| 2010-07-23 |
| → jmcantrell joined | 00:00 |
| ← hasrb left | 00:00 |
| → explody joined | 00:01 |
|
frogonwheels
| hsa2: ok.. on ubuntu it's got stuff in it. | 00:02 |
| hsa2 → hsa2|gone | 00:03 |
|
murz
| now im getting "fatal: pathspec 'origin' did not match any files " when i try to add orign >_< | 00:03 |
| ← p3rg2z left | 00:03 |
|
explody
| is there some way to recursively not ignore a directory, particularly if I don't always know what the subdirectory structure will be? Same as this guy's problem, which doesn't really have an answer yet: http://is.gd/dCHdB | 00:03 |
|
frogonwheels
| hsa2|gone: and it appears under http://localhost/gitweb/ | 00:04 |
|
| explody: did you try !/test ? | 00:05 |
|
explody
| yeah | 00:05 |
|
frogonwheels
| explody: or perhaps a new .gitignore in test with !* ? | 00:05 |
|
explody
| actually I think the last answer in that link might work | 00:06 |
|
frogonwheels
| explody: also I believe that 'absolute' pathnames are relative to the top of the repo. | 00:07 |
| → rlb3_ joined | 00:07 |
|
explody
| hm everything half-works, it still seems like I have to specify all the subdirs under a directory that I want to !ignore | 00:08 |
| ← steveire left | 00:09 |
| rlb3_ → rlb3_home | 00:09 |
| ← ajpiano left | 00:11 |
| → developernotes joined | 00:11 |
| ← developernotes left | 00:11 |
| ← gdw2 left | 00:12 |
| ← Yuuhi left | 00:12 |
|
frogonwheels
| explody: !**/* ? | 00:12 |
| → mitnavn joined | 00:13 |
|
mitnavn
| Can you revert a reversion? :P | 00:13 |
| frakturfreak_ → frakturfreak | 00:14 |
|
AAA_awright
| mitnavn: What does that mean to you? | 00:15 |
|
explody
| frogonwheels: no love, ignore exclusions just aren't recursive, I don't think | 00:15 |
| → steveire joined | 00:15 |
| ← steveire left | 00:15 |
| → steveire joined | 00:15 |
|
mitnavn
| AAA_awright: Get the content of the file back to what it was before reverting. | 00:16 |
| hsa2|gone → hsa2 | 00:16 |
|
AAA_awright
| git checkout the old revision | 00:16 |
| ← sh1mmer left | 00:17 |
|
mitnavn
| Thanks, I will try that. | 00:18 |
|
AAA_awright
| There is git checkout, git reset, or git revert depending on what you want to do to your checkout or the history | 00:18 |
|
murz
| anyone know what "'myrepo.git' unable to chdir or not a git archive " means? | 00:18 |
|
hsa2
| murz, did you initialize a git repo? | 00:19 |
|
murz
| yes | 00:19 |
|
AAA_awright
| .git should exist in your directories or one of the parent directories | 00:19 |
| ← hobodave left | 00:19 |
|
hsa2
| you should enter the path right then | 00:19 |
|
murz
| i tried myrepo.git and /home/git/repositories/myrepo.git (where it is on my filesys) | 00:21 |
| → agile joined | 00:22 |
| ← steveire left | 00:23 |
| → steveire joined | 00:24 |
| ← steveire left | 00:24 |
| → steveire joined | 00:24 |
| → sanjoyd joined | 00:24 |
| hsa2 → hsa2|gone | 00:25 |
| ← legind left | 00:26 |
| ← d0k left | 00:27 |
| ← redbrain left | 00:28 |
| → vals_ joined | 00:28 |
| ← Joeconyers left | 00:29 |
| → Joeconyers joined | 00:29 |
|
sitaram
| Oxf13: how about this weekend? I'll send you email about what tha escape hatch will do (I'll try and keep it as generally useful as possible while serving your purpose; if that doesn't work, I'll make it a patch just for your use) | 00:29 |
|
murz
| what about this one: fatal: pathspec 'origin' did not match any files | 00:30 |
|
sitaram
| Oxf13: oh and BTW the "-s" thing is already in "master" | 00:30 |
|
Oxf13
| sitaram: yeah, I foudn that in master. And this weekend is fine. | 00:30 |
|
sitaram
| cool | 00:30 |
| ← tango_ left | 00:30 |
| ← mitnavn left | 00:31 |
| → stringoO joined | 00:31 |
| ← hebz0rl left | 00:34 |
| → scarabx joined | 00:36 |
| → SlickT10 joined | 00:36 |
| → simplechat joined | 00:36 |
| → anathematic joined | 00:37 |
| ← tobiassjosten left | 00:37 |
| ← steveire left | 00:42 |
| → steveire joined | 00:43 |
| ← tildeequals left | 00:43 |
| ← Luyang left | 00:44 |
| → jtauber joined | 00:46 |
| ← jtauber left | 00:46 |
| ← rlb3_home left | 00:47 |
| → thwart joined | 00:50 |
| ← ricky left | 00:51 |
| ← mva left | 00:51 |
| → mva joined | 00:52 |
| ← anathematic left | 00:54 |
| ← stringoO left | 00:59 |
| → cirwin1 joined | 01:01 |
| → anathematic joined | 01:01 |
| ← cirwin left | 01:03 |
|
Oxf13
| sitaram: still around? Got a bit of a logistical ACL issue to pick your brain about | 01:04 |
|
sitaram
| sure | 01:04 |
| → bcardarella joined | 01:04 |
| ← bcardarella left | 01:04 |
|
Oxf13
| background. | 01:04 |
|
| We have releases, a la Fedora 13, Fedora 12, etc.. We also have "rawhide" | 01:05 |
|
sitaram
| ok | 01:05 |
|
Oxf13
| Rawhide is our ever moving forward development tree, and we "branch" from rawhide when we start to stabalize and prepare a release | 01:05 |
| → ricky joined | 01:05 |
|
Oxf13
| rawhide maps to origin/master | 01:05 |
|
| our releases are polished, released and updated from origin/<release>/master | 01:06 |
|
| eg origin/f13/master | 01:06 |
|
| We allow developers to create their own branches from the release branches to deal with concurrent issues or potentially broken things | 01:06 |
|
| but we require them to use the naming scheme such that if htey wanted to work on a new feature for Fedora 13, they'd use origin/f13/newfeature | 01:07 |
|
sitaram
| wait... dont you mean f13/master? (origin is just the remote, so yes that's what it looks like on a clone). Or do you really have a branch name that has "origin" in it? | 01:07 |
| ← aziz left | 01:07 |
|
Oxf13
| yes, sorry f13/master | 01:07 |
|
| I put origin in there to express remoteness | 01:07 |
|
sitaram
| sure ok... just wanted to make sure I understood; go on... | 01:07 |
|
Oxf13
| We force the naming convention so that my client side tool can look at the local working branch, determine its remote merge point, and deduce the Fedora release from that. | 01:08 |
|
| so local "frobiz" branch merges to origin/f13/frobiz, which I can reduce down to just "f13" | 01:08 |
|
| the ACL structure I have gives people RW access within the repo to "f13" or "f14" or "master" | 01:09 |
| → lanthan_ joined | 01:09 |
|
Oxf13
| If I understand matching, RW access to "f13" would match a write attempt to f13/frobiz or f13/master | 01:09 |
|
sitaram
| yes | 01:09 |
|
Oxf13
| ok, here comes the tricky part | 01:10 |
|
| What is somebody wants to make a branch of rawhide, that is a direct branch of master? | 01:10 |
|
| conceptually they'd just do a local branch of 'rebase', and map that to origin/rebase | 01:10 |
| ← sergio_ left | 01:10 |
| → indygreg joined | 01:10 |
| ← steveire left | 01:11 |
| ← murz left | 01:11 |
| ← sanjoyd left | 01:11 |
|
sitaram
| what would that be in git terms? i.e., what git operation do you want to prevent? | 01:11 |
|
Oxf13
| that's not going to match anything we currently list for ACL | 01:11 |
| ← Tommy[D] left | 01:12 |
|
Oxf13
| people that have RW to master should be able to create branches of master, people without write access to master should not be able to | 01:12 |
| → sergio_ joined | 01:12 |
| ← lanthan__ left | 01:12 |
| ← justin-george left | 01:12 |
|
Oxf13
| I can't think of a way to work this out without forcing branches of master into their own directory eg rawhide/rebase | 01:13 |
| → steveire joined | 01:13 |
| ← steveire left | 01:13 |
| → steveire joined | 01:13 |
|
sitaram
| aah that has to be 2 rules, and adding a couple more for other branches; maybe you want to start by looking at http://github.com/sitaramc/gitolite/blob/pu/doc/3-faq-tips-etc.mkd#separating_create_and_push_rights | 01:13 |
| ← adamm left | 01:14 |
| → Tommy[D] joined | 01:14 |
| ← Vampire0 left | 01:14 |
|
sitaram
| actually I'm not sure what you mean by "rebase" up there. Is that an arbitrary branch name? | 01:15 |
|
Oxf13
| sitaram: arbitrary | 01:16 |
|
sitaram
| Oxf13: you can't prevent what a branch is *based* on via gitolite, only the *name* of the branch that is *pushed* | 01:16 |
|
Oxf13
| Ok, so for a user to be able to create a branch, they need "RWC" ? | 01:16 |
|
sitaram
| Oxf13: well... it's a little tricky. If any ACL in the repo has a "C" in it somewhere, then yes. Else no | 01:16 |
|
| Oxf13: backward compat, you know...! | 01:16 |
|
Oxf13
| eg, for "bob" to create a branch "f13/newbob", he'd need "RWC f13 = bob" ? | 01:17 |
|
sitaram
| or f13/ instead of f13, yeah | 01:17 |
|
Oxf13
| does the / matter? | 01:17 |
|
| I suppose it would allow him to do something like f13foobar | 01:17 |
|
sitaram
| not really; just being pedantic. Maybe leaving it is useful once you get to fedora core 131 :-) | 01:17 |
|
| oh yeah that too | 01:18 |
|
Oxf13
| that's going to be somebody else's problem :) | 01:18 |
|
sitaram
| Oxf13: back to the branch called "rebase"; you cannot prevent "git checkout -b rebase master; git push origin rebase" per se. All you can do is restrict him from pushing a branch called "rebase" | 01:18 |
| → hobodave joined | 01:18 |
|
indygreg
| I'm trying to figure out how to push the entire contents of one repository into a subdirectory of an existing repository while preserving full commit histories and could use some help. there is only one branch on the repository being exported | 01:18 |
|
Oxf13
| sitaram: that's fine. | 01:19 |
|
| sitaram: I don't care what crazy ass things they do to get content into the branch, I basically care about keeping the separation in place. | 01:20 |
| ← leo2007 left | 01:20 |
|
Oxf13
| sitaram: for a repo named foo, there is master, f13/master, and el6/master | 01:20 |
|
| there is users bob and jane | 01:20 |
| ← Mage left | 01:20 |
| → adamm joined | 01:21 |
|
Oxf13
| bob write access to master and f13/ but not el6 | 01:21 |
| ← squi left | 01:21 |
|
Oxf13
| jane has write access to el6/ but nothing else | 01:21 |
| → Yuffster_ joined | 01:21 |
| → Mage joined | 01:21 |
|
sitaram
| Oxf13: by "write" you include "write or create new"? | 01:22 |
| ← Yuffster left | 01:22 |
|
Oxf13
| sitaram: without forcing bob to use rawhide/ as the place to push branches meant to track rawhide (master), I can't think of a way to let him create those branches without also having him have access to el6/ | 01:22 |
|
| sitaram: yes, in this context I do | 01:23 |
|
sitaram
| without also having him access to el6/ in terms of write or create? | 01:23 |
|
Oxf13
| both really | 01:23 |
|
| it would seem that in order for bob to write or create new top level branches he would need "RWC" to the entire repo | 01:24 |
|
sitaram
| Oxf13: ok one final question: is there *any* situation in which someone has write access to a branch but *not* the access to *create* such a branch | 01:24 |
|
Oxf13
| that is branches that are at the same level of master | 01:24 |
|
sitaram
| its possible, except I'm not understanding very precisely what you want that's all | 01:25 |
|
| let me try writing up something... give me 2 min | 01:25 |
|
Oxf13
| sitaram: no, the way our logic works is that if you are allowed to write to a release (eg you can commit stuff for Fedora 13), you're subsequently allowed to create branches for that release as well | 01:25 |
| ← Yuffster_ left | 01:25 |
| ← enherit left | 01:26 |
| → Yuffster joined | 01:26 |
| ← parasti left | 01:26 |
|
Oxf13
| the more I think about this though, the more I allowing people to create branches at the same level as master is a bad idea, because somebody could create the f14 branch too early | 01:27 |
|
| perhaps only git admins should be able to create high level branches, but allow people with write access to master the ability to create rawhide/ branches | 01:28 |
| → Remixman joined | 01:28 |
| → Vampire0 joined | 01:28 |
| ← kpreid left | 01:29 |
| → stringoO joined | 01:30 |
| anjen → food_anjen | 01:31 |
| → byang joined | 01:33 |
|
sitaram
| Oxf13: http://pastebin.com/S5T4yEZa | 01:33 |
| → gbacon joined | 01:33 |
|
sitaram
| oh wait | 01:34 |
|
| Oxf13: http://pastebin.com/LYQcf9t6 -- does that help? | 01:35 |
|
| Oxf13: remember the order matters | 01:35 |
|
| Oxf13: and my "u1" is a super admin who can *create* any branch so that people who are not allowed to create can then push to it | 01:36 |
|
Oxf13
| sitaram: ok, take that one step further. | 01:36 |
| ← hipe_ left | 01:37 |
|
Oxf13
| how would you allow u2 to create branches at the same level of master, but not write to el6/ ? | 01:37 |
| → hipe joined | 01:37 |
|
ju2wheels
| quick question... is projects.list supposed to have two entries per project? Im looking through gitweb code and it looks like its expecting the project name plus the owner but my projects.list generated by gitolite only shows the project name and i think it may be causing my gitweb issues.... | 01:38 |
| ← yngress left | 01:38 |
|
sitaram
| Oxf13: http://pastebin.com/Adgq9pn9 | 01:39 |
|
Oxf13
| ok, I think I see where you're going with that. | 01:39 |
|
| btw, what's with the extra master in there? | 01:40 |
|
| # | 01:40 |
|
| - master f[0-9] = u2 # cant create f14 early | 01:40 |
|
sitaram
| Oxf13: typo/thinko; ignore it :) | 01:40 |
|
Oxf13
| ok. | 01:40 |
| ← rfx left | 01:40 |
|
Oxf13
| so basically I need to have a RW for anything a user has write access to, then a - for any reserved branch names | 01:42 |
|
| let me take that back. | 01:42 |
|
sitaram
| correct, and as I said the order matters (it stops at the first match for the operation being performed or a "-") | 01:42 |
| ← hipe left | 01:42 |
|
Oxf13
| an RWC to any release a user has write access to, a - for all reserved branch names, and finally a RWC to everything, should they have write access to rawhide. | 01:42 |
| → masonkramer joined | 01:43 |
|
Oxf13
| I think I can work with that. | 01:43 |
| → SRabbelier1 joined | 01:43 |
| ← SRabbelier left | 01:43 |
|
Oxf13
| my acl generation script will get a bit more ... complicated | 01:43 |
| ← Mage left | 01:43 |
|
sitaram
| Oxf13: aah well, you could always go with what you said earlier: | 01:44 |
|
| 06:57:17 < Oxf13> the more I think about this though, the more I allowing people to create branches at the same level as | 01:44 |
|
| master is a bad idea, because somebody could create the f14 branch too early | 01:44 |
| → Mage joined | 01:44 |
| ← saiyr9000 left | 01:44 |
|
Oxf13
| sitaram: that's a fallback position | 01:44 |
|
| I like what you've come up with though | 01:44 |
|
sitaram
| Oxf13: well this feature didn't exist last December so you would not have seen it. Remember me telling you gitolite moved a heck of a lot since then? ;-) | 01:45 |
|
Oxf13
| indeed | 01:45 |
|
| just to confirm, what's the difference between "RWC" and "RW+C" ? | 01:45 |
|
sitaram
| the "+" indicates rights to rewind (force push) | 01:46 |
|
Oxf13
| ah right. No chance in hell :) | 01:46 |
|
sitaram
| Oxf13: erm... much as I am confident of gitolite doing the right thing, once you have come up with a canonical ruleset [for *one* repo, I mean], could you email me so I can test it thoroughly? | 01:46 |
|
Oxf13
| I have to make immutable really really mean immutable. Legal issues | 01:46 |
|
sitaram
| sure | 01:46 |
|
Oxf13
| sitaram: indeed! | 01:46 |
|
sitaram
| no "+" then | 01:46 |
|
Oxf13
| an RWC on 'master' just means they could make a branch named "masterblaster" right? | 01:47 |
|
| no other foreseen downsides to RWC master ? | 01:47 |
|
sitaram
| yup. | 01:47 |
|
| you can use "RWC master$" instead, but other than the *first* push the "C" no longer has any meaning anyway | 01:48 |
|
| so if you can pre-create all the "sacred" branches at appropriate times by user "u1" in my example, you dont need a "C" on the sacred branches for anyone else | 01:48 |
| ← mar77i left | 01:48 |
|
sitaram
| but be sure to have a C for the "u1" user or group. Remember, if there is not a single "RWC " or "RW+C", it falls back to backward compat behavious (RW implies C) | 01:49 |
| ← cirwin1 left | 01:50 |
|
Oxf13
| yeah, all the repos will have a c | 01:50 |
|
| and there will be an repo @pkgs; RWC = @admins | 01:51 |
|
| pkgs is every repo, admins are the git admins | 01:51 |
|
sitaram
| ok | 01:52 |
|
| Oxf13: I really gotta go now; be back online in about 2 hours from work (I'm running very late today!) | 01:52 |
|
| I'll email you the patch stuff we discussed earlier sometime Saturday my time, which is prolly 12-ish hours ahead of you | 01:53 |
|
Oxf13
| sitaram: no worries, I gotta run for dinner. Have a good day, and thanks again! | 01:53 |
|
sitaram
| you're welcome! and bon appetit | 01:53 |
| agib|away → agib | 01:53 |
| agib → agib|away | 01:54 |
| ← ju2wheels left | 01:56 |
| ← henkboom left | 01:56 |
| → mdel joined | 01:59 |
| ← byang left | 01:59 |
| ← t0rc left | 01:59 |
| → byang joined | 01:59 |
| → henkboom joined | 02:00 |
| ← thwart left | 02:00 |
| ← comawhite left | 02:01 |
| ← steveire left | 02:02 |
| → steveire joined | 02:03 |
| ← steveire left | 02:03 |
| → steveire joined | 02:03 |
| → sh1mmer joined | 02:05 |
| → yaotti_aw joined | 02:05 |
| yaotti_aw → yaotti | 02:06 |
| ← tar- left | 02:11 |
| ← Blackguard left | 02:12 |
| → kpreid joined | 02:13 |
| ← Mage left | 02:13 |
| → Mage joined | 02:14 |
| ← sh1mmer left | 02:17 |
| hsa2|gone → hsa2 | 02:19 |
|
hsa2
| hi | 02:19 |
|
| i want a full rollback to an old commit | 02:19 |
|
| is it possible? | 02:19 |
| → cytrinox_ joined | 02:20 |
| ← thoraxe left | 02:21 |
| → mandazi joined | 02:22 |
|
hsa2
| any idea? | 02:22 |
|
mandazi
| hey all | 02:23 |
|
hsa2
| mandazi, hi | 02:23 |
|
mandazi
| im having some issues trying checkout a repo from heroku. geting this eror | 02:23 |
|
hsa2
| i want a full rollback to an old commit, is it possible; i said | 02:23 |
|
mandazi
| Agent admitted failure to sign using the key. | 02:23 |
|
| Permission denied (publickey). | 02:23 |
|
| fatal: The remote end hung up unexpectedly | 02:23 |
| ← cytrinox left | 02:23 |
| cytrinox_ → cytrinox | 02:23 |
|
mandazi
| i know i have a public key. i ran ssh-keygen -t rsa | 02:23 |
|
Ilari
| mandazi: Sounds like problem with local ssh agent. | 02:25 |
|
| mandazi: Try cloning without having ssh agent enabled? | 02:26 |
|
mandazi
| how do i do that | 02:27 |
| ← reflectivedev left | 02:27 |
|
thiago
| hsa2: git checkout that-commit | 02:27 |
|
| mandazi: what is the key called? is it id_rsa or id_dsa ? | 02:28 |
|
mandazi
| id_rsa | 02:28 |
| ← kpreid left | 02:28 |
| ← mva left | 02:29 |
|
thiago
| ssh -v into the server | 02:29 |
|
| does it print something? Or does it fail? | 02:30 |
|
mandazi
| im not sure if i can do this with heroku | 02:30 |
| ← agile left | 02:31 |
|
Ilari
| mandazi: The problem looks like client side ssh agent problem. Nothing to do with what runs on the server. | 02:32 |
|
hsa2
| thiago, i did checkout old version but can't commit and push :S | 02:32 |
|
mandazi
| sorry im really new to all this. been googling and all i find is i have to create an ssh key | 02:34 |
| ← brntbeer left | 02:34 |
|
mandazi
| how do i check the client side | 02:34 |
|
thiago
| hsa2: you can commit | 02:34 |
| → agile joined | 02:34 |
|
thiago
| hsa2: what's the error you get when you push? And how did you push? | 02:34 |
|
hsa2
| thiago, after running checkout, i should pull right? | 02:35 |
|
thiago
| hsa2: no | 02:35 |
| → mva joined | 02:35 |
|
thiago
| hsa2: is that repository shared with other people? | 02:35 |
|
hsa2
| for now, no, i am testing it | 02:35 |
|
thiago
| then "roll back" in this case means "throw away all later commits" | 02:35 |
|
hsa2
| yes | 02:36 |
|
| exactly | 02:36 |
|
thiago
| this means rewriting history, so you should use git push -f | 02:36 |
|
| though you probably wanted reset --hard to the commit, not checkout | 02:36 |
| ← flaguy48 left | 02:36 |
|
mandazi
| anyone on here use heroku? | 02:36 |
|
hsa2
| thiago, then i should run git reset --hard <commit> right? | 02:37 |
|
thiago
| hsa2: yes, but first you have to go back to the branch you had checked out first | 02:37 |
|
hsa2
| how? | 02:37 |
|
thiago
| so check out the branch you were in before | 02:37 |
| → hearit joined | 02:37 |
|
hsa2
| :S i am now confused | 02:37 |
|
| by branch you mean master right? | 02:38 |
|
| i think i have only one | 02:38 |
|
thiago
| most likely | 02:38 |
|
| if you have only one, then the one must be the branch you were in before | 02:38 |
| ← tedoc2000 left | 02:38 |
|
hsa2
| thiago, let me tell you | 02:38 |
|
thiago
| note you're not in a branch now if you executed that command I gave you | 02:38 |
|
hsa2
| i deleted everything in directory | 02:38 |
|
| then runned git init | 02:38 |
|
| then added remote and runned checkout | 02:38 |
|
thiago
| "ran" | 02:39 |
|
hsa2
| sorry :) | 02:39 |
| hearit → Iahc | 02:39 |
|
hsa2
| now, in this situation please help me | 02:39 |
|
thiago
| help you do what? | 02:39 |
|
hsa2
| my web application is broken now :P | 02:39 |
|
thiago
| that checkout, is it the command I gave you? | 02:39 |
|
hsa2
| i want to rollback to old commit reverting all other commits after that one | 02:39 |
|
| thiago, git checkout <commit> i ran | 02:40 |
|
thiago
| then run this: git checkout master | 02:40 |
|
| it will undo what I had told you to run | 02:40 |
|
| so let's start over: | 02:40 |
|
| are you ABSOLUTELY sure you want to delete, forever, those commits? | 02:40 |
| ← Elfe left | 02:40 |
|
hsa2
| yes! | 02:40 |
|
thiago
| it's not that they will be reverted. It's like they will never have existed. | 02:41 |
|
| then run: git reset --hard that-commit | 02:41 |
|
| the good commit | 02:41 |
|
hsa2
| i did | 02:41 |
|
frogonwheels
| hsa2: .. though it hasn't really gone until it gets garbage-collected, but then it really will be gone | 02:41 |
|
thiago
| now your local copy should be at the good state. Please verify that. | 02:41 |
|
hsa2
| thiago, ok, verified | 02:42 |
|
thiago
| now push it back to the server, with force | 02:42 |
|
| git push -f | 02:42 |
|
hsa2
| i don't know how, but it's still messed up | 02:45 |
|
| thiago, wait i have another question | 02:45 |
|
| you asked if it is shared | 02:45 |
|
| i am pulling from my web directory to make it synced | 02:46 |
|
| is my web directory in sync now? | 02:46 |
|
| no it's not | 02:47 |
|
thiago
| I didn't understand the question | 02:47 |
|
hsa2
| thiago, i am commiting from my pc to repo, then my public_html pulls from repo | 02:47 |
|
| by this way, i am keeping my project synced | 02:48 |
|
thiago
| ok | 02:48 |
|
hsa2
| now i'll delete my old public_html, and clone git repo | 02:48 |
| ← dreiss left | 02:49 |
|
thiago
| ok | 02:49 |
|
hsa2
| ohhh | 02:51 |
|
| thank god | 02:51 |
|
| thiago, thanks a lot | 02:51 |
| ← indygreg left | 02:52 |
| → timj__ joined | 02:52 |
| ← mandazi left | 02:52 |
| → unreal_ joined | 02:53 |
| ← __name__ left | 02:53 |
| → Elfe joined | 02:54 |
|
hsa2
| now i can sleep in peace :P | 02:55 |
|
| thiago, thanks again | 02:55 |
| ← unreal left | 02:55 |
|
hsa2
| see you all | 02:55 |
| unreal_ → unreal | 02:55 |
| hsa2 → hsa2|gone | 02:55 |
| ← timj_ left | 02:56 |
| → lazylisper joined | 02:58 |
| ← lazylisper left | 02:59 |
| → lazylisper joined | 02:59 |
| → whaley joined | 02:59 |
| ← byang left | 03:01 |
| ← unreal left | 03:03 |
| → kpreid joined | 03:03 |
| → unreal joined | 03:03 |
| → octavio-rdz joined | 03:05 |
| ← zalun left | 03:05 |
| ← dgs left | 03:05 |
