IRCloggy #git 2012-09-06

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2012-09-06

← acl_ left00:00
DrkShadow I did a git-reset HEAD^ to get rid of my last changes, then accidentally did a git reset HEAD^ again. I'm now one behind the branch that I pulled from, with changes in my branch. How do I recheckout the branch that I pulled from? Basically reset, either hard or soft, to the remote branch..00:00
Moult_ → Moult00:01
cmn DrkShadow: check in the reflog for the commit you want and do a reset again00:02
← beneggett left00:03
← BSaboia__ left00:04
DrkShadow ah. Thank you.00:04
→ adamm joined00:05
→ alan_o joined00:05
banghouse → banghouseFOOD00:09
alan_o is there an easy way to uninstall git?00:09
→ juvenal joined00:09
alan_o installed from source00:09
no uninstall target00:09
seems like I might have asked this here before.....00:09
→ gdelfino joined00:09
alan_o seems like I did this manually before.....00:09
← TeckniX left00:10
→ thursday joined00:10
← thursday left00:10
← variousnefarious left00:11
→ variousnefarious joined00:11
← ldlework left00:11
← pfFredd left00:12
cehteh rm is your friend00:13
→ ldlework joined00:13
alan_o patches welcome? :)00:14
ok, figured I'd ask, in case I missed something obvious. Thanks.00:14
→ jason237 joined00:15
← dogarrhea2 left00:17
← AaronMT left00:17
← intripoon_ left00:18
← pdtpatrick left00:18
SigmaVirus24 alan_o: do you have the directory you built it in?00:19
alan_o yes00:19
→ rchavik joined00:20
← rchavik left00:20
→ rchavik joined00:20
SigmaVirus24 I assume you ran configure first, yes?00:20
alan_o I'm sure I did. It's been a while00:20
← hwrd|work left00:20
SigmaVirus24 Well you should probably still have the generated Makefiles for that00:20
alan_o yes00:20
SigmaVirus24 If you can grep through those you can probably find the install locations00:21
→ hwrd|work joined00:21
alan_o yeah00:21
I think it's just a couple spots00:21
← kvanderw left00:21
SigmaVirus24 Yeah so it shouldn't be too bad00:21
→ jkitchen joined00:21
SigmaVirus24 But do yourself a favor, write up some notes on it for your future self as you do it this time00:21
If you have a blog, post the notes there. Easy to find :P00:22
alan_o yeah no doubt00:22
jkitchen so, I'm using git with smarthttp. I have some interesting client behavior. some clients will ask for my password and some will not (and then return a 401 http status code)00:22
→ kvanderw joined00:22
jkitchen is there some setting in git config which I can use to force it to ask for authentication?d00:23
SigmaVirus24 jkitchen: you mean you're administering a server acting as a remote and your clients trying to connect to it are experiencing this? Or are you the client using multiple servers with smarthttp00:24
rohan|dinner → rohan3200:24
jkitchen SigmaVirus24: I'm both, actually. I have multiple client versions (all the official git, but on centos5, centos6, and osx 10.6) and certain ones just never ask me for my password and then fail to push00:25
SigmaVirus24 ah00:25
smart http was introduced in 1.6.6 iirc00:25
So anything older than that cannot use smart http obviously00:25
jkitchen we have it set up so unauthenticated read access to the repository is available, and to push requires auth.00:26
SigmaVirus24: yea, these are all 1.7+ cliesnt00:26
SigmaVirus24 what about the version of git running on the servers?00:26
← astruasdh left00:26
← harish left00:27
jkitchen 1.7.1 (centos 6) works. 1.7.10.2 didn't00:27
1.7.10.2 is from osx 10.600:27
← pantsman left00:28
SigmaVirus24 that's really bizarre00:28
jkitchen the server is also 1.7.1 (it's centos6)00:28
yea00:28
normall, I would say just screw smarthttp and use ssh, but my boss *required* ldap auth. sshkeys were out of the question :/00:28
SigmaVirus24 If I had unicode in this terminal I'd make a face00:29
I could understand kerberos auth00:29
jkitchen the weird thing is though that my 1.7.1 clients are asking for the password immediately before doing anything, but the 1.7.10.2 is never asking for the password00:29
← jn_ left00:29
jkitchen SigmaVirus24: shrug. apparently ssh keys are insecure? dunno.00:30
I just am not sure if I'm missing something in the client config00:30
SigmaVirus24 No they're just a pain in the balls to manage (or so I would imagine)00:30
→ pantsman joined00:30
← pantsman left00:30
→ pantsman joined00:30
jkitchen yea. gitolite makes it easy-ish, but it's still a bit of a pain00:30
SigmaVirus24 jkitchen: yeah the default behavior may have changed between those versions but I don't know00:30
jkitchen ok00:30
→ intripoon joined00:31
SigmaVirus24 This may help: http://git-scm.com/2010/03/04/smart-http.html00:31
← virako left00:31
SigmaVirus24 Don't have the time to read it myself though, sorry00:31
jkitchen SigmaVirus24: it's coo, thanks for your help. I'll take a look at that article00:31
if I do figure it out, I'll let you know the solution :)00:32
SigmaVirus24 Yeah I'd appreciate that. Was trying to get my former place of employment to switch from offering SVN repos to git00:32
This was one of their concerns: authentication00:32
rohan32 → rohan|hw00:32
→ kukks joined00:32
← kukks left00:32
→ kukks joined00:32
→ dhruvasagar joined00:34
← etcetera left00:34
← click left00:34
micrcolonel → kiken00:35
→ bartek joined00:36
EugeneKay I'm pretty sure gitolite includes helper scripts for self-service SSH key management00:36
kiken → microcolonel00:37
→ etcetera joined00:37
jkitchen it does00:38
← hypera1r left00:38
jkitchen trust me, I tried. I tried hard :)00:38
EugeneKay And it even includes setup instructions for smart http mode00:38
← etcetera left00:39
EugeneKay If you were feeling really masochistic you could build a dedicate gitolite VM which uses winbind+krb5 to integrate with AD/LDAP user/passwords, then ForceCommand users into gitolite00:39
(Technically this doesn't need a dedicated VM, just a SSHd)00:40
(but if you're gonna fuck about with winbind+krb5 it's best to put it on its own VM)00:40
This would let you do SSH via password OR key into gitolite00:40
jkitchen yes00:41
that's not the issue at hand though.00:41
← Targen left00:41
jkitchen the git client never prompts me for my password00:41
EugeneKay Update to something modern.00:41
jkitchen ok,.00:41
thanks for the non-help :)00:41
EugeneKay Any time! I hope it was worth every penny00:42
→ stephenjudkins joined00:43
sitaram EugeneKay: that "masochistic" setup you mentioned: any hints on how to do that? Several people I know would jump at it...!00:43
URLs work best, since I won't be the one doing it00:43
EugeneKay sitaram - like example configs?00:43
sitaram anything at all, really (I've looked before; can't remember finding anything that ties ssh -> krb -> AD)00:44
EugeneKay The hard part is making samba+winbind play nicely with your AD domain for enumerating users to UNIX UIDs00:44
→ ehsan joined00:44
EugeneKay Lemme dig up the one I used00:44
sitaram thanks00:44
← Akufen left00:44
EugeneKay I use it primarily so my Desktop & Laptop(which are Windows, AD-managed) can auto-auth to my Samba fileserver00:44
This guide is mostly correct http://wiki.samba.org/index.php/Samba_&_Active_Directory00:45
The idmap it mentioned should be from a higher starting UID - something that won't conflict with auto-created UIDs on the system00:45
sitaram right; I remember seeing that. I was hoping something more direct. Now that you mention winbind, maybe what I was hoping for isn't even possible...00:46
EugeneKay In my environment winbind ends up taking foreeeever to reload UIDs, so I manually mapped the UIDs in AD(via the Advanced mode menu in the MMC snapin), then configured them statically in /etc/passwd00:46
This is a terrible idea for maintainability, but I rarely-to-never add users/groups so it orks00:46
→ gavin_huang joined00:47
EugeneKay For gitolite you wouldn't need to worry about the home user thing at all.... instead the ForceCommand should be set to a wrapper script which invokes gitolite via sudo or such00:47
→ boombatower joined00:47
sitaram EugeneKay: that won't work for the people I am talking about. Think "a few thousands of users, with a fair bit of churn"00:47
EugeneKay In fact, I don't know that you'd even need winbind for this.00:47
Hehe00:48
I can look into it more in about a week or two, if you're after a recipe for making it work ;-)00:48
stephenjudkins i have a repository with submodules. I use git clone --shared on the repo in the deployment process, and it's very fast. however, i must also run `git submodule init` and `git submodule update` on the shared repo, which is slow since it re-fetches from github00:48
sitaram right; I was hoping they could "ssh://" and get asked for a password, which would actually be their AD password but in this case asked by krb00:48
EugeneKay: that would be cool; thanks! I'll owe you more than a few beers :)00:48
stephenjudkins is there a way to have the new shared clone also share all the submodules from its parent?00:48
or, some way to cache submodules?00:49
EugeneKay If it works I could end you an invoice for my consulting time on it :-p00:49
sitaram EugeneKay: sure; I'd pay it the same way I got paid for gitolite ;-)00:49
cmn stephenjudkins: see man git submodule update has --reference00:49
gitinfo stephenjudkins: the git-submodule manpage is available at http://git-scm.com/docs/git-submodule.html00:49
EugeneKay I was thinking more along the lines of the corporate entity :v00:49
sitaram EugeneKay: oh that... hmm worth thinking about. The problem is, when you're an evangelist and trying to convince someone to do the right thing, you're working against everything/everyone00:50
EugeneKay Anyway, I *think* it would be possible to configure PAM such that it would accept any username, check the password against LDAP via krb5, then log you in "as" the gitolite user, and ForceCommand to the gitolite binary.00:50
But I would need to play with it00:50
The advantage of this approach is that you eliminate samba+winbind, which are pretty crap00:50
sitaram has never really understood PAM enough to make statements like that, but it would be real cool to have this!00:51
EugeneKay AND you start off as the gitolite UID/GID, so no sudo tricks are needed00:51
sitaram yup00:51
← pantsman left00:51
EugeneKay You would still want to put this on a dedicated VM, because such hacks are terrible and should not be allowed to escape into the wild00:51
SeoZ-work[AWAY] → SeoZ00:51
← iocor left00:51
EugeneKay But at least it would give you a user+pass-able ssh:// account!00:52
sitaram sure00:52
more important, tied to existing AD00:52
EugeneKay Indeed.00:52
jkitchen ok, so looking more at the issue it looks like it's getting a 401 and then trying to send auth information, but it's sending an empty password00:53
rather than prompting me for the password.00:53
→ cooloney joined00:53
sitaram jkitchen: maybe you have a .netrc file that contains an empty password?00:53
→ _huoxito joined00:53
jkitchen nope, I checked for that00:53
sitaram or an env var of some sort00:53
← bartek left00:53
→ vitor-br joined00:53
sitaram jkitchen: on that same machine, I'd create a brand new user, login to that, then try the http access and see what it does00:54
jkitchen sitaram: good call. trying that now00:54
cmn jkitchen: when is this happening? fetch? push?00:54
jkitchen cmn: push00:54
→ tmcmahon joined00:54
sitaram developers tend to accumulate cruft in their rc files :)00:54
jkitchen our fetch doesn't require auth00:54
sitaram: :)00:54
→ etcetera joined00:54
cmn does your setup ask for the password for the first request already?00:54
or does it wait until the POST with the data?00:54
dendazen is it possible to git pull only one file from the repo and not clone the whole repo?00:55
EugeneKay sitaram - for extra fun, if the user is using PuTTY from an AD-connected box, it ought to be possible to use Kerberos tickets to login. No password-entry needed!00:55
jkitchen cmn: git on this machine has never prompted me for a password00:55
seems to be related to this: http://git.661346.n2.nabble.com/git-no-longer-prompting-for-password-td7565755.html00:55
cmn that's nothing to do with it00:55
sitaram EugeneKay: putty does ticketing stuff? wow... finally something *good* about it00:55
cmn when does git get the 401?00:55
EugeneKay Yup, GSSAPI is on by default, IIRC00:55
← brainbox1100 left00:55
← gdelfino left00:55
cmn on the first request, which is the GET, or in the successive POST?00:55
jkitchen cmn: POST00:56
EugeneKay It "just works" once you have krb5+winbind running00:56
jkitchen and then it tries an authenticated post (with empty password)00:56
cmn then it's a bug that got fixed a few days ago, put your password in the .netrc or fix the apache config00:56
jkitchen cmn: ahh00:56
ok then :)00:56
stephenjudkins cmn: it looks like i'll have to use `git submodule foreach` and do the directory-name munging by hand. does that sound accurate?00:56
jkitchen lemme try that00:56
← Adam_ left00:57
cmn stephenjudkins: yes00:57
stephenjudkins cmn: thanks for your help00:57
→ beneggett joined00:57
jkitchen sitaram: brand new user, same thing. however, cmn may have solved my issue00:57
EugeneKay runs out for dinner00:57
jkitchen lemme see if brew has a new version00:57
cmn jkitchen: basically, git is expecting to be already authenticated by the time it sends the POST00:57
← rohan|hw left00:58
← hobodave left00:58
cmn unless brew uses git's next branch, it won't be there00:58
jkitchen cmn: so it pretty much *is* that email I just posted00:58
cmn: ahh ok00:58
cmn it could be00:58
crispus i just did a few commits to the master branch before realizing i forgot to change to develop first, should i just reset them and do it again manually on the right branch?00:58
jkitchen cmn: looks like brew is giving me 1.7.1200:59
sitaram jkitchen: cmn++ then :)00:59
jkitchen I'll use the .netrc trick00:59
oh, there's a karmabot in here?00:59
cmn right, that's the last release, this came after00:59
jkitchen cmn++00:59
cmn you'll have to wait for 2.0 to come out00:59
jkitchen cmn: ok, cool. thanks again!00:59
cmn np00:59
← guampa left01:00
← bapa left01:00
→ pheaver_ joined01:01
crispus wow01:01
→ luxurymode joined01:01
→ bapa joined01:01
← pheaver left01:02
pheaver_ → pheaver01:02
crispus so i just did reset HEAD~4, stashed the work, checked out the right branch, then grabbed the stash01:02
SigmaVirus24 git is wonderful, isn't it?01:03
crispus that seems really cool, is that right?01:03
yeah01:03
cmn way too much work01:03
crispus im a noob :\01:03
← beneggett left01:03
cmn git branch newbranchname; git reset HEAD~401:03
→ intripoon_ joined01:03
cmn more variations exist, but get complex, like git checkout -b newbranch; git branch -f olderbranch olderbranch~401:04
crispus very cool01:04
← intripoon left01:06
← maletor left01:06
→ beneggett joined01:07
→ pdtpatrick joined01:08
→ Yuuhi` joined01:09
→ hyperair joined01:10
← Yuuhi left01:11
→ ipalaus joined01:11
← fireh left01:13
← vitor-br left01:14
← pdelvo|afk left01:15
→ dir joined01:18
→ gtuckerkellogg joined01:18
→ pdelvo|afk joined01:19
← beneggett left01:19
← hwrd|work left01:19
← bizhanMona left01:20
jkitchen awesome, working great. thanks everyone!01:20
← jkitchen left01:20
← cannonball left01:20
→ maletor joined01:21
→ dozymoe joined01:21
← Sh0rtWave left01:21
→ bizhanMona joined01:22
dozymoe → fireh01:22
← fireh left01:22
→ acl_ joined01:23
→ beneggett joined01:24
← xrq left01:24
← afuentes left01:25
← ipalaus left01:25
← kvanderw left01:25
→ Eryn_1983_FL joined01:29
← towski left01:29
Eryn_1983_FL hello! I am trying to get my git repos to talk toeach other..01:29
i want to be able to push and pull from my puppet repo into my main git repo on my server.. is that possible?01:30
→ vitor-br joined01:30
cehteh pushing to non-bare is not really a good idea, but otherwise pushing/pulling in all kinds of ways is what git is for01:31
→ nitrix joined01:31
Eryn_1983_FL ok..01:31
hang on let me show you the error i get01:31
milki whoa01:31
talking git repos01:31
cehteh espeak :)01:32
→ pandeiro joined01:32
milki Eryn_1983_FL: !bare01:32
gitinfo Eryn_1983_FL: 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/01:32
← gusnan left01:32
Eryn_1983_FL http://sprunge.us/RAPa01:33
milki o01:33
perms01:33
ha01:33
/etc/puppet01:33
its probably by a puppet user or roto01:33
or something01:33
→ xiangfu joined01:33
milki you need to set group permissions01:33
uh01:34
cehteh setgid ..01:34
milki git init --shared=group01:34
or uh01:34
cehteh yes01:34
milki i dunno if you can do it on clone01:34
cehteh well that doesnt look bare ..01:34
milki im not particular sure how to do it on a repo thats already been created01:34
cehteh you can fix the permissions manually01:35
make dirs setgid if you want to push as another user01:35
but pushing into non-bare would be the next problem01:35
milki o, so the config option doesnt really matter?01:35
cehteh config cares for umask01:35
milki ah01:35
← Vampire0 left01:36
Eryn_1983_FL yeah i am going to do some backups... then i will play01:39
→ rohan32 joined01:40
Eryn_1983_FL perhaps tonight or tomorrow01:40
← nitrix left01:41
milki Eryn_1983_FL: include a bare repo in between the repos or use something like !gitolite01:41
gitinfo Eryn_1983_FL: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html01:41
cehteh pushing into non bare with a hard reset works also reasonable well01:42
← pdtpatrick left01:43
→ zhouxiukun joined01:43
→ nitrix joined01:44
→ jackh joined01:45
→ macabre joined01:46
Eryn_1983_FL ok01:47
← e66 left01:47
→ ecksit joined01:48
← nitrix left01:48
← dfr|sf left01:49
→ iori joined01:49
→ nitrix joined01:49
← gtuckerkellogg left01:49
→ pdtpatrick joined01:50
→ Vampire0 joined01:50
← zivester left01:51
→ zivester joined01:52
ecksit when working in a team what is the best way for people to get the latest version of the code to their local machine without causing too many dramas and crazy merges01:54
→ Cache_Money joined01:54
← Davey left01:54
→ luc_ joined01:55
→ Davey joined01:58
← tinti left01:58
← jackh left01:59
→ jackh joined02:00
→ bartek joined02:00
← DanyO83 left02:01
← Cache_Money left02:02
offby1 ecksit: I'd think "git pull" or "git pull -r"02:03
offby1 whistles innocently02:03
← luxurymode left02:04
← dhruvasagar left02:05
← blorbx left02:05
→ jdyin joined02:06
cehteh asciidoc is sometimes a bitch ... grr02:07
→ kerrick joined02:07
← karadoc left02:07
→ ipalaus joined02:09
← etcetera left02:10
→ brainbox1100 joined02:11
← stephenjudkins left02:12
← ecksit left02:12
acl_ can gitolite possibly be made to *run* under Windows?02:13
← Eridius_ left02:13
acl_ or is the effort really just not worth it, so to speak?02:14
→ etcetera joined02:14
cehteh you have an sshd under windows?02:14
← bartek left02:14
acl_ cehteh: not yet, but I'd like to know what's needed to set it up02:15
cehteh i dont know any .. but i didnt used windows in this millenium anyway02:16
gitolite is python .. cygwin may help but an sshd would prolly be the problem02:16
user handling under windows is vastly different02:16
→ luxurymode joined02:16
cehteh i'd say its not worth the effort02:17
← orafu left02:18
← LucasTizma left02:18
acl_ cehteh: that's too bad, but thanks anyway02:18
→ orafu joined02:18
cehteh maybe a virtual machine with linux or colinux may help .. but i have no idea/never used that02:19
← FernandoBasso left02:19
← AlbireoX left02:20
acl_ never really worked with VM's that way, but I'll look into that02:21
← Droolio left02:23
cehteh well if you have a coice then use a real linux server02:23
EugeneKay cehteh - gitolite is PERL.02:23
And gitolite-under-cygwin is.... well, just don't bother.02:23
→ AlbireoX joined02:24
cehteh oops .. ok perl then :)02:24
← brainbox1100 left02:24
cehteh i deployed git at some customer, using windows and thing on windows are quite ugly sometimes02:25
creating processes from a shell is way slower than on linux, things become very slow02:25
rking EugeneKay: I'm curious; why not?02:25
EugeneKay !why02:25
gitinfo Why? Because screw you, that's why.02:25
rking I'd figure a running perl is a running perl.02:25
rking sheds a tear.02:26
EugeneKay gitolite is not a persistent process02:26
rking puts it in bottle with other rejectionTears.02:26
← ipalaus left02:26
cehteh both cygwin and msys do some hard work to shoehorn unix semantics on top of windows02:26
in some cases that just doesnt work well, in others its slow at best02:26
← psoo left02:27
→ Eridius_ joined02:28
→ tijko joined02:30
→ towski joined02:30
tijko i just pushed a file into a repository with a name that already exists in it. how do i remove one of the files with the same name?02:31
cmn tijko: that sentence doesn't make sense in git02:31
you don't push files, you push commits02:31
tijko cmn commit , my fault02:31
cmn that's still impossible, you can only have one file with a given filename02:32
→ bartek joined02:32
→ wolftankk joined02:32
tijko cmn you are right the text is close mulitjack and multijack02:33
cmn so what's the problem?02:33
tijko to remove would be git add remove?02:34
filename02:34
cmn it's rm02:34
tijko cmn thanks02:35
← cakehero left02:35
tijko what if the file isn't in the directory anymore?02:35
cmn then you don't need to remove it02:36
tijko its in the repository thou02:36
cmn ask status02:36
tijko tho02:36
excuse me?02:36
i don't understand02:36
← pheaver left02:38
→ karadoc joined02:39
tijko status lists it as deleted02:39
→ ipalaus joined02:39
cmn and it should tell you what to do with it02:40
see also man git rm02:40
gitinfo the git-rm manpage is available at http://git-scm.com/docs/git-rm.html02:40
← ipalaus left02:40
← cybersphinx left02:40
tijko i should have used git rm to begin with02:40
but i didn't realize that there was a file misspelled02:40
i just git add -u02:41
thanks for the help02:41
← beneggett left02:41
← maletor left02:42
→ maletor joined02:43
→ thursday joined02:44
nitrix Git made by linus, freenode being the official server for geek stuff, gitinfo being op here, hint hint....02:44
gitinfo: That's a very nice hat that you have Linus :>02:44
Kidding xD02:44
The questions I see here are always very similar.02:45
Is it because people don't bother looking or is the documentation missing something / not easy to find ?02:45
tijko i had my question marked as a favorite on stack02:45
nitrix tijko: Oh, seriously don't take it personal, i was speaking generally.02:46
→ Cache_Money joined02:46
tijko i guess its pretty rude/lazy for me to think asking on here for a quik question02:46
nitrix tijko: Nah quit it man, it's fine xD02:46
cmn !botsnack02:46
gitinfo Om nom nom02:46
tijko nitrix no its alright,02:46
diegoviola Linus is the man http://www.youtube.com/watch?v=IVpOyKCNZYw02:47
← macabre left02:47
nitrix Just thinking of a way to maybe improve the git content.02:47
← acl_ left02:47
→ bfig joined02:47
tijko there seems to be the rtfm all over02:47
nitrix Oh noes, gitinfo is a bot ;<02:47
bfig before i lose my job: i need to revert master to this commit 7c3fbba02:47
→ diegovio1a joined02:47
← jason237 left02:47
nitrix bfig: BEFORE you do anything02:47
bfig i already backed up, tell me the magic password02:47
tarballed everything02:47
nitrix okay02:47
bfig the clock's ticking02:48
← kenperkins left02:48
cmn git checkout 7c3f -- .02:48
bfig 's blood pressure goes up exponentially fast02:48
← t0rc left02:48
cmn ten you can commit and say sorry to everyone for screwing up a bunch of commits02:48
diegoviola → Guest3755902:48
→ t0rc joined02:48
diegovio1a → diegoviola02:48
← Guest37559 left02:49
nitrix Don't miss the .02:49
bfig thanks02:49
nitrix I think $ git push -f would force push it remotly (for everyone to use), if that's what you want02:49
