IRCloggy #git 2011-02-06

Provider of IRC logs since 2005.

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

	2011-02-06
	→ urbanmonk joined	00:00
	→ normanrichards joined	00:01
	→ ijuma_ joined	00:01
	← dr_win left	00:01
PovAddict	cirwin: I'm doing just that while trying to see which set commits is equivalent to a squashed commit	00:03
	← normanrichards left	00:03
	→ normanrichards joined	00:04
PovAddict	cirwin: git reset --hard foo^; git cherry-pick A B C D; git diff HEAD foo;	00:04
	→ dr_win joined	00:04
PovAddict	then tweak the A B C D list until the diff is small enough :P	00:04
	← chrischambers left	00:04
	← normanrichards left	00:04
	cirwin writes an alias :p	00:05
	← albel727 left	00:05
cirwin	I used checkout though	00:06
	→ Sgeo_ joined	00:06
	← Sgeo left	00:08
	← infidium left	00:09
	← jpieper left	00:09
	→ chrischambers joined	00:10
	→ infidium joined	00:11
	→ thorat joined	00:14
	→ jpieper joined	00:15
	← ixti left	00:17
	← tizzo left	00:19
	← DiGitalX left	00:21
	← Dave^\| left	00:22
	← ChUrP left	00:22
	← davidc_ left	00:24
	← rgr left	00:24
	→ PovAddict_ joined	00:25
	← PovAddict left	00:25
	PovAddict_ → PovAddict	00:25
	→ normanrichards joined	00:26
	← ch077179 left	00:28
	→ ChUrP joined	00:29
	→ davidc_ joined	00:29
	← normanrichards left	00:35
sitaram	Arrowmaster: what us "SU"? (You said something about my first +10 points on SU)	00:37
PovAddict	superuser?	00:37
	→ Buglouse joined	00:37
	← Elfe left	00:38
	← phaedrix left	00:42
	→ PovAddict_ joined	00:42
	← PovAddict left	00:43
	PovAddict_ → PovAddict	00:43
	→ amerine joined	00:44
	→ zombor joined	00:45
	→ ldh joined	00:46
	← zombor left	00:48
Arrowmaster	yeah superuser.com	00:50
	→ Elfe joined	00:51
	← PovAddict left	00:51
	← aisipos left	00:51
	← priidu left	00:52
	← amondracorp left	00:53
	→ tizzo joined	00:54
	← infidium left	00:56
	→ infidium joined	00:57
	← iAsteris_ left	01:01
	← chrischambers left	01:02
	→ Zeu5 joined	01:02
Zeu5	hi there, i hired 2 freelancers to help me with my project. i instructed them to each create a separate branch from the origin/master branch	01:03
	and commit those branches to the private online git repo	01:03
	on my own laptop i did a git pull	01:04
	followed by git checkout -b freelancer1 origin/freelancer1	01:04
	i want to switch to the other branch to see the changes. but i get an error	01:05
	← stringoO left	01:05
Zeu5	error: You have local changes to 'mockups/wildcard.ombi60.com/admin/products_index.bmml'; cannot switch branches.	01:05
	please help me. i only know enough to do push and commit and status	01:05
sh4rm4	you can move your changes to the stash area	01:05
	git stash	01:06
	→ normanrichards joined	01:08
	← airborn left	01:09
	→ jrmuizel_ joined	01:09
	← jrmuizel_ left	01:10
Zeu5	wat is this stash area sh4	01:11
	wat is this stash area sh4rm4	01:11
	?	01:11
	← ChUrP left	01:12
sh4rm4	an area for temporary stuff	01:12
	that's not to be commited yet	01:12
Zeu5	i see.	01:13
	← lamont left	01:15
	← ludde left	01:15
Zeu5	okie sh4rm4 thanks. but i have another problem	01:16
	because i stashed the changes when i was in one of the branches, hte change was no longer there when i switched back to master branch	01:16
	how do i get back the stashed changed into my master branch?	01:16
sh4rm4	http://linux.die.net/man/1/git-stash	01:18
	← amerine left	01:19
	→ ChUrP joined	01:19
	→ lamont joined	01:19
	→ MikhX joined	01:20
	→ utu joined	01:21
	← mar77i left	01:22
	→ amerine joined	01:22
Zeu5	okie thanks sh4rm4	01:22
	← OOPMan left	01:24
	← oriba left	01:25
	→ OOPMan joined	01:26
	← OOPMan left	01:30
	→ bcardarella joined	01:31
	← d0k left	01:32
	→ OOPMan joined	01:33
	→ homiziado joined	01:36
	← davidc_ left	01:37
	← tewecske left	01:37
	→ oriba joined	01:38
	→ zombor joined	01:39
	→ davidc_ joined	01:42
	→ up_the_irons joined	01:42
	← sako left	01:44
	← oriba left	01:44
	← oracal left	01:45
	← rorist_ left	01:45
	← tizzo left	01:46
	→ kipras joined	01:46
	→ babbage joined	01:53
babbage	Why doesn't this work in my .gitignore, to include to two negated directories? http://pastebin.com/pXU5xsa0	01:54
	Ah. Because there was a .git directory in there, it would appear.	01:56
	← MikhX left	02:03
	← brendon left	02:05
	→ brendon joined	02:05
Arrowmaster	you can remove that * too	02:07
	← jdav_gone left	02:07
	← dools left	02:08
	← becomingGuru_ left	02:09
	→ angelsl joined	02:10
	→ mar77i joined	02:12
	← Zeu5 left	02:12
	← thorat left	02:12
	← silanus left	02:13
	→ DrNick2 joined	02:13
	← DrNick left	02:14
	← panike left	02:15
	← gilaniali left	02:16
	← amerine left	02:16
	← utu left	02:17
	→ Fboy joined	02:18
	← ChUrP left	02:18
	→ MikhX joined	02:18
	→ silanus joined	02:20
	→ becomingGuru_ joined	02:20
	→ Ventrillus joined	02:22
	→ hyperair joined	02:22
	← Ventrillus left	02:22
	DrNick2 → DrNick	02:24
	→ ChUrP joined	02:26
	→ gilaniali joined	02:26
	← vitor-br left	02:26
	← normanrichards left	02:27
	← icwiener left	02:28
	← MattDiPasquale left	02:28
	← infidium left	02:29
	← homiziado left	02:29
	← Titosemi left	02:29
	← xiong left	02:30
	→ infidium joined	02:31
	← aah left	02:33
	→ sako joined	02:33
	→ shiba_yu36 joined	02:36
	← SvenDowideit left	02:39
	→ SvenDowideit joined	02:39
	← AzaToth left	02:39
	→ homiziado joined	02:41
	→ Vampire0_ joined	02:43
	→ gnufied joined	02:44
	← bcardarella left	02:44
	← homiziado left	02:44
	→ homiziado joined	02:45
	→ dl9pf_ joined	02:45
	← dl9pf left	02:46
	→ t3ntman joined	02:47
	← t3ntman left	02:47
	← ipis left	02:47
	← Vampire0 left	02:47
homiziado	noob question: was trying to follow http://carisenda.com/written/gitweb-on-mac-os-x/ but no repository appears on the web interface	02:47
	← phantomcircuit left	02:48
	← infidium left	02:49
	← gnufied left	02:49
	→ gnufied joined	02:50
	→ leo2007 joined	02:51
	→ infidium joined	02:51
	← booginga left	02:54
	→ booginga joined	02:55
	← urbanmonk left	02:56
	→ astm joined	02:56
	→ jrmuizel_ joined	02:58
	← byteframe left	03:00
	→ daze21 joined	03:02
daze21	so I have this remote repo on codaset that's been fine pushing to in the past, but now I get the error described here...http://stackoverflow.com/questions/2816369/git-push-error-remote-rejected-master-master-branch-is-currently-checked-ou	03:03
	I don't understand how to execute "git config --bool core.bare true" in my remote repo folder as the top answer says... I just want this to work. (and also to understand why it suddenly didn't work)	03:03
	← f00li5h left	03:04
	← Elfe left	03:04
	→ Elfe_ joined	03:04
	← dkasak left	03:06
	← mcjb left	03:06
	← jrmuizel_ left	03:07
	→ dkasak joined	03:10
	← shruggar left	03:10
	→ phantomcircuit_ joined	03:10
	← phantomcircuit_ left	03:10
	→ phantomcircuit joined	03:11
	→ intripoon joined	03:11
	← intripoon_ left	03:15
	← kipras left	03:16
	← ChUrP left	03:19
	→ Heimidal joined	03:22
	← daze21 left	03:25
	→ ChUrP joined	03:26
	→ jdav_gone joined	03:30
	← Amorphous left	03:36
	← chessguy left	03:39
	← gilaniali left	03:40
	← homiziado left	03:41
	← ldh left	03:41
	← justin-george left	03:43
	→ ignotus joined	03:44
	← hyperair left	03:45
ignotus	hi, I have some stuff in the index, how can I make the index clear, so I can git add my stuff again, in a better order	03:46
	→ RobertLaptop joined	03:47
	← gnufied left	03:47
	← NickyP left	03:47
ignotus	I think what I need is git reset --mixed	03:48
	oh no, I already did a git reset --soft HEAD^	03:48
	infogulch → infogulch[a]	03:50
	→ Amorphous joined	03:51
	→ gnufied joined	03:51
	← mrconnerton left	03:51
	→ necromancer joined	03:52
	→ timj__ joined	03:52
	→ eno joined	03:54
	← eno left	03:54
	→ eno joined	03:54
ignotus	ah, a simple git reset did it	03:55
	← timj_ left	03:56
	← ChUrP left	04:01
	→ BiggFREE joined	04:01
	← Schmallon left	04:03
	← Fboy left	04:03
	← tstclair left	04:04
	→ _psychic_ joined	04:05
	→ sudonotplease joined	04:07
	← sudonotplease left	04:07
	→ ChUrP joined	04:07
	← ignotus left	04:09
	← jjuran left	04:11
	→ gilaniali joined	04:12
	← gilaniali left	04:13
	→ normanrichards joined	04:13
	← sako left	04:14
	← necromancer left	04:14
	→ sudonotplease joined	04:15
	→ necromancer joined	04:15
	→ bhosie joined	04:16
	← bhosie left	04:16
	← unreal left	04:21
	← necromancer left	04:21
	→ gozoner joined	04:21
	← gnufied left	04:22
	→ unreal joined	04:22
	→ blaines joined	04:24
	→ hyperair joined	04:24
	← gozoner left	04:26
	← stigrett left	04:27
	→ jjuran joined	04:30
	→ simplechat joined	04:34
	← simplechat left	04:34
	→ simplechat joined	04:34
	→ tizzo joined	04:34
	← hyperair left	04:35
	← infidium left	04:36
	← shiba_yu36 left	04:39
	← juangiordana left	04:42
	→ MacGyverNL_ joined	04:45
	← MacGyverNL_ left	04:45
	→ MacGyverNL_ joined	04:45
	← blaines left	04:45
	→ shiba_yu36 joined	04:45
	→ neilkod joined	04:45
neilkod	how do i prepare a local repository for use on github?	04:46
	← BiggFREE left	04:46
	← babbage left	04:48
cirwin	neilkod: git remote add github [email@hidden.address]	04:48
	then just git push github will work	04:48
neilkod	do i have to create the repository in github first?	04:49
cirwin	I think so	04:49
	← ChUrP left	04:53
	→ kelvie joined	04:56
	← nothingHappens_ left	04:58
	← brendon left	04:58
	→ gnufied joined	05:00
	← hagebake left	05:01
	→ hagebake joined	05:01
	← hagebake left	05:01
	→ hagebake joined	05:01
	→ ChUrP joined	05:02
	→ justin-george joined	05:02
	← markelikalderon left	05:02
	← tizzo left	05:04
	← tchan left	05:06
	→ aah joined	05:07
	← gnufied left	05:07
	→ Utkarsh_ joined	05:12
	← Utkarsh left	05:15
	← neilkod left	05:20
	→ tizzo joined	05:21
	→ maleknet joined	05:22
	→ skoop joined	05:23
	← skoop left	05:23
	→ skoop joined	05:23
	← OOPMan left	05:26
	← lfaraone left	05:26
	→ envi joined	05:27
	→ blaines joined	05:27
	→ eletuchy_ joined	05:27
	← adamm left	05:28
	→ adamm joined	05:29
	← eletuchy left	05:30
	← justin-george left	05:30
	→ Utkarsh joined	05:31
	→ dyek joined	05:34
	← Utkarsh_ left	05:34
	→ tchan joined	05:35
	← ceej left	05:36
	← MacGyverNL_ left	05:36
	→ MacGyverNL joined	05:38
dyek	Hi! Given a git commit SHA-1 hash, how do I check if the commit exists in a specific branch? Is there a way I can check if another commit replaced the specified commit?	05:38
	→ sako joined	05:39
cirwin	git branch --contains <commit>	05:39
	what do you mean by "replaced"?	05:39
	→ eletuchy joined	05:40
dyek	cirwin: Thank much. "replaced" means a better fix was developed and the original fix is no longer used.	05:40
cirwin	umm, how would you expect git to know about that?	05:40
	← zivester_ left	05:41
dyek	cirwin: I don't know. I thought git has parent-child tree relationship, but I'm new to git.	05:41
	→ tapank joined	05:41
	← eletuchy_ left	05:42
cirwin	ok — I've probably just misunderstood your question	05:42
	→ zivester_ joined	05:42
	← tapank left	05:42
PerlJam	sounds like looking for a "git revert" to me, but I have no idea	05:43
