IRCloggy #git 2012-08-12

Logs Search ←Prev date Next date→ Channels Documentation

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

2012-08-12

← lupo7 left00:01
→ r0bby joined00:01
← red82 left00:01
← robbyoconnor left00:03
← TomyLobo left00:03
← Guest97802 left00:04
← nobdraisentone left00:05
→ nobdraisentone joined00:06
← nobdraisentone left00:07
→ nobdraisentone joined00:08
→ simesy joined00:08
← nobdraisentone left00:09
→ nobdraisentone joined00:09
← HyperNerdV2 left00:10
→ rmr joined00:10
→ HyperNerdV2 joined00:10
← skyf left00:12
← rmr left00:13
← Raydiation left00:15
→ zastern joined00:16
zastern I'm missing something about how git works - I've got a local "master" branch, and a "master" branch on github. A file is different on my master branch on github, but when I do a git pull origin master, the local one isn't being overwritten. It's saying everything is up to date.00:17
→ JStoker joined00:19
avpx zastern: What does "git status" say?00:19
→ zomg joined00:19
zomg Hello, I wonder if there's a specification of how the git http-backend thingy works anywhere?00:19
zastern avpx: I actually figured out what to do - git reset --hard origin/master00:19
avpx wonders why you would use git over http00:19
avpx zastern: Okay, good. I was going to suggest a git reset --hard :p00:20
zomg Trying to implement a quick repo fetch mechanism in an environment with no git client, nor proper support for anything more complicated than http/https00:20
→ towp joined00:20
zastern avpx: hard wouldn't do it - there were no changes locally00:21
avpx: what I had done, was, while on "masteR", accidentally done "git pull origin testing"00:21
zomg Trying to decipher http-backend.c and not having much luck... testing against github, i'd assume their implementation is compatible00:21
Most I got out so far was just reverse engineering what the actual git clone command does by netcatting a local port and trying to clone that :P00:22
→ cakehero joined00:22
zomg I suppose I should just have something read the network traffic of a full clone-over-http...00:22
← towp left00:23
zomg As far as I can tell, doing github.com/foo/bar.git/objects/xx/xxxxxxxxxx... should allow me to fetch object data but just getting 404...00:24
← Element9 left00:25
→ iamjarvo joined00:25
← dfr|mac left00:27
→ ViniciusPeretti joined00:28
→ lirt joined00:29
zomg Meh, not entirely unexpected for it to lack any and all documentation...00:30
← lirt left00:30
← Ether_Man left00:31
← CpuID3 left00:31
← zii left00:32
→ CpuID3 joined00:32
→ pppaul joined00:32
→ Ether_Man joined00:32
jnewt how do i get a git rebase to quit giving me the "both added" error, and just modify the file with the new branch file? i try checkout branch , then rebase with strategy ours, but the conflict still happens.00:33
← rue left00:34
→ zii joined00:34
→ adrient joined00:34
← kart0ffelsack left00:34
→ luke-jr joined00:35
← mikemol left00:35
← DeviceZer0 left00:35
← jspiros left00:35
← grillo left00:35
← lamont left00:35
← yardenac left00:35
→ kenyon joined00:36
→ FauxFaux joined00:36
→ jspiros joined00:36
→ MacGyver joined00:36
← MacGyver left00:36
→ MacGyver joined00:36
→ Zeeshan_M joined00:36
→ mbreslin joined00:36
→ Targen joined00:37
→ grillo joined00:38
→ lamont joined00:38
→ Kindari joined00:38
→ KBme joined00:38
→ tm1000 joined00:38
→ DeviceZer0 joined00:39
← Ether_Man left00:40
→ cschneid joined00:40
→ squentin joined00:40
→ kart0ffelsack joined00:41
→ Mez joined00:41
← Mez left00:41
→ Mez joined00:41
→ mikemol joined00:42
→ etcetera joined00:42
→ Ether_Man joined00:43
← ViniciusPeretti left00:43
→ yardenac joined00:43
← SamB_MacG5 left00:43
← cybersphinx left00:46
← hyperair left00:48
→ cannonball joined00:49
← thiago left00:49
→ codeshah joined00:49
→ SamB_MacG5 joined00:50
← Bombe left00:51
← evanz left00:51
← cannonball left00:54
← maletor left00:54
← iamjarvo left00:55
← theCole left00:56
→ skyf joined00:57
→ angelsl joined00:58
← etcetera left00:59
← chuckharmston left00:59
→ Bombe joined01:00
← edude03 left01:00
→ macabre joined01:01
→ velikasha1 joined01:06
← velikasha1 left01:06
→ bubbles10101 joined01:08
→ esa joined01:11
← jargon- left01:14
← pppaul left01:14
→ Olipro joined01:14
← kukks left01:14
← kcj left01:15
← eicto left01:16
← diegoviola left01:19
→ etcetera joined01:19
← Lgb left01:20
← bizhanMona left01:20
← palm_vue left01:20
← macabre left01:20
→ palm_vue joined01:20
→ Lgb joined01:20
→ bizhanMona joined01:21
→ ded joined01:21
→ dfr|mac joined01:21
← cakehero left01:22
ded I'd like to do the equivalent of setting receive.denyCurrentBranch to ignore/warn *and* have git update the working tree to what I push. Any git-native way to do that or would I need to do it with e.g. a hook?01:23
(Please don't bother replying if all you have to say is why would you want to do that.)01:23
→ ysyk joined01:23
← guns left01:23
milki o.O01:24
→ dc5ala joined01:24
milki that disclaimer is distracting -.-01:24
CareBear\ ded : way to go to make people really f-ing uninterested in your problems01:24
adrient → finch01:25
milki jnewt: no clue01:25
→ intripoon_ joined01:26
→ Goplat joined01:26
milki ded: why doesnt warn/ignore work?01:26
CareBear\ jnewt : do you also give -m ?01:26
jnewt : sorry, just read, -s implies -m01:26
→ rue joined01:26
ded milki: I can double check, but the error message that git issues is pretty explicit about the fact that only the working tree will not be updated.01:27
← marhaban left01:27
milki Freeaqingme: if the branch is deleted in the remote repo, you cant retrieve it01:27
ah01:27
ded milki, as in this error message. http://stackoverflow.com/questions/804545/what-is-this-git-warning-message-when-pushing-changes-to-a-remote-repository01:27
Freeaqingme milki, I dont need the branch itself, only the name of the deleted branches01:27
milki ded: in that case, a post-receive hook01:27
Freeaqingme: same thing01:27
Freeaqingme milki, kk,, tnx01:28
→ diegoviola joined01:28
ded milki, ok thanks.01:28
milki Freeaqingme: it would be a security hole if you could01:28
→ eicto joined01:28
Freeaqingme milki, why? Even if a branch was deleted, its commits still may be somewhere?01:29
← tshauck left01:29
milki right01:29
ded CareBear, yeah I guess. The commonness of that response is one of the saddest thing about technical mailing lists.01:29
CareBear\ but you need to know the commit then01:29
ded / irc01:29
Freeaqingme CareBear\, there's a command to find dangling commits01:29
CareBear\ ded : do you know why people give you that response?01:29
← intripoon left01:29
milki ded: well, we would probably suggest a smarter way to do things if you tell us what you are trying to accomplish01:30
ded CareBear, yes01:30
CareBear\ ded : but you are not interested?01:30
SeoZ-work[AWAY] → SeoZ01:30
ded milki: most frequently that response comes from know-it-alls who have little else to contribute; not always, of course.01:30
s/milki/CareBear!01:31
CareBear\ dad : is this your first visit to #git ?01:31
s/dad/ded01:31
milki we are all know it alls, and we have tons to contribute01:31
jnewt i would normally just reset the head if i just wanted to overwrite, but the idea is to maintain history which i didn't do correctly and now i can't fix it.01:31
← rue left01:31
→ cakehero joined01:32
