	18:34
atrigent	cbreak: it looks like --full-diff is just ignored when -p is not supplied	18:34
cbreak	onethfour: git cat-file -t blah, it'll tell you what it is :)	18:34
atrigent	just tried without --follow	18:34
onethfour	what you mean?	18:34
	→ Misan joined	18:34
cbreak	onethfour: it's a commit. that commit doesn't have that file.	18:35
	← frgtn left	18:35
atrigent	in any case, yes I was using -p... sorry, I took it out for simplicity	18:35
cbreak	onethfour: if you want to see which files it contains, git ls-tree -r blah	18:35
atrigent	perhaps I shouldn't have	18:35
onethfour	ok so it turns out it didn't copy the file to my current directory like i wanted, but rather in the same path as the other branch it was in	18:35
	→ mdentinho joined	18:35
cbreak	atrigent: then it seems something's wrong. I wonder if the order matters...	18:35
	→ willwh joined	18:35
cbreak	onethfour: of course.	18:35
	onethfour: it won't copy anything anywhere, it'll just check it out.	18:36
	→ zhost joined	18:36
onethfour	what the heck is a "check out"	18:36
cbreak	checkout has many functions	18:37
	→ epicowl joined	18:37
osse	onethfour: the "it" makes all the difference	18:37
onethfour	well git checkout did too many things i didn't want	18:37
cbreak	in file mode, it gets files from the index or from commits and checks them out into the current working directory (and index)	18:37
onethfour	all i wanted to do was copy and paste a file	18:37
cbreak	onethfour: then you should be more specific with what you want :)	18:37
onethfour	and use the file as a template for what i was doing	18:37
	no history nothing	18:38
	git checkout added to staged	18:38
	← FernandoBasso left	18:38
cbreak	then you could have used git cat-file -p Blah:path/to/file \| pbcopy	18:38
	(on OS X)	18:38
onethfour	yeah cat-file might have worked	18:38
	→ lazers joined	18:38
onethfour	just cat it to a file	18:38
	>	18:38
	← aard_ left	18:38
	→ howlymowly joined	18:38
cbreak	pbcopy would be better.	18:38
	that'd allow you to copy&paste the thing	18:39
onethfour	git cat-file -p MyBranch:path/to/file > myfile.txt	18:39
	→ dhruvasagar joined	18:39
onethfour	no need to use clipboard	18:39
cbreak	you can't really copy&paste without it.	18:39
onethfour	you can copy though	18:39
	and effectually that's what i meant	18:39
	← grampajoe left	18:40
	→ fobius joined	18:40
	→ Roots47 joined	18:40
	→ arietis joined	18:40
	→ grampajoe joined	18:41
	→ LuxuryMode joined	18:41
	← demiurgo left	18:42
	← vurma left	18:42
	← LuxuryMo_ left	18:42
	→ ngladitz joined	18:43
	← lazers left	18:43
	→ terminalRepeat joined	18:43
	← willwh left	18:43
	→ demiurgo joined	18:43
onethfour	Why does it take a few operations for git to figure out that you renamed a folder? is there a quick way to tell it that you did that	18:44
osse	onethfour: use git mv from the get-go	18:44
	← ngladitz left	18:44
onethfour	oooh i see	18:44
osse	or regular mv, then git add -A	18:44
	← ryan_turner\|MTW left	18:44
	← chuckharmston left	18:46
onethfour	Is there a way to ignore a file without adding it to .gitignore	18:46
	repo specific?	18:46
	← axl_ left	18:46
	→ chuckharmston joined	18:46
osse	onethfour: add it to .git/info/exclude	18:46
	← freestyl3r left	18:47
jrnieder	jfyi, onethfour has been repeating the same questions over a few days :)	18:47
	← AaronMT left	18:47
jrnieder	perhaps that means the answers weren't helpful enough	18:47
	→ chuckhar_ joined	18:47
	← chuckharmston left	18:47
onethfour	i forget easily :(	18:47
	← dhruvasagar left	18:47
onethfour	and i learned something new with that git cat-file thing (thanks cbreak)	18:48
osse	I have to look up what a "record" and what a "field" is every single time I use awk. It just won't stick	18:48
	← chuckhar_ left	18:48
	← dsdeiz left	18:48
	→ tjbiddle joined	18:49
	→ chuckharmston joined	18:49
	→ tatsuya_o joined	18:50
	← chuckharmston left	18:50
	← victorbjelkholm left	18:50
	→ chuckharmston joined	18:50
jasom	osse: really, I thought those were common terms	18:51
	→ harobed_ joined	18:51
	← demiurgo left	18:51
osse	jasom: they probably are :'(	18:51
jasom	osse: e.g. C has "fields" in structs	18:52
	→ asQuirreL joined	18:52
osse	heh, I call them members.	18:52
	→ demiurgo joined	18:53
osse	I'm not good with lingo	18:53
cbreak	HTML has fields in forms. :)	18:53
	← asQuirreL left	18:53
	Jiffy_ → Jiffy	18:53
jasom	osse: you probably cal RDMBS records "rows" too then	18:53
osse	jasom: depends on what RDMBS is	18:53
atrigent	baseball has fields too	18:53
jasom	that seems to be an old-school new-school divide: fields vs. members and records vs. rows	18:54
	→ dhruvasagar joined	18:54
cbreak	HTML also has rows. In Tables.	18:54
	must be the ultimate language for databases.	18:54
osse	Formula 1 circuits have records	18:55
	type: lap	18:55
	jasom: I may have exhaggerated. These days it's roughly every other time I use awk.	18:56
	→ axl_ joined	18:56
osse	The "member" thing probably comes from OOP	18:56
	though I still call methods functions :P	18:56
cbreak	osse: they are called "member functions" in the C++ standard too :)	18:57
osse	cbreak: brilliant. then I don't have to feel bad	18:57
jasom	called "fields" in the C standard	18:57
osse	At least I don't say bitmember	18:57
jasom	also pascal has "records" which contain "fields"	18:58
cbreak	C++ has all kinds of members. Member types, member functions, member templates, member variables... :)	18:58
osse	commitee members	18:59
jasom	C is just C++ that has been dismembered	18:59
osse	haha	18:59
	← matteo left	19:00
	← karstensrage left	19:00
	→ sparq_ joined	19:00
	← tvw left	19:00
	← HumanSky left	19:00
	← rethus left	19:01
	→ ngladitz joined	19:01
	← ngladitz left	19:01
	→ ngladitz joined	19:01
	← sparq_ left	19:01
	← arietis left	19:02
	→ buibex joined	19:02
	→ nwertman joined	19:03
	← iamjarvo_ left	19:03
	→ iamjarvo joined	19:03
	← pr0ton left	19:03
bremner	so why is C++ so much more painful??	19:03
	→ lifechamp joined	19:03
	← buibex left	19:03
	→ Guest29305 joined	19:04
	← mdentinho left	19:04
	← ngladitz left	19:04
	← rvince left	19:04
cbreak	it's not.	19:05
	I love it.	19:05
	← ewnd9 left	19:06
cbreak	osse: C++ has bit members too :) https://ideone.com/8xyfW9	19:06
jasom	cbreak: We'll have to agree to disagree :)	19:06
cbreak	hardly anyone uses them.	19:06
	(for obvious reasons)	19:06
jasom	bitfields in C should be avoided, since many ABIs require quite suboptimal code gen for them	19:07
	→ ryan_turner\|MTW joined	19:07
jasom	e.g. an int is passed in a register while a bitfield (that would fit in an int) is passed on the stack	19:07
	← nwertman left	19:08
	← nickpresta left	19:08
	→ sq-one joined	19:08
	→ rpg_ joined	19:09
	← rpg left	19:09
	→ rednaks joined	19:10
jasom	the e500 sysV abi even is caller copy, so a copy is made even if the callee doesn't always write to it; (a copy can be avoided if the function is declared const, but nobody actually remembers to add const to their function declarations)	19:10
	← snuffeluffegus left	19:11
	→ snuffeluffegus joined	19:11
	→ AaronMT joined	19:11
cbreak	just use a const ref. :)	19:11
	← snuffeluffegus left	19:11
	→ snuffeluffegus joined	19:12
	→ x2z joined	19:14
	→ mFacenet joined	19:15
	→ FernandoBasso joined	19:16
	← Wilduck left	19:16
davividal	Vampire0: sorry, just curiosity: why don't you recommend use git-svn for one-time conversions?	19:18
	I just got asked from my boss	19:18
	← k4rtik left	19:18
	→ NanoArrow joined	19:18
onethfour	why don't man pages have any examples :(	19:18
	git mv ???	19:18
	→ MrKeys88 joined	19:19