	→ tapank joined	05:43
	→ Elvashi joined	05:45
dyek	To illustrate, lets say a module developer created the first version of a fix for a problem and "submitted" it to be merged to linus's tree. Then, a kernel maintainer refined the fix to be more appropriate and thus replaced the original patch with a new patch. I wonder if git keeps track of such evolving relationship -- given a commit hash, find a subsequent hash that replaced the original one.	05:46
	← ChUrP left	05:46
	← Bass10 left	05:47
cirwin	dyek: sadly, it does not	05:47
dyek	cirwin: OK. Thanks!	05:47
cirwin	it would be awesome though, I agree	05:48
PerlJam	sounds like something that would have to be built on top of git.	05:48
	Something that knows more about the meaning of the changes.	05:49
cirwin	PerlJam: well, rebase should add the link too	05:49
	and cherry-pick etc.	05:49
	and you could then use that information to help with merges too	05:49
	→ Error404NotFound joined	05:50
	← linusoleander left	05:51
	→ juangiordana joined	05:52
	← jfkw left	05:52
	← juangiordana left	05:52
PerlJam	you could use git notes to do it in a poor-man's way	05:52
	→ mikepack joined	05:53
cirwin	hm, I was thinking of just putting a line in the commit	05:53
	probably as part of the message section for backwards compatibility	05:53
	→ juangiordana joined	05:54
PerlJam	That'll work too	05:54
	→ ChUrP joined	05:54
cirwin	it would be very useful for our team if we could rebase, and not lose github comments :)	05:55
	→ gnufied joined	05:55
cirwin	so only backwards pointers are needed	05:55
	though you could use notes for forward pointers too	05:55
	← zombor left	05:55
	→ randomus3r joined	05:59
	← sako left	06:01
	← sudonotplease left	06:02
	→ urbanmonk joined	06:03
	← skoop left	06:03
	← randomus3r left	06:06
	← astm left	06:06
	← zivester_ left	06:06
	← MetaCosm_ left	06:09
	← Elvashi left	06:10
	→ sako joined	06:13
	← sako left	06:13
	→ sako joined	06:13
	→ j416 joined	06:13
	← gnufied left	06:15
	→ amerine joined	06:16
	→ Utkarsh_ joined	06:16
	← Utkarsh left	06:19
	← joel_falcou left	06:19
	→ joel_falcou joined	06:20
	→ Zygo joined	06:20
	← jason237 left	06:20
	→ jason237 joined	06:21
	← ChUrP left	06:22
	← carrus85 left	06:23
	← TML left	06:24
	← Zygo left	06:24
	→ TML joined	06:24
	→ Zygo joined	06:24
	← yhager left	06:25
	→ zombor joined	06:25
	→ shamrock joined	06:26
	→ Elvashi joined	06:27
	← t0rc left	06:27
	→ ChUrP joined	06:29
	→ gnufied joined	06:33
	→ augiedoggie joined	06:34
	← davidc_ left	06:34
	→ smolyn joined	06:35
	← shiba_yu36 left	06:36
	→ davidc_ joined	06:40
	← TanEk left	06:42
	← Heimidal left	06:45
	→ NickyP joined	06:46
	→ yhager joined	06:46
	← adamm left	06:46
	→ MrShmelo12 joined	06:46
	← MrShmelo12 left	06:47
	→ coconutz joined	06:48
coconutz	hey all	06:49
	if i do new repo i should put it in a diffrent dir than the web site source files?	06:49
	or inside of it	06:49
j416	coconutz: not sure what you mean by that	06:50
	← shamrock left	06:50
j416	do you mean if you should put your web root in a subdir or not?	06:50
cirwin	http://toroid.org/ams/git-website-howto	06:50
	^ coconutz	06:50
coconutz	i got files in /var/www/html	06:50
	i got files in /var/www/html/site	06:50
	so i should do init there	06:50
j416	if you do init inside /var/www/html, git will track both html and site	06:51
	if you do init inside /var/html/site, git will track only site	06:51
coconutz	and also im using zend studio , is there a plugin for git?	06:51
cirwin	and in both cases, people will be able to navigate to http://mysite.com/.git/ unless you configure your server right	06:51
j416	I don't know what zend studio is, sorry.	06:51
coconutz	ok	06:52
j416	coconutz: I have a /var/web which contains my git repo. Inside it I have a directory I call 'root', which I symlink to /var/www.	06:52
	→ TheUni joined	06:52
j416	that way I keep the repo outside reach.	06:53
coconutz	is there a way to hook a db dumb on each commit ?	06:53
	→ shiba_yu36 joined	06:53
cirwin	coconutz: you can set up hooks using git, yes	06:53
j416	coconutz: read about git hooks. But I'm not sure why you'd want that.	06:53
cirwin	man githooks	06:53
jast	the 'githooks' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/githooks.html [automatic message]	06:53
cirwin	you also might want to look at cron	06:53
	for that kind of thing	06:53
coconutz	oh ok	06:54
	i subscribed to github	06:54
	is there a better choices?	06:54
	i pay 7$	06:54
wereHamster	define better	06:54
coconutz	faster cheapper	06:54
	→ radarek joined	06:55
	← booginga left	06:55
	→ booginga joined	06:56
	→ OOPMan joined	07:00
coconutz	btw: what is fork?	07:00
cirwin	makes a new copy of the repository	07:02
	→ shamrock joined	07:02
coconutz	if i want to upload changes its "commit" right?	07:03
cirwin	no, "push" uploads "commits"	07:03
	man gittutorial	07:03
jast	the 'gittutorial' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/gittutorial.html [automatic message]	07:03
coconutz	ok and pull is to download right?	07:03
cirwin	with a little bit of extra magic, yes	07:03
coconutz	what software to do use? eclipse? tortisesvn?	07:04
cirwin	umm, git?	07:04
	there is a tortoise git I think	07:04
	← OOPMan left	07:04
	← darsk1ez left	07:06
	← tizzo left	07:10
	← booginga left	07:10
	← dyek left	07:10
coconutz	what is the gems stuff i see it all the time	07:12
	→ t_b joined	07:12
cirwin	gems?	07:13
	their ruby libraries that have been packaged	07:13
	*they're!	07:14
	Elvashi → Textmode	07:14
	→ tty1 joined	07:15
	← radarek left	07:16
	→ OOPMan joined	07:19
	← ChUrP left	07:20
	← gnufied left	07:21
	→ darsk1ez joined	07:21
	← shamrock left	07:24
	→ LongBeach joined	07:25
	→ rgr joined	07:25
	→ pfrenssen joined	07:27
	→ ChUrP joined	07:28
	← _psychic_ left	07:29
	→ Yuuhi` joined	07:35
	→ Chillance joined	07:36
	← leo2007 left	07:37
	← Yuuhi left	07:37
	← becomingGuru_ left	07:52
	→ justin-george joined	07:54
	← pfrenssen left	07:55
	→ wtfness joined	07:57
	← jjuran left	07:59
	→ keyvan- joined	07:59
	→ leo2007 joined	08:02
	→ becomingGuru_ joined	08:05
	→ cesario joined	08:06
	← ChUrP left	08:06
	← zombor left	08:10
	→ zivester joined	08:12
	→ D4rkR4g3 joined	08:13
	← D4rkR4g3 left	08:13
	→ zombor joined	08:13
	→ ChUrP joined	08:13
	← ISF left	08:17
	← Error404NotFound left	08:19
	← hagebake left	08:19
	← Octalot left	08:20
	→ wfriction joined	08:20
	→ jjuran joined	08:20
	→ nathacof joined	08:23
	← nairb774 left	08:23
	← becomingGuru_ left	08:24
	← wfriction left	08:25
	→ wfriction joined	08:27
	← cesario left	08:28
	← zivester left	08:29
	← juangiordana left	08:33
	→ workmad3 joined	08:35
	← zombor left	08:37
	→ nairb774 joined	08:38
	→ hagebake joined	08:39
	← hagebake left	08:39
	→ hagebake joined	08:39
	← nathacof left	08:40
	→ zivester joined	08:40
	← wfriction left	08:43
	→ tewecske joined	08:44
	← tapank left	08:46
	← hagebake left	08:47
	→ hagebake joined	08:48
	← hagebake left	08:48
	→ hagebake joined	08:48
	← shiba_yu36 left	08:48
	→ shiba_yu36 joined	08:49
	← ChUrP left	08:50
	→ tapank joined	08:51
	← keyvan- left	08:52
	→ kumbayo joined	08:55
	← hagebake left	08:55
	→ wfriction joined	08:56
	→ dfr\|mac joined	08:57
	← marc_kdab left	08:57
	→ marc_kdab joined	08:57
	← dfr\|mac left	08:58
	→ dfr\|mac joined	08:58
	← wfriction left	08:58
	→ ChUrP joined	08:58
	← sako left	09:06
	→ TheFox21 joined	09:06
	→ ThomasLocke joined	09:11
	→ PigeonFriend joined	09:12
	← envi left	09:13
	→ envi joined	09:13
	← emptyflask left	09:19
	→ doener joined	09:20
	← cirwin left	09:23
	← madewokherd left	09:25
	← TheFox21 left	09:27
	← eletuchy left	09:28
	→ xiong joined	09:28
	← MikhX left	09:29
	← ChUrP left	09:31
	← leo2007 left	09:31
	→ TrineLocke joined	09:32
	→ chrischambers joined	09:36
	→ ChUrP joined	09:37
	→ iAsterisk joined	09:38
	→ RaceCondition joined	09:38
	→ Zephyrus joined	09:40
	→ anderse joined	09:45
	→ radarek joined	09:47
	→ kar8nga joined	09:49
	→ ixti joined	09:49
	→ squentin joined	09:51
	→ ph^ joined	09:53
	← shiba_yu36 left	09:53
	→ lanthan_afh joined	09:54
	← killfill left	09:56
	→ killfill joined	09:56
	→ priidu joined	09:57
	→ vmiklos joined	09:58
	→ MetaCosm_ joined	09:59
	→ airborn joined	10:00
	← mikepack left	10:00
	← killfill left	10:01
	→ killfill joined	10:02
	← ixti left	10:03
	← Mage left	10:07
	← tapank left	10:07
	→ matjas_ joined	10:08
	→ gnufied joined	10:09
	→ Cvbge joined	10:09
	← killfill left	10:10
	→ killfill joined	10:11
	→ Fboy joined	10:15
	← killfill left	10:16
	→ killfill joined	10:16
	→ xiangfu joined	10:18
	→ poxd joined	10:20
poxd	i have some stuff to be added on the staging area but I'd like to do so on an another branch than I'm on, what's the simplest way to do so?	10:21
	→ ludde joined	10:21
	← ChUrP left	10:21
	← poxd left	10:23
	→ drizzd joined	10:25
	← doener left	10:26
	→ ch077179 joined	10:26
	→ Guest71890 joined	10:26
	← Guest71890 left	10:27
	→ ChUrP joined	10:27
	→ cesario joined	10:29
	← drizzd_ left	10:29
	← giallu left	10:37
	→ doener joined	10:38
	→ goshawk joined	10:38
	→ sontek joined	10:38
sontek	When you guys clone an svn repository, do you keep master a clone of svn? i.e only git svn rebase it in, and any changes you make are always in a branch?	10:39
	← roger_padactor left	10:39
doener	sontek: that would be a pointless exercise. The pristine state from svn is already kept in a remote-tracking branch	10:40
sontek	ok, so I can just keep committing to master and be ok	10:42
	→ verdon joined	10:42
	→ linusoleander joined	10:43
doener	sontek: branching is still useful for organizing your work and strongly recommended. Just the "keep master the same as svn's trunk" part is not a good reason to do so	10:45
	← amerine left	10:47
doener	sontek: you'll just have to be careful about merging. With git-svn I usually do the rebase/dcommit dance each branch and never merge, unless there's also a branch in svn, that I want to merge (i.e. there's gonna be a 1:1 mapping of the merge)	10:47
	→ hohoho joined	10:50
	← xiangfu left	10:50
	← linusoleander left	10:51
	matjas_ → matja	10:53
	matja → matjas	10:53
	← Buglouse left	10:54
	→ f00li5h joined	10:55
	→ ppawel joined	10:57
	← Cvbge left	11:03
	← ChUrP left	11:04
	← intripoon left	11:05
	→ intripoon joined	11:06
	← matjas left	11:10
	→ ChUrP joined	11:10
	→ MAK-Jr joined	11:12
	→ carpet_the_walls joined	11:13
	← blaines left	11:19
	→ blaines joined	11:19
	→ intripoon_ joined	11:21
	→ paszo2008 joined	11:21
	← theoros left	11:22
	← intripoon left	11:23
	← MAK-Jr left	11:24
	→ MAK-Jr joined	11:24
	→ HG` joined	11:24
	→ pdelgallego joined	11:24
	→ darkredandyellow joined	11:25
	← codespectator left	11:30
	← MAK-Jr left	11:36
	→ MAK-Jr joined	11:37
	→ rcu joined	11:41
	→ ExtraSpice joined	11:44
	← emiel_ left	11:45
	← rcu left	11:45
	→ sandbags joined	11:47
	← ExtraSpice left	11:47