| → byang joined | 03:06 |
| ← agile left | 03:08 |
| ← mdel left | 03:11 |
| ← gbacon left | 03:11 |
| → mdel joined | 03:11 |
| ← masonkramer left | 03:11 |
| ← stringoO left | 03:15 |
| → agile joined | 03:15 |
| ← byang left | 03:17 |
| → byang joined | 03:21 |
| ← zenazn left | 03:21 |
| food_anjen → anjen | 03:24 |
| ← byang left | 03:25 |
| → rchavik joined | 03:26 |
| → zenazn joined | 03:28 |
| → schlortoo joined | 03:29 |
| ← Mage left | 03:30 |
| ← welly left | 03:30 |
| → Mage joined | 03:31 |
| ← aresnick left | 03:31 |
| ← dr_win left | 03:31 |
| ← Remixman left | 03:32 |
| → Remixman joined | 03:33 |
| ← Remixman left | 03:33 |
| ← alester left | 03:33 |
| → byang joined | 03:34 |
| ← sagsousuke left | 03:34 |
| ← tstclair left | 03:34 |
| → shade_ joined | 03:35 |
| ← lazylisper left | 03:37 |
| → slakmagik joined | 03:39 |
| → hohoho joined | 03:41 |
| ← ajb left | 03:42 |
|
goddard
| is github the best place to host your project for free? | 03:43 |
| shade_ → SHADE\ | 03:44 |
| SHADE\ → shade\ | 03:44 |
|
slakmagik
| imo | 03:44 |
| ← programble left | 03:45 |
| → SamothUK joined | 03:45 |
| ← Bass10 left | 03:46 |
| → bradallen137_ joined | 03:47 |
| ← mw_ left | 03:51 |
| → enherit joined | 03:52 |
| ← Mage left | 03:52 |
| → Mage joined | 03:53 |
| ← scarabx left | 03:53 |
| → mithro joined | 03:55 |
| ← unreal left | 03:55 |
| → unreal joined | 03:56 |
| → jtauber joined | 03:57 |
| ← bradallen137_ left | 04:00 |
| yaotti → yaotti_aw | 04:04 |
| → dr_win joined | 04:04 |
| → programble joined | 04:05 |
| ← jmcantrell left | 04:06 |
| ← mithro left | 04:06 |
| → spiderr joined | 04:06 |
| ← Joeconyers left | 04:09 |
| → tildeequals joined | 04:10 |
| ← multigrain left | 04:10 |
| ← Mage left | 04:12 |
| ← spiderr left | 04:13 |
| → Mage joined | 04:13 |
| ← vinse left | 04:16 |
| ← shade\ left | 04:18 |
| ← Yuffster left | 04:18 |
| → shade_ joined | 04:19 |
| → marc_kdab_ joined | 04:20 |
| ← vmiklos left | 04:21 |
| → mithro joined | 04:21 |
| ← steveire left | 04:21 |
| → gnrfan joined | 04:21 |
| ← bartman left | 04:21 |
| → steveire joined | 04:22 |
| ← steveire left | 04:22 |
| → steveire joined | 04:22 |
|
gnrfan
| hope my question is ok for this channel.. first time here.. any hints on how to give someone access to a GIT repo in my Linux server over ssh without giving shell access? Been trying rssh + git patch but no luck for now :/ | 04:22 |
| ← byang left | 04:23 |
|
offby1
| gitolite. | 04:23 |
| → byang joined | 04:23 |
|
offby1
| next question. | 04:23 |
|
gnrfan
| hehe cool.. gonna check that out | 04:23 |
|
cbreak
| gitolite requires a shell account | 04:25 |
|
| just only one for everyone | 04:25 |
| → thwart joined | 04:27 |
|
gnrfan
| anyone got rssh to work with git_ | 04:28 |
|
offby1
| don't even know what "rssh" _is_. | 04:29 |
| ← SlickT10 left | 04:33 |
|
Oxf13
| sitaram: does whitespace matter (eg the number of spaces) in the gitolite.conf file, before compiling? | 04:34 |
|
cxreg
| cbreak: if you never get to an actual shell, is it still a shell account? | 04:36 |
|
gnrfan
| cool.. got it to work.. rssh + patch works nicely.. it was me.. had to add user to /etc/ssh/sshd_config.. sorry :/ | 04:37 |
| ← doener left | 04:37 |
| → doener joined | 04:37 |
|
gnrfan
| offby1: a very restricted shell in which you can allow scp, or sftp and with a patch git only | 04:38 |
| ← ProperNoun left | 04:39 |
| → ProperNoun joined | 04:40 |
| ← ProperNoun left | 04:40 |
| → ProperNoun joined | 04:40 |
| → radioman-lt joined | 04:42 |
| → galderz joined | 04:43 |
| ← galderz left | 04:43 |
| → galderz joined | 04:43 |
| ← ericindc left | 04:43 |
| ← dr_win left | 04:46 |
| ← radioman-lt left | 04:47 |
| → MattDiPasquale joined | 04:47 |
|
wereHamster
| gnrfan: well, there's gitolite and git-shell which you can configure as the default shell for the user and he then will only be able to access git repos | 04:47 |
| ← MattDiPasquale left | 04:47 |
| → MattDiPasquale joined | 04:47 |
| → psankar joined | 04:50 |
| ← hyperair left | 04:51 |
| cpg → cpg|away | 04:52 |
| → freimatz joined | 04:57 |
| → indygreg joined | 04:59 |
| ← galderz left | 04:59 |
| ← TDJACR left | 04:59 |
| → daaku joined | 05:00 |
| ← MattDiPasquale left | 05:01 |
| → welly joined | 05:02 |
| ← LRN left | 05:02 |
| ← anathematic left | 05:02 |
| ← shade_ left | 05:03 |
| → LRN joined | 05:03 |
| yaotti_aw → yaotti | 05:04 |
| ← carutsu_ left | 05:05 |
| → carutsu_ joined | 05:06 |
| ← thedeeno left | 05:06 |
| → mar77i joined | 05:06 |
| → comawhite joined | 05:07 |
| → hyperair joined | 05:07 |
| → thedeeno joined | 05:08 |
| → ProperN[out] joined | 05:09 |
| ← ProperN[out] left | 05:09 |
| → ProperN[out] joined | 05:09 |
| ← jmspeex left | 05:09 |
| ← slakmagik left | 05:10 |
| → jmspeex joined | 05:10 |
| ← ProperNoun left | 05:10 |
| ← hsa2|gone left | 05:10 |
| ← bartek left | 05:10 |
| → hsa2|gone joined | 05:11 |
| ProperN[out] → ProperNoun | 05:11 |
| ← Iahc left | 05:11 |
| → vu3rdd joined | 05:11 |
| → bartek joined | 05:12 |
| → hyperboreean joined | 05:12 |
| ← RobertMeta left | 05:12 |
| cpg|away → cpg | 05:13 |
| ← dropdrive left | 05:14 |
| → MetaCosm_ joined | 05:14 |
| → dropdrive joined | 05:14 |
| ← MetaCosm_ left | 05:15 |
| → TDJACR joined | 05:15 |
| → lhz joined | 05:15 |
| → MetaCosm_ joined | 05:15 |
| ← kpreid left | 05:17 |
| ← steveire left | 05:17 |
| → saiyr9000 joined | 05:18 |
| → steveire joined | 05:18 |
| ← TDJACR left | 05:19 |
|
Oxf13
| sitaram: http://fpaste.org/gLCY/ @pkgs has every package listed. (I have to fix the duplicates) | 05:20 |
| → TDJACR joined | 05:21 |
| ← TDJACR left | 05:21 |
| → TDJACR joined | 05:21 |
| → hasen joined | 05:24 |
| ← molokai left | 05:32 |
| ← TDJACR left | 05:33 |
| → molokai joined | 05:34 |
| ← gnrfan left | 05:35 |
| → roop joined | 05:37 |
| → slakmagik joined | 05:42 |
| ← Eaven left | 05:43 |
| → opalka joined | 05:45 |
| ← welly left | 05:51 |
| → Modius joined | 05:51 |
| ← Modius left | 05:52 |
| → skoop joined | 05:53 |
| → Modius joined | 05:53 |
| → aspotashev joined | 05:54 |
| ← psankar left | 05:55 |
| → g0 joined | 05:57 |
|
g0
| I'm compiling git to run from ~, I don't have the zlib1g-dev package on this system, I compiled the source for zlib1g-dev and installed it with ~/bin prefix | 05:58 |
|
| now how do I tell git's installation to look for zlib1g-dev there? | 05:59 |
|
| This is the firsst time i'm trying to install anything on ~ | 05:59 |
| → brntbeer joined | 05:59 |
| → shruggar joined | 06:02 |
| ← octavio-rdz left | 06:02 |
| ← schlortoo left | 06:03 |
| → benzap_ joined | 06:04 |
|
wereHamster
| LDFLAGS="-L$HOME/lib" | 06:05 |
|
frogonwheels
| g0: usually it will be a configure option. ./configure --help | 06:05 |
| → schlortoo joined | 06:05 |
|
frogonwheels
| or ask somebody who really knows. | 06:05 |
|
| :) | 06:05 |
|
wereHamster
| g0: which distro are you using? | 06:05 |
| → Dave^| joined | 06:05 |
| ← benzap left | 06:06 |
| → apsysoev joined | 06:07 |
| ← g0 left | 06:07 |
| ← hohoho left | 06:09 |
| ← aspotashev left | 06:09 |
| → hohoho joined | 06:11 |
| ← thwart left | 06:12 |
| ← pastie left | 06:12 |
|
doener
| in case he returns, it should be (IIRC): echo ZLIB_PATH=$HOME >> config.mak; make && make install | 06:13 |
| ← Haarg left | 06:13 |
|
doener
| assuming that $HOME was used with --prefix for the zlib build (it sounded like he used --prefix=$HOME/bin, though that would be really weird ;-)) | 06:13 |
| ← jones- left | 06:13 |
|
doener
| anyway, gotta go | 06:13 |
| → cbreak_work joined | 06:14 |
| ← brntbeer left | 06:16 |
| → ycyuxin joined | 06:16 |
| → jones- joined | 06:17 |
| → khmarbaise joined | 06:17 |
| → Belna joined | 06:17 |
| → chrislerum joined | 06:19 |
| → Haarg joined | 06:20 |
| → [Po]lentino joined | 06:21 |
| → The_Ball joined | 06:26 |
| → eighty4 joined | 06:27 |
| → carter joined | 06:28 |
| → leo2007 joined | 06:28 |
| → warthog9 joined | 06:31 |
| → fivetwentysix joined | 06:31 |
| ← chrislerum left | 06:33 |
| → poli joined | 06:35 |
| → psoo joined | 06:37 |
| → datapolitical joined | 06:39 |
|
datapolitical
| git-folk, how would I go about removing an update from a remote repository | 06:40 |
| ← mithro left | 06:40 |
|
cbreak_work
| update? | 06:40 |
|
| a commit? | 06:41 |
|
| you can force-push local stuff to remote branches | 06:41 |
|
| (if the remote allows it) | 06:41 |
|
| but of course, if someone else pulled, you get trouble | 06:41 |
|
datapolitical
| nobody else has | 06:41 |
| → j4cbo joined | 06:41 |
| ← unreal left | 06:41 |
|
datapolitical
| so how does force-push work | 06:41 |
| → unreal joined | 06:42 |
|
j4cbo
| i've got a repo (from a project in a class some time ago) that i'd like to save, with all history, in a compact form | 06:42 |
|
cbreak_work
| it's -f | 06:42 |
|
| read the man page | 06:42 |
|
j4cbo
| i've git gc'd it so everything is in a single .pack file; can i just save that file, and later reconstitute it into a proper repo with worktree? | 06:43 |
|
cbreak_work
| j4cbo: git clone --bare and then archive it? | 06:43 |
|
datapolitical
| I did, so it was a silly way to ask that question. force push doesn't delete files already on the remote, it seems | 06:43 |
|
frogonwheels
| datapolitical: so you first reset your head in the local repository to the commit (or rebase if you want to pick one update out and you don't want to just patch reverse it). | 06:43 |
|
j4cbo
| oh, clone to bare seems ideal, actuall | 06:44 |
|
| actually* | 06:44 |
|
cbreak_work
| datapolitical: of course not | 06:44 |
|
| remotes you can push to don't have files | 06:44 |
|
| and if they had, push would not update them anyway | 06:44 |
| ← ignatp left | 06:45 |
|
datapolitical
| wait, what? | 06:47 |
|
| a remote doesn't store files? | 06:47 |
|
| RTFM | 06:47 |
| ← didlix left | 06:47 |
|
frogonwheels
| datapolitical: no, what he means is it doesn't have a working directory | 06:47 |
|
| datapolitical: once a blob doesn't have any references, it will be eventually garbage collected | 06:48 |
|
datapolitical
| okay | 06:48 |
|
cbreak_work
| datapolitical: remotes you push to are bare. they have no working directory by definition | 06:48 |
|
frogonwheels
| datapolitical: You need to be a bit more precise about what you're after. | 06:48 |
|
| datapolitical: ie this commit you want to remove - is it the last commit(s) in the branch? is it an old commit? is it sensitive data? | 06:50 |
| → ljungk joined | 06:50 |
|
datapolitical
| it's the first | 06:51 |
|
| and it's sensitive data | 06:51 |
|
frogonwheels
| datapolitical: ok. is it a file you want to get rid of ? | 06:51 |
|
datapolitical
| I used the web interface to delete the repository, which didn't exactly work | 06:51 |
|
cbreak_work
| datapolitical: recreate the repository. | 06:51 |
|
| if you are really paranoid about the data | 06:51 |
|
datapolitical
| which I did | 06:51 |
|
| and the file was there | 06:51 |
|
| so, i've got a repository | 06:52 |
|
cbreak_work
| I mean, recreate it _without_that_file_ of course :) | 06:52 |
|
datapolitical
| unfuddle is being funny about that | 06:52 |
| → hearit joined | 06:53 |
|
datapolitical
| to which everyone replies, use github! | 06:53 |
| → frakturfreak_ joined | 06:54 |
|
ljungk
| I have two branches and there is a difference between a file. now I want to remove the difference. is there any way of letting git know that it came from the other branch? | 06:54 |
| → galderz joined | 06:54 |
| ← galderz left | 06:54 |
| → galderz joined | 06:54 |
| → cilly joined | 06:54 |
|
cbreak_work
| ljungk: no, git only merges complete branches | 06:54 |
|
| not only files | 06:54 |
| → dreinull joined | 06:55 |
| ← galderz left | 06:55 |
| → mithro joined | 06:55 |
|
ljungk
| cbreak_work: but isn't that what cherry-pick does? | 06:55 |
|
cbreak_work
| no | 06:55 |
| → priidu joined | 06:55 |
|
cbreak_work
| cherry-pick takes changes from a complete commit | 06:55 |
|
| merge takes a complete branch (a chain of commits) | 06:56 |
| ← carter left | 06:56 |
|
ljungk
| okay, so if this difference would have been from a commit, i could have used cherry-pick? | 06:57 |
|
cbreak_work
| a complete commit, yes | 06:58 |
| ← frakturfreak left | 06:58 |
|
cbreak_work
| you can just check out that file from the other branch if you are only interested in the contents, but that doesn't do history tracking | 06:58 |
|
ljungk
| cbreak_work: yeah, that's probably the only thing i can do since the changes have been there since the branching | 07:00 |
| → MattDiPasquale joined | 07:01 |
|
cbreak_work
| you can merge that past? | 07:01 |
|
| well, if you want the whole commit/history | 07:02 |
| → workmad3 joined | 07:02 |
| → aspotashev joined | 07:02 |
|
ljungk
| cbreak_work: i can't since they weren't the only differences in the branching, right? | 07:03 |
| → _stefanos_ joined | 07:03 |
| ← MattDiPasquale left | 07:04 |
| ← aspotashev left | 07:05 |
|
cbreak_work
| it would get all. | 07:05 |
|
| git tracks history on a per-commit basis. | 07:05 |
| ← warthog9 left | 07:09 |
| ← mva left | 07:11 |
| → steffkes joined | 07:11 |
| ← __ed left | 07:11 |
| ← priidu left | 07:11 |
| → priidu joined | 07:12 |
| ← hearit left | 07:12 |
| → gebi joined | 07:14 |
| → giallu joined | 07:15 |
| ← MetaCosm_ left | 07:16 |
| → cjs joined | 07:17 |
| → h32Lg joined | 07:17 |
| ← h32Lg left | 07:17 |
| → h32Lg joined | 07:17 |
| → didlix joined | 07:17 |
|
cjs
| What command I can I use that will print just the commit ID of the current working copy, and whether or not the working copy is modified? | 07:17 |
|
cxreg
| has there ever been talk about a sister channel to this one that talks about git.git hacking? | 07:17 |
|
| i think we need one | 07:18 |
|
cbreak_work
| cjs: git rev-parse can do the former | 07:18 |
|
| I think git diff can do the latter | 07:18 |
|
cjs
| Ah. Good. "git rev-parse HEAD" gives me half of what I need. | 07:19 |
| → mva joined | 07:19 |
|
cbreak_work
| --exit-code | 07:19 |
|
slonopotamus
| cxreg: git-devel? | 07:19 |
|
cbreak_work
| or even --quiet | 07:19 |
|
cxreg
| slonopotamus: ah. well. it's almost entirely empty, but ok :) | 07:20 |
| → trochala joined | 07:20 |
|
selckin
| can bad things happen when you edit files during svn dcommit? | 07:21 |
|
cbreak_work
| the edits might vanish | 07:21 |
| → warthog9 joined | 07:21 |
| → HG` joined | 07:21 |
|
cjs
| Hm. How do I get the current branch name? (And only that, for a shell script.) | 07:21 |
|
cbreak_work
| I think svn dcommit does a rebase&reset after it's done | 07:21 |
|
| cjs: I think git branch can do that | 07:21 |
|
selckin
| as long as they won't get commited i'm happy | 07:21 |
|
cbreak_work
| or rev-parse | 07:21 |
|
cjs
| cbreak_work: I can't seem to find the bit in the git-branch manpage to tell me how to do it. | 07:23 |
|
slonopotamus
| selckin: well, it can fail | 07:23 |
|
cjs
| I guess $(cat .git/HEAD) will do well enough. | 07:24 |
| ← yatiohi left | 07:24 |
|
cbreak_work
| git rev-parse --symbolic-full-name HEAD | 07:24 |
|
leo2007
| when I use git svn to clone http://svn.clozure.com/publicsvn/openmcl/trunk/darwinx86/ccl, it doesn't seems to clone the whole repo. For example, I only have 4 files in the cloned repo. Ideas? | 07:25 |
|
vals_
| leo2007: submodules maybe? | 07:25 |
|
leo2007
| submodules in svn? | 07:25 |
| ← pkrumins left | 07:25 |
|
leo2007
| I don't understand what you mean. | 07:26 |
| → gergap joined | 07:26 |
|
slonopotamus
| leo2007: that url only has 4 files | 07:26 |
|
| leo2007: what else did you expect? | 07:26 |
|
| leo2007: maybe you want "git clone http://svn.clozure.com/publicsvn/openmcl/ --stdlayout" ? | 07:27 |
|
| err, git svn clone | 07:28 |
|
leo2007
| slonopotamus: svn co the same url gives far more than 4 files, it seems svn pulled from external sources. | 07:29 |
|
slonopotamus
| cxreg: i think most git.git action happens on maillist | 07:29 |
|
leo2007
| I saw "Checked out external......" in 'svn co' | 07:29 |
|
slonopotamus
| oh | 07:29 |
| → mw_ joined | 07:31 |
| → __ed joined | 07:31 |
| → Schmallon joined | 07:32 |
| → masterkorp joined | 07:35 |
| → aspotashev joined | 07:35 |
| adamholt → adamholt_away | 07:36 |
| adamholt_away → adamholt | 07:36 |
| → Axius joined | 07:40 |
| ← koollman left | 07:41 |
| ← dreinull left | 07:42 |
| → koollman joined | 07:43 |
| → trivol joined | 07:43 |
| ← hagabaka left | 07:44 |
| → slava_dp joined | 07:47 |
| ← gebi left | 07:47 |
|
slava_dp
| I can't modify a commit's message, can I? | 07:47 |
|
cbreak_work
| you can | 07:48 |
|
| but it's history rewriting | 07:48 |
|
| so the result will be a new commit in place of the old one | 07:48 |
| ← programble left | 07:48 |
|
cbreak_work
| slava_dp: man git-rebase -i | 07:48 |
|
frogonwheels
| slava_dp: you can annotate it. umm I think that's teh correct word.. | 07:48 |
| ← psoo left | 07:49 |
|
shruggar
| slava_dp, what is your goal? Was the commit the last commit you made? Have you pushed it / branched from it, yet? | 07:49 |
|
frogonwheels
| nope. wrong word | 07:49 |
| adamholt → adamholt_away | 07:49 |
| adamholt_away → adamholt | 07:49 |
| ← steveire left | 07:50 |
| ← shruggar left | 07:50 |
|
slava_dp
| I think I'll just leave it as it is :) just wanted to know if there is an easy way. | 07:50 |
|
frogonwheels
| slava_dp: have you pushed it? | 07:51 |
| → steveire joined | 07:51 |
| ← steveire left | 07:51 |
| → steveire joined | 07:51 |
|
Rhonda
| slava_dp: For the last commit there is an easy way (always keep in mind to only do it when you haven't pushed yet): git commit --amend | 07:51 |
|
cbreak_work
| rebase -i is easy | 07:51 |
|
| but it's dangerous and very powerful | 07:51 |
|
frogonwheels
| slava_dp: as in pushed the commit to a server | 07:51 |
| ← unreal left | 07:52 |
| → unreal joined | 07:52 |
|
frogonwheels
| .. only spectacularly dangerous when you rebase already pushed commits.. and force it up | 07:52 |
|
slava_dp
| frogonwheels, no, it's my local git, I don't have a parent server. | 07:52 |
|
frogonwheels
| slava_dp: well then if it's the last one, use Rhonda's suggestion, otherwise rebase -i is really easy and sooo useful | 07:53 |
|
slava_dp
| ok, I'll read help rebase then :) | 07:53 |
| ← gergap left | 07:54 |
|
frogonwheels
| slava_dp: you can also squash up those icky 'woops' bugfix commits.. | 07:54 |
| → gergap joined | 07:54 |
|
frogonwheels
| slava_dp: those 'left off a ;' ones ;) | 07:54 |
| → tizbac joined | 07:54 |
|
| Rhonda doesn't like sqashing commits. People aren't perfect and those are only natural and one shouldn't feel the need to "hide" them. | 07:56 |
|
cbreak_work
| I craft my history. | 07:57 |
|
Rhonda
| And people that laugh at such commits are the ones that usually do it themself all the time . o O ( and squash them themself because otherwise they'd die in embarassment like they do onto others ) | 07:57 |
|
cbreak_work
| I want every commit to be compilable | 07:57 |
|
frogonwheels
| :) I'm with cbreak_work | 07:57 |
|
cbreak_work
| but mostly, I use --amend | 07:58 |
|
hyperair
| Rhonda: would you prefer to have 2-3x more patches to review, each of them correcting a mistake made in a patch before it? | 07:59 |
| ← quaestor left | 07:59 |
| ← bgy left | 07:59 |
| ← workmad3 left | 08:00 |
|
hyperair
| i squash commits and make the messages more descriptive to aid anyone who reviews my patches prior to running format-patch | 08:00 |
|
cbreak_work
| I commit bug fixes as separate commits if they are interesting. | 08:00 |
| → michel___ joined | 08:00 |
|
cbreak_work
| (and in the commit message I boast about how hard the journey to it's squashing was, and how fierce the battle played out) | 08:00 |
| ← Axius left | 08:01 |
|
hyperair
| lol | 08:01 |
| → paulbaumgart joined | 08:01 |
| ← krawek left | 08:02 |
| → gemmes joined | 08:03 |
| ← steveire left | 08:04 |
| → calin joined | 08:04 |
| → shruggar joined | 08:04 |
| → steveire joined | 08:04 |
| ← steveire left | 08:04 |
| → steveire joined | 08:04 |
| → johnw joined | 08:07 |
| ← AAA_awright left | 08:08 |
| → AAA_awright joined | 08:08 |
| → tvw joined | 08:08 |
| ← datapolitical left | 08:10 |
| ← michel___ left | 08:10 |
| ← johnw left | 08:11 |
| ← QwertyM left | 08:11 |
| ← rchavik left | 08:11 |
| ← mutex left | 08:12 |
|
gemmes
| Hi all, noob here, how do I revert back to a previous commit? | 08:12 |
| → steffkes- joined | 08:12 |
|
cbreak_work
| gemmes: more details | 08:12 |
| ← khmarbaise left | 08:12 |
|
cbreak_work
| what do you want to do in the past? | 08:12 |
| → rchavik joined | 08:12 |
|
cbreak_work
| do you want to stay there? | 08:12 |
|
| did you already push your present? | 08:13 |
| → mutex joined | 08:13 |
|
gemmes
| i only installed git yesterday and made one commit, i now have errors, i want to commit the current files then revert back to yesterdays files | 08:13 |
|
shruggar
| gemmes, you want to commit the current files, then discard them? | 08:14 |
| ← doener left | 08:14 |
| ← zorzar left | 08:14 |
| ← d1zzy left | 08:14 |
|
shruggar
| gemmes, sounds like you might want to make a branch | 08:14 |
|
gemmes
| i will only discard them if previous files have bug too. | 08:14 |
| ← steffkes left | 08:14 |
|
shruggar
| ah, so you want to /temporarily/ go back to the previous commit | 08:15 |
|
cbreak_work
| I think making a branch is reasonable anyway. | 08:15 |
|
gemmes
| yes to see if files are working | 08:15 |
| → d1zzy joined | 08:15 |
| → zorzar joined | 08:15 |
| → doener joined | 08:15 |
|
Rhonda
| hyperair: It of course highly depends on the worflow and interaction with others. Still I think people are too keen to hide their mistakes when they are just human, and people should get used to be able to admit them instead of humiliate others that don't hide. | 08:15 |
| ← mithro left | 08:15 |
| → froschi1 joined | 08:16 |
|
shruggar
| commit the files, then "git checkout HEAD^", this will put you into a "detached head" where you can examine the previous state. | 08:16 |
|
| ^gemmes | 08:16 |
|
gemmes
| ok will look into trying that | 08:16 |
| ← vpilo left | 08:17 |
|
paulbaumgart
| gemmes: git stash might be what you're looking for: http://progit.org/book/ch6-3.html | 08:18 |
|
shruggar
| good point, paulbaumgart | 08:18 |
|
gemmes
| thanks for link, will try that too | 08:19 |
| → khmarbaise joined | 08:19 |
| → vpilo joined | 08:20 |
| → nevyn^_ joined | 08:22 |
| ← anjen left | 08:23 |
| → Titosemi joined | 08:23 |
| ← poli left | 08:24 |
|
ericlee
| can anyone show me the difference between git and subversion? | 08:24 |
|
| blueshift shows ericlee | 08:25 |
|
cbreak_work
| what's not different? | 08:25 |
|
Kasreyn
| ericlee: snapshots? | 08:25 |
| → poli joined | 08:25 |
|
mar77i
| ericlee: why do you ask? | 08:26 |
|
ericlee
| because ready to learn | 08:26 |
|
cbreak_work
| there's an intro, git for svn users | 08:27 |
| → ProperN[out] joined | 08:27 |
|
Kasreyn
| clear your mind and press search | 08:27 |
| ← ProperN[out] left | 08:27 |
| → ProperN[out] joined | 08:27 |
| → intripoon joined | 08:28 |
| → MrCheetoDust joined | 08:29 |
|
MrCheetoDust
| Hey guys, I've used git for a while and recently installed gitx -- it was working for a while, but now it's just showing me 'Pieter de Bie' | 08:29 |
