IRCloggy #git 2012-05-26

Provider of IRC logs since 2005.

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

	2012-05-26
	← nitrospectide left	00:01
	→ enmand joined	00:03
	← Pookz left	00:06
	→ fuzzybunny69y joined	00:07
	→ jkyle joined	00:08
	→ beneggett joined	00:09
	→ krz joined	00:10
	← newbie999 left	00:10
	→ newbie999 joined	00:11
	← arietis left	00:11
	← bitslip left	00:12
	→ bitslip joined	00:12
	→ arietis joined	00:12
	← krz left	00:12
	← Y_Ichiro left	00:13
	→ Y_Ichiro joined	00:13
	← Y_Ichiro left	00:13
	→ Y_Ichiro joined	00:13
	← jkyle left	00:13
	← kenperkins left	00:13
	← SkAzZ1 left	00:16
	← jstout24 left	00:16
	← jeax left	00:17
	← aselder left	00:21
	← FernandoBasso left	00:21
	← Targen left	00:21
	→ jkyle joined	00:22
	← btanaka left	00:24
	← jkyle left	00:28
	← Eridius left	00:28
	← neurodrone left	00:28
	→ Eridius joined	00:34
	← dreiss left	00:34
	← arietis left	00:35
	← davidbanham left	00:36
	← ColKurtz left	00:40
	→ DofLeFou joined	00:41
	← eijk_ left	00:42
	→ cagedwisdom joined	00:43
	← phillllijw left	00:45
	← AaronMT left	00:46
	→ igotnolegs joined	00:46
	→ g00s joined	00:49
g00s	hi, i had 2 branches: master and hotfix … i was in master and merged hotfix. noticed a mistake and did commit --amend. now when i try to delete the hotfix branch i see 'error: The branch 'hotfix' is not fully merged.'	00:51
	not sure what to do now :\|	00:51
cbreak	g00s: you amended a merge commit?	00:52
	that's not a good idea	00:52
g00s	it was a fast forward	00:52
cbreak	oh, excellent	00:52
	then you can decide:	00:52
k-man	when you merge a branch into master, does it delete the branch?	00:52
cbreak	are you happy with the new version of the commit you amended?	00:52
	no	00:53
	merging only affects the current branch	00:53
	g00s: if you are happy, you can git branch -D otherbranch	00:53
	that'll delete it for sure.	00:53
	← mikepack left	00:53
g00s	cbreak: ok, thanks … i'm new to git, so i'm worried that there is something on hotfix which actually did not get merged.	00:54
cbreak	yeah, the last commit	00:54
SethRobertson	`git branch -d` if you want to be safe and only delete it if it was properly merged. Note it will not delete on any other system	00:54
cbreak	your amending changed the last commit so the last commit on the original branch is in fact not merged	00:55
g00s	oh	00:56
	i guess i don't understand the data model well enough, i thought that the fast forward just put the hotfix pointer to the last commit, and the -ammend just changed that commit	00:58
	← zarac left	00:58
	← mcluskydodallas left	01:00
cmn	but you only changed the branch you were on	01:01
	← Slix` left	01:01
	← pdtpatrick left	01:01
	← lonewulf` left	01:01
cmn	but if you merged hotfix into master, what the ff merge did was point master to the same commit as hotfix was pointing to	01:01
	← thearthur left	01:01
cmn	and then you changed that commit, so they're no longer pointing to the same commit	01:02
	← Textmode left	01:02
cmn	of course, you can't change a commit, what you do is create a new one with the same parent; this may be where your confusion comes from	01:03
	→ lonewulf` joined	01:03
	← apok left	01:04
g00s	cmn … hmm, ok	01:04
	→ ryez joined	01:05
	← Eridius left	01:06
	→ Eridius joined	01:06
	← diegoviola left	01:06
	→ diegoviola joined	01:07
g00s	cmn, but i only have 2 commits for the whole repo - the first was the initial add, the second was the fix (which was checked in on the hotfix branch). i don't see a 3rd commit	01:08
	← fisted left	01:08
cmn	how have you looked?	01:08
g00s	since it was a simple fast forward, i guess thats why i only see two commits after the merge	01:08
	i just said 'git log'	01:08
cmn	so there's only two commits on that branch	01:08
	→ erkules\|away joined	01:08
cmn	it doesn't mean there aren't more in the repo	01:09
	git log --graph --decorate --all	01:09
	or gitk --all	01:09
g00s	oh my	01:09
	:O	01:09
cmn	you've just experienced what often takes more time to get, to !rewriting_public_history	01:10
gitinfo	Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to `git pull --rebase` and even worse things if they have tagged or branched, so you must publish your humiliation so they know what to do. You will need to `git push -f` to force the push. The server may not allow this. See receive.denyNonFastForwards (git-config)	01:10
cmn	in this case, you did it against yourself	01:10
	← TTSDA left	01:11
	→ fisted joined	01:11
	← beneggett left	01:11
	→ JEntrep joined	01:11
	← erkules left	01:11
	← dermoth_ left	01:12
	← JEntrep left	01:12
	→ mcluskydodallas joined	01:12
	→ Vampire0 joined	01:14
	← kcj left	01:14
	← RaceCondition left	01:15
	→ kcj joined	01:16
	← ill1cit left	01:16
	→ ill1cit joined	01:16
	← Vampire0_ left	01:16
	← jfischoff left	01:16
	→ ColKurtz joined	01:17
	→ TTSDA joined	01:18
	→ Targen joined	01:18
	← ymasory left	01:18
g00s	cmn. i just read through Pro Git , the branching and merging chapter, and it all seemed so straightforward, now i realize that i have no clue what is going on	01:19
	this is just a local repo	01:19
	← Lgb left	01:20
cmn	yes, it's local, but the effect is analogous to what happens when you rewrite public history	01:21
	you now have two slightly different commits	01:21
	it's all straightforward as long as you don't start doing odd things like changing commits after you've considered them good and merged them somewhere else	01:21
	← mcluskydodallas left	01:22
	→ Lgb joined	01:22
	← goshawk left	01:23
	← xiaotian left	01:23
	← mwrosen left	01:24
	→ intripoon joined	01:25
	→ dermoth_ joined	01:26
	← Yuuhi left	01:26
	← kadoban left	01:27
	← intripoon_ left	01:28
	← fuzzybunny69y left	01:29
	→ fayimora_ joined	01:31
	→ kenperkins joined	01:33
	← vitor-br left	01:34
	← williamcotton left	01:35
	← mocramis left	01:35
	← fayimora left	01:35
	fayimora_ → fayimora	01:35
	→ mocramis joined	01:35
	← kenperkins left	01:36
	← optixx left	01:40
	→ pheaver_ joined	01:40
	← pheaver_ left	01:40
	→ kerrick joined	01:41
	← Eridius left	01:42
	→ davidbanham joined	01:42
	→ optixx joined	01:42
	← Chillance left	01:43
	→ vitor-br joined	01:43
	← pheaver left	01:43
	← berserkr left	01:46
	← mocramis left	01:47
	→ mocramis joined	01:48
	← DofLeFou left	01:50
g00s	it would be kinda cool, when i go to git-scm.com, that 'Download for the mac' would not show 'Latest stable release: 1.7.10.3' above it … and then you go click download and it starts downloading 1.7.10.1 :\|	01:51
	← lolfrenz left	01:53
	→ xiangfu joined	01:58
	← gchristensen left	02:00
	← maletor left	02:01
	→ kenperkins joined	02:02
	← chihiro left	02:03
	← jrockway left	02:03
	← xpot-station left	02:03
	← ivan\ left	02:05
	← notmypudding left	02:06
	→ ivan\ joined	02:06
	← kenperkins left	02:08
	→ adityamenon joined	02:09
	→ adityamenon90 joined	02:10
	← JStoker left	02:12
	← davidbanham left	02:13
	← adityamenon left	02:14
	← optixx left	02:15
	← jpieper left	02:16
	← pandeiro left	02:17
	→ optixx joined	02:18
	→ eurek joined	02:20
	→ kenyon joined	02:20
	← lonewulf`tp left	02:21
	→ pandeiro joined	02:22
	→ dhruvasagar joined	02:22
	← twisted` left	02:22
	→ Chib joined	02:23
	→ twisted` joined	02:24
	← eurek left	02:24
	→ hachque joined	02:26
	→ ISF joined	02:26
	← igotnolegs left	02:30
	← hachque left	02:31
	← pandeiro left	02:32
	→ lonewulf`tp joined	02:33
	→ glennpratt joined	02:33
	← cagedwisdom left	02:39
	← eletuchy left	02:39
	← zell left	02:39
	→ eletuchy joined	02:39
	← xiangfu left	02:40
	→ cagedwisdom joined	02:42
	← fayimora left	02:46
	→ fayimora joined	02:46
	← lonewulf`tp left	02:47
	→ HyperNerdV2_ joined	02:48
	← Vinnie_win left	02:49
	→ DofLeFou joined	02:50
	← HyperNerdV2 left	02:51
	→ Vinnie_win joined	02:51
	← HyperNerdV2_ left	02:52
	→ gavin_huang joined	02:56
	← eletuchy left	02:58
	→ williamcotton joined	02:59
	← optixx left	03:00
	← bitkiller left	03:00
	← mtx_init left	03:01
	→ optixx joined	03:01
	→ edude03 joined	03:02
	← Dreamer3 left	03:02
	→ neurodrone joined	03:03
	← fayimora left	03:03
	← glennpratt left	03:04
	← matthiaskrgr left	03:05
	→ dreiss joined	03:08
	→ bubbles10101 joined	03:08
	→ Lord_DeathMatch joined	03:10
	→ zell joined	03:11
	→ matthiaskrgr joined	03:14
	← LekeFly left	03:16
	← w09x left	03:17
	← matthiaskrgr left	03:18
	→ glennpratt joined	03:21
	→ lonewulf`tp joined	03:22
	→ mtx_init joined	03:24
	→ cakehero joined	03:28
	← optixx left	03:28
	→ Sahua joined	03:28
	← twisted` left	03:29
	← traviscj left	03:30
	← williamcotton left	03:31
	← dmnd left	03:34
	→ optixx joined	03:37
	← dreiss left	03:39
	→ rchavik joined	03:39
	← rchavik left	03:39
	→ rchavik joined	03:39
	← brngardner left	03:41
	← Sahua left	03:42
	← kukks left	03:42
	→ Dreamer3 joined	03:44
	→ Sahua joined	03:46
	← kcj left	03:49
	ChanServ set mode: +v	03:49
	← OOPMan left	03:51
	← vitor-br left	03:51
	→ ipoval joined	03:51
	← enmand left	03:55
	→ kcj joined	03:58
	→ rurufufuss joined	04:01
	← neersighted left	04:03
	→ lonewulf`tb joined	04:04
	→ fayimora joined	04:04
	→ traviscj joined	04:05
	→ williamcotton joined	04:09
	← brokenshakles left	04:10
	→ neersighted\|AFK joined	04:10
	neersighted\|AFK → neersighted	04:10
	→ fuzzybunny69y joined	04:11
	→ OOPMan joined	04:12
	← fuzzybunny69y left	04:13
	→ pheaver joined	04:14
	→ HyperNerdV2 joined	04:15
	→ fuzzybunny69y joined	04:15
	→ brokenshakles joined	04:28
	← palyboy left	04:29
	→ palyboy joined	04:29
	← Contexual left	04:29
	← g00s left	04:31
	← madewokherd left	04:31
	→ kadoban joined	04:33
	← adityamenon90 left	04:33
	← glennpratt left	04:33
	→ adityamenon90 joined	04:33
	← Mowsh left	04:41
	→ kenperkins joined	04:41
	← dzonder left	04:42
	→ warthog9 joined	04:45
	EugeneKay drinks	04:49
	→ mangdood_ joined	04:52
	← mangdood_ left	04:54
	← barrbrain_ left	04:54
	→ adityamenon joined	04:57
EugeneKay	Opinion. Should I relicense my !deploy hook as WTFPL?	04:57
gitinfo	Git is not a deployment tool, but you can build one around it for simple environments. Here is an example hook to get you started: http://goo.gl/l68hq	04:57
	← adityamenon90 left	04:59
	← bubbles10101 left	05:01
	→ cirwim joined	05:01
	→ igotnolegs joined	05:02
	→ bubbles1_ joined	05:03
	← HyperNerdV2 left	05:03
	→ efreese joined	05:04
	← klj613 left	05:05
offby1	EugeneKay: I think that's a ringing endorsement	05:05
	offby1 has a drink	05:05
offby1	Richard Gere has a drink	05:06
	Shelley Long has a drink	05:06
EugeneKay	It's my favorite FOSS license	05:06
	← Slush-_ left	05:07
	← snax left	05:08
	← __CA__ left	05:08
	→ AlbireoX`Laptop joined	05:08
	→ WAA1BUC joined	05:09
EugeneKay	Tada.	05:09
	→ __CA__ joined	05:09
	offby1 is enjoying his Foxtrot Tango Whiskey.	05:09
offby1	Very peat-y.	05:09
EugeneKay	Mmmm whiskey	05:09
	← bubbles1_ left	05:13
	→ dreiss joined	05:13
	← cakehero left	05:19
	→ g00s joined	05:19
	→ Sigma joined	05:20
	← __CA__ left	05:21
	→ bubbles10101 joined	05:22
	← bubbles10101 left	05:22
	← tshauck left	05:24
	← dhruvasagar left	05:26
	← adityamenon left	05:26
	→ adityamenon joined	05:26
	→ dhruvasagar joined	05:27
	← optixx left	05:28
	→ maleknet joined	05:29
	← neurodrone left	05:29
	→ optixx joined	05:30
	→ pdtpatrick joined	05:31
	→ HyperNerdV2 joined	05:32
	→ Dave^\| joined	05:33
	← _W_ left	05:34
	→ _W_ joined	05:34
	← lamont left	05:34
	→ goshawk joined	05:35
	→ lamont joined	05:35
	→ dr_win joined	05:39
	← ashirus left	05:40
	→ ashirus joined	05:40
	← fayimora left	05:40
	← ryez left	05:40
	← HyperNerdV2 left	05:41
	← guampa left	05:43
	← adityamenon left	05:45
	→ adityamenon joined	05:46
	← pheaver left	05:46
	→ bigkm joined	05:46
	→ Trodfoot joined	05:48
	→ pheaver joined	05:49
	← dhruvasagar left	05:51
	→ akiress joined	05:51
	← Trodfoot left	05:52
	← dr_win left	05:56
	→ fruers joined	05:57
fruers	I'm looking to setup Git in an enterprise setting. From my research I'm finding Gitolite to the most popular solution for this and I've hooked it into LDAP authentication and authorisation pretty easily	05:58
	I've got one question about the Author field	05:58
	← adityamenon left	05:58
fruers	As it is at the moment, the user sets up their own values locally (in ~/.gitconfig) and that is what appears in the Author field	05:58
	Is there some feature in Gitolite that can overwrite the Author field that the user has supplied?	05:59
	→ adityamenon joined	05:59
fruers	By default it's not the behaviour I'm seeing, but I'm assuming that this is a common issue to overcome	06:00
	But on the other hand, my rudimentary understanding of Git gives me the gut feel that this would be an expensive operation on the Gitolite side because it'll have to dive into each commit that is being pushed and change the Author field	06:01
	← Soliton left	06:02
	← kenperkins left	06:02
	← OOPMan left	06:03
	→ Robbo_ joined	06:03
Robbo_	is there a way I would be able to use github to install stuff on a server that already has content? For example...	06:04
	Say my structure has a library and js folder	06:04
	To install I need to put files/folders in each but there is also other stuff in the root	06:04