	← xiong left	11:48
	← sandbags left	11:48
	→ leo2007 joined	11:48
	← MAK-Jr left	11:49
	→ ldh joined	11:51
	← paszo2008 left	11:52
	→ codespectator joined	11:52
	← jjuran left	11:52
	← blaines left	11:54
	→ zorzar joined	11:55
	← EricInBNE left	11:56
	← zorzar_ left	11:57
	→ paszo2008 joined	11:57
	← ppawel left	11:59
	← Elfe_ left	12:03
	→ Dave^\| joined	12:10
	→ roop joined	12:11
	→ mrconnerton joined	12:15
	← Dave^\| left	12:15
	→ torvalamo joined	12:16
	← ChUrP left	12:19
	← Chillance left	12:20
	→ markelikalderon joined	12:21
	→ jjuran joined	12:21
	→ vitor-br joined	12:22
	→ linusoleander joined	12:23
	→ ChUrP joined	12:26
	← ldh left	12:28
	← davidc_ left	12:29
	← bradleyayers left	12:30
	← roop left	12:32
	← linusoleander left	12:34
	→ davidc_ joined	12:34
	← davidc_ left	12:34
	→ davidc_ joined	12:34
	→ napster_123 joined	12:34
	→ j0su3 joined	12:36
	← radarek left	12:37
	→ klevison joined	12:38
	→ radarek joined	12:38
	← radarek left	12:39
klevison	how can I configure my git to do automatic login	12:39
	every command I need put password	12:39
thiago_home	git doesn't have that option	12:39
	what transport are you using? Ssh?	12:40
klevison	but I need to configure it	12:40
	→ maattd joined	12:41
thiago_home	so?	12:41
klevison	yeah	12:41
thiago_home	then configure ssh not to require a password	12:41
klevison	I use centos + git	12:41
	on github	12:41
thiago_home	use key-based authentication and run ssh-agent	12:41
	→ theoros joined	12:43
	← theoros left	12:43
	→ theoros joined	12:43
	← justin-george left	12:43
	← mrconnerton left	12:43
	→ Klevison_ joined	12:45
	← klevison left	12:45
	→ Titosemi joined	12:46
	→ DivineOmega joined	12:47
	← Klevison_ left	12:47
	→ klevison joined	12:47
klevison	thiago_home: are you brazilian?	12:47
	← coconutz left	12:47
	→ jstemmer joined	12:48
DivineOmega	I'm a bit of a newbie to source control, would anyone mind advising me regarding with an issue I'm having?	12:48
selckin	!ask	12:49
	→ ExpDate joined	12:49
klevison	I've installed this redmine plugin https://github.com/koppen/redmine_github_hook/	12:50
	I works	12:50
	It works	12:50
	→ Dave^\| joined	12:50
DivineOmega	Basically, its a matter of organising. I'm using git in assembla (hosted git repo.). A small group of friends are working on a project within Eclipse.	12:50
klevison	but there is only one problem	12:50
ExpDate	hei, anyone know how to get files from 2 previous commit ?	12:50
FauxFaux	Assembla. /o\	12:50
ExpDate	I meant discarded two previous commit	12:50
DivineOmega	FauxFaux: Really?	12:50
ExpDate	How it is could be done	12:50
klevison	I need to type my password every time there is a new update on source code	12:51
FauxFaux	DivineOmega: Have you tried subscribing to changes on anything?	12:51
	← Dave^\| left	12:51
DivineOmega	FauxFaux: It seems a bit dodgy sometimes, I can get e-mail with the incorrect names occasionally.	12:51
	My main problem is that I'm not sure how to organise the repo to work correctly with Eclipse's 'projects'	12:52
	Is the best idea to open put /src/ as tracked in the repo, and all the dot folders as untracked?	12:52
	only*	12:53
klevison	I've installed this redmine plugin https://github.com/koppen/redmine_github_hook/ . It works, but there is only one problem. I need to type my password for every commit. I don't type it.	12:53
	→ eijk joined	12:54
ExpDate	How to discard two previous commit ?	12:54
bremner	klevison: so people mentioned several times the answer to typing your password is ssh-agent and public key authentication. Is this not clear to you?	12:55
klevison	no, it isn't	12:56
bremner	well, google ssh passwordless login finds quite a few references.	12:56
	actually, to be fair that will probably be keys without passphrases.	12:57
	klevison: do you know what ssh-agent is?	12:57
	→ d0k joined	12:57
klevison	bremner: no :(	12:58
	← OOPMan left	12:58
	→ roger_padactor joined	13:00
	→ Dave^\| joined	13:00
bremner	klevison: ssh-agent is a program that runs in the background on your client (i.e. where the local clone of the git repo is). It is independent of git, so you just need to find out how to get it working for your centos user.	13:00
	← Tommy[D] left	13:01
	→ marlun joined	13:01
	→ d0k_ joined	13:01
marlun	Hi, Could someone take a look at https://github.com/marlun/jquery-megadropdown/commits/master and tell me why those "Merge branch 'master' of github.com:marlun/jquery-megadropdown" commits is happening?	13:02
	← d0k left	13:02
	d0k_ → d0k	13:02
	→ Tommy[D] joined	13:02
	→ OOPMan joined	13:02
	← leo2007 left	13:03
bremner	klevison: http://kimmo.suominen.com/docs/ssh/ looks pretty complete, although I haven't actually used it.	13:03
	klevison: but really it is a question for your centos support channel, whatever that is.	13:04
	← OOPMan left	13:04
FauxFaux	marlun: There's two people committing to the repo; when you pull it defaults to merging their changes with yours; git pull --rebase # will fix the problem with all the normal rebase caveats.	13:04
	→ p3rg2z joined	13:05
	→ emiel_ joined	13:05
	→ OOPMan joined	13:05
	→ malte_ joined	13:05
klevison	bremner: thanks a lot.	13:05
FauxFaux	marlun: http://i.imgur.com/ZU25D.png etc.	13:06
	← codespectator left	13:07
	← wtfness left	13:09
	→ Cvbge joined	13:10
marlun	FauxFaux: Thanks. Just havn't seen it in other repos that I know more then 2 people commit to but I guess they are using rebase then. Will look into it, thanks!	13:10
	ChanServ set mode: +v	13:10
	→ AzaToth joined	13:11
	← OOPMan left	13:12
AzaToth	if branch A has an submodule, and branch B doesn't, is there any way to prevent the directory for the submodule to polute branch B?	13:12
	I don't know if it's a common, or even valid reason/situation	13:14
	but say you branch master, and for a specific functionallity you include submodule X, then when swithcing back to master, the created directory is still there	13:15
	I'm using 1.7.2 by the way, perhaps there's been changes in 1.7.4...	13:16
	→ neilkod joined	13:16
	← p3rg2z left	13:17
	← ExpDate left	13:18
	→ hyperair joined	13:20
	← ChUrP left	13:20
	→ gusnan joined	13:20
	→ Anti-X joined	13:20
	← makubi left	13:20
	← torvalamo left	13:21
	Anti-X → torvalamo	13:21
	← gusnan left	13:22
	→ gusnan joined	13:22
	→ bentob0x joined	13:24
	← friskd left	13:25
	→ MattDiPasquale joined	13:26
	→ ChUrP joined	13:26
	→ OOPMan joined	13:27
	← kumbayo left	13:28
	→ DiGitalX joined	13:29
	→ felipec joined	13:30
felipec	is it possible to always add the 'From' line when sending mails?	13:30
	like it's added when author != committer	13:30
	← neilkod left	13:31
AzaToth	felipe: --envelope-sender=auto perhaps	13:32
	→ sanooj joined	13:34
AzaToth	felipec: *	13:35
	→ adsgb joined	13:35
AzaToth	felipec: ↑ my tab lured me	13:35
	← iAsterisk left	13:35
sanooj	hey. How can I create a new root in a git repository?	13:36
	aha. "Creating New Symbolic Branches" in the git book tells me how.	13:38
felipec	AzaToth: nope... I actually added that 'auto' option	13:38
	AzaToth: that just adds -f $from on the sendmail command	13:38
	→ emptyflask joined	13:39
	← adsgb left	13:41
	← rjnienaber left	13:41
	→ bitkiller joined	13:42
	→ _iron joined	13:44
	← sanooj left	13:46
	→ codespectator joined	13:46
	→ ppawel joined	13:47
AzaToth	felipec: you want the reverse?	13:49
felipec	AzaToth: no, as I said, I want the line "From" line in all the mails, even if author == committer, envelope sender has nothing to do with that	13:50
Cvbge	Is there something like git-new-workdir but for filesystems that do not support links, like fat32?	13:50
	← codespectator left	13:51
AzaToth	felipec: you want what?	13:52
	do you want a list of all From lines?	13:52
	or do you want to fill the "from" line with author?	13:52
	← kar8nga left	13:53
	← mar77i left	13:53
	← t_b left	13:53
	→ t_b joined	13:54
	→ Sgeo__ joined	13:54
felipec	AzaToth: try this: git commit --amend --author="Foo Bar [email@hidden.address] and send that patch to yourself	13:57
	← Sgeo_ left	13:57
	_iron → i_ron	13:57
felipec	AzaToth: you would notice an extra "From" line at the top of the body... that's what I want: http://pastie.org/1533917	13:59
	→ astm joined	14:00
AzaToth	ah	14:01
	felipec: unless I'm mistaken, the commit sha1 will change if you change the author	14:01
felipec	AzaToth: so?	14:01
	AzaToth: pick any commit that you didn't write then	14:02
AzaToth	felipec: http://paste.debian.net/106673/	14:04
	if I change the author, I must change the ID	14:04
	→ Sonderblade joined	14:04
felipec	sigh	14:04
AzaToth	and aint the Signed off line enough?	14:05
felipec	AzaToth: enough for what? Some people send patches through mail, and their corporate servers mangle them Felipe Contreras -> Felipe Contreras (Nokia Helsinki), when the recipient tries to apply the patch, they have to fix the author	14:07
AzaToth	ah	14:07
felipec	but if the "From" line is in the body, they don't	14:08
	← verdon left	14:11
	← maleknet left	14:12
	← napster_123 left	14:12
	← orafu left	14:13
	→ verdon joined	14:13
	→ cdere joined	14:15
	← cdere left	14:15
	→ silverfix joined	14:16
silverfix	hello	14:16
	← felipec left	14:17
silverfix	guys I'd want to ask a question that I can't easily find on the docs: why when I create a new branch e switch to it if I type: git push I give: fatal: No destination configured to push to. ??	14:18
	in the master branch git push works	14:18
	← Titosemi left	14:18
tango_	silverfix: because the new branch doesn't know where to push to	14:20
silverfix	ok, it seems alright so, how can I config this?	14:20
tango_	silverfix: so you have to configure it to know where to push to. assuming you want to push somebranchname to a new branch on the remote origin, you would do something like: git push -u origin somebranchname:somebranchname	14:21
	the -u option sets the new branch as the official remote branch to be tracked by somebranchname	14:21
	of course on the second half of : you would put whatever you want the remote branch to be called	14:21
	man git-push for further details	14:21
jast	the 'git-push' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-push.html [automatic message]	14:21
	← Cvbge left	14:22
silverfix	oh very well tango_, thank you!	14:22
	→ Tixz joined	14:22
	ChanServ set mode: +v	14:23
Tixz	Hi, how do I push local changes from a submodule to it's original (remote) repo?	14:23
silverfix	tango_: but at this point I can also add the default repo directly by git config right?	14:23
tango_	if you know how, sure	14:23
	silverfix: but unless the remote branch exists, you have to create it anyway	14:24
silverfix	I create it by -u?	14:24
tango_	silverfix: no, just push	14:25
silverfix	ok	14:25
tango_	-u sets the upstream reference, i.e. configures git to know that what you push to will be the upstream of what you push from	14:25
silverfix	I understand	14:26
	thank you :)	14:26