|
| for author, date, subject, etc when I select my commits | 08:29 |
|
| anyone know how to fix that? | 08:30 |
|
wereHamster
| fix what? | 08:30 |
| ← Mage left | 08:30 |
| ← ProperNoun left | 08:30 |
| ProperN[out] → ProperNoun | 08:30 |
|
MrCheetoDust
| I'm using GitX | 08:30 |
|
| and now anytime I select a commit | 08:30 |
|
| to view changes, etc | 08:30 |
|
cbreak_work
| does git log show correct? | 08:30 |
|
wereHamster
| what does `git log` show as the author? | 08:30 |
|
MrCheetoDust
| it's blank and just shows 'Peter de Bie' for author/date/etc | 08:30 |
|
| yes git itself | 08:30 |
|
| is perfect | 08:30 |
|
wereHamster
| take it up with the gitx author.. | 08:31 |
| ← MrCheetoDust left | 08:31 |
| → hearit joined | 08:33 |
| hearit → Iahc | 08:33 |
|
Pieter
| what's up? | 08:33 |
| → lbt_ joined | 08:35 |
| → Paraselene_ joined | 08:35 |
| → dreinull joined | 08:36 |
| ← carutsu_ left | 08:42 |
| → bauruine joined | 08:43 |
| → cdamian joined | 08:44 |
|
cdamian
| I am (still) trying to migrate our company from svn to git. Is there any magic to speed up git-svn? I think it is getting slower for every revision and I think it might take a month to finish, if it does at all. | 08:46 |
|
Pieter
| you can try to kill it and run it | 08:47 |
|
selckin
| could try, ^c; git gc; git svn rebase, | 08:48 |
| → galderz joined | 08:50 |
|
slonopotamus
| cdamian: how big svn repo is? (gigs, num of commits) | 08:50 |
|
shruggar
| cdamian, how complex is the clone you're trying to make? Just trunk or tracking multiple-branches, for example? | 08:51 |
|
| shruggar doesn't know where that hyphen came from... | 08:51 |
|
slonopotamus
| shruggar: how that matters? | 08:51 |
|
cbreak_work
| cdamian: do the conversion on the same host | 08:51 |
|
shruggar
| slonopotamus, because if it's just trunk, and he doesn't mind waiting a month or so, git-remote-svn may be done by then ;) | 08:51 |
|
cdamian
| compressed svn dump is ~ 2G, about five branches and 35000 revisions | 08:51 |
|
| I tried it with a local repo and it didn't help much | 08:52 |
| ← enherit left | 08:52 |
|
slonopotamus
| cdamian: that should be <1 day on modern hardware. | 08:52 |
|
wereHamster
| cdamian: are you cloning through svn: or http: or from a local filesystem? | 08:53 |
| ← roop left | 08:53 |
|
wereHamster
| and there are other tools for svn->git migration thatn git-svn | 08:53 |
| → tatsuya joined | 08:53 |
|
cdamian
| At the moment I do with http: , but I also tried with file: | 08:53 |
|
wereHamster
| some much faster than git-svn | 08:53 |
|
cdamian
| I have used svn-dump-fast-import, which is a lot faster and takes the dump. But I think git svn preserves the branches and merges better. | 08:54 |
|
shruggar
| git-svn preserves merges? | 08:55 |
|
slonopotamus
| git-svn doesn't handle merges | 08:55 |
|
cdamian
| ok, I am mistaken then :-) | 08:55 |
|
slonopotamus
| because svn doesn't handle merges | 08:55 |
|
shruggar
| slonopotamus, svn does handle merges, it's just not very elegant ;) | 08:56 |
|
cdamian
| it does store some merge information in the properties, but it is all a bit rubbish | 08:57 |
|
slonopotamus
| shruggar: no, it doesn't | 08:57 |
| ← paulbaumgart left | 08:57 |
|
shruggar
| slonopotamus, in what way does svn not handle merges? | 08:57 |
|
slonopotamus
| shruggar: svn log is linear | 08:58 |
|
| shruggar: for example | 08:58 |
|
cbreak_work
| it has one line | 08:58 |
|
shruggar
| slonopotamus, so your argument is basically "svn doesn't handle feature X because svn has a crappy interface" | 08:58 |
|
cbreak_work
| a feature that's not usable does not exist :) | 08:58 |
|
slonopotamus
| shruggar: merged commits do not appear in history at all. | 08:58 |
| → workmad3 joined | 08:59 |
|
cdamian
| maybe I give one of the other tools another chance and also try to find some better hardware. | 08:59 |
| ← slava_dp left | 08:59 |
|
spb
| when you merge in subversion, the only history you can see is one side of the merge, with hopefully a checkin comment that says what the other side is | 08:59 |
| ← slonopotamus left | 09:00 |
| → slonopotamus joined | 09:00 |
|
cdamian
| and some of the tools do the comment (or storing it in properties) automatically | 09:00 |
|
shruggar
| I'll admit that I've not used svn directly since mergeinfo was added, but it does exist | 09:00 |
|
slonopotamus
| it does exist, yep :) but it doesn't solve. | 09:01 |
| → offby1` joined | 09:01 |
| ← dj_tjerk left | 09:01 |
|
shruggar
| slonopotamus, what doesn't it solve? I'm interested in this because I'm trying (or, should be over the weekend or next week) to write up a description of various svn histories and how they would translate into git histories | 09:02 |
| ← aspotashev left | 09:03 |
| ← offby1 left | 09:03 |
| ← Tricia left | 09:03 |
| ← froschi1 left | 09:03 |
| → aspotashev joined | 09:04 |
| → dj_tjerk joined | 09:07 |
|
cdamian
| well, if I ever manage to migrate I will post something about it :-) there are more problems any way, with client hooks, jira, deployment by svn revision, sparse checkouts and more fun. And everyone here hates the shell. | 09:07 |
| → jargan joined | 09:07 |
| → bburhans_ joined | 09:07 |
| → d0k joined | 09:07 |
| → charon joined | 09:07 |
| → sven^ joined | 09:07 |
| ← koki left | 09:07 |
| ← stockholm left | 09:07 |
| ← bburhans left | 09:07 |
| → dirker joined | 09:08 |
| → koki joined | 09:08 |
|
slonopotamus
| shruggar: there's so much said on this on the web already... | 09:08 |
| → knittl joined | 09:08 |
| → broquaint joined | 09:08 |
| → stockholm joined | 09:08 |
| → roop joined | 09:08 |
| ← Iahc left | 09:09 |
|
slonopotamus
| shruggar: http://designbygravity.wordpress.com/2009/10/19/what-mother-never-told-you-about-svn-branching-and-merging/ | 09:10 |
| ← masterkorp left | 09:12 |
| → Razor__ joined | 09:12 |
| ← broquaint left | 09:13 |
| ← Razor__ left | 09:13 |
| → trash joined | 09:14 |
| → gebi joined | 09:14 |
| → wereHamster joined | 09:15 |
| ← werdan7 left | 09:18 |
|
shruggar
| slonopotamus, as is pointed out in the comments, that appears to be a "pre merge-tracking" workflow | 09:19 |
| → paulbaumgart joined | 09:19 |
|
hyperair
| slonopotamus: that looks like constantly rebasing. | 09:20 |
|
paulbaumgart
| how stable is 1.7.2? any ideas on when it will become officially designated as stable? | 09:20 |
|
shruggar
| paulbaumgart, I think it was released officially on wednesday | 09:21 |
|
| paulbaumgart, afaik "released" means "stable" | 09:21 |
|
sitaram
| Oxf13: white space does not matter (sorry I was in a meeting pretty much the whole darn day till now!) | 09:21 |
|
paulbaumgart
| shruggar: oh, great, apparently git-scm.com hasn't been updated to reflect that | 09:22 |
| → prietopa joined | 09:22 |
|
shruggar
| paulbaumgart, I just saw that, yeah. http://git.kernel.org/?p=git/git.git;a=tag;h=refs/tags/v1.7.2 it's not an "rc" anymore, anyway | 09:22 |
| ← bauruine left | 09:22 |
|
paulbaumgart
| shruggar: great, thanks | 09:22 |
| → shwaiil joined | 09:23 |
| → bluenovember joined | 09:23 |
|
prietopa
| Why GIT not recognise changes in my repository?. I change the file name. MsxHouseFidelity --> MSXHouseFidelity, but no changes in git | 09:24 |
| ← ycyuxin left | 09:24 |
|
shruggar
| prietopa, are you on windows? | 09:24 |
|
prietopa
| yes | 09:24 |
|
Arrowmaster
| thats why | 09:24 |
|
prietopa
| is a windows problem? | 09:24 |
|
Arrowmaster
| yes | 09:24 |
|
shruggar
| prietopa, it's a "git on windows" problem | 09:24 |
|
blueshift
| windows doesn't really understand case | 09:24 |
| → ruskie joined | 09:24 |
| → masterkorp joined | 09:24 |
|
prietopa
| ok, thanks | 09:25 |
|
Arrowmaster
| try git rm --cached MsxHouseFidelity | 09:25 |
| → aziz joined | 09:25 |
|
Arrowmaster
| then git add MSXHouseFidelity | 09:25 |
| ← slonopotamus left | 09:26 |
|
prietopa
| Arromaster: thanks, i try to change file name in ubuntu system. | 09:26 |
| → werdan7 joined | 09:27 |
| ← dreinull left | 09:29 |
| ← cdamian left | 09:29 |
| ← masterkorp left | 09:29 |
| ← aspotashev left | 09:29 |
| → curtana joined | 09:30 |
| → masterkorp joined | 09:31 |
| ← Dave^| left | 09:31 |
| → Dave^| joined | 09:31 |
| → jds joined | 09:32 |
| → jtdowney joined | 09:32 |
| → becomingGuru joined | 09:33 |
| → Fullmoon joined | 09:34 |
| → lorenz joined | 09:34 |
| ← lorenz left | 09:35 |
| → bauruine joined | 09:38 |
| → radioman-lt joined | 09:38 |
| ← masterkorp left | 09:39 |
| → ralf_e joined | 09:39 |
| ← roger_padactor left | 09:39 |
| → roger_padactor joined | 09:40 |
| → entrosca joined | 09:40 |
| → slonopotamus joined | 09:42 |
| ← mw_ left | 09:42 |
| ← curtana left | 09:48 |
| → nicoulaj joined | 09:49 |
| → masterkorp joined | 09:52 |
| → Bennid joined | 09:53 |
| ← slonopotamus left | 09:55 |
| ← steveire left | 09:56 |
| ← masterkorp left | 09:56 |
| → steveire joined | 09:57 |
| ← steveire left | 09:57 |
| → steveire joined | 09:57 |
| ← prietopa left | 09:59 |
| ← fivetwentysix left | 10:00 |
| → dany_ joined | 10:00 |
|
dany_
| hi all | 10:01 |
|
Ilari
| hi | 10:01 |
| ← gergap left | 10:01 |
| → fivetwentysix joined | 10:01 |
| ← fivetwentysix left | 10:01 |
| → gergap joined | 10:01 |
| → JEEBcz joined | 10:01 |
|
dany_
| I'm new with git. Can you tell me if is it possible to commit all the file that you have modified with using add namefile? | 10:01 |
| ← vpilo left | 10:02 |
|
shruggar
| dany_, I expect you're asking "is there a way to commit all changes, regardless of whether or not they've previously been added to the index using "git add <file>"?" | 10:02 |
| → vpilo joined | 10:03 |
| yaotti → yaotti_aw | 10:03 |
| → psankar joined | 10:03 |
|
dany_
| uhm | 10:03 |
| → curtana joined | 10:03 |
|
dany_
| shruggar: yes | 10:03 |
|
| shruggar: sorry for my ugly english | 10:04 |
| ← JEEB left | 10:04 |
| → Mpenz joined | 10:04 |
| ← steveire left | 10:04 |
|
shruggar
| dany_, "git commit -a" will do that for you. "git add -u" will (similarly) add all changes in one go. However, I recommend trying "git add --patch", as it may make the power of the index more obvious. | 10:04 |
|
dany_
| thanks | 10:05 |
| ← becomingGuru left | 10:05 |
| → steveire joined | 10:06 |
| ← steveire left | 10:06 |
| → steveire joined | 10:06 |
|
dany_
| you were very useful :) | 10:06 |
|
Ilari
| Or 'git add --interactive'. | 10:06 |
| → gbacon joined | 10:07 |
|
dany_
| Ilari: what is the difference | 10:07 |
|
| ? | 10:07 |
|
shruggar
| git add --interactive is menu-driven, while git add --patch is a shortcut for one of the more commonly-used options of the interactive mode | 10:07 |
|
dany_
| cool | 10:08 |
|
| thankss | 10:08 |
|
cbreak_work
| git add -p is great | 10:08 |
|
dany_
| cool | 10:08 |
|
cbreak_work
| it's edit mode is... interesting | 10:09 |
| ← bdrewery left | 10:09 |
|
dany_
| thank you guys | 10:09 |
|
shruggar
| I think I'd prefer some sort of dedicated patch-editing tool to vim for editing.. | 10:10 |
|
cbreak_work
| it uses TextWrangler here :) | 10:10 |
|
dany_
| :D | 10:10 |
|
shruggar
| I use the edit mode at least once a week, usually more | 10:10 |
|
charon
| shruggar: well, i don't think there's any particular reason not to support a dedicated config variable for that editor | 10:11 |
| → AAA_awright_ joined | 10:12 |
|
shruggar
| charon, it was more of a pipe-dream, I don't know of any dedicated patch-editor. I assume I could wrap vimdiff in a simple shell script to get what I really want | 10:12 |
| → beni- joined | 10:13 |
|
shruggar
| (a way to say "oops, that's not right, let me edit that hunk quickly" without actually editing the patch by hand) | 10:13 |
|
| shruggar adds it to his big list | 10:13 |
| ← AAA_awright left | 10:15 |
| → AAA_awright joined | 10:16 |
|
shruggar
| hm, though should it be "edit the pristine file, with hunk applied", "edit the currently-staged file, with hunk applied", or "edit a copy of the work-tree file" | 10:17 |
| ← skoop left | 10:17 |
| → sh1mmer joined | 10:17 |
| ← AAA_awright_ left | 10:17 |
|
shruggar
| and potential weirdness if your edits result in more than one hunk | 10:18 |
| → bdrewery joined | 10:18 |
| → girishr joined | 10:18 |
| → Joeconyers joined | 10:19 |
| ← dany_ left | 10:19 |
| → rajeshsr_ joined | 10:22 |
| → JEEB joined | 10:23 |
| ← JEEBcz left | 10:24 |
| ← gbacon left | 10:24 |
| ← rajeshsr_ left | 10:29 |
| ← steveire left | 10:33 |
| → dreinull joined | 10:33 |
| ← entrosca left | 10:33 |
| → entrosca joined | 10:34 |
| ← Joeconyers left | 10:35 |
| → parasti joined | 10:35 |
| → steveire joined | 10:35 |
| ← JEEB left | 10:36 |
| → JEEB joined | 10:36 |
| → ereslibre joined | 10:38 |
| ← ereslibre left | 10:38 |
| → ereslibre joined | 10:38 |
| ← gemmes left | 10:42 |
| ← sh1mmer left | 10:43 |
| → formale joined | 10:44 |
| vals_ → tango_ | 10:44 |
|
pom
| I have changes in a branch which only exists live. In the same branch there are conflicting changes on another repository. They have already been pushed to our bare rep. Hence, the changes that are live cannot be pushed to solve the conflict on a dev-server. Can I rollback the changes pushed to bare? | 10:46 |
| → froschi1 joined | 10:48 |
| → theduke joined | 10:49 |
|
Kasreyn
| how can you apply username/password with git push http://some.site/my.git ? i have added a user to my .htaccess.. | 10:49 |
|
theduke
| a question about submodules: is there a way to only have a subdirectory of a submodule-repository in the working dir? | 10:49 |
| ← jtdowney left | 10:50 |
| → skoop joined | 10:50 |
|
AAA_awright
| Kasreyn: For pushing over WebDAV? | 10:50 |
|
| I thought it asks you for a password, I forget | 10:51 |
|
Kasreyn
| using git-http-backend | 10:51 |
|
AAA_awright
| I tried it once a while ago | 10:51 |
|
Kasreyn
| only password? what about user? | 10:51 |
|
AAA_awright
| Maybe that too | 10:51 |
|
Kasreyn
| if i typed in the URL on a webbrowser it would prompt for username/password but with git i have no idea | 10:52 |
| ← steveire left | 10:52 |
|
AAA_awright
| Oh I see, that's the new HTTP CGI | 10:53 |
| → steveire joined | 10:53 |
| ← steveire left | 10:53 |
| → steveire joined | 10:53 |
| ← ralf_e left | 10:53 |
|
AAA_awright
| I haven't used that but the authentication should be similar | 10:53 |
|
Kasreyn
| AAA_awright: yes indeed. it's smart! | 10:53 |
|
AAA_awright
| Kasreyn: Have you tried pushing yet? | 10:53 |
| ← peper left | 10:53 |
|
Kasreyn
| yes, first i got error 500, now i get 401 | 10:54 |
|
| so, making progress :) | 10:54 |
|
AAA_awright
| Hm how does it know to enable write access? | 10:55 |
|
| Maybe you need http.receivepack | 10:55 |
| ← d1zzy left | 10:55 |
| cpg → cpg|away | 10:57 |
| → thierryp joined | 10:57 |
| ← byang left | 10:57 |
| → level09 joined | 10:57 |
| → nevyn^__ joined | 10:58 |
| ← nevyn^__ left | 10:58 |
| → rseifert joined | 10:59 |
| → ralf_e joined | 11:02 |
| ← nevyn^_ left | 11:02 |
|
Kasreyn
| AAA_awright: i already have my.git/config http.receivepack set true | 11:03 |
| ← sylr left | 11:04 |
|
AAA_awright
| Does it work without the htaccess restriction? | 11:04 |
|
| I'm not sure | 11:04 |
|
Kasreyn
| checkout is enabled for anonymous as default but not write | 11:05 |
| ← jpieper left | 11:08 |
|
Kasreyn
| however, setting http.receivepack should enable write for anonymous | 11:08 |
|
| setting it true that is | 11:08 |
|
| if i read man git-http-backend correctly | 11:08 |
| → jpieper joined | 11:08 |
| ← psankar left | 11:09 |
| → Fernandos joined | 11:12 |
|
Fernandos
| hi | 11:13 |
| → d1zzy joined | 11:13 |
| ← dreinull left | 11:14 |
|
Fernandos
| I've a problem pushing changes :( | 11:15 |
|
| can you help me please http://image-upload.de/image/1gjpKg/66d15cc509.png | 11:15 |
|
| this is what I get when I push | 11:15 |
|
tango_
| Fernandos: you are trying to push to a non-bare repository | 11:15 |
|
Fernandos
| what does that mean? | 11:16 |
|
tango_
| (it's a FAQ, by the way) | 11:16 |
|
| faq non-bare | 11:16 |
|
| is the bot here? | 11:16 |
|
| oh damn | 11:16 |
|
Fernandos
| idk | 11:16 |
|
| how can I solve this? | 11:16 |
| → bartman joined | 11:16 |
|
tango_
| https://git.wiki.kernel.org/index.php/GitFaq#How_do_I_make_existing_non-bare_repository_bare.3F | 11:17 |
| ← steveire left | 11:17 |
| → dreinull joined | 11:18 |
|
Kasreyn
| git push http://myuser@some.site/my.git will prompt for password. not working though | 11:19 |
| → steveire joined | 11:19 |
| ← steveire left | 11:19 |
| → steveire joined | 11:19 |
| ← skoop left | 11:21 |
| → skoop joined | 11:21 |
| ← skoop left | 11:21 |
| ← [Po]lentino left | 11:21 |
|
Fernandos
| tango_: why doesn't this command work? git clone --bare -l C:\Users\Ferhat\My Dropbox\RubyOnRails\VirtualTrading C:\Users\Ferhat\Workspace\internal\git\VirtualTrading | 11:22 |
| → skoop joined | 11:22 |
|
Fernandos
| I get an error saying "too many arguments" | 11:22 |
| → corni joined | 11:22 |
|
tango_
| because you forgot the quotes (") around the path containing spaces | 11:22 |
| → smuf joined | 11:23 |
|
tango_
| iow it should be "c:\users\ferhat\my dropbox\rubyonrails\virtualtrading" | 11:23 |
|
Fernandos
| now I get invalid argument | 11:23 |
|
| ah It worked.. I neeeds forward slashes | 11:24 |
|
| huh? what does it do? | 11:25 |
|
| it copied the contents of the .git folder into a VirtualTrading folder? | 11:26 |
| → tobiassjosten joined | 11:27 |
| → hebz0rl joined | 11:27 |
|
Fernandos
| is this the correct command? git clone --bare -l "C:/Users/Ferhat/My Dropbox/RubyOnRails/VirtualTrading" "C:/Users/Ferhat/Workspace/internal/git/VirtualTrading\.git" | 11:28 |
| ← level09 left | 11:28 |
|
Fernandos
| I have got a bare repo now. what do I do next to work productively. I want to be able to push/pull from the other repo | 11:30 |
| ← roop left | 11:31 |
| → masterkorp joined | 11:32 |
| → winterweaver joined | 11:32 |
| → xiong joined | 11:33 |
| → nataraj joined | 11:34 |
| ← kenyon left | 11:35 |
| ← Pepsi1x1 left | 11:35 |
| → kenyon joined | 11:35 |
| ← steveire left | 11:36 |
|
Fernandos
| I tried: git add remote dropbox "C:/Users/Ferhat/My Dropbox/RubyOnRails/VirtualTrading" in my bare repo now | 11:36 |
| → steveire joined | 11:37 |
| ← steveire left | 11:37 |
| → steveire joined | 11:37 |
|
wereHamster
| you can't work in a bare repo | 11:37 |
|
| work as in make and commit new changes.. | 11:37 |
|
Fernandos
| fellas I'm kinda lost :/ could you help me please to just be able to pull/push with that repo? | 11:37 |
| ← sebastorama left | 11:38 |
|
wereHamster
| make sure the repo on dropbox is bare, then clone it, work on the code, commit and push back | 11:38 |
| → barismetin joined | 11:38 |
|
Fernandos
| but I already have commits in that repo in the dropbox | 11:38 |
|
wereHamster
| so? | 11:38 |
|
Fernandos
| so I tried cloning that repo into a local directory and this resulted in this error http://image-upload.de/image/1gjpKg/66d15cc509.png | 11:39 |
| → ppr joined | 11:39 |
| ← hohoho left | 11:39 |
|
wereHamster
| I can only help you if you tell me which git commands you executed, what the output was or what happened and what you expected to happen | 11:40 |
| ← ppr left | 11:41 |
|
charon
| heh. | 11:41 |
| → peper joined | 11:42 |
|
charon
| the info is all there. it's just completely unreadable because this SmartGit apparenlty isn't so Smart after all. | 11:42 |
| ← hasen left | 11:43 |
|
nataraj
| ! [rejected] master -> master (non-fast forward) , while trying a git push origin master | 11:43 |
| ← moldovean left | 11:43 |
|
wereHamster
| nataraj: faq non-ff | 11:43 |
|
nataraj
| why would it reject? | 11:43 |
|
wereHamster
| nataraj: https://git.wiki.kernel.org/index.php/GitFaq#non-ff | 11:44 |
| → flaguy48 joined | 11:45 |
| → cannonball joined | 11:47 |
| → saurabhb joined | 11:47 |
| → becomingGuru joined | 11:48 |
|
Fernandos
| wereHamster: I typed: git clone -l "C:/Users/Ferhat/My Dropbox/RubyOnRails/VirtualTrading" "C:/Users/Ferhat/Workspace/internal/git/VirtualTrading" | 11:49 |
| → spaceonline joined | 11:49 |
| → davidc_ joined | 11:49 |
|
davidc_
| ok stupid question, I've a remote and I want to pull that remote into my local branch. I've git checkout -b branchname and when I do git pull remotename branchname it doesn't find my remote ref branchname | 11:50 |
| → walk joined | 11:51 |
| ← galderz left | 11:51 |
| → galderz joined | 11:51 |
| ← galderz left | 11:51 |
| → galderz joined | 11:51 |
|
charon
| davidc_: well, in 'git pull remotename branchname' the <branchname> must exist on the *remote* side | 11:53 |
| ← galderz left | 11:53 |
|
bremner
| davidc_: simple answer: git fetch remotename && git merge remotename/branchname | 11:54 |
|
charon
| since it specifies which branch of the <remotename> you are merging. | 11:54 |
| → rajeshsr_ joined | 11:55 |
|
davidc_
| hmm ok | 11:55 |
|
| what if I want to pull a remote master into my local branch? | 11:55 |
|
| would that be possible or the branch *has* to exist on the remote? | 11:55 |
| ← marc_kdab_ left | 11:55 |
|
bremner
| davidc_: see what I wrote above | 11:56 |
| ← flaguy48 left | 11:56 |
| → hasen joined | 11:56 |
|
bremner
| after you have fetched, you have all of the remote branches, and you can see them with git branch -r; then just merge which one you want. | 11:57 |
| ← RobertLaptop left | 11:57 |
|
davidc_
| oh nice cheers man :) | 11:58 |
| ← steveire left | 11:58 |
| → steveire joined | 11:59 |
| ← steveire left | 11:59 |
| → steveire joined | 11:59 |
| ← becomingGuru left | 12:01 |
| ← dreinull left | 12:02 |
| → roop joined | 12:04 |
| ← Fernandos left | 12:05 |
| ← hyperair left | 12:06 |
| ← drubies left | 12:06 |
| → flaguy48 joined | 12:08 |
| ← opalka left | 12:09 |
| → tar- joined | 12:11 |
| → razvand joined | 12:11 |
| → Bass10 joined | 12:11 |
| → hyperair joined | 12:11 |
| ← ralf_e left | 12:11 |
| ← formale left | 12:12 |
| → aspotashev joined | 12:13 |
| → marc_kdab_ joined | 12:14 |
| → Ryback_ joined | 12:15 |
| ← unreal left | 12:16 |
| ← hyperair left | 12:17 |
| → unreal joined | 12:17 |
| → galderz joined | 12:17 |
| ← galderz left | 12:17 |
| → galderz joined | 12:17 |
| ← froschi1 left | 12:17 |
| → IslandUsurper joined | 12:19 |
| ← peper left | 12:20 |
| → bodom joined | 12:20 |
| → _Vi joined | 12:20 |
| ← warthog9 left | 12:22 |
| → peper joined | 12:23 |
| ← h32Lg left | 12:23 |
| → SlickT10 joined | 12:25 |
| → roop_ joined | 12:26 |
| ← roop left | 12:27 |
| ← saurabhb left | 12:27 |
| → jaql joined | 12:27 |
| → ralf_e joined | 12:28 |
| → gantman joined | 12:28 |
| → dr_win joined | 12:28 |
| → froschi1 joined | 12:29 |
| adamholt → adamholt_away | 12:31 |
| adamholt_away → adamholt | 12:31 |
| ← spaceonline left | 12:33 |
| ← walk left | 12:34 |
| → walk joined | 12:35 |
| ← hebz0rl left | 12:36 |
| ← girishr left | 12:38 |
| → spaceonline joined | 12:38 |
| → bgy joined | 12:38 |
|
bgy
| Hi | 12:38 |
| → StrongOrder joined | 12:39 |
|
bgy
| I'd like to experiment with rebase, but I don't want to break my repository, and I'd like to do on a real repo rather than one done for the occasion, can I copy my branches, both master and experiments to try it out ? | 12:39 |
|
cbreak_work
| just use git clone | 12:40 |
|
| it clones repositories | 12:40 |