	However I want a way so I can just install from the CLI since I setup so many test/dev sites	06:05
	tommyvyo_ → tommyvyo	06:05
	← asakura left	06:08
	← edude03 left	06:08
	→ asakura joined	06:09
	→ dr_lepper joined	06:10
	← dr_lepper left	06:10
	→ dr_lepper joined	06:10
cirwim	Robbo_: see !website	06:11
gitinfo	Robbo_: [!deploy] Git is not a deployment tool, but you can build one around it for simple environments. Here is an example hook to get you started: http://goo.gl/l68hq	06:11
cirwim	grr, people keep changing my favourite bot answers: the website I wanted to link to was: http://toroid.org/ams/git-website-howto	06:12
	→ Soliton joined	06:12
	→ jgoss joined	06:12
	← jgoss left	06:12
	→ jgoss joined	06:12
Robbo_	Yeah I already do that sort of thing with hg. Was just thinking of a more general thing so I can quickly setup a dev/test site. What I might do is a have a repo with everything installed on it and sub modules and then whenever I want a new site I can do that and then maybe some hooks for config	06:14
cirwim	hmm	06:15
	I don't have enough information to help you further	06:15
Robbo_	Oh I don't really need more, was just thinking out loud	06:15
cirwim	:)	06:15
	everyone is happy then	06:15
Robbo_	Was just wondering if there was a hidden feature I could use heh	06:15
	→ adityamenon90 joined	06:19
	→ mishok13 joined	06:21
	← adityamenon left	06:22
	← fuzzybunny69y left	06:24
	← Fandekasp left	06:24
	→ nahhh joined	06:25
	← optixx left	06:29
	→ optixx joined	06:30
	→ glcrazy joined	06:33
	→ davidbanham joined	06:34
	→ chayin joined	06:37
	← chayin__ left	06:39
	→ gchristensen joined	06:39
	→ drizzd joined	06:41
	← DofLeFou left	06:43
	← Axsuul left	06:43
	→ fayimora joined	06:45
	→ fuzzybunny69y joined	06:45
	← drev1 left	06:46
	→ fayimora_ joined	06:47
sitaram	fruers: could you write up how you did the LDAP hookup and send it to me? Someone was asking for more details than the sparse amount of stuff I had in the docs and since I have no experience with that I could not help	06:48
	→ ehsan joined	06:48
	← optixx left	06:48
sitaram	fruers: as far as the author field is concerned, even if gitolite (or some server side program) could change it, it would change all the SHAs and create conflicts.	06:49
	← fayimora left	06:49
	fayimora_ → fayimora	06:49
fruers	sitaram: yeah, I assume you only want the authentication information?	06:50
	I'm using smart http so it's just apache configuration	06:50
	← kerrick left	06:50
	← BlindOracle left	06:51
	→ optixx joined	06:51
	← lonewulf`tp left	06:51
	→ BlindOracle joined	06:51
sitaram	fruers: oh for some stupid reason i thought LDAP-based pubkeys or whatever; nvm then	06:51
fruers	I think this represents a problem for Git in an enterprise setting as it essentially means that changes can't be tracked	06:51
	← igotnolegs left	06:52
	← efreese left	06:52
fruers	Or at least, it relies on potentially bogus information from developers	06:52
sitaram	fruers: see bottom of src/VREF/EMAIL-CHECK for why I disagree with you	06:52
	→ igotnolegs joined	06:52
	← igotnolegs left	06:52
fruers	ok	06:53
	→ igotnolegs joined	06:53
	← lonewulf`tb left	06:54
	aLeSD → Guest88029	06:56
	→ b1rkh0ff joined	06:58
	→ ToBeFree joined	06:59
	← meridion left	07:00
	→ adityamenon joined	07:01
	← adityamenon90 left	07:01
	→ ph^ joined	07:04
	← scientes left	07:05
	← adityamenon left	07:05
	→ tvw joined	07:07
	← dvide left	07:07
	← optixx left	07:09
	← glcrazy left	07:09
	→ glcrazy joined	07:10
	→ optixx joined	07:10
	→ ferdna joined	07:11
	→ whitman joined	07:15
sitaram	fruers: also understand that most people use gitolite's access control to enforce "who can push to master", and that is always a tested+QA/reviewed series of commits. In some places only the QA guys can push versioned tags (like v1, v2, v3.5, etc), and the presumption is that if they did, they have tested it and are "signing off" on it	07:15
	← g00s left	07:16
	← dr_lepper left	07:18
	← Lgb left	07:20
	→ g00s joined	07:20
	→ raatiniemi joined	07:21
fruers	sitaram: I'm in agreement with your comments in in the email check, in that it undermines the fundamentals of DVCS	07:21
	→ Lgb joined	07:22
fruers	I think that my workplace would be satisfied with the enterprise ID that pushed a changeset	07:22
	← h0bbit left	07:22
	→ bubbles10101 joined	07:22
sitaram	which gitolite logs for you	07:23
	← ph^ left	07:24
fruers	I guess it also depends on the visibility of the author of the push versus the visibility of the author of the commit	07:24
	← pdtpatrick left	07:25
fruers	If bogus names are appearing everywhere -- which they will with 100+ developers -- then it's a concern	07:25
	← nutron left	07:26
	← gavin_huang left	07:27
fruers	Hmm... the author email is what I'm actually seeing in `git log`, so the visibility is high	07:27
	→ thiago joined	07:28
	→ lonewulf`tp joined	07:28
	← BrainFreak__ left	07:29
fruers	I could enhance the email-check vref to accept any valid username/email address which would bring back some of the DVCS functionality (like offline merges)	07:29
	→ DuzyZiemniak joined	07:29
	→ gavin_huang joined	07:30
sitaram	fruers: there's an author id and a committer id. Both of these are stored in the actual repo (and affect the SHA for example). The person who pushed could be someone quite different, although usually it will be the committer. The ID of who pushed a particular commit series is in the gitolite log files. They are plain text files to if this is a concern protect them -- in any case no one should have access to the server (ideally)	07:30
fruers	What would my last proposal mean for the "pusher"? Say, their push were rejected because a commit they merged had a bogus author info... could they easily fix those commits before pushing?	07:33
	Just looking up difference between author and committer	07:33
	→ Gentlecat joined	07:34
	← Gentlecat left	07:35
	→ Gentlecat joined	07:36
	← gchristensen left	07:36
	→ danielpunt joined	07:39
	→ ph^ joined	07:46
	→ Davey joined	07:46
	→ Codif joined	07:46
	→ adityamenon joined	07:47
	← bolt left	07:47
thiago	fruers: it could be easy or it could be difficult	07:48
	it's hard to say generically	07:48
	if you have a development model like the kernel where there is a lot of merging going on, the "bogus" commit may be a long way down the line	07:48
	fixing it wouldn't be easy	07:48
	→ lg188 joined	07:49
	← danielpunt left	07:49
lg188	When I try to fetch from Bitbucket.org I get this error http://pastebin.mozilla.org/1650218	07:49
sitaram	anyone can change anything and push. Identity is not established by what the author and committer fields say. That is why you have signed tags for specific commits (which then implicitly validate all commits below it), or in the corporate world you let the ssh/http userid be the "authenticator"	07:49
thiago	lg188: are you sure you got the URL right?	07:49
	← optixx left	07:49
lg188	thiago: yes i'm right	07:50
	thiago: sure*	07:50
thiago	lg188: then it's a server-side problem	07:50
	lg188: prefer the git:// URL instead	07:50
lg188	thiago: i'll try to use it	07:50
	thiago: ssh right?	07:50
fruers	I'm doing quite a bit of reading and I'm agreeing with this in a philosophical sense, but I'm wondering how to overcome the negatives	07:50
	→ optixx joined	07:50
thiago	lg188: no, git://	07:51
fruers	Primarily, developers will want to look at a change and be able to identify who made it	07:51
sitaram	fruers: most people eventually realise that if you care too much about author/committer info and not about the code they got bigger problems :)	07:51
	← _W_ left	07:51
sitaram	fruers: yes and that is the author ID	07:51
	→ _W_ joined	07:51
thiago	fruers: that's the author field, usually	07:52
sitaram	fruers: if you have an author who is consistently using wrong emails he needs to be kicked out. If you cant trust them to get that right how can you accept code from them?	07:52
thiago	fruers: that's why Git keeps it separate from the committer: so the original author, who usually owns the copyright anyway, is tracked.	07:52
lg188	well anyway how can I reset my local source to my latest commit? like forget all my changes?	07:52
	← adityamenon left	07:53
sitaram	thiago: he's talking about people who use a bous email ID as the author (IIUC)	07:53
	bogus*	07:53
thiago	if your problem is that you have people who are submitting using bogus, other author addresses, you should kick them out too	07:53
sitaram	nod vigorously	07:53
thiago	e.g., my commits have Author: Sitaram Chamarty so that sitaram gets the blame, not I :-)	07:53
	then you should kick me	07:53
	there are two ways of making sure that doesn't happen, from what I can tell	07:54
fruers	I agree, it would be great if everyone didn't suck ;)	07:54
thiago	in a distributed environment, like the kernel, use commit tags: the tags sign the history up until that point (usually, up until another tag) and it's unambiguous who made it	07:54
	← adamm left	07:55
fruers	But providing DVCS for a large enterprise development team with high turnover I can tell you that they stuff up simpler stuff than this	07:55
thiago	then you say "I take responsibility for these commits, whether I made them or I applied patches by these other people"	07:55
sitaram	fruers: enforce the author ID for junior developers (who should be pushing only their own stuff), and make the senior developers accountable for entire commit series not just their commits	07:55
thiago	the other way I see is when you have a central repository, common in corporate environments	07:55
	for those, just make a hook enforce that one can only push commits with their own names as Author	07:56
sitaram	fruers: I have several large teams using git in $DAYJOB (one particularly large one too) and they don't have this issue.	07:56
fruers	thiago: yeah, this is what sitaram is saying gitolite provides, which is the ultimate auditing, but it's the actual visibility that is the problem. If people have to jump through hoops to find the real committer/pusher/whatever then they're going to freak	07:57
sitaram	thiago: gitolite has such code you can enable	07:57
thiago	fruers: the point is that people shouldn't have to verify, if they trust the system	07:57
fruers	It could be 1 or all 100 that are juniors	07:57
thiago	fruers: with a central repository, you _have_ to trust the system	07:57
et	it's kind of an issue in university context. some smartass students will inevitably think "i can put anything as committer? cool, let	07:57
	← glcrazy left	07:57
et	+'s do that"	07:57
sitaram	fruers: you're overestimating the problem. No one does this accidentally -- it will be intentional and it will get caught pretty quick	07:58
	→ glcrazy joined	07:58
thiago	well, no one does it accidentally for long	07:58
fruers	Yeah, I agree it's mostly done by accident	07:58
thiago	each new installation's first few commits might have a bogus author and email address	07:58
	but it's usually caught quite quickly and, moreover, developers usually want the credit	07:59
fruers	heh	07:59
thiago	e.g., I can produce nice graphs and tell my manager, "look, I'm the top contributor this month again!"	07:59
fruers	I'm thinking that some kind of out-of-band nightly job that trawled commit authors could solve the problem	07:59
thiago	http://www.macieira.org/~thiago/qt-stats/current/qt-all.author.absolute.png :-)	07:59
et	also, someone might have rsync'ed a repository instead of git cloning it and then accidently commit as the one that he copied from	07:59
fruers	If it's just a few commits and you can always go back to the gitolite logs to get the pusher, then that's fine	08:00
et	(if the original one had local name/email settings, as i have for most repos)	08:00
fruers	It really comes down to having to accept this if you want to use DVCS	08:00
	← davidbanham left	08:00
thiago	fruers: you don't need to do it nightly. Doing it that way requires matching the push logs to the commits.	08:00
	fruers: if you do it in a hook, the information is there, at the moment of the push.	08:01
	→ pantsman joined	08:01
	← pantsman left	08:01
	→ pantsman joined	08:01
thiago	and gitolite has that setting	08:01
fruers	So let it pass, but fire off an email to the pusher?	08:01
thiago	I'd say, no. Block on pushing.	08:01
	then the dev who has the accidental mistake will simply fix it.	08:01
	it's usually a new installation and won't be more than a few commits.	08:01
	→ davidbanham joined	08:02
fruers	Ah I see, that would cover 99% of cases. The other 1% of a developer offline merging or cherry picking commits locally would be a minor edge case anyway	08:02
thiago	for merging, it might be that the hook already does the right thing, if the merged commits are present elsewhere in the repository (a shared branch, for example)	08:03
fruers	So if they have made 10 commits locally, is it easy enough for them to change the author on those 10 commits?	08:03
thiago	it's not trivial, but it's doable	08:03
	and you'll probably have a wiki page somewhere telling them what to do	08:03