tango_	yw	14:26
	← klevison left	14:26
	→ Titosemi joined	14:26
Tixz	Submodules anyone? :)	14:27
	← malte_ left	14:28
silverfix	Tixz: submodules?	14:28
	Tixz: what do you mean?	14:28
Tixz	git submodule, I need to know how to push updates to the 'mother' repo	14:29
	→ Ruudjah joined	14:29
	← Titosemi left	14:29
Tixz	Can't find any concrete info on it though :	14:29
	but maybe I'm using the wrong keywords for searches	14:29
	← ch077179 left	14:33
	← darkredandyellow left	14:35
	← LongBeach left	14:35
	→ darkredandyellow joined	14:36
	← darkredandyellow left	14:37
	→ ch077179 joined	14:37
	← Zephyrus left	14:41
	→ Buglouse joined	14:42
	→ napster_123 joined	14:45
	← silanus left	14:45
	→ codespectator joined	14:46
	← kpreid left	14:47
	← davidc_ left	14:48
	→ tizbac joined	14:49
	← hyperair left	14:49
	← tizbac left	14:49
	← Fboy left	14:49
	← ChUrP left	14:50
	silverfix → cdere^Away	14:50
	← cdere^Away left	14:50
	→ hyperair joined	14:50
	→ Fboy joined	14:51
	→ Mohjive joined	14:52
	→ t0rc joined	14:53
	→ zivester_ joined	14:53
	← angelsl left	14:54
	→ tizbac joined	14:56
	→ ChUrP joined	14:57
	← rgr left	14:58
	← emiel_ left	15:02
	→ sudonotplease joined	15:03
	→ silanus joined	15:04
	→ rgr joined	15:05
	→ radarek joined	15:08
	→ MattDiPa_ joined	15:10
Mohjive	I need a tip on project layout and how git handles my problem - the premise is that I have multiple platforms for some projects and files can be shared between platforms	15:11
	→ xorAxAx joined	15:11
bremner	Mohjive: use branches?	15:12
Mohjive	with Visual SourceSafe (VSS) - don't say anything as it was not my idea, I just inherited this - it is possible to "link" files between directories, so that whenever one files change it is seen in other places	15:12
	bremner: I thought of this	15:12
	← ch077179 left	15:12
	← MattDiPasquale left	15:12
bremner	Or use symlinks, but this has nothing to do with git	15:12
Mohjive	and, yes, that's one of the solutions	15:13
	symlinks are not possible since we're including windows here	15:13
bremner	afaict, those are the options	15:14
	s/options/non-crazy options/	15:14
	← MattDiPa_ left	15:14
Mohjive	I just need some good (rather great) solution to this so we (I) can start moving into git (instead of this horrible piece of software called vss)	15:14
	→ MattDiPasquale joined	15:15
	→ ch077179 joined	15:15
Mohjive	the one weak point with branches I can see is you have to merge back manually, right?	15:15
bremner	yes	15:15
Mohjive	but that could be a good thing actually :)	15:16
	← ch077179 left	15:16
bremner	that is kind of the point of using version control, instead of symlinks, yes	15:16
j416	Mohjive: if it was automatically updated, testing would be a jungle	15:16
Mohjive	j416: yea, it is already...	15:17
j416	better to merge, test, distribute, than auto-update, not test, and distribute buggy versions....	15:17
	not saying that the latter is what you're doing now, but just for the point of it.	15:17
Mohjive	hehe, sure :)	15:18
j416	Mohjive: I'd create a 'master' branch that contains the common code for all platforms	15:18
	→ booginga joined	15:18
j416	work on that for common code	15:18
	then create one branch for each platform (or version of that platform even)	15:18
	merge master into that, or rebase onto master if that's what you prefer	15:19
	optionally extend your workflow with topic branches, bug branches etc.	15:19
	→ ch077179 joined	15:19
j416	Mohjive: this for example http://nvie.com/posts/a-successful-git-branching-model/	15:20
	but adapted to your case.	15:20
bremner	or, in non-sucky platforms/languages, just have a common directory for common files	15:20
Mohjive	that doesn't sound like a bad approach	15:20
bremner	using include / make paths etc...	15:20
	← theoros left	15:21
j416	the problem with this approach though, is that it tends to be tricky to develop on that common branch, because you'll likely need a specific version for testing	15:21
Mohjive	bremner: mm, but I also have to take the current workflow into account, even though a common directory would be possible even here	15:21
	→ theoros joined	15:21
	→ oriba joined	15:22
j416	and by specific version I mean you're likely on some platform and you want to try out your code there first	15:22
	but you can't commit it to the branch of that platform because it's common code.	15:22
	go figure	15:22
bremner	Mohjive: one thing to keep in mind, is what your deliverables are. I am used to open source, where a source distribution probably includes all platforms. Your case may be different of course	15:22
	→ MattDiPa_ joined	15:23
	← booginga left	15:23
j416	Mohjive: submodules is another option.	15:23
	keep the common code in a submodule	15:23
	might be more work than it's worth.	15:23
	j416 shrugs	15:23
Mohjive	thanks for the helpful tips - back to the drawing table to produce a presentable idea/solution :)	15:25
	← xorAxAx left	15:25
j416	Mohjive: this is in fact a problem I'm facing too	15:26
	← MattDiPasquale left	15:26
Mohjive	j416: that branch figure was nice	15:26
j416	having to develop on a common branch, but needing to have another branch checked out so that things'll work properly	15:26
	I'm considering making an alias that'll commit to the common branch, then automatically rebase the current branch on top of that	15:27
	but it seems a bit ugly.	15:27
	→ MattDiPasquale joined	15:27
j416	s/seems/feels/	15:27
Mohjive	bremner: mostly embedded systems, but software is developed on pc (windows or linux) before being moved "down"	15:27
	→ darkredandyellow joined	15:28
	→ MattDiP__ joined	15:28
bremner	Mohjive: I guess the point is that branching on platform makes it a bit harder to deliver snapshots of all platforms. But you have to decide if that is an issue.	15:29
	← MattDiPa_ left	15:29
j416	Mohjive: another solution is to have the "common" branch be a branch configured for the platform you're developing on, and have the other branches reconfigure that to their respective platforms	15:29
Mohjive	not when I think of it, we mostly focus on one platform at a time	15:29
j416	(having the platform-branches always rebased on top of your dev branch, or a stable master of that)	15:30
	→ RvnPhnx joined	15:30
Mohjive	I think branches is what I'll go for (for now)	15:30
	→ MattDiPa_ joined	15:30
Mohjive	it seems possible to move branches into directories later (I'm not that experienced with git yet)	15:31
	→ MattDi___ joined	15:31
	← MattDiPasquale left	15:32
	← Fboy left	15:32
j416	Mohjive: why not set them into directories in the first place and skip the work of moving them later	15:32
	→ GaVrA joined	15:32
Mohjive	please elaborate	15:32
	← MattDiP__ left	15:33
j416	I'm not sure what you are after	15:33
	but doing things the right way from the start tends to avoid some headaches in the future	15:33
	just that.	15:33
	:)	15:33
Mohjive	there are three issues, main issue being I'm currently forced to use vss	15:34
Tixz	Hi, How do I push updates from a submodule to it's 'parent' repo?	15:34
RvnPhnx	Tixz Are you in the submodule?	15:35
Mohjive	second is finding a way to make the managers accept git (for this I need a way to "translate" our vss repo to git)	15:35
Tixz	yes	15:35
	← MattDiPa_ left	15:35
j416	k hehe	15:35
	gl with that	15:35
	→ lu_zero joined	15:35
	← lu_zero left	15:35
	→ lu_zero joined	15:35
Tixz	RvnPhnx Yes it's from submodule to where I've cloned it from	15:35
Mohjive	which is my third issue	15:35
RvnPhnx	Mohjive vss?	15:35
Mohjive	visual sourcesafe	15:36
RvnPhnx	Tixz Is the upstream repo set?	15:36
Mohjive	something MS developed early nineties, it's a bit worse than CVS	15:36
RvnPhnx	Mohjive Oh, ok... the "v" was clogging my brain	15:36
	ugh	15:36
Tixz	RvnPhnx What do you mean? Sorry, I'm really new to Git	15:37
tango_	well if somebody could do worse than cvs, it had to be ms	15:37
RvnPhnx	Tixz You cloned into that submodule from another repo when you created it, right?	15:37
Tixz	Yes	15:37
	I used the git submodule init	15:38
	I have both repo's on my local machine	15:38
Mohjive	j416: my "private" issue is to translate our <platform>/<project>/<[shared] file> into something workable in git but that the management accepts	15:38
	← MattDi___ left	15:38
Mohjive	though I gtg, bbl and will read backlog	15:38
	→ MattDiPasquale joined	15:38
j416	Mohjive: quite a task you have there :)	15:39
RvnPhnx	Tixz So that means you have both "parents" on the local machine?	15:39
j416	convincing management that is	15:39
	good luck.	15:39
Tixz	RvnPhnx Yes, the repo that I have 'submoduled' is a set of classes that I want to maintain on it's own so I can easily update it across several projects	15:39
RvnPhnx	Tixz Ok, so which direction are we trying to send changes? From the "global" to the "local" or the other way around?	15:40
Mohjive	j416: yeah..., thanks for the help	15:41
	:)	15:41
j416	hope it works out. :)	15:41
RvnPhnx	j416 Sounds like he's got his work cut out for him...	15:41
j416	not sure what that means	15:41
	but I'll say that's a definite maybe	15:41
	← darkredandyellow left	15:42
Tixz	RvnPhnx From the local to the global... I just use git submodule update when I start out, but hten as I go along I might edit the cloned files, but I would like to commit them to the repo that I have cloned	15:42
	so I can update the other projects that depend on it and get the latest version	15:42
	→ kpreid joined	15:43
Tixz	RvnPhnx Do you get what I mean? I'm new to git, so I'm not sure I use the right terminology all the time	15:44
RvnPhnx	Tixz What does your .gitmodules in the topmost directory contain (please use a paste-bin; ask if you don't know what that means)?	15:46
	→ d0m joined	15:46
Tixz	RvnPhnx https://gist.github.com/813456	15:47
	All I want to do is to push/add/commit my local changes to the global repo	15:49
	RJD22 → RJD22\|away	15:50
RvnPhnx	Tixz I found what I was looking for: http://blog.jacius.info/git-submodule-cheat-sheet/	15:50
Tixz	So I should use git push origin master?	15:51
RvnPhnx	Sounds like what you're trying to do...	15:52
Tixz	Thanks a lot :)	15:52
	I'll try it out	15:52
	→ Zephyrus joined	15:54
Tixz	RvnPhnx I get an error :S	15:54
	RvnPhnx I get this error: https://gist.github.com/813456	15:55
	← yhager left	15:56
	→ yhager joined	15:56
	→ stigrett joined	15:56
RvnPhnx	Tixz, that's what I thought it would do.	15:57
	← stigrett left	15:57
	→ stigrett joined	15:57
Tixz	RvnPhnx How excatly do I fix it? I'm not sure I understand the error message	15:57
	← MattDiPasquale left	15:58
RvnPhnx	What will best serve you is to have a bare repository for that shared module and use THAT as the upstream for all instances of that module--including that which is TECHNICALLY your master copy.	15:58
	Then you can push/pull at will and get the expected results.	15:59
	← codespectator left	15:59