davividal	onethfour: are you talking specifically about git-mv?	19:19
	→ Lekensteyn joined	19:19
onethfour	yeah nevermind	19:20
	pretty simplistic	19:20
	→ asQuirreL joined	19:20
	← Lekensteyn left	19:20
	← kcm1700 left	19:21
	→ kcm1700 joined	19:21
canton7	git mv [-v] [-f] [-n] [-k] <source> <destination> - pretty clear :P	19:21
	→ valgrind joined	19:22
	← stamina left	19:22
onethfour	examples are very helpful to me	19:22
	← d2dchat left	19:22
onethfour	So i did git mv oldfile.php newfile.php and now the file is gone completely!!	19:22
	what the heck??	19:22
	→ BSaboia joined	19:23
	→ Mode-M joined	19:23
	← x2z left	19:23
cbreak	it should now be called newfile.php	19:23
onethfour	its not	19:23
	its gone	19:23
	and it says taht the file is deleted when i do git status	19:23
canton7	it'll show 'oldfile.php' as being deleted, because it has been	19:24
	a mv is a delete of the old file,and an add of the new one	19:24
	← eddyb left	19:24
	← Sendoushi left	19:24
canton7	there should be a corresponding 'add' for newfile.php	19:25
	← MrKeys88 left	19:25
	→ MrKeys88 joined	19:25
canton7	sure you didn't type 'rm' instead of 'mv'? got your shell history you can pastie?	19:26
	→ Pyrogerg joined	19:26
	→ multiverse joined	19:26
	→ andrewSC joined	19:27
	← MrKeys88 left	19:27
andrewSC	hi all	19:27
	could someone recommend me a good gui mergetool?	19:27
	Guest58300 → Freeaqingme	19:27
canton7	OS?	19:27
andrewSC	i use diffmerge now and it's on my last nerve	19:27
	os x	19:27
canton7	hmm not sure about mac. I like p4merge for windows - not sure if there's a mac version. Have a look at the supported tools in man git mergetool	19:28
gitinfo	the git-mergetool manpage is available at http://jk.gs/git-mergetool.html	19:28
	← cmpera left	19:28
	→ Wilduck joined	19:28
tang^	andrewSC: Kaleidoscope?	19:29
	→ mattmcclure joined	19:29
	← LuxuryMode left	19:30
	← sq-one left	19:30
	→ khushildep joined	19:30
	← khushildep left	19:30
	← blackstar257 left	19:31
	→ reim joined	19:31
breakingthings	I have branch foo that branched from branch bar. Branch bar has new commits. How can I make branch foo also have these changes (preferably by changing foo's parent commit to the newest one)	19:31
	→ sq-one joined	19:32
	← mattmcclure left	19:32
delianides	andrewSC: +1 Kaleidoscope	19:32
canton7	breakingthings, so you want to rebase? has branch foo been pushed anywhere?	19:32
	→ ngladitz joined	19:32
	← ngladitz left	19:32
	← NanoArrow left	19:33
cbreak	andrewSC: OS X has one that comes with Xcode	19:33
breakingthings	canton7: no it has not. and yes, I think I do, but I've only used rebase for reordering commits on a single branch yet so I don't know how.	19:33
	→ shadok_ joined	19:33
	→ y_nk joined	19:33
	← shadok_ left	19:33
	→ shadok_ joined	19:33
	← y_nk left	19:33
	→ y_nk joined	19:33
canton7	breakingthings, cool. git checkout foo; git rebase bar	19:33
cbreak	andrewSC: it's the default, so you don't have to configure anything	19:33
breakingthings	and this will keep foo branch?	19:33
canton7	that was rebase in its original incantation, before -i was introduced	19:34
	← Muad`Dib left	19:34
breakingthings	well, keep both branches in tact, that is.	19:34
canton7	yup, rebase rebase creates a new set of commits, then moves the branch ref to point to the new version of the branch	19:34
	→ LuxuryMode joined	19:34
breakingthings	alright, thanks.	19:34
canton7	so any commits which are referred to by any other branches/tags stick around	19:34
	← mizu_no_oto left	19:35
breakingthings	cool, didn't realize that's how that worked.	19:35
	→ cmpera joined	19:35
	→ sparq_ joined	19:35
	← infynyxx left	19:35
	← sparq_ left	19:36
	← bbloom left	19:36
	→ rat_thing joined	19:36
	→ sparq_ joined	19:36
	→ sparq__ joined	19:36
	← sparq_ left	19:37
	← sparq__ left	19:37
	→ d2dchat joined	19:38
	→ bbloom joined	19:38
	← worblehat left	19:38
	→ InvidFlower joined	19:38
	→ JohnBat26 joined	19:38
canton7	breakingthings, btw, rebase -i does the same with regard to changing what a branch is based on. it's just the common use-case involves rewriting a branch to sit on the same commit that it was sitting on before	19:38
	← mFacenet left	19:38
breakingthings	yeah	19:38
	thanks	19:39
	→ fragamus joined	19:39
	→ sparq__ joined	19:40
	→ sparq_ joined	19:40
	← tallest_redd left	19:40
	← minim left	19:40
	← zacts left	19:40
	→ maslowbeer joined	19:40
banzounet	Hey guys, If I pused a commit to github with api key, how can I remove them from my commit history?	19:40
	← sparq__ left	19:40
banzounet	(reference to my api key in certains files)	19:41
	← el3ctron_ left	19:41
	← sparq_ left	19:42
lov	banzounet: !filter-branch	19:42
	→ zacts joined	19:42
lov	.... really?	19:42
	argh	19:42
	!rewrite	19:42
gitinfo	[!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://goo.gl/waqum	19:42
	→ Sendoushi joined	19:42
Norrin	I'm doing something wrong here?	19:42
	$ git checkout -p 1330e660bee1fb43d2e0cc5139532781b172c27b BurstlyPlugin.jar	19:42
	No changes.	19:42
	$ git checkout -p ec12e9b79ac5afb1313a0ad50a40f0d69e8d76ef BurstlyPlugin.jar	19:42
	No changes.	19:42
	→ mklappstuhl joined	19:42
	→ sparq_ joined	19:42
	→ sparq__ joined	19:43
Norrin	btw, the file is different between the two commits	19:43
lov	that depends, is- oh.	19:43
	← sparq__ left	19:43
	→ el3ctron joined	19:43
	← apipkin left	19:43
	→ tagava joined	19:43
	→ apipkin joined	19:43
	→ tagava\|2 joined	19:43
	← babilen left	19:45
	← sparq_ left	19:45
banzounet	lov: Thanks	19:45
	← rpg_ left	19:45
	← luxgeek left	19:46
	← jfinn left	19:46
jrnieder	Norrin: what does 'git diff --stat 1330e6 ec12e9b7 -- BurstlyPlugin.jar' say?	19:47
	← whunt left	19:47
	→ arrosado joined	19:47
	→ mizu_no_oto joined	19:49
	← asQuirreL left	19:49
	← Sendoushi left	19:49
	→ sparq_ joined	19:49
	← sparq_ left	19:49
	→ ngladitz joined	19:49
	← ngladitz left	19:49
	→ sparq_ joined	19:50
	← sparq_ left	19:50
	← a-l-e left	19:50
	← chuckharmston left	19:51
	← dwmw2_gone left	19:51
	← hyperair left	19:52
Norrin	jrnieder, no output	19:52
	strange, because github is showing the file changed on that ec12 commit	19:52
	→ willwh joined	19:52
Norrin	jrnieder, https://github.com/burstly/BurstlyUnityPlugin/commit/ec12e9b79ac5afb1313a0ad50a40f0d69e8d76ef	19:53
	→ BiggFREE joined	19:53
	← starfox21 left	19:53
	→ infynyxx joined	19:53
	← byprdct left	19:54
	← d2dchat left	19:54
	← Soliton left	19:54
	→ dwmw2_gone joined	19:56
	← fragamus left	19:57
	← Jackneill left	19:57
	jrnieder looks	19:58
	← rfoust left	19:58
andrewSC	tang^, delianides: maybe free/open source? I'd like to exhaust free options before I pay. Not that I have a problem with paying but why pay if something else I see fits the bill just fine, you know?	19:59
	← OOPMan left	20:00
	→ Soliton joined	20:00
	→ danicampa901 joined	20:00
jrnieder	Norrin: BurstlyPlugin.jar is in a subdirectory	20:00
	Norrin: so perhaps you want 'git diff --stat 1330e6 ec12e9b7 -- "**/BurstlyPlugin.jar"'	20:00
	Norrin: likewise for the checkout command	20:00
	← mklappstuhl left	20:01
	← bamj0rb left	20:01
jrnieder	that said, the message from checkout was kind of unhelpful	20:01
	← reim left	20:01
jrnieder	it would be better to notice that there were no special characters in the filenames and give a hint that there's no such file	20:02
	→ minim joined	20:02
	→ delinquentme joined	20:02
delinquentme	made a commit w too many files ... how can I undo JUST the commit ... so that I can redo the commit and select the specific files that SHOULD be added	20:03
jrnieder	delinquentme: did you commit with or without "-a"?	20:03