	→ TheJH joined	08:03
	→ Guest1598791926 joined	08:03
fruers	Ok, I'm convinced, but this will be the first thing in the getting started guide!	08:04
	← davidbanham left	08:04
sitaram	fruers: I have to go now but bottom line: use the VREF I sent you for junior developers, and as they gain seniority remove them from that list. Although IME/IMO all this is overkill and even in teams of 800+ people working on large products with multiple modules I have never heard this complaint from my users (at $DAYJOB)	08:05
fruers	sitaram: Thanks for the help and great work with Gitolite, it was much easier to stand up than I was expecting	08:06
	→ blacktulip joined	08:08
	← mishok13 left	08:09
	← tvw left	08:11
	→ tvw joined	08:11
	→ intransit joined	08:13
	← optixx left	08:13
	← kadoban left	08:15
	→ optixx joined	08:15
	← ferdna left	08:15
	← g00s left	08:17
	← intransit left	08:17
	← ehsan left	08:17
	← tvw left	08:17
	Guest42683 → chrisward	08:19
	← bigkm left	08:20
	← NimeshNeema left	08:25
	→ notmypudding joined	08:27
	← luke-jr left	08:29
	→ luke-jr joined	08:29
	← optixx left	08:30
	← ipoval left	08:30
	← lonewulf`tp left	08:31
	→ optixx joined	08:32
	→ fayimora_ joined	08:34
	← pheaver left	08:36
	← fayimora left	08:37
	fayimora_ → fayimora	08:37
	→ PapaSierra joined	08:38
	→ Eax joined	08:43
Eax	Hey there Gentlefolks - I have tried creating my own remote repo, and it's actually gone quite okay. BUT! The files in the remote repo aren't changed. I can see in git log that the commit/push has gone through, but no files have changed - Have anything heard of this before?	08:44
milki	Eax: !bare	08:46
gitinfo	Eax: an explanation of bare and non-bare repositories (and why pushing to a non-bare one causes problems) can be found here: http://bare-vs-nonbare.gitrecipes.de/	08:46
milki	most likely your remote is a bare repo	08:46
Eax	It is! So it shouldn't be bare when it's a remote repo?	08:46
milki	no, it should be bare	08:46
	are you trying to use git to !deploy a website?	08:47
gitinfo	Git is not a deployment tool, but you can build one around it for simple environments. Here is an example hook to get you started: http://goo.gl/l68hq	08:47
	← ph^ left	08:47
Eax	milki: Heh, yeah, I am :)	08:47
milki	then check out that link above	08:47
	→ sq-one joined	08:47
Eax	milki: Great, many thanks :) Have a swell saturday	08:48
milki	you too	08:48
Eax	Thanks :)	08:48
	← Eax left	08:48
	← glcrazy left	08:51
	→ SkAzZ joined	08:52
	→ lkong joined	08:52
	→ Quanteek joined	08:54
	→ SkAzZ1 joined	08:55
	→ dailylinux joined	08:55
	→ lonewulf`tp joined	08:55
dailylinux	hi, i have a git URL: git://gitorious.org/owncloud/apps.git, and wonder if there is a way to git pull just an application from there without getting all?	08:56
	← SkAzZ left	08:56
	← AlbireoX`Laptop left	08:58
	← goshawk left	09:00
	→ ixti joined	09:01
wereHamster	dailylinux: no	09:01
dailylinux	hmm, ok	09:01
milki	well, not with the git url	09:01
	go to the source tree via browser	09:02
	and get the ddl link for it	09:02
	→ d0k joined	09:03
	→ khmarbaise joined	09:04
	→ w09x joined	09:04
	→ _iron joined	09:05
dailylinux	milki, ok. like http://gitorious.org/owncloud/apps/trees/master/files_odfviewer ?	09:07
	that's the app i need	09:07
	← zell left	09:07
	← khmarbaise left	09:07
	→ h0bbit joined	09:08
dailylinux	just thinking if i can add this url in my .git/config file and get updates :)	09:09
wereHamster	dailylinux: no	09:09
	git doesn't work that way	09:09
dailylinux	ok	09:10
	i can't even add in that url in my config file and do something like 'git pull origin files_odfviewer' ?	09:10
	→ Shusshu joined	09:11
	→ Jackneill joined	09:11
	← Jackneill left	09:11
	→ Jackneill joined	09:11
	← Codif left	09:11
dailylinux	ah, i can't	09:12
	this is not a .git url	09:12
wereHamster	no	09:12
	git doesn't work that way	09:13
	→ ryez joined	09:13
dailylinux	any idea how i can achieve this?	09:14
	can i 'merge' this app into my git config?	09:14
	maybe?	09:14
	← Sahua left	09:14
wereHamster	just clone the whole repo	09:14
dailylinux	ok, then?	09:15
wereHamster	then cd files_odfview; then do whatever you need to do	09:15
dailylinux	ok	09:15
	i'll try that	09:15
	thanks a lot! wereHamster	09:15
wereHamster	was that 'solution' not obvious?!?	09:16
	→ adityamenon joined	09:16
dailylinux	yes it was	09:16
	but i was looking for a more "correct" one	09:16
	since i already a git config and wanted to get all updates together when i do git pull	09:17
ryez	hmm, cherry-pick doesn't have a `-N` option for not even stage the picked change	09:18
milki	cherry-pick and soft reset?	09:19
ryez	milki, yes, looks that's way to go	09:20
	I tried `git cherry-pick -n`, but still have to soft reset	09:20
	to unstage it	09:21
	← adityamenon left	09:22
ryez	what about proposing such option?	09:23
milki	good luck?	09:24
	feel it out on the mailing list if you want	09:24
	seeing as its only two commands, i dont know if people will care enough	09:24
ryez	yes, we need luck, anyway, I can try	09:26
	→ Mkaysi joined	09:26
ryez	this way cherry-pick would only be faster with `-N`	09:26
	← blacktulip left	09:28
ryez	will pull request from github be accepted?	09:28
	→ Dimeetrees joined	09:29
	← ill1cit left	09:29
	← igotnolegs left	09:31
	← Shusshu left	09:31
_br_	Does anyone here have an opinion about git-annex? I mean in general for usage or you suggest an alternative?	09:31
ryez	ah, git is already on github, I remember it was on kernel.org in the past, I'm out	09:32
	→ martinjlowm joined	09:36
	→ BarryCrayon joined	09:37
	← optixx left	09:39
	→ optixx joined	09:40
	← SkAzZ1 left	09:41
	→ ivan12348 joined	09:41
	← BarryCrayon left	09:41
	→ ivan12349 joined	09:41
	→ ivan12350 joined	09:41
	← ivan12345 left	09:44
	← ivan12347 left	09:44
	→ bigkm joined	09:45
	← ivan12346 left	09:45
	← TheJH left	09:48
	← Davey left	09:49
	→ RaceCondition joined	09:49
	→ barrbrain_ joined	09:51
	→ balau joined	09:51
	← cirwim left	09:52
	→ enr joined	09:55
	← drizzd left	09:57
	← Quanteek left	09:57
	gitinfo set mode: +v	10:00
lkong	hi, where do you see the message identifier for an email send with git-send-email in order to use it in --in-reply-to ?	10:01
	is there an option that gives you an history of the sent emails ?	10:02
	→ nioncode joined	10:06
	→ bolt joined	10:06
	→ mishok13 joined	10:08
	→ Davey joined	10:10
	← bolt left	10:11
	→ bolt joined	10:11
	← Davey left	10:12
	→ Element9 joined	10:14
	→ intransit joined	10:15
	← dailylinux left	10:17
	← nahhh left	10:18
	← intransit left	10:19
	← sq-one left	10:20
	→ gemmes joined	10:20
	← gemmes left	10:21
	→ bjensen joined	10:22
	→ airborn joined	10:23
bjensen	can anyone point me to a resource that explains the most safe way to merge branches? I don't want any auto merging of files this time..	10:23
	← dreiss left	10:23
	← RaceCondition left	10:24
	→ twisted` joined	10:24
	← pantsman left	10:25
	→ matthiaskrgr joined	10:25
	→ Sky[x] joined	10:26
	← martinjlowm left	10:27
	← twisted` left	10:28
	← Bigcheese left	10:29
	← Y_Ichiro left	10:29
	→ ph^ joined	10:30
	← nioncode left	10:30
	→ nioncode joined	10:30
	← jspiros left	10:33
milki	so you want to discard all changes that touch existing files in the destination branch?	10:34
	thats...odd	10:34
	→ jspiros joined	10:34
milki	one way you could do that is add another commit to the branch that simply deletes files that exist in the destination branch...	10:35
	→ Tinti_ joined	10:36
	← nioncode left	10:36
	→ nioncode joined	10:36
	← matthiaskrgr left	10:38
	← jspiros left	10:40
	→ silv3r_m00n joined	10:40
	→ jpieper joined	10:40
silv3r_m00n	if I go a git fetch , and then I don't want the changes that came in , can I use git reset --hard HEAD to simply unfetch ?	10:40
canton7	silv3r_m00n, 'git fetch' doesn't touch any of your local branches	10:40
silv3r_m00n	what does it do then ?	10:41
canton7	it updates the relevant remote-tracking branches (the ones that start with e.g. 'origin/'), which act as a record of the state of the remote branches	10:41
wereHamster	ryez: no, pull requests on github will not be accepted	10:41
	ryez: see the repo description on github...	10:42
silv3r_m00n	canton7: I did a git fetch and it showed 0d27780..42b9cec master -> origin/master now how to do I view what came in ?	10:42
canton7	silv3r_m00n, git log origin/master	10:42
silv3r_m00n	and by the way when I do git branch , it shows * and master but nothing called origin/master	10:42
canton7	that line says that the local branch origin/master was updated to contain what the remote master branch contains	10:43
	silv3r_m00n, git branch -a	10:43
silv3r_m00n	what does that do ?	10:43
wereHamster	silv3r_m00n: man git-branch	10:43
gitinfo	silv3r_m00n: the git-branch manpage is available at http://git-scm.com/docs/git-branch.html	10:43
	← silv3r_m00n left	10:43
canton7	lists all branches, including remote ones	10:43
	→ silv3r_m00n joined	10:44
canton7	in case you missed that, git branch -a lists all branches, including remote-tracking ones	10:44
	← hwrdprkns left	10:44
wereHamster	in case you missed that, the man page describes each and every single option that git branch supports	10:45
	→ hwrdprkns joined	10:45
silv3r_m00n	so once I fetched , I should do git checkout origin/master to check the incoming things ? and how do I compare the differences between my local master and origin/master ?	10:45
wereHamster	git diff master origin/master	10:46
canton7	no. never checkout a remote-tracking branch (e.g. origin/master)	10:46
	(unless you know about detached heads)	10:46
silv3r_m00n	canton7: then ?	10:46
	I just did git checkout origin/master	10:46
canton7	oh dear, if it's going to be one of these cases, I'm not going to get involved. Listen to wereHamster and read man git-log and man git-branch	10:47
gitinfo	the git-log manpage is available at http://git-scm.com/docs/git-log.html	10:47
	the git-branch manpage is available at http://git-scm.com/docs/git-branch.html	10:47
silv3r_m00n	canton7: fatal: ambiguous argument 'master': both revision and filename	10:47
	Use '--' to separate filenames from revisions	10:47
	I get that when I do a git diff master origin/master	10:47
wereHamster	do as git tells you	10:47
	→ Yuuhi joined	10:49
	→ ashirus_ joined	10:51
	← ashirus left	10:53
	→ jspiros joined	10:54
	← silv3r_m00n left	10:54
	→ arietis joined	10:54
	← bigkm left	10:55
	→ bigkm joined	10:56
	→ adityamenon joined	10:56
	→ JStoker joined	10:56
	← preflex left	10:57
	→ preflex_ joined	10:58
	preflex_ → preflex	10:58
	→ lolfrenz joined	10:59
	→ stephanmg joined	11:01
	→ RaceCondition joined	11:02
	→ umc joined	11:02
	← DuzyZiemniak left	11:03
	← enr left	11:04
	→ enr joined	11:04
	→ Raging_Hog joined	11:06
	→ Resure joined	11:06
	→ mneorr joined	11:10
bob2	and try not to make a file called 'master'	11:10
	→ flavius joined	11:14
	← ryez left	11:15
	→ project2501a joined	11:16
	→ txomon\|home joined	11:17
	→ paul424 joined	11:19
	← fuzzybunny69y left	11:19
paul424	git branch --track development refs/remote/origin/development does not work	11:20
	what's wrong ?	11:20
	← raatiniemi left	11:20
cmn	remotes is misspelt	11:21
	→ ns5 joined	11:21
ns5	how to find out the origin repo location/url?	11:22
	← Tinti_ left	11:22
	→ Tinti_ joined	11:22
cmn	git remote -v	11:23
ns5	origin [email@hidden.address] (fetch)	11:23
	origin [email@hidden.address] (push)	11:23
	← adityamenon left	11:23
ns5	how do I modify the ip address?	11:23
	the git server's IP has changed	11:23
cmn	man git remote	11:24
gitinfo	the git-remote manpage is available at http://git-scm.com/docs/git-remote.html	11:24
cmn	and using DNS is also a good idea	11:24
bob2	maybe look at creating some sort of system to map permanent names to ip addresses somehow	11:24
cmn	you should make a startup out of that idea, sounds useful	11:25
sitaram	ns5: you could just edit .git/config	11:30
	carefully... :)	11:30
ns5	sitaram: thanks, and thanks for creating gitolite	11:31
	← harish left	11:31
sitaram	you're welcome!	11:32
	← flavius left	11:35
	← bigkm left	11:36
	← jspiros left	11:36
	→ TheJH joined	11:37
	→ felipec joined	11:38
	→ raatiniemi joined	11:38
	← Sigma left	11:38
felipec	how can I find the root commits of a list of revisions? master..topic (suppose there's 3 commits whose parent is the merge base of master and topic)	11:39
cmn	what do you mean by root commits here?	11:40
	→ madsy joined	11:41
	← madsy left	11:41
	→ madsy joined	11:41
	← stephanmg left	11:43
felipec	cmn: the last ones in the graph	11:44
	← RaceCondition left	11:45
felipec	cmn: like, if you just do 'git log' you will see all the commits until there's a commit without parent (root)	11:45
	and there might be more than one root	11:45
	→ RaceCondition joined	11:46
cmn	yeah, but that doesn't have much to do with what you get when you say master..topic	11:46
	→ nahhh joined	11:46
felipec	cmn: how is it any different? it's still a graph, which can have more than one root	11:46
cmn	because the commits themselves do have paretns	11:47