ded milki, sure, what I'm trying to accomplish is keeping two working trees in sync. It's purely a development context, and I am the only person who is using these repos.01:32
← FernandoBasso left01:32
milki jnewt: perhaps you need to clean up the history of one of the branches before you try rebasing01:32
CareBear\ ded : why are there two development trees?01:33
milki ded: is there a particular need to not use bare repos?01:33
ded one is on my laptop and one is on a test server01:33
→ scotepi joined01:33
ded I need the code to be checked out in both places because it is being executed in both places01:33
CareBear\ ded : do you know http://joemaller.com/990/ ?01:33
scotepi anyone have any advice to get git to run faster from windows over a samba share?01:33
milki that doesnt mean you cant have bare repos01:34
ded: check out !deploy01:34
gitinfo ded: Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html01:34
CareBear\ scotepi : don't use windows and don't use samba shares :p01:34
scotepi thanks ><01:34
milki >.>01:34
← cakehero left01:34
CareBear\ scotepi : windows shouldn't be the problem, but I would do some comparison with and without the shared folder, to see performance differences01:35
← AaronMT left01:35
→ Vampire0_ joined01:36
scotepi CareBear\, it is definitly the samba share.. I have my web dev server shared and I work on it as a maped drive on windows. However im trying to transition to git and its painfuly slow to do status and the like01:36
i took 3mins to do it on windows on 2sec on the server01:36
milki have a local copy that you work on and then push to samba share when you need to01:36
CareBear\ scotepi : stop doing it that way01:37
ded milki, CareBear thanks. I'm going to try a post-receive hook using git reset --hard HEAD.01:37
CareBear\ scotepi : you too, have a look at http://joemaller.com/990/01:37
milki kk01:37
CareBear\ ded : use a third repo like joemaller01:37
ded : save yourself the friction01:37
scotepi CareBear\, that would meen i need to install a web server matching my webhost on every workstation, not an ideal solution for me01:37
← darren left01:38
CareBear\ scotepi : huh? no?01:38
ded CareBear, ok, this is not a production deploy, just looking for something lightweight to help iterate on debugging01:38
CareBear\ ded : yes, joemaller01:38
ded ok01:38
CareBear\ ded : production deploy is better left for another tool01:38
milki ded: that doesnt really matter01:38
→ havenn joined01:38
← etcetera left01:38
CareBear\ (that other tool may also be git based, but git itself isn't such a great fit)01:39
milki if you need your code executing on the server, just have a post-receive that "deploys" the code to where you need it01:39
thats all01:39
→ darren joined01:39
← Vampire0 left01:39
CareBear\ as a bonus of joemaller you also get a backup of your repo01:39
for when your laptop is stolen01:39
milki by CareBear\01:40
:o01:40
→ tqmz___ joined01:40
CareBear\ umm..01:40
ded ok right, I see the "I tried it and things do get weird fast" line. I'm very happy to take that advice.01:40
CareBear\ milki : you promised you wouldn't tell anyone where you bought that cheap laptop I set you up with01:40
milki i didnt01:40
:o01:41
CareBear\ ded : please do ask also if you have concrete issues - they are usually even easier to answer01:41
ded : and !welcome01:41
gitinfo Welcome to #git, a place full of helpful gits. If you have a question, feel free to just go ahead and ask—somebody should answer shortly. For more info on this channel, see http://jk.gs/git/ Take backups (type !backup to learn how) before taking advice.01:41
ded yep, sorry about my opening. Not sure what I was thinking of.01:42
← tqmz left01:43
CareBear\ thanks for that01:43
→ ISF joined01:44
→ Lord_DeathMatch joined01:48
→ ipoval joined01:49
→ xiangfu joined01:49
→ maxwellcorbin joined01:50
← Dimitr left01:51
→ red82 joined01:53
← ded left01:54
→ Cerise joined01:55
→ cakehero joined01:58
→ lcc joined01:59
→ reds joined01:59
← havenn left01:59
reds → Guest361201:59
← matthiaskrgr_ left02:03
← Guest3612 left02:03
→ kukks joined02:04
← kukks left02:04
→ kukks joined02:04
← fission6 left02:06
→ jargon- joined02:06
← Twey left02:06
→ Twey joined02:06
← scientes left02:08
→ cagedwisdom joined02:09
→ landypro joined02:11
→ iamjarvo joined02:11
← xiangfu left02:13
← dr_lepper left02:15
→ jiangenj joined02:15
gitinfo set mode: +v02:16
jiangenj Hi cbreak, yesterday git project object broken has been fixed; just clone a mirror then put it back can fix, not sure why02:16
← leeN left02:17
← cakehero left02:18
jiangenj Hi all, where can I find a git book for admin? git help is seperated by commands which I think can only be used if knows what to do02:18
frogonwheels jiangenj: !doc02:19
gitinfo jiangenj: A list of useful documentation of all kinds is here: http://git-scm.com/documentation -- or try the wiki at http://git.wiki.kernel.org/. Also try typing "!book" "!cs" "!bottomup" "!parable" "!best_practices" or "!vcbe" or "!designers" here in IRC. !book is probably the most helpful.02:19
frogonwheels jiangenj: yes, the git man pages become more useful once you know the broader outline of how git operates.02:19
jiangenj thanks02:20
← dorkmafia left02:20
→ matthiaskrgr joined02:20
jiangenj !book02:21
gitinfo There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable02:21
→ beneggett joined02:22
→ evanz joined02:23
→ reactormonk joined02:23
→ e0je joined02:24
→ r0bby_ joined02:25
← jargon- left02:25
← bubbles10101 left02:25
r0bby_ → robbyoconnor02:27
← r0bby left02:28
→ dr_lepper joined02:28
← dr_lepper left02:28
→ dr_lepper joined02:28
→ Targen_ joined02:29
← Targen left02:29
← e0je left02:29
← Araxia left02:30
← maxwellcorbin left02:30
← segher left02:31
→ bubbles10101 joined02:32
→ segher joined02:32
→ maxwellcorbin joined02:32
→ vitor-br joined02:34
→ albel727 joined02:34
← Lord_DeathMatch left02:35
→ Chib joined02:36
← spyro left02:36
→ jargon- joined02:38
→ r0bby joined02:39
→ xiangfu joined02:40
← robbyoconnor left02:41
jnewt ok, i've started with a clean repo, and have added 3 commits. r1, r3, r2 in that order. now let's say i want to put them in the order r1,2,3. i'd think that i would want to rebase r2 onto r1, then r3 onto r2.02:42
→ r0bby_ joined02:43
jnewt git rebase --no-ff --strategy ours --onto r1 r3 r2. this just moves the branch pointer r2 to r1, and the r2 commit goes away02:43
← red82 left02:43
← xiangfu left02:44
→ Eridius joined02:45
jnewt i get the "already applied" message. then All done. as if i had actually rebased that commit onto r1.02:45
← orafu left02:45
→ macIvy joined02:45
← shruggar left02:45
← beneggett left02:45
→ orafu joined02:45
→ soooga joined02:45
→ aviraldg joined02:46
← r0bby left02:46
→ spyro joined02:46
← spyro left02:46
→ spyro joined02:46
→ xiangfu joined02:47
← ysyk left02:48
jnewt and now i'm getting a couple of "deleted by us" errors, for two files that have never been deleted in any commit02:49
← Twey left02:52
→ Lord_DeathMatch joined02:54
→ wolftankk joined02:54
← iamjarvo left02:55
← Chib left02:55
→ Chib joined02:56
r0bby_ → robbyoconnor02:56
jiangenj !cs02:57
gitinfo "Git for Computer Scientists" is a quick introduction to git internals for people who are not scared by phrases like Directed Acyclic Graph. http://eagain.net/articles/git-for-computer-scientists/ See also !concepts !bottomup02:57