	→ rpg joined	20:03
jrnieder	to keep the index as-is but just undo the commit, you can do "git reset --soft HEAD^"	20:04
_ikke_	delinquentme: git reset HEAD^	20:04
	← DoniRS left	20:04
	→ stamina joined	20:04
	← surfdue left	20:04
_ikke_	That undoes the last commit, but keeps the changes	20:04
	→ aard_ joined	20:04
delinquentme	git add -A	20:04
	→ pr0ton joined	20:04
jrnieder	to undo the commit and reset the index (but keep your changes in the worktree), "git reset HEAD^" does the trick, like _ikke_ said	20:04
	← wrouesnel left	20:04
jrnieder	!reset is a nice quick-reference card :)	20:05
gitinfo	A good resource explaining git-reset is http://git-scm.com/2011/07/11/reset.html	20:05
	→ sparq_ joined	20:05
jrnieder	gah, that's the wrong one	20:05
	← sparq_ left	20:05
	jrnieder hunts it down on jk.gs	20:05
	← maslowbeer left	20:05
	← willwh left	20:05
	← Soliton left	20:05
jrnieder	!resets	20:05
gitinfo	tl;dr of man git-reset: --soft moves HEAD, --mixed moves HEAD+index, --hard moves HEAD+index+work-tree, -- foo.txt will !unstage	20:05
	→ howlymowly_ joined	20:05
jrnieder	there it is :)	20:05
	← howlymowly left	20:05
	← Mode-M left	20:06
	→ Devastator_ joined	20:06
	→ ngladitz joined	20:07
	← ngladitz left	20:07
	→ ngladitz joined	20:07
	← ngladitz left	20:07
Norrin	jrnieder, well I was in the dir with the file, so no additional path was needed	20:08
	→ sparq_ joined	20:08
	← mribeirodantas left	20:08
	← Devastator left	20:08
Norrin	but i figured it out. the file is in two places. it didn't change in the place I was lookijng. but the commit history shows the other location is where the change happened	20:08
	→ chuckharmston joined	20:08
	← sparq_ left	20:08
	→ surfdue joined	20:09
	← delinquentme left	20:09
	→ worblehat joined	20:11
	← FernandoBasso left	20:11
	← Devastator_ left	20:11
	→ Devastator_ joined	20:11
	Devastator_ → Devastator	20:11
	← Robarelli left	20:12
	→ MacWinner joined	20:13
	→ Robarelli joined	20:14
	← cmpera left	20:15
	→ mattmcclure joined	20:15
	← TheMoonMaster left	20:16
	→ peroht joined	20:17
	→ cmpera joined	20:17
	→ pepper_chico joined	20:18
	← lifechamp left	20:18
	→ arvind_khadri joined	20:18
	→ SHyx0rmZ joined	20:18
	← h0bbit_ left	20:19
	→ TheMoonMaster joined	20:19
	← atc3030 left	20:19
	← iamjarvo left	20:20
	← s0ber left	20:20
	→ iamjarvo joined	20:21
	← iamjarvo left	20:21
	← skorgon left	20:21
	→ iamjarvo joined	20:22
	→ whunt joined	20:22
	→ s0ber joined	20:22
	→ sparq_ joined	20:22
tang^	andrewSC: oh, yeah, I forgot how much that one was.	20:23
	andrewSC: they do have a 15 day trial, though. you might find it worth it	20:23
	→ ngladitz joined	20:24
	→ mozzie_ joined	20:25
	→ beastd joined	20:25
	→ Sycokinetic joined	20:25
	→ maslowbeer joined	20:26
	← JohnBat26 left	20:26
Sycokinetic	alright, i need more newbie help. i'm trying to push my project to github, but i keep getting a permission denied error :c	20:28
canton7	Sycokinetic, output of 'git remote -v'?	20:28
	→ willwh joined	20:28
	← willwh left	20:28
	→ willwh joined	20:28
	← mozzie_ left	20:28
	← cmpera left	20:28
	→ DanielHollands joined	20:29
Sycokinetic	canton7:	20:29
	origin git@github.com:Sycokinetic/Random-Stuff.git (fetch)	20:29
	origin git@github.com:Sycokinetic/Random-Stuff.git (push)	20:29
	← sparq_ left	20:29
canton7	looks sensible. If you push, do you get a password prompt?	20:29
	→ pr0t joined	20:29
Sycokinetic	canton7: nope	20:29
	← axl_ left	20:29
canton7	exact error?	20:30
	sounds like a key problem	20:30
	← LimeBlast left	20:30
	← zivester left	20:30
p3lim_	Is it possible to specify a specific directory or a file when embedding other repos with gitmodules?	20:30
	p3lim_ → p3lim	20:30
	← saml left	20:31
Sycokinetic	canton7: here's the pastebin http://pastebin.com/U0vb6dcL	20:31
	← aard_ left	20:31
	→ JohnBat26 joined	20:31
canton7	Sycokinetic, you set up your keys right?	20:31
Sycokinetic	canton7: i'm honestly not sure :/	20:31
canton7	what did you do to set up the keys?	20:32
	← maslowbeer left	20:32
	← Hetu left	20:32
Sycokinetic	canton7: errrr... i'm not even sure i did anything honestly. the first time i ran the command, bash gave me the RSA key and asked if i wanted to continue connecting	20:33
	← grilix left	20:33
	← tatsuya_o left	20:34
	← willwh left	20:34
canton7	Sycokinetic, https://github.com/settings/ssh - have you added a key there?	20:34
Sycokinetic	canton7: ah! no... i have not	20:34
	→ Soliton joined	20:34
canton7	that's very, very important ;)	20:34
	read the two guides linked at the top of that page	20:34
	→ tatsuya_o joined	20:35
	← chrisss404 left	20:35
	→ willwh joined	20:35
	← willwh left	20:35
	→ willwh joined	20:35
	→ grilix joined	20:35
	← grilix left	20:36
Sycokinetic	canton7: is there any particular difference between the ssh and https methods?	20:36
canton7	Sycokinetic, with https, you need to enter your password on every push, or store your github password in plaintext somewhere	20:37
	→ ingof joined	20:37
canton7	with keys, you can always have a password-less push, or use ssh-agent to cache your key's passphrase after you've entered it once	20:37
	also with keys you can revoke access for particular keys	20:37
Sycokinetic	canton7: ahhhh	20:37
canton7	but https can often get through firewalls with ssh blocks... basically, ssh is preferable, if it works	20:37
	→ msheiny joined	20:37
	← mizu_no_oto left	20:38
Sycokinetic	canton7: right, i used to have a similar ssh setup for my pc's here	20:38
canton7	s/with/which	20:38
	← tatsuya_o left	20:39
Sycokinetic	... this is the same ssh key guide i used to set that up too XD	20:39
canton7	cools	20:39
	← arvind_khadri left	20:40
	→ circ-user-5rMaA joined	20:40
	← Soliton left	20:41
	← circ-user-5rMaA left	20:41
	→ JosephFerano joined	20:41
	→ nt4cats joined	20:41
	← dreiss_ left	20:42
	→ chrisss404 joined	20:42
	← willwh left	20:42
	← zacts left	20:42
	→ victorbjelkholm joined	20:43
nt4cats	I am looking for a website I found a while back. It was a list of git problems or challenges broken down by their difficulty. I've been asking the Google for 15 minutes and haven't been able to find it again ...	20:43
canton7	nt4cats, is it that interactive introduction to branching thing?	20:43
nt4cats	Let me try that in Google and see what it comes back with	20:44
onethfour	to scriptomaniacs: is there a way to get autocompletion within another branch, like FOOBAR:path/to/file	20:44
	→ zivester joined	20:44
canton7	nt4cats, I'm thinking of http://pcottle.github.io/learnGitBranching/	20:44
	→ axl_ joined	20:44
	← mattmcclure left	20:44
	← LukeKetley left	20:45
	→ tvw joined	20:45
nt4cats	canton7: that site looks pretty good, but it isn't the one I was thinking of. I know I'm doing a terrible job of describing the site (which is probably why Google can't help me either)	20:46
canton7	yeah, I don't think I've seen the one you're trying to describe, I'm afraid	20:46
	← rat_thing left	20:46
nt4cats	canton7: I figured it couldn't hurt to try here, thanks for jogging your brain and offering a link	20:46
	→ \|Freeman\| joined	20:47
	→ reim joined	20:47
	→ starfox21 joined	20:47
canton7	someone else might know what you're talking about :P	20:47
	→ Soliton joined	20:47
	← nt4cats left	20:47
	← rednaks left	20:47
	← geekbri left	20:47
	← starfox21 left	20:47
osse	onethfour: yes. zsh	20:48
	← maroloccio left	20:48
	← TheBrainWork left	20:50
	← mishok13 left	20:50
	← chuckharmston left	20:51
	← iamjarvo left	20:51
	→ chuckharmston joined	20:51
Sycokinetic	canton7: thar! now it's working. thank you :3	20:52