cmn there is no need to force anything02:50
nitrix okay sweet02:50
← t0rc left02:50
nitrix didn't know about that02:50
It simply moves the HEAD ?02:50
bfig i'm getting merge conflicts now O_O02:50
cmn what does?02:51
you can't get merge conflicts from that command02:51
bfig when trying to push02:51
cmn you don't get merge conflicts when pushing either02:51
bfig seems somebody pushed something T_T02:51
cmn if someone else already did more work, then you'll have to be smarter about it and selectively revert commits02:52
rohan32 → rohan|sleep02:55
← goshawk left02:55
← cehteh left02:55
→ cehteh joined02:56
← kukks left02:56
← Eridius left02:56
← thursday left02:56
← jcao219 left02:57
bfig feels as the pressure goes down02:59
bfig i'll have to review all commits02:59
fucking shit, dammit.02:59
infogulch → infogulch|afk03:00
→ AaronMT joined03:00
cmn this is why you use branches03:00
mh` tell us how you really feel03:00
bfig i feel angry and tired03:01
and stupid as hell03:01
← rohan|sleep left03:02
→ brainbox1100 joined03:06
home branches? :o03:06
← zhouxiukun left03:06
→ zhouxiukun joined03:07
← bartek left03:11
cmn what about them?03:12
→ Targen joined03:12
milki lol03:13
nitrix Indeed, branches would avoid errors like that.03:13
← home left03:13
nitrix Instead of everyone working in master.03:13
At least makes it easier when a last-minute decision about the project happens.03:13
← towski left03:13
→ jason237 joined03:13
bfig well, i had branched03:13
← dvide left03:13
bfig the problem is i thought the refactoring was done when it actually had broken everything03:14
but for some reason it kinda worked so i didn't realize before merging the branch03:14
nitrix ohhww03:14
cmn so undo the changes from the merge03:14
→ Goplat joined03:16
← brainbox1100 left03:16
← Eridius_ left03:17
→ solofight joined03:17
solofight people - i was exploring the advantages of Git over CVS - read through lot of mamterials over internet including this http://www.ericsink.com/entries/dvcs_dag_1.html03:18
← rc_96 left03:19
solofight what would be the top 3 or 5 points that differentiate Git from CVS from development stand point ?03:19
offline, speed of checkout ..03:19
→ nickpresta joined03:19
solofight Ive not used CVS - can somebody here who have used it give weightage for those points please ?03:20
→ anish[1] joined03:20
milki o.O03:20
speed of checkout = fast03:20
cmn CVS has no concept of commit across files, for one03:20
solofight cmn: commit across files ?03:20
cmn a commit that affects more than one file03:20
← cehteh left03:20
solofight cmn: oh03:21
cmn: please proceed03:21
→ apok_ joined03:22
← apok_ left03:22
cmn with what? that's all you need to know03:22
solofight milki: no no, am asking how and with what to give weightage and order them when compared to CVS03:22
milki o..03:22
→ cehteh joined03:22
milki does anyone even recommend cvs these days?03:23
cmn every file is versioned independently, it simply has no place in this century03:23
solofight cmn: so the only highlight would be that CVS cannot have commits which involves multiple files .. ?03:23
cmn no03:23
solofight milki: no idea - but am facing a team who wont come out of CVS03:23
cmn see svn feature list for the main diffrences between CVS and something sane03:23
← hyperboreean left03:23
milki solofight: oo, thats awesome03:23
→ hyperboreean joined03:24
solofight milki: yeah - tell em about it03:24
← artnez left03:24
← albel727 left03:24
→ Chib joined03:24
→ Eridius joined03:24
milki at least they are using version control03:24
:o03:24
← apok left03:24
SigmaVirus24 still probably better than rcs :P03:25
cmn does CVS know anything about merging?03:25
or about branches?03:26
frogonwheels solofight: I 'transitioned' my team from a file-based VCS to Git. It has been a long process, but so worth it.03:26
→ towski joined03:26
solofight milki: i earlier faced one more team which says they dont want Git because of its collaborative nature - Dev A does some changes and pushes, When Dev B pulls from server and faces a conflict - he has 3 options usually 1. overwrite existing with this 2. ignore yours and go with existing 3. have both03:26
cmn does it have commit messages?03:26
adityamenon → adityamenon_03:26
cmn solofight: that's using git badly03:26
frogonwheels solofight: I wrote my own converter that tried to convert file-based check-ins to GIT commits - collecting stuff with the same desc together.03:26
solofight now they are saying that each developer feels his code important and they will keep overwriting with theirs - i mean wtf03:26
frogonwheels solofight: lol. That's just screwed.03:27
solofight cmn: beanching and merging are not available in CVS ?03:27
→ chrisshattuck joined03:27
frogonwheels solofight: we didn't even allow multiple check-outs of the same file (check-out being a 'lock' in effect)03:27
→ pheaver joined03:27
solofight frogonwheels: how did the transition happened ?03:27
cmn solofight: there was a queston mark for a reason03:28
← chrisshattuck left03:28
solofight how did you get them open their minds first of all ?03:28
← glennpratt left03:28
frogonwheels solofight: I talked it over with a few, and then just said, "We're gonna use it"03:28
solofight frogonwheels: good team you have03:29
cmn: why did you say - that's using git badly ?03:29
frogonwheels solofight: I had a process of incrementally updating Git from the old repo, so they could play with the real version of it.03:29
→ intripoon joined03:29
frogonwheels solofight: yeah. There was much dragging of feet from one developer. Didn't even want to try. That's the battle/risk.03:29
solofight: A couple of the support guys got the hang of it quicker than this one developer. :|03:30
← pandeiro left03:30
solofight cmn: seems cvs can create branch https://kb.wisc.edu/middleware/page.php?id=4087#creating03:30
?03:30
cmn: question mark for a reason ? please explain03:31
frogonwheels solofight: not helped by the fact that I am 2700kms away from my team.03:31
solofight frogonwheels: the culture here is wrongly built - they dont want to try out anything new03:31
frogonwheels solofight: patience.03:32
← intripoon_ left03:32
cmn solofight: question marks is how you indicate that you're asking a question, rather than making an assertion...03:32
solofight cmn: so you are asking me whether cvs knows about branching, merging and has commit messages ?03:33
cmn: [08:56:18] <solofight> milki: i earlier faced one more team which says they dont want Git because of its collaborative nature - Dev A does some changes and pushes, When Dev B pulls from server and faces a conflict - he has 3 options usually 1. overwrite existing with this 2. ignore yours and go with existing 3. have both [08:56:20] <cmn> does it have commit messages? [08:56:35] <cmn> solofigh03:34
cmn: why you said this ?03:34
← nickpresta left03:35
solofight that's using git badly ? because the way i explained merging to them ?03:35
← SigmaVirus24 left03:35
milki i think thing comparing features is enough to convince cvs folks to switch if they dont understand the features03:35
i dont think*03:35
its a very different mindset03:35
cmn I'm not asking directly, I'm posing questions that should lead to the answers you're seeking03:36
the using git badly refered to your comment about why they didn't want to use it03:36
solofight cmn: ok understood03:36
cmn: then i told them not one single version control in the world can control your developers - if they want to prioratize their code on top of others03:37
milki i think it would be good to explore some workflows with git and show how the workflow is better managed with git rather than cvs03:37
← dir left03:37
solofight think their line of work is sequential development for which Git wont work out03:37
cmn what you have to find is the pain points and explain how a workflow involving git would ease that03:37
frogonwheels solofight: get support on side too.03:38
milki solofight: git pretty much manages most workflows you can think of03:38
→ sebastorama joined03:38
solofight cmn: any particular source which talks about such pain points in general about using CVS ?03:38
frogonwheels solofight: I had lots of issues with support selectively get files to include in a build - so we couldn't get a particular build reliably from source control!03:38
cmn your developers03:38
solofight frogonwheels: how did you implement this for support ?03:38
cmn why do you want to move them?03:39
solofight GIt as ticketing tool ?03:39
milki ?03:39
cmn don't use git as an issue tracker, it doesn't end well03:39
milki but it can be done!03:39
frogonwheels solofight: Not that aspect, more that I could tell support exactly what went into a particular build (I embeded the sha1 in the version resource), and reproduce it easily03:39
solofight frogonwheels: ohh you are talking about post-production support ?03:39
i had IT support in mind03:39
:p03:39
cmn: na we use bugzilla for that03:40
mh` "maam if you could just rebase the ticket and push i'll be happy to take a look at that for you"03:40
frogonwheels solofight: We've made use of jira for ticketing now (just moved across)03:40
mh`: lol03:40
milki works for github03:42
cmn what works?03:42
frogonwheels solofight: We're pretty small, so our inhouse IT support and post-production support is the same team.03:42
solofight oh03:42
milki pull requests which are issues03:42
cmn but those aren't in git03:43
solofight cmn: this link talks about branching, merging in CVS https://kb.wisc.edu/middleware/page.php?id=408703:45
← Chib left03:45
→ brainbox1100 joined03:45
← toothr left03:46
cmn ok03:46
solofight cmn: what ! they way you said i thought it was not there ! ? or am i missing something ?03:47
cmn: i remember linus saying its a foreever operation to merge branches in cvs and svn - thats what you are refering to ?03:48
→ toothr joined03:49
cmn no03:49
I had the impression that CVS didn't have merges03:49
frogonwheels solofight: old system - merging dev/release - a week or 2 set asside. Git - 5 mins for a simple branch merge 2 to 4 hours for a really complex merge involving 2 repositories that have no common tree, but a lot of common code.03:50
→ iamjarvo joined03:50
solofight frogonwheels: oh03:51
so its basically the time aspect (which is more critical ofcourse)03:51
iamjarvo hi all, not trying to start anything. friend and I were talking about merge commits. we were trying to think of a reason why they are viewed as a bad idea03:51
frogonwheels solofight: oh, and about 1000% more reliable.03:52
solofight frogonwheels: oh - why do you say that ?03:52
CVS merges are not reliable?03:52
← Fing left03:52
frogonwheels solofight: dunno. wasn't CVS. depends if it works out a real common base,03:53
← wolftankk left03:53
frogonwheels solofight: if you're 'merging' two streams using 2-way diffs only, then you're in trouble.03:53
→ arietis joined03:53
→ thirdknife joined03:54
→ Fingolfin joined03:55
→ lun_ joined03:55
← toothr left03:56
solofight frogonwheels: not sure i understood what you meant by "using 2-ways diffs only" - can you please explain a bit ?03:57
← brainbox1100 left03:57
frogonwheels solofight: when you merge, you should be using a common base file (the point where the 2 files diverged). If you are just comparing the file fro mbranch A and the file from branch B without reference to the common base, then that's a crappy way to merge is all I'm sayin'03:58
→ home joined03:58
solofight frogonwheels: understood - Git uses a 3 way diff (with a common base file) - CVS also does a 3 way diff ?03:59
→ artnez joined04:00
← thirdknife left04:01
← Tabmow left04:01
→ Hasbro joined04:01
← Hasbro left04:01
→ Hasbro joined04:01
← Corey left04:01
→ toothr joined04:04
← milleja46 left04:04
← home left04:07
→ home joined04:07
→ milleja46 joined04:09
← razdaz left04:09
→ DancingBear joined04:10
→ brainbox1100 joined04:12
← home left04:13
→ codeshah joined04:14
← keltvek left04:15
← codewrangler left04:16
← pheaver left04:17
← karadoc left04:18
→ pheaver joined04:19
← eletuchy left04:21
← tmcmahon left04:21
← crankharder left04:21
← webguynow left04:22
← tjholowaychuk left04:23
← AaronMT left04:23
→ Chib joined04:23
← tijko left04:23
solofight Thank you all for your inputs04:24
frogonwheels: cmn mliki thank you very much04:24
← solofight left04:25
→ duckx joined04:26
→ Ivoz joined04:27
← duckxx left04:27
← anish[1] left04:29
← diogogmt left04:29
← jason237 left04:29
→ Cache_Money1 joined04:30
→ Sigma joined04:30
adityamenon_ → adityamenon04:31
← Cache_Money left04:31
← duckx left04:32
→ Tabmow joined04:33
→ Corey joined04:34
→ shirish joined04:34
→ beautifulmind joined04:35
← etcetera left04:36
← luxurymode left04:36
← Chillance left04:37
← shirish left04:38
← boombatower left04:39
← Esa_ left04:39
→ pretty_function joined04:40
→ etcetera joined04:40
→ boombatower joined04:41
← Hasbro left04:41
→ shirish joined04:42
← SeoZ left04:42
← Zenmaker__ left04:42
→ Zenmaker__ joined04:43
← thebitguru left04:43
→ albel727 joined04:43
← albel727 left04:43
→ albel727 joined04:43
→ am1ja2Vu_ joined04:43
← etcetera left04:46
← boombatower left04:46
← Cache_Money1 left04:46
→ codewrangler joined04:47
→ boombatower joined04:48
← am1ja2Vu_ left04:48
← codewrangler left04:48
→ shruggar joined04:49
← ThibautD left04:50
← Fingolfin left04:50
→ Merwin joined04:50
← Sigma left04:51
→ caseymcg joined04:51
← boombatower left04:55
← bruce__zu left04:55
→ boombatower joined04:55
← boombatower left04:55
← shirish left04:57
→ boombatower joined04:57
→ karadoc joined05:02
→ thirdknife joined05:02
→ SeoZ joined05:03
→ irqq joined05:03
→ wolftankk joined05:03
← madewokherd left05:04
← adityamenon left05:05
→ adityamenon joined05:07
← juvenal left05:09
← alan_o left05:09
← diegoviola left05:09
← maletor left05:11
← Antisoche left05:11
← xiangfu left05:11
← iamjarvo left05:11
→ Antisoche joined05:12
→ diegoviola joined05:13
→ ronnieBoy joined05:13
→ private joined05:14
private → Guest2764805:14
← shruggar left05:16
← ronnieBoy left05:17
→ lng joined05:22
← vitor-br left05:22
← gerard0 left05:23
→ gerard0 joined05:23
→ xiangfu joined05:23
← Eridius left05:24
→ OOPMan joined05:24
← SeySayux left05:24
lng Hi! Do these two commands do the same thing, `git checkout -t origin/foo` and `git checkout -b foo origin/foo`?05:24
→ acl_ joined05:25
frogonwheels lng: By default, I think it does do magic to make them the same.05:25
lng: git checkout foo should work as well05:26
lng frogonwheels: thanks05:26
← cehteh left05:26
frogonwheels lng: there are config options that disable some of the magicness iirc (hence the extra options from when the magic didn't happen)05:26
→ maqr joined05:26
lng frogonwheels: `git checkout foo` - you are right05:27
→ ka2be joined05:27
← ka2be left05:28
lng then, whait is the difference between: `git checkout foo` and `git checkout -b foo`?05:28
← DrkShadow left05:28
→ shruggar joined05:29
sitaram lng: -b creates a new branch from current HEAD. without that, it creates a new branch from origin/foo (if it exists) or errors05:33
lng I see, thank you so much05:34
→ thirdknife_ joined05:34
sitaram lng: the convenience aspects of "git checkout foo" meaning "git checkout -b -t foo origin/foo" (or some such) came later; in older gits it was more explicit05:35
lng that's nice we have such defaults05:35
← thirdknife left05:36
← shruggar left05:37
milki its like magic05:37
lng how to check if branch is remote tracking one?05:37
→ beneggett joined05:37
milki git branch -v05:38
→ Eridius joined05:38
lng milki: it shows the most latest commit hash and message, right?05:40
→ piggybox_ joined05:41
→ harshpb joined05:43
← piggybox left05:43
piggybox_ → piggybox05:43
sitaram I use 'git branch -a -v -v'; no idea why05:43
← brainbox1100 left05:43
lng yea, I used -vv05:44
→ tjadc joined05:44
← RickXy left05:45
→ rrs joined05:45
→ SeySayux joined05:46
milki when in doubt, add more v's05:47
!05:47
lng does this part '[origin/analytics]' idicate the branch is tracking remote?05:47
← jgoss left05:47
lng I guess05:47
not sure it stays permanent05:47
← pheaver left05:47
← beneggett left05:48
→ longbeach joined05:53
← igotnolegs left05:54
→ weiyang joined05:55
← tjadc left05:55
→ hwrd|work joined05:56
← resmo left05:56
The24z → The24z_away05:58
pdelvo|afk → pdelvo05:59
→ arkiver joined06:00
→ maletor joined06:00
pdelvo → pdelvo|afk06:01
→ jgoss joined06:01
→ AlbireoX`Laptop joined06:01
→ gelatocar joined06:02
→ Tegi joined06:03
← gelatocar left06:03
→ _iron joined06:04
→ und3f joined06:04
EugeneKay Unless you change it, it will continue to be that branch06:04
← AlbireoX left06:04
→ flijten joined06:05
← legnaleurc left06:05
← hwrd|work left06:05
→ noyb joined06:06
→ _eMaX_ joined06:07
→ herdingcat joined06:07
→ legnaleurc joined06:07
→ ThomasLocke joined06:08
← mritz left06:11
→ soee joined06:11
→ tsdh joined06:12
The24z_away → The24z06:13
The24z Can anybody explain this flow06:13
http://pastie.org/467240606:14
→ duckxx joined06:14
The24z If pull origin master says everything is up to date, why can't I push back?06:14
← weiyang left06:14
→ weiyang joined06:15
→ cbreak-work joined06:15
murr4y does it work without tags?06:15
maybe the tags are in conflict06:16
→ iHrushikesh joined06:16
The24z yap. I guess I had local changes that were not committed.06:16
← iHrushikesh left06:16
The24z → The24z_away06:17
→ mishok13 joined06:17
← adityamenon left06:18
→ lb joined06:18
← Synthead left06:18
← thirdknife_ left06:19
→ hashset1 joined06:19
→ simesy_ joined06:20
→ papegaaij joined06:20
← yoklov left06:21
← hashset left06:21
→ thirdknife_ joined06:22
← simesy left06:23
← hyperboreean left06:24
→ pheaver joined06:24
← _iron left06:24
→ robbyoconnor joined06:24
← mishok13 left06:25
→ ph^ joined06:25
→ r0bby joined06:27
← giany left06:27
← ph^ left06:27
→ thiago joined06:27
→ skorgon joined06:27
← skorgon left06:27
→ skorgon joined06:27
← r0bby left06:28
← robbyoconnor left06:29
← unreal left06:29
→ unreal joined06:31
→ r0bby joined06:31
← adamm left06:31
← irqq left06:32
← letharion left06:34
→ adityamenon joined06:34
→ pppaul joined06:35
← pppaul left06:35
r0bby → robbyoconnor06:36
→ pppaul joined06:36
← pppaul left06:36
→ pppaul joined06:36
← pppaul left06:37
→ pppaul joined06:37
← pppaul left06:37
→ pppaul joined06:38
← pppaul left06:38
← boombatower left06:38
→ pppaul joined06:39
← pppaul left06:39
→ pppaul joined06:39
← pppaul left06:39
→ pppaul joined06:40
← pppaul left06:40
← weiyang left06:40
→ pppaul joined06:41
← pppaul left06:41
→ weiyang joined06:41
→ sebrock joined06:41
→ pppaul joined06:41
← pppaul left06:41
→ pantsman joined06:42
← pantsman left06:42
→ pantsman joined06:42
→ pppaul joined06:42
← pppaul left06:42
→ hyperboreean joined06:42
→ pppaul joined06:43
← pppaul left06:43
→ giallu joined06:43
→ pppaul joined06:43
← pppaul left06:43
← duckxx left06:43
→ pppaul joined06:44
→ gun-lpt joined06:44
← pppaul left06:44
→ pppaul joined06:45
← pppaul left06:45
→ pppaul joined06:45
← pppaul left06:45
→ pppaul joined06:46
← pppaul left06:46
← Guest27648 left06:48
← mjt left06:48
← Davey left06:48
← harshpb left06:51
→ mishok13 joined06:51
← Maccer left06:52
← simesy_ left06:53
→ simesy joined06:54
→ letharion joined06:54
→ Maccer joined06:55
← towski left06:56
→ tjadc joined06:57
→ leokaito joined06:57
→ mark_vh joined06:58
← noyb left06:58
→ peterpan joined06:58
← maletor left06:59
← acl_ left07:00
→ specialGuest joined07:01
→ maletor joined07:02
→ mischat joined07:05
← cworth left07:05
→ resmo joined07:06
→ afuentes joined07:07
← mischat left07:07
→ mischat joined07:07
→ harshpb joined07:08
→ gerryvdm_mbp joined07:10
← gerryvdm_mbp left07:10
→ gerryvdm_mbp joined07:10
→ kkszysiu_work joined07:11
bigmeow git init failed, what's wrong?07:15
← kkszysiu_work left07:15
bigmeow http://stackoverflow.com/questions/12294787/git-init-failed-whats-wrong07:15
→ Anders_J joined07:16
skorgon simple seems not to be a valid value for push.default07:18
→ hoschi joined07:18
skorgon on my system man git config lists: nothing, matching, upstream, tracking and current as valid values07:18
gitinfo the git-config manpage is available at http://git-scm.com/docs/git-config.html07:18
← bizhanMona left07:20
→ cworth joined07:20
← adityamenon left07:20
bigmeow skorgon: should i delete that line in ~/.gitignore?07:20
→ adityamenon joined07:20
← arkiver left07:20
skorgon either delete it or alter it to whatever you want this option to be. but simple is invalid07:20
→ kkszysiu_work joined07:20
bigmeow skorgon: http://linux.die.net/man/1/git-init cannot see such lists:(07:21
skorgon: simple is valid, it is generated by git07:21
skorgon: not me07:21
→ osxorgate joined07:21
skorgon it's not in man git init, but in man git config07:21
gitinfo the git-init manpage is available at http://git-scm.com/docs/git-init.html07:21
the git-config manpage is available at http://git-scm.com/docs/git-config.html07:21
← adityamenon left07:21
→ bizhanMona joined07:22
→ adityamenon joined07:22
skorgon and the error message you posted explicitly lists the valid option values. simple is not amongst them07:22
← beautifulmind left07:23
→ dr_lepper joined07:24
← gun-lpt left07:24
→ gun-lpt joined07:24
← mischat left07:26
crocket How do I sync the latest commit snapshot of a bare git repo with an FTP server?07:26
→ mischat joined07:26
← kkszysiu_work left07:26
EugeneKay crocket - !deploy07:27
gitinfo crocket: Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html07:27
crocket EugeneKay, You're a genius.07:28
← hashset1 left07:28
EugeneKay That's beside the point07:29
← thiago left07:29
→ kkszysiu_work joined07:29
→ Cache_Money joined07:30
← Element9 left07:30
bigmeow why i failed to push to git repo?07:30
http://iyanwu.com/pastebin/showthread.php?tid=177307:30
crocket "checkout -- the "fight club" of git deployment"07:31
→ ph^ joined07:32
← OOPMan left07:33
→ _Qman joined07:33
Bombe bigmeow, so, you clone a repository, then _delete_ that repository and are wondering why you can’t push anything?07:34
lb bigmeow: you dleeted your local clone07:34
← specialGuest left07:34
lb ah, what Bombe said ;)07:35
← sh4rm4 left07:35
← Vinnie_win left07:35
lb bigmeow: besides, you'll have to "cd $thatcloneddir" before doing any git command that should affect this repo07:37
→ sh4rm4 joined07:37
bigmeow Bombe: it is just a test07:37
lb ok bigmeow short version: you've failed, because you deleted your local repo before working with it.07:38
Bombe bigmeow, okay, if it’s just a test then simply continue to do stuff that doesn’t make any sense and doesn’t work. After all, it’s just a test.07:38
→ Vinnie_win joined07:39
← Vinnie_win left07:39
→ Vinnie_win joined07:40
← Vinnie_win left07:40
bigmeow why i failed to push to git repo? (with more detail)07:40
← aLeSD left07:40
bigmeow http://iyanwu.com/pastebin/showthread.php?tid=177407:40
→ Vinnie_win joined07:40
← Vinnie_win left07:40
→ stodan joined07:40
→ Vinnie_win joined07:40
← Vinnie_win left07:40
Bombe bigmeow, are you thick? You are _deleting_ your local repository before working with it. What do you expect to happen?07:40
→ Vinnie_win joined07:41
← Vinnie_win left07:41
→ Vinnie_win joined07:41
← Vinnie_win left07:41
→ Vinnie_win joined07:41
← Vinnie_win left07:41
→ Vinnie_win joined07:41
→ shruggar joined07:41
← Vinnie_win left07:42
← weiyang left07:42
→ Vinnie_win joined07:42
← Cache_Money left07:42
← Vinnie_win left07:42
→ Vinnie_win joined07:42
lb Bombe: bigmeow btw your url https://code.google.com/p/prjname/ retuns 40407:42
← Vinnie_win left07:42
→ Vinnie_win joined07:42
Bombe lb, hey, it’s just a test, didn’t you hear him?07:42
← Vinnie_win left07:42
→ simesy_ joined07:42
lb ahh sorry ;)07:42
→ weiyang joined07:43
→ Vinnie_win joined07:43