← wolftankk left02:59
→ wolftankk joined03:00
→ Twey joined03:00
← ipoval left03:02
→ caseymcg joined03:02
← blupp left03:04
→ cwang joined03:04
→ blupp joined03:05
← macIvy left03:07
← Chillance left03:07
→ Anxi80 joined03:09
→ ysyk joined03:10
← ysyk left03:10
← jey_ left03:11
→ jey joined03:12
→ centipedefarmer joined03:12
→ duckxx joined03:12
→ Yuuhi` joined03:13
→ ysyk joined03:14
← Drool left03:14
← ysyk left03:14
← localhost left03:14
← Yuuhi left03:15
→ maletor joined03:16
→ localhost joined03:17
← wolftankk left03:17
avpx jweinberg: The best way to reorder commits this way is using an interactive rebase.03:18
rebase -i03:18
jnewt**03:18
← trtzrdb left03:20
→ ysyk joined03:22
← ysyk left03:23
← metasansana left03:23
→ duckx joined03:23
← yoklov left03:25
← duckxx left03:26
← dfr|mac left03:28
← Lord_DeathMatch left03:29
→ metasansana joined03:30
← metasansana left03:31
→ etcetera joined03:31
→ scriabin joined03:33
← bubbles10101 left03:34
← duckx left03:35
← jiangenj left03:39
→ kcj joined03:39
← Eridius left03:40
← lcc left03:41
← caseymcg left03:42
→ jackh joined03:42
← jackh left03:42
← etcetera left03:43
→ mh` joined03:46
→ rohan32 joined03:47
→ scholchr joined03:48
→ gavin_huang joined03:50
← rohan32 left03:50
← nwest left03:51
→ yoklov joined03:52
→ lcc joined03:54
→ andy77 joined03:58
← centipedefarmer left03:59
→ reds joined03:59
→ iamjarvo joined04:00
reds → Guest8374304:00
→ dleonardi joined04:00
→ beneggett joined04:00
→ centipedefarmer joined04:01
← tjadc left04:02
→ nothingbutlove joined04:03
← Guest83743 left04:04
← kukks left04:05
← Anxi80 left04:06
← bartman left04:07
← gavin_huang left04:10
← beneggett left04:11
→ bluekelp joined04:11
← mbreslin left04:13
← intransit[w] left04:13
→ chuckf joined04:14
→ bartman joined04:14
← bluekelp left04:14
→ adiabatic joined04:15
← maletor left04:17
→ dfr|mac joined04:18
→ unnamedwill joined04:18
← nothingbutlove left04:18
→ duckxx joined04:19
→ john_anderson_ joined04:20
→ nothingbutlove joined04:21
→ scientes joined04:23
→ Putr joined04:23
← dleonardi left04:28
← goshawk left04:33
← scientes left04:35
← Twey left04:39
→ Guest74974 joined04:39
→ gavin_huang joined04:39
→ joepie91 joined04:39
← joepie91 left04:39
→ Davey joined04:41
→ igotnolegs joined04:41
← gavin_huang left04:42
← andrewhl_ left04:44
→ Twey joined04:44
→ mbreslin joined04:45
→ scientes joined04:47
← xiangfu left04:49
← nor5trom left04:50
← zastern left04:53
→ Transcended joined04:55
→ Fullmoon joined04:57
→ ipoval joined04:58
→ rue joined05:03
← cesc left05:06
→ r0bby joined05:07
← rue left05:07
→ helgikrs joined05:08
luke-jr Is there any "multi bisect" where you can bisect in parallel?05:09
ie, if I have 7 machines for testing, give me 7 revisions to test at a time05:09
(for bugs that take days to reproduce)05:10
← robbyoconnor left05:10
← scotepi left05:11
← kpreid left05:17
→ kpreid joined05:18
→ cesc joined05:18
→ smw joined05:19
← yoklov left05:20
→ towski joined05:21
← simesy left05:23
→ Araxia joined05:23
← neurodrone left05:24
→ xiangfu joined05:24
→ rue joined05:25
→ noyb joined05:25
← noyb_ left05:26
← towski left05:26
← thelazydeveloper left05:28
→ etcetera joined05:29
← xiangfu left05:29
→ opus joined05:30
gitinfo set mode: +v05:30
← etcetera left05:31
offby1 never heard of such a thing05:31
→ ovnicraft joined05:33
ovnicraft hello i want to create a new branch and import from existing local ? how is this possible with git branch ?05:33
← Ether_Man left05:36
← marc_kdab_ left05:36
→ marc_kdab_ joined05:37
FauxFaux ovnicraft: git checkout -b newbranchname # is probably what you want.05:37
← maxwellcorbin left05:38
ovnicraft FauxFaux: yes in know but i get cloned in another local branch and i want to "import" in my new branch05:38
← scientes left05:39
ovnicraft when i say another local i means another repo05:39
FauxFaux Okay, I clearly have no idea what you mean by import. You have a server repo, A, and a local clone of it in B, currently on branch "foo", and you want to...05:40
→ xiangfu joined05:41
ovnicraft i will explain it better, i clone 2 branches from bzr to git but are in diff repos now i want to put them in just one i create05:43
in my local repo A i want to create a new branch and clone my repo B there05:43
loops ovnicraft, you can clone a branch from repo B into a branch in repo B, but not clone an entire repo to a single local branch05:45
clone a branch from repo B into a branch in repo A*05:46
→ etcetera joined05:46
FauxFaux ovnicraft: git fetch ../otherrepo master && git checkout -b otherrepobranch FETCH_HEAD05:47
ovnicraft yes i want to clone a branch from repo B in to branch in repo A05:47
← ipoval left05:49
← Goplat left05:49
← etcetera left05:50
→ uzyn joined05:50
loops git fetch remote_repo_url remote_branch:new_local_branch05:52
→ scientes joined05:52
→ etcetera joined05:53
→ private joined05:53
private → Guest6805:54
← ahayweh left05:55
← john_anderson_ left05:57
→ LeMike joined05:58
← iamjarvo left05:59
→ reds joined06:00
reds → Guest2478606:00
→ r0bby_ joined06:03
r0bby_ → robbyoconnor06:03
← dfr|mac left06:04
← duckxx left06:05
← Guest24786 left06:05
← scientes left06:06
ovnicraft loops: applies too git fetch local_path local_branch:new_local_branch ?06:07
← r0bby left06:07
← gtrak left06:07
← aviraldg left06:12
→ aviraldg joined06:14
→ ahayweh joined06:15
← Error404NotFound left06:18
← darren left06:22
← david1010 left06:23
→ david1010 joined06:23
→ scientes joined06:26
→ AmirBehzad joined06:27
← fireh left06:28
← jargon- left06:31
← scientes left06:31
→ shreya joined06:31
→ scientes joined06:33
→ jtri joined06:36
jtri hello, I'm getting a strange message on my server when i try to use git commit: *** Please tell me who you are.06:37
erynofwales Probably have to set name and email06:37
jtri yes, but to what?06:38
does it matter?06:38
erynofwales git config --global user.name "<your name>"06:38
No06:38
Whatever you want06:38
git config --global user.email "<your email>"06:38
→ jargon- joined06:38
erynofwales Though I think it's generally best to use your real name or a username and your real email06:38
jtri so global sets it global for the machine i'm working on...06:40
thanks06:40
erynofwales Almost. Global sets the option for the user06:40
jtri check06:41
erynofwales Git puts those settings in .gitconfig06:41
jtri i see06:41
erynofwales in your home directory06:41
← Targen_ left06:42
→ Targen joined06:45
← andy77 left06:46
← AmirBehzad left06:47
← avpx left06:47
→ toabctl joined06:48
← scientes left06:49
→ pppaul joined06:49
→ iori joined06:52
→ Lord_DeathMatch joined06:54
← aviraldg left06:55
← toabctl left06:56
→ aviraldg joined07:00
→ pantsman joined07:00
→ scientes joined07:01
← ISF left07:05
→ varg joined07:06
← Guest68 left07:09
→ subhojit777 joined07:13
← ovnicraft left07:18
← bizhanMona left07:20
← palm_vue left07:20
← Lgb left07:20
← iori left07:20
→ Lgb joined07:21
→ palm_vue joined07:21
→ bizhanMona joined07:21
← jargon- left07:21
← scientes left07:22
→ sh0ne joined07:27
→ simesy joined07:28
→ _iron joined07:30
→ theCole joined07:31
← nothingbutlove left07:32
← madewokherd left07:33
← JStoker left07:34
← varg left07:37
← pppaul left07:37
→ toabctl joined07:37
← sh0ne left07:38
→ sh0ne joined07:38
→ ph^ joined07:38