Tixz	So instead of using submodules, have a repo that I can push/pull from?	15:59
	→ DaveIngram joined	15:59
	→ codespectator joined	16:01
RvnPhnx	No, not quite	16:02
	Tixz you can still use submodules, but you can't push the changes back DIRECTLY to a non-bare shared master.	16:02
Tixz	Hmm, I'll have to read up on all this origin/master relationship stuff, so I can understand who it all works	16:03
RvnPhnx	So you'll need a global bare repository for that.	16:03
	Sounds like a good idea	16:04
	Good Luck!	16:04
SethRobertson	Tixz: Perhaps you should consider gitslave as an alternative to git-submodule, depending on your development workflow it may be better (or worse). However, it doesn't alter what RvnPhnx is talking about WRT a shared bare local master repo	16:04
	→ ExtraSpice joined	16:05
Tixz	Do you guys know a good place to get an easy introduction to Git? I mean, the Git Book is pretty heavy for beginners, better for reference	16:05
RvnPhnx	There was a posting a couple of weeks back to the git mailing list that had a ton of resources listed, but I don't have the reference URL for that handy right now (it is at work, and I'm not).	16:06
	→ LongBeach joined	16:06
	→ kukks joined	16:07
SethRobertson	The ProGit (or GitPro) book is good	16:07
RvnPhnx	The git wiki probably has most of them listed however	16:07
SethRobertson	People also recommend git for computer scientists	16:07
	Both listed in the wiki	16:07
	Or google	16:07
Tixz	Yeah, I tried Google, but it seems to turn up too many random blog posts	16:08
	I'll check out the ProGit book	16:08
SethRobertson	No, search for the book and document I mentioned	16:08
	← ExtraSpice left	16:09
Tixz	Cheers guys :)	16:09
	→ eekTheCa1 joined	16:12
eekTheCa1	I added a git repo as a sub folder of my repo	16:13
	the sub repo didn't get commited	16:13
	← eekTheCa1 left	16:13
	→ roop joined	16:13
	→ eekTheCat joined	16:13
	← eekTheCat left	16:16
	← gnufied left	16:16
	→ eekTheCat joined	16:16
	← gusnan left	16:16
	dwmw2_gone → dwmw2_AVF	16:16
	← stamina_ left	16:17
	→ gnufied joined	16:17
	← envi left	16:17
eekTheCat	I added a git repo as a sub folder of my repo. The folder did not show up in git status and did not get commited. How do I commit it?	16:18
	I removed the .git folder	16:18
	in the subfolder	16:18
	but it's still not showing up	16:18
	→ stamina_ joined	16:19
eekTheCat	in git status	16:19
charon	what are you really trying to achieve?	16:20
	eekTheCat: usually when importing whole repos, people would either make them a submodule (see git-submodule manpage) subtree-merge (see git-subtree on google)	16:20
eekTheCat	charon: i want to add the contents of the folder to my repository	16:20
charon	without history?	16:20
eekTheCat	charon: yup	16:21
	no need for a submodule	16:21
thiago_home	cp the folder, then git add	16:21
eekTheCat	it somehow automagically became a sub module	16:21
	← Buglouse left	16:21
	→ mithun joined	16:21
charon	if you add a directory that is in itself a git repo, it does become a submodule	16:21
eekTheCat	how do I unmake it a submodule?	16:22
	→ Anti-X joined	16:22
charon	you remove its entry from the index: git rm --cached $dir	16:22
SethRobertson	You may need to modify .gitmodules as well	16:22
	→ booginga joined	16:22
charon	(from the description of his steps i'm inferring .gitmodules was never involved)	16:23
eekTheCat	yeah. never made a .gitmodules :)	16:23
	charon: that seems to have done the trick. thank you	16:24
	← torvalamo left	16:26
	← NickyP left	16:26
	← SvenDowideit left	16:27
	→ rseifert joined	16:29
	→ rjnienaber joined	16:30
	→ becomingGuru_ joined	16:30
	→ puffandstuff joined	16:30
puffandstuff	hi	16:30
	→ awilkins joined	16:32
puffandstuff	Is there a way to find commits where function_foo() was changed (added, modified, deleted etc..) ?	16:32
	→ jmaya joined	16:33
	← roop left	16:33
	← d0m left	16:36
charon	puffandstuff: not yet, but you can try merging the WIP log -L feature and see if it works for you	16:37
	other than that the usual approach is to blame around the function, find the last commit C that changed it, look at C, and blame again in C^	16:37
awilkins	puffandstuff, You could `git annotate` the file	16:37
puffandstuff	thank you	16:38
awilkins	puffandstuff, If you're looking for a specific bug, and you have a test for it, you could also use bisect to find the revision that introduced it	16:38
thiago_home	puffandstuff: git log -p path/to/file, then grep for the function name	16:38
puffandstuff	thank you all :)	16:39
	← booginga left	16:40
	→ Ins\|de joined	16:40
	→ booginga joined	16:40
	← Utkarsh_ left	16:40
Ins\|de	hello, i have a project with some submodules, most of their origin url's changed, how do i proceed?	16:41
	should i just edit the .gitmodules manually ?	16:41
	→ tizzo joined	16:42
	← tizzo left	16:42
	← gnufied left	16:44
	→ kipras joined	16:44
	→ jfkw joined	16:44
	→ Error404NotFound joined	16:45
	← Error404NotFound left	16:45
	→ Error404NotFound joined	16:46
	→ skoop joined	16:48
	← tewecske left	16:52
	← jmaya left	16:52
	← Error404NotFound left	16:52
SethRobertson	You can do that, delete the existing repos, and ask to check them out again	16:53
	Or you can do that and then also change the upstream URLs in the sub-repos	16:53
	→ Buglouse joined	16:54
	→ ixti joined	16:55
	→ gnufied joined	16:57
	→ madewokherd joined	16:57
	→ luis_ joined	16:59
	← huehnts left	17:00
	← gnufied left	17:01
	← radarek left	17:06
	→ SvenDowideit joined	17:06
	→ knittl_ joined	17:06
	← hohoho left	17:06
	← knittl_ left	17:06
	← tizbac left	17:07
	→ sudonotp1ease joined	17:07
	← storrgie_ left	17:08
	← sudonotplease left	17:09
	← sudonotp1ease left	17:09
	→ ivanx joined	17:09
	← aah left	17:10
	→ aah joined	17:10
	← ivanx left	17:11
	→ tapank joined	17:11
	→ gnufied joined	17:11
	← pielgrzym left	17:13
	← tapank left	17:17
	→ tizbac joined	17:17
	← bitkiller left	17:17
	→ bitkiller joined	17:17
	← chrischambers left	17:18
	← madsy left	17:18
	← verdon left	17:18
	→ verdon joined	17:18
	→ electrogeek joined	17:20
	← electrogeek left	17:20
	→ electrogeek joined	17:20
	← puffandstuff left	17:21
	→ leo2007 joined	17:21
	→ madsy joined	17:21
	RJD22\|away → RJD22	17:22
	← electrogeek left	17:22
	← Tixz left	17:22
	← verdon left	17:25
	← awilkins left	17:25
rlb	What's the best way to programmatically determine if a tree is clean, i.e. that git commit would complain "nothing to commit"?	17:26
	← Ins\|de left	17:26
	→ chrischambers joined	17:27
doener	rlb: commit would complain about that when the index is clean, the working tree doesn't matter (unless you use commit -a). And you can check whether the index is clean using: git diff-index --cached --quiet	17:28
	→ verdon joined	17:28
	→ ISF joined	17:29
	→ bodom joined	17:29
	→ jhelwig_ joined	17:29
rlb	doener: OK, thanks -- I just want to check whether or not the tree is clean after applying an arbitrary patch (or archive), i.e. did it do anything git would commit.	17:29
	← Buglouse left	17:30
	← DiGitalX left	17:30
	→ pfrenssen joined	17:30
	← i_ron left	17:30
bodom	Hi!	17:31
	I've setup a plain ssh git server: i created a folder and git --bare inited it. Then I created some system users and added them to the group git and changed the folder permissions to rwxrws---	17:31
	→ ldh joined	17:31
	→ neilkod joined	17:31
bodom	Now my problem is than when a users makes a push the objects folder is created with rwxr-x---- permissions and other users will be unable to push until I do a chmod -R g+w *	17:32
rlb	doener: actually, that won't reallly work -- doesn't deal with new files, and I don't think I want --cached. Basically I want to test if there are any differences (incl new files) between the working tree and HEAD.	17:33
	→ radarek joined	17:33
bodom	Is there any way to set git default file permissions to 770?	17:33
	→ juangiordana joined	17:33
	→ puffandstuff joined	17:34
	← simplechat left	17:35
rlb	OK, I suppose 'test -z "$(git status --short)"' might be sufficient.	17:36
	← pfrenssen left	17:38
	← neilkod left	17:38
	← becomingGuru_ left	17:40
	← normanrichards left	17:41
	← rgr left	17:41
	← project2501a left	17:42
	→ ceej joined	17:42
	← MacGyverNL left	17:42
	→ MacGyverNL joined	17:42
	→ icwiener joined	17:43
	← bodom left	17:44
	← napster_123 left	17:46
	← anderse left	17:47
	← MacGyverNL left	17:48
CrazyGir	how does one clone a specific branch from a remote repo, as opposed to the master?	17:49
	← ixti left	17:50
doener	CrazyGir: clone creates remote-tracking branches for all remote branches by default	17:50
	CrazyGir: you can control for which of them a loacl branch head gets created using -b	17:51
	← zivester_ left	17:51
	→ zivester_ joined	17:53
	→ mikepack joined	17:53
	→ tcurdt joined	17:56
	← irfaN left	17:59
	← goshawk left	17:59
CrazyGir	ah	18:01
	ty	18:01
	→ electrogeek joined	18:03
	← electrogeek left	18:03
	→ electrogeek joined	18:03
	→ project2501a joined	18:04
	← project2501a left	18:04
	← dr_win left	18:04
	→ Davey_ joined	18:04
	← Davey_ left	18:04
	→ Davey_ joined	18:04
	→ Buglouse joined	18:06
	→ dr_win joined	18:07
	→ CyberDoc joined	18:08
	← CyberDoc left	18:09
	← kpreid left	18:10
Buglouse	Can i issue git pull out side of the 'git clone' dir?	18:16
	somehow specifying to check /dir	18:16
	← doener left	18:17
	← tizbac left	18:17
	→ doener joined	18:17
thiago_home	Buglouse: cd into the directory, then pull	18:18
	→ MrGlass joined	18:18
	→ xerora joined	18:19
	→ Slix` joined	18:20
	← Slix` left	18:20
Buglouse	trying to perform operation on git pull. pushd dir; git pull; popd; will return exit code from popd. just wanted to know if there was a more direct way of acting on git pull exit code. Ill just write more.	18:20
thiago_home	set GIT_DIR and GIT_WORKTREE	18:21
	or GIT_WORK_TREE	18:21
	Buglouse: or check the arguments to the git part of git pull (man git)	18:21
jast	Buglouse: the 'git' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git.html [automatic message]	18:21
	← zivester_ left	18:22
MrGlass	is there a way to use .ignore so that my folders are added to the repo, but none of the folder contents?	18:22
	sry, .gitignore	18:23
rlb	MrGlass: https://git.wiki.kernel.org/index.php/GitFaq#Can_I_add_empty_directories.3F	18:24
	← setmeaway left	18:24
MrGlass	hrm. it says I can use .gitignore as the file it needs to add the directory. my .gitignore files are just ''. is there a way to set that up as except .gitignore, or manually add .gitignore?	18:25
	add, git add -f	18:27
	thanks!	18:27
	→ normanrichards joined	18:28
	← RvnPhnx left	18:29
	← dfr\|mac left	18:31
	→ setmeaway joined	18:32
	→ sako joined	18:33
	→ cirwin joined	18:34
	→ matjas joined	18:35
	→ Fboy joined	18:35
	← PigeonFriend left	18:36
	← tcurdt left	18:39
	→ orafu joined	18:39
	← eekTheCat left	18:39
Buglouse	so git --work-tree=dir pull should pull from dir/.git into dir?	18:44