onethfour	osse, what about bash?	20:52
	← pr0ton left	20:53
	→ dreiss_ joined	20:53
	→ Octalot joined	20:54
	← Soliton left	20:54
canton7	Sycokinetic, coolio! With gitolite you get a password prompt if you don't have your keys set up right, and for some reason I assumed github would do the same (though it makes sense that they've bothered to configure that away), hence the slight red herrings at the beginning :P	20:55
	← BiggFREE left	20:55
onethfour	does zsh allow normal bash scripts? i'm so used to bash i dont think i could make the switch	20:55
p3lim	Is it possible to specify a specific directory or a file when embedding other repos with gitmodules?	20:57
	← acharlieh left	20:57
	→ Mars joined	20:57
bekks	onethfour: Its a different shell. If zsh doesnt support all bash constructs, it does not support bash scripts fully.	20:57
	← raonyguimaraes left	20:57
	Mars → Guest53254	20:57
	→ infynyxx_ joined	20:58
tang^	canton7: you only get that password prompt if your sshd is set up to allow password connections.	20:58
	← andrewSC left	20:58
osse	onethfour: there is a git completion script for bash included with git itself. I don't know if it has that feature	20:59
banzounet	c	20:59
osse	onethfour: running a bash script (or a zsh script) is typically do different than running a python script or a perl script (or ruby or tcl)).	20:59
	onethfour: if the script has #!/bla/bla/bla line a process will be run for it so it doesn't matter what shell you use to execute it	21:00
SamB	onethfour: when I hear "bash script", I think "#!/bin/bash", which zsh obviously doesn't have to do anything about at all ...	21:00
grawity	osse: if it also has the +x permission	21:01
SamB	darn osse beat me by 30 seconds	21:01
	← infynyxx left	21:01
	← apipkin left	21:01
	← kmctown left	21:01
	← msheiny left	21:01
osse	grawity: sure. But in case it doesn't: bash path/to/script :P	21:01
	→ pr0ton joined	21:02
osse	SamB: my chronograph says 31 second. hah!	21:02
	→ zacts joined	21:03
	→ tpollard joined	21:03
	→ acharlieh joined	21:03
	← ingof left	21:04
	→ apipkin joined	21:04
	← panpainter left	21:04
	← Sycokinetic left	21:04
	→ iderik joined	21:06
	← rainbyte16 left	21:07
	← AzizLight left	21:08
	← King_Hual left	21:08
	→ sgu joined	21:08
	→ King_Hual joined	21:08
	← King_Hual left	21:09
onethfour	oh	21:09
	turns out git auto completion DOES work for BRANCH:	21:09
	how do io copy a whole directory from branch A to current branch	21:09
	← jason237 left	21:09
	← cannonball left	21:09
	← staafl left	21:10
PerlJam	onethfour: git checkout branch -- path/to/dir	21:10
	(that doesn't add it to the branch automatically, but you can figure that part out)	21:10
onethfour	but i want to specify a destination folder	21:10
	→ staafl joined	21:10
	← BSaboia left	21:11
PerlJam	oh. Hmm.	21:11
	← albel727 left	21:11
osse	onethfour: git archive branch -- path/to/dir \| tar -xf - -C desti/nation	21:11
	← staafl left	21:11
	← arrosado left	21:12
onethfour	that's quite a line	21:12
	i'll give it a shot thanks	21:12
	→ arrosado joined	21:12
	→ staafl joined	21:12
PerlJam	osse++	21:12
	→ OMGOMG joined	21:12
	← snuffeluffegus left	21:12
	→ asQuirreL joined	21:12
	← arrosado left	21:12
osse	\o/	21:12
	→ willwh joined	21:12
	← freeed left	21:13
osse	maybe something like cd destinaton; GIT_DIR=path/to/.git git checkout branch -- path/to/dir could work too.	21:13
	But then git-archive is simpler, imho	21:13
	← peroht left	21:14
	← _iron left	21:14
	→ LuxuryMo_ joined	21:14
	← dhruvasagar left	21:15
onethfour	git archive didnt work, i resorted to cat-file	21:16
	→ davemq joined	21:16
osse	lies	21:16
onethfour	fatal: current working directory is untracked	21:16
	tar: This does not look like a tar archive	21:16
	tar: Exiting with failure status due to previous errors	21:16
	→ tinti joined	21:16
osse	where did you run it?	21:16
onethfour	there was probably a typo in your answer but i was to unknowledgable to fix it	21:17
	→ Soliton joined	21:17
	← delianides left	21:17
onethfour	git archive HOTFIX -- src/foo/bar/ \| tar -xf - -C .	21:17
	← LuxuryMode left	21:17
onethfour	is what i ran	21:17
osse	onethfour: you have to run it from the git repo dir	21:17
	← gzg left	21:17
onethfour	i was in a git repo	21:18
	← Dougie187 left	21:18
osse	hmm. works here.	21:18
	← y_nk left	21:18
	→ breck7_ joined	21:18
tang^	might have to define 'the git repo dir'	21:18
	→ y_nk joined	21:18
osse	onethfour: was src/foo/bar relative to where you were?	21:19
	← breck7 left	21:19
onethfour	osse, no it was relative to the root of the branch	21:19
	errr root of the repo	21:19
	→ koda joined	21:20
onethfour	i'm on bash though	21:20
	is that why?	21:20
	→ gzg joined	21:20
	← breakingthings left	21:20
osse	onethfour: make it relative to the current dir	21:20
	bash has nothing to do with it	21:20
	← LuxuryMo_ left	21:20
onethfour	osse, that doesnt make sense, because the path doesnt even exist in the current branch, only in the other one i'm trying to get it from	21:21
joshc	passing '-f -' to tar is a bit redundant	21:21
tang^	onethfour: but what if git is using your current location as its starting point to find the files in the other branch?	21:22
osse	onethfour: I see, I get the same message now	21:22
onethfour	tang^, oh yeah that's a possibility	21:22
	← asQuirreL left	21:23
	← willwh left	21:23
	← y_nk left	21:23
osse	onethfour: if you're in the root of the repo it'll work	21:23
	This is a bug, imo	21:23
onethfour	but....?	21:23
osse	no but	21:23
onethfour	i can't pass it relative path?	21:23
osse	apparently not	21:23
onethfour	well you can make a script to CD to the repo root	21:24
osse	my instinct was that it would behave like tang^ said. I think many other git commands do.	21:24
	← infynyxx_ left	21:24
osse	onethfour: luckily that's simple: cd "$(git rev-parse --show-toplevel)"	21:25
	→ NoiseEee joined	21:25
	→ mizu_no_oto joined	21:25
onethfour	cool	21:26
	← Soliton left	21:26
	← worblehat left	21:26
	← zarubin left	21:26
NoiseEee	hey folks, if i'm working on a branch, and I have a bunch of commits on it, and then I want to merge that into Master... whats the merge option to NOT include the log of all branch commits?	21:26
	← mizu_no_oto left	21:26
osse	onethfour: I see you want to extract to the current dir (-C .). Is that dir ideally outside the repo completely?	21:26
	← gzg left	21:27
osse	NoiseEee: git merge --squash	21:27
NoiseEee	thanks osse	21:27
osse	NoiseEee: that'll make one jumbo-commit at the tip of master containing everything you did in the other branch.	21:27
NoiseEee	osse: cool thanks :)	21:27
osse	Are you sure that's what you want? If you delete the other branch the history is gone	21:27
NoiseEee	osse: prettty sure, is that an unusual request?	21:28
	← atSetKey_ left	21:28
osse	No	21:28
tang^	looking at some of the commits i've made, yeah, I'd say I want to hide all those half-witted attempts and make it look like I did everything in one clean pass :-P	21:28
onethfour	honestly though it seems like a bit of a hackery, i just want to cp -r a folder from a git branch to a/path/on/my/system/, git checkout would be pretty annoying	21:28
	→ gzg joined	21:29
	← NoiseEee left	21:29
osse	onethfour: (cd repo; git archive branch -- path) \| tar -xf -	21:29
	that could work too	21:29
	← donmichelangelo left	21:30
tang^	onethfour: have you looked at man git-readtree yet?	21:30
onethfour	you mean all on 1 line ?	21:30
	→ NanoArrow joined	21:30
onethfour	tang^, not recently	21:30
tang^	man git-read-tree	21:30