| → coderdad joined | 12:40 |
|
bgy
| cbreak, Won't it affect my bare remote repository ? | 12:40 |
|
cbreak_work
| only if you push back | 12:41 |
|
| clones are independent | 12:41 |
| → mbroeker joined | 12:41 |
|
bgy
| Ok, I though rebase did it alone | 12:41 |
|
cbreak_work
| each repo is independent unless you clone it with --shared | 12:41 |
|
bgy
| Ok | 12:41 |
|
| How could I know my rebase worked properly ? a git log would be enough ? | 12:42 |
| → airborn joined | 12:42 |
|
cbreak_work
| if it is as you want it to be then it worked | 12:42 |
|
bgy
| clear enough : ) | 12:42 |
| ← roger_padactor left | 12:43 |
|
bgy
| thank you | 12:43 |
| → tstclair joined | 12:43 |
|
magcius
| hm | 12:44 |
| ← gergap left | 12:45 |
| ← spaceonline left | 12:46 |
| → felipec joined | 12:47 |
| ← workmad3 left | 12:47 |
|
magcius
| I can't clone git://git.gnome.org/gtk+ since the upgrade to 1.7.2 | 12:48 |
|
| can anybody else? | 12:48 |
| → scarabx joined | 12:49 |
|
charon
| magcius: git clone 'git://git.gnome.org/gtk%2B' | 12:50 |
| → Rearden joined | 12:50 |
| ← StrongOrder left | 12:50 |
|
magcius
| charon: aha, is there a reason now? | 12:50 |
|
| charon: it wasn't in the release notes, and it's backwards-incompatible | 12:50 |
|
cbreak_work
| is + a permitted char in URLs? | 12:51 |
|
magcius
| Indeed it is. | 12:51 |
|
| You've never seen + used instead of spaces? | 12:51 |
|
cbreak_work
| I've seen %20 used instead of spaces | 12:51 |
|
charon
| well, if you look in rfc2396 you will find that + is a reserved character | 12:52 |
|
| but i agree on the release notes point | 12:52 |
| → Yuuhi joined | 12:52 |
|
bgy
| How to cancel a git pull ? | 12:52 |
|
cbreak_work
| ctrl-c? | 12:53 |
|
magcius
| ctrl-\ | 12:53 |
|
bgy
| I mean after the pull is finished.. | 12:53 |
|
magcius
| bgy: revert back to where it was before? | 12:53 |
|
cbreak_work
| git reset --hard HEAD@{1} | 12:54 |
|
| note that this will screw up your history if the pull didn't do anything | 12:54 |
|
charon
| magcius: what was the last version that worked? | 12:54 |
| ← Friction left | 12:55 |
|
magcius
| charon: 1.7.1? | 12:55 |
| → Luyang joined | 12:55 |
|
Luyang
| has anyone tried http://code.google.com/p/tortoisegit/ | 12:55 |
|
| using it | 12:56 |
| → workmad3 joined | 12:57 |
|
charon
| magcius: apparently it's from 9d2e942 (decode file:// and ssh:// URLs, 2010-05-23) | 12:57 |
|
cbreak_work
| Luyang: a tiny bit | 12:58 |
|
Luyang
| cbreak_work: did it feel stable? | 13:03 |
|
cbreak_work
| no. | 13:03 |
|
| it felt like windows | 13:03 |
| ← marienz left | 13:03 |
|
cbreak_work
| I have no idea how people can work with windows | 13:03 |
|
| switching between more than two or three applications with one or two windows each is a pain | 13:03 |
|
| I so missed exposé | 13:04 |
|
| anyway, I like git command line more | 13:04 |
|
divVerent
| isn't expose dead in 10.6? | 13:04 |
|
cbreak_work
| no. why? | 13:04 |
|
divVerent
| or rather, replaced by a simple thumbnail view? | 13:04 |
|
| I tried it | 13:04 |
|
cbreak_work
| it just looks different | 13:04 |
|
divVerent
| I no longer can get normal expose like in 10.4 | 13:04 |
|
| it now is just a thumbnail view on a rectangular grid | 13:04 |
|
cbreak_work
| I liked the older layout more | 13:04 |
|
divVerent
| can't vista do THAT too? :P | 13:04 |
|
cbreak_work
| I use 7 | 13:04 |
|
| no idea how | 13:05 |
|
divVerent
| sure | 13:05 |
|
| just saying | 13:05 |
| ← tizbac left | 13:05 |
|
divVerent
| apparently, Apple turned Expose into a very standard rectangular grid | 13:05 |
|
cbreak_work
| expose on os x shows all windows | 13:05 |
|
| which is what I want | 13:05 |
|
| both the old and new layout | 13:05 |
|
divVerent
| sure | 13:05 |
|
cbreak_work
| I have yet to find out how windows can do that | 13:05 |
|
divVerent
| but doesn't vista do somethign similar in, IIRC, alt-tab? | 13:05 |
| → MattDiPasquale joined | 13:05 |
|
divVerent
| some time ago since I last used it, though | 13:05 |
| → hyperair joined | 13:06 |
|
cbreak_work
| I'll try next time I start the VM | 13:06 |
| ← jnl left | 13:06 |
|
divVerent
| I am normally on Linux, and never could get to like Expose | 13:06 |
| → mw_ joined | 13:06 |
| ← mw_ left | 13:06 |
| → mw_ joined | 13:06 |
|
cbreak_work
| I use 2 real screens, 6 virtual screens and expose | 13:06 |
|
divVerent
| it's one of the cool effects but seems entirely useless - ESPECIALLY now that the location of a window in the expose view is EVEN LESS predictable | 13:06 |
| ← airborn left | 13:06 |
|
| cbreak_work has way too many windows | 13:06 |
|
divVerent
| I now use virtual desktops only :P | 13:06 |
| ← MattDiPasquale left | 13:07 |
| ← nevyn left | 13:07 |
|
divVerent
| and on OS X, I _would_ have preferred the dock, had it been easily keyboard accessible | 13:07 |
|
| too bad OS X lacks a good alt-tab | 13:07 |
|
magcius
| charon: did you just report that? | 13:07 |
| ← roop_ left | 13:07 |
|
cbreak_work
| it is keyboard accessible | 13:07 |
|
divVerent
| it can switch between applications, but how to switch between windows is application dependent | 13:07 |
|
cbreak_work
| the default commands are stupid though | 13:08 |
|
| no, in german layouts, you can switch between windows with cmd-< and cmd-> | 13:08 |
| ← ereslibre left | 13:08 |
|
divVerent
| but yes, back then on that 10.4 mac, I had simply installed "Witch" go get a sensible alt-tabbing | 13:08 |
|
| speaking of german layout | 13:08 |
| → nevyn joined | 13:08 |
| ← calin left | 13:08 |
| → airborn joined | 13:08 |
|
divVerent
| why is that intentionally evil? | 13:08 |
| → marienz joined | 13:08 |
|
| cbreak_work uses swiss german layout | 13:09 |
|
charon
| magcius: well i'm typing up an email | 13:09 |
|
divVerent
| windows user on osx machine tries to enter email address, hits command-q (german windows altgr-q), and quits the browser without asking | 13:09 |
| ← neunon left | 13:09 |
|
magcius
| charon: it was already fixed | 13:09 |
| → hugoxrosa joined | 13:09 |
|
cbreak_work
| cmd-q is quit. | 13:09 |
|
magcius
| charon: http://git.kernel.org/?p=git/git.git;a=commit;h=ce83eda15554ffd859ecc9f4c427303d460dab60 | 13:09 |
| ← winterweaver left | 13:09 |
|
divVerent
| osx user on windows machine tries to enter email address, hits winkey-l (apple alt-l), and locks the screen | 13:09 |
|
cbreak_work
| it's like that since ... since I can remember | 13:09 |
|
divVerent
| cbreak_work: sure, but THE SAME combination is @ on DOS/Windows | 13:09 |
|
| this is the one thing that makes Apple computers ABSOLUTELY unsuitable for users who want a computer to "just work" | 13:10 |
|
cbreak_work
| the cmd key is in the same spot as the windows key on windows keyboards | 13:10 |
|
charon
| magcius: now i'm confused. that does fix a bug in the original commit but it does not fix your gripe with the backwards-incompatible chagne. | 13:10 |
|
divVerent
| cbreak_work: no, the alt key is | 13:10 |
|
magcius
| charon: it doesn't? | 13:10 |
|
divVerent
| everyone who had previously used Windows, will quit his browser all the time on OS X | 13:10 |
|
| cbreak_work: but in scancode, the cmd key is :P | 13:10 |
|
| (if you plug PC keyboard into Apple) | 13:10 |
|
| Apple keyboard has the codes of the keys between ctrl and space swapped | 13:11 |
|
cbreak_work
| I have a cmd key left and right from the space key | 13:11 |
| → LiamH joined | 13:11 |
|
divVerent
| cbreak_work: exactly | 13:11 |
|
| and on german pc layout, on the right of space is AltGr | 13:11 |
|
cbreak_work
| isn't that where the windows key is? | 13:11 |
| ← nataraj left | 13:11 |
|
divVerent
| which you press together with q to enter @ | 13:11 |
|
magcius
| charon: oh shoot. | 13:11 |
|
| charon: I misread "Jun 23", thinking it was today | 13:11 |
|
charon
| magcius: that commit is in 1.7.2 too. | 13:11 |
|
magcius
| yeah, I'm an idiot | 13:11 |
|
cbreak_work
| anyway, @ is alt-g on my layout :) | 13:12 |
|
| you can use either alt key | 13:12 |
|
magcius
| charon: so why doesn't the fix work? | 13:12 |
|
divVerent
| cbreak_work: http://www.buessert.de/Trs-80/GermanTRS80/PC_GermanKeyboard_Make_large.gif | 13:12 |
|
charon
| magcius: because it fixes something else. read the commit message | 13:12 |
|
cbreak_work
| hmm. interesting. | 13:12 |
|
divVerent
| this @ issue is so weird, it looks INTENTIONALLY evil | 13:12 |
|
| maybe to "lock in" users :P | 13:12 |
| → zimbres joined | 13:12 |
|
divVerent
| as it makes both switching from PC to Apple, and from Apple to PC, needlessly hard | 13:12 |
|
thiago
| Alt+G or AltGr+g? | 13:12 |
|
cbreak_work
| the windows key was the most anoying key I've noticed while using windows | 13:13 |
|
divVerent
| true | 13:13 |
|
magcius
| charon: yeah, keep writing up the email | 13:13 |
|
cbreak_work
| never found a use for it | 13:13 |
|
divVerent
| on Linux I don't use it much :P | 13:13 |
|
| on my notebook actually I do use it | 13:13 |
|
cbreak_work
| it just randomly opened the start menu or locked the screen | 13:13 |
|
divVerent
| I xmodmap'ed them to pageup/pagedown | 13:13 |
|
| so I don't need the FN key to read long text on the eeepc :P | 13:13 |
| ← hyperair left | 13:13 |
|
cbreak_work
| divVerent: j and k for up and down... :) | 13:14 |
| frakturfreak_ → frakturfreak | 13:14 |
|
divVerent
| cbreak_work: tell that Chromium :P | 13:14 |
|
charon
| magcius: have an address for a cc? (pm it if you don't want it logged but it'll be in the git list archives...) | 13:14 |
|
cbreak_work
| shift-j and shift-k for pagewise :) | 13:14 |
|
magcius
| charon: [email@hidden.address] | 13:14 |
|
cbreak_work
| there's a vim extension for firefox which I liked a lot | 13:14 |
|
magcius
| [email@hidden.address] | 13:14 |
| ← steveire left | 13:14 |
|
magcius
| sorry | 13:14 |
|
divVerent
| cbreak_work: sure, but firefox is SLOW :P | 13:14 |
|
| also... does j and k work in the PDF viewers? | 13:14 |
|
cbreak_work
| there's a firefox extension that makes it fast | 13:14 |
|
curtana
| wow, what an awful place to put the menu key :) | 13:14 |
|
cbreak_work
| don't know. | 13:15 |
|
divVerent
| pgup/pgdn have the advantage of working almost universally | 13:15 |
|
| curtana: where? | 13:15 |
|
cbreak_work
| macs don't have that problem anyway | 13:15 |
| → steveire joined | 13:15 |
| ← steveire left | 13:15 |
| → steveire joined | 13:15 |
|
cbreak_work
| you can scroll with multi-touch touchpads. :) | 13:15 |
|
divVerent
| oh, wtf | 13:15 |
|
| on my screenshot | 13:15 |
|
curtana
| right end of shift key ;) | 13:15 |
|
divVerent
| haha | 13:15 |
|
| never noticed THAT | 13:15 |
|
magcius
| shift-jk? | 13:15 |
| → opalka joined | 13:15 |
|
magcius
| seriously? That's a major stretch. | 13:15 |
|
curtana
| i always press right shift with my little finger, so i would constantly hit menu and get ENRAGED | 13:15 |
|
divVerent
| menu key is probably the most useless key ever | 13:16 |
|
| 1. it duplicates shift-f10 | 13:16 |
|
curtana
| it's good as a Compose key | 13:16 |
| → kukks joined | 13:16 |
|
divVerent
| 2. just like nobody ever used shift-f10 for context menu, nobody ever uses the menu key for context menu | 13:16 |
|
cbreak_work
| at the moment I use a US keyboard (with swiss-german layout), so I often type $ instead of newline :/ | 13:16 |
|
divVerent
| wtf, weechat highlights the words "nobody" and "context" | 13:16 |
|
| ah, there's users of these names here | 13:16 |
|
curtana
| nah, here's an evil keyboard... https://robots.org.uk/stuff/odd-keyboards/norskdata.jpg | 13:17 |
| ← JEEB left | 13:17 |
|
divVerent
| a key for GOTO | 13:17 |
|
| a computer scientist's nightmare | 13:17 |
|
| (especially Dijkstra's) | 13:17 |
|
curtana
| hehe. i like the light that lgihts up when you press a key! | 13:18 |
|
| in case you didn't realise, perhaps | 13:18 |
|
divVerent
| but that keyboard is not even that evil... | 13:18 |
|
| it's not as bad as a Sun keyboard :P | 13:19 |
|
thiago
| even though that says "Norsk Data", it's actually a Swedish keyboard | 13:19 |
|
cbreak_work
| http://store.apple.com/ch-de/product/MB110SM/A?fnode=MTY1NDA1Mg&mco=MTMzNzcyMDA&s=topSellers | 13:19 |
|
magcius
| curtana: oh man, you don't remember the APL keyboard? | 13:19 |
|
cbreak_work
| that's the layout I am used to | 13:19 |
| → tizbac joined | 13:19 |
|
curtana
| magcius, ;) | 13:19 |
|
magcius
| curtana: some of the coolest features were on old IBM keyboards. Like the key map button | 13:20 |
|
curtana
| did it change the keyboard layout | 13:20 |
|
magcius
| Yes. | 13:20 |
|
curtana
| comedy | 13:20 |
|
divVerent
| Stop-A | 13:20 |
|
| oops | 13:20 |
|
| now I have to know I have to type "go" quickly to not break the file system | 13:20 |
|
magcius
| And it had a 8-direction arrow key | 13:20 |
|
| with an enter in the middle | 13:20 |
|
cbreak_work
| isn't that what's num-lock is good for on pcs? | 13:21 |
| → hyperair joined | 13:21 |
|
divVerent
| cbreak_work: you mean alt-sysrq? | 13:21 |
| → hohoho joined | 13:21 |
|
curtana
| see, things are better than they used to be. remember that the next time you quit your browser by accident? | 13:21 |
|
divVerent
| oh, the 8-dir arrow | 13:21 |
|
curtana
| (shtty web apps. a real program would ask if you are sure you want to quit....) | 13:21 |
|
divVerent
| except that the "diagonal" directions have home, end, pgup, pgdn mapped | 13:21 |
|
| curtana: 1. it is common OS X interface guideline to NOT ask questions on Cmd-Q :P | 13:22 |
|
cbreak_work
| you're supposed to ask if there's unsaved state | 13:22 |
|
divVerent
| curtana: 2. on web sites, thank the advertisement annoyance coders for that | 13:22 |
|
| originally onUnload was there for exactly that reason | 13:22 |
|
| cbreak_work: then why does Safari not ask? | 13:22 |
|
| when there are filled out form fields :P | 13:23 |
|
cbreak_work
| because it's a browser. it doesn't have state | 13:23 |
|
divVerent
| of course it has | 13:23 |
|
| unsent form fields | 13:23 |
|
cbreak_work
| (none it considers valuable) | 13:23 |
|
divVerent
| THAT is the problem | 13:23 |
|
cbreak_work
| I think safari can ask if you have multiple opened tabs or something | 13:23 |
|
| maybe taht was firefox | 13:23 |
|
divVerent
| a registration form that took 2 minutes to fill out, and 5 minutes to solve the CAPTCHA, CONTAINS such data | 13:23 |
|
curtana
| what, that sounds a bit dangerous. back when i had a mac you'd get ask if you e.g., pressed cmd+q without saving in clarisworks, photoshop, teachtext, etc... | 13:23 |
|
Luyang
| I just ran /share/msysGit/initialize.sh ... EXACTLY what's that doing? Is it downloading the whole Git repository? | 13:24 |
| ← steveire left | 13:24 |
|
divVerent
| but... chromium also does not ask when closing the tab with ctrl-w | 13:24 |
|
cbreak_work
| curtana: if you have documents that aren't saved, then you have unsaved state | 13:24 |
|
| divVerent: if you had firefox, it would restore the form content next launch :) | 13:24 |
|
divVerent
| cbreak_work: of course, that would be even better :P | 13:24 |
| → steveire joined | 13:24 |
|
divVerent
| just, teach that to Safari | 13:25 |
|
cbreak_work
| apple did some anouncement for a plugin api | 13:25 |
|
curtana
| yeah at least on chromium you can press ctrl-shift-tab to get backa closed tab, with the form state preserved( i think...) | 13:25 |
|
cbreak_work
| I have no idea if you can write session restore in it though | 13:25 |
|
divVerent
| curtana: chromium does not preserve form content in shift-ctrl-t | 13:25 |
|
| at least not on google's search entry line | 13:25 |
|
curtana
| maybe in TEXTAREAs then | 13:26 |
|
divVerent
| maybe, yes | 13:26 |
|
| or maybe only in POST forms :P | 13:26 |
|
curtana
| e.g., replying to LWN.net comment; the text is preseved but not the subject. how annoying | 13:26 |
|
| bloody web browsers, they are all shit! | 13:26 |
|
divVerent
| yes | 13:27 |
| ← hohoho left | 13:27 |
|
divVerent
| still, putting "quit" on a very common key combo of another layout is a bad idea | 13:27 |
|
cbreak_work
| blame windows | 13:27 |
|
divVerent
| why? | 13:27 |
|
cbreak_work
| they put their alt on that spot | 13:27 |
|
divVerent
| since when does APple have quit on Cmd-Q? | 13:27 |
|
cbreak_work
| since the macintosh I think | 13:27 |
|
| so 1984 or so? | 13:27 |
|
| I don't actually use macs since then though so I don't know for sure | 13:28 |
|
divVerent
| was that before or after the first MS-DOS with german layout? | 13:28 |
|
cbreak_work
| I have no idea. | 13:28 |
|
divVerent
| (I assume that already had @ on the final location) | 13:28 |
|
cbreak_work
| did people use the @ key back then? | 13:28 |
|
shruggar
| hey, Linux keeps its "send SIGINT" key on the standard windows "copy selected text" key ;P | 13:28 |
|
divVerent
| not much | 13:28 |
|
| IIRC some config.sys entries had needed it | 13:28 |
| ← benzap_ left | 13:28 |
| ← froschi1 left | 13:29 |
| → flazz joined | 13:30 |
|
divVerent
| DOS 2.1 added the keyb command to load a keyboard layout | 13:30 |
|
| Oct 1983 | 13:30 |
|
| shruggar: true that | 13:30 |
|
| except that this combination does nothing in GUI apps | 13:30 |
|
| but, unix CLEARLY was first :P | 13:30 |
|
| and macintosh 128k was after 1984, as it seems | 13:31 |
|
| 24 Jan 1984 | 13:31 |
|
magcius
| curtana: the RFC doesn't define a meaning to +, so it shouldn't be parsed as a space | 13:31 |
|
| curtana: also, the %xx hack doesn't work for something older than 1.7.2 | 13:32 |
|
divVerent
| I think DOS assigned AltGr-Q before Apple assigned Cmd-Q... | 13:32 |
|
| however, not a huge time interval :P | 13:32 |
| → yngress joined | 13:32 |
|
cbreak_work
| what's the difference between the alt keys anyway? | 13:32 |
|
| I always found that iritating | 13:32 |
| ← leo2007 left | 13:32 |
|
divVerent
| on German layout? | 13:33 |
| → Vojto joined | 13:33 |
|
divVerent
| left alt corresponds to apple Cmd key | 13:33 |
|
| right alt corresponds to apple Alt key | 13:33 |
| → hebz0rl joined | 13:33 |
|
divVerent
| (called AltGr on PC) | 13:33 |
|
cbreak_work
| they should have picked different names. | 13:33 |
|
divVerent
| whatever AltGr stands for | 13:33 |
| ← Belna left | 13:33 |
|
divVerent
| sure, and it IS a good idea to have two of them | 13:34 |
|
mapreduce
| Can we have symbol shift back please? | 13:34 |
|
Vojto
| Hey guys, can anyone help me with this one please? I created a new branch from master, but forgot to pull first. So it created a new branch, but it doesn't have latest changes from master. How do I get those changes from master to my new branch? | 13:34 |
|
cbreak_work
| anyway, apple's US keyboard layout seems to have @ on shift-2 | 13:34 |
|
divVerent
| note that German layout is one of the few PC layouts that actually HAVE an alternate-character-table key (like german right alt, Apple alt) | 13:34 |
|
| yes, US layouts are similar | 13:34 |
|
cbreak_work
| Vojto: you can try git fetch remotename | 13:34 |
|
divVerent
| also, when will Apple finally release a working VNC server? | 13:34 |
|
cbreak_work
| then you can git rebase remotename/master | 13:34 |
|
divVerent
| one that works with non-US keyboard layouts :P | 13:34 |
|
cbreak_work
| to rebase your own master on top of the one on remotename | 13:34 |
|
Vojto
| cbreak_work: remotename would be what? the master branch? | 13:34 |
|
cbreak_work
| Vojto: the name of the remote | 13:35 |
| ← yngress left | 13:35 |
| → yngress joined | 13:35 |
|
cbreak_work
| if you cloned, probably origin | 13:35 |
|
Luyang
| OK guys I've just installed msysgit. I wanna clone a Git repository from code.google.com Where should I clone it to? Which dir? | 13:35 |
|
| I'm in /git | 13:35 |
|
cbreak_work
| git has no restrictions to repo location I am aware of | 13:35 |
|
Vojto
| cbreak_work: yea it's origin | 13:35 |
|
Luyang
| uhm... But what's a good place | 13:36 |
|
Vojto
| but it didn't work | 13:36 |
|
charon
| Luyang: your earlier comment about initialize.sh seems to indicate you installed the msysgit development environment (msysgit-*.exe) instead of a normal git install (git-*.exe)? | 13:36 |
|
Vojto
| those changes are in master branch | 13:36 |
|
Luyang
| charon yeah I installed the full installer | 13:36 |
|
cbreak_work
| Luyang: in ~/Documents/Projectname? | 13:36 |
|
parasti
| Vojto: fetch is one half of pull, the other half is merge | 13:36 |
|
charon
| Luyang: but anyway, clone it wherever you want to work with it | 13:36 |
|
workmad3
| Luyang: somewhere sensible... I normally have a projects directory somewhere | 13:36 |
|
cbreak_work
| Vojto: sorry | 13:36 |
|
parasti
| Vojto: so, git merge origin/master | 13:36 |
|
cbreak_work
| Luyang: ~/Documents/Projectname? | 13:36 |
|
Luyang
| tell cbreak: oh ok sounds good | 13:36 |
|
| oh | 13:37 |
|
| I'm not using FICS now so I don't have to write "tell" :P | 13:37 |
|
Vojto
| parasti: gotcha, git merge worked | 13:37 |
|
cbreak_work
| Vojto: you shoudl use rebase instead of merge | 13:37 |
|
Vojto
| thanks a lot guys | 13:37 |
|
cbreak_work
| if you want a nice, linear history | 13:37 |
|
Vojto
| cbreak_work: why? | 13:37 |
|
| oh | 13:37 |
|
cbreak_work
| it's more beautiful and simpler to look at/understand | 13:37 |
|
| what did not work with rebase? | 13:37 |
|
Vojto
| I should really read some book on git :-) | 13:37 |
|
| I didn't try rebase | 13:37 |
|
| just fetch | 13:38 |
|
cbreak_work
| then why did you say it didn't work? | 13:38 |
|
workmad3
| Vojto: http://progit.org/ | 13:38 |
|
Vojto
| I did :o) | 13:38 |
| → brntbeer joined | 13:38 |
|
Vojto
| anyway thanks guys | 13:38 |
|
Luyang
| Darn I just installed Git and all that and I just realize the project is using Subversion! :D http://code.google.com/p/opensudoku-android/wiki/SourceCode?tm=4 | 13:38 |
|
Vojto
| this channel always helps whenever I'm stuck on something :-) | 13:38 |
|
divVerent
| Luyang: git-svn ;) | 13:38 |
|
workmad3
| Luyang: git-svn :) | 13:38 |
| ← daed left | 13:38 |
|
cbreak_work
| Luyang: different project | 13:38 |
|
Luyang
| can I use that? | 13:38 |
|
workmad3
| although you'll need to be careful | 13:38 |
|
cbreak_work
| I find git-svn a pain to use | 13:39 |
|
workmad3
| Luyang: yup... it comes with git | 13:39 |
|
divVerent
| git-svn clone --stdlayout http://opensudoku-android.googlecode.com/svn/ | 13:39 |
|
cbreak_work
| it resets history behind your back | 13:39 |
|
Luyang
| I see | 13:39 |
|
| oh | 13:39 |
| → RobertLaptop joined | 13:39 |
|
shruggar
| that's because it's a pain to use ;) | 13:39 |
|
Luyang
| :P | 13:39 |
|
divVerent
| cbreak_work: yes, mostly useful for setting up a public read-only git mirror of the repo | 13:39 |
|
Luyang
| I think I'm gonna use tortoisesvn | 13:39 |
|
workmad3
| cbreak_work: that's because it's having to interface with the PoS that is SVN ;) | 13:39 |
|
cbreak_work
| I know why it does it. But it still sucks :) | 13:39 |
|
divVerent
| but git-svn also is slow :P | 13:40 |
|
cbreak_work
| (I blame SVN for it!) | 13:40 |
|
shruggar
| anyone who wants to help with the git-remote-svn project is very welcome. It's in need of manpower | 13:40 |
|
divVerent
| that is actually what sucks MORE | 13:40 |
|
| I mean | 13:40 |
|
| why does it take so long PER REVISION? | 13:40 |
|
cbreak_work
| it's svn. | 13:40 |
|
divVerent