← Vinnie_win left07:43
→ Vinnie_win joined07:43
bigmeow lb: it is https://code.google.com/p/phpsdks3/07:43
Bombe: git clone, then delete it, just to assure the repo exsists07:44
Bombe Well, it’s empty, so there’s nothing to clone.07:44
bigmeow Bombe: why i cannot push to it?07:45
→ leeN joined07:45
← letharion left07:46
Bombe bigmeow, have you tried “git push <repo> master”? The first push needs a branch name, I think.07:46
→ Cache_Money joined07:46
← simesy left07:46
← simesy_ left07:47
→ Pugnax joined07:48
← pantsman left07:49
→ jn_ joined07:50
bigmeow Bombe: git push origin master07:51
fatal: remote error: Repository not found07:51
jn_ Hi I have a branch which I want to update with the latest changes from master, would it be correct to stand on my branch and do "git rebase master" ?07:51
→ pantsman joined07:53
← pantsman left07:53
→ pantsman joined07:53
→ irqq joined07:53
EugeneKay jn_ - git supports merge commits. Don't be afraid of them - use `git merge master` instead.07:53
Flattening your history with rebase for the sake of flattenign history is absurd and SVNish07:54
← harshpb left07:54
Bombe bigmeow, I think Google does something strange with their repositories, I can not reproduce that situation locally. Please contact Google for assistance, or read their help pages. I’m pretty sure they cover this situation.07:54
→ harshpb joined07:54
← mark_vh left07:54
bigmeow Bombe: http://stackoverflow.com/questions/10202015/code-google-com-git-fatal-remote-error-repository-not-found07:55
others have alsow confronted with this issue07:55
← Cache_Money left07:56
Bombe bigmeow, then go ask Google about it.07:56
bigmeow Bombe: maybe the reason is that i have two lines in .netrc, and git push only use the first line07:57
Bombe bigmeow, you could ask Google about it.07:57
Also, reading the Stack Overflow page might help you. One of the answers might be a solution for you.07:58
But of course you have already seen that.07:58
crocket EugeneKay, I'm going to modify your script and use it.07:58
EugeneKay crocket - try not to let it eat your head07:59
crocket EugeneKay, ???07:59
EugeneKay crocket - !weasels ;-)07:59
gitinfo crocket: The consequences of this proposal are not well-defined. A band of furious weasels may infest your undergarments, or it might work just fine. You should !backup then !tryit and let us know what happens.07:59
EugeneKay I license my scripts under WTFPL for a reason07:59
← xxtjaxx left08:00
← pantsman left08:00
Anders_J I'd check if the backup works first <.<08:00
→ xxtjaxx joined08:01
← xxtjaxx left08:01
→ xxtjaxx joined08:01
→ LimeBlast joined08:01
← harshpb left08:02
→ thiago joined08:03
→ harshpb joined08:03
← harshpb left08:03
jargon- i've reinstalled my netbook. is there anyway to get to reuse the ssh key that's stored in github?08:03
or do i need to create a new pair?08:03
→ harshpb joined08:03
→ kngspook joined08:04
← Milossh left08:04
Bombe jargon-, if your private key is lost then, well, your private key is lost.08:04
That’s what “lost” means, I guess.08:04
jargon- Bombe: aah. ok.08:04
gitinfo set mode: +v08:05
jargon- too bad the priv key is not L O S T08:05
then we could just go back08:05
;-)08:05
Bombe Indeed. :)08:05
jn_ <EugeneKay> thanks =)08:05
kngspook Will setting "-diff" on a filetype in .gitattributes in any way change how git stores the changes to the file? The git diff then claims it's a binary file, but I'm wondering if that's just an unhelpful message rather than an actual change in classification.08:05
→ Xethron joined08:06
kngspook (Otherwise it sounds like it would be synonymous with setting the filetype to "binary" in gitattributes...)08:06
→ blami_ joined08:06
EugeneKay git stores files as they existed at the time of commit08:06
kngspook Regardless of binary or text settings?08:07
EugeneKay Correct08:07
bigmeow Bombe: .netrc have something to do with curl, maybe git invoke curl when use http or https08:07
→ letharion joined08:07
bigmeow Bombe: curl just don't fetch the right line08:07
Bombe bigmeow, and maybe Google knows more.08:07
EugeneKay Internally git will repack these file blobs into packfiles, but it retains the ability to unpack them 100% as they were commited08:08
bigmeow Bombe: i am sure the problem is curl related08:08
EugeneKay Whenever possible you should NOT mark files as !binary08:08
gitinfo Storing binary files in git causes repo balloon, because they do not compress/diff well. In other words, each time you change a file the repo will grow by the size of the file. See !annex for some solutions08:08
Bombe bigmeow, great. Problem solved, next!08:08
kngspook EugeneKay: Well, I'm not really looking to mark files as binary. I have text files that I just want not to be diffed by default.08:09
EugeneKay Define "not to be diffed"08:10
You can't change how git creates packfiles, and worrying about plumbing that low-level for ideological reasons is psychopathic08:10
bigmeow Bombe: not solved yet08:11
Bombe: can i specify which .netrc to use when i use git push?08:11
Bombe bigmeow, but it’s a curl problem. Maybe #curl or ##curl can help. Or the curl developers.08:11
drarok → Drarok08:11
EugeneKay bigmeow - it sounds like you're having problems pushing via http:// ?08:12
bigmeow Bombe: not a pure curl problem , since git invoke it, and the issue maybe git don't invoke curl smartly08:12
→ lun__ joined08:12
Bombe bigmeow, git does not invoke curl.08:12
Why would git invoke curl?08:12
bigmeow EugeneKay: yes, dude.. are you admin herer?08:12
Bombe That’s crazy.08:12
EugeneKay bigmeow - technically yes, but only by virtue of being an overhelpful regular. See also the !stats08:13
gitinfo bigmeow: [!irclog] Public logs of #git are kept at: http://colabti.org/irclogger/irclogger_log/git or look at stats for the last 30 days: http://itvends.com/irc/git.html08:13
← Eridius left08:13
kngspook EugeneKay: I want git to do everything it's doing now, except show files with the extension .x when a user does git-diff.08:13
EugeneKay (which I maintain)08:13
bigmeow Bombe: maybe git use libcurl and don't invoke curl through curl cli08:13
kngspook EugeneKay: treat them exactly the same in every respect except for displaying.08:13
← lun_ left08:13
→ Milossh joined08:13
EugeneKay bigmeow - git definitely does not `curl`; it is built with libcurl.08:14
crocket Does git allow custom hooks?08:14
→ tjadc_ joined08:14
EugeneKay kngspook - then specifying the -diff option as you were saying is in fact what you want to do.08:14
crocket a hook such as "custom"08:14
bigmeow EugeneKay: git is not smart enough08:14
EugeneKay crocket - tl;dr: no.08:14
bigmeow EugeneKay: when there are two lines in .netrc08:14
EugeneKay bigmeow - git is stupid. It says so in the man git page ;-)08:14
gitinfo bigmeow: the git manpage is available at http://git-scm.com/docs/git.html08:14
← tjadc left08:15
EugeneKay bigmeow - git's support for http(s):// is, to not beat about the bush, total crap. Use ssh:// like a sane person and abuse ~/.ssh/config to your heart's content.08:15
If your $REPOHOST doesn't support ssh:// then you should stop using them, because that is stupid and backwards and not gitty.08:16
bigmeow EugeneKay: but i am using google code, which do not support ssh08:17
EugeneKay bigmeow - see above re:$REPOHOST08:17
→ chimay joined08:17
← chimay left08:17
→ chimay joined08:17
← diegoviola left08:17
crocket EugeneKay, I can handle your script.08:17
Just replacing rsync with lftp is enogh.08:18
← Goplat left08:18
crocket enough08:18
EugeneKay makes a rude noise about FTP08:18
crocket EugeneKay, I have no other choice.08:19
EugeneKay, I don't want to use FTP, either.08:19
← sirdancealot7 left08:19
→ dagerik joined08:19
EugeneKay Get a better webhost that uses a protocol updated since 2000 ;-)08:19
dagerik when i do git log and search with /, the first hit's hashvalue is not shown because all of the above is "skipped". halp08:20
jn_ How do I change the default mergetool when using git mergetool ?08:20
→ dpino joined08:20
EugeneKay jn_ - man git-config, ctrl-f for "mergetool"08:20
gitinfo jn_: the git-config manpage is available at http://git-scm.com/docs/git-config.html08:20
jn_ ok thanks08:21
crocket EugeneKay, Which protocol?08:22
EugeneKay crocket - SFTP, rsync-over-ssh, scp.... anything that's, you know, secure.08:23
→ stepnem joined08:24
→ diegoviola joined08:24
→ tinti joined08:25
← karadoc left08:25
← cedr left08:26
← pdtpatrick left08:29
← mischat left08:30
bigmeow how to tell git not to use .netrc?08:31
← Chib left08:31
wereHamster bigmeow: it's curl that uses it.08:32
→ RoyBellingan joined08:33
← awx left08:33
→ karadoc joined08:33
← ThomasLocke left08:34
→ awx joined08:34
bigmeow wereHamster: i know, is there any method to tell git which.netrc should be used?08:35
wereHamster: i know, is there any method to tell git which.netrc curl should used?08:36
Bombe bigmeow, no, because curl uses it, not git.08:36
→ ThomasLocke joined08:36
bigmeow Bombe: but git should tell which .netrc curl should use, otherwise curl will use the default config in .netrc08:36
wereHamster bigmeow: curl always uses the one in $HOME, it seems.08:37
← xxtjaxx left08:37
bigmeow Bombe: if there are more than one .netrc files, git has the responsibility to tell curl which config it should use08:37
wereHamster: curl have option to define which .netrc file it use,08:38
← daniel-s__ left08:38
→ daniel-s joined08:38
wereHamster bigmeow: which one?08:38
→ w41_ joined08:38
→ Eridius joined08:39
Bombe bigmeow, I am still a little bit convinced that you are doing this all wrong and that the Stack Overflow page you posted earlier does indeed contain a solution for you.08:40
Apparently, using Google Code is already a “doing it wrong.”08:40
bigmeow wereHamster: --netrc-optional08:40
Bombe: no solution yet08:40
Bombe: or i should remove .netrc, which is not convinent08:40
when i do git push08:40
→ jceb joined08:41
bigmeow Bombe: but with it, i cannot push using two google account08:41
Bombe bigmeow, you know what? I suddenly stopped caring. Please go away.08:41
← w41 left08:41
wereHamster bigmeow: why do you need to change the netrc?08:42
bigmeow shrug08:42
wereHamster: use more than one google code account08:42
FauxFaux .netrc supports the alias hack, wherein you use a bogus url and override the hostname, doesn't it?08:42
→ _eMaX_1 joined08:43
wereHamster like in ssh_config?08:43
(though there it's not really a hack, but the way you should use the config file..)08:43
← _eMaX_ left08:43
← fisted left08:44
← Fuuzetsu left08:44
← sebrock left08:44
→ xxtjaxx joined08:44
← xxtjaxx left08:44
→ xxtjaxx joined08:44
bigmeow wereHamster: i can use more than one github account if i change .ssh/config08:44
FauxFaux: how to hack .netrc08:45
thiago use different "Host" in .ssh/config08:45
thiago has "Host qt-gerrit" and "Host qt-gerrit-personal" so that two different keys are used for the same server08:45
→ Shanachan joined08:46
← Shanachan left08:46
→ Shanachan joined08:46
bigmeow thiago: this can be done for ssh, but can it for http？08:47
thiago: now we talk about libcurl not ssh08:47
crocket EugeneKay, What is if [ ! -d "${dest}" ] ?08:47
${dest} is not a folder.08:47
thiago bigmeow: http (usually) has no keys08:47
EugeneKay A weasel08:47
thiago bigmeow: besides, you should not push via HTTP08:47
← skorgon left08:47
thiago bigmeow: push via SSH08:47
FauxFaux wereHamster: Mmm, I just dislike having "ssh example.com" going to example.com normally, and "ssh foo" going to example.com through a different proxy (for when I'm at a different location). Yes, I actually use 'foo'. >.>08:47
crocket EugeneKay, Is it a trap?08:48
wereHamster FauxFaux: why dislike?08:48
thiago <ackbar>it's a trap!</ackbar>08:48
EugeneKay It's a check that the destination dir exists on the local filesystem08:48
FauxFaux Seems like there should be a neater solution.08:48
EugeneKay If you're FTPing or such to a remote you won't want it ;-)08:48
crocket EugeneKay, It's for a remote directory.08:49
→ Element9 joined08:49
crocket Did you ever see it run without an error?08:49
git archive is run outside a git repo.08:49
→ fisted joined08:49
thiago git archive is run inside the repo08:49
you can give it a --remote= option though08:50
→ skorgon joined08:50
← skorgon left08:50
→ skorgon joined08:50
← xxtjaxx left08:50
Kiall → zz_Kiall08:50
→ robotmay joined08:52
bigmeow \does .netrc support alias?08:52
crocket thiago, If $GIT_DIR is set, does "git archive" run on $GIT_DIR?08:53
thiago crocket: you can test08:54
→ mjt joined08:54
crocket EugeneKay, I'm sure ${dest} wouldn't exist.08:54
${dest} directory08:54
→ eletuchy joined08:56
← thorbjornDX left08:57
← eletuchy left08:57
zz_Kiall → Kiall08:58
→ eletuchy joined08:58
→ polymar joined08:58
← dpino left08:58
→ thorbjornDX joined08:59
→ gnoshi joined08:59
→ SkAzZ joined09:00
← RoyBellingan left09:00
→ RoyBellingan joined09:01
← harshpb left09:01
→ psoo joined09:02
→ harshpb joined09:02
← oojacoboo left09:03
← adityamenon left09:03
crocket EugeneKay, Your script doesn't seem to check if an updated branch is deploy.${branch}09:04
EugeneKay It doesn't do a lot of things.09:04
Hence the caveats09:04
crocket EugeneKay, oops.09:04
EugeneKay Patches welcome09:04
crocket I misunderstood.09:04
EugeneKay sleeps09:04
crocket EugeneKay, Why umode 007?09:04
jast hey, no sleeping on the channel!09:04
EugeneKay Because that's the umode that suited /me/09:05
→ oojacoboo joined09:05
EugeneKay sticks jast with a caffiene IV and then wanders off to bed09:05
crocket me!!09:05
jast not good. I'm at work. I kind of need focus more than hyperactivity.09:05
→ manolo joined09:06
← dsantiago left09:07
→ karadoc_ joined09:07
→ noisebleed joined09:07
← noisebleed left09:07
→ noisebleed joined09:07
← karadoc left09:09
← maletor left09:09
← jackh left09:10
← harshpb left09:11
← weiyang left09:11
→ dsantiago joined09:11
← polymar left09:12
← diegoviola left09:12
→ FasTTo joined09:12
→ KungFuPanda joined09:12
← piti left09:12
→ karadoc joined09:12
← jargon- left09:12
→ xxtjaxx joined09:12
KungFuPanda a09:12
← xxtjaxx left09:12
→ xxtjaxx joined09:12
← KungFuPanda left09:12
→ i42n joined09:13
gnoshi Hey: probably a really simple question, but one I haven't found an answer for: is it possible to disconnect a repo from a cache (i.e. if created with git init --reference) without doing a repack (i.e. git repack -a)? I don't want the repo packed (because other repos --reference it, and packing seems to screw that up), but need to disconnect it from what it refers to.09:13
→ PigeonFriend joined09:14
← peterpan left09:14
← karadoc_ left09:14
← kngspook left09:15
→ mabrand joined09:16
→ athit joined09:16
→ talk joined09:18
→ weiyang joined09:18
selckin sounds like you made a fun little chain09:19
talk Hello I have already committed some changes and I have to do some additional changes on this commit.And after this both changes,I need to send this as a single commit.How can it be done?09:20
skorgon talk, commit --amend09:20
→ mischat joined09:21
→ harshpb joined09:23
jast gnoshi: copy everything from the referenced object store into the target object store09:23
and you *can* repack... just make sure you don't prune anything (e.g. git repack -Ad)09:24
gnoshi cool, thnx09:27
→ polymar joined09:28
talk skorgon: great.I just had to add git add -f file_name and do git commit --amend.09:29
it worked.09:29
← xxtjaxx left09:29
→ catphish_ joined09:30
→ mark_vh joined09:30
→ FernandoBasso joined09:30
→ codeM0nK_ joined09:30
→ pfFredd joined09:31
← codeM0nK3Y left09:31
codeM0nK_ → codeM0nK3Y09:31
DeadZen there a way to get a short status list of different files between branches?09:33
canton7 DeadZen, diff --name-status or --name-only09:34
DeadZen ah thx09:35
→ xxtjaxx joined09:36
bigmeow EugeneKay: hi, dude09:38
← herdingcat left09:39
← Element9 left09:39
← gun-lpt left09:39
Shanachan → Fuuzetsu09:39
← gnoshi left09:40
→ gun-lpt joined09:41
→ dsirijus joined09:41
→ SegFault1X|work2 joined09:42
← karadoc left09:42
← FasTTo left09:43
← leeN left09:43
→ mklappstuhl joined09:44
← SegFaultAX|work2 left09:44
bigmeow lol09:44
https://secure.gravatar.com/avatar/5150ef019ba97242bbbf09d9db371e04?s=140&d=https://a248.e.akamai.net/assets.github.com%2Fimages%2Fgravatars%2Fgravatar-user-420.png09:45
who is this?09:45
→ leeN joined09:45
jast I believe that's Eugene09:45
→ SegFaultAX|work2 joined09:47
FauxFaux Looks like the frogman to me: http://i.imgur.com/Elo3V.gif09:47
← SegFault1X|work2 left09:48
bigmeow FauxFaux: yep, frogman:)09:48
← xxtjaxx left09:49
→ KungFuPanda joined09:49
KungFuPanda hi guys, I got "fatal: CRLF would be replaced by LF in index.php." when I try to add index.php. I already created .gitattributes and set text=auto, and I also set core.autocrlf input.09:50
but I still got this error can not add the file.09:51
→ xxtjaxx joined09:51
← xxtjaxx left09:51
→ xxtjaxx joined09:51
selckin the best way is to not have git mess with it, and disable autocrlf, and use a non braindead editor09:51
canton7 git config core.safecrlf09:51
→ nathandim joined09:51
selckin any editor that still doesn't understand lineendings is not wroth uhsing09:52
KungFuPanda my safecrlf is true09:52
canton7 selckin, it can be handle to get git to ensure that no crlfs are committed09:52
KungFuPanda, read core.safecrlf in man git-config09:52
gitinfo KungFuPanda: the git-config manpage is available at http://git-scm.com/docs/git-config.html09:52
→ arkiver joined09:52
selckin it doesn't mather what it has, only that it doesn't change09:52
→ FasTTo joined09:52
← weiyang left09:53
→ beautifulmind joined09:53
jast !crlf09:54
gitinfo to fix problems with line endings on different platforms, check out http://line-endings.gitrecipes.de/09:54
KungFuPanda thanks, I will have a look the safecrlf, because I want to git to handle the line ending, we got linux, max and window developer. so it should be good,if git can handle it09:54
masak I'm wondering about git's tendency to store a SHA1 in a two-letter directory, like .git/objects/af/5626b4a114abcb82d63db7c8082c3c4756e51b -- this is so as not to overwhelm the .git/objects/ directory with too many nodes, yes?09:54
jast masak: yes09:54
← i42n left09:54
KungFuPanda thanks everyone, I will have a look09:54
jast it's somewhat nicer for people trying to sift through, too09:55
canton7 KungFuPanda, the gist is that git's being extra-safe about changing line endings, and thinks your file might be binary, as it contains both lf and crlf, so is refusing to do the conversion09:55
masak jast: oh, true.09:56
jast: does it change the directory structure further as more nodes are added?09:56
← arkiver left09:56
masak hm, doesn't seem that way.09:56
jast that's because it doesn't ;)09:57
→ i42n joined09:57
jast in any case, you're supposed to run 'git gc' or something like that every now and then, and that usually gets rid of most of the loose files09:58
→ luc__ joined10:01
← luc_ left10:01
→ jargon- joined10:06
→ Hail_Spacecake joined10:06
← talk left10:06
→ gbjk joined10:07
← caseymcg left10:09
← xxtjaxx left10:09
gbjk Hi. I accidentally deleted a branch locally. I thought I'd pushed it to origin but it doesn't seem to be there either. I have the sha, but it's reported as a bad object. Is here anything I can do?10:10
cbreak-work gbjk: did you use -d or -D?10:10
catphish_ if you checked it out lately try git reflog10:11
to find the sha10:11
← spyro left10:11
lb catphish_: thought he has the sha10:11
catphish_ oh sorry10:11
that's very odd10:11
gbjk cbreak-work: -D I think.10:11
catphish_: Why is that very odd?10:11
catphish_ the sha shouldn't go away immediately when deleteing a branch pointed to it10:12
ie commits arent deleted10:12
canton7 double-check the reflog, you might have the sha1 wrong10:12
gbjk realises something.10:12
FauxFaux !gka10:12
gitinfo For a better way to view the reflog, try: gka() { gitk --all $(git log -g --format="%h" -50) "$@"; }; gka10:12
gbjk FACEPALM.10:12
← _eMaX_1 left10:12
gbjk It's in another local repository. Nevermind.10:12
catphish_ lol10:12
cbreak-work good10:13
→ tomaw_ joined10:13
← FernandoBasso left10:13
← gavin_huang left10:15
← eletuchy left10:19
→ mastro joined10:20
→ caddoo joined10:22
→ whitman joined10:24
→ spyro joined10:24
Hail_Spacecake I created a gh-pages branch on github10:26
and for some reason it doesn't exist locally10:26
how do I create a local version of the branch?10:26
canton7 Hail_Spacecake, git branch -r ?10:26
→ xxtjaxx joined10:26
← xxtjaxx left10:27
→ xxtjaxx joined10:27
Hail_Spacecake yeah I see the branch there10:27
github/gh-pages10:27
but I tried checking it out10:27
canton7 so, you can create a new local branch based on that remote on with 'git checkout -b gh-pages github/gh-pages'10:27
Hail_Spacecake and I somehow made a local branch literally called "remotes/github/gh-pages"10:27
canton7 oh, nice. get rid of that10:28
(easily done btw)10:28
Hail_Spacecake I did10:28
canton7 in general, !local_branch_from_remote10:28
gitinfo The following commands are all equivalent, assuming <branch> doesn't yet exist: 'git checkout -b <branch> <remote>/<branch>', 'git checkout -t <remote>/<branch>', 'git checkout <branch>'. The latter invokes some magic.10:28
Hail_Spacecake so, what happens (or is supposed to happen) if you just do git checkout -b <branch>?10:28
→ ro_st joined10:28
→ mritz joined10:28
Hail_Spacecake where <branch> is the name of a remote branch?10:28
canton7 it creates <branch>, based on the current commit10:28
← KungFuPanda left10:28
canton7 if <branch> is the name of a remote branch, it doesn't do what you expect: it creates a branch called e.g. 'github/gh-pages' based on the current commit10:29
Hail_Spacecake so that gh-pages branch I just made locally has no relation to the gh-pages branhc on github I care about?10:29
canton7 correct10:29
Hail_Spacecake oh -b is what you pass to checkout to make a new branch without first using git branch10:29
bah10:29
yeah10:29
canton7 yup :010:30
* :)10:30
→ fluxdude joined10:30
→ dvide joined10:30
canton7 some kind soul made just 'git checkout gh-pages' work the same as 'git checkout -b gh-pages github/gh-pages', provided a set of assumptions are met10:30
← jargon- left10:30
→ jargon- joined10:31
Hail_Spacecake canton7 huh?10:31
I think that will work for me now10:31
that I've set up the remote tracking10:31
jargon- is there a way to clone a mercurial repo with git? i don't want to install hg10:31
canton7 jargon-, hg-git? (or it might be git-gh, can't remember)10:32
Hail_Spacecake but before I was getting something like this ref doesn't exist10:32
canton7 Hail_Spacecake, cool10:32
→ swair joined10:32
Hail_Spacecake well anyway pushed10:32
so I know it takes a few minutes for github to build the page10:32
but I thought it at least put up the notification saying it was building the page10:32
pretty quickly10:32
← hyperair left10:33
→ gnoshi joined10:34
← gnoshi left10:34
← harshpb left10:35
jargon- canton7: ok let me look into that10:35
canton7 *git-hg10:35
→ harshpb joined10:35
→ bfgun joined10:36
← Xethron left10:38
← i42n left10:38
← swair left10:39
→ swair joined10:39
← heroux left10:39
← swair left10:39
← bfig left10:39
← luc__ left10:39
→ swair joined10:39
→ MaximLacrima joined10:40
ro_st is there a simple way to achieve automated build numbering with git in a particular branch?10:40
→ luc_ joined10:40
ro_st i've looked into semantic versioning but i get hopelessly lost in theory and what looks like a legal document10:41
skorgon ro_st, i think some projects use something like the output of 'git describe'10:42
← tinti left10:42
← dr_lepper left10:43
bremner ro_st: yeah, git describe | sed should work10:43
ro_st ok so we need to be using tags10:43