gitinfo	the git-read-tree manpage is available at http://jk.gs/git-read-tree.html	21:30
	→ donmichelangelo joined	21:30
osse	onethfour: it doesn't have to be on one line. But it's a pipeline so in my opinion it looks nicer on one line as long as it isn't very long	21:31
	← rainbyte left	21:31
tang^	I've been using a tip to merge repositories together that reads the files from one branch into a subdirectory elsewhere	21:31
	→ rainbyte joined	21:31
osse	onethfour: the point is, it doesn't matter what the current directory is. As long as you know the path to the repo and the path to the destination everything's possible	21:31
	→ infynyxx joined	21:31
	→ SporkWitch joined	21:32
onethfour	osse, well ideally i would alias it or make another script	21:32
	branchcp	21:32
	← apipkin left	21:32
	← irqq left	21:33
osse	sure, no problem	21:34
	as of 1.8.4 Git has a -C switch akin to tar. that'll make it even easier	21:34
	← ojacobson left	21:34
	← tbmiller left	21:34
	→ frgtn joined	21:34
	← psynor left	21:35
	→ tbmiller joined	21:35
	← tbmiller left	21:35
onethfour	i'm on debian which is liek 1.7.....	21:36
	→ tallest_red joined	21:36
onethfour	s/liek/like	21:36
jrnieder	onethfour: can you say again what you're trying to do?	21:36
	→ lifechamp joined	21:36
jrnieder	onethfour: is it !deploy or similar?	21:36
gitinfo	onethfour: Git is not a deployment tool, but you can build one around it(in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/the-list-and-irc/deploy.html	21:36
SporkWitch	onethfour: sounds like a job for make && make install ^^	21:37
onethfour	i was just trying to copy a directory from another branch into my current dir	21:37
jrnieder	onethfour: okay, that's how you were planning to do it	21:37
	but it still doesn't tell me the goal	21:37
onethfour	no i was going to start coding but i wanted a starting point	21:37
jrnieder	okay, the idea is you have a repo	21:37
onethfour	so i was going to use a similar dir to start from	21:37
jrnieder	and you want to copy a directory to start with on another project?	21:37
onethfour	I'm sorry i dont have time to discuss i got to get going	21:38
tang^	judging from the branch named HOTFIX, I'd guess he's after just a part of that branch	21:38
jrnieder	I'd usually do that by checking out the branch I am interested in in the repo with the first project	21:38
	and then "cp -R" into the second one	21:38
	← InvidFlower left	21:38
onethfour	yeah thats the ~slow~ way tod o it JrCs_	21:38
	← longbeach left	21:38
onethfour	;)	21:38
	jrnieder, but you'd have to use a temp dir	21:38
jrnieder	I guess that depends on the context	21:38
onethfour	or i guess you could float the changes over	21:38
jrnieder	e.g., if I want to copy some files from glibc into my-special-project	21:39
	then I probably already have a clone of glibc around	21:39
	there's no temp dir involved.	21:39
tang^	jrnieder: this is between branches of a single project	21:39
jrnieder	in that case, it's not clear to me what it means to take a subdirectory of one branch as a starting point for another	21:39
	in any "how do I" question, there are two ways to answer	21:40
	← DonRichie left	21:40
tang^	one branch has a directory that isn't contained in another and the desire is to copy just that directory into another branch	21:40
	← tinti left	21:40
jrnieder	one is to figure out the goal and figure out a good way to do that	21:40
	the other is to ignore the goal and answer it as a technical question	21:40
SporkWitch	jrnieder: maybe not the cleanest approach, but couldn't you just fork the branch, delete what you don't need, and if/when you merge back later you just make sure to keep the files from main branch that you removed from your working branch?	21:41
	SporkWitch new to git, so assumes there's a cleaner way than this	21:41
jrnieder	SporkWitch: maybe? It's not clear to me what onethfour is trying to do	21:41
	once I know, I can give a command.	21:41
	like many people have already above :)	21:41
SporkWitch	jrnieder: fair enough, just testing my own understanding; i learn a lot from looking at what others are trying to do heh	21:41
	→ Georg3 joined	21:42
onethfour	honestly though there should be something like cp BRANCHB:path/to/file .	21:42
jrnieder	SporkWitch: if the idea is 'I want a new branch based on a subdirectory of an existing branch, and I want the history to record that', I think your method makes sense	21:42
	onethfour: depending on what you're trying to do, maybe there should be	21:42
	it's still not clear at all to me what you want the effect to be	21:43
	if you're just trying to grab a file from another branch, the usual way is "git show BRANCHB:path/to/file"	21:43
onethfour	sorry i'll get in trouble if i spend more time on irc, you can try reading the logs above	21:43
jrnieder	you can redirect the output to a file	21:43
	onethfour: ok :)	21:43
	← Amandil left	21:43
onethfour	and yes that works for files, but what about directories	21:43
	← el3ctron left	21:43
onethfour	ack really leaving	21:43
	SporkWitch does not miss jobs that were that draconian	21:44
jrnieder	onethfour: if you're trying to grab a directory from another branch, the usual way is "git archive BRANCH:path/to/directory"	21:44
	you can pipe the output to 'tar'	21:44
tang^	which has been offered	21:44
onethfour	can someone make a "git cp" command	21:44
	to do just that	21:44
jrnieder	onethfour: sure, you can :)	21:44
	← minim left	21:45
onethfour	would take me proabbly 5 hours	21:45
SporkWitch	jrnieder: ++	21:45
jrnieder	if you make a script in the $PATH named git-cp, then running "git cp" will do what you're asking for	21:45
onethfour	you're kidding	21:45
	that's awesome!	21:45
SporkWitch	onethfour: 5 hours to save how many future hours for how many other people out there? :)	21:45
osse	onethfour: my pipeline above works just fine	21:45
jrnieder	if you want it to be included in the standard installation of git, [email@hidden.address] is the place to suggest it	21:45
	← howlymowly_ left	21:46
	SporkWitch is totally the guy from XKCD writing the mouse-wiggle script and the arbitrary-condiment-passing script...	21:46
	→ g3orge joined	21:46
	→ atSetKey joined	21:47
	→ jon_r joined	21:47
	← jon_r left	21:47
	← glosoli left	21:47
	← jnavila left	21:48
	← axl_ left	21:49
	→ breck7 joined	21:49
	← breck7_ left	21:50
	→ axl_ joined	21:50
	← natechan left	21:50
	→ zoli joined	21:50
	← pr0t left	21:51
	→ h0bbit_ joined	21:51
	← Misan left	21:51
	← davemq left	21:53
	→ willwh joined	21:53
	← willwh left	21:53
	→ willwh joined	21:53
	← leeN left	21:53
g3orge	how can I make git grep look into submodules as well??	21:54
	← axl_ left	21:55
	← Sigma left	21:55
m0viefreak	g3orge: you could probably do something along 'git submodule foreach --recursive git grep ...'	21:56
	← TomyLobo left	21:56
m0viefreak	but the output you are getting will not indicate from what submodule the results are coming from	21:56
	← beastd left	21:57
	← infynyxx left	21:57
	→ jason237 joined	21:57
	→ rat_thing joined	21:57
g3orge	yeah. or use plain grep. I guess I am asking why doesn't git grep recurse through submodules...	21:58
m0viefreak	why should it?	21:58
	→ infynyxx joined	21:58
m0viefreak	after all, the main repo and the submodules are kind of intedepdent	21:58
	→ minim joined	21:59
	← reim left	21:59
	← infynyxx left	21:59
	← jerius left	22:00
	← kcm1700 left	22:00
	→ frgtn_ joined	22:02
	← NanoArrow left	22:02
	← ngladitz left	22:02
	→ eletuchy joined	22:02
	← rat_thing left	22:02
g3orge	hmm, yeah I guess	22:03
	← acharlieh left	22:03