| is that really the best way to download all the history from svn? | 13:40 |
|
| if yes - how do people back up their svn repos then? | 13:40 |
|
workmad3
| divVerent: it's the only way AFAIK | 13:40 |
|
| well, without being on the server to be able to just copy the directory it's stored in | 13:40 |
|
cbreak_work
| I bet git-svn has some lines of code like: sleep(4242); // Make it look more like the real svn client | 13:41 |
|
divVerent
| I mean... is e.g. svnsync that slow too? | 13:41 |
|
workmad3
| cbreak_work: heh :) | 13:41 |
|
divVerent
| sounds more like svnserve has such a line ;) | 13:41 |
|
shruggar
| divVerent, git-remote-svn is much faster (obscenely fast, by comparison) | 13:41 |
|
charon
| divVerent: you can get a huge speedup if you can get an rsynced copy of the svn server's backing store | 13:41 |
|
divVerent
| shruggar: how well does git-remote-svn already work, BTW? | 13:41 |
|
cbreak_work
| svn server has sleep(4242); // Do not touch this line, or the program will break. Don't ask me why. | 13:42 |
|
divVerent
| I don't know svn protocol, so I can't help much with it probably... but I do have a git-svn managed repo that I could "rebuild" using read-only git-remote-svn | 13:42 |
| ← brntbeer left | 13:42 |
|
workmad3
| cbreak_work: I think the comment is wrong... it's actually '// Only touch this line when the hordes are outside crying for speedups' | 13:42 |
|
divVerent
| still | 13:42 |
|
| WHY is git-svn so slow? | 13:42 |
|
| does it send a request, wait for reply, and only then sends the request for the next revision | 13:43 |
|
shruggar
| divVerent, brainpower works too. For example, a lot of thought it needed for the revision mapper, and none of the "thinking"/"design" part needs any code work | 13:43 |
|
divVerent
| naively asking, why doesn't it simply send all the requests for all the revisions at once? | 13:43 |
| ← gantman left | 13:43 |
|
workmad3
| divVerent: 'check out revision 1, receive revision 1, commit revision 1 to git repo, repeat ad nauseam' | 13:43 |
|
divVerent
| in HTTP they call this "pipelining" | 13:43 |
|
| sure, but why not query the number of revisions, and then send "check out revision 1" to n all at once? | 13:44 |
|
| or is svn protocol so bad that this can't be done? | 13:44 |
|
shruggar
| divVerent, the short answer is "because it's not written in a way that's very maintainable" | 13:44 |
| → brntbeer joined | 13:44 |
| → gergap joined | 13:44 |
|
workmad3
| divVerent: wouldn't surprise me if the SVN protocol just doesn't support that | 13:44 |
|
shruggar
| divVerent, git-remote-svn will probably wind up being the back-end for git-svn, in addition to being its own thing | 13:44 |
|
workmad3
| divVerent: it isn't a typical use-case for SVN after all... | 13:44 |
| ← brntbeer left | 13:45 |
| → WebDragon joined | 13:45 |
| → steven_t joined | 13:46 |
|
steven_t
| hello | 13:46 |
|
shruggar
| 1.6's protocol supports an "svnadmin dump"-like feature, and artagnon has just about finished "svnrdump", an interface which produces near-identical output (which is due to be out in 1.7, and will be maintained separately for 1.6) | 13:46 |
|
steven_t
| how do you completely checkout the previous commit and discard all current changes in the index? | 13:46 |
| ← thiago left | 13:46 |
|
steven_t
| including changes like deleted files etc | 13:46 |
|
shruggar
| steven_t, git reset --hard HEAD | 13:46 |
| → thiago joined | 13:46 |
|
steven_t
| ah thanks | 13:46 |
|
workmad3
| I don't type fast enough :) | 13:46 |
|
steven_t
| ahhhh much better thanks! | 13:47 |
|
divVerent
| shruggar: also I wonder how you are going to handle svn tags in git-remote-svn | 13:47 |
|
workmad3
| steven_t: also, if you want to do that but not discard all your work, use 'git stash' | 13:47 |
|
steven_t
| ahh | 13:47 |
|
divVerent
| or even svn branching at all :P | 13:47 |
| → Jferro joined | 13:48 |
| ← xiong left | 13:48 |
|
divVerent
| there are so many things that svn still can't handle well... | 13:48 |
|
shruggar
| divVerent, it's one of the things that needs people other than me to think about it :P | 13:48 |
| ← scarabx left | 13:48 |
|
divVerent
| also, will git-remote-svn support stone age svn servers? :P | 13:48 |
|
shruggar
| I'm guessingthe support won't go back beyond 1.6 | 13:48 |
|
divVerent
| icculus.org e.g. uses svn 1.3.2 | 13:48 |
|
workmad3
| steven_t: that'll store your work in a local stash, revert back to the previous commit and then later on you can do 'git stash apply' to apply your changes back on top of your working directory | 13:48 |
|
Luyang
| shruggar: some people want auto lock on files before edit | 13:48 |
|
divVerent
| which is the svn server which hosts the repo I am git-svn mirroring | 13:48 |
| → jani_ joined | 13:49 |
|
divVerent
| and the admin refuses to upgrade svn because "who knows what it may break" | 13:49 |
|
jani_
| hi | 13:49 |
|
divVerent
| and yes, this means mergeinfo ALSO does not work | 13:49 |
|
workmad3
| divVerent: bleh, sounds hellish :( | 13:49 |
|
divVerent
| well, git-svn can deal with this repo :P | 13:49 |
|
| if git-remote-svn cannot, I won't be able to use it | 13:49 |
|
| and if git-svn takes over git-remote-svn's access code, I'd sadly have to keep an old version of git-svn then | 13:50 |
| → mar77i_ joined | 13:50 |
|
shruggar
| I assume for older versions someone will make a slower interface | 13:50 |
| → jrmuizel joined | 13:50 |
| ← smuf left | 13:50 |
|
shruggar
| join #git-devel if you want to discuss and/or are interested in helping with git-remote-svn | 13:50 |
|
divVerent
| regarding git-remote-svn, it's not just about speed though | 13:51 |
|
| the "slower interface" idea would probably just work for git-svn | 13:51 |
| → ehabkost joined | 13:51 |
|
shruggar
| it's not just about speed. We're aiming for 1:1 mapping of much more than git-svn currently handles | 13:51 |
|
divVerent
| right | 13:51 |
|
| and THAT is what seems impossible on old svn servers | 13:51 |
|
| that do not even have mergeinfo | 13:51 |
| → brntbeer joined | 13:52 |
|
divVerent
| unless you store stuff as blobs in svn properties, or even files... and THAT would be insane | 13:52 |
|
| and then think of all the mess that might ensue if you then upgrade svn to a version supporting that stuff :P | 13:52 |
| → ereslibre joined | 13:52 |
| ← ereslibre left | 13:52 |
| → ereslibre joined | 13:52 |
| ← mar77i left | 13:52 |
|
shruggar
| at the moment EVERY repository which "supports merginfo" hast history with no mergeinfo data in it | 13:53 |
|
divVerent
| sure | 13:53 |
|
| I just am saying... either git-remote-svn would have to continue treating the repo as a "stupid old" repo forever | 13:53 |
|
| or it'd cause a funny mix of old "git-only" and new "svn compatible" merge info data | 13:54 |
|
| (and then it'd have a lot of fun parsing that mix) | 13:54 |
| ← _stefanos_ left | 13:54 |
|
divVerent
| of course... all this assuming that one CAN represent the info in a svn-compatible way :P | 13:54 |
| → scarabx joined | 13:54 |
| ← mar77i_ left | 13:54 |
|
divVerent
| but I think one can, if one creates "weird-named" branches | 13:55 |
| ← steven_t left | 13:55 |
| → mar77i joined | 13:55 |
|
shruggar
| all git history can be represented as svn history. git's history model is much simpler than svn's, so that's a given | 13:55 |
|
divVerent
| oh? | 13:55 |
| ← rchavik left | 13:55 |
| → froschi1 joined | 13:55 |
|
divVerent
| what can svn represent that git can't? | 13:55 |
| ← cjs left | 13:55 |
|
divVerent
| merges from remote URLs come to my mind, but I have no idea if mergeinfo stores them | 13:56 |
|
shruggar
| mostly things having to do with subtrees. I expect that will involve creating "fake history" in various places to translate it into something git friendly | 13:56 |
|
divVerent
| oh, yes | 13:56 |
|
| nothing prevents me to svn merge from another source than a branch | 13:56 |
|
shruggar
| and that's a good thing | 13:56 |
|
divVerent
| and create messy mergeinfo all over the place :P | 13:56 |
|
| even then, doesn't each commit in svn have a clearly defined set of parents? | 13:58 |
|
| IIRC that should suffice to make git happy | 13:58 |
|
| subtree merges are ALSO very weird in git, but in the end, git doesn't care if the trees of commit objects part of a merge have anything to do with each other or not | 13:58 |
| ← nicoulaj left | 13:58 |
|
cbreak_work
| commits in subversion have always one parent | 13:59 |
|
divVerent
| it does not record any metadata on HOW the merge was performed | 13:59 |
|
cbreak_work
| the previous revision | 13:59 |
| → loincloth joined | 13:59 |
|
divVerent
| cbreak_work: plus possible "extra parents" from mergeinfo | 13:59 |
|
| (if you want to map it back to git) | 13:59 |
| ← loincloth left | 13:59 |
|
divVerent
| but yes, this is highly asymmetric | 13:59 |
|
cbreak_work
| I don't think mergeinfo is about parents. shouldn't it be about folders? | 13:59 |
| → loincloth joined | 14:00 |
| → mar77i_ joined | 14:00 |
| ← freimatz left | 14:00 |
| → lucsky joined | 14:00 |
| ← mar77i left | 14:00 |
|
divVerent
| mergeinfo basically contains a path, and a revision range | 14:00 |
|
| of what has been merged from the path | 14:00 |
| → dreinull joined | 14:00 |
|
shruggar
| I think mergeinfo can represent the same "multiple parent" things that git can | 14:01 |
| → lucianogp joined | 14:01 |
| → hohoho joined | 14:01 |
|
divVerent
| shruggar: if you create history that is "weird to svn", yes :P | 14:01 |
|
| (to deal with git's "nameless branching" that can happen) | 14:01 |
|
lucianogp
| hi everyone, how would you set .gitignore to ignore files (not directories) under "tmp/" ? | 14:02 |
|
shruggar
| by the time svn sees it, all branches are nameless | 14:02 |
| ← corni left | 14:02 |
|
divVerent
| shruggar: so you are not exporting a git branch to branches/foo in svn? | 14:02 |
|
| the problem with svn mergeinfo is that to svn, merge and cherry-pick are the same | 14:03 |
|
shruggar
| we're not doing anything yet, because I'm the only one who has yet said "sure, I'm interested in this aspect of it enough, I'll write something up on it", and I'm a lazy bastard. For the GSoC term, branches will not be supported in any form | 14:03 |
|
divVerent
| (in fact, "svn merge" is no different from cherry-picking all "missing" revisions from the other side, and squashing all that into a single commit, recording what you have merged) | 14:03 |
|
| and an obvious heuristics would be to assume that a git-merge had been performed if ALL revisions from the source branch have been merged | 14:04 |
| ← frakturfreak left | 14:05 |
|
shruggar
| divverent, you seem to have thoughts about the matter, so I encourage you to /join #git-devel ;) | 14:05 |
| ← scarabx left | 14:05 |
|
divVerent
| okay, did. Still, mostly have them from the "user's point of view" (and mathematician's) | 14:06 |
|
lucianogp
| (solved) | 14:06 |
|
shruggar
| that's all I've got on the svn side. I haven't looked at any svn code outside of svnrdump | 14:06 |
|
divVerent
| you know, the main incentive to switch my projects from svn to git was the old svn server on icculus.org :P | 14:07 |
|
| (and that mergeinfo still does not work well) | 14:07 |
| ← loincloth left | 14:07 |
|
divVerent
| with svn I can't simply sync it to a more capable server, work there, and commit back to the original one :P | 14:07 |
|
| (which is exactly the decentral workflow git is good at) | 14:07 |
| ← Fullmoon left | 14:08 |
| → loincloth joined | 14:08 |
| ← dreinull left | 14:09 |
| → herdingcat joined | 14:09 |
| ← sondreb left | 14:10 |
| ← froschi1 left | 14:11 |
|
avar
| charon: that + -> %20 thing looks like a nasty bug, replied on-list | 14:11 |
| ← lucianogp left | 14:12 |
| ← bodom left | 14:13 |
| → frakturfreak joined | 14:13 |
| ← steveire left | 14:13 |
| ← poli left | 14:14 |
| → Tureba joined | 14:14 |
| → steveire joined | 14:15 |
| ← walk left | 14:16 |
| → walk joined | 14:17 |
| ← theduke left | 14:18 |
| ← mar77i_ left | 14:18 |
| → rayners joined | 14:18 |
| → unreal_ joined | 14:19 |
| ← SlickT10 left | 14:19 |
| → [Po]lentino joined | 14:20 |
| → mar77i joined | 14:20 |
| → tjholowaychuk joined | 14:21 |
| ← unreal left | 14:22 |
| unreal_ → unreal | 14:22 |
| → JEEB joined | 14:23 |
| ← unreal left | 14:23 |
| → _stefanos_ joined | 14:24 |
| → corni joined | 14:24 |
| ← mar77i left | 14:24 |
| → sukima joined | 14:25 |
| ← crazy2k left | 14:25 |
| ← travlr left | 14:25 |
|
sukima
| can you stash only one file. Need to do a reset/rebase/loose changes except on file that I want to stash. | 14:26 |
| ← Titosemi left | 14:26 |
| ← schlortoo left | 14:26 |
|
sukima
| s/on file/on one file/ | 14:26 |
| → schlortoo joined | 14:27 |
| → crazy2k joined | 14:27 |
| ← jrmuizel left | 14:27 |
| → warthog9 joined | 14:30 |
| ← ehabkost left | 14:31 |
| → Titosemi joined | 14:31 |
| → ju2wheels joined | 14:32 |
|
avar
| you can do it by adding only that file to the index, then stashing the index iirc | 14:32 |
|
| (or was it the other way around) | 14:32 |
| → ehabkost joined | 14:33 |
| → aaron_ joined | 14:33 |
|
ju2wheels
| Hi, has anyone encountered this error with gitweb 1.7.2: Can't call method "http" on an undefined value at /usr/share/gitweb/index.cgi line 3401. | 14:33 |
| offby1` → offby1 | 14:33 |
| ← offby1 left | 14:33 |
| → offby1 joined | 14:33 |
| aaron_ → dropandho | 14:33 |
| → byang joined | 14:33 |
|
dropandho
| Ilari- you wouldn't believe what the issue was! | 14:33 |
|
ju2wheels
| it started happening after I upgraded git/gitweb yesterday using the git ppa | 14:33 |
| → ajpiano joined | 14:35 |
| ← razvand left | 14:35 |
| → ppawel joined | 14:36 |
| → mdwright_laptop joined | 14:36 |
|
sukima
| avar: that work thanks (former is correct) | 14:37 |
|
| avar: It actually isn't wheather it's in the index or not it weather a file is tracked or not | 14:37 |
| → kpreid joined | 14:38 |
| → brizly joined | 14:38 |
|
mdwright_laptop
| are there any tricks to using git_prompt_info in a huge code base? currently whenever I do anything in the directory, I just lose my terminal for ~20-30 seconds while it runs | 14:39 |
| ← ehabkost left | 14:39 |
| ← ajpiano left | 14:39 |
| ← rajeshsr_ left | 14:39 |
| → aresnick joined | 14:39 |
| → altrux joined | 14:40 |
| → ehabkost joined | 14:40 |
|
avar
| ju2wheels: what's on that line in your gitweb? | 14:40 |
|
| if (defined $cgi->http('HTTP_ACCEPT') && ? | 14:41 |
|
| sukima never heard of git_prompt_info and is now furiously googling the term | 14:41 |
|
koruptid
| I'm new to git and working off a remote repo... I just tried to push to it and was denied. something about "remote: error: refusing to update checked out branch: refs/heads/master" | 14:41 |
|
| what do I need to do to be able to push? | 14:41 |
|
avar
| koruptid: don't push to non-bare repositories | 14:41 |
|
divVerent
| do you have to? | 14:42 |
|
koruptid
| avar: what the hell is a non-bare versus bare repo? | 14:42 |
|
divVerent
| you have the option to push to another named branch on the non-bare remote repository | 14:42 |
|
avar
| koruptid: see man git-init --bare | 14:42 |
|
divVerent
| koruptid: bare repo = "server stored repository" (only ever accessed via git commands) | 14:42 |
|
| non-bare repo = "working repository" (has a work tree where you can work normally) | 14:42 |
|
cbreak_work
| bare repos are used for communication | 14:42 |
|
| non-bare repos are used for working | 14:42 |
| → jani__ joined | 14:42 |
| → hearit joined | 14:43 |
|
divVerent
| a non-bare repo "almost" (apart from a config setting) contains a bare repo in its .git subdirectory | 14:43 |
|
koruptid
| avar: I imported this from an SVN repo... is there a way to turn it into a bare repo? | 14:43 |
|
sukima
| koruptid: non-bare == working repository AKA local. the directory your working in. bare == on a server with no working files | 14:43 |
|
divVerent
| koruptid: yes, but then git svn will not like much to work on it any more | 14:43 |
| → roop_ joined | 14:43 |
|
avar
| koruptid: with git-svn ? | 14:43 |
|
divVerent
| are you still going to use git svn on it? | 14:43 |
|
koruptid
| avar... yes | 14:43 |
|
| divVerent: no | 14:44 |
|
divVerent
| that is good :P | 14:44 |
|
koruptid
| divVerent: I'm making a clean break from subversion | 14:44 |
|
divVerent
| then just git clone --bare it | 14:44 |
|
koruptid
| okay | 14:44 |
|
divVerent
| wait | 14:44 |
|
| no, clone --mirror | 14:44 |
| ← warthog9 left | 14:44 |
|
koruptid
| annnnnnnnd then re-remote the working copy? | 14:44 |
|
divVerent
| you could | 14:44 |
|
| you could also just rename the current repo's dir, and put the new bare repo in that place :P | 14:44 |
|
koruptid
| I have changes that need to go bad to the master and this is over a gig of files | 14:45 |
|
| *s/bad/back/ | 14:45 |
| ← hearit left | 14:45 |
| ← henkboom left | 14:45 |
| → unreal joined | 14:45 |
| → ajpiano joined | 14:45 |
|
divVerent
| of course, you also can quite easily "bare-ify" the existing repo you have, by editing .git/config to set it as bare, delete everything but .git, and move everything in .git to .., then delete the empty .git | 14:45 |
|
| but clone --mirror is so much easier | 14:46 |
| ← jani_ left | 14:46 |
| → hearit joined | 14:46 |
|
divVerent
| and also gets rid of (but not irreversibly) the git-svn metadata | 14:46 |
|
| (git-svn can quickly restore it if you set up git-svn on your cloned repo again, though) | 14:46 |
| ← gergap left | 14:46 |
| ← kpreid left | 14:46 |
|
avar
| ju2wheels: still there? | 14:46 |
|
koruptid
| divVerent: already removed the metaaaaaaaaaadata with a clone... followed instructions from a site. they just failed to mention making it a bare repo | 14:47 |
|
divVerent
| oh | 14:47 |
|
| ok | 14:47 |
|
| clone --mirror would have been the way to go :P | 14:47 |
|
koruptid
| yeah, I know that now. :P | 14:47 |
|
divVerent
| did you have any branches in the svn repo you want to keep? | 14:47 |
|
| or do you only want master anyway? | 14:47 |
|
mdwright_laptop
| sukima: git_prompt_info is just a wrapper for __git_ps1 is appears. | 14:47 |
| → thwart joined | 14:47 |
|
koruptid
| no, just master | 14:48 |
|
divVerent
| good, then you lost nothing important :P | 14:48 |
|
| with the non-mirror clone | 14:48 |
| → jstemmer joined | 14:48 |
| → henkboom joined | 14:48 |
| ← unreal left | 14:48 |
|
divVerent
| still, if you DO want to use git-svn again on the repo, for whatever reason - "git svn fetch" works on a bare repo too | 14:48 |
|
| it's how I do my git-svn-mirrors | 14:49 |
|
cbreak_work
| can you dcommit back or is that one-way? | 14:49 |
|
divVerent
| I can dcommit back | 14:49 |
|
| but from my client | 14:49 |
|
| I don't push to the master branch and have the server dcommit back | 14:49 |
| ← apsysoev left | 14:49 |
|
cbreak_work
| hmm. | 14:49 |
|
| doesn't dcommiting require meta? | 14:49 |
|
divVerent
| yes | 14:49 |
|
| but it generates it quickly | 14:49 |
|
| it doesn't check out every single rev from svn for that :P | 14:50 |
|
| it just parses the git-svn commit messages to regenerate it | 14:50 |
|
koruptid
| mirror worked great... thanks so much! | 14:50 |
|
divVerent
| to dcommit, my "client" repos needed two git-config settings | 14:50 |
|
| [svn-remote "svn"] | 14:51 |
|
| url = svn://svn.icculus.org/twilight | 14:51 |
| → Isaiahs joined | 14:51 |
|
divVerent
| fetch = trunk/darkplaces:refs/heads/master | 14:51 |
|
| that's all I need | 14:51 |
|
| note that git-svn would never create this "fetch" line :P | 14:51 |
|
Isaiahs
| Sendme a pic of ur dic 214916 0073 | 14:51 |
| ← steveire left | 14:51 |
|
divVerent
| (refs/heads/master as "target") | 14:51 |
| → steveire joined | 14:51 |
| ← Isaiahs left | 14:51 |
|
divVerent
| but then dcommit knows where the svn history is, and works fine | 14:52 |
|
koruptid
| wtf was that? | 14:52 |
|
divVerent
| a spambot | 14:52 |
|
| well, looks like it | 14:52 |
|
koruptid
| a really crappy one | 14:52 |
|
bgy
| Is it possible after a rebase is done to come back to the previous state ? like no rebasing happened | 14:52 |
|
divVerent
| bgy: yes | 14:52 |
|
| git reflog | 14:52 |
|
| locate the state you want | 14:52 |
|
ju2wheels
| @avar: yes sorry, had stepped away. that is the line on mine | 14:52 |
|
divVerent
| git reset --hard thatstate | 14:52 |
|
koruptid
| anyway... thank you so much... that worked perfectly | 14:52 |
|
bgy
| divVerent, is it reversible ? | 14:53 |
|
avar
| ju2wheels: How do you reproduce it, and how are you running gitweb? | 14:53 |
|
divVerent
| bgy: "not really" | 14:53 |
| ← _stefanos_ left | 14:53 |
|
divVerent
| or what | 14:53 |
|
avar
| ju2wheels: Anyway, gitweb had a bunch of changes recently. If you can bisect that one that'd probably be very helpful. were you using 1.7.1 before? | 14:53 |
|
divVerent
| the reset? it is | 14:53 |
|
| just use the same method again | 14:53 |
| → gbacon joined | 14:53 |
|
avar
| ju2wheels: That's definitely some internal error in it | 14:53 |
|
divVerent
| but this only works for a certain time till the reflog is pruned | 14:53 |
|
| and not over the network | 14:53 |
|
| .oO( I wish bare repos had a reflog too, when pushing e.g. ) | 14:54 |
|
bgy
| divVerent, Ok, thank you | 14:54 |
| ← hearit left | 14:54 |
|
ju2wheels
| avar: yes i was using 1.7.1 and im actually trying to revert, i just removed the git ppa but apt is being a pain and wont let me revert | 14:54 |
| ← thwart left | 14:54 |
|
koruptid
| one last question hopefully for a while... is there a way to "roll up" all your existing commits to a certain point? I have a lot offff SVN commits where I didn't have comments | 14:54 |
|
ju2wheels
| so far I have removed the ppa, removed all the git packages and done a apt update but now it says it has no source for git packages | 14:54 |
|
divVerent
| cbreak_work: anyway, if you want, I can write up how to make this "as good as possible git-svn mirror" that even lets people work on git branches on the repo | 14:55 |
|
| (just not push to master) | 14:55 |
| → unreal joined | 14:55 |
|
divVerent
| it involves a cronjob with git svn fetch AND git fetch, gitolite to do this access restriction so nobody messes with git-svn-owned history, and two git config commands to set up the client side to be able to dcommit | 14:55 |
| agib|away → agib | 14:55 |
| ← ehabkost left | 14:55 |
| → koruptid_ joined | 14:56 |
|
ju2wheels
| well more specifically it says "E: Package git has no installation candidate" | 14:56 |
|
divVerent
| try git-core | 14:56 |
| → ehabkost joined | 14:56 |
| ← koruptid left | 14:56 |
|
avar
| ju2wheels: well, a bisect would be really useful, if you can manage that. | 14:56 |
| koruptid_ → koruptid | 14:56 |
| → WebDragon|laptop joined | 14:57 |
|
cbreak_work
| ju2wheels: apparently git is not this git but some other git no one else has used in this century | 14:57 |
|
avar
| aoeu git (master) $ git bisect start master v1.7.1 -- gitweb/ | 14:57 |
|
| Already on 'master' | 14:57 |
|
| Bisecting: 12 revisions left to test after this (roughly 4 steps) | 14:57 |
|
| ju2wheels: like that | 14:57 |
| ← vu3rdd left | 14:58 |
| ← rseifert left | 14:58 |
| ← WebDragon left | 14:59 |
| WebDragon|laptop → WebDragon | 14:59 |