← Twey left07:39
→ JStoker joined07:41
→ blacktulip joined07:41
→ dleonardi joined07:42
→ airborn joined07:43
→ Twey joined07:43
← toabctl left07:46
← Davey left07:50
← cwang left07:51
← skyf left07:51
→ hyperair joined07:51
→ thiago joined07:54
SeoZ → SeoZ-work[AWAY]07:54
→ Davey joined07:56
← dleonardi left07:59
→ dleonardi joined07:59
→ reds joined08:00
reds → Guest4038708:01
← lonewulf` left08:01
← HyperNerdV2 left08:03
← LeMike left08:03
← ph^ left08:03
→ lonewulf` joined08:03
← dleonardi left08:04
← Guest40387 left08:05
← simesy left08:06
→ rehat joined08:10
← scholchr left08:11
→ dleonardi joined08:11
→ dfr|mac joined08:12
→ Eridius joined08:12
→ scholchr joined08:13
← dleonardi left08:15
→ dleonardi joined08:16
← xiangfu left08:16
→ w41_ joined08:17
← rehat left08:17
← w41 left08:20
→ Error404NotFound joined08:20
← dleonardi left08:20
→ simesy joined08:21
→ dleonardi joined08:21
← igotnolegs left08:22
→ gavin_huang joined08:24
→ n1x joined08:25
→ jjido joined08:25
← gavin_huang left08:27
← dleonardi left08:34
← subhojit777 left08:35
← theCole left08:35
→ dleonardi joined08:35
→ skorgon joined08:39
← pantsman left08:40
← Araxia left08:42
← adiabatic left08:42
← dleonardi left08:45
← eletuchy left08:46
→ Element9 joined08:48
→ dleonardi joined08:50
← kaen left08:50
→ nischayn22 joined08:51
← PLU420 left08:53
← dleonardi left08:55
← Lord_DeathMatch left08:55
← angelsl left08:55
→ angelsl joined08:55
→ leeN joined08:56
→ dleonardi joined08:56
← aviraldg left08:56
→ xiangfu joined08:57
← Guest74974 left08:59
→ DrHouse|TARDIS_ joined09:00
← jjido left09:01
→ aviraldg joined09:01
← angelsl left09:01
→ angelsl joined09:01
← xiangfu left09:02
← DrHouse|TARDIS_ left09:04
← dleonardi left09:05
→ xiangfu joined09:06
← Twey left09:06
→ Twey joined09:10
→ swair joined09:11
← blacktulip left09:12
→ gavin_huang joined09:16
→ timthelion joined09:17
← Eridius left09:18
← DrHouse|TARDIS left09:19
← tqmz___ left09:19
→ tqmz joined09:20
→ AlecTaylor joined09:22
AlecTaylor hi09:22
How do I recover a file that was deleted in the latest push?09:23
cbreak check it out.09:23
← subbyyy left09:24
→ eletuchy joined09:26
→ mmc1 joined09:26
← jtri left09:30
→ OOPMan joined09:34
AlecTaylor How do I recover a file that was deleted in the latest push?09:34
cbreak check it out.09:34
you're kind of slow, aren't you? :)09:35
FauxFaux The funk-soul cbreak, right about now.09:36
→ subhojit777 joined09:37
cbreak hmm? :)09:37
→ jceb_ joined09:38
AlecTaylor How do I recover a file that was deleted in the latest push?09:38
FauxFaux cbreak: Classic dance music reference; nevermind. =p09:38
AlecTaylor: Jesus fucking christ.09:38
cbreak AlecTaylor: should I kick you out of the channel or do you want to keep spamming?09:39
AlecTaylor "check it out" isn't giving me any information. How do I find the revision hash? - What's the command I run to checkout that specific revision?09:39
cbreak see man pages of git log and git checkout09:39
you can pick any version that had that file09:40
and then use the normal file mode checkout: git checkout ref -- file09:40
AlecTaylor How do I revert the last commit?09:42
FauxFaux man git revert09:42
gitinfo the git-revert manpage is available at http://git-scm.com/docs/git-revert.html09:42
← Targen left09:44
← gavin_huang left09:45
→ rchavik joined09:47
← arcane left09:49
→ Targen joined09:50
← OOPMan left09:52
→ d3vic3 joined09:54
→ j416 joined09:54
The_Ball For http read-only access, can I just stick a git repo in a http accessible path and use 'git clone http://foo/bar.git/ ?09:57
Or is the git-http-backend required in all http cases?09:58
→ gavin_huang joined09:58
← n1x left09:59
→ aspotashev joined09:59
FauxFaux The_Ball: "dumb http" like that is supported, but you need to google dumb http and man git update-server-info.10:00
gitinfo The_Ball: the git-update-server-info manpage is available at http://git-scm.com/docs/git-update-server-info.html10:00
The_Ball FauxFaux, thanks!10:00
→ reds joined10:01
← jonn left10:01
reds → Guest3983710:01
FauxFaux The_Ball: man git config receive.updateserverinfo10:01
gitinfo The_Ball: the git-config manpage is available at http://git-scm.com/docs/git-config.html10:01
→ marhaban joined10:02
→ RaceCondition joined10:03
→ Lord_DeathMatch joined10:04
→ nor5trom joined10:05
← Guest39837 left10:06
← mklappstuhl left10:06
→ babilen joined10:06
→ Davey_ joined10:11
← Davey left10:11
Davey_ → Davey10:11
→ dleonardi joined10:12
→ tjadc joined10:12
← Davey left10:14
← aspotashev left10:15
→ Davey joined10:16
→ longbeach joined10:16
← dleonardi left10:18
← dfr|mac left10:18
← gavin_huang left10:19
→ mklappstuhl joined10:19
→ varg joined10:20
← marhaban left10:22
→ bubbles10101 joined10:24
← bubbles10101 left10:24
→ bubbles10101 joined10:25
← swair left10:26
→ swair joined10:27
→ fornext joined10:32
→ tweedledumber joined10:32
← tweedledumber left10:37
→ tweedledumber joined10:38
← opus left10:38
← Lord_DeathMatch left10:38
FauxFaux http://faux.uwcs.co.uk/sexy-gitk-beat.png apparently gitk really likes underage Japanese girls. (Yes, I know how prefetching works on Windows.)10:39
→ drougge joined10:40
← simesy left10:40
→ adlwalrus joined10:40
adlwalrus How can I see all the branches that exist in my repo?10:41
← varg left10:41
→ beawesomeinstead joined10:41
FauxFaux adlwalrus: git branch -avv10:41
← Twey left10:45
← capiira left10:45
← airborn left10:46
→ capiira joined10:48
← rchavik left10:48
→ Twey joined10:48
adlwalrus FauxFaux: thanks :)10:51
→ Enzo_cc joined10:53
→ specialGuest joined10:53
→ arcane joined10:53
← specialGuest left10:54
→ specialGuest joined10:54
→ naee joined10:54
← d3vic3 left10:55
← eean left10:55
→ ReekenX joined10:55
← khaije|minder left10:56
← Targen left10:57
→ bulletmark joined10:57
← surma left10:58
← MetaCosm left10:58
← actel left10:58
← brownies left10:58
← nischayn22 left10:58
← beawesomeinstead left10:58
← tfnico left10:58
← ladder4 left10:58
→ khaije|minder joined10:58
← bubbles10101 left10:58
→ actel joined10:58
← scholchr left10:59
← swair left10:59
← eicto left10:59
→ varg joined10:59
→ Targen joined11:00
→ beawesomeinstead joined11:02
→ tubbo joined11:02
→ CEnnis91_ joined11:03
→ marhaban joined11:03
→ jtri joined11:04
← bulletmark left11:04
← fornext left11:05
→ xerxas joined11:05
jtri I've noticed that sometimes when i change a file "git status" shows the changes as not staged for commit, while sometimes they are... what decides?11:05
canton7 jtri, when you change a file, those changes are never staged automatically. use of 'git add' (or commit -a) does the staging11:07
→ simesy joined11:08
→ tickbrown joined11:09
→ Droolio joined11:09
→ tfnico joined11:09
→ offbyone joined11:09
→ MetaCosm joined11:09
jtri canton7: if I do a commit -a, those files wouldn't show up with "git status" though, right?11:10
mh` they'll show up as tracked and modified, or untracked11:10
jtri i.e. once committed they don't show up with a check on status... b/c they are committed11:10
mh` they won't show up on status if they haven't been modified since the last commit11:11
jtri ah, so they move from untracked to tracked and modified if they were previously committed and re-modified without a push?11:11
mh` push isn't what matters, it's the commit11:12
→ surma joined11:12
mh` if your workflow was like:11:12