	though you may be able to get an empty %P if the expansion is done after the simplification	11:47
felipec	cmn: but they are not part of the committish query	11:47
cmn	which is why you may get empty %p or %P after simplication	11:48
	but if you look at the actual commit, you won't be able to tell it's a root commit	11:48
	→ simesy joined	11:49
felipec	cmn: why would %p change after simplification?	11:49
	→ dinduks_ joined	11:49
felipec	(it doesn't)	11:50
cmn	because the commit's metadata are temporarily rewritten so they fit in the graph you asked for	11:51
	so it's possible that the parent commits would change	11:51
felipec	cmn: but that's not what --format prints; it prints the real metadata	11:52
	← Lord_DeathMatch left	11:52
cmn	then you'll have to add your negative commits to a list of hashes you consider to make a commit parentless	11:52
	maybe the simplication only happens at that scale when you filter by paths... I have a vague idea that it should work	11:53
	→ ryez joined	11:54
	← chrisward left	11:54
	→ jspiros joined	11:56
	← sirdancealot2 left	11:56
	→ AeroNotix joined	11:56
	→ chrisward joined	11:56
	→ sirdancealot2 joined	11:57
	chrisward → Guest23058	11:57
AeroNotix	I'm trying to get apache+git working. I've used the config which is on the archwiki for Git. I've mv'd some local copies of a git repo into my $GIT_SRV directory, and then tried to pull/clone using the url but it fails with "request not supported: $GIT_SRV_DIR" in the apache log	11:58
felipec	I guess the only way is to log with --boundary, track each commit's parent, and find out if that parent is a boundary	11:59
thiago	AeroNotix: first of all, reconsider and use ssh	12:00
	← TheJH left	12:01
	← jspiros left	12:01
AeroNotix	thiago: why?	12:02
thiago	AeroNotix: because it's much easier to use, more secure and easier to make work	12:02
	git was designed to work with ssh	12:02
	→ Chillance joined	12:02
	→ gusnan joined	12:02
AeroNotix	thiago: Do you have some tutorials/resources that you could link to ?	12:02
thiago	nothing except the git basics	12:03
AeroNotix	right	12:03
	← RaceCondition left	12:03
	← setmeaway left	12:04
	→ RaceCondition joined	12:05
	← lamont left	12:09
bob2	and gitolite if you want people writing to it	12:09
	← felipec left	12:09
	← optixx left	12:09
	← gavin_huang left	12:10
	→ lamont joined	12:10
	← h0bbit left	12:10
AeroNotix	thiago: well, that took all of 5 seconds to do	12:11
	→ optixx joined	12:11
AeroNotix	thiago: thanks	12:11
	thiago: Now, using this set up. I.e. a directory which contains git --bare init'd folders. Is there a web-frontend available for it?	12:12
	I'm trying to move my workplace onto using version control	12:12
thiago	gitweb	12:13
AeroNotix	thiago: thanks	12:13
	→ erichynds joined	12:13
AeroNotix	ahmahgahd this is pretty ql	12:15
	I think I'll finally be able to start using git at work	12:15
	→ xjiujiu joined	12:15
	→ enmand joined	12:16
	→ jspiros joined	12:17
	← RaceCondition left	12:19
	→ neurodrone joined	12:19
	← enmand left	12:19
	← enr left	12:20
	→ enr joined	12:20
	← erichynds left	12:20
	← Milossh left	12:21
	← jspiros left	12:21
	← lamont left	12:22
	→ priori99 joined	12:22
	→ jspiros joined	12:22
	← maleknet left	12:22
	→ RaceCondition joined	12:23
	→ felipec joined	12:24
	→ moltar_net1 joined	12:25
	gitinfo set mode: +v	12:25
moltar_net1	Hi! I'm getting the "To prevent you from losing history, non-fast-forward updates were rejected" error, but when I do a pull, it creates an empty merge commit. Why such a thing?	12:25
nahhh	Is it possible to add directories which has git repositories within git repository _without_ submodles? For some reason that directory which is subrepository is empty...	12:26
	→ kerrick joined	12:27
	← spyro left	12:27
	→ DoniRS joined	12:27
	→ lamont joined	12:28
bob2	moltar_net1, the content is not the important thing	12:28
	← lamont left	12:28
	→ lamont joined	12:28
bob2	nahhh, yes but git will ignore them	12:28
nahhh	Why?	12:28
moltar_net1	bob2: yeah the ref I guess is the important thing, but I still can't wrap my head around why this has happened. I am the only one using it.	12:29
bob2	why what?	12:29
	← RaceCondition left	12:29
bob2	moltar_net1, well, you can see why it happened	12:29
	moltar_net1, 'git fetch --all ; git log master ; git log origin/master'	12:29
moltar_net1	However "git pull --rebase" solved the problem :)	12:29
nahhh	bob2, take a look at if you can: http://pastebin.com/H5wV1URS	12:29
bob2	moltar_net1, the second will show a commit (or commits) you didn't have in the first log	12:29
	moltar_net1, don't blindly do that, you need to know why extraneous commits appeared	12:29
nahhh	"vendor/doctrine" is a subrepository, but it's not managed by git submodules.	12:30
	→ spyro joined	12:30
	← spyro left	12:30
	→ spyro joined	12:30
	→ Lord_DeathMatch joined	12:30
	→ enmand joined	12:31
moltar_net1	hm I don't even see any differences :/	12:31
bob2	git log master..origin/master	12:32
	← enmand left	12:32
bob2	git log origin/master..master	12:32
	note that this isn't of any use now if you if blindly did a rebase	12:32
moltar_net1	oh I see yeah, I guess rebase killed all the differences, eh?	12:32
	I wonder if I lost anything :\|	12:32
	now I am paranoid :\|	12:32
	ha I do however have another copy on production server, but its not entirely up to date.	12:33
	← priori99 left	12:33
	← jspiros left	12:34
	← fruers left	12:35
	→ herock joined	12:37
	← herock left	12:37
	→ trying2code joined	12:37
	→ zinjacoder joined	12:37
trying2code	I copied pasted new folders and git doesnt recognize them	12:38
	nor does it do git add .	12:38
	how do i track all the files/folders from beginning?	12:38
bob2	need to be more specific	12:38
	since 'git add .' will add them	12:38
trying2code	git status says, on add portion, add ./	12:39
	i do that, git add .	12:39
	and git push	12:39
	it only pushes 3 files	12:39
	not anything else	12:39
bob2	erm	12:39
cbreak	git add . will only add the current folder	12:39
bob2	stop and think for a sec	12:39
	since you didn't commit anything	12:39
cbreak	and git add . has NOTHING to do with pushing	12:39
trying2code	(forgot to mention, i did commit after adding)	12:39
cbreak	also, git doesn't track folders directly, only files	12:39
	if you committed, do a git log -p	12:40
	are the files you expect there?	12:40
	if not then you failed to add them	12:40
	→ kenperkins joined	12:40
	→ klj613 joined	12:40
trying2code	the 3 files are there,	12:40
	so how do i add the folders to my repo	12:41
cbreak	not at all	12:41
	git only tracks files	12:41
	put files into them and add the files	12:41
trying2code	got it	12:41
	thank you	12:41
cbreak	they can be empty	12:41
	(the files)	12:42
	← ToBeFree left	12:43
	← flaguy48 left	12:46
trying2code	if I care about the folder structure, just as being able to recover folders (non empty)	12:46
	should i be using svn and not git?	12:46
	→ martinjlowm joined	12:46
	→ khmarbaise joined	12:48
bob2	depends	12:48
	← ns5 left	12:49
	→ Vampire0_ joined	12:49
	→ ns5 joined	12:49
trying2code	If I store everything in remote directory, I guess git pull should resolve that	12:50
bob2	no	12:50
trying2code	even if I have not made new pushes?	12:50
	← neurodrone left	12:51
	→ dfr\|mac joined	12:51
	→ adityamenon joined	12:51
	← Vampire0 left	12:53
	← raatiniemi left	12:53
	→ jspiros joined	12:54
Element9	in which situation can the HEAD get detached on cheching out the branch?	12:54
	← Vampire0_ left	12:54
	→ Vampire0 joined	12:54
kpreid	Element9: Precisely when you specify something other than a branch name.	12:54
thiago	Element9: when you check out something that isn't a branch	12:55
Element9	but it is a branch. a remote branch	12:56
	← Lord_DeathMatch left	12:56
thiago	then it's not a branch	12:56
	only local branches are branches	12:56
	← sirdancealot left	12:57
Element9	thiago: aha got it. thanks	12:58
AeroNotix	hmm, I can't get gitweb working. I have apache serving the pages, that works. Except it keeps saying '404 no projects found' the directory is chmod 777 so it's (hopefully) not a permission issue. Any ideas?	12:58
bremner	other than doublechecking the paths in gitweb.conf, not really.	12:58
AeroNotix	bremner: the path is the same path I'm using to ssh and push/pull from with that	12:59
bremner	and an absolute path?	12:59
AeroNotix	bremner: yup	12:59
	$projectroot = "/opt/gitrepos";	13:00
bremner	are you using the "scan the directory" option, or an explicit project file?	13:00
AeroNotix	bremner: neither?	13:00
	bremner: is this in the gitweb.conf?	13:00
	← bjensen left	13:00
bremner	AeroNotix: yeah, there should be a comment	13:00
AeroNotix	hmm	13:01
	2 sec	13:01
	bremner: what do you mean "should be a comment" ?	13:02
	← airborn left	13:02
bremner	in gitweb.conf, explaining which is the default.	13:02
AeroNotix	bremner: I didn't have a default gitweb.conf	13:03
	bremner: I had to create it	13:03
	→ NimeshNeema joined	13:03
bremner	$projects_list = $projectroot;	13:03
AeroNotix	bremner: I'll try :)_	13:04
	bremner: no. How can I make sure that the gitweb.conf is being read?	13:04
	it doesn't make any logs :/	13:05
bremner	dunno. You might have to look at the gitweb source to turn on debugging.	13:05
AeroNotix	aha	13:05
	→ jackh joined	13:06
	→ davidbanham joined	13:07
AeroNotix	:(	13:08
	→ ronniy joined	13:09
AeroNotix	omfg	13:10
	OMFG	13:11
	I had created a file named git-web.conf instead of gitweb.conf	13:11
	Thanks all for putting up with that stupidity	13:11
	← paul424 left	13:11
	→ Sahua joined	13:12
felipec	how can I find out the commits in certain branch that are not in any other branches?	13:12
thiago	git log thatbranch ^otherbranch1 ^otherbranch2 ^otherbranch3	13:14
	be careful if ^ is special in your shell	13:14
felipec	thiago: thats not what I want	13:14
thiago	felipec: that's what you asked	13:15
	← spyro left	13:15
thiago	felipec: if it's not what you wanted, please rephrase differently	13:15
	→ niki joined	13:15
	→ TheJH joined	13:16
felipec	thiago: I'm not going to list 100 branches	13:17
	git show-branch --independent	13:17
	→ pantsman joined	13:17
thiago	felipec: then write a script	13:18
	→ spyro joined	13:18
	→ royale1223 joined	13:18
thiago	felipec: maybe what you want can be done with git branch --merged or --contains	13:18
bremner	felipec: if you want to catch cherry picking, you'll need to use git patch-id	13:19
felipec	thiago: why? the command I'm looking for is 'git show-branch --independent' <- that is what I wanted	13:19
	← dfr\|mac left	13:19
bremner	and a brute foce loop.	13:19
	→ enr^ joined	13:19
bob2	that seems pretty niche	13:19
	← Lgb left	13:20
bob2	since it just tells you which branches aren't prefixes of the others	13:20
thiago	felipec: well, that's not exactly what you asked for originally, but ok	13:20
	← w09x left	13:20
	← enr left	13:21
	→ Lgb joined	13:22
felipec	thiago: that is exactly what I asked: --independent: Among the <reference>s given, display only the ones that cannot be reached from any other <reference>.	13:22
	← ruli left	13:22
felipec	thiago: replace reference with branch and it's exactly what I asked	13:22
m1sc	felipec: you asked for commits, no.?	13:22
	Guest23058 → chrisward	13:22
	→ ruli joined	13:23
thiago	felipec: you asked "how can I find out the commits in certain branch that are not in any other branches?"	13:23
	felipec: there are multiple ways to do that	13:23
	→ flaguy48 joined	13:23
	← Gentlecat left	13:24
	← nahhh left	13:25
	← lkong left	13:26
	→ ehsan joined	13:27
	→ Lord_DeathMatch joined	13:27
	← Resure left	13:29
felipec	thiago: of course, I could fire gitk and find out, but obviously I would want the simplest	13:29
	m1sc: it shows the commits	13:30
thiago	felipec: most people have a handful of branches, so my first answer would be ok	13:30
	felipec: git show-branch isn't a common command	13:30
	→ RaceCondition joined	13:30
	→ madewokherd joined	13:30
	→ adityamenon90 joined	13:31
felipec	I didn't ask for a common command	13:31
thiago	I'm not going to discuss with you	13:32
	→ vekexasia joined	13:32
vekexasia	hello all	13:33
	I've followed the git subtree	13:33
	guide	13:33
	← fisted left	13:33
vekexasia	and I've 2 subtrees	13:33
	for 2 different external repos	13:33
	the last one I setted up	13:33
	← adityamenon left	13:33
vekexasia	doesnt merge in the correct directory	13:34
	Can anyone help?	13:34
	→ bitkiller joined	13:34
	→ Milossh joined	13:34
EugeneKay	I've found that Johnny Walker makes it a lot easier	13:34
	→ fisted joined	13:34
vekexasia	EugeneKay: are you talking to me?	13:35
EugeneKay	Sure	13:35
	→ Bigcheese joined	13:35
vekexasia	EugeneKay: hmm can you explain it better ?	13:36
EugeneKay	Whiskey. Ethanol. Intoxicating bev-e-rages.	13:36
vekexasia	lol?	13:37
	← bubbles10101 left	13:37
EugeneKay	http://eugenekay.com/img/codingdrunk.png	13:37
vekexasia	Lol eugene i would love to talk about drunkness with you	13:37
	but I'd really like to figure out how to solve my issue :)	13:37
	and i'll buy the book :P	13:38
PapaSierra	by N. E. Briated?!	13:38