	← h0bbit_ left	22:03
	← rpg left	22:04
	← sdx23 left	22:04
	← frgtn left	22:04
	→ kcm1700 joined	22:04
	← pr0ton left	22:04
	← valgrind left	22:05
	→ sdx23 joined	22:05
	← nemesit\|znc left	22:06
	→ Fandekasp joined	22:06
	← gitpicker left	22:07
	← kcm1700 left	22:07
	→ robbyoconnor joined	22:08
	→ nemesit\|znc joined	22:08
	→ gitpicker joined	22:09
	→ _nova joined	22:09
	→ kermit joined	22:09
	← multiverse left	22:09
	→ pr0ton joined	22:09
jrnieder	g3orge: forward-port the patches from https://github.com/jlehmann/git-submod-enhancements/commits/grep-recursive?	22:11
	← nova_ left	22:12
	→ nova_ joined	22:12
	→ raonyguimaraes joined	22:12
Vampire0	davividal, svn2git is a gazillion faster as git-svn has to do a manual svn checkout for each and every revision to build up the repository, the history that comes out is more correct as far as I remember and especially you have much more control over how the conversion is done, like what path in SVN lands in which path in which Git repository, and so on, like e. g. splitting one big SVN repo with many modu	22:12
	les into several small Git repositories and so on.	22:12
	→ kcm1700 joined	22:13
	← _nova left	22:14
	← pantsman left	22:14
	→ HumanSky joined	22:15
	← zhost left	22:17
	← Swayz left	22:17
	→ Soliton joined	22:18
	← harobed_ left	22:19
sgu	Hi all, when I do `git pull -s recursive -X theirs` and I get this message "XX file would be overwritten by merge, commit your changes or stash them before you can merge". What should I do if I just want it to be overwritten.	22:19
	← kcm1700 left	22:19
	→ functorfun joined	22:20
	functorfun → funfunctor	22:20
	← jaequery left	22:20
osse	sgu: git checkout HEAD -- file	22:20
	→ BiggFREE joined	22:22
	← grampajoe left	22:24
	← Soliton left	22:24
	← dreiss_ left	22:24
	→ Gamecubic joined	22:24
	← boombatower left	22:25
	→ freestyl3r joined	22:26
	→ DanielKarp joined	22:27
	← hholtmann left	22:27
	→ panpainter joined	22:28
	→ Archweasel joined	22:28
	← willwh left	22:29
	← v0n left	22:29
sgu	osse: is there an option of `git pull` let me save this step?	22:30
	→ boombatower joined	22:30
	→ Soliton joined	22:31
	→ purpleidea joined	22:31
	← zoli left	22:31
	← Guest29305 left	22:32
purpleidea	hey, something has gone wrong. i was working on master, i had some files added and some modified to staging. i did git stash and then git stash list. i then wrote a patch, and merged it. it is seen in my log. i then did git stash apply, and some of the earlier git add -p changes i've made aren't seen. help!	22:32
	→ dreiss_ joined	22:32
	← Destos left	22:33
iveqy	purpleidea: !repro	22:34
	← Huevos left	22:34
gitinfo	purpleidea: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.	22:34
	← druonysus left	22:34
	→ knuth joined	22:34
	→ druonysuse joined	22:34
knuth	hi guys.	22:34
gitinfo	knuth: 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.	22:34
	← phile left	22:35
	← vinc left	22:35
purpleidea	iveqy: any idea what happened in my particular scenario if i did something the wrong way ?	22:35
	→ willwh joined	22:36
	← normanrichards left	22:36
	← danicampa901 left	22:36
iveqy	purpleidea: I can't understand your explanation, that's why I asked you to repro	22:36
purpleidea	iveqy: ok	22:36
	← DanielKarp left	22:37
	← mitz_ left	22:37
	→ Narven joined	22:37
	→ Orbitrix joined	22:38
	→ bwreilly_ joined	22:39
purpleidea	http://fpaste.org/45956/14447751/	22:39
	→ bhutjolokia joined	22:39
purpleidea	iveqy: http://fpaste.org/45956/14447751/	22:39
	→ madewokherd joined	22:40
bhutjolokia	Git can detect file integrity problems, right?	22:40
	→ Guest53432 joined	22:41
	← HicksD left	22:41
purpleidea	bhutjolokia: git-fsck ?	22:41
	← mobileblue left	22:42
	← bwreilly left	22:42
bhutjolokia	purpleidea, no, git in general. If a bit gets corrupted somewhere it should know, should it not?	22:42
	→ mitz_ joined	22:42
	← Soliton left	22:42
grawity	yes, it should know	22:42
sayd	a/close	22:42
purpleidea	bhutjolokia: yes, see git-fsck	22:42
iveqy	purpleidea: but 00:40 -!- purpleidea [~james@unaffiliated/purpleidea] has joined #git	22:43
	purpleidea: sorry..	22:43
purpleidea	iveqy: ?	22:43
	→ jaequery joined	22:44
iveqy	purpleidea: What I can see from git status the exact same things are added to the index both before and after your stash. Is it so that part of stuffxyz isn't added	22:44
	?	22:44
	→ Droolio joined	22:44
grawity	hmm, interesting	22:44
	→ natechan joined	22:44
grawity	apparently git does not check if a blob it just read actually matches its hash	22:44
	oh, it does	22:44
purpleidea	iveqy: yes, exactly. part of my changes to stuffxyz aren't seen after the git stash apply. in fact, some of those changes included 'e' edits, so data seems to be lost.	22:44
	← rnsanchez left	22:44
grawity	just `git cat-file` doesn't, but things like `git show` still do.	22:45
	$ git show fd247ec09b6e1d5b4189b4fc1248ff898adb9a4f	22:45
	error: sha1 mismatch fd247ec09b6e1d5b4189b4fc1248ff898adb9a4f	22:45
	→ cotton joined	22:45
grawity	so yeah, bhutjolokia, git would know that.	22:45
bhutjolokia	Curious. I have a .jar file tracked in a git repo. The file began failing out of nowhere after a system reboot. Git sees no difference in the file. When I recompile the jar from source and replace it, git _does_ see a difference and the file runs.	22:45
	← sq-one left	22:45
bhutjolokia	How do I explain this?	22:46
	← jaequery left	22:46
grawity	Does it work if you delete the file from working tree, then check out a good copy from the repo?	22:46
	gitinfo set mode: +v	22:46
	← jason237 left	22:46
	← SHyx0rmZ left	22:46
fobius	What's the difference between (merging topic_branch and master and then deleting topic_branch) and (rebasing topic_branch onto master)?	22:47
iveqy	purpleidea: this is not good at all, please send this to the !list	22:47
gitinfo	purpleidea: [!mailing_list] The mailing list can be reached via [email@hidden.address] You don't need to subscribe to the list, you will always be put in cc on reply. Read archives at http://j.mp/gitlist	22:47
grawity	I think things like `git status` have some optimizations – e.g. they won't compare the contents if the timestamp is still the same	22:47
bhutjolokia	grawity, yes, it does.	22:47
purpleidea	iveqy: :(	22:47
bhutjolokia	How can there be no difference between a broken file and a running one?	22:47
grawity	fobius: you're basically asking "What's the difference between merging and rebasing in general", right?	22:47
purpleidea	iveqy: can you reproduce this issue ?	22:47
grawity	bhutjolokia: see what I said about optimizations	22:48
	→ bdavenport joined	22:48
bhutjolokia	grawity, Even git diff?	22:48
grawity	hmm, not sure	22:48
	→ normanrichards joined	22:48
iveqy	purpleidea: sorry, I don't have a possibility to reproduce rightnow	22:49
	← m0ltar left	22:49
	→ kcj joined	22:49
	→ DanielKarp joined	22:49
fobius	grawity, pretty much. I mean, from what I gather they produce the same end result on the branch being merged into/rebased onto. According to git-scm it causes the histories to differ but I'm confused as to how. It seems like in both situations git log would show the merged/rebased commit and then the commits on master prior to that commit.	22:49
bhutjolokia	grawity, you were right.	22:49
purpleidea	anyone have an idea how to fix this please? http://fpaste.org/45956/14447751/	22:49
jrnieder	purpleidea: use "git stash --keep-index"	22:50
	iiuc	22:50
	→ Soliton joined	22:50
knuth	Can someone point me to a good article or two that I can use to help convince my org to convert from vss to git?	22:50