← marhaban left11:12
mh` new file && edit that file && commit && git status would be clean here && edit that file && git status shows modified && commit && git status is clean again11:12
it was untracked up until the point you commit -a11:13
(which just combined git add <file> and git commit <file> into one step)11:13
← Enzo_cc left11:14
→ blacktulip joined11:16
→ marhaban joined11:17
← aviraldg left11:17
← simesy left11:17
→ lectrick joined11:19
→ oxidative_ joined11:20
→ theCole joined11:21
→ jaequery joined11:26
→ TomyLobo joined11:27
← oxidative_ left11:27
← varg left11:28
→ grandy joined11:28
→ wbs joined11:30
→ aviraldg joined11:31
→ cybersphinx joined11:31
wbs what's the most straightforward way of checking whether a certain branch (or HEAD) contains a particular commit? like git branch --contains <hash>, but also for checking HEAD... I can do e.g. "git branch --contains hash | grep ^\*" to see if it matches "* branchname" (or "* (no branch)" if not on a branch), but it feels a bit ugly11:32
→ kleist joined11:32
→ osse joined11:33
← SiDz left11:34
→ SiDz joined11:35
→ macIvy joined11:36
← zomg left11:36
→ carloshh joined11:37
→ nischayn22 joined11:37
wereHamster wbs: test -z "$(git rev-list <hash>..<branch>)"11:37
→ jab416171|Cloud joined11:38
wbs wereHamster: thanks!11:39
→ skorgon_ joined11:40
← osse left11:41
→ osse joined11:41
→ Psycho_pr joined11:43
← Psycho_pr1 left11:43
← skorgon left11:43
← Davey left11:43
loops wbs, probably want to add -n 1 to that rev-list11:45
← FauxFaux left11:46
← Droolio left11:47
← ajpiano left11:47
→ ajpiano joined11:48
→ dotemacs joined11:48
→ arietis joined11:49
→ Neptu_ joined11:49
wbs loops: yes, that also sounds like a good idea. thanks!11:49
← Neptu_ left11:50
→ ehsan joined11:51
→ davidbanham joined11:52
← eletuchy left11:53
← carloshh left11:53
→ FauxFaux joined11:54
← marhaban left11:54
→ carloshh joined11:54
← ehsan left11:55
→ jackarain joined11:59
← aviraldg left12:00
→ reds joined12:01
reds → Guest735012:01
→ heidar joined12:02
→ the_cybersphinx joined12:04
→ FernandoBasso joined12:04
→ scholchr joined12:06
← carloshh left12:06
← Guest7350 left12:06
← cybersphinx left12:08
→ Droolio joined12:10
← jackarain left12:10
→ aviraldg joined12:12
→ Jackarain joined12:12
→ DarkSector joined12:13
→ lupo7 joined12:14
FauxFaux Oooh, I like the "last change" header on http://git-scm.com/docs/git-push.html12:16
→ NimeshNeema joined12:16
canton7 ooh, that is nice12:18
→ ZyX-I joined12:19
ZyX-I Hello. How can I list all ignored files?12:20
FauxFaux git clean -nX12:20
cbreak ZyX-I: man git-ls-files12:21
gitinfo ZyX-I: the git-ls-files manpage is available at http://git-scm.com/docs/git-ls-files.html12:21
ZyX-I cbreak: You mean ls-files --ignored? It does not work, even with --exclude-standard.12:22
FauxFaux Monospace is still fucking broken under windows, though. Cunts.12:22
→ carloshh joined12:22
thiago please mind your language12:22
→ nifty_ joined12:22
ZyX-I By the way, why “--exclude-standar*d*”?12:23
cbreak ZyX-I: git ls-files -o -i --exclude-standard12:23
what else?12:23
ZyX-I cbreak: Thanks.12:23
cbreak --exclude-special? :)12:23
-o will show untracked12:23
thiago as opposed to excluding from a different list12:24
cbreak if you want to show tracked files that would be ignored but are not, use -c instead12:24
ZyX-I cbreak: For some reason I thought it is standart (that is how written in Russian). Dictionary tells I was wrong.12:24
cbreak standart is a kind of flag12:24
thiago it's still "standard"12:25
← Droolio left12:27
→ rjgonza joined12:27
→ simesy joined12:28
→ denysonique joined12:28
→ twisted` joined12:28
← dvide left12:30
← Jackarain left12:30
→ Jackarain joined12:31
→ Dimitr joined12:31
← theCole left12:31
← simesy left12:33
← tickbrown left12:34
← ketas left12:34
← babilen left12:35
→ ps_jinx joined12:35
← _eMaX_1 left12:37
← adlwalrus left12:37
→ thelazydeveloper joined12:38
→ marhaban joined12:39
→ ketas joined12:39
→ adlwalrus joined12:39
→ dleonardi joined12:41
→ fayimora joined12:47
→ codeM0nK3Y joined12:48
← carloshh left12:48
→ smellynoser joined12:49
smellynoser "Your local changes to 'cgi/web/account/login.php' would be overwritten by merge. Aborting."12:49
I don't give a **** about that file being overwritten, I just want the changes to go up without issue12:49
It's a remote server, so I just want whatever is the current git structure on the server12:50
I don't care about anything else12:50
thiago then undo your changes12:50
← adlwalrus left12:50
thiago git status will tell you what to do12:50
smellynoser There aren't any changes12:51
← shreya left12:51
thiago there are12:51
smellynoser Not actual changes, there could be changes git thinks because somebody else is pushing changes, but not to that file12:51
thiago it thinks there are local changes, by you, on that repository, to that file12:52
← pierre__ left12:52
→ pierre__ joined12:52
← capiira left12:53
smellynoser Yes, I'm aware of that - there isn't - this server is only used to run 'git pull origin master' by two people12:53
→ shreya joined12:54
← shreya left12:54
thiago it doesn't matter what you think12:55
it only matters that git thinks there are modifications12:55
you have to throw them away12:55
wereHamster smellynoser: running 'git pull origin master' to deploy the code is asking for trouble. Don't do that.12:56
smellynoser wereHamster, Thankyou - what is the best way of doing it? Should I just remove the directory and readd/init/pull ?12:56
wereHamster smellynoser: !deploy12:57
gitinfo smellynoser: Git is not a deployment tool. You can build one around it for simple environments. http://sitaramc.github.com/the-list-and-irc/deploy.html12:57
← Chappo left12:57
→ bartek joined12:57
thiago smellynoser: was that line prefixed with "remote:" ?12:57
← dleonardi left12:57
← specialGuest left12:58
← arcane left12:59
→ shruggar joined13:00
→ ladder4 joined13:01
AlecTaylor How do I revert the last commit?13:01
← ladder4 left13:01
→ neurodrone joined13:01
wereHamster AlecTaylor: git revert HEAD13:02
→ ladder4 joined13:02
← ladder4 left13:02
wereHamster AlecTaylor: are you sure 'revert' is what you actually want?13:02
→ eletuchy joined13:04
→ simesy joined13:04
← cyen left13:04
→ chuckharmston joined13:04
→ rohan32 joined13:05
← scholchr left13:06
→ ladder4 joined13:07
← eletuchy left13:08
← aviraldg left13:08
→ _Vi joined13:11
→ duckx joined13:11
cbreak I think AlecTaylor has a brain capacity of less than 10 minutes :(13:12
we already told him how to revert the last commit this morning13:13
after telling him three times how to check out a file from the past13:13
← RaceCondition left13:13
← jtri left13:15
← macIvy left13:16
→ shreya joined13:16
→ brownies joined13:19
← Shapeshifter left13:19
← Lgb left13:20
← bizhanMona left13:20
← palm_vue left13:20
← marhaban left13:20
→ ISF joined13:20
→ palm_vue joined13:20
→ Lgb joined13:20
→ bizhanMona joined13:21
→ madewokherd joined13:21
← kcj left13:23
→ marhaban joined13:25
← fayimora left13:25
→ fayimora_ joined13:25
← jdav_gon` left13:26
→ varg joined13:27
skorgon_ → skorgon13:28
→ fission6 joined13:29
→ ph^ joined13:30
→ nyxx_ joined13:31
→ jdav_gone joined13:31
← jceb_ left13:32
← marhaban left13:32
→ jceb joined13:32
→ specialGuest joined13:33
← shruggar left13:35