	→ herdingcat joined	13:38
EugeneKay	My experience is that with anything sub-(modules, trees, marines), it's best to start with the bottle.	13:38
vekexasia	Already tried :P	13:38
	thiago goes for the ice tea then	13:38
	EugeneKay spikes thiago's with gin	13:38
vekexasia	got drunk threw up and here again	13:38
	:P	13:38
	→ matthiaskrgr joined	13:39
EugeneKay	Sadly, however, I must get dressed and go work the fish fry at my VFD.	13:39
vekexasia	:(	13:39
	can anyone else help me with git subtree?	13:39
	→ tshauck joined	13:41
	→ brngardner joined	13:41
	← Tinti_ left	13:44
	→ JoeDevC joined	13:45
	← xjiujiu left	13:45
	← JoeDevC left	13:45
	→ xjiujiu joined	13:46
	← royale1223 left	13:46
PapaSierra	when working with a cms you're managing files and data. i've got my files in git, but what about the data? what are the cool kids doing/	13:46
thiago	files contain data	13:47
PapaSierra	sorry, rephrase: "database contents"	13:48
felipec	hmm, no, actually 'git show-branch --independent' doesn't list the commits, it's pretty easy to find out with 'git show-branch', but doesn't look like there's a direct command	13:48
	← matthiaskrgr left	13:48
	← kcj left	13:49
	← ehsan left	13:49
	→ Sigma joined	13:50
	← RaceCondition left	13:50
	→ royale1223 joined	13:51
PapaSierra	thiago, any ideas how to handle the database contents question?	13:52
thiago	PapaSierra: dump the contents to files and commit them to git	13:52
	← tyatpi left	13:53
PapaSierra	thiago, ok that was my fallback plan. i was just wondering if there was a framework that hooks into a git pre-commit or something, if not your idea is just fine, even if manual	13:53
	→ ehsan joined	13:54
	→ DuzyZiemniak joined	13:55
	← Milossh left	13:56
	← royale1223 left	13:56
	← diegoviola left	13:58
	→ squeeks joined	13:59
	← pantsman left	13:59
	→ dr_lepper joined	13:59
squeeks	Is there some sane way to apply config to all git repos in various subfolders?	14:00
	→ lkong joined	14:00
	← Robbo_ left	14:00
cmn	a for loop or an include statement in their configs	14:02
	include is new in 1.7.10	14:03
felipec	I guess there is no log format to display what name-rev displays =/	14:03
	← nioncode left	14:05
	→ tyatpi joined	14:05
	← ryez left	14:07
	← dr_lepper left	14:09
	→ blacktulip joined	14:09
	→ sh0ne joined	14:13
	→ iAFK joined	14:14
	← optixx left	14:15
	→ edude03 joined	14:15
	→ Milossh joined	14:15
	→ dr_lepper joined	14:16
	← lapistano left	14:16
	→ lapistano joined	14:16
	→ optixx joined	14:17
	← iAFK left	14:18
	→ iAFK joined	14:18
	→ Robbo_ joined	14:18
	← ns5 left	14:18
	← iAFK left	14:18
	← herdingcat left	14:18
	→ asparagui joined	14:20
	← lkong left	14:21
	→ herdingcat joined	14:21
	→ enmand joined	14:23
	→ dralston joined	14:23
	← jason237 left	14:25
	← kerrick left	14:26
	→ j416 joined	14:27
dralston	Hello, all. I have had a thought, and I want to know if it's stupid. So, rebasing a branch that you push upstream borks things up, since the resulting commit isn't a fast-forward of the upstream branch.	14:27
	← herdingcat left	14:28
	→ FernandoBasso joined	14:28
dralston	So, why not 1) create a temporary branch 2) rebase it 3) Get a diff/patch between the rebased branch and the original branch, and 4) apply that patch to the original branch?	14:28
	Wouldn't that get the the same result as a rebase without the messy upstream business?	14:29
thiago	no	14:29
bob2	ie bonghits	14:30
	→ herdingcat joined	14:30
thiago	it would get you the same tree state at the end, but it would not be the same thing	14:30
	for one, you'd have lost all the commits	14:30
	→ ipoval joined	14:30
m1sc	dralston: it's not about "isn't a fast-forward" -- !rewrite	14:30
gitinfo	dralston: [!rewriting_public_history] Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to `git pull --rebase` and even worse things if they have tagged or branched, so you must publish your humiliation so they know what to do. You will need to `git push -f` to force the push. The server may not allow this. See receive.denyNonFastForwards (git-config)	14:30
	← edude03 left	14:30
	← jackh left	14:31
cmn	dralston: the point of rewriting the history is for you to have a series of patches that make more sense to the reader	14:31
dralston	Thanks, important points I had forgotten.	14:32
	cmn: If you've made drastic enough changes, from upstream, though, merge can botch things. Rebase nicely avoids that, but obviously brings its own issues to the table.	14:33
	(Please forgive the comma incontinence)	14:34
	← simesy left	14:34
cmn	huh? rebasing is a smaller set of merges	14:34
	from the conflict point of view	14:34
	→ simesy joined	14:34
	← kenperkins left	14:34
	→ RaceCondition joined	14:34
	→ dinduks joined	14:35
cmn	and if you rebase you have to resolve them over and over, merging only the one time	14:35
	→ kenperkins joined	14:35
dralston	Yes, true. Alright, I am convinced that my idea is merely "sleepy logic". Thanks! :)	14:36
	← dinduks_ left	14:37
canton7	moltar_net1, what command are you typing to do a pull?	14:38
	moltar_net1, oops, was looking at some dim and distant part of the scrollback, ignore me	14:38
moltar_net1	canton7: thanks anyways :)	14:38
	← RaceCondition left	14:38
	→ shreya joined	14:40
	→ bondar joined	14:40
	← Sigma left	14:40
	→ kukks joined	14:41
	→ kadoban joined	14:41
	← bondar left	14:41
	← alberto56 left	14:43
	→ guampa joined	14:44
	← dralston left	14:45
	← rchavik left	14:45
	→ alberto56 joined	14:45
	→ gchristensen joined	14:46
	→ edude03 joined	14:47
	← edude03 left	14:48
Vinnie_win	Is there a simple way to create a new branch with no ancestry?	14:48
bob2	yes	14:49
Vinnie_win	...and?	14:49
bob2	but that's exceedinly rare to want and most of the time people ask about it it is because they're very confused	14:49
Vinnie_win	"gh-pages".	14:49
bob2	git branch --help	14:49
canton7	it's in man git-checkout	14:49
gitinfo	the git-checkout manpage is available at http://git-scm.com/docs/git-checkout.html	14:49
bob2	oh, oops	14:50
	← squeeks left	14:50
	← alberto56 left	14:50
Vinnie_win	git checkout -b branch --orphan followed by git rm -rf	14:50
	→ berserkr joined	14:50
	← herdingcat left	14:51
Vinnie_win	nope its "git checkout --orphan branch"	14:51
	sigh thta didn't create the branch...	14:52
	← cagedwisdom left	14:52
	→ Slix` joined	14:52
canton7	Vinnie_win, you need to commit before the branch is born	14:52
Vinnie_win	Yeah that was the ticket	14:54
	→ RaceCondition joined	14:55
	→ hellopat joined	14:55
	← mh_le left	14:56
	← juvenal left	14:56
	→ dtang joined	14:57
	← adityamenon90 left	15:02
	→ adityamenon joined	15:02
	→ Sigma joined	15:02
	← davidbanham left	15:03
	← trying2code left	15:04
	← xjiujiu left	15:05
	→ AlbireoX`Laptop joined	15:05
	← gusnan left	15:05
	→ jaseemabid joined	15:07
	→ Fernando1asso joined	15:07
	→ HyperNerdV2 joined	15:08
	→ dubellz joined	15:09
	→ adityamenon90 joined	15:10
	→ johnanderson joined	15:10
	← Chaser left	15:10
	← FernandoBasso left	15:10
	← dubellz left	15:10
	→ oriba joined	15:11
	← adityamenon left	15:13
	→ cakehero joined	15:14
	← Targen left	15:17
sh0ne	why pushing to remote repo (if a repo was not created bare but with the existing source) is not recommended?	15:18
thiago	because people expect something that git doesn't do	15:18
	→ adityamenon joined	15:19
thiago	did you read the FAQ?	15:19
	← adityamenon90 left	15:19
cbreak	there should be a robot link on faq non-bare	15:19
gitinfo	Pushing to non-bare repositories is discouraged; more details available at http://jk.gs/gitfaq#non-bare	15:19
	→ FernandoBasso joined	15:20
	← jaseemabid left	15:20
	← akiress left	15:22
	← Fernando1asso left	15:22
	← fayimora left	15:23
sh0ne	thiago: thank you, today is my first day with Git (well, first day with any CVS), so I do have some confusion. I DLed 'Pro Git' and started to read it, but I hurried with some assumptions I suppose.	15:23
	← zinjacoder left	15:23
	← cakehero left	15:23
	→ fayimora joined	15:23
	← thiago left	15:23
	← p3rror left	15:24
	← FernandoBasso left	15:24
	→ FernandoBasso joined	15:24
	→ zinjacoder joined	15:25
	→ adityamenon90 joined	15:25
	← johnanderson left	15:25
	← troyt left	15:26
	→ cakehero joined	15:27
_br_	sh0ne: gitcasts is also pretty neat for beginners	15:28
	← adityamenon left	15:29
	→ _mpu joined	15:30
_mpu	Hi, how can I see the number added/removed between working tree and HEAD?	15:30
	And between two commits?	15:30
cmn	diff --stat HEAD	15:30
	→ neurodrone joined	15:30
	← brngardner left	15:31
_mpu	Oh, thank you, that's it.	15:31
	→ brngardner joined	15:32
	→ adityamenon joined	15:34
	→ gusnan joined	15:35
	→ drigo joined	15:37
	← brngardner left	15:37
	← adityamenon90 left	15:37
	← ehsan left	15:38
	→ jaseemabid joined	15:40
	→ dreiss joined	15:40
	→ p3rror joined	15:41
AeroNotix	another small problem. I've added gitosis and that's working fine. I have access and can create and push/pull to the remote repo	15:42
	however, I cannot get the repos to show up in gitweb	15:42
	← ph^ left	15:42
	→ tdubz joined	15:42
AeroNotix	I even added the http user to the git grou	15:42
	group	15:42
	That didn't work	15:42
	I have the $projectroot pointing to the repo which gitosis uses	15:42
	→ ph^ joined	15:42
AeroNotix	then I have the gitweb = yes in the gitosis config	15:42
	→ fayimora_ joined	15:43
AeroNotix	anything I am missing?	15:43
	← fayimora left	15:43
	fayimora_ → fayimora	15:43
	← preflex left	15:43
	→ Shusshu joined	15:43
	← jgoss left	15:43
	→ preflex joined	15:44
	← sporous left	15:45
bob2	need to clarify what "didn't work" means	15:45
	→ trying2code joined	15:45
	→ sporous joined	15:45
	← fayimora left	15:45
	→ krnl joined	15:45
	→ fayimora joined	15:46
	← j416 left	15:46
AeroNotix	bob2: as in it tells me 404 no projects found	15:46
	bob2: interesting... It's a permissions issue	15:47
	← ph^ left	15:47
AeroNotix	bob2: I just ln -s the /gitosis/repositories into a chmod 777 folder outside that dir... it works	15:47
	→ gasbakid joined	15:48
AeroNotix	bob2: but it's throwing a python traceback for the gitosis-run-hook	15:48
bob2	that no one but you can see	15:48
	bpaste.net	15:48
	← Milossh left	15:48
AeroNotix	bob2: No, I read/write python quite well, I can debug from here. Thanks	15:48
	← vekexasia left	15:49
	← tdubz left	15:49
	→ tdubz joined	15:49
	← tdubz left	15:49
	← trying2code left	15:49
	→ cybersphinx joined	15:50
	→ fayimora_ joined	15:52
	← preflex left	15:52
	← the_cybersphinx left	15:52
	← fayimora_ left	15:53
	→ preflex joined	15:53
	← sporous left	15:53
	→ fayimora_ joined	15:53
	← FernandoBasso left	15:54
	→ sporous joined	15:54
	→ duckxx joined	15:55
	← fayimora left	15:55
	→ fayimora joined	15:56
	← Dave^\| left	15:57
	← AeroNotix left	15:57
	→ adityamenon90 joined	15:57
	← fayimora_ left	15:58
	← adityamenon left	16:00
	← jexmex left	16:03
	→ juvenal joined	16:03
	→ pheaver joined	16:07
	← gasbakid left	16:08
m1sc	AeroNotix: !gitosis	16:11
gitinfo	gitosis is no longer maintained and supported by the author; we usually recommend gitolite instead which has much better documentation and more features: http://github.com/sitaramc/gitolite	16:11
	← agile left	16:11
	→ matthiaskrgr joined	16:13
	→ fuzzybunny69y joined	16:13
	→ spaceonline joined	16:15
	← sporous left	16:16
	→ sporous joined	16:16
	← preflex left	16:17
	→ preflex_ joined	16:17
	← Chib left	16:18
	preflex_ → preflex	16:18
	← matthiaskrgr left	16:18
	→ hitman_ joined	16:19
	← sirdancealot2 left	16:20
	→ preflex_ joined	16:20
	gitinfo set mode: +v	16:20
hitman_	I'm trying to apply a patch to my git repo with "git am --signoff < '/home/miles/messaging.diff'" but I get an error	16:20
	→ glennpratt joined	16:20
	→ agile joined	16:21
hitman_	it says "previous rebase directory /var/www/html/fusem/.git/rebase-apply still exists but mbox given."	16:21
	what am I doing wrong?	16:21
offby1	hitman_: I wonder if git thinks you're still in the middle of a rebase	16:21
	try "git status"	16:22
hitman_	`nothing to commit` offby1	16:22
sitaram	git rebase --abort	16:22
hitman_	have never tried rebasing and not sure what that is.. it was in advanced section in a git book so I figured I can read about it later	16:23
	← preflex left	16:23
sitaram	if you're sure you didn't have one going already	16:23
	preflex_ → preflex	16:23
offby1	it's possible that "git am" does a rebase behind your back	16:23
hitman_	sitaram: It says "It looks like git-am is in progress. Cannot rebase."	16:23
sitaram	git am --abort	16:23
hitman_	no output.. must have worked sitaram	16:24
	→ sirdancealot2 joined	16:24
	← tshauck left	16:24
hitman_	btw, the patch is for an older version of my code.. should I rever to that version and then try applying?	16:24
	→ gasbakid joined	16:25
	← JackWinter left	16:25
sitaram	depends... try it on the current version	16:25
	→ trying2code joined	16:25
offby1	I never know what to do in that case.	16:25
sitaram	if you get conflicts solve them	16:26
offby1	patches don't, as far as I know, include a reference to the commit from which they were made.	16:26