thiago_home	Buglouse: to be on the safe-side, set both	18:44
	--git-dir and --work-tree	18:45
Buglouse	just doesn't work as expected	18:45
CrazyGir	for those that might know and work with pip/python & git	18:45
	→ {ixti} joined	18:45
CrazyGir	how do I know pip is using the correct branch in git?	18:46
	→ kpreid joined	18:46
thiago_home	Buglouse: what did you expect and what happened?	18:46
cirwin	CrazyGir: if you don't want it to be master you have to put @branch on the end of the git repo uri	18:47
CrazyGir	sweet	18:47
	→ zombor joined	18:47
CrazyGir	ty!	18:47
	→ Elfe joined	18:48
Buglouse	i want to pull a git repo from outside the dir containing .git. using GIT_WORK_TREE=relative_dir GIT_DIR=relative_dir git pull. I receive either: fatal: no a repo: .git or fatal: no a repo 'relative_dir'	18:48
	no/not	18:49
cirwin	Buglouse: then just "git pull ../other/directory/"	18:50
	→ koo5 joined	18:50
	← {ixti} left	18:50
	← radarek left	18:50
Buglouse	thats what the man says i can do, but i am in an empty dir with only ./dir/.git as contents, and no go.	18:51
	→ SimonP86 joined	18:51
cirwin	what error do you get?	18:51
	→ domino14 joined	18:51
doener	Buglouse: GIT_DIR should likely be relative_dir/.git -- but IIRC pull doesn't correctly work with GIT_DIR/GIT_WORK_TREE	18:52
	→ pfrenssen joined	18:52
doener	cirwin: Buglouse doesn't want to pull from the repo given by relative_path, but pull without changing into that directory	18:52
domino14	i have a very simple question. if i do a few git commits, and i forget whether i did git push, is there a way to ask git how many commits ago i did a push?	18:53
cirwin	"""i want to pull a git repo from outside the dir containing .git. using""" is ambiguous :)	18:53
domino14	or whether the push is up to date	18:53
doener	Buglouse: I'd just do (cd whatever && git pull) the subshell avoids the problem with popd overwriting the exit code	18:53
	← ptl left	18:54
	← iravid left	18:54
cirwin	domino14: first do a "git fetch" to see what's around on the remote, and then use "git branch -v" to see which branches are ahead or behind	18:55
	→ iravid joined	18:55
	← JohnFlux_ left	18:55
	← jpieper left	18:55
	→ JohnFlux__ joined	18:55
	← dkasak left	18:55
	→ pantsman joined	18:55
	→ amerine joined	18:55
	← kpreid left	18:56
domino14	ok i did git branch -v and i get a little log line	18:56
	* master 12bfce5 removing vestigial 'everything is unchanged' message	18:57
	how do i know if this is in my own copy of the repository?	18:57
thiago_home	Buglouse: use a subshell then	18:58
	→ mikemeyer joined	18:58
Buglouse	doener: huh, git --git-dir=dir/.git pull works fine.. thought i tried that one. thanks	18:58
domino14	oh i understand, it's just listing the branches, there's just one	18:58
	→ resmo joined	18:59
domino14	so should i always do git pull before i start work? this is similar to svn update?	18:59
thiago_home	domino14: only if you want to	18:59
	→ dkasak joined	18:59
	← gnufied left	18:59
doener	Buglouse: that might have messed up the directory you were in, though. As you didn't set the work tree directory in that command	19:00
domino14	well, to avoid the pain of merging later, i guess	19:00
	does anyone find something bad with the idea of using a dropbox directory as a remote git repo?	19:00
thiago_home	what's a dropbox directory?	19:01
_rane	this comes up from time to time	19:01
thiago_home	and how do you access it?	19:01
_rane	thiago_home: dropbox.co	19:01
	+m	19:01
CrazyGir	cirwin: are you familiar enough with pip & git that I may ask you a few more such questions?	19:01
thiago_home	_rane: do you access it via DAV?	19:01
	→ neilkod joined	19:01
_rane	you access it via filesystem	19:01
CrazyGir	I'm having a difficult time getting my git repo to play nice with pip	19:01
cirwin	CrazyGir: I've no idea	19:01
CrazyGir	hah	19:01
	ok :)	19:01
thiago_home	_rane: so, via DAV?	19:02
jast	thiago_home: dropbox automatically mirrors changes to all subscribers to a folder	19:02
domino14	its kind of like using git on git	19:02
jast	I have no idea what the underlying protocol is	19:02
domino14	i'm a git n00b	19:02
	← pdelgallego left	19:02
thiago_home	sounds like a bad idae then	19:03
	some files in git are modified all the time	19:03
	← timj__ left	19:03
cirwin	the fs plugin for dropbox is pretty silly — it even tells everyone that I keep creating .swp files and deleting them	19:03
CrazyGir	hah	19:03
thiago_home	you can probably use it as a backup, but not for sharing	19:03
jast	cirwin: you can tell vim to store .swp files in a fixed location. it's what I do.	19:03
CrazyGir	jast: that's a great idea..	19:03
cirwin	jast: that would spoil my story :p	19:03
CrazyGir	thank you.	19:03
	→ magn3ts joined	19:03
CrazyGir	cirwin: do you use interesting names to give them clues to follow?	19:04
magn3ts	Can I somehow get git to automatically recognize which files I 'rm'ed rather than 'git rm'ed?	19:04
CrazyGir	:P	19:04
	← MetaCosm_ left	19:04
jast	anyway. the main disadvantage of storing it in a dropbox is that you couldn't do very efficient packs, because re-packing a pack might end up removing a pretty large file and replacing it with a new pretty large file	19:04
cirwin	CrazyGir: interesting names?!	19:04
	← koo5 left	19:05
jast	magn3ts: if you're fine with getting git to pick up all changes (modifications and deletions) at the same time, just use git add -u from the top level of the repository	19:05
magn3ts	jast, okay. also, now that I just accidently commited my entire target folder, can I undo the latest local commit before pushing this to github since it would still push them if I remove it from next commit?	19:05
cirwin	magn3ts: git reset HEAD^	19:06
jast	magn3ts: sure. get your index all fixed up to the right state, then use git commit --amend	19:06
	that will replace the previous commit with the new stuff	19:06
magn3ts	jast cool!	19:06
cirwin	python programmers should not like git — there's more than one way to do it..	19:06
jast	cirwin's suggestion pretends that you never had the previous commit, so you can use it to do the same thing	19:06
	→ intripoon joined	19:06
	→ ptl joined	19:06
jast	well, curiously enough I like git but not Python ;)	19:07
magn3ts	jast, but does his command cause me to loose local changes since the last commit?	19:07
jast	magn3ts: no, neither of them.	19:07
magn3ts	oh	19:07
jast	git reset only destroys uncommitted changes if you use the --hard flag	19:07
magn3ts	:) nice, thanks for the tips jast, cirwin.	19:07
jast	'git reset' without flags will just change the index contents and the state of the branch	19:08
	you can even keep the index intact, too, by using git reset --soft	19:08
	Ingmar_FOSDEM → Ingmar_home	19:08
	← theoros left	19:09
	← intripoon_ left	19:09
	→ kpreid joined	19:10
	→ radarek joined	19:10
	→ Cvbge joined	19:11
	ChanServ set mode: +v	19:11
	→ Thell joined	19:11
	→ rgr joined	19:12
	← Arrowmaster left	19:13
	→ engla joined	19:13
	← engla left	19:14
	→ engla joined	19:14
	→ EugenMayer joined	19:14
domino14	so what if i forget to do a git push after committing all day? is there a way to make git automatically push commits?	19:15
	→ timj__ joined	19:15
	→ dtillberg joined	19:15
bremner	you could use a cron job	19:15
cirwin	you could set up a post-commit hook, but you shouldn't	19:15
_rane	i ssh to my work system and push	19:15
	cirwin: worst idea ever	19:16
	← dtillberg left	19:16
	→ dantillberg joined	19:17
cirwin	_rane: you just wait for my next one :p	19:17
Buglouse	is this expected? git clone git://repo; cd dir; rm file dir; git pull; file and dir are not recreated. this is what git pull is supposed to do, correct?	19:17
bremner	Buglouse: rm or git rm?	19:18
Buglouse	rm	19:18
	/bin/rm	19:18
thiago_home	Buglouse: git pull won't recreate th files	19:19
jast	Buglouse: pull updates the branch with upstream changes; uncommitted changes, however, are not part of the branch (because, you know, they are not committed to it ;)).	19:19
thiago_home	that's the expected behaviour	19:19
	→ hagebake joined	19:19
	← hagebake left	19:19
	→ hagebake joined	19:19
jast	even so, "delete" is a change, too. if you delete a file, commit then, then pull updates that change the file, you'll get a so-called delete/modify conflict (git doesn't know whether you want the file back with the new changes or whether it should stay dead)	19:19
	→ storrgie_ joined	19:20
	→ elmob joined	19:20
jast	if you want to recreate the file, use something like git checkout -- <file>	19:20
Buglouse	jast: commit is not done though	19:20
jast	yes, hence what I said first applies	19:20
Buglouse	yeah	19:21
thiago_home	never leave uncommitted changes	19:21
jast	oh, well, it depends on the changes, right?	19:21
	I don't usually commit extremely experimental stuff or trace commands, for example :)	19:22
	→ mrconnerton joined	19:22
	→ theoros joined	19:22
	→ {ixti} joined	19:22
	→ comawhite joined	19:22
	← dantillberg left	19:22
bremner	as long as you are not upset when git reset --hard destroys them, feel free not to commit	19:22
	you presumably being the generic you, rather than jast ;)	19:23
jast	aww, I don't even get personalized advice anymore :(	19:23
	perhaps I should give myself advice instead	19:24
	hey jast, get back to working on your project	19:24
	oh, hey, thanks for the reminder!	19:24
bremner	jast: you should make the bot nag you	19:24
jast	jast: move move move [automatic message]	19:24
	I can't highlight myself :(	19:25
bremner	jast: that is why you need a bot with a different nick. offby1 has hours of fun with that.	19:25
jast	too much work :(	19:25
	Sgeo__ → Sgeo	19:27
	→ linusoleander joined	19:27
	→ domino141 joined	19:28
	← domino14 left	19:28
	← pfrenssen left	19:29
	← {ixti} left	19:31
	→ andrewvos joined	19:31
andrewvos	can anyone let me know what this means git push origin master:gh-pages	19:32
	is it saying push to "master and gh-pages"?	19:32
thiago_home	no	19:32
	it's source:destination	19:32
	push the contents of the local master branch to origin's gh-pages branch	19:32
andrewvos	ohhhh ok	19:32
	thanks!	19:32
	← mrconnerton left	19:35
	→ mizar99 joined	19:35
	→ MetaCosm_ joined	19:36
	← mizar99 left	19:36
	→ darkredandyellow joined	19:36
	→ emiel_ joined	19:37
	← bentob0x left	19:37
	← Davey_ left	19:38
	→ Kwpolska joined	19:40
	→ boombatower joined	19:42
	← boombatower left	19:42
	→ boombatower joined	19:42
Kwpolska	I have a problem with git. http://paste.pocoo.org/show/333213/	19:42
	← domino141 left	19:43
cirwin	Kwpolska: that looks normal	19:43
	what did you expect?	19:43
	→ koo5 joined	19:43
	→ domino14 joined	19:43
	← kpreid left	19:45
	← ptl left	19:45
	→ Davey_ joined	19:45
	← Davey_ left	19:45
	→ Davey_ joined	19:45
	← verdon left	19:45
	← domino14 left	19:46
	→ goshawk joined	19:46
Kwpolska	cirwin: I want to get rid of it	19:47