| → shade_ joined | 14:59 |
|
ju2wheels
| cbreak_work: lol ty very much, thats what i get for using tab completion. i got it reverted. | 15:00 |
|
| avar: readding git ppa now to get you that bisect | 15:00 |
| ← jds left | 15:01 |
| → AAABeef joined | 15:02 |
| → callumacrae joined | 15:02 |
| ← brizly left | 15:03 |
|
callumacrae
| Does anyone have any experience with GitX? | 15:03 |
| ← ehabkost left | 15:03 |
|
SamothUK
| callumacrae: A bit, what about it? | 15:04 |
|
callumacrae
| Well, trying to get it to work | 15:04 |
|
| t won't | 15:04 |
|
| It* | 15:04 |
|
cbreak_work
| divVerent: I hope I'll never have to work with subversion again. | 15:04 |
| → roger_padactor joined | 15:04 |
|
avar
| ju2wheels: You can't bisect with the ppa, you have to check out and build from source | 15:04 |
|
callumacrae
| http://gitx.frim.nl/user_manual.html It says "When you open for the first time"., but I didn't see it | 15:04 |
|
SamothUK
| callumacrae: in what way does it 'not work'? | 15:04 |
|
avar
| ju2wheels: But how does the error happen? | 15:05 |
|
SamothUK
| callumacrae: ahhh | 15:05 |
| → bentob0x joined | 15:05 |
|
callumacrae
| Clicking git:// links doesn't work either | 15:05 |
|
koruptid
| anybody use tortisegit? | 15:05 |
|
cbreak_work
| once. | 15:05 |
| → jrmuizel joined | 15:05 |
|
cbreak_work
| I didn't like it much though | 15:05 |
|
SamothUK
| callumacrae: does the program start at all? | 15:06 |
|
callumacrae
| Yup | 15:06 |
|
| It's in the dock | 15:06 |
|
koruptid
| cbreak_work: I'm used to using tortise so I'm thinking about giving it a try | 15:06 |
|
ju2wheels
| avar: i honestly have no idea, i had gitweb working fine yesterday before upgrading to 1.7.2, with the exception that i couldnt get it to see any of my repositories (which is really what im trying to fix) | 15:06 |
|
| im using the same config i had then | 15:06 |
| ← jtauber left | 15:06 |
|
avar
| ju2wheels: yes, but what are you *doing* with gitweb when you get the error? | 15:07 |
|
| i.e. what request url etc. | 15:07 |
|
SamothUK
| callumacrae: if you get a menu, File->Open and then select a directory with a .git directory in it, or the .git directory itself | 15:07 |
|
avar
| ow is it set up, what runs it | 15:07 |
|
callumacrae
| Okay, thanks | 15:07 |
| ← simplechat left | 15:07 |
| → Belna joined | 15:07 |
| shade_ → shade\ | 15:08 |
| → ehabkost joined | 15:08 |
|
ju2wheels
| avar: all i do is try to load localhost/gitweb/ in my browser and nothing came up | 15:08 |
| ← mdwright_laptop left | 15:08 |
|
SamothUK
| callumacrae: if that doesn't work, it might take a bit more thinking :) | 15:08 |
| → shader joined | 15:08 |
| → drubies joined | 15:09 |
|
callumacrae
| Yeah it's worked :) | 15:09 |
|
| Thanks | 15:09 |
| → jtauber joined | 15:09 |
|
shader
| how do I see which commits I haven't pushed to remote? | 15:09 |
|
avar
| ju2wheels: What webserver do you use for gitweb, what exec model? | 15:09 |
|
| i.e. cgi, fastcgi, etc.? | 15:09 |
|
SamothUK
| callumacrae: hurrah :) | 15:09 |
|
shader
| I've merged in a few times, but I don't know how far back the last push was | 15:09 |
|
avar
| shader: git log remote/branch..branch | 15:09 |
|
shader
| thanks | 15:09 |
|
SamothUK
| callumacrae: doing the "Enable Terminal Usage" from the menu, so you can run 'gitx' from the command line is really useful | 15:09 |
|
ju2wheels
| avar: im using an unmodified apache install, only change was enabling rewrite and cgi modules. Heres some more info on what i was doing http://ubuntuforums.org/showthread.php?t=1536963 | 15:10 |
|
callumacrae
| I don't really use git that often, I probably won't need it, but thanks LD | 15:10 |
|
| :D* | 15:10 |
| ← hyperair left | 15:10 |
| → hyperair joined | 15:11 |
|
bgy
| How to squash commits ? | 15:12 |
|
| I took a look to rebasing, but I already rebased my branch with my master, so I'm not sure if I can rebase master master to squash my commits | 15:13 |
| → smuf joined | 15:13 |
| → kipras joined | 15:13 |
| ← cbreak_work left | 15:14 |
| ← ereslibre left | 15:14 |
| ← callumacrae left | 15:14 |
|
shader
| is there a way to see a log of dangling commits? | 15:14 |
|
ju2wheels
| avar: apache, cgi | 15:14 |
|
shader
| bgy, how many commits do you want to squash? | 15:15 |
| → Heimidal joined | 15:15 |
|
bgy
| shad, ± a dozen | 15:16 |
|
| shader | 15:16 |
|
| sorry shad : ) | 15:16 |
|
shader
| try "git rebase -i head~12" while on master | 15:16 |
| ← herdingcat left | 15:17 |
|
bgy
| shader, but my commits where inserted between others commits when rebasing | 15:17 |
|
shader
| that's fine | 15:17 |
| → nimrod10 joined | 15:18 |
|
shader
| it'll give you the option to rebase interactively, and you can reorder them before squashing | 15:18 |
|
| you can put a particular commit id in place of head~12 if you know where you want to rebase onto (i.e., the furthest back commit you want to touch) | 15:19 |
| ← Luyang left | 15:19 |
| ← sftp left | 15:19 |
|
sitaram
| Oxf13: ping | 15:20 |
|
bgy
| shader, ok, i'll give it a try | 15:20 |
|
sitaram
| Oxf13: take the latest master, set all 4 variables described in http://github.com/sitaramc/gitolite/blob/master/doc/big-config.mkd#other_optimisations to 1 in the rc file | 15:20 |
|
shader
| you might want to check your EDITOR env var first | 15:20 |
| ← jani__ left | 15:21 |
|
shader
| bgy, it will default to vim, if you're ok with that | 15:21 |
|
avar
| ju2wheels: Iteresting, can you bisect it or not, then the bug will be fixed upstream | 15:21 |
|
bgy
| shader, yep it's ok : ) | 15:21 |
|
| Is there a way to know the ~n with git log ? | 15:22 |
| → poli joined | 15:22 |
| → sftp joined | 15:22 |
|
traviscline
| what's the cleanest way to to find the most recent ancestor of a branch? | 15:22 |
|
shader
| I usually just use the first few digits of hash | 15:22 |
|
| easier than counting | 15:22 |
| ← steveire left | 15:22 |
| → ericindc joined | 15:22 |
|
shader
| @ bgy | 15:23 |
| ← ericindc left | 15:23 |
|
avar
| bgy: this in your gitconfig: abbrev-commit = "!f() { git name-rev --tags \"$@\" | sed 's|.*tags/||;s|\\([0-9a-f]\\{7\\}\\).*|\\1|'; }; f" | 15:23 |
|
| ? | 15:23 |
|
| or git describe | 15:23 |
|
ju2wheels
| avar: i dont have it installed from source, whats the general process there? Clone the git repo and just run that command you gave me? | 15:24 |
|
bgy
| i don't have such alias (abbrev-commit) | 15:24 |
| → MattDiPasquale joined | 15:25 |
| → steveire joined | 15:25 |
| ← steveire left | 15:25 |
| → steveire joined | 15:25 |
|
avar
| ju2wheels: clone the git.git repo, then start the bisect I gave you | 15:25 |
| ← schlortoo left | 15:25 |
| ← hobodave left | 15:25 |
|
avar
| then for each bisect version compile it (described in INSTALL), install it and try if the problem goes away | 15:25 |
| → SlickT10 joined | 15:25 |
|
avar
| if so => git bisect good, if not => git bisect bad | 15:25 |
|
| keep doing that until it isolates the bad commit | 15:25 |
|
sitaram
| Oxf13: the pastebin you sent: I assume that was a sample of the config you plan to use? Looks good. Duplicates -- dont worry too much; gitolite unique's them anyway (the beauty of using hashes in most places) | 15:26 |
| → hagabaka joined | 15:26 |
| ← skoop left | 15:27 |
| ← SlickT10 left | 15:27 |
| → slonopotamus joined | 15:28 |
| → andrewl joined | 15:29 |
| ← hadees left | 15:29 |
| → Sho_ joined | 15:30 |
| ← ehabkost left | 15:31 |
| → ehabkost joined | 15:32 |
|
Sho_
| hi guys. we got a server hosting master repositories that also allows developers to make personal clones on the server (thanks to gitolite). a common scenario is for these to become out of sync, and the developers having to push up a lot of history the master repo on the same storage already has. I was wondering if it's possible for the server to check for any objects it could hardlink from the master repo during the push negotiation, and then tell the client it doesn't | 15:33 |
|
| need to push those, but only the objects not shared between the master and the server-side personal clone. does anyone know? | 15:33 |
| → adamv joined | 15:34 |
| ← thiago left | 15:35 |
| → Adriannom joined | 15:35 |
| ← eighty4 left | 15:35 |
| → spaceonline joined | 15:36 |
| → bdimcheff joined | 15:36 |
| ← steffkes- left | 15:37 |
|
Adriannom
| hi. i just did git reset --hard, but i'd forgotten about a change i still needed. i've done a commit since then, is there any way of recovering the unstaged changes i lost? i'm presuming no... | 15:37 |
| ← byang left | 15:38 |
|
slonopotamus
| Adriannom: no. that's why you usually don't want reset --hard | 15:38 |
|
Adriannom
| aye. luckily it wasn't too much work. thanks | 15:39 |
| ← rajeshsr left | 15:39 |
| → rajeshsr joined | 15:40 |
| ← ppawel left | 15:40 |
| → kpreid joined | 15:40 |
| agib → agib|away | 15:41 |
| → neunon joined | 15:42 |
| ← neunon left | 15:42 |
| → neunon joined | 15:42 |
| ← barismetin left | 15:43 |
| → tedoc2000 joined | 15:43 |
| → schlortoo joined | 15:43 |
| ← David-T left | 15:44 |
| → David-T joined | 15:44 |
| → sh1mmer joined | 15:45 |
| ← kpreid left | 15:45 |
| → grey-monkey joined | 15:45 |
|
ju2wheels
| does anyone know if gitolite supports LDAP users and groups? and if it does is there any documentation for how to set that up? | 15:46 |
| ← walk left | 15:46 |
| ← entrosca left | 15:46 |
| IslandUsurper → IslandUsurperAFK | 15:47 |
|
grey-monkey
| I have a git repo on a server that 4 devs ssh into and pull the current branch. The permissions on the files pulled in .git/objects keeps changing to g-w. Is there a way to change this so that every time a user pulls on the server the .git/ojects that are created have g+w permissions? | 15:47 |
| → walk joined | 15:47 |
| ← tedoc2000 left | 15:47 |
| ← schlortoo left | 15:48 |
|
ju2wheels
| grey-monkey: are you using gitolite? It has a UMASK setting in .gitolite.rc that may fix that i think. | 15:48 |
|
grey-monkey
| we are not using gitolite | 15:49 |
| → WebDragon|laptop joined | 15:49 |
|
sitaram
| ju2wheels: gitolite does not have to support it. If you have an sshd that used LDAP as its authkeys database (there is one such project on github) and that sshd has the eqvt of "command=" in authkeys file, it will work anyway | 15:49 |
| ← WebDragon left | 15:49 |
| WebDragon|laptop → WebDragon | 15:49 |
| ← squentin left | 15:49 |
|
sitaram
| grey-monkey: chmod g+ws, and make all users part of that group | 15:50 |
|
| chmod -R g+ws I think | 15:50 |
| ← hyperair left | 15:50 |
| ← dr_win left | 15:50 |
| → hyperair joined | 15:51 |
| → dr_win joined | 15:51 |
| ← Heimidal left | 15:51 |
|
grey-monkey
| sitaram: I'll try that. Thanks! | 15:52 |
| → jeromegn joined | 15:52 |
| → coderdad_ joined | 15:53 |
|
jeromegn
| what could be the cause of git pulling telling me everything is up to date, but when I view the repository on github, the files aren't the same? | 15:53 |
| → raichoo joined | 15:54 |
| ← ehabkost left | 15:54 |
| ← ralf_e left | 15:54 |
| ← coderdad left | 15:54 |
| → carter joined | 15:54 |
| → ehabkost joined | 15:55 |
| → ralf_e joined | 15:55 |
|
ju2wheels
| sitaram: could you elaborate on what you mean by equivalent of command=, i thought that was added by gitolite automatically when you add the user keys? I guess im misunderstanding how that works. So, my company does have sshd setup to use ldap, but i what i actually want is for all users to push through git@server but still be able to differentiate who is pushing based on their gitolite keys. Can i just add a key named after an LDAP use | 15:56 |
| ← matty[pac] left | 15:56 |
| ← slonopotamus left | 15:57 |
|
ju2wheels
| so say, can i have "git" setup as the gitolite user, and in /keys add pub keys for ldapuser1, ldapuser2 and setup gitolite.conf repo perms for ldapuser1/2 and it will work automatically? | 15:58 |
| → schlortoo joined | 15:59 |
|
ju2wheels
| so far ive only setup gitolite on my personal server so thats why im asking about ldap since my final setup will be on my companies server | 16:00 |
| → squentin joined | 16:00 |
| ← squentin left | 16:00 |
| → squentin joined | 16:00 |
| ← roop_ left | 16:00 |
| ← Oszo left | 16:01 |
| ← Elfe left | 16:01 |
| → Oszo joined | 16:02 |
| → juanjoc joined | 16:02 |
| ← Oszo left | 16:04 |
| → thiago joined | 16:04 |
| ← thiago left | 16:04 |
| → thiago joined | 16:04 |
|
sitaram
| ju2wheels: step back a bit | 16:04 |
|
| ju2wheels: think about what normal sshd gets, and what it passes to gitolite | 16:04 |
| ← tstclair left | 16:04 |
| ← ehabkost left | 16:05 |
| → tstclair joined | 16:05 |
|
sitaram
| ju2wheels: sshd gets a pubkey from the user. It matches it in the authkeys file (this is the part that needs patching to use LDAP, AFAIK). After that it should just run the "command=" in the authorized_keys file. | 16:05 |
| → ehabkost joined | 16:05 |
|
sitaram
| ju2wheels: but when the authkeys file itself does not exist (pubkeys are stored in LDAP, by some other, company wide, means), where does the "command=" parameter come from? Somewhere in LDAP, thats where | 16:06 |
|
| sitaram looks for Grum... | 16:06 |
|
sitaram
| he's done LDAP + git at least, if not gitolite | 16:06 |
| ← jeromegn left | 16:07 |
| tchan1 → tchan | 16:07 |
| ← tchan left | 16:07 |
| → tchan joined | 16:07 |
| → Yuffster joined | 16:08 |
|
ju2wheels
| sitaram: ok that makes sense to me so far. | 16:08 |
| ← posciak left | 16:08 |
| agib|away → agib | 16:08 |
|
sitaram
| so far? that's all I have :) | 16:10 |
| ← schlortoo left | 16:10 |
| → engla joined | 16:10 |
| ← engla left | 16:10 |
| → engla joined | 16:10 |
|
ju2wheels
| so i guess what i really should be looking for is how to define/change command= ldap | 16:10 |
|
| ill start doing some research | 16:10 |
|
sitaram
| ju2wheels: your LDAP-backed ssh must somehow be able to infer the userid from the pubkey match, and invoke "[path]gl-auth-command user" | 16:10 |
|
ju2wheels
| k | 16:11 |
| → schlortoo joined | 16:11 |
| → kiliman joined | 16:12 |
| → hobodave joined | 16:14 |
| → copelco joined | 16:15 |
| → mkramer joined | 16:15 |
|
pcc1
| I'm seeing post-receive-email mails classified as spam. anyone ever seen this? | 16:15 |
|
mkramer
| hi... | 16:15 |
| → steveire_ joined | 16:15 |
| ← steveire_ left | 16:15 |
| → steveire_ joined | 16:15 |
|
mkramer
| I'd like to get a git status-like output for the difference between HEAD~3 and HEAD~2 | 16:16 |
|
| not a git diff-like output | 16:16 |
| ← giallu left | 16:16 |
| ← beni- left | 16:17 |
| ← workmad3 left | 16:18 |
|
rane_
| "--stat" or something.. | 16:18 |
| ← steveire left | 16:18 |
| ← walk left | 16:19 |
|
engla
| mkramer: you could use diff with --name-status to get a file list | 16:19 |
| → nevyn^_ joined | 16:19 |
| → steveire joined | 16:19 |
| ← steveire left | 16:19 |
| → steveire joined | 16:19 |
| ← schlortoo left | 16:19 |
| → stephenp joined | 16:20 |
| → walk joined | 16:20 |
| → schlortoo joined | 16:20 |
|
stephenp
| how can i copy the contents of HEAD exactly as they are to a different branch? I have a working HEAD,and I just want to make that HEAD the new tip of master. | 16:21 |
| → Eaven joined | 16:22 |
|
Fangorn_
| stephenp: isn't that what merge is for? | 16:22 |
|
wereHamster
| stephenp: checkout the different branch, then reset --hard or read-tree or checkout the other ref | 16:22 |
| → rajeshsr_ joined | 16:22 |
| → beni- joined | 16:23 |
|
wereHamster
| if you want to preserve history, you should use the 'ours' merge strategy | 16:23 |
|
stephenp
| Fangorn_: for some reason that i fail to comprehend, merge screws it all up, and deletes lines, files, etc. it doesn't actually tell me it does this. | 16:23 |
|
sitaram
| wereHamster: you were trying to use LDAP auth + git weren't you? | 16:23 |
|
wereHamster
| sitaram: yes. DOes it work now? :) | 16:24 |
|
engla
| merge keeps the history and changes of both branches | 16:24 |
| ← lbt_ left | 16:24 |
|
stephenp
| so, if I'm on "live" and i want to copy it's head to master, i should do "git checkout master", "git checkout <sha1 of live head>" then commit? | 16:24 |
| ← intripoon left | 16:25 |
| ← ehabkost left | 16:25 |
|
wereHamster
| stephenp: no. git merge -s ours master; git checkout master; git merge live | 16:25 |
|
sitaram
| wereHamster: how does LDAP authenticate and pass to git? Can it be made to pass a "command" to run also? | 16:25 |
|
wereHamster
| stephenp: that would preserve history | 16:25 |
|
stephenp
| alright, let me try that. | 16:25 |
| ← corni left | 16:26 |
|
wereHamster
| sitaram: I solved it by using ForceCommand in /etc/ssh/sshd_config | 16:26 |
| → ehabkost joined | 16:26 |
|
wereHamster
| sitaram: so users authenticate with their own login and my script takes from there | 16:26 |
|
stephenp
| wereHamster: i should start out on branch "live" right? | 16:26 |
| ← tstclair left | 16:27 |
|
wereHamster
| stephenp: yes, as you said (if I'm on "live" and ..) | 16:27 |
|
stephenp
| i've been futzing around with this for about two and half hours, and i'm starting to doubt my own sanity. | 16:27 |
|
wereHamster
| if you don't want to preserve the history, there are easier ways to do that | 16:27 |
| ← schlortoo left | 16:27 |
|
engla
| stephenp: if you don't like the history graph after that you can turn the merges into a squash merge/normal commit if you reset --soft on master back to before the merge (then commit) | 16:28 |
|
sitaram
| wereHamster: so your LDAP-backed sshd still honors /etc/ssh/sshd_config then... hmm | 16:28 |
|
wereHamster
| sitaram: yes. LDAP is integrated through pam | 16:28 |
| ← sukima left | 16:29 |
| ← Cesario left | 16:29 |
|
stephenp
| okay, that seems to have worked. | 16:29 |
|
sitaram
| wereHamster: ok | 16:30 |
| ← nevyn^_ left | 16:30 |
| → kenichi joined | 16:30 |
| → rubydiamond joined | 16:31 |
|
wereHamster
| sitaram: that solution, however, has its own 'problems'.. Namely, how to make sure users can clone each others' repos | 16:31 |
|
stephenp
| jesus. that was hard. | 16:31 |
| → kpreid joined | 16:31 |
| ← shruggar left | 16:32 |
| → rajeshsr__ joined | 16:32 |
| ← rajeshsr_ left | 16:32 |
|
stephenp
| is there some kind of special trick to merging two branches where each have had independent changes made to them, without losing your mind? | 16:33 |
| → stringoO joined | 16:33 |
| → schlortoo joined | 16:33 |
|
selckin
| git merge branch1 branch2 | 16:34 |
|
wereHamster
| that would merge three branches ;) | 16:34 |
| ← ehabkost left | 16:34 |
|
selckin
| right | 16:34 |
|
stephenp
| yeah, that's what i did. | 16:34 |
|
sitaram
| wereHamster: unless your force_command basically runs a setuid-to-"git" and passes to the normal gitolite auth command with the current $USER as arg1 | 16:34 |
| ← beni- left | 16:34 |
|
engla
| stephenp: look at gitk branch1...branch2 | 16:35 |
|
wereHamster
| sitaram: or init all repos with --shared=all | 16:35 |
|
stephenp
| yeah, no x-windows. | 16:35 |
|
engla
| stephenp: "X" or X Window System | 16:36 |
|
stephenp
| right, none of that. | 16:36 |
|
| headless server. | 16:36 |
| ← steveire left | 16:36 |
| ← koruptid left | 16:37 |
|
stephenp
| probably why it's giving me such fits, because i normally have gitx to look at and keep track of the branches and where everything is. | 16:37 |
|
sitaram
| wereHamster: but you're saying this does preserve SSH_ORIGINAL_COMMAND env var at least? | 16:37 |
| → steveire joined | 16:37 |
| ← steveire left | 16:37 |
| → steveire joined | 16:37 |
|
engla
| git log --graph --oneline <branches> and so is pretty nice | 16:37 |
| → ehabkost_ joined | 16:38 |
| ← schlortoo left | 16:38 |
|
wereHamster
| yes | 16:38 |
|
stephenp
| engla: that's much better. | 16:39 |
| → schlortoo joined | 16:39 |
|
stephenp
| now, should i bother getting rid of this rats nest of commits and merges i seem to have made, or is it safer just to leave it? | 16:39 |
|
Fangorn_
| stephenp: add --decorate=full to see tags,branches,etc | 16:39 |
| ← RobertLaptop left | 16:39 |
|
wereHamster
| sitaram: sysadmins try to avoid suid, so I've chosen init --shared=all. But I think running all git commands suid-to-git would be much easier | 16:39 |
|
| stephenp: or clone the repo to a host where you have X11 | 16:40 |
|
| and then inspect it with gitk | 16:40 |
| ← alin-gym left | 16:40 |
| → alin joined | 16:40 |
| ← aspotashev left | 16:42 |
| → shruggar joined | 16:42 |
|
stephenp
| yarg... "git push origin" error: unable to create temporary sha1 filename ./objects/79: File exists | 16:42 |
|
wereHamster
| stephenp: permission problems | 16:42 |
| → elmob joined | 16:44 |
| ← ralf_e left | 16:44 |
| ← alin left | 16:44 |
|
thierryp
| hi folks - a newbie question | 16:44 |
|
| I accidentally ran 'git add' on a file that I was not intending to add to the repo | 16:45 |
|
| how ca i undo that ? | 16:45 |
|
FauxFaux
| thierryp: git status tells you. | 16:45 |
| ← ehabkost_ left | 16:45 |
| → aspotashev joined | 16:46 |
|
thierryp
| excellent - thanks | 16:46 |
| → ehabkost_ joined | 16:46 |
| → Andy joined | 16:46 |
| → jacobwg joined | 16:46 |
| ← coderdad_ left | 16:46 |
| Andy → Guest24432 | 16:46 |
| ← hagabaka left | 16:47 |
|
jacobwg
| I can't get a post-commit hook to work - no matter what I do, I don't think it's being called. Any ideas? | 16:47 |
| → hagebake joined | 16:47 |
| ← hagebake left | 16:47 |
| → hagebake joined | 16:47 |
| → coderdad joined | 16:47 |
|
wereHamster
| jacobwg: is the hook executable? | 16:47 |
|
jacobwg
| permissions are rwxr-xr-x | 16:48 |
| ← aspotashev left | 16:48 |
|
jacobwg
| So, I would think so. | 16:48 |
|
wereHamster
| jacobwg: does it work if you execute it manually? | 16:49 |
| ← ehabkost_ left | 16:49 |
|
jacobwg
| Yes, it does what is supposed to do. | 16:49 |
| → ehabkost_ joined | 16:49 |
| ← poli left | 16:49 |
| WebDragon → WebDragon|away | 16:50 |
| → tstclair joined | 16:50 |
| → tedoc2000 joined | 16:50 |
| → MrGlass_Work joined | 16:50 |
| ← coderdad left | 16:51 |
|
jacobwg
| wereHamster: I'm not exactly sure if I'm doing this the right way... I have a special user for git and have setup ssh access with my local account (so I'm pushing to gituser@localhost:repo.git). The post-commit hook is located at the server and is owned by the git user. | 16:51 |
| → macek joined | 16:52 |
|
wereHamster
| yeah, that won't work | 16:52 |
|
| because commits are *local* in git | 16:52 |
|
MrGlass_Work
| hi, how do I commit a version of a file, then have git stop tracking the changes to it but not remove it from the repo? for instance, my project has a file with DB username and pass, I want to commit it with "user" and "pass" as those fields but then change it in my project. Similarly, my project has a /log folder that needs to exist, but I want it to not have contents | 16:52 |
|
wereHamster
| jacobwg: man githooks -> post/pre-receive/update | 16:52 |
| → RobertLaptop joined | 16:52 |
|
jacobwg
| So, I want post-receive? | 16:52 |
| → amerine joined | 16:53 |
|
jacobwg
| wereHamster: Do you know the difference between update and post-receive? | 16:53 |
|
wereHamster
| MrGlass_Work: usually you wouldn't track that config file at all. And only keep a template in the repo which each user can copy and edit | 16:53 |
| ← macek left | 16:53 |
|