bhutjolokia	Thanks guys.	22:50
purpleidea	jrnieder: at which point?	22:50
	← m8 left	22:51
purpleidea	jrnieder: at the git stash 'save' part ? is it too late to get my data back ?	22:51
jrnieder	purpleidea: ah, let me back up. are you trying to see what went wrong or to recover?	22:51
	okay, to recover	22:51
knuth	looking to compare to svn.	22:51
purpleidea	jrnieder: i want my data and changes back	22:51
grawity	from what little I know about VSS, "Git/SVN won't corrupt your data" seems a good argument..	22:51
	← HumanSky left	22:51
jrnieder	purpleidea: ok, do I understand correctly that what's in the worktree is right but it forget what was in the index?	22:52
	or is it something else you've lost?	22:52
knuth	lol, how about svn vs git.	22:52
	→ Siecje joined	22:52
purpleidea	jrnieder: yeah, well everything that i git add -p 'ed is gone. including changes made with git add -p 'e' (edits)	22:52
jrnieder	purpleidea: ok, good	22:52
Siecje	if I add a file to staging area why can't I use the command git diff on it?	22:53
jrnieder	I'd use "git fsck --lost-found"	22:53
	← willwh left	22:53
	→ multiverse joined	22:53
jrnieder	purpleidea: it will take some sifting --- sorry I don't have a better idea :/	22:53
knuth	my understanding is that git is going to be harder to use then svn.	22:53
jrnieder	purpleidea: (when you add something to the index, that forces git to record its content, which is good)	22:53
knuth	I want to go with git but need reasons on paper for people.	22:54
	→ AWinterman joined	22:54
jrnieder	purpleidea: (but there is nothing like a reflog for the index, so there's no record of which unreferenced blob in particular is the one you're interested in)	22:54
purpleidea	jrnieder: dangling blobs and other things... now what?	22:54
jrnieder	purpleidea: grep through the blobs :(	22:54
	purpleidea: they should be in .git/lost-found	22:54
	→ mobileblue joined	22:54
	← Pyrogerg left	22:55
knuth	.	22:55
	is git more easy to run on windows then svn?	22:56
	← epicowl left	22:56
	← Soliton left	22:56
jrnieder	knuth: they're both pretty easy to run on windows	22:56
grawity	these days both are equally easy	22:56
jrnieder	knuth: for svn lots of people like tortoisesvn	22:56
	knuth: (a shell extension)	22:56
	← g3orge left	22:56
jrnieder	knuth: there's no single standard gui tool like that which everyone loves for git	22:56
grawity	[although in the past, I used to hear that TortoiseGit was really inferior compared to TortoiseSVN]	22:56
	→ v0n joined	22:56
jrnieder	exactly, I think gitcheetah's a bit better	22:57
purpleidea	jrnieder: nope. none of those have any useful data	22:57
	← panpainter left	22:57
jrnieder	purpleidea: your file isn't in .git/lost-found/other/?	22:57
grawity	but either way – those are just GUIs	22:57
purpleidea	jrnieder: nope	22:57
	jrnieder: and not in .git/lost-found/commit either	22:57
jrnieder	purpleidea: that's strange. anything you added to the index would be in the object db	22:57
	purpleidea: meaning it would either show up in the lost + found or be referenced by some actual commit	22:58
purpleidea	git --version 0.31415	22:58
	just kidding git version 1.8.3.1	22:58
knuth	Do you guys keep a central repository on a server or just depend on developers boxes?	22:58
	→ Soliton joined	22:58
jrnieder	knuth: at work, central repo	22:58
purpleidea	jrnieder: beats me. i'm a little pissed. git stash has generally worked before. i haven't used it in a while, but still. wtf	22:58
grawity	I think most Git projects have some kind of a central repo	22:58
Siecje	if I add a file to staging area why can't I use the command git diff on it?	22:59
jrnieder	purpleidea: hm, let me try to reproduce it	22:59
purpleidea	jrnieder: sure, thanks	23:00
	→ jason237 joined	23:00
jrnieder	Siecje: I don't know. why can't you?	23:00
	← tang^ left	23:00
grawity	Siecje: you can, after you understand how `git diff` works	23:00
	by default, it compares the working tree with the staging area	23:00
purpleidea	jrnieder: any idea if git stash show can show the long output of the diff as opposed to the shortlog	23:00
grawity	if you add a file to the staging area, then there's no difference anymore	23:00
jrnieder	purpleidea: sure, git stash show -p	23:00
purpleidea	s/shortlog/diffstat	23:00
knuth	but pull requests are still handled by the hithub account?	23:00
grawity	you might be looking for `git diff --cached` or other options in the manpage	23:00
purpleidea	jrnieder: ah thanks	23:00
Siecje	grawity: how do I compare it to the last commited version	23:00
grawity	Siecje: man git diff	23:01
gitinfo	Siecje: the git-diff manpage is available at http://jk.gs/git-diff.html	23:01
knuth	on entiprise how do you deal with code promotion?	23:01
jrnieder	Siecje: git diff HEAD	23:01
Siecje	ty	23:01
jrnieder	knuth: mostly, give talks	23:01
	← DanielKarp left	23:01
jrnieder	knuth: or do you mean as part of release management?	23:01
	knuth: or advocacy?	23:02
knuth	not sure honestly. I have only played with git with some simple open source projects.	23:02
	tt-rss.	23:02
purpleidea	jrnieder: so looking through git stash show -p i don't see any of the individual changes that i had previously made with git add -p, (particularly the 'e' stuff which was what mattered the most)	23:02
	→ mFacenet joined	23:03
jrnieder	purpleidea: ok, let me walk through the same steps you did	23:03
	← Soliton left	23:03
knuth	I got the project made a change and made a pull request.	23:03
jrnieder	purpleidea: first you made some change to a file	23:03
	→ druonysus joined	23:04
jrnieder	purpleidea: then "git add -p", edited the file in the index that way	23:04
knuth	when you have a room of coders how is checking in code delt with... (yes i'm talking like a vss noob...)	23:04
jrnieder	purpleidea: then, "git stash"	23:04
	← Shaladdle left	23:04
knuth	does each programer make a pull request and the lead approve them?	23:04
	← iderik left	23:04
purpleidea	jrnieder: (i'm doing it too)	23:05
jrnieder	purpleidea: it looks like it does save the index along with everything else. I learn something new every day :)	23:05
	← g0bl1n left	23:05
jrnieder	purpleidea: now, "git stash pop"	23:05
	← mFacenet left	23:06
jrnieder	it unfortunately doesn't apply the change to the index	23:06
	if I do "git reset <stash commit>^2 -- ." then the change gets applied to the index	23:06
	this UI is crazy	23:06
	← druonysuse left	23:06
knuth	jrnieder: release	23:06
purpleidea	jrnieder: so afaict, you're seeing the same issue, correct ?	23:07
jrnieder	purpleidea: yeah, the stash records the old index but "git stash apply" doesn't reapply it	23:07
grawity	knuth: I have no experience at all, but from what I've heard, it really varies... some projects do pull requests – Github, Gerrit, Phabricator, or just email; others have a "review" repo to push to; yet others just push directly to the main repo.	23:07
purpleidea	jrnieder: will reset-ing to the stash fix this?	23:07
jrnieder	purpleidea: looks like 'git stash apply' takes a --index option	23:07
	← lavaman left	23:07
purpleidea	jrnieder: i don't know the ^2 -- . syntax	23:07
jrnieder	purpleidea: "git stash pop --index" would have done the trick	23:08
	why in the world that's not the default I don't know	23:08
purpleidea	jrnieder: it's a HUGE bug atm.	23:08
	sounds like mailing list material	23:08
jrnieder	purpleidea: UI bug, yes, I agree	23:08
	← jason237 left	23:08
jrnieder	purpleidea: working as intended, but the intention is wrong :)	23:09
purpleidea	jrnieder: agreed	23:09