	→ MrShmelo12 joined	19:48
cirwin	get rid of the file?	19:48
	→ ptl joined	19:48
cirwin	git rm privpastebin	19:48
	← MrShmelo12 left	19:48
	→ Pupeno joined	19:48
Kwpolska	of the message	19:49
	I want to have privpastebin on the server	19:49
	→ {ixti} joined	19:49
	→ eago joined	19:50
Kwpolska	and by server, I mean both the github server and my shell server where I have my github repo	19:50
eago	hi, I have a repository in a usb flash drive... pulling works fine but pushing is very slow.... is there a way to make push faster?	19:50
cirwin	echo privpastebin >> .gitignore	19:51
	or to .git/info/exclude if you don't want to have a .gitignore file	19:51
	→ MacGyverNL joined	19:51
	← MacGyverNL left	19:51
	→ MacGyverNL joined	19:51
cmyers	eago: is it possible the write perf of the flash drive is just way slower than the read?	19:51
eago	nope, works fine when copying files manually	19:51
Kwpolska	cirwin: won't work	19:52
	→ oracal joined	19:52
cirwin	oh — you have local modifications to a committed file that you want git to ignore?	19:52
	git fails for doing that	19:52
	← dkasak left	19:53
	→ BlueAidan joined	19:54
	→ dkasak joined	19:54
BlueAidan	how does git calcuate the sha of a tree?	19:54
	I've been reading http://stackoverflow.com/questions/552659/assigning-git-sha1s-without-git for blob objects, but I can't find something similar for trees.	19:54
	← BlueAidan left	19:55
	→ BlueAidan joined	19:55
	← sako left	19:56
doener	BlueAidan: just use "tree" instead of "blob"	19:57
	← neilkod left	19:57
	← stamina_ left	19:57
doener	BlueAidan: but I guess you're actually looking for the format of trees, instead of just how to get the sha1 for objects storing them, right?	19:58
	→ psoo joined	19:58
	→ neilkod joined	19:58
	→ mjv joined	19:59
	← darkredandyellow left	20:01
	← neilkod left	20:01
	← zombor left	20:02
	→ NickyP joined	20:04
	← ptl left	20:04
	← mjv left	20:05
	→ necromancer joined	20:05
	→ ptl joined	20:05
	→ mjv joined	20:05
BlueAidan	doener: so the sha of a tree object is just based on the name? it doesn't reference any of the sha's that are inside it?	20:07
	→ Kimmono joined	20:07
cirwin	it's based on the contents	20:07
BlueAidan	I'm reading http://book.git-scm.com/7_raw_git.html , the "Creating Trees" bit.	20:08
	→ Chillance joined	20:08
BlueAidan	how does git go from	20:08
	100644 blob 6ff87c4664981e4397625791c8ea3bbb5f2279a3 file1	20:08
	100644 blob 3bb0e8592a41ae3185ee32266c860714980dbed7 file2	20:08
	to f66a66ab6a7bfe86d52a66516ace212efa00fe1f ?	20:08
	I tried just making a sha of the those 2 sha's in python, and didn't get that value.	20:09
	→ Mage joined	20:10
	← dkasak left	20:10
	→ dkasak joined	20:11
thiago_home	BlueAidan: it's the sha-1 of the binary representation	20:12
	the mode, sha-1 and the name	20:12
BlueAidan	but in that example from the book, mktree doesn't have the name of the tree object yet. Just its contents.	20:13
	→ tiagobutzke joined	20:13
	→ kar8nga joined	20:14
	→ davetoo joined	20:14
thiago_home	huh?	20:14
	← davetoo left	20:14
	← tiagobutzke left	20:14
engla	BlueAidan: you need to get the contents right, it's in a binary format	20:14
thiago_home	the binary format is:	20:14
engla	BlueAidan: but the format is pretty much based on your plain text there	20:14
	→ iksik joined	20:15
iksik	hello	20:15
thiago_home	<mode in ASCII> SP <filename> NUL <SHA-1>	20:15
iksik	is it possible to set an alias on commiter name/email ?	20:15
thiago_home	iksik: explain	20:15
	BlueAidan: try it: git cat-file tree f66a66ab6a7bfe86d52a66516ace212efa00fe1f	20:16
	→ kumbayo joined	20:16
BlueAidan	thiago_home: thanks	20:16
	← JohnFlux__ left	20:16
	→ JohnFlux__ joined	20:16
	← magn3ts left	20:17
iksik	for example i'm using few machines from where i'm commiting/pushing my stuff into central repo... on one machine i'm x@localhost and i'm tx@hub on another - there is a little mess in repo now ( when i'm generating statistics from commits )	20:17
	→ tewecske joined	20:18
thiago_home	iksik: ok	20:18
	iksik: it's only because you configured git to do that	20:18
	(or lack of proper configuring it to do otherwise)	20:18
iksik	yes, there WAS a lack of some user.name/user.mail there, but.. now, can i fix it somehow? ;]	20:19
thiago_home	you can use git filter-branch	20:20
	please note you'll need to re-clone in all of the machines	20:20
	and everyone else who has the repository will need to do that too	20:20
iksik	hm	20:20
	this filter-branch i need to run on machine where is the main repository hosted, right?	20:22
	→ tiagobutzke joined	20:25
BlueAidan	thiago_home: so from that example, would the first one as a python string be:	20:25
	'100644 file1\06ff87c4664981e4397625791c8ea3bbb5f2279a3'	20:25
	?	20:25
	because I'm still not getting the same sha as the book when calculating it via python.	20:27
	→ amondracorp joined	20:27
	← [jason63] left	20:29
	← tiagobutzke left	20:31
BlueAidan	trying to make sense of http://git.kernel.org/?p=git/git.git;a=blob;f=builtin/mktree.c;h=098395fda1932674e29f8d2cb332c722a6cb275f;hb=HEAD#l45 now	20:31
Caelum	will I get some sort of reply for my patch to the git list, even if no one likes the patch?	20:32
BlueAidan	seems to align with the format you said	20:32
jast	BlueAidan: git includes a small header indicating the object type in what it hashes	20:32
	← ChUrP left	20:33
BlueAidan	right, I gathered that from searching on the net for blob types, but I can't find what that header is for tree types	20:33
jast	the header is basically a null-terminated string containing the type, a space, and the length as a decimal number	20:33
	I don't know whether the null byte is included in the hash	20:33
BlueAidan	is the length just 0 for tree types?	20:33
jast	no, it's the size of the tree object in bytes	20:34
	← workmad3 left	20:34
BlueAidan	so the sum of the objects contained in it?	20:34
jast	formatted as a stringified number in that small header	20:34
	no	20:34
	→ workmad3 joined	20:35
BlueAidan	then how do you calculate it?	20:35
jast	suppose the tree object says "100644 a\0" plus 20 bytes sha1 sum, then the length is 29 bytes	20:35
	→ pdelgallego joined	20:35
jast	well, if you have the object, you know its length :)	20:35
BlueAidan	what is "a" here?	20:35
jast	the filename	20:35
	→ neilkod joined	20:35
jast	of the blob referenced in that tree	20:35
BlueAidan	but my point is that mktree doesn't accept a name, but still outputs a sha of the contents	20:35
	→ linusoleander_ joined	20:35
jast	not the name of the tree, but the name of one of the objects in the tree	20:36
BlueAidan	so if there are multiple files, it's the sum of the length of the names?	20:36
jast	(this particular contains only one object, called a)	20:36
	→ sgh_ joined	20:36
	← boombatower left	20:36
jast	no	20:36
	just take the tree object as a byte stream. count the bytes. that's the length of the tree object.	20:36
	→ steph021 joined	20:36
	← linusoleander left	20:37
	linusoleander_ → linusoleander	20:37
BlueAidan	ah well, I have to go to the grocery store.	20:37
jast	let's take your example tree: '100644 file1\06ff87c4664981e4397625791c8ea3bbb5f2279a3'	20:37
BlueAidan	bbl	20:37
jast	its length is 33 bytes	20:38
BlueAidan	sorry, I'll read when I get back	20:38
jast	so the header is "tree 33"	20:38
	plus maybe a null byte	20:38
	right, definitely plus a null byte	20:38
	but actually your example didn't encode the sha1 sum correctly. you used a hex string representation, but it's actually stored as bytes in the tree object	20:39
	→ xiong joined	20:39
jast	so, '100644 file1\x00\x6f\xf8\x7c' and so on	20:39
	→ ChUrP joined	20:40
	← pdelgallego left	20:40
	← elmob left	20:43
sgh_	Hi! I'm wondering if it is possible to get at which merge a particullar commit has been merged in. Just like git-merge-base, just the other way ?	20:43
thiago_home	sgh_: not easily, no	20:44
sgh_	thiago_home: Though so ;( ... Do you know a hard way to do it?	20:45
	→ moosey joined	20:45
	→ coderdad_ joined	20:46
thiago_home	sgh_: walk forward from your commit until the other commit becomes past history	20:46
	the "walk forward" part is the hard part	20:47
	→ zombor joined	20:48
	← zombor left	20:48
	→ zombor joined	20:48
	← marc_kdab left	20:49
	→ da5id joined	20:49
sgh_	thiago_home: ok ... so a shell-script would be the best solution. Thanks.	20:50
	→ mDuff joined	20:51
mDuff	Can I hand-edit the staged diff without unstaging it and re-adding with "git add -p"?	20:51
teuf	mDuff: there's a git reset -p	20:51
mDuff	teuf, ahh; thanks!	20:52
	→ pheaver joined	20:52
	→ elmob joined	20:53
cirwin	can I persuade git to do rename detection whitespace insensitively?	20:53
	← koo5 left	20:53
thiago_home	cirwin: you can persuade git to do rename detection, or not do it.	20:54
cirwin	so if I do git show -M -w, that is as good as it gets	20:54
	?	20:54
thiago_home	yes	20:54
	→ koo5 joined	20:54
cirwin	I am surprised by the result	20:54
	ah well	20:55
thiago_home	try --find-copies-harder	20:55
	← pheaver left	20:55
cirwin	no change,maybe there's just too much deleted stuff	20:56
thiago_home	the copy/rename detection is based on file similarity	20:56
	that doesn't do whitespace trimming, it has to be the exact same contents	20:57
cirwin	that's annoying	20:57
	← kumbayo left	20:57
thiago_home	hint: rename, commit, modify the file, commit	20:58
cirwin	but then you have a commit at which the tests don't pass ;)	20:58
	it's not a big issue, but it seems like a fairly fixable one	20:58
thiago_home	not all commits pass tests	20:58
	→ koltroll joined	20:58
cirwin	no, and it's not even a requirement for me	20:58
	though I know of teams where it is	20:58
thiago_home	it's more important to have the logical changes than testable	20:59
cirwin	that depends entirely on your use case	20:59
thiago_home	then copy the file, commit, modify	21:00
	then remove the other	21:00
cirwin	doable too	21:01
	→ malte_ joined	21:02
	← normanrichards left	21:02
	→ giallu joined	21:03
	← giallu left	21:03
	→ giallu joined	21:03
	← Fboy left	21:03
	← Chillance left	21:04
	← coderdad_ left	21:05
	→ coderdad joined	21:05
	necromancer → necromancer`bzzy	21:05
	{ixti} → ixti	21:05
	→ marc_kdab joined	21:06
	→ stamina_ joined	21:07
	→ boombatower joined	21:07
	← boombatower left	21:07
	→ boombatower joined	21:07
	← kar8nga left	21:07
	← ppawel left	21:08
	→ shine-neko joined	21:08
	→ da5id_ joined	21:08
	→ da5id__ joined	21:08
	ChanServ set mode: +v	21:09
shine-neko	Ha !	21:09
	Hello all	21:09
	→ pgib joined	21:09
j416	死ね	21:09
shine-neko	I have a problem with git push for a http://pastebin.com/fKCXkb10	21:11
	← neilkod left	21:11
thiago_home	you can't push to a pastebin	21:12
	can you summarise what the problem was?	21:12
	→ Takechan joined	21:12
	→ artefon joined	21:12
	← Takechan left	21:12
cirwin	thiago_home: maybe the pastebin contains some relevant info :p	21:13