← FernandoBasso left13:36
← teotwaki_ left13:36
← skorgon left13:37
← ISF left13:37
← simesy left13:38
← jceb left13:38
→ AmirBehzad joined13:39
← chuckharmston left13:39
→ andrewhl joined13:40
← specialGuest left13:41
→ beneggett joined13:41
→ dleonardi joined13:42
← ps_jinx left13:42
→ Chappo joined13:43
→ simesy joined13:43
→ Lord_DeathMatch joined13:44
→ giallu joined13:44
→ FernandoBasso joined13:45
→ gtrak joined13:47
← dleonardi left13:47
← AmirBehzad left13:49
← mmc1 left13:50
→ pppaul joined13:52
→ gigamonkey joined13:52
→ arturaz joined13:55
← blacktulip left13:55
← ph^ left13:59
← beneggett left13:59
→ samyotte joined13:59
← simesy left14:01
→ reds joined14:02
→ iori joined14:02
reds → Guest4740814:02
→ Sebo joined14:02
→ Parhs8 joined14:02
← Parhs8 left14:03
→ Parhs joined14:03
Parhs Hello guys.. How to cancel the changes i made in some all files?14:03
→ simesy joined14:04
rue Parhs: You can checkout the file from HEAD, if you’ve not committed yet14:04
← samyotte left14:05
wereHamster Parhs: read the output of 'git status'14:05
← _Vi left14:05
→ _Vi joined14:06
← Sebo left14:06
← Guest47408 left14:07
← arturaz left14:07
← codeM0nK3Y left14:08
→ Anxi80 joined14:08
→ ISF joined14:11
→ taohappy joined14:12
← cesc left14:12
← taohappy left14:13
← shreya left14:13
→ cakehero joined14:13
→ lirt joined14:15
← simesy left14:16
← nyxx_ left14:17
→ dleonardi joined14:21
← smellynoser left14:22
← pppaul left14:23
→ Superlokkus joined14:24
← dc5ala left14:25
← Chappo left14:25
→ simesy joined14:25
← nor5trom left14:25
→ Dimitr|2 joined14:26
← lacrymology left14:26
→ RaceCondition joined14:28
← dleonardi left14:28
← Dimitr left14:28
→ arturaz joined14:32
→ Eridius joined14:33
→ cwang joined14:33
Bilge >some all files14:33
wat14:33
→ w41__ joined14:34
→ marhaban joined14:36
← lirt left14:37
→ RaceCondishun joined14:37
← w41_ left14:37
← lonewulf` left14:38
→ lonewulf` joined14:39
← w41__ left14:40
→ pppaul joined14:40
→ psoo joined14:40
← RaceCondition left14:40
RaceCondishun → RaceCondition14:40
← RaceCondition left14:40
← marhaban left14:41
→ yoklov joined14:41
→ dleonardi joined14:41
← simesy left14:43
→ OOPMan joined14:45
→ tshauck joined14:46
← fayimora_ left14:47
→ RaceCondition joined14:49
→ aspotashev joined14:50
← dleonardi left14:51
← varg left14:51
→ personal joined14:53
← mklappstuhl left14:53
→ marhaban joined14:53
← marhaban left14:54
→ marhaban joined14:54
→ herdingcat joined14:54
→ dleonardi joined14:54
→ hp joined14:56
→ kukks joined14:56
← hp left14:56
→ pngl joined14:56
pngl How can I set master to be a specific commit? I've checkout out an old revision to see some changes and now git says this revision is 'master'. Checkout out the most recent commit puts me in 'detached HEAD state'.14:57
← Anxi80 left14:58
→ AmirBehzad joined14:58
pngl In other words: I was at new_rev, I did `checkout old_rev`, now old_rev is master. I'd like new_rev to be master.14:59
→ flavius joined14:59
← blupp left15:00
← OOPMan left15:00
grawity I thought a plain `git checkout` doesn't normally update the branch...15:00
kpreid pngl: git checkout can't change what master is; you must have been already detached15:00
← scriabin left15:00
grawity anyway, `git checkout master`, `git reset --hard new_rev` should work15:00
kpreid yep15:01
pngl Yeah, I probably did something else15:01
kpreid but I would recommend checking on what the overall situation is with some visualization tool, log --graph or gitk15:01
make sure the relations among commits are what you think they are, before you go overriding things15:02
pngl worked, thanks. (gitk gave me older_rev ----> old_rev/master ----> some_rev ----> new_rev)15:02
grawity Hmm, trying to write a history-graft script with `git replace`. Is it a good idea to just run `sed` against the raw commit object? (Intention is to add a "parent ..." line.)15:03
→ mattih joined15:03
→ jargon- joined15:03
← Parhs left15:04
→ Parhs8 joined15:04
← gtrak left15:05
← _Vi left15:05
Mpenzi → mpenz15:06
mpenz → Mpenzi15:06
→ _Vi joined15:06
→ simesy joined15:06
← the_cybersphinx left15:06
→ arturaz_ joined15:09
← iori left15:09
← RaceCondition left15:09
→ RaceCondishun joined15:09
← uzyn left15:09
← cwang left15:10
← arturaz left15:10
← duckx left15:10
→ cybersphinx joined15:10
→ subbyyy joined15:13
→ vmiklos joined15:13
← Lord_DeathMatch left15:15
← simesy left15:17
← Dimitr|2 left15:17
→ drizzd joined15:17
→ Dimitr joined15:17
← cybersphinx left15:18
← mattih left15:18
rohan32 → rohan|afk15:18
← Milossh left15:20
← davidbanham left15:23
← neurodrone left15:24
→ cyen joined15:25
← dleonardi left15:25
← soooga left15:25
← AmirBehzad left15:26
→ jazper- joined15:27
← xiangfu left15:29
rohan|afk → rohan15:30
← esa left15:31
→ martinjlowm joined15:31
→ jonn joined15:32
→ esa joined15:32
→ kaen joined15:32
← psoo left15:33
← jey left15:33
→ nor5trom joined15:35
→ cantoma joined15:38
→ simesy joined15:39
→ d3vic3 joined15:40
← pppaul left15:40
← martinjlowm left15:41
→ reds joined15:42
reds → Guest9066015:42
→ martinjlowm joined15:43
→ mike007 joined15:45
→ HyperNerdV2 joined15:45
→ fayimora joined15:46
→ Chappo joined15:47
← tshauck left15:47
Freeaqingme I have 4 branches; development, testing, acceptance and production. The repo contains config files for various separate projects, whenever something is pushed to a branch, it is automatically deployed (puppet). whenever some functionality for a project has been found stable, I want that functionality to be merged with the testing branch. Because the development branch receives continues rolling updates, we simply cannot merge development with testing.15:49
→ darren joined15:49
→ skyf joined15:49
Freeaqingme How could we accomplish this best, merges of feature branches, cherry picking, or sth else? Also, what'd be the best way to keep track of stuff to be merged/picked/sth else into the next branch (once it's been found stable)?15:50
→ sirdancealot7 joined15:50
→ pppaul joined15:52
cbreak there should NEVER be more than one single project in any repository15:53
← pppaul left15:54
Freeaqingme cbreak, we have one puppet server that distributes config files across all our servers. So it's not the projects themselves, it's all the server configs that are stored in one repo15:54
cbreak so you're developing config files?15:55
Freeaqingme versioning them ;)15:55
cbreak you can just merge15:56
ideally, without crossing boundaries15:56
→ tshauck joined15:56
Freeaqingme that'd then be using feature branches. which is fine; but how could I then best keep track of what featurebranch has (or has not) yet been merged into what branches?15:56
cbreak git branch --merged15:57
or --no-merged15:57
Freeaqingme brilliant15:57
tnx15:57
← jargon- left15:58
→ identd joined15:59
← Guest90660 left16:01
← marhaban left16:01
← Superlokkus left16:04
→ jargon- joined16:05
← ISF left16:05
→ russt joined16:06
← identd left16:06
→ Zespre joined16:06
→ AmirBehzad joined16:09
CEnnis91_ → CEnnis9116:10
← AmirBehzad left16:10
→ marhaban joined16:10
→ AmirBehzad joined16:10
← AmirBehzad left16:10
← mike007 left16:14
← RaceCondishun left16:15
→ Shusshu joined16:16
→ skyf_ joined16:17
← kleist left16:18
← skyf left16:18
→ varg joined16:20
→ drizzd_ joined16:20
← Shusshu left16:21
→ OOPMan joined16:21
→ dropdrive joined16:22