bremner yes10:43
→ i42n joined10:43
ro_st oh nice, it returns tag commit count10:43
how simple and elegant10:43
thanks guys. that sorts it out!10:44
← tomaw left10:45
← kkszysiu_work left10:45
tomaw_ → tomaw10:45
← luc_ left10:45
← Hail_Spacecake left10:46
→ Droolio joined10:46
jargon- canton7: look what i found :-) http://pypi.python.org/pypi/git-remote-hg/0.1.110:47
← polymar left10:47
→ kkszysiu_work joined10:49
← harshpb left10:49
→ harshpb joined10:49
← dsirijus left10:50
← osxorgate left10:52
← SkAzZ left10:52
→ arvind_khadri joined10:53
→ Vile joined10:56
→ osxorgate joined10:56
→ osxorgat_ joined10:59
← osxorgate left11:00
← harshpb left11:01
→ gavin_huang joined11:01
→ harshpb joined11:02
→ osxorgate joined11:03
→ jn__ joined11:03
← osxorgat_ left11:03
← gbjk left11:03
← jn_ left11:07
→ visualinfo joined11:07
→ karadoc joined11:07
← bloony left11:10
→ Sunny_ joined11:11
→ cebor joined11:11
→ bloony joined11:14
← bloony left11:14
← visualinfo left11:16
← jdyin left11:16
← mabrand left11:17
← nathandim left11:18
← lun__ left11:19
→ cakehero joined11:20
← robbyoconnor left11:20
← gun-lpt left11:23
→ bloony joined11:24
→ stevedomin joined11:24
← jceb left11:24
← ro_st left11:25
← SeoZ left11:25
→ rwestphal joined11:25
→ MWinther joined11:25
→ SeoZ joined11:26
→ dfr|sf joined11:28
→ eletuchy joined11:30
→ keltvek joined11:32
→ sgambino joined11:32
→ rgsteele joined11:34
← eletuchy left11:34
← mritz left11:35
← keltvek left11:37
→ keltvek joined11:38
← wolftankk left11:39
← xxtjaxx left11:39
← Sunny_ left11:40
← thiago left11:40
→ SkAzZ joined11:41
→ shineworld joined11:43
← shineworld left11:43
→ jaseemabid joined11:46
← mark_vh left11:46
← i42n left11:50
→ skyf joined11:50
→ skyf_ joined11:50
→ a-l-e joined11:51
→ mritz joined11:51
→ alberto56 joined11:52
← SeoZ left11:52
← zhouxiukun left11:52
← TheFuzzball left11:53
→ TheFuzzball joined11:53
→ cannonball joined11:54
→ iocor joined11:54
← harshpb left11:54
← TheFuzzball left11:54
→ TheFuzzball joined11:55
← TheFuzzball left11:55
→ TheFuzzb_ joined11:56
← skyf left11:56
→ the_cybersphinx joined11:56
← skyf_ left11:57
→ Ryback_ joined11:58
← Ryback_ left11:58
→ Ryback_ joined11:58
→ geekbri joined11:58
→ bartek joined11:58
→ koltroll joined12:01
→ SeoZ joined12:01
← comps left12:02
← xiangfu left12:03
← Nibato left12:03
→ mark_vh joined12:03
→ i42n joined12:04
→ rurufufuss joined12:04
→ rjgonza joined12:05
→ juvenal joined12:09
← athit left12:09
→ Nibato joined12:09
← nightcrawler786 left12:10
→ darkestkhan-work joined12:10
→ pandeiro joined12:10
→ athit joined12:11
gitinfo set mode: +v12:11
← iori left12:11
darkestkhan-work is it possible to git-svn dcommit from repository w/o its entire history? (i.e. from snapshot)12:12
→ xxtjaxx joined12:12
→ hyperair joined12:13
← dfr|sf left12:13
→ LekeFly joined12:14
← banghouseFOOD left12:15
→ banghouse joined12:15
← ehsan left12:16
→ tinti joined12:16
→ Tjoppen joined12:17
← i42n left12:18
→ freeed joined12:19
→ zhouxiukun joined12:19
→ codewrangler joined12:20
→ i42n joined12:20
→ jcao219 joined12:21
selckin what's snapshot?12:22
you can clone from a certain rev, and dcommit fine12:22
csmrfx a snapshot contains all the current / given state, I think12:22
not just the changes from previous step, but all data12:23
→ MWinther_ joined12:25
→ gun-lpt joined12:25
→ ascii` joined12:25
→ darwinbits joined12:26
ascii` hi, i did a git add . before editing .gitignore and now i'm asked to add thing like "# modified: .bash_history", how do i remove them before pushing?12:26
canton7 ascii`, .gitignore doesn't apply to files which are already tracked12:27
← jaseemabid left12:27
ascii` canton7, exactly, how do i untrack them?12:27
canton7 you'll probably want to stop tracking those files altogether? git rm --cached <file>12:27
← darwinbits left12:27
SeoZ → SeoZ-work[AWAY]12:27
← MWinther left12:27
MWinther_ → MWinther12:27
ascii` canton7, thanks! trying now12:28
→ pppaul joined12:30
← pppaul left12:30
→ ipalaus joined12:30
→ pppaul joined12:30
← pppaul left12:30
→ pppaul joined12:31
← pppaul left12:31
darkestkhan-work selckin: think about situation where you don't have entire history of repository but you have most recent version and you want to dcommit... is it possible?12:31
← geekbri left12:31
→ geekbri joined12:31
→ pppaul joined12:31
← pppaul left12:31
→ pppaul joined12:32
← pppaul left12:32
crocket If I invoke git-push on windows, what would be the resulting permissions of the remote repository?12:32
→ pppaul joined12:32
crocket Last time I did it, it removed read and execute permissions from others.12:32
← pppaul left12:32
darkestkhan-work I think 77712:32
→ ffunenga joined12:33
→ pppaul joined12:33
← pppaul left12:33
nevyn crocket: depends.12:33
canton7 crocket, git only tracks the execute bit12:33
nevyn what are you pushing to?12:33
→ pppaul joined12:33
← pppaul left12:33
crocket nevyn : a debian server.12:34
ascii` canton7, thanks works perfectly!12:34
cmn push doesn't affect the permissions of the remote repo12:34
→ pppaul joined12:34
← pppaul left12:34
canton7 ascii`, np12:34
nevyn so the files will be whatever the umask of the remote user is.12:34
crocket I executed "git archive | tar xf -" and synced the files with an FTP server.12:34
And the files on FTP server didn't have read permissions on files.12:34
→ pppaul joined12:34
nevyn unless you use a git server like !gitolite12:34
← pppaul left12:34
gitinfo Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html12:34
crocket read permission for others12:34
selckin darkestkhan-work: it sounds completely impossible to get into that store to me12:35
crocket nevyn, I use gitolite.12:35
→ nwest joined12:35
nevyn crocket: please explain the end to end dataflow.12:35
→ pppaul joined12:35
← pppaul left12:35
cmn crocket: and how is that related to git?12:35
nevyn you're confusing me.12:35
I thought you said you ran git push?12:35
→ pppaul joined12:35
← pppaul left12:35
crocket windows git repository$ git push debian12:35
← nigelb left12:36
→ pppaul joined12:36
← pppaul left12:36
crocket On debian, the repo has post-receive that executes "git archive master | tar xf -" on a directory and copies it to an FTP server with lftp.12:36
lftp executes "mirror -R local remote"12:36
→ pppaul joined12:37
← pppaul left12:37
→ iamjarvo joined12:37
→ pppaul joined12:37
nevyn ok.12:37
← pppaul left12:37
nevyn git doesn't track permissions.12:37
cmn and that is related to permission on the repo how?12:37
→ pppaul joined12:38
← pppaul left12:38
→ nigelb joined12:38
nevyn whe n you say permissions on the repo.12:38
crocket Maybe I should use "git checkout -f" instead.12:38
nevyn what files are you refering to?12:38
← bartek left12:38
→ Desproges joined12:38
crocket every file.12:38
→ pppaul joined12:38
← pppaul left12:38
ffunenga Hello everyone. I'm using ubuntu 12.04 and I'm new to git. I have some questions: To make a commit I use the command 'git commit -a -m "msg"'. How do I execute a redo action? Are there any guis for ubuntu that make this? I've installed Git-Cola and gitg and I was not able to understand how to do it...12:38
nevyn because I think that means something different to what you're saying12:38
so.12:38
canton7 crocket, did you pay attention to when we said that git only tracks the execute bit, and the rest depend on the umask of the user doing the checkout?12:39
nevyn permissions on the repo == the permissions of the files in the bare repo on the server.12:39
→ pppaul joined12:39
← pppaul left12:39
canton7 ffunenga, !book12:39
gitinfo ffunenga: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable12:39
nevyn redo?12:39
→ glennpratt joined12:40
nevyn ffunenga: what do you mean by a "redo action"12:40
→ Twey joined12:40
ffunenga nevyn: I mean undo12:40
nevyn oh that's like the total opposite.12:40
→ bartek joined12:40
→ pppaul joined12:40
← pppaul left12:40
crocket nevyn, ok12:40
nevyn ffunenga: there are multiple ways to undo with git. depending on what you want to happen.12:40
crocket: not the files you'd edit or deploy12:40
→ pppaul joined12:41
← pppaul left12:41
crocket nevyn, So the permissions of the bare repo files are determined by umask of the machine that holds the bare repo?12:41
nevyn nope12:41
← teotwaki left12:41
→ pppaul joined12:41
nevyn those are managed by git.12:41
← pppaul left12:41
crocket nevyn, how does git determine permissions?12:41
ffunenga nevyn: I don't understand what you mean. After a commit -a -m (...) how do I undo the last commit?12:41
nevyn but when I say the bare repo files I don't mean the things you edit.12:42
canton7 ffunenga, you really need to read a good git resource before trying to dive in too far12:42
→ pppaul joined12:42
← pppaul left12:42
nevyn ffunenga: so do you want it to never have existed. or do you want to reverse it?12:42
← cakehero left12:42
Twey Hi. I have a centralized remote repository whose working tree represents a ‘staging’ server — a testing server for demonstrating for clients. The idea is that we should be able to push our work to the staging server & have it accessible by the client (a directory in the working tree is accessible by a webserver). However, when I try to push, I get the ‘won't push onto a checked-out tree’ message. The odd thing is that even when I git config…12:42
… receive.denyCurrentBranch ignore on the remote server (I have a receive hook), git still won't let me push, & gives me the same message.12:42
nevyn Twey: !deploy12:43
gitinfo Twey: Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html12:43
→ jola joined12:43
nevyn Twey: what's in your recieve hook?12:43
→ pppaul joined12:43
← pppaul left12:43
canton7 nevyn, doesn't sound like he has one: he's just pushing to a non-bare repo12:43
← rchavik left12:43
→ pppaul joined12:44
← pppaul left12:44
nevyn canton7: 2:50 < Twey> … receive.denyCurrentBranch ignore on the remote server (I have a receive hook), git still won't let me push, & gives me the same message.12:44
see the bit in brackets ;)12:44
→ pppaul joined12:44
canton7 ah yes, you're right12:45
Twey, if you try and push to the checked-out branch of a non-bare repo, git won't update the checked-out files. this leads to lots of confusion, so the operation is disabled by default12:45
nevyn Twey: poke.12:45
crocket Does "git diff" track only changes in execute permission?12:45
bremner wat?12:46
nevyn git doesn't care about permissions AT ALL12:46
→ mgaunard joined12:46
bremner except execute bit12:46
canton7 crocket, yes, git only tracks the execute bit12:46
mgaunard hi, how can I update a branch where the remote has been rebased? I don't care about my local history12:46
canton7 mgaunard, no local work you want to save? git reset --hard remote/branch12:46
← i42n left12:47
Twey nevyn: git reset --hard12:47
mgaunard canton7: thanks12:47
nevyn Twey: so on the remote. git config recieve.denyCurrentBranch ignore12:47
is that syntax right?12:48
Twey nevyn: I did, but I still get the message12:48
nevyn Twey: man git-config12:48
gitinfo Twey: the git-config manpage is available at http://git-scm.com/docs/git-config.html12:48
Twey I think so. It now shows it in my git config -l.12:48
→ i42n joined12:48
crocket I think file permissions are set according to OS rules if git ignores permissions on bare repos.12:49
→ lun_ joined12:49
nevyn crocket: yay. you've finally got to the water ;)12:49
taste good does it?12:49
crocket nevyn, the water tastes juicy.12:49
nevyn excellent12:49
← ffunenga left12:50
nevyn crocket: so. if you care about permissions of files you may need a real deployment tool12:50
crocket I'll have to debug on my side tomorrow.12:50
nevyn like cogit12:50
not cogit what's it called?12:50
canton7 crocket, can you not see the many messages telling you exactly that? :P12:50
← hyperair left12:51
crocket nevyn, read permission was removed on files for others, so I couldn't access anything on the website.12:51
← SkAzZ left12:51
masak given a repository, is there a way to create a my-repository.git file for upload to some HTTP server for later cloning?12:52
crocket Where is cogit?12:53
nevyn umm..12:53
→ SkAzZ joined12:53
nevyn capistrano.12:53
was what I was thinking of.12:53
cmn masak: man git bundle12:53
gitinfo masak: the git-bundle manpage is available at http://git-scm.com/docs/git-bundle.html12:53
nevyn !capistrano12:53
→ SkAzZ1 joined12:53
nevyn no factoid for capistrano BAD BOT12:54
crocket cogito?12:54
masak cmn: thanks! I already had my eye on git-bundle. good to know that is it.12:54
nevyn crocket: http://capistranorb.com/12:54
← jola left12:54
nevyn gitinfo: capistrano is a application deployment tool written in ruby12:55
!capistrano12:55
!factoid12:55
→ acekim joined12:55
gitinfo [!triggers] Please don't spam me! I'm just a poor bot! Here's everything I know: http://jk.gs/git/bot/trigger.php12:56
cmn there's a bit of magic, something like12:56
← bartek left12:56
cmn .learn capistrano is an application deployment tool in ruby, which has support for git12:56
maybe not that magic12:56
→ kvanderw joined12:57
jast it's .trigger_edit <name> <output>12:57
this is all documented btw :}12:57
cmn right, learn is for another bot12:57
.trigger_edit capistrano capistrano is an application deployment tool in ruby, which has support for git12:57
gitinfo cmn: Okay.12:57
jast might wanna include a link or something12:58
← jargon- left12:58
← SkAzZ left12:58
crocket Does anyone use post-receive to deploy?12:59
csmrfx Sounds like somewhat workable idea13:00
jast !deploy13:00
gitinfo Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html13:00
darkestkhan-work selckin: rm -rf .git ; git init ; git add . ; git ci -m "initial commit"13:01
cmn .trigger_edit capistrano capistrano is an application deployment tool in ruby, which has support for git, see http://capistranorb.com for more information13:01
gitinfo cmn: Okay.13:01
selckin darkestkhan-work: if you expect that to work there is no hope for you13:02
← ascii` left13:02
jast as usual, the capistrano website is missing a crucial section13:02
cmn crocket: you don't use post-receive to deploy, you use it to trigger a deployment with whatever tool fits your needs13:02
jast "what is capistrano?"13:02
also fairly useful yet missing: "why would you want to use capistrano?"13:02
csmrfx selckin: please elaborate13:02
cmn it's an island with a beach, didn't you see the picture at the top?13:02
jast cmn: ohhh. my mistake.13:02
so it has something to do with fishing?13:03
← kcj left13:03
cmn that's my best guess from the site13:04
← und3f left13:04
cmn you can do it on rails, which is handy, I've never gone fishing on a train13:04
sitaram jast: yeah but it's *ruby* -- why would you NOT want to use it?13:05
csmrfx I propose a new topic: make jokes about pr of open source ruby projects13:05
sitaram runs and hides13:05
jast sitaram: excellent point13:05
but let's be honest here13:05
→ und3f joined13:05
jast anything that isn't written in all of: python, ruby, haskell, erlang, java... can't possibly be any good13:06
also, http://linuxhaters.blogspot.com/13:06
csmrfx java?13:06
→ macabre joined13:06
csmrfx geebus13:06
jast java: slow is good13:06
sitaram cmn: surprisingly even the "getting started" link doesn't say anything, and the "from the beginning"" link is only marginally better13:06
jast slow is still competing with overly verbose for top spot, though13:07
cmn I think their README is better13:07
jast ohh13:07
the page title explains what capistrano is13:07
sort of13:08
← Stereo left13:08
cmn the title on capistranorb.com you mean?13:08
← afuentes left13:08
→ luc_ joined13:08
→ Zenmaker_ joined13:08
jast yes13:08
cmn so it does; I don't have the top bar on my windows, so I never see those things13:09
→ tmcmahon joined13:09
cmn so it should probably link to the github page, which is a lot better at saying what it is and does13:09
.trigger_edit capistrano capistrano is an application deployment tool in ruby, which has support for git, see http://github.com/capistrano/capistrano for more information13:10
gitinfo cmn: Okay.13:10
← Desproges left13:10
→ jargon- joined13:10
→ hyperair joined13:11
→ Aleric joined13:11
→ Desproges joined13:11
Aleric Hi. How can I change the default branch to push to?13:11
cmn set push.default to upstream and set your branch's upstream13:12
← Zenmaker__ left13:12
jast cmn: yeah, much better page13:13
masak hm, git-clone doesn't expect a bundle on the server, does it? it seems to expect a .git directory.13:14
← Desproges left13:14
→ ehsan joined13:14
← rurufufuss left13:15
jast masak: correct13:15
→ fireh joined13:15
cmn you download a bundle to your local disc, that's what they're for13:16
masak so... the most expedient way to provide a git repository from a server is to just upload the .git directory as-is?13:16
cmn otherwise you'd just use git clone against the remote repo, wouldn't you?13:16
do it with push, but yeah13:16
masak ah, ok.13:16
thank you.13:16
← pretty_function left13:16
Aleric cmn: How can I do that? :/13:17
cmn with git config and git branch --set-upstream13:17
sitaram masak: the main purpose of a bundle file seems to be to allow a restartable, single-file download to seed large repo clones. Whether this was the original intent or not I can't say though :)13:17
canton7 Aleric, read push.default in man git-config to understand what that key does. then read about what branch --set-upstream and push -u do13:18
gitinfo Aleric: the git-config manpage is available at http://git-scm.com/docs/git-config.html13:18
sitaram masak: because 'git clone' is not resumable13:18
crocket Is git-deploy https://github.com/mislav/git-deploy ?13:18
masak sitaram: oh!13:18
← chimay left13:18
csmrfx crocket: no13:18
crocket csmrfx, tell me.13:18
sitaram so people on low/bad connections suffer when trying to clone a large repo; this helps13:18
cmn sitaram: not just clone priming, you can make bundles that are basically what you'd send over on a git-push13:18
← catphish_ left13:19
sitaram right; for offline repos13:19
→ jceb joined13:19
csmrfx crocket: https://github.com/avar/git-deploy13:19
sitaram true13:19
csmrfx crocket: afaik13:19
csmrfx should do more homework on git-deploy to see if it would work for him13:19
← bizhanMona left13:20
crocket Isn't https://github.com/git-deploy/git-deploy the real thing?13:20
→ jaseemabid joined13:20
← jaseemabid left13:20
csmrfx the one and.. the same13:20
cmn meaning what?13:20
→ AaronMT joined13:21
→ erichynds joined13:21
crocket It seems that umask is 007 in git.13:21
→ Ecesis joined13:21
crocket I mean post-receive13:21
post-receive is on a secret mission to kill me.13:21
→ bizhanMona joined13:22
← macabre left13:22
← manolo left13:22
→ macabre joined13:22
← Ecesis left13:23
← osmosis left13:23
crocket umask is set to 027 in post-receive. I confirmed it with umask.13:23
→ duckxx joined13:23
→ jackh joined13:24
← sdfgsdjf left13:24
→ etcetera joined13:24
← Gitzilla left13:24
→ piti joined13:24
crocket umask is 027 in post-receive hook.13:24
Why is it so?13:24
Did I fool myself?13:25
cmn if you want a particular umask, set it13:25
Aleric Sorry... I don't seem to understand this :(. .. The "problem" is that when I do 'git checkout master' I get the message "Your branch is ahead of 'siana/master' by 39 commits.", but that is an old repository, it should read 'singu/master'. How can I change that?13:26
crocket cmn : Do you know how git sets umasK?13:26
cmn with 2 umask13:26
← trungqtran left13:26
cmn that's how you set the umask from a program13:26
← ehsan left13:27
cmn Aleric: man git branch --set-upstream13:27
gitinfo Aleric: the git-branch manpage is available at http://git-scm.com/docs/git-branch.html13:27
→ ddelrio1986 joined13:27
crocket cmn : Do you know where git sets umask?13:28
cmn I don't know every piece of code off by heart13:28
→ les_sylvains joined13:28
cmn and it's irreleveant, if you need a particular umask, set it13:28
jargon- damnit. the git-remote-hg isn't working as advertised in the readme. http://pypi.python.org/pypi/git-remote-hg/0.1.1 anybody know how i can clone a remote mercurial repo into my local git repo, without my needing to install hg?13:28
→ bartek joined13:29
crocket cmn : I think gitolite set it to 027.13:29
cmn : .gitolite.rc has UMASK 027.13:29
→ tplaner_ joined13:29
cmn that seems a reasonable assumption13:29
canton7 Aleric, git can keep a bit of information about a branch, saying which remote branch that branch is related to, called the branch's 'upstream'. To generate that message, git uses this information to decide which branch to compare your current branch to. This information is changed with branch --set-upstream13:30
Aleric http://codepad.org/vpjxwSdm13:30
What am I doing wrong?13:30
← pppaul left13:31
→ paul_andrew joined13:32
Aleric It still says siana/master13:32
← ConstantineXVI left13:32
→ chx joined13:32
→ ConstantineXVI joined13:32
← ConstantineXVI left13:32
chx how can i clone a repository locally keeping the remotes?13:32
Bombe chx, you can’t. Information about remotes is not cloned.13:32
chx i see that but is there a way to say list all the remotes and recreate them?13:33
Aleric canton7: Can you look at http://codepad.org/vpjxwSdm please, and tell me what I did wrong?13:34
jast chx: yes... write a script13:34
chx haha13:34
certainly13:34
jast chances are you can just copy the .git/config file13:34
← tmcmahon left13:35
chx oh13:35
→ LmAt joined13:35
jast Aleric: git branch --set-upstream localbranchname remote/branchname13:35
chx jast: that's what i was looking for, very awesome13:35
→ osmosis joined13:35
LmAt → face2face13:36
jast Aleric: also note that you can run 'git checkout' without args to get the same line about up-to-date-ness13:36
← kermit left13:36
Aleric >git branch --set-upstream master singu/master13:36
warning: refname 'singu/master' is ambiguous.13:36
fatal: Ambiguous object name: 'singu/master'.13:36
:/13:36
← face2face left13:36
→ LmAt joined13:36
Aleric Is that because of my previous command?13:36
→ Davey joined13:36
jast yeah, because your failed attempt added wrong stuff13:36
LmAt → face2face13:37
→ jaseemabid joined13:37
jast try this: git branch -d singu/master; git config --remove-section branch.singu/master13:37
Aleric :(13:37
ok13:37
jast then try again13:37
Aleric error: The branch 'singu/master' is not fully merged.13:38
If you are sure you want to delete it, run 'git branch -D singu/master'.13:38
→ hellopat joined13:38
jast do what it says ;)13:38
Twey nevyn: So, do you know why my config setting was ignored?13:38
Aleric it won't delete my master right?13:38
← mark_vh left13:38
jast Aleric: it will delete the local branch 'singu/master' which you created by accident13:38
Aleric >git config --remove-section branch.singu/master13:39
fatal: No such section!13:39
← chx left13:39
← ishi left13:39
cmn that's because the branch -d deleted it13:39
Aleric ok..13:40
jast oh, right13:40
so just move on to the next step13:40
→ kermit joined13:40
→ martinjlowm joined13:42
→ afuentes joined13:42
→ pppaul joined13:43
← pppaul left13:43
→ pppaul joined13:43
← pppaul left13:43
Aleric Thanks... Now, what I really trying to do :p... How can I push a new branch to upstream? I thought it was: git push -f <remote>, but that doesnt' seem to work13:44
→ pppaul joined13:44
← pppaul left13:44
→ pppaul joined13:45
← pppaul left13:45
→ jmd joined13:45
→ pppaul joined13:45
← pppaul left13:45
cbreak-work Aleric: it's git push remotename branchname13:45
no -f required13:45
jmd Is there an easy way to git a list of commits which are on branchA but not on branchB ?13:45
cbreak-work Aleric: how do you think git knows which branch to push if you don't tell it?13:45
jmd: yes.13:46
→ pppaul joined13:46
← pppaul left13:46