knuth	grawity: Can you have repos that get auto pull aproved. The "review" repo?	23:10
	→ mdiers joined	23:10
	→ Risto joined	23:10
	← Hello71 left	23:11
	← boombatower left	23:11
jrnieder	knuth: for large projects, I like the way the kernel is maintained	23:11
	knuth: but lots of companies can't afford to have one person be full-time maintainer	23:11
purpleidea	jrnieder: what would be the correct git reset command to pull out the stash correctly on master?	23:11
	→ boombatower joined	23:12
grawity	knuth: sounds like a repo that just allows everyone to push to	23:12
jrnieder	knuth: and prefer to have developers make changes with some autonomy, possibly after code review	23:12
	→ Hello71 joined	23:12
jrnieder	purpleidea: do you still have the stash you were trying to apply?	23:12
knuth	our shop is a bit of a free for all..	23:12
purpleidea	jrnieder: yeah, i never pop-ed it.	23:12
jrnieder	purpleidea: then I'd do "git reset --hard" followed by "git stash apply --index"	23:13
	purpleidea: note that will make the worktree and index look exactly like HEAD before applying the stash	23:13
	← fobius left	23:13
jrnieder	purpleidea: so if you had local changes it wouldn't be a good thing to do	23:13
knuth	book suggestions I'm looking at pro git and the link.	23:13
	http://refcardz.dzone.com/refcardz/git-patterns-and-anti-patterns	23:13
purpleidea	<stdin>:322: new blank line at EOF.	23:13
	+	23:13
	warning: 1 line adds whitespace errors.	23:13
	error: The following untracked working tree files would be overwritten by merge:	23:13
	...	23:13
	Please move or remove them before you can merge.	23:14
	Aborting	23:14
	Index was not unstashed.	23:14
	← funfunctor left	23:14
grawity	knuth: that sounds like a simple Git or SVN setup then... a central repo where anyone can push anything. Easy to set up with Gitolite or whatever other tool you choose	23:14
jrnieder	purpleidea: ah, sure, "git reset --hard && git clean -nd && git clean -fd"	23:14
	← Siecje left	23:15
	← falconfox left	23:15
	→ tbmiller joined	23:15
	→ falconfox joined	23:15
	← victorbjelkholm left	23:16
purpleidea	jrnieder: i think i'll just remove some of those files manually, because some of those aren't tracked or saved anywhere...	23:16
	so the git clean would delete them i think.	23:16
knuth	grawity: cool thanks :) One more question for now (besides the request for book suggestion) can I hook github into this setup. We are taking over a project from an outside vendor. We will still use them but often will be updating the code ourselves.	23:16
	← pr0ton left	23:16
jrnieder	purpleidea: yeah	23:17
knuth	How do you suggest setting up that so we can push but they have to be approved/put in a testing/staging area.	23:17
	← nova_ left	23:17
	← joshsmith left	23:18
	← Wilduck left	23:18
grawity	knuth: I'm guessing you could create an 'organization' on Github, then you'll be able to give push access to a specific group, and everyone else will have to do pull-requests	23:18
purpleidea	jrnieder: well it seems things are almost back to normal. thanks for the chat. who wants to report this?	23:19
jrnieder	purpleidea: you can :)	23:19
	purpleidea: thanks	23:19
purpleidea	jrnieder: balls :(	23:19
	← yacks left	23:19
purpleidea	jrnieder: thank goodness for irssi history	23:19
	$ git stash apply --index	23:20
	<stdin>:322: new blank line at EOF.	23:20
	+	23:20
	warning: 1 line adds whitespace errors.	23:20
	Auto-merging ...	23:20
	i don't remember having a EOL problem before...	23:20
	← pepper_chico left	23:21
	→ mizu_no_oto joined	23:21
purpleidea	git d --cached --check	23:21
	← primitive left	23:21
purpleidea	file.foo 162: new blank line at EOF.	23:22
knuth	anyone useing https://enterprise.github.com/	23:22
	← frgtn_ left	23:22
purpleidea	but it's not there in the cache or the regular file	23:22
	→ Soliton joined	23:22
	→ frgtn joined	23:22
	→ mklappstuhl joined	23:23
	← boombatower left	23:23
	→ willwh joined	23:23
	← koda left	23:24
	→ luxgeek joined	23:24
knuth	grawity: could I have a local repo that is slave to the github repo so that if github is down we still are up?	23:24
	→ tinti joined	23:24
knuth	have the local repo be a single user?	23:25
grawity	knuth: sure	23:25
	← farn left	23:25
knuth	then let anyone local push to that as they wish?	23:25
	gitinfo set mode: +v	23:25
purpleidea	sweet	23:25
grawity	yeah, just need to find a way to automate the synchronization	23:25
knuth	with github?	23:25
grawity	automating a fetch, or push, is easy	23:26
	it's a bit harder when people can push to both repositories and the script has to decide whether to push from local-copy to github or to fetch instead	23:26
	← tagava\|2 left	23:26
	← tagava left	23:26
purpleidea	jrnieder: if you do a git stash pop instead of apply, and you don't use the --index, does that mean the index data (that you helped me find) would be "deleted" forever ?	23:27
knuth	ya that is why I figured it would be managed by a person.	23:27
grawity	knuth: but, technically, you can still be 'up' even without that	23:27
knuth	though honestly most changes could come from local.	23:27
	← frgtn left	23:27
	← raonyguimaraes left	23:27
	← decotii left	23:27
grawity	since everyone just keeps committing locally – to their own clone – and when Github goes back up the next day, they can continue as if nothing happened	23:28
knuth	yep yep. but this is probably needed anyhow with the outside vendor.	23:28
	have two local repos one to pull back the vendors stuff to test?	23:29
	the other our groups current copy?	23:29
	→ el3ctron_ joined	23:29
	← tbmiller left	23:30
	→ tbmiller joined	23:31
	← Robarelli left	23:32
	→ raf joined	23:33
	→ boombatower joined	23:35
	→ farn joined	23:35
	← tbmiller left	23:35
	← willwh left	23:36
	← Soliton left	23:38
	← tvw left	23:40
	→ Soliton joined	23:40
	← kcj left	23:41
	← mizu_no_oto left	23:41
	→ infynyxx joined	23:43
	← zacts left	23:44
	← Archweasel left	23:44
	← Soliton left	23:45
	→ functorfun joined	23:45
	functorfun → funfunctor	23:45
	← natechan left	23:45
	← Zarthus left	23:45
sgu	Hi all, A particular version may stay unchanged for many commits, so the last version of a particular may not be in HEAD~1. Any git command can give me the file before its last change?	23:47
	→ adamm joined	23:47
	→ pr0ton joined	23:47
sgu	Hi all, A particular file may stay unchanged for many commits, so the last version of a particular may not be in HEAD~1. Any git command can give me the file before its last change?	23:47
	← bwreilly_ left	23:48
purpleidea	is there a way to do an interactive merge similar to how git add --patch works or the 'e' works, so that i can make a few simple changes before the merge applies?	23:49
	→ grilix joined	23:50
	→ joshsmith joined	23:50
	→ Soliton joined	23:50
	→ rchavik joined	23:53
iveqy	sgu: not afaik. You can always do something like this:	23:54
	FILE=<filename> && git diff $(git rev-list HEAD "$FILE" \| head -2 \| tail -1) "$FILE"	23:54
	sgu: are you used to work with cvs?	23:54
	purpleidea: I don't know, see man git merge	23:55
gitinfo	purpleidea: the git-merge manpage is available at http://jk.gs/git-merge.html	23:55
iveqy	sgu: usually you use git log to see the history of a file	23:56
	→ epicowl joined	23:56
	→ chirayuk joined	23:57
iveqy	sgu: like git log -p <filename> see man git log	23:57
gitinfo	sgu: the git-log manpage is available at http://jk.gs/git-log.html	23:57
	→ cagedwisdom joined	23:57
	→ slakmagik joined	23:57
	← slakmagik left	23:58
	← tjbiddle left	23:59

Logs Search ←Prev date Next date→

Channels Documentation