← angelsl left16:22
← herdingcat left16:23
← drizzd left16:24
→ Pupeno joined16:24
← Pupeno left16:24
→ Pupeno joined16:24
→ cybersphinx joined16:24
→ KeepSake joined16:25
← marhaban left16:25
← OOPMan left16:25
← russt left16:26
→ ISF joined16:29
→ eicto joined16:29
→ Goplat joined16:31
→ NodeJohnny joined16:32
← tshauck left16:33
← gigamonkey left16:34
→ RaceCondition joined16:35
← RaceCondition left16:35
→ specialGuest joined16:35
→ Anders_J joined16:36
← NodeJohnny left16:36
← letharion left16:36
→ mmc1 joined16:36
→ whitman joined16:37
← specialGuest left16:37
→ specialGuest joined16:37
→ rehat joined16:37
← d3vic3 left16:38
← pngl left16:38
← nor5trom left16:38
→ iori joined16:38
→ RaceCondition joined16:39
← FernandoBasso left16:39
← personal left16:40
→ monsieurlu joined16:40
→ OOPMan joined16:41
→ duckxx joined16:41
← Chappo left16:41
← skyf_ left16:41
→ Chappo joined16:41
← Anders_J left16:45
→ spyro_ joined16:46
→ letharion joined16:46
← spyro_ left16:46
← cybersphinx left16:48
→ DrHouse|TARDIS joined16:49
→ w41_ joined16:50
nyxx If I have a branch on github called feature, and I just cloned the repo onto my system, how can I work on the branch feature? Doing `git branch` only shows `master`.16:53
i tried `git pull origin feature` but im still in the branch `master`16:54
← rohan left16:54
← subhojit777 left16:55
→ cybersphinx joined16:55
cbreak nyxx: you just merged feature into master...16:56
if you want to switch branches, use the branch switching command, git checkout16:56
→ igotnolegs joined16:57
← HyperNerdV2 left16:57
← lupo7 left16:58
nyxx cbreak: did i merge feature into master on the local repo?16:58
or on the github repo as well?16:58
→ sq-one joined16:58
← evanz left16:59
→ pelao joined16:59
← darren left17:00
nyxx cbreak: I want to redo that merge! I tried `git pull origin master` but it says its up to date17:01
oh no17:01
cbreak merges are local only17:03
you tried to merge the master from github into your local one17:03
no one pushed anything to master, so it was already up to date17:03
you should maybe reset --hard HEAD~1 the merge away17:03
(note that this will nuke away uncommitted changes too)17:03
→ GNUcifer joined17:04
cbreak it will remove the topmost commit17:04
← _Vi left17:06
← Parhs8 left17:06
→ Parhs joined17:06
→ _Vi joined17:07
nyxx ok fixed it... thanks!17:08
when I clone a repo, is it possible to also have all the branches created?17:08
cbreak you have tracking branches for all of them17:09
git will create local branches if you try to check out a tracking branch's basename17:09
with newer git17:09
→ oriba joined17:09
cbreak (if git checkout feature worked, then all should be well. if you are on a detached HEAD, ask again)17:09
nyxx i did git checkout -b feature origin/feature and it seems to work17:10
← arietis left17:10
cbreak that's what you have to do with older git17:10
if you add -t, it'll also track upstream17:10
← Chappo left17:10
nyxx im not familiar with the term tracking17:10
→ Chappo joined17:10
nyxx im on git version 1.7.917:11
→ arietis joined17:11
nyxx Can I use git to update git?17:13
AlecTaylor How do I revert the last commit?17:14
nyxx I tried following http://johnreilly.tumblr.com/post/40076200/update-git-using-git, but I cant find the directory src/vendor/git17:14
→ afuentes joined17:14
← drizzd_ left17:15
→ ology joined17:17
← Chib left17:17
CareBear\ AlecTaylor : git revert HEAD17:17
AlecTaylor : but maybe you want to use some of the other git features17:18
AlecTaylor : why do you want to revert?17:18
← GNUcifer left17:18
→ shreya joined17:20
→ ehsan joined17:20
← cybersphinx left17:21
← Onkeltem left17:21
← ehsan left17:22
→ dv310p3r joined17:24
→ shruggar joined17:25
← simesy left17:25
→ thirdknife joined17:25
← thirdknife left17:26
cbreak AlecTaylor: are you dumb?17:26
nyxx: you can update git via git if you compile git yourself17:26
AlecTaylor: you've been told by three different people how to do it and you still keep asking.17:27
maybe you should actually read the answers instead of spamming around like a brainless script.17:27
← wbs left17:28
← Parhs left17:30
← letharion left17:32
→ vasanth joined17:32
← vasanth left17:33
→ tshauck joined17:33
← centipedefarmer left17:34
← cagedwisdom left17:34
← Chappo left17:35
→ Chappo joined17:36
→ macIvy joined17:36
→ Parhs joined17:36
→ letharion joined17:40
CareBear\ sorry for not spotting the troll17:42
→ pppaul joined17:50
→ skyf joined17:53
→ iwoj joined17:53
← iwoj left17:53
← RaceCondition left17:54
← j416 left17:54
→ j416 joined17:55
→ subhojit777 joined17:55
→ Araxia joined17:56
← iori left17:57
→ darren joined17:57
→ aspotashev| joined17:57
← aspotashev left17:58
→ BadDesign joined18:00
→ cesc joined18:00
→ reds joined18:02
← darren left18:02
reds → Guest7144218:02
→ babilen joined18:03
→ pngl joined18:03
→ nor5trom joined18:04
→ pantsman joined18:05
← pantsman left18:05
→ pantsman joined18:05
← Guest71442 left18:06
← _Vi left18:06
→ Transcended_ joined18:07
→ v0n joined18:07
→ _Vi joined18:07
→ maletor joined18:08
← yoklov left18:08
← Jackarain left18:08
stillLotR → LotR18:09
← Transcended left18:10
← fission6 left18:10
v0n hi18:13
git log --stat doesn't add an entry for empty new files18:13
is there a way to display it?18:13
← Transcended_ left18:15
← subhojit777 left18:16
wereHamster nyxx: you can't find src/vendor/git because it's the directory where the author has the git source checked out. If you haven't done that, you first have to clone the git repo18:16
→ tureba joined18:16
← reactormonk left18:17
← skyf left18:17
← subbyyy left18:19
→ subbyyy joined18:19
← subbyyy left18:19
→ iwoj joined18:20
→ subbyyy joined18:20
← aspotashev| left18:21
← esa left18:21
→ gigamonkey joined18:24
← iwoj left18:24
→ darren joined18:24
→ skyf joined18:25
→ snearch joined18:27
→ TeamESA joined18:27
→ simon joined18:28
→ subhojit777 joined18:29
→ Davey joined18:29
← dr_lepper left18:31
→ TheJH joined18:33
mh` v0n: add --summary18:33
you'll get some extra info there, but that's one way to do it that i can think of18:33
v0n mh`, yeah I've just seen that. There's also --numstat which might be useful18:33
→ iwoj joined18:34
mh` if you want to normalize the output of them, and you dont actually need the line counts from stat18:34
you can do --stat --summary --name-status18:34
← AlecTaylor left18:35
← iwoj left18:35
→ simesy joined18:36
→ davetoo joined18:38
davetoo ok my brain/git-fu is not working. on my workstation I have succeeded in deleting a local tracking branch, and the remote (on github)... but on the other clients where I've cloned from github, I can't figure out how to get rid of the reference to the now-empty/gone remote branch18:40
mh` v0n: ah yeah, numstat looks best if you're trying to parse18:40
← simesy left18:40
mh` (combined with some other options for parse-ability)18:40
davetoo: git remote prune18:40
→ FatDarrel joined18:40
mh` if i'm understanding your question correctly18:40
← skyf left18:41
davetoo mh: I think you're right18:41
mh` "Deletes all stale remote-tracking branches under <name>."18:41
can add --dry-run to be safe18:41
that's a noop18:41
davetoo yep,18:42
-n is my friend :)18:42
mh` fwiw i usually do git remote update --prune18:42
davetoo and ... reveals that "you are correct, sir!"18:42
→ aspotashev| joined18:42