jmd cbreak-work: Namely?13:46
canton7 jdm, !dots13:46
gitinfo In the log family (git-log, gitk, etc.) A..B means "everything in B but not in A" [formally: ^A B] and A...B means "everything in A or B but not in both" [formally: A B --not $(git merge-base --all A B)]. An empty "side" of the dots implies HEAD, so 'git log master..' is very different from 'git log master'!13:46
cbreak-work git rev-list branchA ^branchB13:46
→ pppaul joined13:46
← pppaul left13:46
Aleric cbreak-work: I assumed it always works on the one you have checked out :/13:46
→ pppaul joined13:47
← pppaul left13:47
→ harish joined13:47
cbreak-work Aleric: nope, doesn't.13:47
jmd cbreak-work: Thanks.13:47
cbreak-work pushing is a bit convoluted13:47
→ pppaul joined13:47
← pppaul left13:47
skorgon $ 50 :O13:48
sorry, wrong #13:48
→ pppaul joined13:48
← pppaul left13:48
cbreak-work skorgon: #git: priceless13:48
→ pppaul joined13:48
← pppaul left13:48
→ shreya joined13:48
→ pppaul joined13:49
→ itatitat joined13:49
← pppaul left13:49
→ pppaul joined13:49
← pppaul left13:49
→ pppaul joined13:50
← pppaul left13:50
cmn Aleric: see man git config push.default13:50
gitinfo Aleric: the git-config manpage is available at http://git-scm.com/docs/git-config.html13:50
→ pppaul joined13:51
← pppaul left13:51
← hyperair left13:51
→ pppaul joined13:51
← pppaul left13:51
→ pppaul joined13:52
← pppaul left13:52
← cooloney left13:52
← und3f left13:52
→ pppaul joined13:52
← pppaul left13:52
→ hyperair joined13:53
→ pppaul joined13:53
← pppaul left13:53
→ ehsan joined13:53
← r126t_ left13:53
→ pppaul joined13:53
← pppaul left13:53
← prettyrobots left13:54
→ pppaul joined13:54
← pppaul left13:54
jargon- how do u find out the total size of a remote repo?13:54
bigmeow anybody here who have used bitbucket?13:54
← itatitat left13:54
→ pppaul joined13:54
Aleric bigmeow: a little bit13:55
lb bigmeow: yes13:55
bigmeow is gmail supported byt bitbucket?13:55
if yes, why i cannot receive confirmation email?13:55
→ cakehero_ joined13:55
→ cakehero joined13:55
bigmeow Aleric: bitbucket seems to be more friendly than github:)13:55
→ trungqtran joined13:55
bigmeow lb: do you register with gmail>?13:56
lb no13:56
bigmeow lb: which kind of mail do you use?13:56
tell me :)13:56
← face2face left13:57
lb i'll never share my private stuff ... go away you spy :D (for this bb account i used live.com)13:57
← cakehero_ left13:57
← athit left13:58
lb but i use openid for login13:58
→ athit joined13:59
← soee left13:59
← Pugnax left13:59
← lng left14:00
← flijten left14:00
→ neurodrone joined14:00
→ codeM0nK_ joined14:02
← ph^ left14:02
Twey So, I'm now trying to use ‘git checkout’ in my receive hook as a simple deployment method. However, when I push, it says ‘remote: already in 'master'’ and doesn't update the files.14:02
(git checkout -f master, that is)14:02
← swair left14:02
← erichynds left14:02
← codeM0nK3Y left14:02
codeM0nK_ → codeM0nK3Y14:02
→ razdaz joined14:03
cmn Twey: see !deploy and do what it says14:03
gitinfo Twey: Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html14:03
Twey cmn: Yes, that's why I'm using ‘git checkout’ now14:03
But it's not working as the page advertises, unless I misread the page14:03
← zhouxiukun left14:03
cmn then it doesn't sound like you're using -f14:04
Twey cmn: But I am :þ14:04
cmn then the error message doesn't make sense, double-check everything14:04
jast what version of git is running on the server?14:05
Twey cmn: http://hpaste.org/7433814:05
jast: 1.7.114:05
→ chimay joined14:05
→ alexisabril joined14:05
← SkAzZ1 left14:05
← beautifulmind left14:06
← chimay left14:06
→ chimay joined14:06
bigmeow lb: dude, hell is in front of you, step forward:)14:06
jast that's definitely recent enough14:06
bigmeow jast: dude, do you use bitbucket?14:06
→ authmillenon joined14:08
Twey When I run ‘git checkout -f master’ on the remote I just get ‘Already on 'master'’14:08
‘git checkout -f’ without specifying a branch does what I expect14:08
‘git branch’ shows only one branch, named ‘master’14:09
← tureba left14:10
→ dzonder joined14:13
sitaram Twey: "already on master" is not an error; at least it doesn't seem to set the exit code to non-014:15
Twey It's not an error, it just doesn't do anything14:15
sitaram Twey: and you're missing "unset GIT_DIR"14:15
oh wait14:15
you're missing GIT_WORK_TREE=/deploy/dir14:16
Twey Sorry, I added it later14:16
← _jesse_ left14:16
← duckxx left14:16
← Zenopus left14:16
sitaram Twey: well, all I can say is if this doesn't work, gitolite has been doing something very mysterious for the last 3 years ;-)14:16
'cos that's what I use within gitolite14:17
→ dhruvasagar joined14:17
Twey I've made a change locally, added, committed14:17
← luc_ left14:17
← ipalaus left14:17
→ TeckniX joined14:18
Twey Pushed; got ‘already on master’; GIT_WORKING_TREE seems to be what I want it to be (I echo'd it)14:18
← athit left14:18
Twey Files haven't been updated14:18
→ athit joined14:18
→ dir joined14:19
Twey Running ‘git checkout -f master’ manually on the server causes the changes to happen.14:19
→ chuckharmston joined14:19
sitaram well that's why there's that line in there: you don't talk about it or ask questions about it :)14:19
pdelvo|afk → pdelvo14:20
Twey T_T14:20
sitaram has to go have dinner; can't help right now14:20
→ sbroffo joined14:20
← hyperair left14:24
← sbroffo left14:24
→ anish[1] joined14:24
Twey sitaram: 'kay ☹14:24
← paul_andrew left14:25
← jackyyll left14:25
← Enlik left14:26
← razdaz left14:26
← darkestkhan-work left14:26
→ torjus_ joined14:27
→ Cache_Money joined14:27
→ jackyyll joined14:27
→ hyperair joined14:27
→ paul_andrew joined14:27
→ canton8 joined14:28
← jmd left14:28
jast bigmeow: no14:30
← thirdknife_ left14:30
← gavin_huang left14:30
← canton7 left14:30
→ Enlik joined14:31
→ crashsky joined14:31
→ fly9 joined14:31
← stodan left14:31
crocket What's good about git-deploy?14:32
It seems bloated.14:32
→ gusnan joined14:32
← acekim left14:32
→ ph^ joined14:32
cmn don't use it then14:32
← jn__ left14:34
cmn deployment is a complex field where a lot can go wrong, you need to account for all of those and figure out how to recover14:34
← karadoc left14:34
crocket cmn : Do you use it?14:36
← kpreid left14:36
← MWinther left14:36
→ kpreid joined14:36
cmn no, I've mentioned already that I oursource any deployment I need to make14:36
crocket cmn : Isn't git-deploy a form of outsourcing?14:37
Which deployment tool do you use?14:37
→ macBigRig joined14:38
→ heroux joined14:38
_ikke_ outsourcing?14:39
bremner he just said he uses make14:40
crocket bremner, "GNU make"?14:40
bremner sure14:40
← iamjarvo left14:40
crocket cmn : I thought "make" was a verb.14:40
bremner I thought you were a linux admin ;)14:40
crocket bremner, Who am I/14:40
sitaram Twey: I'm back. But truthfully, it's not easy to debug this but I'd start with "pwd >&2; env | grep GIT >&2" just before the "git checkout"14:41
← neurodrone left14:41
← mritz left14:42
crocket sitaram, I've had good time with gitolite.14:42
cmn crocket: make is a verb, yes, and I used ti as such14:43
→ ottoshmidt joined14:43
crocket cmn : I want to know what you use to deploy.14:43
cmn I outsource it14:44
I don't do anything14:44
crocket hmm...14:44
ok14:44
You let others do it.14:44
cmn I get others to do it, yes14:44
crocket clever.14:44
cmn : Currently, post-receive syncs the master branch with FTP server.14:45
It turns out to be slow.14:45
cmn it would be14:45
crocket I'm not sure if deployment should be done on every commit on master.14:45
← shreya left14:45
→ mritz joined14:45
ottoshmidt gitweb lists my repos but no summary or log or whatever else info is available14:45
← bitslip_ left14:45
ottoshmidt I get 404 - Unknown commit object type of things14:45
I have setup gitweb for the first time so I am newbie in this14:46
→ bitslip joined14:46
← jceb left14:46
← ThomasLocke left14:48
← Milossh left14:49
← koltroll left14:49
← lb left14:50
← jackh left14:50
Twey sitaram: Might the issue be that I'm not using a bare repository?14:50
→ Desproges joined14:50
Twey My work tree is a user's homedir, with a lot of files & directories, many of which the user doesn't own & can't manipulate (Plesk; don't ask)14:51
cmn crocket: that is something you and your team needs to decide14:51
→ iamjarvo joined14:51
sitaram Twey: shouldn't. The GIT_WORK_TREE in gitolite is not even a repo, bare or not14:51
→ monsieurlu joined14:52
← paul_andrew left14:52
→ v0n joined14:53
→ tigris_ joined14:54
→ brainbox1100 joined14:54
← brainbox1100 left14:54
→ scarlisle joined14:54
crocket !deploy14:55
gitinfo Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html14:55
→ shreya joined14:55
→ rmr joined14:56
DeadZen the fight club ;p14:56
→ neurodrone joined14:57
→ Milossh joined14:58
→ pdtpatrick joined14:59
tigris_ is there another command like git-cat-file -p SHA1 that can show me the info a git log --stat shows for a commit? e.g. the blob changes14:59
→ znake joined15:00
cbreak-work tigris_: git show shows that.15:01
git cat-file isn't supposed to show any diffs15:01
→ thiago joined15:02
→ zhouxiukun joined15:02
Twey sitaram: Gah, I think it was just a sequence of stupid errors15:02
Forgot to add here, forgot to chmod +x there, edited the wrong file over there…15:03
→ rath_ joined15:03
Twey It seems to be working now *crosses fingers*. Thanks for your very useful page!15:03
→ goshawk joined15:03
← Cache_Money left15:04
← goshawk left15:04
← znake left15:04
→ Jay_Levitt joined15:04
→ znake joined15:04
cmn cat-file can't show changes, it works on single objects15:04
tigris_: what are you actually trying to get?15:05
← znake left15:05
← shreya left15:06
← gun-lpt left15:06
→ jceb joined15:07
tigris_ cmn: i was trying to figure out what files changed in a particular commit, using low level plumbing tools, but i just realised dealing with things like empty commits is going to screw me anyway, so i am gonna have to think of something new15:07
avar crocket, csmrfx it's git-deploy/git-deploy15:07
crocket avar : ???15:08
← iamjarvo left15:08
cmn tigris_: that's what diff-tree is for15:08
avar crocket: wrt previous discussion about git-deploy15:08
crocket avar, Do you use it?15:09
cmn but you'd be better off using a real programming language rather than bash to achieve whatever it is15:09
→ duckxx joined15:09
← ddelrio1986 left15:09
← zivester left15:10
→ revagomes joined15:10
→ nizeguy joined15:10
← chimay left15:10
← Moussekateer left15:10
→ chimay joined15:11
← chimay left15:11
→ chimay joined15:11
jast cmn: you mean zsh? :^D15:11
→ zivester joined15:11
→ Zenopus joined15:11
cmn obviously :)15:11
tigris_ cmn: i'm using ruby, with the rugged gem, built on the libgit2 c bindings, so wherever i can avoid shelling out to things like git-log is much faster for me, but i'm finding there are just times it's easier to use git itself15:11
→ jason237 joined15:11
cmn rugged unfortunately doesn't have diff support yet (though you can grab some rudimentary one from a PR)15:11
it might seem easier to shell out to git, but as soon as something unexpected happens, you're going to crash and burn15:12
→ Synthead joined15:12
tigris_ i wrote a very basic one just using the Rugged::Walker class, capturing a list of blobs that changed in each commit compared to previous commit, but dealing with empty commits is proving an issue... i'll think of something15:13
cmn you just have to check if they're treesame15:13
which should just be next if last.tree_id == current.tree_if15:14
← cbreak-work left15:14
→ zatan joined15:15
→ jola joined15:15
zatan hey i have dont git reset --hard sad2s its reseted but when I try to push something its telling me to pull first ??15:15
← pppaul left15:15
→ kukks joined15:15
cmn yes15:15
because you've changed that branch's history, so it no longer build it up, but changes it15:16
tigris_ cmn: yep, you are right... empty commits the tree_oid matches, that should help me a bit15:16
← psoo left15:17
→ pppaul joined15:18
→ babilen joined15:20
→ goshawk joined15:21
← canton8 left15:21
→ canton7 joined15:23
canton7 zatan, !rewrite15:23
gitinfo zatan: [!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)15:23
→ SirCmpwn joined15:24
← stanzgy left15:24
← aspiers left15:25
← goshawk left15:26
→ jakob-_ joined15:27
← etcetera left15:28
← jakob-_ left15:28
→ _iron joined15:29
→ aspiers joined15:29
← blami_ left15:29
a-l-e hi... at scribus (http://scribus.net) we have have started using a bit of git on top of svn...15:30
← osxorgate left15:30
← hyperair left15:30
a-l-e ... for now we can't move away from svn for the main development... but we will try to move to git after the next release.15:31
← pppaul left15:32
a-l-e the consequence is that the current situation is a bit complicated and nobody really has the skills to manage it: we are a bit going through tries and errors...15:32
→ maestrojed joined15:32
a-l-e and, now, i could need some help :-)15:32
← fread2281 left15:32
a-l-e we have two repositories: trunk.git and scribus.git.15:32
→ catphish_ joined15:33
a-l-e in trunk.git the only interesting branch is "svn" which is syncronized with the svn server.15:33
catphish_ is there any way to reduce the memory used by git gc?15:33
→ proradii joined15:33
→ pppaul joined15:34
a-l-e our git admin has created a scribus.git/master branch which i could successfully update with a "pull trunk svn".15:34
the newes files are in there.15:34
→ etcetera joined15:34
→ kenperkins joined15:35
a-l-e if i do the same with i have created my self and which should also based on trunk/svn (at this point i'm not sure of nothing anymore...)...15:35
i don't get the newest commits...15:35
can anybody help me find out why in this second case "git pull trunk svn" (in reality git "pull svn svn" since i've called the trunk repository "svn"...)15:36
does not pull the latest changes?15:36
→ tjholowaychuk joined15:36
a-l-e first: how can i be sure that the "indic" branch is indeed based on "trunk/svn"?15:37
(the "scribus.git/indic" branch...)15:37
← cakehero left15:37
← pdtpatrick left15:38
← pfFredd left15:38
a-l-e if i do "git remote show origin"15:38
← _huoxito left15:38
→ goshawk joined15:38
a-l-e i get: http://pastebin.com/6CjXL2Q815:39
→ adityamenon joined15:39
← gerryvdm_mbp left15:39
a-l-e why is the the "indic" branch missing in the list "Local branches configured for 'git pull'"?15:39
what did do wrong?15:39
← skorgon left15:40
a-l-e (or better: what could be wrong?)15:40
→ brainbox1100 joined15:40
→ am1ja2VuemllCg joined15:40
cmn is your local indic branch set up with any upstream information?15:40
a-l-e cmn: how can i tell?15:40
cmn does it say it's tracking anything when you use git branch -vv?15:40
a-l-e t$ git branch -vv15:41
← dhruvasagar left15:41
a-l-e indic af8ed4f [svn/svn: ahead 9] Merge branch 'indic' of scribus.net:scribus into indi15:41
cmn that's the reason then15:41
it's not trackign anything from 'origin'15:42
catphish_ is there any way of passing options to git-repack when it's executed by git-gc?15:43
← goshawk left15:43
← ottoshmidt left15:43
a-l-e this is the full list of the branches:15:43
http://pastebin.com/82Dmg2cc15:44
→ nightcrawler786 joined15:44
→ jwmann joined15:44
cmn and that is is for what?15:44
→ ShadeTornado joined15:45
a-l-e should i do a "git diff indic..master" and then recreate the indic branch based on master and apply the patch?15:45
or is there another way to "relink" the branch?15:46
cmn man git rebase see --onto (and be careful, the syntax is tricky)15:46
gitinfo the git-rebase manpage is available at http://git-scm.com/docs/git-rebase.html15:46
canton7 but the examples help a lot15:46
a-l-e well, is there anything without a tricky syntax where i have to be careful? :-)15:46
cmn that you choose the right commits15:47
the example with the graph is very helpful15:47
make sure you map that into your needs, then it's just replacing the arguments15:47
→ fread2281 joined15:48
→ cehteh joined15:48
→ gdelfino joined15:48
← rath_ left15:48
a-l-e i don't understand half of the words in the git-rebase man page...15:49
... a sign that it's not the best way for me :-(15:49
cmn you can do cherry-pick manually15:49
basically the same thing, but more manual work15:49
← shruggar left15:49
bremner a-l-e: check the examples in !book15:52
gitinfo a-l-e: There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable15:52
← trungqtran left15:53
a-l-e i fear that reading that book won't allow me to solve the problem in a usable time frame...15:54
← brainbox1100 left15:54
catphish_ hmm, still can't figure out how to limit memory consumed by git-repack :(15:54
a-l-e in two month i would be able to do what i need...15:54
catphish_ --window-memory=<n> doesn't do what i think it does15:54
(or is ignored)15:54
a-l-e but until then we will have lost the new contributor :-)15:54
← whitman left15:54
a-l-e i love the few git features i'm using... but its complexity is killing me :-(15:55
→ goshawk joined15:55
bremner a-l-e: I did not suggest you read the whole book, although that would probably be a good idea in the long run15:56
a-l-e the whole thing looks like as if my goal would be to learn git...15:56
→ JPohlman1 joined15:56
→ jakob- joined15:56
cmn what is the actual problem? there's a branch that's based on the wrong commit?15:57
a-l-e ... but in reality i just want to be able to work on code and let other problems work on code...15:57
cmn a-l-e: if you want to use git, you'll need to learn how to use it15:57
it's not different from anything else15:57
→ prudnikov joined15:57
a-l-e yeah, it's clear to me that i have to learn how to use it...15:57
← ology left15:57
← fread2281 left15:57
masak what's the appropriate way to resolve a conflict during a 'stash apply'?15:57
→ gregm_ joined15:57
a-l-e but reading a full books seems to be a big first step.15:57
... most of all because it's not the first step...15:58
bremner in the time you have talking about it, you could have found the example15:58
→ dogarrhea2 joined15:58
cmn masak: you make the file look like it should, see man git merge for an explanation of the conflict representation15:58
gitinfo masak: the git-merge manpage is available at http://git-scm.com/docs/git-merge.html15:58
a-l-e i still don't know which example i should be looking from :-(15:58
looking for...15:58
→ trungqtran joined15:58
cmn what is the actual problem?15:58
← am1ja2VuemllCg left15:59
a-l-e the problem is that i've created a branch based on the wrong remote branch15:59
masak cmn: oh, that bit is fine. I'm just confused by git's advice to commit my changes to mark the conflict resolved.15:59
← JPohlman1 left15:59
a-l-e and i should change the branch it's based upon.15:59
cmn masak: you need to use add to tell git that you solved it15:59
a-l-e: then that example in git-rebase is exactly what you're looking for15:59
or use cherry-pick15:59
masak cmn: right. so I just use 'add', and then unstage the change again?16:00
cmn unstage what change?16:00
← zhouxiukun left16:00
a-l-e cmn: i believe you, that it must be the correct tool to use..16:00
→ zhouxiukun joined16:00
a-l-e ... but from the man page i can't tell how to use it!16:00
← tjadc_ left16:01
cmn yes, you can16:01
masak cmn: if I do git-add on the file, my conflict-resolved stashed changes end up in the staging area. I want those changes to remain unstaged.16:01
cmn have you looked at where we told you?16:01
a-l-e most of all because the next tip was to be very careful...16:01
← jakob- left16:01
a-l-e it would be wonderful if somebody could help me with some more step by step hints...16:01
cmn right, so? you should always be careful16:01
→ hyperair joined16:01
cmn the exact command is in that manpage16:01
if you refuse to look at it, there's nothing we can do16:01
a-l-e no, cmn, i'm looking at it!16:01
cmn masak: then yes, I guess16:01
← tgunr left16:02
a-l-e i don't refuse at all to look at it and try to understand...16:02
→ jakob-_ joined16:02
a-l-e ... i'd love to be able to figure out what is in there...16:02
← neurodrone left16:02
a-l-e the problem is that it starts with "DESCRIPTION16:02
If <branch> is specified, git rebase will perform an automatic git checkout <branch> before doing anything else. Otherwise it remains on the current branch.16:02
"16:02
← robotmay left16:02
← kurrataW left16:02
a-l-e which still does not tell me what it does :-(16:02
what it's good for.16:03
← Milossh left16:03
a-l-e (or "Forward-port local commits to the updated upstream head " in the short version... which does not tell me much either"16:03
← mklappstuhl left16:03
tigris_ a-l-e: do you understand what a <branch> is?16:03
← juvenal left16:03
cmn that is absolutely not where we told you to look16:03
a-l-e yep, i know what a branch is.16:03
canton7 a-l-e, read the examples starting at "Here is how you would transplant a topic branch"16:04
up until (but not including) "A range of commits could also be removed with rebase"16:04
a-l-e ok, canton7, doing it...16:04
one problem:16:05
← pppaul left16:05
→ shruggar joined16:05
a-l-e "to pretend that you forked the topic branch from the latter branch, using rebase --onto." ... i want to pretend tha i did "git branch --track indic origin/master"16:06
→ Milossh joined16:06
cmn yes16:06
a-l-e i have not idea if it's the same.16:06
→ gigamonkey joined16:06
canton7 if you're having trouble, just look at the before/after pictures16:06
and find a pair that match your situation16:06
a-l-e but otherwise it looks indeed similar to what i want to do.16:06
→ boombatower joined16:06
wereHamster if you need before/after pictures: http://before-and-after-pictures.tumblr.com/16:07
→ kevwil joined16:07
catphish_ does anyone know why "git repack -d -l -A --window-memory=50m" might consume 5+GB of RAM?16:07
← cenuij left16:08
a-l-e or http://cdn.itproportal.com/photos/nokia-lumia-920-pureview-2-640x725_original.jpg ...16:08
i guess i stick to the ascii art, wereHamster16:08
tigris_ what's the command to get back all the space of a blob file you know you never want back? is it gc, prune or repack or something along those lines?16:08
FauxFaux catphish_: How big is the pack file? I suspect it's unavoidable.16:08
tigris_: man git gc, speficfically the --expire kind of options.16:08
gitinfo tigris_: the git-gc manpage is available at http://git-scm.com/docs/git-gc.html16:08
cmn the window isn't all that affects memory usage16:09
a-l-e git rebase --onto origin/master svn/svn indic16:09
does this look correct?16:09
i want to move the indic branch which i based on svn/svn to be based on origin/master ...16:09
← leokaito left16:09
← goshawk left16:09
← tsdh left16:09
canton7 looks right16:09
→ pppaul joined16:09
→ gfontenot joined16:10
→ pdtpatrick joined16:10
catphish_ FauxFaux: the repository is very large and probably has some very large existing pack files16:10
without --window-memory=50m a "git gc" consumed 12GB of RAM16:10
with it, it's still hitting about the size of the repo (7GB)16:11
a-l-e ok, i guess that this will only change my local repository, and that it will spread to the server when i do a "git push"16:11
tigris_ FauxFaux: thanks, i had read that but wasn't sure on the terminology around "loose objects"... i assumed a `git rm` would leave that blob attached to some commit trees, and therefore they'd never be "loose"16:11
→ djh__ joined16:11
→ TheJH joined16:11
a-l-e so i can duplicate the directory where my local repository is, to get back to the previous state if something goes wrong.16:11
← djh__ left16:11
FauxFaux tigris_: Oh, you want to purge a /committed/ object from history? That's much harder and can't be done without !rewriting. man git filter-branch16:11
gitinfo tigris_: [!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)16:11