thiago_home	he didn't make a full sentence in the channel...	21:13
shine-neko	thiago, error: unable to find 95bda9a7456665dc4910ea06ec8126ccfd82cb45	21:14
	error: pack-objects died with strange error	21:14
thiago_home	try git fsck on your repository	21:14
	it looks broken	21:14
shine-neko	thiago, How can I fix it?	21:15
thiago_home	has fsck finished running?	21:19
	→ panike joined	21:19
shine-neko	thiago, I already tried : git fsck --full	21:19
	← koltroll left	21:19
thiago_home	it finished without errors?	21:21
shine-neko	thiago_home, yes	21:22
thiago_home	I don't know then	21:22
shine-neko	:s	21:22
	← iksik left	21:23
	→ Titosemi joined	21:23
	→ tizbac joined	21:24
	← ijuma_ left	21:25
	← sgh_ left	21:25
	→ montylounge joined	21:28
	→ ereslibre joined	21:30
	← ereslibre left	21:30
	→ ereslibre joined	21:30
	→ justin-george joined	21:32
	← montylounge left	21:32
	← amerine left	21:34
	→ primalmotion joined	21:35
	← ChUrP left	21:36
	← doener left	21:39
	← Cvbge left	21:39
	← tizbac left	21:39
	→ bradleyayers joined	21:40
	← LongBeach left	21:40
	← pgib left	21:41
	→ ChUrP joined	21:42
	← radarek left	21:43
	→ MattDiPasquale joined	21:44
	→ harryman100 joined	21:46
harryman100	Hi, when I do git am something.mbox, i get a 'mkdir: cannot create directory `.dotest': Permission denied'. Anyone know why?	21:47
	← theoros left	21:48
	→ treak007 joined	21:48
	→ sterlo joined	21:48
	→ radarek joined	21:49
	← necromancer`bzzy left	21:50
charon	presumably because git am wants to create a .dotest directory and you don't have permission to do so?	21:55
	← emiel_ left	21:57
	← paszo2008 left	21:57
	→ p3rg2z joined	21:58
	← TrineLocke left	21:58
ceej	is there a way to update all submodules ?	21:58
harryman100	charon: that's the obvious answer, but I can't work out where it's trying to create it	21:59
	→ theoros joined	21:59
	→ Fboy joined	21:59
	← goshawk left	22:00
	→ tiagobutzke joined	22:01
	→ LaChouffe joined	22:01
	← primalmotion left	22:03
	← warthog9 left	22:03
	← eago left	22:03
	→ primalmotion joined	22:03
	→ gantman joined	22:03
gantman	is this the right place to get help for gitolite?	22:04
	← puffandstuff left	22:04
	→ amerine joined	22:05
	← Fboy left	22:06
	← icwiener left	22:06
charon	harryman100: i think it was in the toplevel of the worktree or something	22:06
	and unless i'm mistaken it also means you have a very old git	22:06
	since it was changed to .git/dotest or something ages ago	22:06
harryman100	charon: thanks, that would make sense	22:07
	← theoros left	22:09
	← zombor left	22:09
	← sterlo left	22:10
	→ theoros joined	22:11
	← theoros left	22:11
	→ theoros joined	22:11
	→ darkredandyellow joined	22:14
	← treak007 left	22:16
ChrisPitzer	hey - at some point in the history of my project someone deleted a function called foo_bar_asdf (a relatively specific name that doesn't matter). How can I grep the changes to find a changeset that involved that text?	22:17
	the tricky thing is I don't know what file it was in.	22:17
	← ThomasLocke left	22:17
	→ pdelgallego joined	22:18
	← MetaCosm_ left	22:18
panike	git rev-list HEAD \| while read commit; do git grep foo_bar_asdf $commit && break; done	22:18
	← ChUrP left	22:20
	← matjas left	22:20
cirwin	git log -Sfoo_bar_asdf	22:20
	← leavittx left	22:20
	→ MetaCosm_ joined	22:22
panike	That is probably better	22:22
cirwin	:)	22:22
	← MetaCosm_ left	22:23
	← MetaCosm left	22:23
	→ MetaCosm_ joined	22:23
	← MetaCosm_ left	22:23
	→ MetaCosm_ joined	22:23
	← radarek left	22:24
	← darkredandyellow left	22:25
	← bradleyayers left	22:25
	→ Heimidal joined	22:26
	→ zenom joined	22:28
	→ LotR joined	22:29
	→ zombor joined	22:30
	← j0su3 left	22:30
	→ MetaCosm joined	22:30
	← MetaCosm left	22:30
	→ MetaCosm joined	22:30
	← Kimmono left	22:31
	← Zephyrus left	22:31
BlueAidan	jast: sorry I had to run earlier. You had said that the length of '100644 file1\06ff87c4664981e4397625791c8ea3bbb5f2279a3' was 33 bytes. how did you calculate that? roger on the hex version of the digest.	22:33
	it should have been '10064 file1\0o\xf8\|Fd\x98\x1eC\x97bW\x91\xc8\xea;\xbb_"y\xa3'	22:34
	← zombor left	22:34
	← eijk left	22:35
	← p3rg2z left	22:35
	← LaChouffe left	22:36
	← steph021 left	22:36
	→ NewtonianB joined	22:36
	← coderdad left	22:37
	← j416 left	22:37
	→ airborn_ joined	22:37
	→ opalka joined	22:37
	← opalka left	22:37
	→ opalka joined	22:37
	← airborn left	22:38
	← malte_ left	22:38
	→ darkredandyellow joined	22:41
	← EugenMayer left	22:41
	→ j0su3 joined	22:42
	← urbanmonk left	22:42
	← NewtonianB left	22:43
	← ereslibre left	22:44
	← ixti left	22:44
	→ eletuchy joined	22:47
	→ malte_ joined	22:49
	→ DiGitalX joined	22:49
	← oriba left	22:51
	→ angelsl joined	22:52
	← cesario left	22:52
	← storrgie_ left	22:54
	← leo2007 left	22:57
	← nairb774 left	22:58
	→ nairb774 joined	22:58
	← amerine left	22:59
	← marc_kdab left	22:59
	← andrewvos left	23:02
	← ph^ left	23:02
	→ storrgie_ joined	23:03
	← orafu left	23:05
	← ludde left	23:06
	→ marc_kdab joined	23:07
	← elmob left	23:08
	← artefon left	23:08
	→ neilkod joined	23:10
	→ nothingHappens_ joined	23:11
	← shine-neko left	23:12
	← xiong left	23:13
jast	BlueAidan: well, 33 seemed to me to be the length of that string.	23:13
	← LouisJB left	23:13
	← rgr left	23:13
jast	BlueAidan: (using the correct way to encode the sha1 ID, that is)	23:13
BlueAidan	it was 52. :) so, I just need to add up the lengths of the mode, space, filename, nul byte, digest version of the sha ?	23:15
	← bitkiller left	23:15
	→ shruggar joined	23:15
BlueAidan	so when there are multiple files, do you include a \n between each?	23:15
	← DiGitalX left	23:15
	← opalka left	23:15
	→ tiagobutzke_ joined	23:16
	← Heimidal left	23:16
	→ opalka joined	23:17
	← opalka left	23:17
	→ opalka joined	23:17
BlueAidan	nvm, you were right! not sure what I was looking at	23:17
	it was 33.	23:17
	← tiagobutzke left	23:18
	→ carlocci joined	23:19
	→ molney joined	23:19
	← molney left	23:19
	← MacGyverNL left	23:20
	← ptl left	23:21
	← tiagobutzke_ left	23:21
BlueAidan	jast: so, to calculate the sha of a tree, it'd be: sha1("tree " + sum_of_all_file_descriptors + "\0" + all_file_descriptors) ?	23:21
jast	it's not the number of tree entries, it's the size of all tree entries taken together	23:22
	← d0k left	23:22
jast	truth be told, I don't look at the internal formats much, so I can't tell you about every last detail	23:22
	I had to look up the part about the header	23:23
thiago_home	BlueAidan: the contents of a tree object are what I told you several hours ago. The SHA-1 of a tree object is "tree " + the size of the object + NUL + the contents of the object	23:23
	BlueAidan: why do you want to calculate the SHA-1 of a tree object outside of Git?	23:23
	→ MacGyverNL joined	23:23
	← MacGyverNL left	23:23
	→ MacGyverNL joined	23:23
BlueAidan	thiago_home: granted. I think I was having trouble because I was using the hex version of the digest	23:24
jast	to make sure that git doesn't make mistakes ;o)	23:24
BlueAidan	thiago_home: I like to know how things work. :)	23:24
thiago_home	the SHA-1 of any object is: object name + space + object size + nul + object contents	23:24
	s/name/type/	23:24
	← rseifert left	23:25
	← Ruudjah left	23:26
	← boombatower left	23:26
	← neilkod left	23:27
	← marc_kdab left	23:27
	→ orafu joined	23:27
	→ jlewis joined	23:27
	→ normanrichards joined	23:29
	→ ptl joined	23:29
	→ xiong joined	23:30
	→ marc_kdab joined	23:30
	← kipras left	23:31
	→ tiagobutzke joined	23:32
	→ bitkiller joined	23:32
	→ stringoO joined	23:34
	→ urbanmonk joined	23:34
BlueAidan	gah, I can't duplicate the sha from http://book.git-scm.com/7_raw_git.html for creating the first tree of the 2 files	23:34
	→ boombatower joined	23:35
	← boombatower left	23:35
	→ boombatower joined	23:35
FauxFaux	BlueAidan: You can't guess the contents of myfile.txt that has a hash of 6ff87c4664981e4397625791c8ea3bbb5f2279a3?	23:36
	→ deck_ joined	23:36
	← aro left	23:36
	← DivineOmega left	23:37
	→ aro joined	23:37
BlueAidan	huh?	23:37
FauxFaux	That was my interpretation of your questions.	23:38
	Question.	23:38
deck_	Sorry for newbie quistion. ive just clone a project by ssh, and cant write to there. should i add ssh key by ssh-add ?	23:38
	→ theoros_ joined	23:38
FauxFaux	deck_: You've successfully cloned it over ssh:// and can't push to it?	23:39
deck_	FauxFaux, i just cant creat a file in the folder	23:40
	only read	23:40
	← urbanmonk left	23:40
FauxFaux	You've cloned it to a local directory and you can't write to said local directory (the one you've just written to)?	23:40
BlueAidan	FauxFaux: I'm trying to duplicate the f66a66ab6a7bfe86d52a66516ace212efa00fe1f sha by calculating it manually. Do I have to have to know the contents of the 2 files to do that? (I.e., their sha is not enough info)?	23:40
	→ urbanmonk joined	23:40
FauxFaux	BlueAidan: f66a66ab6a7bfe86d52a66516ace212efa00fe1f is the hash of the contents of the file. If you don't have the same file, you won't get the hash.	23:40
	← theoros left	23:41
deck_	FauxFaux, yes	23:41
	→ mrconnerton joined	23:42
FauxFaux	deck_: Odd. Unless you ran the clone as another user, the clone was obviously able to write to the directory... SSH keys won't do anything for a local directory.	23:42
	← linusoleander left	23:43
	← HG` left	23:43
	→ linusoleander joined	23:44
	← tiagobutzke left	23:44
	← pantsman left	23:44
	← storrgie_ left	23:46
	← Buglouse left	23:46
deck_	FauxFaux, sorry, but what should i do to start to work with a project ?	23:46
	← darkredandyellow left	23:47
FauxFaux	You have a local filesystem problem, not a git problem; assuming you're on linux you can go familiarise yourself with ls -al and chown/chmod and try and work out what you did wrong.	23:47
	→ darkredandyellow joined	23:48
	→ tiagobutzke joined	23:48
	→ riddochc joined	23:49
deck_	FauxFaux, yes, i did what i want. thx. and now, what should i do not to write password for each action (push,pull)	23:54
	→ josue joined	23:55
	→ ping-pong joined	23:55
	← pedrobelo left	23:56
carlocci	how would you migrate to git? The guys I'm trying to help wrote a web framework, never used anything to handle the code	23:58
	← j0su3 left	23:58
carlocci	sometimes they fix stuff on the production server without "front"porting it to the dev server, sometimes they forget to push stuff to the production server	23:58
	so the codebases are diverging over time	23:59
	→ storrgie_ joined	23:59
cirwin	carlocci: if you don't have any history	23:59
	just start using git	23:59
FauxFaux	carlocci: Ouch. /o\ There's various guides on how to host websites directly from git; so they can push to the production server.. allow no interactive logons to the production server, only allow pushing there from the dev server etc?	23:59

Logs Search ←Prev date Next date→

Channels Documentation