hitman_	oh.. it says "Patch does not have a valid e-mail address." ... it had said that once before the rebase error crept in	16:26
offby1	if that "diff" was created by the "diff" program, then you shouldn't use "git am"	16:26
sitaram	then maybe you want 'git apply' instead of 'git am'	16:26
	← juvenal left	16:26
offby1	get your own brain etc etc	16:26
hitman_	how do I refer to that commit?	16:27
	in git apply?	16:27
	← AlbireoX`Laptop left	16:28
	→ nadavoid joined	16:28
	→ juvenal joined	16:29
offby1	dunno	16:29
	I could read the man page -- but then so could you	16:29
	← asparagui left	16:29
	← sirdancealot2 left	16:30
offby1	hitman_: in fact, I don't even understand your question. What does "refer to that commit" even mean here?	16:30
	→ sirdancealot2 joined	16:30
	← hitman_ left	16:32
	← duckxx left	16:33
	→ JackWinter joined	16:33
	← nadavoid left	16:33
milki	some asked something similiar	16:34
	how do "apply" a commit rather than than apply a patch	16:34
	closest thing is cherry-pick -n and reset	16:34
	← TheJH left	16:35
	← spaceonline left	16:37
	→ jfischoff joined	16:37
	← gasbakid left	16:37
	→ j416 joined	16:37
	→ darkmagex6 joined	16:38
	← dreiss left	16:38
	← glennpratt left	16:38
	← HyperNerdV2 left	16:38
	→ rchavik joined	16:39
	→ duckxx joined	16:42
	→ Targen joined	16:43
	→ ehsan joined	16:43
	← dinduks left	16:44
	← jeppy left	16:44
	→ tshauck joined	16:44
	→ l0ll0lll joined	16:44
	← enr^ left	16:45
	← ehsan left	16:45
l0ll0lll	hi all. I've done several stashes, and now need to view the diff for one before the last stash. I use "git stash show -p stash@{1}", but it prints nothing. What am I doing wrong?	16:46
	→ duckx joined	16:46
	→ fayimora_ joined	16:47
	← berserkr left	16:49
	← fayimora_ left	16:49
	← duckxx left	16:49
	← adityamenon90 left	16:49
	→ fayimora_ joined	16:49
	← fayimora_ left	16:49
	← fayimora left	16:49
	→ adityamenon90 joined	16:50
	→ berserkr joined	16:51
	← j416 left	16:51
	→ Textmode joined	16:53
	← JackWinter left	16:55
	→ fayimora joined	16:55
	← adityamenon90 left	16:59
	→ adityamenon90 joined	16:59
	← rchavik left	17:00
	→ JackWinter joined	17:01
	→ lenswipe joined	17:03
	← lenswipe left	17:03
	→ lenswipe joined	17:03
lenswipe	hey guys	17:03
gitinfo	lenswipe: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.	17:03
lenswipe	is it possible to remove something so that it doesnt appear to one remote but it does to another	17:03
	i have a Boxfile for pagodabox that i want to remove so that it stops appearing in future commits on github but i kinda need it for pagodabox	17:04
	its not private, its just that people downloading my code on github wont give a damn about it	17:04
	→ dvide joined	17:05
	→ dhruvasagar joined	17:05
	→ neektza joined	17:06
	← Sahua left	17:07
cmn	you can remove a file from history with filter-branch, but that's !rewriting	17:08
gitinfo	[!rewriting_public_history] Rewriting public history is a very bad idea. Anyone else who may have pulled the old history will have to `git pull --rebase` and even worse things if they have tagged or branched, so you must publish your humiliation so they know what to do. You will need to `git push -f` to force the push. The server may not allow this. See receive.denyNonFastForwards (git-config)	17:08
	→ Sahua joined	17:08
lenswipe	cmn, i dont want to remove it from history	17:08
	← adityamenon90 left	17:08
lenswipe	cmn, i just want to do git rm but only on one branch if you see what i mean	17:09
	err not one branch	17:09
	one remote*	17:09
	→ UNIXgod joined	17:09
lenswipe	i have github and my hosting repo	17:09
cmn	what do you mean one remote?	17:09
	→ Davey joined	17:09
lenswipe	ever heard of pagodabox@?	17:09
	→ adityamenon90 joined	17:09
cmn	no	17:09
	but that's irrelevant	17:09
lenswipe	no its not	17:09
	its a git based PHP scaling cloud	17:09
cmn	so waht?	17:09
lenswipe	i have a config file in my repo for that and i push to it with git	17:09
cmn	if you want to remove a file, remove it	17:10
	but a remote is only a place where commits appear	17:10
lenswipe	well no because i need it for pagodabox	17:10
	i want to stop pushing it to github	17:10
cmn	then you don't want to remove it, don't	17:10
lenswipe	sigh	17:10
cmn	then branch	17:10
lenswipe	so i have 2 branches?	17:10
	sounds reasonable	17:10
cmn	however many you need	17:10
lenswipe	ill look that up thanks	17:10
	← lenswipe left	17:10
l0ll0lll	ok, i've solved my problem by upgrading to latest version of git	17:11
	→ macabre joined	17:12
	← l0ll0lll left	17:12
	→ gasbakid joined	17:12
	← tshauck left	17:14
	→ Milossh joined	17:14
	← williamcotton left	17:16
	← dhruvasagar left	17:16
offby1	neektza: welcome	17:16
	← adityamenon90 left	17:16
	← shreya left	17:16
	→ adityamenon90 joined	17:17
neektza	offby1: thx	17:17
	← jfischoff left	17:18
	← tmz left	17:18
	← warthog9 left	17:18
	← optixx left	17:20
	← project2501a left	17:21
	→ jgoss joined	17:22
	← darkmagex6 left	17:22
	→ optixx joined	17:22
	→ scientes joined	17:25
	← adityamenon90 left	17:27
	→ adityamenon90 joined	17:27
	→ codeshah joined	17:29
	← gusnan left	17:30
	→ Adrao joined	17:31
	← tyatpi left	17:31
	← juvenal left	17:32
	→ tmz joined	17:33
	← adityamenon90 left	17:34
	← Davey left	17:34
	→ adityamenon90 joined	17:35
	← pheaver left	17:35
	← trying2code left	17:36
	→ juvenal joined	17:37
	→ BarryCrayon joined	17:38
	← adityamenon90 left	17:40
	→ adityamenon90 joined	17:41
	← BarryCrayon left	17:42
	→ und3f joined	17:43
	← enmand left	17:44
	→ xpot-station joined	17:46
	← heinrich5991 left	17:47
	→ Goplat joined	17:47
	→ arvind_khadri joined	17:49
	→ alberto56 joined	17:49
	→ Davey joined	17:49
	← adityamenon90 left	17:49
	→ adityamenon90 joined	17:50
	→ tshauck joined	17:51
	← zinjacoder left	17:52
	→ heinrich5991 joined	17:53
	→ adityamenon joined	17:53
	← calimocho left	17:54
	→ matthiaskrgr joined	17:54
	→ Dave^\| joined	17:54
	← adityamenon90 left	17:55
	→ jason237 joined	17:56
	← Sahua left	17:56
	← Davey left	17:57
	→ erichynds joined	17:57
	→ flippo joined	17:57
	→ Sahua joined	17:59
	← erichynds left	17:59
	→ vitor-br joined	18:00
	→ calimocho joined	18:00
	→ j416 joined	18:02
	→ adityamenon90 joined	18:03
	← albel727 left	18:03
	← ology left	18:03
	← arvind_khadri left	18:04
	→ albel727 joined	18:05
	← adityamenon left	18:05
Milossh	anyone around to help me with git basics? (ie. branching and remotes)	18:07
	→ jceb_ joined	18:07
jaseemabid	Milossh: !progit	18:08
gitinfo	Milossh: [!book] 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 !designers !gitt !vcbe and !parable	18:08
offby1	Milossh: ask ask	18:09
	→ glennpratt joined	18:10
	← Lord_DeathMatch left	18:10
	→ jstoone joined	18:11
Milossh	offby1: so, I have a repo(on github) where my coworkers and I work on a project, and I have forked it to my personal account. Now, what I want to get is: 2 branches on project repo, have the same in my own fork, and then have a project repo as a remote, so that I can pull in changes others made	18:11
	how'd I do that?	18:11
jstoone	Hi, my friend and I are working on a project and I just got a mail+notification at github that i can now contribute to the project, but when i push my changes it says that i cant push to the project. Any Ideas?	18:12
Milossh	do I first create a desired branch on project repo, then fork, and then add a remote? if so, how do I add remotes per branch, or it automagically gets branches, too	18:12
	jstoone: make sure you're pushing with proper credentials	18:13
	also, check the remote url(make sure it's not read only)	18:13
	← juvenal left	18:13
jstoone	Milossh: thanks, I'll just check.	18:13
	← calimocho left	18:15
	← glennpratt left	18:15
	← jaseemabid left	18:15
	→ calimocho joined	18:16
jstoone	Milossh: So, first I clone the repo, and then i change, commit/add, and then push? am i right?	18:17
Milossh	jstoone: yeah, that should be it, except that you add before you commit	18:17
	git clone http://...	18:17
	cd repo	18:17
	echo something something.txt	18:17
	git add something.txt	18:18
	git commit -m "my revision"	18:18
	git push	18:18
cbreak	you should consider writing a proper commit message	18:18
	one line title, an empty line, then body.	18:18
	→ jaseemabid joined	18:19
	← vitor-br left	18:19
jstoone	cbreak: so more like `git commit -m "Jstoone test \n\n This is a little test to see if I can commit"	18:20
	`	18:20
cbreak	no idea. I use an editor to write the message.	18:20
jstoone	oh.	18:20
	how do you commit using an editor?	18:21
	when i use `git commit -a` i get Nano popping up.	18:21
cmn	so write the message, save it and exit	18:21
cbreak	nano is an editor	18:22
	I use TextWrangler	18:22
	← jgoss left	18:22
jstoone	how do i specify which editor to use?	18:23
cbreak	jstoone: man git-config	18:23
gitinfo	jstoone: the git-config manpage is available at http://git-scm.com/docs/git-config.html	18:23
cbreak	look for core.editor	18:23
jstoone	thanks	18:23
	→ glennpratt joined	18:23
	← glennpratt left	18:25
	← lg188 left	18:25
	→ Bass10 joined	18:25
	→ Mowsh joined	18:26
	← adityamenon90 left	18:26
	→ adityamenon90 joined	18:27
	→ mikepack joined	18:27
	← jceb_ left	18:28
	→ jazz264 joined	18:29
	→ sq-one joined	18:29
	← adityamenon90 left	18:29
	← jaseemabid left	18:30
	← duckx left	18:30
	← bitkiller left	18:30
	→ adityamenon90 joined	18:31
	← lapistano left	18:31
	→ lapistano joined	18:32
	→ dreiss joined	18:32
	← jazz264 left	18:32
	← mneorr left	18:33
	→ ToBeFree joined	18:33
	← sq-one left	18:33
	→ r0bby joined	18:34
	→ LekeFly joined	18:34
	← jstoone left	18:35
	→ pheaver joined	18:35
	← robbyoconnor left	18:36
	← adityamenon90 left	18:36
	→ adityamenon90 joined	18:37
	← dreiss left	18:37
	→ duckx joined	18:37
	→ _huoxito joined	18:37
	→ dreiss joined	18:37
	→ r0bby_ joined	18:39
	← calimocho left	18:39
	→ Davey joined	18:42
	← adityamenon90 left	18:42
	→ adityamenon90 joined	18:42
	← r0bby left	18:42
	→ calimocho joined	18:42
adityamenon90	why is selecting a branch called checking out a branch?	18:43
	is that in reference to "I'm checking out this code"..?	18:43
	← hiryu_ left	18:43
	← fuzzybunny69y left	18:44
	adityamenon90 → adityamenon	18:44
cmn	selecting in what sense?	18:45
	checking out a branch means setting that branch as the current one and checking out the files from that branch	18:45
	← adityamenon left	18:45
	→ boombatower joined	18:46
	→ adityamenon joined	18:46
	← matthiaskrgr left	18:46
	← RaceCondition left	18:46
	← duckx left	18:46
	← calimocho left	18:47
	→ OOPMan joined	18:47
	→ goshawk joined	18:48
	← DuzyZiemniak left	18:49
	← Sigma left	18:50
	← umc left	18:50
	→ maletor joined	18:51
adityamenon	selecting in the sense, selecting for development	18:51
	git checkout branch_name	18:51
	now whatever I commit goes to branch_name	18:52
	so I 'selected' branch_name for development	18:52
	→ airborn joined	18:52
	← adityamenon left	18:52
	→ TomyLobo joined	18:52
TomyLobo	hi	18:52
cmn	if you want to look at it that way, I guess	18:52
	→ adityamenon joined	18:52
adityamenon	sorry cmn, I got disconnected - did anyone say anything about what I said while I was away?	18:53
cmn	no	18:53
	I just answered now	18:53
TomyLobo	i only have like 10 revisions in my repo and i want to apply some hardly automatable --tree-filter to all of them. so i figured i'd just make filter-branch wait for user input. tried "read", which failed	18:54
cmn	the name probably comes courtesy of bitkeeper and possibly subversion or older systems	18:54
adityamenon	oh - thank you...	18:54
	← cakehero left	18:55
	→ dinduks joined	18:55
	← OOPMan left	18:58
	← dreiss left	18:58
	→ HyperNerdV2 joined	19:01
	← stepnem left	19:02
	← ronniy left	19:03
	← adityamenon left	19:04
	→ adityamenon joined	19:05
	→ stepnem joined	19:06
	→ TheJH joined	19:07
	→ enmand joined	19:10
	→ pantsman joined	19:11
	← pantsman left	19:11
	→ pantsman joined	19:11
	← LekeFly left	19:11
	→ OOPMan joined	19:12
	← adityamenon left	19:13
	← optixx left	19:13
	→ adityamenon joined	19:13
	→ optixx joined	19:15
	→ johtso joined	19:16
	→ fuzzybunny69y joined	19:16
	← Lgb left	19:20
	→ fAz4 joined	19:20
fAz4	does git clone uses ssl ?!	19:20
cbreak	depends.	19:21
	← fayimora left	19:21
fAz4	cbreak: how ?	19:21
kpreid	adityamenon: originally when vcs's used exclusive access you "checked out" the file you were working on, so metaphorically you have it rather than the repository having it, and "checked in" when you had made your changes, returning it to the repository	19:21
	← enmand left	19:21
cbreak	fAz4: git only uses ssl when cloning via https	19:21
	→ gusnan joined	19:22
fAz4	cbreak: can i choose the protocol ?	19:22
	→ Lgb joined	19:22
	← rurufufuss left	19:22
kpreid	adityamenon: the name has persisted as the operation of selecting the file(s) you are going to perform work on and have "in hand" rather than stored in whatever fancy delta data store the repository uses	19:22
	← adityamenon left	19:22
cbreak	fAz4: no, you can only pick between what the remote offers	19:22
	milki picks ip over dns!	19:22