wereHamster
| jacobwg: the man page explains it | 16:54 |
|
| MrGlass_Work: is it a rails project by any chance? | 16:54 |
|
jacobwg
| wereHamster: is that the man page like as in the terminal command man? | 16:54 |
|
wereHamster
| jacobwg: yes, a normal unix man page | 16:54 |
| → slonopotamus joined | 16:55 |
|
jacobwg
| wereHamster: Perfect! Thank you very much for your help! | 16:56 |
|
MrGlass_Work
| wereHamster: i dont want it tracked, i just want that tempalte to be in the repo | 16:56 |
|
| im new to git | 16:56 |
|
| and nto rails, symfony | 16:56 |
|
| but ive needed to do this for lots of diff things | 16:57 |
|
| and ive seen it done, jsut dont know how | 16:57 |
|
wereHamster
| MrGlass_Work: git mv the/config.file the/config.file.template; echo 'the/config.file' >> .gitignore; git add .gitignore; git commit -m 'Provide a template config file for the/config.file' | 16:58 |
|
| theen instruct your users to 'cp the/config.file.template the/config.files; vi the/config.file' | 16:58 |
| → eighty4 joined | 16:59 |
|
MrGlass_Work
| ah i see | 16:59 |
|
| hmm | 16:59 |
| ← slonopotamus left | 16:59 |
| ← ehabkost_ left | 16:59 |
| → florianb joined | 17:00 |
|
avar
| ju2wheels: get anywhere with bisecting? | 17:00 |
| → ehabkost_ joined | 17:00 |
| ← steveire left | 17:01 |
| → macek joined | 17:01 |
| → poli joined | 17:01 |
|
ju2wheels
| avar: im actually @ work now, so i wont be able to do that testing until later tonight or tomorrow. | 17:02 |
| IslandUsurperAFK → IslandUsurper | 17:03 |
| ← shruggar left | 17:03 |
| ← neunon left | 17:03 |
| ← Dreamer3 left | 17:04 |
| → leo2007 joined | 17:04 |
| ← ehabkost_ left | 17:04 |
| ← LiamH left | 17:05 |
| → tanoku joined | 17:05 |
| ← gebi left | 17:05 |
| → steveire joined | 17:05 |
| ← steveire left | 17:05 |
| → steveire joined | 17:05 |
|
avar
| great, anyway, if you mail the list now with the error perhaps someone will have fixed it by that time anyway | 17:05 |
| ← brntbeer left | 17:05 |
| ← unreal left | 17:06 |
|
avar
| ju2wheels: I recomend that, make sure to cc Jakub Narebski [email@hidden.address] | 17:06 |
| → unreal joined | 17:06 |
| ← tstclair left | 17:06 |
| → apropos joined | 17:06 |
|
ju2wheels
| avar: ok, will do. | 17:06 |
| → Dreamer3 joined | 17:07 |
| → ehabkost_ joined | 17:07 |
| ← khmarbaise left | 17:08 |
| → rajeshsr_ joined | 17:08 |
| hsa2|gone → hsa2 | 17:08 |
|
hsa2
| hello | 17:08 |
| ← hsa2 left | 17:08 |
| → hsa2 joined | 17:08 |
|
hsa2
| i am having trouble :P | 17:08 |
|
apropos
| Aren't we all | 17:08 |
|
hsa2
| i am not in sync with git repo, i want to ignore the changes i made and get the last version from git repo | 17:09 |
|
| i tried git checkout -f | 17:09 |
|
| but it didn't work | 17:09 |
|
Ilari
| dropandho: Sorry, was away... | 17:09 |
| ← David-T left | 17:09 |
|
Ilari
| hsa2: fetch first. Checkout is local-only operation. | 17:10 |
| ← Dreamer3 left | 17:10 |
|
avar
| git fetch && git reset --hard origin/master | 17:10 |
|
hsa2
| why git rest is necessary avar? | 17:11 |
| ← Vojto left | 17:11 |
|
wereHamster
| to update your working tree | 17:11 |
|
engla
| and change the whole branch's state | 17:12 |
| → karamorf joined | 17:12 |
| → level09 joined | 17:12 |
|
wereHamster
| alternatively you could checkout origin/master insteat of the reset | 17:12 |
| ← daaku left | 17:12 |
|
level09
| how do I make a non-master branch a Active branch (instead of master) | 17:12 |
| ← rajeshsr__ left | 17:12 |
|
wereHamster
| level09: git checkout <that other branch> | 17:13 |
| → Elfe joined | 17:13 |
|
apropos
| I'm using gitolite, and I tried to push, but: "remote: Use of uninitialized value in do "file" at hooks/update line 32. | 17:13 |
|
| remote: Null filename used at hooks/update line 32." | 17:13 |
|
| The strange thing is though, from a different UNIX user account it works perfectly. | 17:13 |
|
karamorf
| ok I'm completely lost with git. I made a repo on github, managed to commit various config files from my home folder (.vimrc). Now I'm trying to pull those files down into my home folder on a second computer. first, i guess I can't create a WC on a non empty folder? symbolic links can fix that i suppose ... Did a clone but I can't get it to pull down the files from the repo (which i see are there from github.com). Also a svn convert if that matters... | 17:14 |
| ← schlortoo left | 17:14 |
| → schlortoo joined | 17:14 |
|
level09
| wereHamster: thanks, however, when I clone the master will still be the default branch , no ? | 17:14 |
|
wereHamster
| karamorf: first push the canhnges back to github (git push), the go to the otheer computer and do 'git clone [email@hidden.address] | 17:14 |
|
bdrewery
| karamorf: did you git push to github? | 17:14 |
|
wereHamster
| level09: yes | 17:14 |
|
level09
| can i edit the config so the head is pointing to the non-master | 17:15 |
| → David-T joined | 17:15 |
|
wereHamster
| level09: for that you'd have to edit the config file on the server | 17:15 |
|
| level09: on the server: git checkout <the other branch> | 17:15 |
|
karamorf
| Yes, I did a git push | 17:15 |
|
level09
| okiez thanks | 17:15 |
| ← tobiassjosten left | 17:15 |
|
ju2wheels
| does anyone know if i can get gitolite to list the actual repos available under a wildcard directory instead of listing the wild card dir itself when you do ssh git@server? | 17:16 |
| ← sh1mmer left | 17:16 |
|
sitaram
| ju2wheels: http://github.com/sitaramc/gitolite/blob/pu/doc/4-wildcard-repositories.mkd#reporting | 17:17 |
| ← EricInBNE left | 17:17 |
| → theoros joined | 17:17 |
|
mkramer
| engla: that works, thanks | 17:18 |
|
| engla: re: diff --name-status | 17:18 |
| Guest24432 → AndyJ | 17:18 |
| ← j4cbo left | 17:18 |
| → slonopotamus joined | 17:19 |
| → t0rc joined | 17:19 |
|
dropandho
| Ilari- no problem...i got it to work! there is something funny going on...i need to use the server's ip | 17:19 |
|
Ilari
| dropandho: Vhosts. | 17:19 |
| ← davidc_ left | 17:20 |
| ← jstemmer left | 17:20 |
| ← pipegeek left | 17:20 |
| ← hagebake left | 17:20 |
| → neunon joined | 17:21 |
|
ju2wheels
| sitaram: awesome ty | 17:21 |
| ← kipras left | 17:22 |
| → brntbeer joined | 17:22 |
| ← [Po]lentino left | 17:22 |
| ← Sho_ left | 17:22 |
|
Ilari
| dropandho: Can you also try what happens if you add some weird name for correct address in local /etc/hosts (or find one from DNS) and try to use that. What happens? | 17:22 |
|
dropandho
| def...i'll try that out! | 17:23 |
| → alex__c2022 joined | 17:24 |
|
Ilari
| dropandho: Since that sort of thing causes that weird name to be sent to server as vhost. | 17:24 |
| ← trivol left | 17:24 |
| ← level09 left | 17:25 |
| ← tatsuya left | 17:25 |
| ← thiago left | 17:25 |
| ← engla left | 17:25 |
| ← schlortoo left | 17:25 |
|
dropandho
| gotcha....it worked with that random entry in my hosts file | 17:26 |
| ← ehabkost_ left | 17:26 |
| ← curtana left | 17:26 |
| ← bluenovember left | 17:26 |
| → ehabkost_ joined | 17:27 |
| → krawek joined | 17:27 |
|
dropandho
| the very first line of my 000default file for apache has: NameVirtualHost *:80 and all the directives are wrapped in <VirtualHost *:80> | 17:27 |
| ← opalka left | 17:30 |
| ← tanoku left | 17:30 |
| ← macek left | 17:32 |
| → daaku joined | 17:33 |
| → roop_ joined | 17:33 |
|
Ilari
| dropandho: Seems that one vhost entry is somehow treated specially... | 17:34 |
| ← neunon left | 17:35 |
|
dropandho
| yeah, it seems i'm missing a ServerAlias or ServerName somewhere | 17:35 |
|
| i'll look into that | 17:35 |
| → level09 joined | 17:36 |
|
dropandho
| in the meantime, I have all my git repos in their respective subdirectory under /home/shared | 17:36 |
|
apropos
| It looks like gitolite has forgotten to add the command= bit in the .authorized_keys file. | 17:37 |
| → thiago joined | 17:37 |
|
dropandho
| however I only want some of the repos to be accessable via httpbackend...and i'm having trouble | 17:37 |
|
bdrewery
| works great for me | 17:37 |
|
dropandho
| i thought it only works for repos that have the following set: [http] receivepack = true | 17:38 |
| ← kiliman left | 17:38 |
| → brizly joined | 17:38 |
| ← ehabkost_ left | 17:38 |
|
dropandho
| but httpbackend works for anything in that home/shared dir because that's where i'm pointing its ScriptAlias | 17:39 |
| ← kpreid left | 17:39 |
| ← hohoho left | 17:39 |
| ← apropos left | 17:40 |
| ← Elfe left | 17:41 |
| → ehabkost_ joined | 17:41 |
| → Elfe_ joined | 17:41 |
| ← _Vi left | 17:41 |
| → dreiss joined | 17:41 |
| ← brizly left | 17:42 |
| → _Vi joined | 17:42 |
| ← flaguy48 left | 17:42 |
| ← steveire left | 17:43 |
| ← radioman-lt left | 17:43 |
| → Sigma joined | 17:43 |
| → radioman-lt joined | 17:43 |
| ← ehabkost_ left | 17:44 |
| ← tildeequals left | 17:44 |
| → koruptid joined | 17:44 |
| → ehabkost_ joined | 17:45 |
|
koruptid
| is it possible to put a git repo inside an ignored directory within another repo? | 17:45 |
| ← level09 left | 17:46 |
|
leo2007
| koruptid: did you try it first? | 17:46 |
|
| koruptid: clone a repo and abuse it. | 17:46 |
|
koruptid
| leo2007: just tried it... messed up on a command and that's why I got confused for a sec... forgot to cd into the repo | 17:47 |
|
leo2007
| ;) | 17:47 |
| ← jacobwg left | 17:47 |
| ← henkboom left | 17:48 |
| ← ehabkost_ left | 17:49 |
| → macek joined | 17:50 |
| → ehabkost_ joined | 17:50 |
| ← ehabkost_ left | 17:50 |
| → henkboom joined | 17:51 |
| → steveire joined | 17:51 |
| ← steveire left | 17:51 |
| → steveire joined | 17:51 |
| → slonopotamus_ joined | 17:52 |
| ← zimbres left | 17:52 |
| ← karamorf left | 17:52 |
| → ehabkost_ joined | 17:53 |
| WebDragon|away → WebDragon | 17:54 |
| ← slonopotamus left | 17:54 |
|
Ilari
| dropandho: Don't override the receivepack setting unless you know what you are doing. | 17:54 |
| ← brntbeer left | 17:55 |
| slonopotamus_ → slonopotamus | 17:56 |
| ← shwaiil left | 17:56 |
|
dropandho
| I'm not trying to override the settings...I just noticed that it doesn't matter if I have the [http] receivepack = true in my configs they all seem to be accepting httpbackend....i was expecting only the repos with receivepack set to true to accept httpbackend | 17:57 |
| → loinclot_ joined | 17:58 |
| → KelebekSohbet935 joined | 17:58 |
| → schlortoo joined | 17:58 |
|
dropandho
| any idea how to stop the other repos from accepting httpbackend...even though they are all under the same main directory that I pointed the httpbackend ScriptAlias to? | 17:59 |
| hsa2 → hsa2|gone | 17:59 |
| ← loincloth left | 18:00 |
| ← parasti left | 18:01 |
| ← hobodave left | 18:01 |
| → hobodave joined | 18:01 |
| ← Belna left | 18:02 |
| → babusri joined | 18:03 |
| ← indygreg left | 18:04 |
| ← slonopotamus left | 18:06 |
| → parasti joined | 18:06 |
| ← hyperair left | 18:06 |
| ← schlortoo left | 18:06 |
| → hyperair joined | 18:06 |
| → rseifert joined | 18:07 |
| ← gbacon left | 18:07 |
| ← daaku left | 18:07 |
| → Jippi_mac joined | 18:08 |
| ← hyperair left | 18:09 |
| → brizly joined | 18:09 |
| ← galderz left | 18:09 |
| ← Yuffster left | 18:10 |
| ← poli left | 18:12 |
| → poli joined | 18:12 |
|
KelebekSohbet935
| Selam.. | 18:12 |
|
| nasýlsýnýz | 18:12 |
|
| www.trsohbetci.com | 18:12 |
| → hyperair joined | 18:12 |
| → schlortoo joined | 18:12 |
|
KelebekSohbet935
| 94.102.5.93 | 18:12 |
|
leo2007
| spammer? | 18:12 |
| → sh1mmer joined | 18:13 |
| → gebi joined | 18:13 |
|
rajeshsr_
| hi all, is it possible to have something like an update-only branch. That is i have a branch, which will "track" certain files only from some branch and updates it say with something like git pull when am in that branch.. | 18:13 |
|
| ? | 18:13 |
| → tstclair joined | 18:14 |
| ← florianb left | 18:15 |
| ← loinclot_ left | 18:16 |
| → loincloth joined | 18:16 |
| ← roop_ left | 18:17 |
| ← steveire_ left | 18:18 |
| ← hyperair left | 18:18 |
| → WebDragon|laptop joined | 18:20 |
| ← WebDragon left | 18:20 |
| WebDragon|laptop → WebDragon | 18:21 |
| → whatdoesitwant joined | 18:21 |
| ← KelebekSohbet935 left | 18:21 |
| ← Rearden left | 18:21 |
| ← redondos left | 18:22 |
| → redondos joined | 18:22 |
| → hyperair joined | 18:22 |
| → corni joined | 18:23 |
| → aspotashev joined | 18:23 |
| ← _Vi left | 18:24 |
| agib → agib|away | 18:25 |
| ← mbroeker left | 18:25 |
| → tildeequals joined | 18:25 |
| ← tjholowaychuk left | 18:25 |
| → tjholowaychuk joined | 18:26 |
| → ProperN[out] joined | 18:26 |
| ← ProperN[out] left | 18:26 |
| → ProperN[out] joined | 18:26 |
| ← ProperNoun left | 18:27 |
| ProperN[out] → ProperNoun | 18:27 |
| ← felipec left | 18:28 |
| → CoinOpeBoy joined | 18:28 |
| ← schlortoo left | 18:29 |
| ← walk left | 18:29 |
| → justin-george joined | 18:30 |
| ← gebi left | 18:30 |
| → workmad3 joined | 18:32 |
|
charon
| rajeshsr_: you cannot merge only a subset of files (you can merge a full sub*tree* with the subtree strategy) | 18:32 |
|
| other than that it seems your problem is already handled by the remote-tracking branches shown by 'git branch -r' | 18:33 |
| ← thiago left | 18:33 |
| → thiago joined | 18:33 |
| slyphon → slyphon_away | 18:34 |
| → nkuttler joined | 18:35 |
| ← BlindOracle left | 18:35 |
| → vinse joined | 18:35 |
| ← bburhans_ left | 18:35 |
| → bburhans joined | 18:35 |
|
nkuttler
| can git svn give me a list of revisons on the remove server? | 18:36 |
| → Titosemi_ joined | 18:36 |
|
wereHamster
| nkuttler: svn log $(git config --git svn-remote.svn.url) | 18:37 |
| ← Titosemi left | 18:38 |
| Titosemi_ → Titosemi | 18:38 |
|
wereHamster
| or, git log git-svn (or whatever the branch name is that git-svn is importing into your git repo | 18:38 |
| → ska-fan joined | 18:38 |
| → BlindOracle joined | 18:38 |
| → gebi joined | 18:38 |
| agib|away → agib | 18:39 |
| ← krawek left | 18:40 |
| → krawek joined | 18:40 |
|
nkuttler
| wereHamster: thanks! | 18:41 |
| → schlortoo joined | 18:45 |
| ← SvenDowideit left | 18:45 |
| → SvenDowideit_ joined | 18:45 |
| SvenDowideit_ → SvenDowideit | 18:45 |
| → brntbeer joined | 18:46 |
| ← mw_ left | 18:46 |
|
rajeshsr_
| charon, can you throw more light? i don't quite get it.. | 18:46 |
| → w00tz joined | 18:46 |
| ← marc_kdab_ left | 18:47 |
| → marc_kdab_ joined | 18:48 |
|
aresnick
| Hi! I have a directory which is full of git repositories. I'd like to create a single git repository out of all them. What's the right way to do this? | 18:48 |
|
Grum
| not to do that, its not svn :( | 18:49 |
| → elitecoder joined | 18:49 |
|
aresnick
| Grum: Right =) Those repositories no longer represent separate projects, so they don't need to be managed separately. Does that make sense? | 18:49 |
|
rajeshsr_
| aresnick: copy all of them into a folder and do git init! :) | 18:49 |
|
theoros
| modulify? | 18:49 |
|
rajeshsr_
| u want commits to be preserved? | 18:49 |
|
Grum
| do you care for the history? | 18:49 |
|
aresnick
| rajeshsr_: I'd like to keep their commits, history, etc. | 18:49 |
|
| Grum, rajeshsr_: yeah | 18:50 |
|
rajeshsr_
| a grand merge? | 18:50 |
|
elitecoder
| I have a setup where developers are using a shared login on a dev server to work, and using public keys to commit would make all commits done by the same user... is there anything I can do? | 18:50 |
|
selckin
| plenty of google htis on how to do that | 18:50 |
|
| elitecoder: gitolite | 18:50 |
|
theoros
| elitecoder: there are ways of modifying the commit user depending on which key is in use | 18:51 |
| ← jtauber left | 18:51 |
|
elitecoder
| theoros: is there a manual page I can look at? | 18:51 |
|
aresnick
| selckin, rajeshsr_: I should be looking for how to do a "grand merge?" | 18:51 |
|
elitecoder
| I prefer not to setup additional software, and I may not have the option as our origin repo is on a hosted service... | 18:52 |
|
rajeshsr_
| nope, i didn't mean that :) | 18:52 |
|
elitecoder
| beanstalk | 18:52 |
|
selckin
| aresnick: http://www.google.com/search?hl=en&q=git+combine++multiple+repositories&aq=f&aqi=&aql=&oq=&gs_rfai= | 18:52 |
| → p3rg2z joined | 18:52 |
|
aresnick
| selckin: Oh =p Thanks | 18:52 |
| ← babusri left | 18:53 |
| → npgoog joined | 18:53 |
| ← tvw left | 18:53 |
| → ShooterM_ joined | 18:54 |
|
npgoog
| Is there a git command to do the equivalent to `git branch | egrep '^\*' | cut -c3`? | 18:54 |
|
wereHamster
| npgoog: what does that do? what are you trying to get? the current brnach name? | 18:55 |
|
| what if you are on detached head? | 18:55 |
|
npgoog
| Yeah, I want the current branch name, and I want git to tell me if I am on a detached head automatically. I have some scripts that want this information. | 18:56 |
| ← poli left | 18:56 |
|
wereHamster
| npgoog: git symbolic-ref HEAD | 18:56 |
|
| observe its output and return value | 18:57 |
|
npgoog
| Thanks | 18:57 |
|
theoros
| git rev-parse HEAD? | 18:57 |
|
wereHamster
| exit value | 18:57 |
|
| theoros: rev-parse gives you the sha | 18:57 |
|
theoros
| oh the name, right | 18:57 |
|
| name-rev is alright too if you don't want the full thing | 18:58 |
| → rorist joined | 18:58 |
| → ralf_e joined | 18:58 |
|
npgoog
| I want it to cut out the remotes/ or refs/heads/ part, so name-rev seems the closest, but not precisely what I am looking for. | 19:00 |
| → unix_lappy joined | 19:00 |
|
unix_lappy
| anyone know how to delete a directory under the auspices of git without having to kill gitcache.exe? | 19:01 |
|
npgoog
| I need to try these out some more and come back later once I know what I want. | 19:01 |
| ← npgoog left | 19:01 |
| → dfr|work joined | 19:01 |
|
dfr|work
| hey all. I am trying to drop a couple of commits via git reabse -i master | 19:01 |
| → flaguy48 joined | 19:02 |
|
dfr|work
| but for some reason I get "nothing to commit" after "Finished one cherry pick". Here's a pastie: http://pastie.org/1057412 | 19:02 |
|
| help? :) | 19:02 |
|
workmad3
| dfr|work: doesn't that work like a merge... it commits the result automatically, as long as there weren't any conflicts? | 19:03 |
|
unix_lappy
| halp? | 19:03 |
|
| is this a windows specific git error? | 19:04 |
|
dfr|work
| workmad3, I was hoping that it would just work like a regular rebase, except not worrying about patches by those commits | 19:04 |
|
workmad3
| unix_lappy: what's the error? | 19:04 |
|
patrikf
| dfr|work: does that commit introduce any changes that are not already on master? | 19:04 |
| → vmiklos joined | 19:04 |
| ← sh1mmer left | 19:04 |
|
unix_lappy
| rm: cannot remove directory `ninject1': Permission denied | 19:05 |
|
patrikf
| (or in the commits before it) | 19:05 |
|
dfr|work
| patrikf, the ones I threw out? um.. it's a bit tricky. probably not the ref specs | 19:05 |
|
patrikf
| dfr|work: what happens if you try to manually cherry-pick it? | 19:06 |
|
| (in the state git rebase leaves you in) | 19:06 |
|
Grum
| unix_lappy: if a folder is open or a file in it you might not be able to delete it yeah | 19:06 |
|
unix_lappy
| Grum: even if Im' in the parent? | 19:06 |
| → jtauber joined | 19:06 |
|
dfr|work
| patrikf, seems fine: Finished one cherry-pick" | 19:07 |
|
unix_lappy
| rm: cannot remove directory `ninject1/.git/objects/pack': Directory not empty | 19:07 |
|
| rm: cannot remove directory `ninject1/.git/objects': Directory not empty | 19:07 |
| → slonopotamus joined | 19:07 |
|
patrikf
| dfr|work: then just "rebase --skip" and go on | 19:07 |
|
unix_lappy
| do I have to remove it from source control first? | 19:07 |
| ← rayners left | 19:07 |
|
dfr|work
| patrikf, should i restart the process? | 19:07 |
|
workmad3
| unix_lappy: git rm -r directory | 19:07 |
|
dfr|work
| patrikf, can you perhaps explain what's going on? | 19:07 |
|
unix_lappy
| workmad3: I thought rm just unstaged things? | 19:08 |
|
| fatal: Not a git repository (or any of the parent directories): .git | 19:08 |
|
workmad3
| unix_lappy: no, rm removes it from the commit, | 19:08 |
|
unix_lappy
| ^ | 19:08 |
|
patrikf
| dfr|work: nah, I don't see what's going on, what I proposed might "just work". if you want to find out why it's acting so strange, start a new branch at master and cherry-pick the commits by hand and see if you get an error | 19:09 |
|
dfr|work
| patrikf, so what I did right now is did git rebase --continue That would pick a commit and apply it alright, and then do same thing. So I had keep doing until I hit a conflict. Perhaps that conflict has something to do with it. | 19:10 |
|
| patrikf, does git rebase -i just cherry-pick each of the commits and applies? | 19:10 |
|
unix_lappy
| workmad3: killing tgitcache.exe worked... | 19:10 |
|
| I might need to update the git implementation for windows...it's been a few months. | 19:10 |
| → programble joined | 19:11 |
| ← thiago left | 19:11 |
|
patrikf
| dfr|work: yeah, everything you can do with rebase can also be done with checkout, reset, cherry-pick, commit | 19:12 |
| ← phantomcircuit left | 19:12 |
|
theoros
| dfr|work: git rebase rebases | 19:12 |
|
patrikf
| dfr|work: hmh... what does "git status" say after rebase stops? | 19:12 |
|
theoros
| and before | 19:12 |
|
dfr|work
| patrikf, it says I'm not on a branch | 19:12 |
|
patrikf
| dfr|work: paste it, please | 19:13 |
|
| pastebin, that is | 19:13 |
| → thiago joined | 19:13 |
| ← jtauber left | 19:15 |
|
dfr|work
| patrikf, http://pastie.org/1057432 | 19:16 |
| → Guest85240 joined | 19:17 |
| → entrigan joined | 19:17 |
|
patrikf
| dfr|work: well, the only thing I can think of is that you have a commit in there that doesn't introduce any new changes. but try manual cherry-picking and see what you ge | 19:18 |
|
| get* | 19:18 |
| ← hobodave left | 19:18 |
| ← Jferro left | 19:19 |
| → carutsu_ joined | 19:19 |
|
dfr|work
| patrikf, mhmm.. For some reason when I cherry-pick the only active commit on a separate branch off master/sync, it doesn't show up anything in logs. | 19:21 |
|
| weird. | 19:21 |
| → rajeshsr__ joined | 19:23 |
|
unix_lappy
| fatal: Not a git repository (or any of the parent directories): .git | 19:23 |
| ← rajeshsr_ left | 19:23 |
|
unix_lappy
| when I do git rm -r ninject | 19:23 |
|
| on a new cloned repo. | 19:23 |
|
wereHamster
| unix_lappy: did you cd into the directory you cloned? | 19:23 |
| → rajeshsr_ joined | 19:23 |
|
unix_lappy
| wereHamster: nope, still in the parent dir. | 19:24 |
|
wereHamster
| and what is ninject? the repo you just cloned? | 19:24 |
|
unix_lappy
| yar. | 19:24 |
|
workmad3
| unix_lappy: oh, so you're not in the repository | 19:25 |
|
wereHamster
| uhm, why did you clone it if you want to remove it right away? | 19:25 |
|
unix_lappy
| nope. | 19:25 |
|
workmad3
| unix_lappy: then why would you expect git commands to work? | 19:25 |
|
dfr|work
| patrikf, figure what was going on. One of the patches I accidently was gonna let to keep going was modifying files introduced by the skipped commits | 19:25 |
|
unix_lappy
| fatal: This operation must be run in a work tree | 19:26 |
|
dfr|work
| patrikf, once I skipped them as well, it started behaving more expectedly :D | 19:26 |
|
unix_lappy
| workmad3: when in the .git directory. | 19:26 |
|
wereHamster
| unix_lappy: what was your clone command you used? | 19:26 |
|
unix_lappy
| git clone [github-uri].git | 19:26 |