tigris_: the git-filter-branch manpage is available at http://git-scm.com/docs/git-filter-branch.html16:11
← MaximLacrima left16:11
← EyesIsMine left16:12
tigris_ FauxFaux: these repos will never ever be cloned by anyone, so rewriting is fine16:12
← thorbjornDX left16:12
→ thorbjornDX joined16:13
catphish_ FauxFaux: i wonder how others get around this memory limitation, i guess limit the size of pack files from day 116:13
→ koltroll joined16:14
→ fread2281 joined16:14
cmn you should hardly ever hit memory limits, specially not on day 116:14
catphish_ cmn: it's not day1 that's the problem16:14
→ shamrock joined16:14
catphish_ it's when the repack reaches 10GB :)16:14
→ gigamonkey_ joined16:15
tigris_ consider the situation where you accidentally commit a 50mb file, make a bunch more commits to other files, then want to remove that 50mb file but have not pushed yet and don't want that 50mb file in the history16:15
cmn don't repack everythign then16:15
shamrock → Guest7303016:15
cmn tigris_: man git filter-branch16:15
gitinfo tigris_: the git-filter-branch manpage is available at http://git-scm.com/docs/git-filter-branch.html16:15
← Guest73030 left16:15
catphish_ cmn: i have a 5GB packfile16:15
→ tgunr joined16:15
catphish_ in this particular repo16:15
→ rgr joined16:15
catphish_ now that i'm in this situation i'm unsure what the solution is16:15
maybe a full unpack and a repack with a limit on the packfile size16:16
it seems that repack never breaks packfiles up16:16
cmn that's precisely what's going to cause you to use a lot of memory16:16
← gigamonkey left16:16
gigamonkey_ → gigamonkey16:16
cmn what are you actually trying to achieve? repack everything into one pack?16:16
catphish_ cmn: i'm trying to run git gc without using 12GB of RAM16:17
cmn that's not a goal16:17
a-l-e ok16:17
catphish_ cmn: yes it is16:17
a-l-e git rebase --onto origin/master svn/svn indic16:17
gave me:16:17
http://pastebin.com/LfH5uXgW16:17
cmn catphish_: fine, then have fun with your experiments and don't forget to look at man git config16:17
gitinfo catphish_: the git-config manpage is available at http://git-scm.com/docs/git-config.html16:17
a-l-e which somehow looks correct (i can't really tell... but i don't see big problems)16:17
catphish_ cmn: gitinfo: i know where the manuals are16:18
canton7 a-l-e, looks aright16:18
a-l-e but git branch --v -a still shows:16:18
catphish_ but they don't offer any specific advice on reducing packfile size16:18
a-l-e but git branch -vv -a still shows:16:18
cmn catphish_: good for you16:18
→ TalleyHo joined16:18
a-l-e indic e38e57f [svn/svn: ahead 14] never select CAIRO_PRIVATE any more - someone should probably clear out CMakeLists.txt16:18
canton7 a-l-e, the indic branch will still be tracking whatever it was befoer16:18
but it'll be somewhat different16:19
cmn catphish_: yes, it does16:19
a-l-e well, that's not really the goal...16:19
← athit left16:20
a-l-e but well, i would already be happy if i could successfully pull the changes that are happening in svn/svn ...16:20
catphish_ i assume that using --max-pack-size=<n> and --window-memory=<n> will stop large packfiles being generated, but doesn't seem to help when existing large files exist16:20
← fluxdude left16:20
canton7 a-l-e, "tracking" is a bit of config which says, among other things, what to merge in when 'git pull' is run, and what the branch is compared to when git branch -vv is run16:20
a-l-e, the commit that your branch is based on is an entirely different thigns16:21
catphish_ cmn: but if you know the answer, as your comments imply, why not just tell me?16:21
a-l-e canton7: however, the change i'm testing to see if the current state contains the latest commits in svn/svn still does not show up...16:21
→ ncd|leeN joined16:21
canton7 a-l-e, it's easy to change what branch your branch is tracking. see git branch --set-upstream16:21
cmn because you dismiss everything16:21
also see !fish and all that goodness16:21
gitinfo In the long run, it is much better to give pointers to answers than the answers themselves.16:21
cmn what? no fishes in ! fish? I am diasspoint16:21
canton7 a-l-e, so what are you expecting to see when you run git branch -vv?16:22
a-l-e !dead16:22
(in the long run we are all dead :-)16:22
catphish_ cmn: what have you suggested that i have dismissed?16:22
a-l-e canton7: my main goal is to find out how to pull the changes from svn/svn into my indic branch...16:22
the mean i thought i should go, was to get the indic branch based on on origin/master like all the other branches.16:23
← tgunr left16:23
a-l-e so, right now i was expecting to see the indic branch to point to "origin/master" instead of "svn/svn"16:23
← _Qman left16:23
catphish_ apart from reading the manual, which i have clearly done16:23
→ anabelle joined16:23
a-l-e another possible way would be to create a new branch with the same content as indic but based on origin/master.16:24
← leeN left16:25
a-l-e i think i'll try a manual way...16:25
← pppaul left16:26
← ncd|leeN left16:26
→ txomon|nqas joined16:26
→ leeN joined16:27
txomon|nqas hi, I have a tree from a spot of the history, and I would like to know where in the history that tree is16:27
a-l-e i'm creating a indic2 branch based on origin/master and i will try to apply each change made in indic to it.16:27
→ EyesIsMine joined16:27
→ nicoulaj joined16:27
a-l-e is there any visual tool to do it?16:28
txomon|nqas gitk16:28
a-l-e, ^16:28
→ starlays joined16:29
txomon|nqas speak about commit16:29
← LimeBlast left16:29
txomon|nqas changes... is not the same as a commit16:29
a-l-e ... i only knew what i wrote...16:29
catphish_ cmn: in future, be more polite, since nobody is both willing and able to answer me directly, i will test every memory mapping and packfile sizing option to see which resolve the problem with existing packfiles16:29
→ dsirijus joined16:29
a-l-e (what's the difference between changes and commits?)16:29
starlays is there an command that can show me only the commits for a certain file, so i can see the history of that file from beginning to the end? thank you in advance16:29
canton7 a-l-e, "so, right now i was expecting to see the indic branch to point to "origin/master" instead of "svn/svn"" << I've already told you that your assumptions here are flawed, how they're flawed, and how to correct the situation16:30
milki starlays: man git-log16:30
gitinfo starlays: the git-log manpage is available at http://git-scm.com/docs/git-log.html16:30
EugeneKay starlays - git-log will accept !dashes16:30
gitinfo starlays: Use a double-dash(--) to separate refs or arguments from file paths. This is especially useful when dealing with ambiguous names. Ex: `git checkout origin -- master` will check out the file "master" from branch "origin"16:30
canton7 a-l-e, that "pointing to" is a trivial little bit of config which it's dead easy to change. and i've told you how16:30
a-l-e canton7: somebody else told me that i should use the git rebase command and that the command i pasted looked ok...16:31
canton7 a-l-e, yup, and that rebased your branch onto origin/master correctly16:31
← authmillenon left16:31
txomon|nqas a-l-e, change means the changes itself, but commit means patches, in the way that 5 commits can have 5 different changes on the same thing, but there would be just 1 resulting change16:31
canton7 a-l-e, but, for the 3rd time, there's still a little bit of config saying that git should, in its status message, compare your branch to svn/svn16:31
← zatan left16:31
canton7 a-l-e, which you need to change16:31
milki a-l-e: rebase doesnt change the upstream pointer16:32
← lun_ left16:32
→ am1ja2VuemllCg joined16:32
→ tgunr joined16:33
← i42n left16:34
canton7 a-l-e, once you understand more about git you'll realise how simple this is. for now, please recognise that your assumption is wrong, and that the branch that git compares your branch to when you run 'git branch -vv' is configured by a line of config, and is essentially arbitrary16:34
← mritz left16:34
← irqq left16:34
a-l-e canton7: i have no problem in recognizing it :-)16:34
→ juvenal joined16:34
a-l-e i believe everything you say and i try to do it as good as possible...16:35
canton7 a-l-e, now run 'git branch --set-upstream indic2 origin/master16:35
a-l-e but i still don't have my indic branch updated to the content of svn/svn or origin/master...16:35
← Desproges left16:36
a-l-e i have no religious feeling about which way leads to the solution :-)16:36
canton7 a-l-e, but you do. you're just assuming you don't16:36
→ mritz joined16:36
a-l-e no, the change i'm checking *is not* in the local directory.16:36
→ keyzs joined16:36
← sylr left16:36
canton7 a-l-e, git log --graph --oneline --decorate indic2 origin/master svn/svn16:37
pastie that16:37
a-l-e ... gonna try to change the upstream16:37
→ fornext joined16:37
← jola left16:37
a-l-e ok, now i have:16:38
indic e38e57f [origin/indic: ahead 14, behind 9] never select CAIRO_PRIVATE any more - someone should probably clear out CMakeLists.txt16:38
(after git branch --set-upstream indic origin/indic ... and this matches how the other branches are defined)16:38
canton7 where did origin/indic come from? it was origin/master before :P16:39
keyzs is there any channel on freenode related more to project management?16:41
a-l-e i can give you an example with a branch wich i have correctly created:16:41
bremner #phb ?16:42
EugeneKay keyzs - I know of none dedicated to it, but we're usually willing to chat16:42
canton7 a-l-e, now you've introduced origin/indic, i'm again unclear as to what you want16:42
a-l-e git checkout --track -b statusbar origin/master16:42
← RoyBellingan left16:42
canton7 (that --track isn't needed btw)16:42
a-l-e i made my changes, pushed it to the server and now it looks like this:16:42
statusbar16:42
remotes/origin/statusbar16:43
canton7 what looks like that? what are you looking at?16:43
a-l-e well, when i do "git branch -a" i see that...16:43
and git branch -vv -a shows:16:43
→ DanyO83 joined16:43
a-l-e statusbar fca7048 [origin/statusbar] prepare moving the page selector functionality to the toolbar16:43
remotes/origin/statusbar fca7048 prepare moving the page selector functionality to the toolbar16:44
← Davey left16:44
catphish_ FauxFaux: thanks :)16:44
canton7 a-l-e, ok. now do you understand what the [origin/statusbar] in the above output means?16:44
catphish_ looks like this memory usage during packing can be significantly reduced, though when writing existing pack files, their size doesn't seem to be reduces, and the write still seems to have significant memory needs, unless i just haven't found the right limit yet16:45
a-l-e that there is a link between both... but no, i don't really understand what the links means.16:45
← bartek left16:45
← rgr left16:45
canton7 a-l-e, ok, for the 4th time, that link is entirely arbitrary. it exist only for convenience16:46
*exists16:46
a-l-e voilà, i did a "git pull origin master" while being in the statusbar branch and, there, the change i'm tracking shows up16:46
canton7: as i told, all this is a mean to the goal.16:46
← kevwil left16:46
→ bartek joined16:46
canton7 there's nothing special about what's shown in the square brackets16:46
now, what do you think is wrong about how the indic branch is configured? exactly?16:47
a-l-e until now "git pull origin master" while being in the indic branch did not pull that change!16:47
i can retry it now...16:47
canton7 cool! that's a result of the rebase16:47
→ Davey joined16:47
→ gatinha joined16:47
← kerrick left16:47
a-l-e no, after the rebase the change was not in the indic branch (and it still is not)16:47
the other branches get it, but not the indic one.16:48
but i'll try again now after the rebase.16:48
canton7 define "get it"16:48
a-l-e pull from the server.16:48
→ brainbox1100 joined16:48
canton7 after the rebase, 'git checkout indic && git pull origin master' would have done what you wanted16:48
→ skyf_ joined16:48
→ skyf joined16:48
keyzs EugeneKay, wich platform is probably best on world to project management?16:48
canton7 (I think. I'm still not entirely clear what you want)16:48
a-l-e somebody committed the change to the svn server, it was automatically syncronized with svn/svn and manually pulled into origin/master.16:48
→ apok joined16:49
a-l-e well, the canton7, the two other commands "after the rebase" were still missing for me...16:49
← codeshah left16:49
a-l-e canton7: i want a way to pull the changes from the svn server into my indic branch.16:49
canton7 'cos you enver said you wanted to pull :)16:50
*never16:50
a-l-e canton7: those may be magical words for you, that are very clear...16:50
... on my side i try to explain what i want to do as good as possible: and it's hard!16:50
→ apok_ joined16:51
a-l-e but if "pull" is more clear to you, well, yes, i want to pull16:51
:-)16:51
canton7 you said you wanted to rebuild the indic branch on top of origin/master. we told you how to do that16:51
← Mannequin left16:51
→ maletor joined16:51
a-l-e well, this is what i want to do...16:51
canton7 the "re-linking" you were talking about was the branch --set-upstream16:51
→ ipalaus joined16:52
→ JotaLambda joined16:52
a-l-e it should be based on origin/master. this means for me that it should be "easy" to get the changes that happen in origin/master in the indic branch.16:52
← mishok13 left16:52
canton7 yup. and the rebase achieved that16:52
a-l-e but it may be time that you tell me what "based" really means in git-language :-)16:52
→ Akufen joined16:52
← Akufen left16:52
→ Akufen joined16:52
canton7 before that, you were complaining that a 'git pull' didn't work. we told you how to make it work. since then you've changed the goal from "get pull to work" to "get the upstream changes into my branch". the solution to the latter is the former, as you've found out16:53
← apok left16:54
apok_ → apok16:54
← etcetera left16:54
canton7 a-l-e, heh, if I checkout master, then create a new branch, that new branch is based on master16:54
master is the commit on which the commits which comprise my branch are built16:54
a-l-e no, i did not change my goal at all :-)16:54
but i may have not expressed my goal in a way that was understandable for you.16:55
→ etcetera joined16:55
txomon|nqas so can anyone help me discovering a tree in the history?16:55
canton7 <a-l-e> why is the the "indic" branch missing in the list "Local branches configured for 'git pull'"?16:55
^^ that's the first thing I can identify as a goal16:55
→ SparFux joined16:55
SparFux Hi! Can I undo git reset --hard HEAD^ somehow?16:55
a-l-e no that was a first specific question after i have described the problem...16:56
... however, this won't change what i thought i was saying and what you did understand...16:56
txomon|nqas SparFux, did you stash/commit?16:56
← DanyO83 left16:56
SparFux txomon|nqas: no.16:56
a-l-e my goal is still to get the thing to work... not to be right :-)16:56
canton7 a-l-e, yeah, dead horses and all that16:56
← etcetera left16:56
← catphish_ left16:56
txomon|nqas and add?16:56
SparFux no.16:56
canton7 a-l-e, are things working as you'd expect now?16:56
← nightcrawler786 left16:57
SparFux no add, git diff says, there are no diffs.16:57
txomon|nqas SparFux, then I think you can't recover16:57
a-l-e let me try... i had to first commit the changes i made to statusbar to be able to switch back to indic16:57
SparFux txomon|nqas: ok, thx for answer.16:57
txomon|nqas but not totally sure, as I am not a geek16:57
→ arkiver joined16:57
→ dr_lepper joined16:57
← dr_lepper left16:57
→ dr_lepper joined16:57
a-l-e no, canton7, the string i'm looking for is still not in the file i'm checking.16:57
← gatinha left16:58
← Vile left16:58
a-l-e and it was in the statusbar branch after having pulled origin master16:58
canton7 a-l-e, which remote branch is the string available in?16:58
ok, so the string is in origin/master?16:58
← macabre left16:58
canton7 git checkout indic && git pull origin master16:58
a-l-e now, i guess at least in origin/master and origin/statusbar16:58
git pull origin master16:58
From scribus.net:scribus16:58
* branch master -> FETCH_HEAD16:58
Already up-to-date16:58
→ iamjarvo joined16:59
a-l-e Scribus/scribus/commonstrings.cpp16:59
E486: Pattern not found: \<trMonochrome\>16:59
git checkout statusbar16:59
canton7 ok, so everything present in the master branch is present in the indic branch. git log -p -- path/to/file , anything relating to the string?16:59
a-l-e QString CommonStrings::trMonochrome = "";16:59
← Yuuhi` left17:00
← dsirijus left17:00
→ pfFredd joined17:00
→ cenuij joined17:00
← cenuij left17:00
→ cenuij joined17:00
a-l-e ok, i probably see the problem...17:01
→ Hasbro joined17:01
← arkiver left17:01
→ gatinha joined17:01
← iamjarvo left17:01
a-l-e "git log -p -- path/to/file" shows that a patch applied to the indic branch has deleted the changes done in the svn repository.17:02
→ g33k joined17:02
a-l-e it's a "misunderstanding" between the guy who did the patch and me...17:02
→ towski joined17:02
canton7 misunderstanding by you guys or a misunderstanding by git? :P17:03
→ nightcrawler786 joined17:03
→ Geralt joined17:03
← Geralt left17:03
→ Geralt joined17:03
a-l-e well, the problem is that the guy who has sent the patch has not based his work on the git repository...17:03
← zhouxiukun left17:04
← Geralt left17:04
a-l-e the workflow is still less then perfect... it's quite bad actually...17:04
but i have to live with it...17:05
i guess i've lost a full work day on this...17:05
canton7 ouch17:05
→ rgr joined17:05
→ Fandekasp joined17:05
a-l-e yep, that's the price to pay to learn how to learn git in a too complex environment :-(17:06
but well, if i don't do it, we lose contributors...17:06
→ _Vi joined17:06
a-l-e ... so, at the end, the time is well invsted... but it hurts!17:06
← apok left17:07
→ kevwil joined17:07
a-l-e and the step from "knowing nothing" to "managing and helping other people" is too steep...17:07
→ Haix64 joined17:07
← shruggar left17:07
→ etcetera joined17:07
Haix64 Hi. is it possible to create a system wide ".gitignore"?17:07
canton7 Haix64, man git-config core.excludesfile17:07
gitinfo Haix64: the git-config manpage is available at http://git-scm.com/docs/git-config.html17:07
constant → trout17:07
Haix64 thanks ;-)17:08
a-l-e however, thans to everybody and especially to canton7 and cmn!17:08
← arvind_khadri left17:08
← Haix64 left17:08
a-l-e i've learnt several things this evening...17:08
canton7 it's really worth digging into some git literature before trying to do too much17:09
nitrix Well thats why people stays on IRC channels.17:09
canton7 personally I prefer !bottomup to the book (and it's shorter!) but each to their own17:09
gitinfo 'Git from the bottom up' starts with explaining the building blocks of git and proceeds to tell you how they fit together. http://ftp.newartisans.com/pub/git.from.bottom.up.pdf17:09
a-l-e ... but now i will ask the two guys who have made their patches to redo them and apply them again to a clean branch.17:09
nitrix You get help and eventually help others with problems that you had before17:09
→ skywalker joined17:09
skywalker → Guest9741117:09
a-l-e well, i've been using svn for years and i understand the basics of git (i've read much about it... but never a full book)... but it's hard to learn advanced things before you need them...17:10
← Maccer left17:10
canton7 yeah, svn is another problem: almost everything from it has no relevance to git. and a lot of it is misleading :P17:10
and yeah I know it's hard, but having gone through it it's really worth it17:10
a-l-e yep, i love the new chances git gives to me!17:11
it makes many things much easier!17:11
← metcalfc left17:11
← gatinha left17:11
a-l-e but it's also a bit too hard to learn for something that is to be seen as a basic need for a free software project.17:12
canton7 otoh, once you dig into it, git becomes a powerful, active, productivity-helping tool, and not just a background chore like svn was17:12
← gdelfino left17:13
← angelsl left17:13
← fornext left17:14
← ipalaus left17:15
PerlJam a-l-e: git takes the same approach as perl. Simple things are simple, hard things possible. And you don't have to learn the hard things up front; you can learn as you go.17:15
tigris_ true that, i know people who their sole job is "release management" with CVS... don't see that shit with git17:16
a-l-e well, PerlJam, i think that you could have brought a better example than perl :-)17:16
PerlJam a-l-e: better in what way?17:16
→ neurodrone joined17:16
PerlJam a-l-e: (note my nick ;)17:16
a-l-e well, i would not sugges you to use perl as an example to convince me that something is good :-)17:16
← kevwil left17:16
txomon|nqas smells war17:17
a-l-e ... nothing against it and people using it, but i don't see it as a friendly thing :-)17:17
tigris_ a-l-e: no one said it was good, they said "simple things are simple, hard things are possible"17:17
txomon|nqas says: 3, 2, 1... Fight!17:17
PerlJam txomon|nqas: heh, not quite :)17:17
← Tegi left17:18
a-l-e nor i said that perl good or not: just not the good example to convince me :-)17:18
bremner you find praising git in #git to be flamebait? interesting.17:18
a-l-e fyi, beer as also a bad example :-)17:18
PerlJam a-l-e: okay, pick some technolog you like and I'll relate git to it :-)17:18
technology even17:18
tigris_ bacon17:18
a-l-e better :-)17:18
PerlJam *everything* is better with bacon! ;)17:19
→ hwrd|work joined17:19
→ rath_ joined17:19
→ macabre joined17:19
a-l-e however, thanks for you help and i will have a look at the git bottom up book17:19
tigris_ git is like bacon, you know you want some even when you don't need it17:19
← alberto56 left17:19
← glennpratt left17:20
→ neurodrone_ joined17:21
txomon|nqas so, has git any tool to help me searching the tree I have where it is in the history of a repo (I know it comes from some point of it)17:21
→ paul_and1ew joined17:21
tigris_ txomon|nqas: you have the sha1 of a tree and want to know what commit put it there?17:22
→ beneggett joined17:22
a-l-e thanks for all the hints!17:22
→ glennpratt joined17:22
← etcetera left17:24
→ alberto56 joined17:24
→ zastern joined17:25
← adityamenon left17:25
← les_sylvains left17:26
zastern Does creating a .gitignore file in a directory, that contains just *, also ignore everything in all subdirectories?17:26
→ metcalfc joined17:27
→ etcetera joined17:27
tigris_ txomon|nqas: try this $ git log --format='%T %H' | grep '^$TREE_SHA1'17:28
← jaseemabid left17:29
← rgr left17:30
← brainbox1100 left17:31
← PigeonFriend left17:32
→ sorahn joined17:32
sorahn hey guys, i'm trying to add a custom function to my .gitconfig, and i'm not sure whats wrong with it.17:33
https://gist.github.com/be506df5af48217094ee17:33
the 'rmm' function is the new one that it's complaining about.17:33
← g33k left17:33
← anabelle left17:34
← gregm_ left17:35
→ Yuuhi joined17:35
← metcalfc left17:35
txomon|nqas tigris_, I just have the tree, is not part of a git repo itself, just the tree without the .git17:35
→ stephenjudkins joined17:35
milki so like a bunch of files17:35
tigris_ sorahn: it will be that \*17:36
you may need another escape on it17:36
← neurodrone left17:36
milki you might be able to generate an actual git tree from those files and obtain a sha that should match a git object in the repo17:36
sorahn ah!17:36
that was totally it! Thanks tigris_!17:37
txomon|nqas milki, how may I do that?17:37
milki txomon|nqas: i dunno17:37
o hey17:37
txomon|nqas: theres a man git-mktree17:38
gitinfo txomon|nqas: the git-mktree manpage is available at http://git-scm.com/docs/git-mktree.html17:38