	→ adityamenon joined	19:22
cmn	but among those, you are usually free to choose, depending on your firewall rules	19:23
fAz4	cbreak: its git:// ...	19:23
cmn	that won't use ssl	19:23
cbreak	that's unencrypted unauthenticated read only	19:23
adityamenon	kpreid, that makes sense thank you :)	19:23
kpreid	adityamenon: you're timing out, did you get my second message?	19:23
adityamenon	no, I only saw the first one...	19:23
kpreid	adityamenon: the name has persisted as the operation of selecting the file(s) you are going to perform work on and have "in hand" as actual files rather than stored in whatever fancy managed storage the repository uses	19:24
	← Tinti left	19:24
adityamenon	okay - so the name has persisted, but git has changed the rules because 'everything is in-hand' now, right?	19:25
kpreid	no, I meant what git calls the working tree	19:25
	i.e. plain files	19:25
	the exclusive access and remote aspects have died out, but every vcs has a "give me a set of ordinary files I am going to modify" operation, and "checkout" is the conventional name for that	19:27
	not entirely consistent, of course	19:27
	e.	19:27
	g. i	19:27
	n	19:27
	(sorry) e.g. in Subversion, checkout is what git calls clone	19:27
cmn	with many many caveats	19:28
kpreid	er, yeah.	19:28
comps	kpreid: clone clones the object database and performs a checkout ("export" from the database to plain files)	19:29
adityamenon	ohkay	19:29
cmn	clone gets a copy of the repository, svn checkout grabs one version of the files, which is more similar to what git checkout does	19:29
comps	kpreid: you can do the checkout by yourself as well	19:29
kpreid	yes, exactly. I should have said, perhaps, that Subversion by contrast does not call the "move from one revision to another" operation "checkout"	19:29
comps	git clone --bare avoids the checkout, for example	19:29
cmn	the --bare makes the checkout senseless, it doesn't avoid it	19:30
cbreak	git doesn't call "moving from one revision to an other" checkout either	19:30
	git calls it commit	19:30
	← adityamenon left	19:30
cbreak	or reset	19:30
comps	cmn: --bare makes the checkout senseless and thus avoids it	19:30
cmn	that's not avoidance	19:30
	→ adityamenon joined	19:30
	→ linuxdaily joined	19:31
	← fuzzybunny69y left	19:31
	← Jackneill left	19:31
kpreid	cbreak: er, no. I mean the operation of selecting a different version in history as what you currently have-as-plain-files and diffing against, not the operation of defining a new version. git certainly calls that checkout; reset is either half of that or that-plus-altering-a-head depending.	19:31
	but we're arguing names for things, so I'll just shut up now	19:32
comps	kpreid: there are quite a lot of "svn->git" command cheat sheets, repo.or.cz had one of them, IIRC	19:32
cbreak	all of them suck	19:33
adityamenon	would any of you store a private repo on GitHub or BitBucket?	19:35
	is there a chance of stuff being 'stolen'?	19:35
	I can simply maintain a remote on the VPS of a customer, for example	19:35
	but these sites provide a nice interface for the non-hardcore clients	19:36
cbreak	I don't trust them	19:36
adityamenon	who still want to keep track of what's happening	19:36
cbreak	but some other people in my company uses github's private features	19:36
adityamenon	cbreak, hmm yeah, it's really hard to just put code in someone else's hands	19:36
cbreak	nah	19:37
	just set up your own server	19:37
comps	depends on how much "private" do you want to be, I don't mind putting my fluxbox config there, but I'd rather not put any "saved passwords" to anything I don't have absolute control of :)	19:37
	← krnl left	19:38
cbreak	I would not put the windows source code on github	19:38
	they'd make fun of it endlessly I bet.	19:38
	and one of the admins might leak the worst parts...	19:38
	→ BarryCrayon joined	19:38
cmn	but only for the lolz	19:39
	← angelsl left	19:39
cmn	adityamenon: the question is how much you trust them vs how much support burden having something else is	19:40
cbreak	at work I have my own gitolite repository	19:40
	but that's only for people in my branch, quite easy to administrate	19:40
	← adityamenon left	19:40
	→ adityamenon joined	19:41
	← BarryCrayon left	19:43
adityamenon	cmn, Indeed - for very small projects, I don't think there's anything to fear, there's hardly any "breakthroughs" happening	19:43
	→ jn_ joined	19:44
cmn	and you also have to keep in mind that any VPS provider is probably also able to steal your code	19:44
	or even allow third parties to do it if they're being stupid about it; there was this case with Linode, I think	19:45
sh0ne	is it good practice to add '.gitignore' file to the ignore list?	19:45
	→ DuzyZiemniak joined	19:45
SethRobertson	sh0ne: Probably not	19:46
	You typically want everyone else in the project to be able to ignore the same files	19:46
sh0ne	SethRobertson: so I should track my .gitignore file	19:46
	← balau left	19:46
SethRobertson	Most people do. There is a .git/info/excludes or some filename like that for private ignores	19:46
cbreak	depends	19:46
sh0ne	SethRobertson: thank you for clearing that up for me	19:46
	← brokenshakles left	19:47
cbreak	I put mac specific files into the global git ignore file	19:47
	← airborn left	19:47
cbreak	(Like .DS_Store)	19:47
adityamenon	cmn, come again - Linode allowed code stolen?	19:47
SethRobertson	Remembering that there is no global git ignore file that you do not specifically configure	19:47
adityamenon	was it like, they got hacked?	19:47
cbreak	I put local build specific files into .git/info/exclude (like generated .xcodeproj and build/ folders)	19:47
	→ glennpratt joined	19:48
	→ bubbles10101 joined	19:48
	← adityamenon left	19:48
	→ adityamenon joined	19:49
	→ jaseemabid joined	19:49
	← glennpratt left	19:50
	→ YPY joined	19:51
	← YPY left	19:51
felipec	I'm writting a blog post comparing git and mercurial branching models, feedback would be appreciated: http://felipec.wordpress.com/?p=1202&shareadraft=4fc1346553287	19:53
jaseemabid	I wrote a post for git beginners about aliases and shortcuts. Feedback will be appreciated. http://jaseemabid.github.com/05-07-2012/git-for-the-lazy-ones.html	19:55
	← adityamenon left	19:55
	→ adityamenon joined	19:55
	← dinduks left	19:58
	→ fuzzybunny69y joined	19:58
offby1	Milossh: did you get your question answered	19:58
Milossh	offby1: I figured it out	19:59
offby1	good	19:59
	← linuxdaily left	20:00
	→ fayimora joined	20:01
	→ Dulak joined	20:03
	← adityamenon left	20:03
	← Dulak left	20:03
	→ Dulak joined	20:03
	← Davey left	20:04
cbreak	felipec: looks ok	20:04
	→ adityamenon joined	20:04
	→ fukushim_ joined	20:04
	← fukushima left	20:05
felipec	cbreak: :)	20:06
SethRobertson	felipec: My comments are: 1) argumentative, and 2) I'm not perfectly sure that it is a stand-alone blog post or whether you need to read the post(s)? you are referring to in order to understand it	20:06
	→ Davey joined	20:07
cbreak	I only checked if it says something wrong about git	20:07
	← simesy left	20:07
cbreak	can't say anything about the hg part	20:07
	→ simesy joined	20:07
SethRobertson	I can't either, for that matter	20:07
	I will note that when you merge with git you need to merge with --no-ff to prevent the fast-forward anonymous branch problem with shoots down some of your arguments	20:08
cbreak	jaseemabid: I like git log --graph --oneline --decorate --all	20:08
	SethRobertson: personally, I only use --no-ff when merging feature branches with more than two or three commits, and a really distinct feature that I actually might care about later	20:09
	→ rcu joined	20:09
cbreak	jaseemabid: also, not sure if you use it, but there's a git completion bash file in contrib	20:10
	← adityamenon left	20:10
felipec	SethRobertson: what's wrong with being argumentative?	20:10
	→ adityamenon joined	20:11
SethRobertson	cbreak: ff is fine when you are pulling, but if there was any simultaneous development, I don't want to lose track of which branch was which.	20:11
	→ klj613_ joined	20:11
	cbreak pulls with --rebase	20:11
	← Raging_Hog left	20:11
felipec	SethRobertson: I never mentioned anything about fast-forwardness	20:11
cbreak	the projects I am involved with are two or three people at most, 90% of which is me :)	20:12
	← simesy left	20:12
bremner	90% of 3 people is you?	20:12
	that explains a lot.	20:13
cbreak	people area measured by their code size :)	20:13
	← fAz4 left	20:13
SethRobertson	felipec: Nothing per se, but I think you went over the top when you treat what I can only assume are follow-on paragraphs as if they occurred temporally after your git example. It is fair to complain that his justification doesn't match what he wants, but saying "How convenient; first he ...." seems overdone if the second paragraph was already present before you started	20:14
	← klj613 left	20:15
	← adityamenon left	20:15
	← EyesIsMine left	20:15
	← klj613_ left	20:16
	← TheJH left	20:16
	→ adityamenon joined	20:16
	→ FernandoBasso joined	20:17
felipec	SethRobertson: how would you make the same point?	20:17
	→ cakehero joined	20:17
	← boombatower left	20:17
	← fayimora left	20:18
	→ TheJH joined	20:18
	→ brngardner joined	20:18
	→ Pookz joined	20:19
SethRobertson	Just note that the justification doesn't match the desires. In the second example it is less overdone and you mention a reason why it wouldn't matter.	20:19
	Also "dependent con commits your colleague has no visibility from" is a typo and "from" is probably wrong as well	20:20
felipec	jaseemabid: have you seen __git_ps1?	20:20
SethRobertson	felipec: You could also mention that you could create a prepare-commit-msg hook to encode the branch name in the commit message if he is so sure he wants it.	20:20
	→ klj613 joined	20:21
	← adityamenon left	20:22
	← idahoev left	20:22
	→ adityamenon joined	20:22
	→ idahoev joined	20:24
	← jaseemabid left	20:24
	→ juvenal joined	20:27
	→ afuentes joined	20:29
	→ jfischoff joined	20:29
felipec	SethRobertson: dependent on commits your colleague has no visibility to?	20:29
	← stamina left	20:29
	→ Slim_ joined	20:30
SethRobertson	Perhaps "of"? Of course, what this may really be about is that the other, intermediary branch, has already been reviewed and only the new stuff on topic has not been reviewed.	20:30
	That might be where something which actually formally tracks what has been reviewed and what has not might come into play.	20:31
	← adityamenon left	20:31
	→ adityamenon joined	20:32
	← dtang left	20:32
	← moltar_net1 left	20:34
	← Dimeetrees left	20:34
	→ L0cky joined	20:35
	← idahoev left	20:36
cbreak	if you have a code review system, you'd probably tag all commits that are reviewed	20:36
	or something like gerrit	20:36
	that just rewrites history as it pleases anyway	20:36
	→ drev1 joined	20:36
felipec	SethRobertson: updated	20:37
	← pheaver left	20:38
cbreak	I've not used gerrit myself but Qt uses it	20:38
	and apparently it is made for reviewing changes in a more centralized workflow with patches instead of manually managed feature branches on the server	20:39
	→ macmartine joined	20:42
	→ resistivecorpse joined	20:43
	← Dreamer3 left	20:43
	← flaguy48 left	20:45
	← oriba left	20:47
	→ SkAzZ joined	20:47
	→ Sigma joined	20:49
	→ SkAzZ1 joined	20:49
	← Davey left	20:49
	→ pheaver joined	20:50
	→ cagedwisdom joined	20:51
	← SkAzZ left	20:52
sitaram	felipec: they have repository namespacing now? wow... I am even more convinced they knew long ago that git was doing the right thing and chose to come to it over 5 (or more?) revisions, albeit with multiple mis-steps along the way	20:56
	(ultimate WTF was moving all bookmarks when you commit, in early days of bookmarks)	20:57
	← pheaver left	20:58
	← blacktulip left	21:01
	→ blacktulip joined	21:04
	→ sorin joined	21:04
	← txomon\|home left	21:05
	← kpreid left	21:05
	→ nadavoid joined	21:06
	→ kpreid joined	21:06
sorin	Hi. I have a problem with Git mergetool. If I define a mergetool called vimdiff3 with cmd = vim -f -d -c \"wincmd J\" -c \"wincmd =\" \"$MERGED\" \"$LOCAL\" \"$BASE\" \"$REMOTE\", it fails to launch it when I execute git mergetool. But, if I rename vimdiff3 to foosbar, git mergetool works. As far as I can tell, the only available git mergetool definitions are vimdiff and vimdiff2. Why doesn't naming my tool vimdiff3 work?	21:06
	← xxtjaxx left	21:07
	→ xxtjaxx joined	21:08
	← xxtjaxx left	21:08
	→ xxtjaxx joined	21:08
	→ cirwim joined	21:09
	→ n1x joined	21:12
	← lolfrenz left	21:15
	← Sahua left	21:16
	← Vooloo left	21:18
	→ stkrzysiak joined	21:18
	← hotspant1 left	21:18
	→ hotspants joined	21:19
	← mishok13 left	21:20
	→ mxweas_ joined	21:21
	← und3f left	21:21
	→ EyesIsMine joined	21:24
	← jn_ left	21:26
	→ matthiaskrgr joined	21:27
	→ cirwin joined	21:28
	← brngardner left	21:29
	← ixti left	21:31
	← matthiaskrgr left	21:32
	← rcu left	21:34
	← Sigma left	21:38
	→ rryoumaa joined	21:38
	gitinfo set mode: +v	21:39
	→ BarryCrayon joined	21:39
rryoumaa	is there a way to keep a git repo at a depth of 2 or so?	21:39