|
wereHamster
| unix_lappy: git clone ...github.com:yourusername/theproject.git; cd theproject; git whatever-command-you-are-trying-to-run | 19:27 |
| ← rajeshsr__ left | 19:27 |
| → datapolitical joined | 19:28 |
|
unix_lappy
| wereHamster: ok, so git rm -r ran, but it still says another process is tracking the files. | 19:29 |
|
| how can I get tgitcache.exe to release the files? | 19:29 |
|
wereHamster
| ask whoever wrote tgitcache.exe | 19:29 |
|
unix_lappy
| wereHamster: yea it seems it's a tgitcache.exe specific issue. | 19:29 |
|
| http://code.google.com/p/tortoisegit/issues/detail?id=401 | 19:30 |
| → _Vi joined | 19:30 |
| slyphon_away → slyphon | 19:30 |
|
dfr|work
| patrikf, all in all, I got it to work! thanks :D | 19:31 |
| ← schlortoo left | 19:31 |
| → xiong joined | 19:31 |
| → Alagar joined | 19:31 |
| → schlortoo joined | 19:32 |
| ← workmad3 left | 19:33 |
|
unix_lappy
| heh, all this work for nothing. | 19:35 |
|
| seems ninject only has compiled builds and not source for windows phone 7 up on github. | 19:35 |
| ← rpg left | 19:38 |
| → Heimidal joined | 19:39 |
| ← Heimidal left | 19:39 |
| ← MrGlass_Work left | 19:40 |
| → alester joined | 19:40 |
| → mw_ joined | 19:41 |
| → peta joined | 19:41 |
| → tvw joined | 19:41 |
|
peta
| hello everybody | 19:41 |
| → name joined | 19:42 |
| name → __name__ | 19:42 |
|
elitecoder
| How do I fix this? lol fatal: cannot do a partial commit during a merge. | 19:42 |
|
wereHamster
| elitecoder: run 'git commit' instead of 'git commit <file>' | 19:42 |
| → MrGlass_Work joined | 19:43 |
| → zenazn_ joined | 19:43 |
| ← zenazn left | 19:43 |
| → jtauber joined | 19:43 |
|
peta
| i'm new to git and have the following problem ... i dunno how i managed to commit on "(no branch)". now i commited a bunch of changes to "(no branch)" and want to merge all but a few files into the other branch called "master". how can i do this? | 19:45 |
|
wereHamster
| peta: https://git.wiki.kernel.org/index.php/GitFaq#detached | 19:45 |
|
peta
| bingo ... "detached" was the keyword | 19:45 |
| ← rubydiamond left | 19:46 |
| ← schlortoo left | 19:47 |
| ← theoros left | 19:50 |
| → schlortoo joined | 19:50 |
| → beni- joined | 19:52 |
| → Risar joined | 19:52 |
|
peta
| wereHamster: you saved my life :) | 19:52 |
|
| thanks | 19:52 |
| ← schlortoo left | 19:55 |
| → Guest79374 joined | 19:55 |
| → poli joined | 19:58 |
| ← Guest85240 left | 19:58 |
| → lulinha joined | 19:59 |
| ← lulinha left | 19:59 |
| ← drubies left | 19:59 |
| ← Guest79374 left | 20:01 |
| → lulinha joined | 20:02 |
| ← lulinha left | 20:02 |
| → theoros joined | 20:02 |
| → sagsousuke joined | 20:03 |
| ← gebi left | 20:04 |
| → schlortoo joined | 20:06 |
| ← squentin left | 20:07 |
| ← flaguy48 left | 20:11 |
| → WebDragon|laptop joined | 20:11 |
| → gebi joined | 20:11 |
| → jhunter joined | 20:12 |
| → squentin joined | 20:12 |
| → fod joined | 20:13 |
| ← WebDragon left | 20:14 |
| WebDragon|laptop → WebDragon | 20:14 |
| ← unix_lappy left | 20:14 |
|
jhunter
| newbie q: how can I visually see I've got stuff in previous branches I need to merge down to master? | 20:14 |
|
| ie - branch X has changed - so when i do a git merge i know what head to reference | 20:15 |
| ← schlortoo left | 20:15 |
| → schlortoo joined | 20:15 |
| ← jtauber left | 20:17 |
| slyphon → slyphon_away | 20:17 |
| ← rajeshsr_ left | 20:18 |
|
patrikf
| jhunter: you can check if anything changed with git log master..branch | 20:19 |
| ← macek left | 20:20 |
|
patrikf
| but i usually don't have too many branches, as you can safely delete them once they're merged | 20:20 |
| → hipe joined | 20:21 |
| ← TML left | 20:22 |
|
wereHamster
| also: git branch --[no-]merged | 20:22 |
|
ajpiano
| hey, i am trying to git reset --hard HEAD a repo that i just cloned, cause i forgot to setup my crlf stuff properly first | 20:22 |
|
| but i'm getting Permission denied when trying to create file as part of that reset | 20:22 |
|
| only on one particular file...what can i check on the file to see what's wrong | 20:23 |
|
| it's not set to readonly | 20:23 |
| ← fod left | 20:24 |
| ← schlortoo left | 20:24 |
| ← cannonball left | 20:25 |
|
elitecoder
| wereHamster: Thanks! :] | 20:26 |
| → heewa joined | 20:26 |
| ← ralf_e left | 20:27 |
|
heewa
| I know I can specify a path for log (eg: git log -- path/), but that's relative to my current dir. Any way to specify relative to git repo root? | 20:27 |
| slyphon_away → slyphon | 20:29 |
| ← leo2007 left | 20:29 |
|
wereHamster
| heewa: git log -- (git rev-parse --cdup)path/ | 20:29 |
|
| $(git rev-parse --show-cdup) | 20:29 |
| ← poli left | 20:29 |
|
heewa
| ewww...fine. Thanks :) | 20:29 |
| → schlortoo joined | 20:30 |
| → jtauber joined | 20:31 |
| → kpreid joined | 20:35 |
| → flaguy48 joined | 20:35 |
| ← schlortoo left | 20:36 |
|
jhunter
| ty | 20:40 |
| → Jferro joined | 20:44 |
| → Octalot joined | 20:49 |
| ← jaql left | 20:49 |
| ← elitecoder left | 20:54 |
| ← WebDragon left | 20:55 |
| ← murr4y left | 20:56 |
|
ljungk
| I'm trying out git-cherry-pick for the first time. It seems like it messes up the connection between local and remote branches. | 20:57 |
| agib → agib|away | 20:57 |
|
ljungk
| so I have a branch A which has a number of commits on top of master | 20:58 |
|
| now I want to cherry pick the first of those commits into master, and then rebase A on master | 20:58 |
| → hasen_ joined | 20:58 |
| ← peta left | 20:59 |
| ← shader left | 20:59 |
| ← zenazn_ left | 20:59 |
| → flazz_ joined | 20:59 |
|
ljungk
| this, however, results in the corresponding remote branches NOT being rebased | 20:59 |
|
ajpiano
| if when i clone a repo, every single is instantly marked changed, how should i set up my CRLF settings on the machine to make sure that doesn't happen | 21:00 |
| agib|away → agib | 21:00 |
| → zenazn joined | 21:00 |
| ← hasen left | 21:02 |
| ← flazz left | 21:02 |
| ← dreiss left | 21:02 |
|
charon
| ljungk: can you spell that in git commands? | 21:03 |
| → bewst joined | 21:03 |
|
vinse
| ajpiano: we use auto.crlf = input. this works for everyone here as long as no one is actually using an editor that needs CRLF | 21:03 |
|
charon
| (so that we know precisely what you tried) | 21:03 |
|
ajpiano
| hm, that's what i'm trying . | 21:04 |
|
vinse
| ajpiano: i'm pretty sure it means that all CRLF will be converted to LF any time you pull/clone etc | 21:04 |
|
charon
| ajpiano: if you clone a repository with CRLFs in it on a machine with autocrlf enabled, madness will ensue | 21:04 |
|
vinse
| ajpiano: ah wait, if some one already comitted stuff that's using CRLF then you're out of luck | 21:04 |
|
ajpiano
| ohhhh | 21:04 |
| ← flazz_ left | 21:04 |
|
ajpiano
| i'm in the very early stages of setting up a repo | 21:05 |
|
charon
| disclaimer: i don't use this stuff since i don't work on windows and refuse to work with anyone who does (well, almost ;) | 21:05 |
|
vinse
| ajpiano: when you commit it, then it'll be fine. but if they dont have the 'input' setting also, then you'll eventually need to hunt them down and kill them | 21:05 |
|
charon
| but basically you have to either disable autocrlf at least for this repo, or fix the repo. | 21:05 |
| ← alester left | 21:05 |
| ← ProperNoun left | 21:06 |
|
ljungk
| charon: sure | 21:06 |
|
vinse
| ajpiano: yeah, i did all that a while ago, i remember lots of problems like you're having, but they eventually worked themselves out, but everyone needs to be using a setting that will *not* commit CRLF. if just you have that setting i dont think it'll help | 21:06 |
| ← whatdoesitwant left | 21:07 |
| ← MrGlass_Work left | 21:07 |
|
ljungk
| git checkout master | 21:07 |
|
| git cherry-pick 6ed128e | 21:07 |
|
| git checkout A | 21:07 |
|
| git rebase master | 21:07 |
| ← koruptid left | 21:07 |
|
charon
| ljungk: and A is a local (not remote-tracking) branch? | 21:08 |
|
ajpiano
| vinse, here's the deal... i have a brand new set of files that i want to start tracking with git. on a mac, i ran git init, and then pushed to a bare repo that i set up on windows running virutal ubuntu. now, in another place on that windows machine, i want to clone that repo, and when i do so, everything comes up marked changed | 21:08 |
| ← lhz left | 21:08 |
|
vinse
| ajpiano: so git status shows all files? | 21:09 |
| ← HG` left | 21:09 |
|
charon
| ljungk: because so far you haven't involved remotes or remote branches at all. where would you expect them to change? | 21:09 |
|
vinse
| ajpiano: we also had strange cases where gitk showed files as changed, but the command line did not | 21:09 |
|
ajpiano
| i'm on command line | 21:09 |
| → schlortoo joined | 21:09 |
|
ljungk
| charon: yes | 21:10 |
| ← heewa left | 21:11 |
|
ljungk
| and then I end up with this: http://tinypic.com/r/2uti8g0/3 | 21:11 |
|
vinse
| ajpiano: heh we had that exact issue at first, but sadly i dont remember how we eventually solved it | 21:11 |
|
ajpiano
| :/ | 21:11 |
| ← stringoO left | 21:11 |
|
charon
| ljungk: well, the purpose of remote-tracking branches (remotes/$nickname/$branchname) is to store git's current idea of the contents of remote $nickname | 21:12 |
|
ljungk
| charon: well, both A (which is really called vega-local) and master have remote branches | 21:12 |
|
charon
| ljungk: hence, they only update if you either change the remote yourself (git push $nickname $branchname) or if you update git's idea of what the remote has (git fetch $nickname) | 21:12 |
|
| so this will only change when you push to the remote | 21:12 |
| → friskd joined | 21:13 |
|
charon
| (since until then, it will keep the old version of the branch) | 21:13 |
|
ljungk
| okay, which was not something I wanted to do until I knew things worked :) | 21:13 |
|
vinse
| ajpiano: asking if anyone else remembers, or if some one took notes (unlikely!), i'll get back to you if i remember anything useful ;p | 21:14 |
|
ajpiano
| vinse, thx | 21:14 |
| ← schlortoo left | 21:14 |
| → alester joined | 21:15 |
| agib → agib|away | 21:15 |
| agib|away → agib | 21:15 |
| → ehabkost joined | 21:17 |
| ← ehabkost_ left | 21:17 |
| ← Tureba left | 21:17 |
| ← Dave^| left | 21:17 |
| → Dave^| joined | 21:17 |
| ← lucsky left | 21:17 |
| → jtdowney joined | 21:18 |
| → ProperNoun joined | 21:18 |
|
ljungk
| okay, when i try to push, i get 'error: failed to push some refs to <servername>' | 21:20 |
| → drizzd joined | 21:20 |
| ← ehabkost left | 21:20 |
|
ljungk
| and then git status yields: Your branch and 'origin/vega-local' have diverged, and have 14 and 14 different commit(s) each, respectively. | 21:21 |
| ← brntbeer left | 21:21 |
| ← airborn left | 21:21 |
|
ljungk
| so somhow the remote branch is not rebased, if i understand things correctly? | 21:21 |
|
Ilari
| ljungk: What is the branch update message when you attempt to push? | 21:22 |
| → dreiss joined | 21:24 |
| → TeckniX joined | 21:25 |
|
ljungk
| look: http://pastebin.com/kU8RReCU | 21:25 |
| → nurey joined | 21:26 |
|
charon
| ljungk: "look" seems a fairly appropriate response... | 21:27 |
| ← hipe left | 21:27 |
|
charon
| as in, read the last paragraph, and see that note about fast-forwards. | 21:27 |
| → hipe joined | 21:28 |
|
charon
| ljungk: anyway, it's expected since you used rebase so you need to override the check eventually. | 21:28 |
| hsa2|gone → hsa2 | 21:28 |
| ← Schmallon left | 21:28 |
|
ljungk
| charon: heh okay | 21:29 |
| agib → agib|away | 21:29 |
| agib|away → agib | 21:30 |
| ← Dave^| left | 21:31 |
| ← Aikawa left | 21:31 |
| → Dave^| joined | 21:31 |
| agib → agib|away | 21:31 |
| → Aikawa joined | 21:32 |
| agib|away → agib | 21:32 |
| → sh1mmer joined | 21:32 |
| ← hipe left | 21:32 |
|
ljungk
| ah , rebase works quite differently from what i thought. seems there is much yet to learn for me. | 21:32 |
| → ignatp joined | 21:35 |
| ← alex__c2022 left | 21:37 |
| ← eighty4 left | 21:38 |
| → davidc_ joined | 21:39 |
| ← dropandho left | 21:40 |
| ← sh1mmer left | 21:41 |
| ← spaceonline left | 21:42 |
| ← adamm left | 21:43 |
| → adamm joined | 21:43 |
| ← tstclair left | 21:43 |
| ← thierryp left | 21:46 |
| ← gebi left | 21:46 |
| → thierryp joined | 21:46 |
| → sh1mmer joined | 21:46 |
| ← Jferro left | 21:47 |
| ← davidc_ left | 21:47 |
| → davidc_ joined | 21:47 |
| → alexmchale joined | 21:47 |
| → Blackguard joined | 21:47 |
| → bcchocie joined | 21:48 |
| ← henkboom left | 21:48 |
| ← alexmchale left | 21:48 |
| → nevyn^_ joined | 21:49 |
| ← IslandUsurper left | 21:49 |
|
charon
| ljungk: as tv said it: you cannot change. you can only rewrite (make a new object) and forget | 21:50 |
| ← nevyn^_ left | 21:50 |
| → henkboom joined | 21:51 |
| ← thierryp left | 21:51 |
| → jtbandes joined | 21:52 |
| ← nurey left | 21:52 |
|
jtbandes
| How do submodules work with shared repos? if I add one, will people get it immediately when they pull? | 21:53 |
| ← jrmuizel left | 21:53 |
| → gebi joined | 21:53 |
| → zackattack joined | 21:56 |
|
zackattack
| what is an insertion/deletion? it's certainly not a diret measure of LOC (lines of code)? | 21:56 |
|
Ilari
| zackattack: Insertions/deletions count? | 21:57 |
| ← raichoo left | 21:57 |
| → hipe joined | 21:57 |
|
zackattack
| yeah. | 21:57 |
|
| 4 files changed, 27 insertions(+), 15 deletions(-) | 21:58 |
| ← justin-george left | 21:58 |
|
zackattack
| Does that mean Only 12 LOC were added? | 21:58 |
| → justin-george joined | 21:58 |
|
Ilari
| zackattack: For text files, additions are + lines in diff and deletions are - lines. No idea what it does with binary files. | 21:58 |
|
zackattack
| cool | 21:58 |
|
| is there a neat way to check out many LOC I modified in the last 12h, for example? | 21:59 |
|
Ilari
| zackattack: The line count change is 27-15 = 12, so target has 12 lines more than base. | 21:59 |
|
zackattack
| :) | 21:59 |
|
| I wonder if github keeps track of those stats | 21:59 |
|
Ilari
| zackattack: 'git diff --stat master@{12.hours.ago} master'? | 21:59 |
|
zackattack
| wow | 22:00 |
|
| *very cool* | 22:00 |
|
Ilari
| zackattack: Actually, that counts all changes in master in last 12 hours. | 22:00 |
|
| zackattack: Including possible merges, pulls and rebases during that time. | 22:01 |
| ← copelco left | 22:02 |
| → nurey joined | 22:02 |
|
zackattack
| Ilari: are there any cool GUI tools (web or desktop) that have these kinds of visuals? | 22:02 |
| ← grey-monkey left | 22:04 |
|
Ilari
| zackattack: No idea. | 22:04 |
|
zackattack
| :) | 22:04 |
|
bcchocie
| All, I've got a bit of a complicated question about pushing a git repo into svn. Anyone familiar with that? | 22:05 |
| ← jtbandes left | 22:06 |
|
bremner
| bcchocie: it is hard to judge whether what you think is complicated is the same as what we do, so give it a try. | 22:06 |
|
bcchocie
| The problem is that it's a kernel tree, and I'd really like the SVN version to have everything prior to a certain commit squashed | 22:07 |
|
| Yet still, I would hope to occasionally push my git-side changes to svn | 22:07 |
| ← adamm left | 22:07 |
| ← masterkorp left | 22:08 |
| ← rseifert left | 22:08 |
| ← loincloth left | 22:08 |
| → adamm joined | 22:08 |
| ← alester left | 22:08 |
|
bcchocie
| bremner, thanks... I'm just afraid that I might be asking the impossible :) | 22:08 |
| ← AAABeef left | 22:09 |
| ← p3rg2z left | 22:10 |
|
patrikf
| bcchocie: i don't know an easy way to do that. it would be easier if you'd apply the git changes as one svn commit | 22:10 |
| → gbacon joined | 22:10 |
|
patrikf
| (the real solution being of course not to have two revision control systems) | 22:10 |
| ← altrux left | 22:11 |
|
markatto
| is there a way to checkout a single file from a different branch? | 22:11 |
|
bcchocie
| patrikf: Ah, yes. That seems obvious now that you've said it. Looks like I just needed another person's perspective. Thanks. | 22:11 |
|
bremner
| markatto: man git-checkout | 22:11 |
|
starcoder|Svr
| test | 22:12 |
|
bcchocie
| patrikf: a single vcs would be my choice too, but i've got to scratch the back of those who sign the checks around here... :) | 22:13 |
| cpg|away → cpg | 22:13 |
| ← bcchocie left | 22:14 |
| → ivanoats joined | 22:15 |
| ← mkramer left | 22:15 |
| ← ajpiano left | 22:16 |
|
avar
| charon: the + patch looks good | 22:17 |
| ← Bennid left | 22:18 |
| → galderz joined | 22:19 |
| ← gebi left | 22:19 |
| ← smuf left | 22:22 |
| → tongueroo joined | 22:22 |
| → eighty4 joined | 22:23 |
| ← hipe left | 22:25 |
| → hipe joined | 22:25 |
| ← TeckniX left | 22:27 |
| → iruediger joined | 22:27 |
| ← bdimcheff left | 22:28 |
| ← hipe left | 22:29 |
| → ralf_e joined | 22:31 |
| → sneakret_ joined | 22:32 |
| cpg → cpg|away | 22:32 |
| ← sneakret left | 22:33 |
| → tanoku joined | 22:33 |
| ← nurey left | 22:34 |
| ← kenichi left | 22:34 |
| → zorzar_ joined | 22:36 |
| ← ignatp left | 22:36 |
| → Tuller joined | 22:37 |
| ← beni- left | 22:38 |
| ← zorzar left | 22:39 |
| ← Ryback_ left | 22:39 |
| → mdel_ joined | 22:41 |
| ← Titosemi left | 22:41 |
| ← slakmagik left | 22:41 |
| ← MattDiPasquale left | 22:41 |
| ← mdel left | 22:41 |
| → slakmagik joined | 22:43 |
| ← Dave^| left | 22:44 |
| ← dreiss left | 22:45 |
| ← stepnem left | 22:46 |
| → irc2samus joined | 22:47 |
|
irc2samus
| hi guys, I have a series of commits on a branch that was created from master, I want to apply only those commits in the branch to a different branch that diverged from master (so I can't just merge) what is the right way to do it? | 22:48 |
| → frooh joined | 22:48 |
| ← thiago left | 22:48 |
| ← ivanoats left | 22:49 |
| → stepnem joined | 22:50 |
|
patrikf
| irc2samus: look into git rebase | 22:50 |
| → tsbtmn joined | 22:50 |
| ← carutsu_ left | 22:51 |
| ← frooh left | 22:51 |
| → Heimidal joined | 22:52 |
| ← galderz left | 22:52 |
| ← Heimidal left | 22:54 |
| ← t0rc left | 22:54 |
| ← juanjoc left | 22:55 |
| ← Tuller left | 22:57 |
| ← bentob0x left | 22:58 |
| → hipe joined | 22:59 |
|
sitaram
| irc2samus: how many commits in master, out of which how many you want to carry across to the other branch? you should use cherry-pick, if it's just a few | 22:59 |
| ← elmob left | 23:01 |
| ← slakmagik left | 23:01 |
| cpg|away → cpg | 23:01 |
| ← AndyJ left | 23:01 |
| ← Andy_ODT_1 left | 23:01 |
|
charon
| avar: i'm afraid junio is right actually | 23:01 |
|
irc2samus
| sitaram: yeah I could but I think it's a good oportunity to learn something new, in fact I've just screwed the rebase :) but I'll figure it out | 23:01 |
|
charon
| avar: i thought the extent of 9d2e942 was right up into this + case, but it only patches url_decode into the client call-chain | 23:02 |
| ← tjholowaychuk left | 23:02 |
|
avar
| hrm, yes. | 23:03 |
| → murr4y joined | 23:04 |
| ← xiong left | 23:05 |
| ← vinse left | 23:05 |
| → slakmagik joined | 23:06 |
| → koruptid joined | 23:06 |
|
irc2samus
| ah I see I have to use --onto :) I got the rebase to do what I want but there's merge conflicts I think I'll merge into master and wait for the next deploy instead | 23:06 |
|
| thanks guys! | 23:06 |
| ← koruptid left | 23:06 |
| ← hugoxrosa left | 23:07 |
|
aresnick
| I'd like to pull in an hg repo as a git submodule into an existing git repo. Is there a better way to do this than host my own git repo clone of the hg repo? | 23:08 |
| → disappearedng joined | 23:08 |
|
disappearedng
| I have 3 mirrors of my repos in my computer. How can I do this legally? git config --add alias.backup "push --mirror /mount/1t/backup/spontivity;push --mirror /mount/950gb/backup/spontivity;push --mirror /mount/350gb/backup/spontivity;" | 23:08 |
| ← eletuchy_ left | 23:09 |
| ← hipe left | 23:09 |
|
disappearedng
| I want to push to simultaneously 3 repos | 23:10 |
| → SlickT10 joined | 23:10 |
| ← datapolitical left | 23:12 |
| → ignatp joined | 23:14 |
| ← irc2samus left | 23:15 |
| ← tsbtmn left | 23:15 |
| → ivanoats joined | 23:17 |
| ← tizbac left | 23:18 |
| ← ignatp left | 23:19 |
| ← zenazn left | 23:19 |
| → thiago joined | 23:20 |
| ← jtauber left | 23:20 |
| → Andy joined | 23:22 |
| Andy → Guest68810 | 23:22 |
|
ju2wheels
| does anyone know how dependent gitolite is on git 1.6.2? Im installing on RHEL5 but i can only find git 1.5.5.6 packages and id rather not install from source | 23:22 |
| ← iruediger left | 23:22 |
| ← Modius left | 23:22 |
| ← Jippi_mac left | 23:22 |
|
selckin
| that's ancient man | 23:22 |
|
charon
| disappearedng: git config alias.backup "!git push ...; git push ...; ..." | 23:23 |
|
ju2wheels
| i know lol | 23:23 |
|
charon
| disappearedng: i doubt --add makes sense with aliases. and the !... form runs a shell command | 23:23 |
| → datapolitical joined | 23:24 |
| → murz joined | 23:25 |
| ← Paraselene_ left | 23:25 |
| ← mdel_ left | 23:26 |
| ← didlix left | 23:26 |
|
ju2wheels
| if i do it from source the prefix is the install root right? so if i use prefix=/usr/share i should get /usr/share/git-core, /usr/share/* etc for all the other git stuff? | 23:26 |
| ← amerine left | 23:27 |
| ← aspotashev left | 23:27 |
|
sitaram
| ju2wheels: 1.5.x won't work well at all. Just try and install from source; it's not hard | 23:28 |
| ← eighty4 left | 23:28 |
|
ju2wheels
| k | 23:29 |
| ← d0k left | 23:29 |
|
sitaram
| ju2wheels: prefix=/usr/local is best; /usr/share won't do since /usr/share/bin is likely not in the path | 23:29 |
| → ProperN[out] joined | 23:29 |
| ← ProperN[out] left | 23:29 |
| → ProperN[out] joined | 23:30 |
|
ju2wheels
| sitaram: ok thx, i did see the general install command was /usr/local but i checked on my system and found git-core in /usr/share which is why i was asking. thx | 23:31 |
| → jtauber joined | 23:31 |
| ← zackattack left | 23:31 |
| ← sh1mmer left | 23:33 |
| ← ProperNoun left | 23:33 |
| ProperN[out] → ProperNoun | 23:33 |
| ← justin-george left | 23:34 |
| → zackattack joined | 23:37 |
| → ignatp joined | 23:38 |
| ← drizzd left | 23:38 |
| ← SlickT10 left | 23:42 |
| → ajpiano joined | 23:47 |
| ← ShooterM_ left | 23:48 |
| ← ralf_e left | 23:48 |
| ← priidu left | 23:50 |
| ← ju2wheels left | 23:50 |
| → ju2wheels joined | 23:51 |
| → rorist_ joined | 23:52 |
| → leonid joined | 23:52 |
| → JohnnyL joined | 23:53 |
|
JohnnyL
| i can download the kernel fine, but when I try to grab a read/write proejct allowed for my account i get:i can download the kernel fine. What do | 23:54 |
|
| i mean: fatal: Unable to look up [email@hidden.address] (Unknown error) | 23:55 |
|
| butter fingers tonight. | 23:55 |
|
| am I missing a key or something? | 23:55 |
|
frsk
| Use [email@hidden.address] | 23:55 |
| ← rorist left | 23:55 |
|
frsk
| And you can try "ssh [email@hidden.address] to make sure the key works :-) | 23:56 |
|
JohnnyL
| frsk: i get : http://nopaste.dk/p2213 | 23:57 |
|
| ok i'll try | 23:57 |
|
frsk
| ah | 23:57 |
|
| JohnnyL: Try [email@hidden.address] | 23:57 |
|
| You're supposed to use ssh, not the git protocol. | 23:57 |
| ← zackattack left | 23:58 |
|
frsk
| (iow, remove "git://") | 23:58 |
|
JohnnyL
| permission denied (publickey) | 23:58 |
|
frsk
| okey, and that goes for "ssh [email@hidden.address] as well? | 23:59 |
| ← tanoku left | 23:59 |
| ← molokai left | 23:59 |
| ← RobertLaptop left | 23:59 |