tigris_ txomon|nqas: http://git-scm.com/book/en/Git-Internals-Git-Objects17:38
milki though..ls-tree might need blob shas too...17:38
→ Eridius_ joined17:38
milki whic means you need to make those too...17:38
just toss them into a repo?17:38
and get the sha17:39
→ stevedomi joined17:39
milki and see if it matches on in the original repo?17:39
in your tree, git init, git add ., git commit, grab the tree sha17:39
← papegaaij left17:40
txomon|nqas good idea17:40
Drarok → drarok17:40
→ whitman joined17:40
← SparFux left17:41
→ artem joined17:41
← artem left17:41
→ Maccer joined17:41
← stevedomin left17:42
→ shruggar joined17:43
sorahn hey tigris_ do you know of a similar command I could run to remove all branches that have been pushed to origin?17:43
local branches that is.17:43
← stevedomi left17:43
milki oo, i made one17:44
hmm17:44
$GIT branch --merged master | $SED 's/\s*\///' | $GREP -v 'master$' | $XARGS -I% $GIT branch -d %17:44
o17:44
no...17:44
thats for what was merged to master17:44
uh17:44
wrong repo17:44
hmm17:44
← artnez left17:44
← rmr left17:44
milki hm..., i guess in these, i assumed whatever was in master is pushed and individual branches are not pushed17:45
so nevermind -.-17:45
txomon|nqas ... git-diff-tree would do something in my case?17:45
→ diegoviola joined17:45
→ luxurymode joined17:45
→ Bass10 joined17:46
→ Moussekateer joined17:46
← thiago left17:46
txomon|nqas has to go home (19:48 at work)17:46
→ thiago_ joined17:46
txomon|nqas bb!17:47
← Bass10 left17:47
← shruggar left17:47
milki pfft, thats not late17:47
→ zmo joined17:47
milki unless home is far17:47
tigris_ can't help you on that one sorry sorahn17:47
zmo hi17:47
gitinfo zmo: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.17:47
thiago_ → thiago17:48
→ Bass10 joined17:48
zmo urgh, gitinfo is frightning !17:48
tigris_ more frightening is the fact he's not even a bot17:48
zmo tigris_ - ahah :)17:49
← Hasbro left17:49
→ fly9_ joined17:49
← fly9_ left17:49
← fly9 left17:49
→ fly9 joined17:49
zmo ok, so I got a question: I actually want to move a few commits from master to a branch (I noticed I changed a full subtree of my code to a new API that should have been in another branch), I know for sure those commits do not interfere with the other commits (starting from the moment I did the mistake)17:50
tigris_ man git-cherry-pick17:50
gitinfo the git-cherry-pick manpage is available at http://git-scm.com/docs/git-cherry-pick.html17:50
→ metcalfc joined17:51
zmo so basically, I'd like to have something like git rebase -i %mycommit%~10 ; remove the cherry picked commits ; generate a new commit with the difference between origin/master and my local master17:51
→ x3ro joined17:51
← giallu left17:51
SethRobertson You want your local master to look exactly like origin/master does? You can do that several ways. !fixup describes several of them17:52
gitinfo So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!17:52
← tigris_ left17:52
gitinfo set mode: +v17:53
zmo SethRobertson - I want my orinig/master look like my local master, after I've removed those erroneous commits17:53
revagomes I'm using submodules on my project and I'm wondering to know why sometimes there is the 'modified: path/to/the/submodule (new commits)' message and I need to commit this on the main repo...17:53
x3ro Hi everyone… I'm having serious trouble with git-svn and I'd like to know if there is _any_ way to at least what the heck is happening, if anything, because "git svn fetch" and "git svn rebase" just get stuck without doing anything most of the time :'(17:53
SethRobertson zmo: Ah, well step one is to fix up your local branch (and the link I sent will tell you how) and then you can either merge that into upstream or you can simply reset upstream to that location17:54
x3ro, strace, tcpdump, and `GIT_TRACE=1 git svn command` are all good first steps17:54
revagomes: !submodules_changes17:55
→ shruggar joined17:55
SethRobertson !submodule_change17:55
revagomes Is this the normal workflow when using submodules?17:55
gitinfo In order to change a submodule you must go into the submodule repository, check it out to the appropriate branch, make the needed change (possibly involving git pull), commit the change, cd .. (out of the submodule), git commit -m "Updated submodule" submodulepath17:55
x3ro SethRobertson: Thank you so much. That way I can at least see if it does anything at all :)17:55
zmo SethRobertson - ok (just before going to RTFM your link), I've already done a 'git rebase -i ...' where I actually removed the commits, and new local HEAD look the way I want... but git status tells me : "Your branch and 'origin/master' have diverged, and have 94 and 102 different commit(s) each, respectively" do you think I shall revert to HEAD my local index, or I can do what I want from where I am ?17:56
← koltroll left17:56
← Ivoz left17:56
→ nioncode joined17:56
→ kevwil joined17:57
SethRobertson zmo: You have several choices. You can blow away all other changes that are on upstream (!rewriting_public_history) or you can update your local branch to include all work other people have done.17:57
← TalleyHo left17:57
SethRobertson !rewrit?17:57
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)17:57
SethRobertson odd17:57
revagomes SethRobertson, Tks! I thought that the submodule workflow was like almost invisible on the main repo...17:57
← pfFredd left17:57
revagomes SethRobertson++17:58
SethRobertson revagomes: You might want to consider !gitslave for something more invisible. submodules is a pain17:58
gitinfo revagomes: gitslave (http://gitslave.sf.net) is useful to add subsidiary git repositories to a git superproject when you control and develop on the subprojects at more or less the same time as the superproject, and furthermore when you typically want to tag, branch, push, pull, etc. all repositories at the same time.17:58
→ thirdknife_ joined17:58
← keyzs left17:58
→ les_sylvains joined17:58
→ sirdancealot joined17:58
zmo SethRobertson - ah, ok, I never intended to rewrite history, so it seems that as sexy "git rebase -i" looked like, it was a terrible option17:58
ok, I'm reverting to origin/master and RTFMing your link17:59
thank you for your insights17:59
← caddoo left17:59
→ robotmay joined18:00
← bfgun left18:00
→ bfig joined18:00
bfig Hello, i just made a fast forward merge and broke everything, how can i go back to the master before the ff merge?18:00
← EyesIsMine left18:01
→ goshawk joined18:01
→ fusion271 joined18:01
canton7 bfig, git reflog master will give you the sha1 of the commit which master was at before the merge18:02
→ comps joined18:02
← trungqtran left18:02
← prudnikov left18:03
← skyf_ left18:04
← skyf left18:04
← v0n left18:04
← RaceCondition left18:05
→ trungqtran joined18:05
→ tjadc joined18:05
→ irqq joined18:05
→ opus joined18:06
→ astruasdh joined18:06
→ cakehero joined18:07
← nioncode left18:08
→ trollboy joined18:08
trollboy Howdy. I've got a head repo, and I've pushed several files for next rollout, however I just pushed a file that needs to roll for emergency rollout. How can I update just that file and leave everything else in repo?18:09
I tried "git pull ./path/to/my/file" and "git checkout ./path/to/my/file" with no love18:10
← fusion271 left18:10
revagomes SethRobertson, Nice! Tks. ;]18:11
SethRobertson By "pushed" do you mean `git push`?18:11
infogulch|afk → infogulch18:12
→ HAL_9001 joined18:12
← HAL_9001 left18:12
→ HAL_9001 joined18:12
→ RaceCondition joined18:12
← RaceCondition left18:12
SethRobertson trollboy: If I understand you (and I may not), you want to `git fetch` to get the change and then `git checkout origin/master -- path/to/file` to get the file committed upstream18:12
← Samual left18:12
← metcalfc left18:12
→ steelwil joined18:13
bfig canton7, great, now how do i remove the last commit?18:13
SethRobertson bfig: !fixup18:14
gitinfo bfig: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!18:14
→ Sigma joined18:14
trollboy and git fetch won't touch any other files?18:14
← monsieurlu left18:15
← ladder4 left18:15
SethRobertson It only modifies the remote tracking branches (!tracking)18:15
!tracking18:15
gitinfo [!tracking_branches] Remote tracking branches are read-only copies of your last information on what branches the remote last held. The "origin's master" branch is stored as the remote tracking branch origin/master. Local tracking branches are your writable copies of the remote tracking branch. http://www.gitguys.com/topics/tracking-branches-and-remote-tracking-branches/18:15
SethRobertson Nothing in the working directory or the index18:15
→ monsieurlu joined18:15
→ KWhat_Work joined18:15
→ peterpan joined18:16
→ nioncode joined18:16
← pheaver left18:17
→ v0n joined18:17
trollboy that worked like a champ! Thanks!18:17
← chimay left18:18
→ pantsman joined18:19
← pantsman left18:19
→ pantsman joined18:19
→ converge joined18:21
→ RaceCondition joined18:21
→ imachuchu joined18:21
← RaceCondition left18:22
→ jaseemabid joined18:23
→ ology joined18:23
→ _md joined18:24
→ rgr joined18:25
→ metcalfc joined18:26
← _md left18:26
→ Stego joined18:26
← sebastorama left18:28
← jaseemabid left18:28
→ Hasbro joined18:28
← Fandekasp left18:31
→ RaceCondition joined18:31
← Stego left18:32
→ mwrosen joined18:32
← Hasbro left18:33
→ Hasbro joined18:33
← mwrosen left18:33
→ aldodelgado joined18:33
← shruggar left18:33
→ Guest91045 joined18:34
← monsieurlu left18:34
→ chrisgeorge joined18:34
← chrisgeorge left18:34
→ chrisgeorge joined18:34
← mischat left18:35
← Milossh left18:35
→ monsieurlu joined18:35
← nightcrawler786 left18:35
← Eridius_ left18:35
→ nightcrawler786 joined18:36
← am1ja2VuemllCg left18:36
→ stevedomin joined18:36
→ artnez joined18:37
HAL_9001 → Samual18:38
← anish[1] left18:38
→ Protux joined18:39
→ am1ja2VuemllCg joined18:40
→ Eridius_ joined18:40
→ jaseemabid joined18:40
→ atrigent joined18:40
← ehsan left18:40
→ ehsan joined18:40
← jason237 left18:40
← FasTTo left18:40
atrigent are there any patch queue things/features in git which would allow you to freely move diff chunks between different patches, between a patch and the working tree, etc, kind of like what git add -p can do?18:41
→ lacrymology joined18:41
atrigent sorry if this is a very basic question, I don't know much about patch queues18:41
PerlJam atrigent: I'm not sure what you're asking.18:44
SethRobertson atrigent: Why not apply the patch and then use `git add` and `git reset` -p?18:44
← Guest91045 left18:44
PerlJam what do you mean "between different patches"?18:44
atrigent: you have patch files that you want to apply partially?18:45
atrigent like say I put a change in the wrong patch18:45
SethRobertson patch? Do you mean commit?18:45
PerlJam atrigent: when you say "patch" I think you really mean "commit"18:45
atrigent well yes, it would depend on the terminology of the patch queue software...18:46
I guess the ones that interface directly with git probably do use commits18:46
← steelwil left18:46
← tjholowaychuk left18:46
SethRobertson atrigent: Step one, turn the patches into git commits. Step two, use !sausage to make each commit look the way you want. Step three, use the various git commands to extract the new patches18:47
gitinfo atrigent: [!sausage_making] Some developers like to "hide the sausage making", transforming their commits before presenting them to the outside world. See http://sethrobertson.github.com/GitBestPractices/#sausage and !perfect18:47
SethRobertson !perfect18:47
gitinfo [!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.18:47
PerlJam atrigent: git doesn't manage "patches", so you're looking for some external tool that uses git and manages patches to your repo?18:47
atrigent I know that git itself does not deal in "patches"18:47
but, for example, StGit is described in terms of patches18:48
SethRobertson You can use the StGit workflow if you want, or the one I described18:48
bremner well, there is e.g. git-format-patch and git-am which are very much about patches18:48
← nioncode left18:49
→ archdori joined18:50
← am1ja2VuemllCg left18:50
PerlJam bremner: sure, but that's more like import/export than "manage"18:50
→ mrOrangeChuck joined18:50
← Zespre_ left18:51
→ jason237 joined18:51
← franck__ left18:51
→ pppaul joined18:51
← pppaul left18:51
→ pppaul joined18:52
← nitrix left18:52
← pppaul left18:52
→ pppaul joined18:52
← pppaul left18:52
→ Zespre joined18:52
→ pppaul joined18:53
← pppaul left18:53
← KWhat_Work left18:53
← jaseemabid left18:53
→ pppaul joined18:53
← pppaul left18:53
← doomMonkey left18:54
→ pppaul joined18:54
← pppaul left18:54
goshawk how can i show a file from another revision on emacs with git.el ?18:54
→ pppaul joined18:55
← pppaul left18:55
→ pppaul joined18:55
← pppaul left18:55
→ jaypipes joined18:55
→ pppaul joined18:56
← pppaul left18:56
jaypipes anybody a submodule expert that might be able to explain wtf git add isn't working here? http://paste.openstack.org/show/20719/18:56
→ pppaul joined18:56
← pppaul left18:56
→ mischat joined18:57
→ pppaul joined18:57
← pppaul left18:57
SethRobertson It says "changes to be committed". What did you expect?18:57
jaypipes SethRobertson: it didn't add cookbooks/jenkins when I git add'd it.18:57
→ pppaul joined18:57
← pppaul left18:57
atrigent SethRobertson: I am aware of interactive rebasing, but I'm looking for a different workflow than that18:58
→ pppaul joined18:58
← pppaul left18:58
atrigent I'm thinking something like "git move <some identifier of a patch or commit> <some other identifier>"18:58
jaypipes SethRobertson: per tutorials like https://git.wiki.kernel.org/index.php/GitSubmoduleTutorial, doing a git submodule add should add the submodule to the local index... but I'm not seeing that happen for some reason. :(18:58
atrigent and then it would run the interactive adder between those two commits18:58
→ pppaul joined18:58
← pppaul left18:58
atrigent and then do all of the rebasing stuff itself18:59
SethRobertson jaypipes: Did cookbooks/jenkins already appear there before the git-submodule-add?18:59
jaypipes SethRobertson: no.18:59
→ pppaul joined18:59
← pppaul left18:59
SethRobertson Was .gitmodules updated?18:59
← aldodelgado left18:59
SethRobertson Ah, yes it was18:59
→ bakineggs joined18:59
→ pppaul joined19:00
← pppaul left19:00
→ jaclinuxhelp joined19:00
SethRobertson Try `git show HEAD cookbooks/jenkins` to see if it was committed?19:00
jaypipes SethRobertson: one sec19:00
SethRobertson Also, what was the .gitmodules change19:00
→ pppaul joined19:00
← pppaul left19:00
jaclinuxhelp we have a branch named v11.x.x, which has become the "defacto" master branch19:00
jaypipes SethRobertson: nope:19:00
jpipes@uberbox:~/repos/att-cloud/chef-repo$ git show HEAD cookbooks/jenkins19:00
jpipes@uberbox:~/repos/att-cloud/chef-repo$19:00
jaclinuxhelp can i rename that v11.x.x to master ?19:01
→ pppaul joined19:01
← pppaul left19:01
jaclinuxhelp i mean, delete the master, then renaming v11.x.x as master19:01
just like that ?19:01
← _Vi left19:01
← tjadc left19:01
jaypipes SethRobertson: and doing a git diff shows nothing at all, even though git status shows the .gitmodules has changed :(19:01
→ pppaul joined19:01
← pppaul left19:01
SethRobertson jaypipes: Also also check to see if that path is .gitignored (by removing the directory and the .gitmodules changes and then making a directory in that spot and then putting a file in there and running `git status`)19:01
jaypipes SegFaultAX|work2: http://paste.openstack.org/show/20721/19:01
oops, sorry SegFaultAX|work219:02
SethRobertson: http://paste.openstack.org/show/20721/19:02
→ pppaul joined19:02
← pppaul left19:02
SethRobertson jaclinuxhelp: Are you talking about branches?19:02
jaypipes SethRobertson: yeah, already checked the .gitignore. Nothing in there.19:02
→ pppaul joined19:02
← pppaul left19:02
jaclinuxhelp SethRobertson: yeah19:02
SethRobertson jaypipes: do a `git reset --hard` and rm the directory and try the test I suggested, then delete the dir and try again19:02
→ teotwaki joined19:02
→ pppaul joined19:03
← pppaul left19:03
SethRobertson jaclinuxhelp: then yes, yes you can19:03
jaclinuxhelp: man git-branch19:03
gitinfo jaclinuxhelp: the git-branch manpage is available at http://git-scm.com/docs/git-branch.html19:03
← rgr left19:03
→ pppaul joined19:03
← pppaul left19:03
→ pppaul joined19:04
← pppaul left19:04
→ pppaul joined19:04
← pppaul left19:04
→ pheaver joined19:05
→ pppaul joined19:05
← pppaul left19:05
→ pppaul joined19:05
← pppaul left19:06
→ pppaul joined19:06
← pppaul left19:06
← macBigRig left19:06
→ pppaul joined19:07
jaypipes SethRobertson: http://paste.openstack.org/show/20722/19:07
← hwrd|work left19:07
← rgsteele left19:07
→ am1ja2VuemllCg joined19:07
← luxurymode left19:08
→ nitrix joined19:08
SethRobertson jaypipes: Odd. You have exhausted the limits of my knowledge. Can I suggest !gitslave instead?19:08
gitinfo jaypipes: gitslave (http://gitslave.sf.net) is useful to add subsidiary git repositories to a git superproject when you control and develop on the subprojects at more or less the same time as the superproject, and furthermore when you typically want to tag, branch, push, pull, etc. all repositories at the same time.19:08
← nitrix left19:09
jaypipes SethRobertson: sure, but I don't actually work on the jenkins cookbook :) that's why I wanted to just use a submoduile19:09
→ jjido joined19:09
→ ChristopheT joined19:09
← bakineggs left19:10
← jjido left19:10
→ alrs joined19:10
← dougsland left19:10
→ eletuchy joined19:10
→ jjido joined19:10
→ dougsland joined19:11
← stevedomin left19:11
← antix left19:13
← noisebleed left19:13
← Hasbro left19:14
← jaclinuxhelp left19:14
→ hidekin joined19:14
→ doomMonkey joined19:15
← mritz left19:15
← stephenjudkins left19:15
← mischat left19:16
jaypipes SethRobertson: so... I had my colleague execute the exact same commands (http://paste.openstack.org/show/20723/) and his worked just fine. So looks like my git is just broken. :(19:16
→ jaseemabid joined19:16
→ Sonderblade joined19:16
jaypipes SethRobertson: is there any way to get some verbose/debug output from the git submodule add command?19:16
SethRobertson `GIT_TRACE=1 git submodule add` might help, but will probably be too high level. Try upgrading your git.19:17
→ pfFredd joined19:18
jaypipes SethRobertson: cheers, will do. thx for your help mate19:18
← bizhanMona left19:20
← jjido left19:21
→ bizhanMona joined19:22
← converge left19:22
→ jjido joined19:23
→ rohan32 joined19:23
← ehsan left19:24
← LittleFool left19:26
→ mritz joined19:27
← beneggett left19:27
→ LittleFool joined19:27
← opus left19:28
→ beneggett joined19:28
→ specialGuest joined19:28
← eletuchy left19:28
→ hwrd|work joined19:28
→ eletuchy joined19:29
← specialGuest left19:29
→ specialGuest joined19:29
← nizeguy left19:29
← bartek left19:30
→ bartek joined19:31
← kevwil left19:32
→ athit joined19:33
→ kevwil joined19:33
→ stephenjudkins joined19:33
← Anders_J left19:34
→ anabelle joined19:35
← jim left19:38
← rath_ left19:38
→ adityamenon joined19:38
← hidekin left19:40
← Beej left19:40
→ dunpeal joined19:40
dunpeal Yo. Is there a way to just get a script-friendly list of branches?19:41
I just want the names, one per line, no additional decorations.19:41
_ikke_ dunpeal: git branch -a?19:41
← cakehero left19:41
dunpeal _ikke_: colors, and marks the current one.19:41
I actually don't need the -a, but I don't want the extra decorations.19:42
_ikke_ dunpeal: add --no-color to get rid of the colors19:43
→ tjadc joined19:43
→ rgr joined19:44
→ cirenyc joined19:45
_ikke_ dunpeal: Another option would be git for-each-ref refs/heads | awk '{print $3}'19:45
dunpeal _ikke_: thanks, my key issues is to get rid of prefixes19:45
← bartek left19:45
_ikke_ prefixes?19:45
refs/heads you mean?19:45
cirenyc I stashed changes that include a file I still wanted. Before popping the stash, I removed an unwanted copy of a file w/ git reset --hard. I thought popping the stash would replace that file, but I'm getting a "deleted by us" conflict. Is there a way to get it back?19:46
SirCmpwn I have a branch that is not an orphan19:46
I want to make it an orphan19:46
→ nitrix joined19:49
dunpeal _ikke_: nah, I mean the whitespace and * before branch names that `git branch` shows, but I guess for-each-ref is exactly what I need19:49
→ eletuchy_ joined19:49
← archdori left19:49
→ Ysgard joined19:49
→ YoMomma joined19:50
← paul_and1ew left19:50
dunpeal btw, is there a `git push --branches` like there is `git push --tags`?19:50
← alrs left19:50
_ikke_ dunpeal: No, git pushed branches by default19:50
← eletuchy left19:51
← nicoulaj left19:51
canton7 dunpeal, see push.default in man git-config19:52
gitinfo dunpeal: the git-config manpage is available at http://git-scm.com/docs/git-config.html19:52
→ bartek joined19:52
banghouse → banghouseAFK19:53
← Ysgard left19:53
dunpeal thanks, apparently in newer git you may need to do a `push --all` to actually push all branches to the remote.19:53
_ikke_ dunpeal: See the config option canton7 linked to19:54
canton7 yeah, that's documented there too19:54
→ Element9 joined19:54
← longbeach left19:54
→ stevedomin joined19:55
dunpeal ok, thanks.19:56
← cirenyc left19:56
→ srayner joined19:57
→ ehsan joined19:57
← Samual left19:57
→ j416 joined19:57
← bartman` left19:58
→ Samual joined19:58
← ehsan left19:58
← maestrojed left19:59
← gigamonkey left19:59
← crashsky left20:00
→ bartman` joined20:00
→ gigamonkey joined20:01
← neurodrone_ left20:01
← les_sylvains left20:02
← stevedomin left20:02
← iocor left20:03
→ _Vi joined20:03
→ skywalker joined20:03
← jjido left20:03
← am1ja2VuemllCg left20:03
skywalker → Guest8457120:03
← Guest97411 left20:05
→ hidekin1 joined20:05
→ ringlej joined20:05
← thirdknife_ left20:06
← arietis left20:06
← anabelle left20:06
→ mishok13 joined20:06
← Poapfel left20:06
← sawdisk_ left20:07
→ sawdisk_ joined20:08
ringlej I had a file in my git repo that I want to remove from git tracking it. This file is an autogenerated file that is developer specific. I used 'git rm' to remove the file from all branches that was tracking the file, but now when I try to 'git checkout' a branch that previously tracked this file, it complains that the file will be removed. I want git to forget about this file (untrack it). how can I do this?20:08
← hidekin1 left20:09
Moussekateer untrack it by adding it to .gitignore20:09
← jaeckel left20:09
_ikke_ Moussekateer: That won't untrack it20:10
← proradii left20:10
← cebor left20:10
Moussekateer oh? Is that because it's already commited?20:11
ringlej Moussekateer: adding the file to .gitignore seemed to accomplish it. _ikke_, why wont it?20:11
→ jjido joined20:11
→ ehsan joined20:11
_ikke_ ringlej: adding tracked files to .gitignore won't make them untracked20:11
→ Ysgard joined20:12
_ikke_ The only thing it does is preventing git status from reporting a file as being untracked20:12
ringlej when I did a 'git checkout branch' it no longer complained that it would delete that file20:12
Bombe It also prevents you from staging modifications to that file.20:12
_ikke_ Bombe: well, it warns you20:12
Bombe: You can still add it anyway if you want20:13
→ gdelfino joined20:13
ringlej _ikke_: so what would your solution be?20:14
→ paul_andrew joined20:14
Moussekateer Is there any way to change the behaviour of git diff on a binary file that's been deleted? I don't want it to through an error, but a helpful message instead20:14
_ikke_ ringlej: copy the file out, checkout -f and then fix the branch20:14
→ iocor joined20:14
ringlej _ikke_: and by "fix the branch", what do you mean? I already did 'git rm file'20:15