	i don't need history, period	21:39
	← mxweas_ left	21:40
cmn	you can re-clone each time or use tarballs	21:40
	if you don't need history, git isn't going to help you much	21:40
	← bubbles10101 left	21:41
rryoumaa	i just want it for pull	21:41
	i suppose clone would work, but it might be slower	21:41
cmn	slower than what?	21:41
felipec	rryoumaa: git clone --depth=2	21:41
cmn	and how do you expect a merge to to work without any history?	21:42
felipec	rryoumaa: I suppose you meant for 'git fetch'	21:42
rryoumaa	dunno :)	21:42
cmn	if what you want is a tarball, grab a tarball	21:42
rryoumaa	um, there is a huge package out there, and it's available by git, and it's too huge, and gc takes forever	21:42
	so i just want the minimum history	21:43
	← BarryCrayon left	21:43
	← DrNick left	21:44
	← sorin left	21:44
	← benbergman left	21:46
	→ blast_hardcheese joined	21:49
	→ mxweas_ joined	21:49
	→ Dreamer3 joined	21:49
	← mxweas_ left	21:49
	← Dreamer3 left	21:50
	→ airborn joined	21:51
	→ mxweas_ joined	21:51
	→ fayimora joined	21:51
	← neurodrone left	21:52
	← Guest1598791926 left	21:52
	← j416 left	21:53
	→ benbergman joined	21:53
	→ millerti joined	21:53
millerti	Well, I've posted about my problem to the git mailing list. No solutions there. And I've put a bounty on my stackoverflow question, which usually gets some quick responses, and nothing. Maybe holiday weekend? http://stackoverflow.com/questions/10740601/cannot-push-to-remote-git-repo-over-ssh-stuck-at-writing-objects-11	21:54
	← nadavoid left	21:54
cehteh	that strace is from the client? .. how about figuring out whats happening on the server	21:56
	quota or volume limit, permissions problem whatever	21:57
	(firewall?)	21:57
	tried to push the same thing to another machine?	21:57
	and you got one response there .. check df and df -i	21:58
	← lapistano left	21:58
	→ brentratliff joined	21:59
	→ lapistano joined	21:59
	← tmz left	21:59
	← Dave^\| left	21:59
	← Goplat left	22:02
	← d0k left	22:02
	→ stamina joined	22:03
	← jfischoff left	22:04
millerti	I did strace on the server too...	22:07
	Let me give you the links.	22:08
	← Element9 left	22:08
	← Pookz left	22:08
	← mxweas_ left	22:09
cmn	from what you posted on the ML, it looks like the kernel keeps returning EINTR, so it doesn't let git continue	22:09
	why that would happen, however, is much more tricky	22:09
	→ Element9 joined	22:10
cehteh	anythings in dmesg?	22:10
millerti	server: http://www.cse.ohio-state.edu/~millerti/foo2.out	22:10
	client: http://www.cse.ohio-state.edu/~millerti/revue_strace3.txt	22:11
	← n1x left	22:11
millerti	dmesg has some ATA errors I should investigate, but nothing relating to git	22:12
	← _mpu left	22:12
cehteh	well ATA errors could be it ..	22:12
	← ISF left	22:13
	→ akiress joined	22:13
	→ mxweas_ joined	22:13
	← mxweas_ left	22:13
cmn	it's odd that they'd only affect it over ssh, though	22:13
cehteh	kernel cant deliver data .. read from disk runs in circles (maybe it will eventually give up with an EIO)	22:13
	you pushed the same repo to another server?	22:13
	'some ATA errors' would make me nervous anyways ..	22:14
	→ ill1cit joined	22:20
	← felipec left	22:26
millerti	Yeah, but I don't think these are related. All push/pull by path works fine.	22:26
	All other disk access works fine.	22:26
	→ vitor-br joined	22:26
millerti	It's only over ssh (even to/from the same machine) where there is a problem.	22:27
	My guess is that there's a git bug triggered by some commit I made, and it triggers when using ssh.	22:27
	I might want to try git://	22:27
cmn	have you tried pushing master~ instead of master?	22:27
millerti	Would that work any differently?	22:27
	What is master~ ?	22:27
cmn	the first parent of the commit at the tip of master	22:28
	i.e. not pushing all at once, but in steps	22:28
millerti	fatal: remote part of refspec is not a valid name in master~	22:28
cmn	when you push locally, what path/url do you use?	22:28
	that would be the local	22:28
	push master~ to master	22:28
millerti	cmn: I've tried two things: by path name /home/blah/blah, and by ssh. By path works fine.	22:29
cmn	by path it doesn't use the git protocol	22:29
millerti	Sure.	22:29
cmn	have you tried file:///path/to/repo ?	22:29
millerti	cmn: That works except that it complains that master is checked out over there.	22:31
	→ DrNick joined	22:31
millerti	How do I find out the current branch?	22:31
cmn	erm... what?	22:31
millerti	The cureent checked out branch...	22:31
cirwin	millerti: git symbolic-ref HEAD	22:32
millerti	Interesting. If I push via ssh (back when it worked), I never had to deal with complaints about the remote being checked out on the same branch I'm pushing. But if I do it by file, it does complain.	22:32
cmn	that's very odd	22:33
	because it should always complain in the same way	22:33
	→ setmeaway joined	22:33
	← setmeaway left	22:33
cmn	hm, unless you turned off the warnings and the error in the repo itself rather globally, maybe	22:33
	← flippo left	22:34
millerti	To start with, I'm trying to commit not to the original origin but to another repo in the same place so that I don't mess up the origin.	22:34
	So I started with a bare repo, but it's complaining about me committing to a non-bare repo, even though it should be bare.	22:34
cmn	and what does the repo say it is?	22:34
millerti	refs/heads/master	22:34
	But there are no files.	22:35
	It is something that I have TRIED to push to but always failed.	22:35
cmn	I mean what the repo says about being bare	22:35
	the one you're trying to push to	22:35
millerti	Do you want the whole error messsage?	22:35
	→ codemaster__ joined	22:35
cmn	no, I want the core.bare config option for that repo	22:35
millerti	http://pastebin.com/74VQrhaN	22:36
	I don't know how to get config options..	22:36
cmn	man git config <config option>	22:36
gitinfo	the git-config manpage is available at http://git-scm.com/docs/git-config.html	22:36
	→ bubbles10101 joined	22:36
cmn	when in that repo	22:36
millerti	false for core.bare	22:37
	Even though there are no files. But like I say, I have tried, unsuccessfully, to push to that repo.	22:37
cmn	so it thinks it's a non-bare repo, so it's right complaining	22:37
	it doesn't matter whether there are files	22:37
millerti	I'll delete it and try again.	22:37
	Ok, so I deleted the repo, did "git init" in it, and then tried pushing again. Same error.	22:38
	Should I not git init?	22:38
cmn	git --bare init is what creates a bare repo, which is the ones you generally want if you're going to push to them	22:39
millerti	Oh, ok. I probably did that with the other repo but forgot to do this one.	22:39
	→ Dreamer3 joined	22:39
	← Slim_ left	22:39
cmn	or to put it another way, if it's not in your workstation, it's probably a bad idea to make it non-bare	22:39
	← codemaster__ left	22:39
	→ Slim_ joined	22:39
millerti	Yeah. I followed instructions for the original origin.	22:40
	→ kcj joined	22:40
millerti	Anyhow, so this command works: git push file:///home/git/repos/revue_daemons_tmp/ master	22:40
cmn	hm... if that works, it means that the git protocol isn't the problem	22:41
	assuming it did what I expect it to do, and it did generate the pack; the output should be the same as over ssh, just faster	22:41
	→ setmeaway joined	22:41
	← setmeaway left	22:42
millerti	Are you sure that file:///path is really different from /path ?	22:42
	← _iron left	22:42
FauxFaux	Yes. man git filter-branch	22:42
gitinfo	the git-filter-branch manpage is available at http://git-scm.com/docs/git-filter-branch.html	22:42
millerti	I mean, what's it talking to?	22:42
	→ Krenair joined	22:42
cmn	it's talking to git	22:42
millerti	Ok, so it's having a problem with ssh then?	22:42
	FauxFaux drops a bear.	22:43
cmn	file:// uses the git protocol, and the different over ssh:// is just that it's talking over a pipe all on the same computer; over ssh it would talk over the network, but the protocol is the same	22:43
	poor bear	22:44
	it certainly points to ssh being the thing that breaks it	22:44
	← berserkr left	22:44
cmn	can you scp a large file over?	22:45
	hm, that might not prove anything...	22:45
	FauxFaux wonders if you can get git to directly use a pipe, so manually run it over ssh.	22:45
cmn	you can change GIT_SSH to a script, and it probably sends its output to its stdin	22:45
FauxFaux	Mmm.	22:46
millerti	I'm trying to set up git-daemon. Do I have to explicitly export each repo?	22:46
cmn	so it would work like tar cf - something/ \| ssh machine tar xf -	22:46
	millerti: you can tell git-deamon to export everything	22:46
millerti	I have no other problems with ssh to/from this machine.	22:46
	--export-all is in there...	22:47
	fatal: remote error: access denied or repository not exported: ~git/repos/revue_daemons_tmp	22:47
	git push git://localhost/~git/repos/revue_daemons_tmp	22:47
	I ee what's wrong.	22:47
	← gusnan left	22:47
FauxFaux	Read the git daemon log.	22:47
cmn	two things there 1) you might need to set the path for git-daemon and 2) you need to explicitly tell it to allow pushes	22:48
	← wmp left	22:48
millerti	cmn: I got that working. Now I'm going to try to push to the original master, but only after I make a backup copy.	22:48
	→ neurodrone joined	22:49
millerti	No problem there.	22:49
	If I use git protocol, it works fine!	22:49
cmn	then it's ssh's fault	22:49
millerti	Yeah. Appears so!	22:50
	← sh0ne left	22:51
millerti	I can push remotely too.	22:51
	From other machines. No problems at all.	22:51
	→ erichynds joined	22:51
	← martinjlowm left	22:52
millerti	For now, I'll just change the origins. I still want to figure out where ssh is going wrong. Other people use this repo!	22:52
	I really don't want to export the git port to the internet. I want ssh for this reason.	22:53
	← ToBeFree left	22:53
	→ JEntrep joined	22:54
cmn	letting anybody push is certainly a bad idea	22:54
	← blacktulip left	22:55
millerti	Exactly.	22:55
cmn	have you tried sending a large file over ssh in the tar fashion?	22:55
	→ comboy_ joined	22:55
millerti	I can try that, but the pregnant wife wants me to get groceries now. :)	22:55
	Thanks for the help!	22:55
cmn	like, grab 1G from /dev/urandom, and then tar cf - thatfile \| ssh thatcomputer tar xf -	22:55
	← akiress left	22:55
millerti	Good idea. Will you still be here in a few hours? :)	22:56
cmn	alright, see you around, good that we managed to track it down to something	22:56
millerti	Yes! Thanks again!	22:56
cmn	I may or may not be, it's already 1am in my TZ	22:56
millerti	Oh, wow.	22:56
cmn	but my IRC client will be here	22:56
	← toothr left	22:58
	← erichynds left	22:59
	← JEntrep left	22:59
	→ tmz joined	23:00
	→ sebastian_o\| joined	23:01
	→ SkAzZ joined	23:02
	→ project2501a joined	23:03
	← SkAzZ1 left	23:03
	← CodeVision left	23:03
	← TheJH left	23:03
	← cytrinox left	23:03
	← jaeckel left	23:03
	← d1b left	23:03
	→ d1b joined	23:03
	→ cytrinox joined	23:03
	→ CodeVision joined	23:03
	→ jaeckel joined	23:04
	→ akiress joined	23:04
	→ TheJH joined	23:04
	← mtytel left	23:04
	← cirwim left	23:04
	← brentratliff left	23:06
	← cirwin left	23:07
	← pantsman left	23:07
	→ urbanmonk joined	23:10
	← rryoumaa left	23:10
	← CodeVision left	23:10
	→ CodeVision joined	23:10
	→ twisted` joined	23:11
	← project2501a left	23:12
	→ project2501a joined	23:12
	← DoniRS left	23:16
	→ idahoev joined	23:18
	← zivester left	23:18
	→ JEntrep joined	23:20
	→ toothr joined	23:20
JEntrep	Does anyone here know any good resources for directed graph theory?	23:20
	→ teweWORK joined	23:23
	→ eletuchy joined	23:24
	← stkrzysiak left	23:25
	← afuentes left	23:25
	← Slix` left	23:26
	← Targen left	23:28
	← whitman left	23:28
	→ Slix` joined	23:28
	← ajpiano left	23:29
	→ ajpiano joined	23:29
esc	JEntrep: what do you need to know?	23:33
JEntrep	Anything and everything :p	23:34
milki	...	23:34
esc	i guess the wikipedia article would be a good start	23:34
	→ xiaotian joined	23:35
	→ codewrangler joined	23:35
JEntrep	I was told this goes over it a lot:http://en.wikipedia.org/wiki/Introduction_to_Algorithms	23:35
esc	well, that is an elementary test, really	23:36
	s/test/text	23:36
JEntrep	Oh ...	23:36
	I geuss wikipedia surfing then is my main option? :/	23:37
	← sebastian_o\| left	23:37
	← idahoev left	23:37
	← bubbles10101 left	23:37
	← giallu left	23:37
esc	well, start from there and then complement it with whatever else you find	23:37
JEntrep	Wolfram more likely actually	23:37
	→ BarryCrayon joined	23:39
	← TheJH left	23:40
	← Bass10 left	23:40
	← b1rkh0ff left	23:41
	← BarryCrayon left	23:43
	→ williamcotton joined	23:43
	→ khmarbaise_ joined	23:44
	← ajpiano left	23:46
	← khmarbaise left	23:46
	khmarbaise_ → khmarbaise	23:46
	← urbanmonk left	23:46
	→ ajpiano joined	23:49
	← Shusshu left	23:50
	→ Bass10 joined	23:50
	→ w09x joined	23:50
	← barrbrain_ left	23:55
	→ b1rkh0ff joined	23:56
	→ benjo2 joined	23:56
	← SkAzZ left	23:56
	← sirdancealot2 left	23:57
	→ sirdancealot2 joined	23:58

Logs Search ←Prev date Next date→

Channels Documentation