on the offending branch20:15
← maqr left20:16
→ hidekin1 joined20:16
→ nioncode_ joined20:18
→ anabelle joined20:18
→ am1ja2VuemllCg joined20:19
← x3ro left20:20
→ Poapfel joined20:20
→ jaeckel joined20:21
← jceb left20:22
→ cirenyc joined20:23
→ mwrosen joined20:23
jaypipes SethRobertson: FYI, in case you're interested, blowing away my local repo and starting over with a fresh one seems to have fixed everything... weird.20:23
← rgr left20:25
← jjido left20:26
→ maqr joined20:28
→ axl__ joined20:30
→ Chillance joined20:31
← cirenyc left20:32
→ stevedomin joined20:32
← axl_ left20:33
axl__ → axl_20:33
← Will| left20:35
← _iron left20:35
← athit left20:35
→ luxurymode joined20:37
→ aidz joined20:37
← aidz left20:38
→ Destos joined20:39
← stevedomin left20:39
→ mritz_ joined20:39
← duckxx left20:40
→ duckxx joined20:40
→ converge joined20:41
← converge left20:41
← TheJH left20:41
→ opus joined20:44
← eletuchy_ left20:44
← AlbireoX`Laptop left20:45
← am1ja2VuemllCg left20:45
→ eletuchy joined20:45
→ eletuchy_ joined20:46
→ codeFiend joined20:47
← mritz_ left20:47
→ am1ja2VuemllCg joined20:48
← adityamenon left20:48
← tplaner_ left20:48
← stepnem left20:49
← eletuchy left20:50
← Zenopus left20:50
← monsieurlu left20:50
← Sigma left20:51
← starlays left20:52
→ AdrienBrault joined20:52
← hegge left20:54
← fisted left20:54
→ mritz_ joined20:55
→ fisted joined20:56
→ psoo joined20:58
← Bass10 left20:58
→ cakehero joined21:00
← chrisgeorge left21:00
← cannonball left21:00
→ Christop` joined21:01
← ChristopheT left21:01
← torjus_ left21:01
→ marced joined21:01
← geekbri left21:02
← _Vi left21:02
→ monsieurlu joined21:02
→ _Vi joined21:03
← nioncode_ left21:03
→ jagster` joined21:03
← freeed left21:04
→ nioncode_ joined21:04
→ chrisgeorge joined21:04
← Guest84571 left21:04
→ Zenopus joined21:04
jagster` can anyone recommend any graphical clients for git on windows?21:05
→ skywalker joined21:05
_ikke_ GitExtensions21:05
skywalker → Guest1578021:05
→ gregm_ joined21:06
← kevwil left21:06
→ hobodave joined21:06
→ heyanabelle joined21:06
haarg might look at the github app. it works on repos that aren't hosted on github too.21:06
→ MB_wrk joined21:06
← rwestphal left21:07
← alberto56 left21:07
→ kerrick joined21:07
MB_wrk How do I undo the last commit and then merge the current uncommitted changes back in?21:07
← Zenmaker_ left21:08
→ am1ja2Vu_ joined21:08
→ Zenmaker_ joined21:08
SethRobertson MB_wrk: Stash or commit your current changes, and then !fixup21:08
gitinfo MB_wrk: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!21:08
→ _adnasa_ joined21:09
← _adnasa_ left21:09
MB_wrk Thanks SethRobertson21:09
← slakmagik left21:10
← am1ja2VuemllCg left21:10
am1ja2Vu_ → am1ja2VuemllCg21:10
→ kevwil joined21:10
→ slakmagik joined21:10
← slakmagik left21:10
→ slakmagik joined21:10
→ AlbireoX joined21:11
→ stevedomin joined21:11
← Christop` left21:12
← hobodave left21:12
→ alberto56 joined21:12
← kerrick left21:13
→ _adnasa_ joined21:14
← _adnasa_ left21:14
← irqq left21:15
→ kerrick joined21:15
→ _adnasa_ joined21:15
← _adnasa_ left21:15
← slakmagik left21:15
→ mischat joined21:15
→ Guest16975 joined21:15
← Guest16975 left21:15
← martinjlowm left21:16
← eletuchy_ left21:17
← stevedomin left21:17
← opus left21:19
→ Vile joined21:19
← mishok13 left21:20
MB_wrk thanks again SethRobertson. that's a great site.21:20
← luxurymode left21:20
← nioncode_ left21:20
MB_wrk And..it's yours. So great work!21:20
→ slakmagik joined21:20
milki heh21:20
→ Fandekasp joined21:20
← maletor left21:21
← mischat left21:21
← alexisabril left21:22
→ cirenyc joined21:23
← cirenyc left21:23
→ maletor joined21:23
← srayner left21:26
← hoschi left21:27
→ john_anderson joined21:27
← hellopat left21:28
← ShadeTornado left21:28
← kortti left21:29
← axl_ left21:29
← mritz_ left21:30
→ axl_ joined21:30
→ Foxandxss joined21:30
← afuentes left21:30
Foxandxss hello21:30
gitinfo Foxandxss: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.21:30
← Guest15780 left21:30
→ mritz_ joined21:32
← stephenjudkins left21:32
Foxandxss Im kinda noob with git and well, I use it with a rails project and I have a question: In my server I cloned my github repo and made changes to 2-3 files (you know, config that I can share in github :P), what's the correct way to make updates to that project using git?21:32
pdelvo → pdelvo|afk21:32
Foxandxss I tried a git pull but it cries because I modified some files21:32
_ikke_ Foxandxss: You first have to commit those files21:32
Foxandxss locally?21:32
→ opus joined21:33
Foxandxss git add files && git commit ?21:33
banghouseAFK → banghouse21:33
← opus left21:33
→ opus joined21:34
← maletor left21:34
← gfontenot left21:34
_ikke_ yes21:34
← j416 left21:34
→ fly9_ joined21:34
← fly9_ left21:35
← kevwil left21:35
← fly9 left21:35
_ikke_ But it's not adviced to commit changes to config files to git21:35
→ fly9_ joined21:35
→ obiwahn joined21:35
canton7 !config21:35
gitinfo [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas21:35
obiwahn hey i look for a projet to use git in a more centralized manner21:35
← fly9_ left21:35
→ opus__ joined21:36
← opus__ left21:36
obiwahn i think i heard of a project but forgot the name21:36
Foxandxss let's me see21:36
thiago obiwahn: are you looking for a project that uses Git in a centralised manner?21:36
obiwahn: or a project that will help you do that?21:36
Foxandxss I commited that files because the app cannot run without them21:36
→ fly9 joined21:36
milki and your app lives in a git repo?21:36
obiwahn a project that will help me to do that21:36
milki obiwahn: !gitolite21:37
gitinfo obiwahn: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html21:37
milki and/or github21:37
and/or gitorious21:37
obiwahn i am not the gitolite page:)21:37
i think it was something else21:37
← bartek left21:37
milki Foxandxss: !deploy21:38
gitinfo Foxandxss: Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html21:38
→ maletor joined21:38
Foxandxss no is not, capistrano is, I have to learn it, for sure :P21:38
milki indeed21:39
→ pantsman- joined21:40
← pantsman left21:42
← whitman left21:43
← nwest left21:44
→ noone4 joined21:44
→ rubious joined21:46
obiwahn http://code.google.com/p/gerrit/21:46
← faen left21:47
milki o, thats code review21:48
→ faen joined21:48
← am1ja2VuemllCg left21:49
→ jasmina87SA__ joined21:50
← imachuchu left21:50
← MB_wrk left21:50
→ madewokherd joined21:51
← optixx left21:51
← hwrd|work left21:51
→ optixx joined21:51
← maletor left21:53
← mwrosen left21:53
← joschi left21:53
→ hwrd|work joined21:55
→ joschi joined21:55
← cakehero left21:58
← kermit left21:58
← mritz_ left21:58
→ eletuchy joined22:00
→ cakehero joined22:01
→ FernandoBasso joined22:01
← Destos left22:01
← babilen left22:01
→ kermit joined22:02
← specialGuest left22:02
← _Vi left22:02
→ _Vi joined22:03
rubious My push keeps getting refused due to "non fast-forward update". But I run git pull, and it tells me I'm up to date! What the eff?22:04
_ikke_ rubious: Did something went wrong earlier with a pull?22:04
canton7 rubious, check where your branch is being pushed to, and what's being merge din when you run 'git pull'22:04
rubious canton7: I'm running 'git pull origin master'22:04
canton7 (the former: look at the output of 'git push'. the latter: look at git branch -vv)22:04
_ikke_ !four22:05
gitinfo [!fetchfour] Never use the four-word version of git-fetch or git-pull (e.g. git fetch remote refspec). It always ends in tears. Yes, if you understand the implications of FETCH_HEAD it can technically be done, but really it is easier to just fetch the whole remote (or perhaps edit the fetchspec if you never want other bits). If you must, see !fetch4why22:05
→ maletor joined22:05
canton7 but that's not the problem here22:05
rubious what?22:05
!fetch4why22:05
gitinfo Firstly, the four-word version of git-fetch does not update the remote-tracking branches or tags. The default git fetch origin master fetches into FETCH_HEAD - ie the equivalent of git fetch origin master:FETCH_HEAD. If you use pull in this case, it will then do the equivalent of git merge FETCH_HEAD. Even updating the remote tracking branch with git fetch origin master:remotes/origin/master it still doesn't update tags22:05
rubious ah22:05
well, either way22:05
Still can't push22:05
canton7 yeah, it's a valid point, but not your problem22:05
← kerrick left22:05
canton7 where are you trying to push to? what's the output of 'git push'?22:05
rubious Does it matter than I'm working in a submodule?22:05
canton7 maybe. check you're on a branch22:06
submodules have a nasty habit of detaching the HEAD22:06
rubious nope22:06
canton7 (well it's by design, but it catches people off-guard)22:06
rubious oooh, yeah, no branch.22:06
But it's a submodule? It shouldn't be on a branch of its umbrella project, right?22:06
canton7 way off the mark :P22:06
rubious canton7: Not sure what that means22:07
← dagerik left22:07
← monsieurlu left22:07
canton7 the supermodule ("umbrella project") tracks the commit which the submodule is at. when you run 'git submodule update', or one of the other commands which updates submodules, the submodule's HEAD is detached, and points at that commit22:07
rubious canton7: I'm getting this terrible feeling I'm missing something supoer obvious22:07
ahhhh22:07
canton7 so the branches inside the submodule are utterly unrelated to the supermodule22:08
← chuckharmston left22:08
rubious wow, yeah22:08
totally detached from head22:08
had to checkout master in my submodule22:08
← pandeiro left22:08
rubious These things can be kind of a pain, amirite22:08
canton7 there're some options: submodule.$name.update and 'git submodule update --rebase or --merge', which can keep the branches attaches22:09
*HEAD attaches22:09
rubious Nice22:09
canton7 yeah, they're not likely by everyone :P but once you understand what they're doing and how they're doing it, they make more sense22:09
rubious Just to clarify, I can use branches in a submodule, right?22:10
canton7 !gitslave and !subtree anyway, fyi22:10
gitinfo gitslave (http://gitslave.sf.net) is useful to add subsidiary git repositories to a git superproject when you control and develop on the subprojects at more or less the same time as the superproject, and furthermore when you typically want to tag, branch, push, pull, etc. all repositories at the same time.22:10
The subtree merge method is great for incorporating a subsidiary git repo into your current one with "unified" history. Read http://git-scm.com/book/en/Git-Tools-Subtree-Merging for more info, or try one of the !subtree_alternatives22:10
canton7 rubious, yeah sure. just expect the HEAD to be detached if you update the submodule22:10
rubious canton7: yeah, just read about those two in the git submodule readme under "haha, so here's other stuff that's just as useful and maybe not as weird".22:10
canton7 they have different aims, so pick whatever suits your usecase22:11
rubious canton7: cool, thanks very much22:11
canton7 no worries. you're probably the quickest person to grasp that particular problem I've had :P22:11
just to make sure: if you're updating a submodule, you know to cd back to the supermodule, add the submodule, commit?22:12
(that updates the supermodule's record of which commit the submodule's at)22:12
→ mritz_ joined22:12
→ branch joined22:13
← dvide left22:17
→ weiyang joined22:17
← jaseemabid left22:18
← macabre left22:19
→ _adnasa_ joined22:19
← _adnasa_ left22:19
← jwmann left22:20
→ _adnasa_ joined22:20
← eletuchy left22:20
→ thursday joined22:20
← mritz_ left22:20
← thursday left22:21
← gtrak left22:21
← nitrix left22:22
→ nitrix joined22:22
← nitrix left22:22
→ eletuchy joined22:23
→ Destos joined22:26
→ lun_ joined22:27
→ stephenjudkins joined22:27
→ chuckharmston joined22:28
→ nitrix joined22:28
← pppaul left22:29
→ chuckhar_ joined22:30
← chuckharmston left22:30
→ tinti_ joined22:31
→ pppaul joined22:32
stephenjudkins is there a way to suppress the "detached head" warning when using `checkout`?22:34
← hidekin1 left22:35
← goshawk left22:35
EugeneKay stephenjudkins - '2>/dev/null' on the command, or checkout to a branch(using '-b foobranch') instead of detaching22:36
stephenjudkins EugeneKay: thanks22:36
← bfig left22:39
← Destos left22:40
→ tjholowaychuk joined22:40
← tmz left22:41
← sgambino left22:44
← pantsman- left22:44
← boombatower left22:44
← jason237 left22:45
→ lustucru joined22:46
← AdrienBrault left22:47
← psoo left22:47
→ jimm joined22:48
← jimm left22:48
→ jimm joined22:48
→ AdrienBrault joined22:50
SeoZ-work[AWAY] → SeoZ22:51
← revagomes left22:51
← Protux left22:53
ameoba merging master into a branch for testing purposes & the merge completely flopped. In one file, it just dropped an entire method body. Is there anything to do other than hand merge from two clean checkouts?22:54
← mritz left22:54
← branch left22:54
→ bfig joined22:54
→ tmz joined22:55
→ Rolybrau joined22:56
← axl_ left22:56
banghouse → banghouseFOOD22:57
← paul_andrew left22:57
← dzonder left23:00
← Rolybrau left23:02
← _Vi left23:03
→ _Vi joined23:04
→ BiggFREE joined23:04
→ Hasbro joined23:05
← TeckniX left23:05
← Hasbro left23:06
→ Hasbro joined23:06
→ mischat joined23:06
← thiago left23:07
← chuckhar_ left23:08
← Davey left23:08
→ Hans_Henrik joined23:09
← trungqtran left23:09
Hans_Henrik installing git on windows, it ask me if i want "Checkout windows-style, commit unix-style line endings", or "checkout as-is, commint unix-style" or "checkout as-is, commit as-is", now, if i happen to put binary files, or any files that git does not understand, any chances it will start corrupting my files?23:10
← AdrienBrault left23:10
← obiwahn left23:10
← pfFredd left23:11
← dangerousdave left23:11
→ AdrienBrault joined23:11
→ Davey joined23:11
→ Hemebond joined23:11
Hemebond I need help understanding a "git diff". I've done "git diff upstream master" and it shows a function being removed from a file. But that function doesn't exist in my file or the file in upstream/master. What does it mean?23:12
Hans_Henrik whatever the case, aparently Git installer "requires that i shut down XChat irc client" to update some files during install x.x byebye23:13
← Hans_Henrik left23:13
← orafu left23:14
← lustucru left23:15
Hemebond ah23:15
nevermind23:15
I had the command wrong23:15
again23:15
Will I ever learn how to "fetch" properly?23:16
cmn fetch everything23:17
milki fetch the world23:17
dunpeal hack the planet23:17
milki CONQUER ALL23:17
→ jnewt joined23:17
← hwrd|work left23:17
→ hwrd|work joined23:18
jnewt i'm getting an error when trying to push. error: failed to push some refs...non-fast-forward updates were rejected. then i read the section about this in git push --help, where they discussed the portion in the context of two or more people using the remote. i'm the only one using it. i don't understand why i'm getting this, and if i do pull, what will i be getting?23:19
← madsy left23:19
← jpreyer left23:19
canton7 jnewt, take a look! git fetch origin && git log --graph --oneline --decorate --all23:20
dunpeal jnewt: I bet you rebased23:21
cmn or ammended, or reset23:21
→ Destos joined23:22
jnewt I agree i've probably done one of those things (rebase would be my guess). if i git fetch origin, will it change my local history?23:23
canton7 nope. fetching updates the remote-tracking branches: the branches starting with .e.g origin/, which are local records of the states of remote branches23:24
← Fandekasp left23:25
→ Fandekasp joined23:26
dunpeal Hey. anyone ever encounter this error message on fetch: 'git fetch' command failed. stderr: Short read of block.23:26
← stephenjudkins left23:27
← kvanderw left23:28
→ BSaboia joined23:28
cmn your network connection probably got cut23:28
banghouseFOOD → banghouse23:29
→ stevedomin joined23:29
← zastern left23:29
dunpeal mhm, I think it's JGit messing up again. Love those error reports from users not even specifying what they're using.23:30
← SirCmpwn left23:30
cmn there are some weird-ass git servers out there23:30
→ mritz joined23:30
cmn yesterday I was trying to clone nuget and it would always finish fine, but somehow get an error from the server making it delete the repo23:31
← weiyang left23:31
dunpeal yeah, I've never had many issues with the vanilla C git clients over vanilla SSH23:31
even though many of my users insist on stuffing their repos with gigabytes of binaries.23:32
cmn that would just need a load of RAM or CPU if the files are large23:32
dunpeal yeah, the boxes are reasonably fat, so it's not an issue23:33
both servers and clients are reasonably powerful workstations. that said, the servers aren't server-class boxes, and they still work well.23:33
→ GregZen joined23:33
← pppaul left23:34
← Hemebond left23:34
cmn with the cloud and all, it seems that most servers aren't more powerful than your average laptop these days23:35
← AdrienBrault left23:35
→ stephenjudkins joined23:35
→ AdrienBrault joined23:36
→ pppaul joined23:36
→ chuckharmston joined23:37
dunpeal well, laptops are ridiculously powerful. my MBPr has 16GB RAM, a quad core i7 with a maximum cycle speed of 3.7GHz, and an SSD that's faster than most server-class HDDs.23:37
GregZen Consider a relatively small central Git server with no more than 500 repositories. I am trying to create a graph to show submodule dependencies. It is easy for me to show the dependencies for each repository, but I want to go the other way. Given a repository, how would I show all of the repositories that depend on it?23:37
cehteh my new laptop is way more powerful than our server23:38
cmn I wouldn't call that an average laptop, though23:38
cehteh which is a small 12" laptop :P23:38
→ cannonball joined23:38
dunpeal cmn: if not for Apple's overpricing, you could get that for a bit over $200023:38
→ SirCmpwn joined23:39
cmn though possibly with less build quality and whatnot (they're expensive but are pretty sweet machines)23:39
dunpeal GregZen: visit each and construct a graph.23:39
cmn: yeah, I agree. it's beautiful hardware, especially since you can't get anything quite like the Retina display anywhere else.23:40
← peterpan left23:40
dunpeal "beautiful" pretty much sums it up though... I don't think it's incredibly more reliable than, say, a good lenovo.23:40
cmn I have a pretty high-resolution display on my 13', not retina, but definitely better than most (the laptop is also more expensive than most)23:41
dunpeal but yeah, finish quality of everything is top notch, everything always works, very reliable hardware even under stress, while staying super light and small.23:41
← chuckharmston left23:41
dunpeal cmn: what's your native resolution?23:41
cmn GregZen: that's just a query in your DB, store the relationship and and ask "what repos have this one as dep"?23:42
→ simesy joined23:42
GregZen dunpeal: yeah, that's one way… periodically go through all the repos. I'm considering doing it more real time by only looking at a repository when a repository is changed, via a post-receive hook.23:42
→ Goplat joined23:42
→ cameroncb1` joined23:42
cmn dunpeal: it's somethingx1080, basically full HD, which IIRC is the limit for DVI, which is is the likely connection that it has to the GPU23:42
dunpeal GregZen: sure, that would work as well23:42
← pppaul left23:42
← tjadc left23:43
dunpeal cmn: cool, is it a Zenbook by any chance? I thought of getting one of those instead of an Air for an ultra-portable.23:43
GregZen dunpeal, cmn: ok. figured I would have to go that route. thanks.23:43
cmn dunpeal: yeah, it's the Zenbook prime, it was either this one or an Air, but because I'm in Europe, the price for the similarly-spec'd Air was ~2400 EUR23:45
← cenuij left23:45
bremner hah, revenge of the colonies23:45
cmn more like Apple being bastards23:45
dunpeal cmn: yeah, Apple is way overpriced, in the US too, though the difference is not as jarring23:45
cmn even if it were $2400 it would be significantly cheaper23:46
dunpeal you ALWAYS pay more for the same hardware when it's from Apple23:46
← scarlisle left23:46
dunpeal to be fair, at least in the current generation, like for the MBPr, you can actually get current generation top notch hardware, even if you do have to pay through the nose for it23:47
← cameroncb1` left23:47
→ cameroncb1` joined23:47
← Chaser left23:47
dunpeal last generation, the best i7 you could get was several notches below the top, which other laptops (e.g. Lenovo) featured.23:47
→ PatF joined23:48
dunpeal that said, I don't know that you can get the same build quality anywhere else, though I thought the Zenbook came close23:48
← cameroncb1` left23:48
dunpeal cmn: which Linux did you install on it?23:48
→ cameroncb1 joined23:48
cmn Debian23:48
I'm pretty happy with it, just a shame you can't put more RAM on it23:49
dunpeal which Debian? does it work well? I want to get one myself and install Ubuntu on it.23:49
cmn Debian sid, the only real one :p23:49
you have to build the kernel yourself right now, as there's a missing patch23:49
dunpeal does sleep / brightness / audio / volume / all the other small laptopy things work well?23:49
cmn but after the next stable release, the patch should go right in, though23:50
← pdtpatrick left23:50
← Ryback_ left23:50
← Destos left23:50
PatF → Destos23:50
← AdrienBrault left23:50
cmn you need the patch that adjusts to the new address ASUS decided to use for the BIOS stuff, once that's in, only backlight hardware keys don't work23:50
← stevedomin left23:51
cmn because they're implemented in some backward way23:51
dunpeal wow, that's better than I expected for Debian, not a classic laptop distro23:51
→ storrgie joined23:51
cmn it's mostly just good ACPI support on part of the hardware maker and the kernel23:51
← chrisgeorge left23:52
← mritz left23:53
Layke → chris_acky23:53
chris_acky → Layke23:53
→ ecksit joined23:54
→ sgambino joined23:54
ecksit hi, having a small issue with my .gitignore not stopping the files from being tracked. https://gist.github.com/5306948f2ae0d59664b5 is my tree listing and the gitignore.23:54
cmn ignore rules don't stop files from being tracked23:55
dunpeal cmn: cool, since you had goo experience, I think I'm going to buy one too.. thanks!23:55
← GregZen left23:55
stephenjudkins suppose I have a git repo somewhere. I have a full directory somewhere else. I'd like to check the entire directory into the git repo as a branch or a commit. is this possible?23:55
milki ecksit: git ignore doesnt work for already tracked files23:55
EugeneKay ecksit - !untrack23:55
gitinfo To remove a file from git's tracking, without deleting it from your working tree, `git rm --cached <file>`. Note that any repo which pulls this change will delete their local copy of that file. You can "bring it back" using `git checkout HEAD^ file`23:55
ecksit ah!23:55
→ GregZen joined23:55
cmn stephenjudkins: copy the files over and commit23:55
← Akufen left23:55
→ rchavik joined23:56
← rchavik left23:56
→ rchavik joined23:56
stephenjudkins cmn: is it possible without copying the files? as in, tell git to treat a different directory as its "working tree" for the purposes of one commit?23:56
milki o thats interesting23:56
could try that23:56
← gdelfino left23:56
← GregZen left23:57
milki GIT_WORK_TREE and create an orphanzed branch?23:57
← optixx left23:57
cmn stephenjudkins: yeah, but then subtle things happen, see man git23:57
gitinfo stephenjudkins: the git manpage is available at http://git-scm.com/docs/git.html23:57
→ optixx joined23:57
→ Esa_ joined23:57
→ wh1rr joined23:58
→ adamm joined23:59
stephenjudkins milki: oh, that's exactly what i'm looking for. thanks.23:59

Logs Search ←Prev date Next date→ Channels Documentation