mh` probably not a good idea on shared repos :)18:42
davetoo thanks; I've used this before but those brain cells are gone, I guess18:42
mh` haha, no worries, most of mine are too18:42
v0n mh`, --name-status is nice too! Thanks18:43
mh` v0n: np18:43
davetoo indeed it is; I need more aliases18:43
→ adamm joined18:43
v0n that's actually for machine parsing so I'll check which is best18:43
← FatDarrel left18:43
← aspotashev| left18:43
← maletor left18:45
→ skyf joined18:45
→ rath joined18:46
← sh0ne left18:47
→ aspotashev| joined18:47
→ FatDarrel joined18:47
← tshauck left18:47
→ maletor joined18:47
→ lmayorga joined18:48
→ skorgon joined18:48
← aspotashev| left18:52
← skyf left18:53
← Chappo left18:53
→ Gitzilla joined18:54
← cakehero left18:56
→ hostilefork joined18:56
← lmayorga left18:57
← macIvy left18:57
→ jobin joined18:57
→ AaronMT joined18:58
← specialGuest left18:58
← skorgon left18:59
jobin → joey7718:59
→ Psycho_pr1 joined18:59
← hostilefork left19:00
← Psycho_pr left19:00
→ hostilefork joined19:01
← hostilefork left19:02
← fayimora left19:02
v0n it's strange that --shortstat or --summary have no effect with --names-only19:03
--name-status*19:04
cbreak maybe because of the -only part :)19:04
← tureba left19:04
→ hostilefork joined19:04
hostilefork Hello room… I am having my first experience merging someone else's changes with Git (as opposed to just cloning/pushing my own in a trivial way!)19:05
← FatDarrel left19:06
→ tureba joined19:06
hostilefork I'm a little confused, though. I'm sure this is a relatively simple question: but why is it that after I pull and get someone's changes, it shows those files as being "staged" in my branch…even when I didn't touch the files at all?19:06
It's as if Git wants to have the same files checked in for more than one patch set. Or am I just seeing something that it's going to sort out when I push my own commit?19:08
cbreak what's a patch set?19:09
← rath left19:09
cbreak if you merge other people's changes, git will create a merge commit, or fast forward19:10
if there's a conflict, git will tell you that there's one and you have to resolve it and then git add and git commit19:10
hostilefork Yup that part I get. Those show up as "files in the working tree"19:10
( I'm using SourceTree: http://www.sourcetreeapp.com/ )19:10
In the window for "files staged in the index" it has (for instance) files that this other person added to the project that I've never touched19:11
I'm just worried that when I do my commit it will be like those files are being…checked in again?19:11
CareBear\ maybe sourcetree support can help you?19:12
I think we primarily know git tools19:12
→ program247365 joined19:12
→ rtjure joined19:13
hostilefork Well, all source tree appears to be doing is showing the output of "git status" with some icons19:13
But I did a pull, and then it's showing "new file" etc from this other person's changes.19:14
→ fayimora joined19:14
→ skorgon joined19:14
hostilefork If after I commit and push, I don't want my checkin to appear to have been the one that added those files.19:14
→ irqq joined19:14
→ specialGuest joined19:15
hostilefork I'm just wondering if this is just the nature of how git works under the hood, that I can just trust it will realize what's going on.19:15
← specialGuest left19:15
→ specialGuest joined19:15
← program247365 left19:16
→ pxdr joined19:17
pxdr whois cmn19:18
← arturaz_ left19:18
cmn me19:18
← palm_vue left19:20
← bizhanMona left19:20
← Lgb left19:20
→ Lgb joined19:20
→ palm_vue joined19:20
→ bizhanMona joined19:21
← subhojit777 left19:21
← pngl left19:27
← pcapriot` left19:27
← pxdr left19:29
gigamonkey Say I've got a branch that I've done a push --force from one machine to a central repo. Now I want to pull it to another machine and just take the version from the central repo. What's the right way to do that?19:29
(I'm the only person using this branch, which is why it seems okay to do the push --force in the first place.)19:30
→ rohan32 joined19:30
→ dfr|mac joined19:31
rohan32 hey, im working with the repo tool in conjunction with git, and whenever i repo sync it discards the commits i've cherry picked. is there any way to keep the cherry picks and make them not affected by a repo sync?19:31
← davetoo left19:31
wereHamster rohan32: ask the people who wrote that tool19:32
← jrdnn left19:32
rohan32 wereHamster, it is Google :D just wondering if any one else works with the tool (for Android) and knows of a solution19:33
→ jey joined19:34
gigamonkey Hmmm. It seems like git reset --hard origin/mybranch does the trick. I guess that makes sense.19:37
→ Kyzz joined19:39
← joey77 left19:39
→ axl_ joined19:39
→ converge joined19:39
← ReekenX left19:41
← snearch left19:41
→ aspotashev| joined19:41
→ joey77 joined19:41
→ snearch joined19:41
← hostilefork left19:41
→ bubbles10101 joined19:41
→ RaceCondition joined19:42
← Goplat left19:42
→ nyx joined19:43
→ w41__ joined19:43
← snearch left19:43
→ jrdnn joined19:43
← axl_ left19:44
← nyxx left19:44
← w41_ left19:46
→ yoklov joined19:46
→ chrisgeorge joined19:46
→ kewah joined19:47
← w41__ left19:47
→ FernandoBasso joined19:48
→ AmirBehzad joined19:50
← yoklov left19:50
→ theCole joined19:51
← AmirBehzad left19:52
→ FatDarrel joined19:52
← ISF left19:53
→ yoklov joined19:53
← simon left19:54
← chrisgeorge left19:54
→ ISF joined19:54
← yoklov left19:54
→ yoklov joined19:55
← ISF left19:57
← tnzr left19:58
→ tnzr joined19:58
→ eph3meral joined19:58
eph3meral how can I git add --patch but without interactive mode?19:58
EugeneKay rohan32 - no, `repo` sucks. Sorry. Bitch to Google :-p19:59
eph3meral like, can I choose specifically, which patch to add, using the command line19:59
given that, for example, I know the line numbers or data in the diff?19:59
gitg seems to do this19:59
emacs's magit seems to do this19:59
← aspotashev| left20:00
← yoklov left20:00
→ nizeguy joined20:02
→ reds joined20:02
← KeepSake left20:03
reds → Guest9583820:03
← nizeguy left20:03
→ aspotashev| joined20:04
→ jceb joined20:04
← oakdog8 left20:04
→ _oakdog8 joined20:04
→ cybersphinx joined20:05
→ an0maly_ joined20:05
← _oakdog8 left20:07
→ oakdog8 joined20:07
← Guest95838 left20:07
cmn eph3meral: interactive add, if you want to go directly to the patch part use -p20:08
← oakdog8 left20:08
← flippo left20:08
mh` he's saying he wants to provide specific line ranges from the command line, instead of interactively adding hunks20:08
(i think)20:08
eph3meral cmn, i don't think you've understood my question, -p and --patch are the same, I understand20:08
i don't want any interactivity20:08
--patch just bypasses the menu during --interactive and goes straight to --patch mode20:08
where it asks me about every change20:09
cmn ah, that line was above the cut20:09
eph3meral i don't want it to ask me about every change, i want to just be able to tell it which hunk to use20:09
cmn if you want to do that, then you'll need to program it20:09
mh` which hunk or which arbitrary lines?20:09
those are two separate goals20:09
← RaceCondition left20:09
eph3meral mh_le, um... hunk I think20:09
whatever gitg does20:09
← CpuID3 left20:09
eph3meral not sure if you guys have used it, but it's got a nice little 'commit' tab where I can cherry pick parts of my files to stage20:10
← jrdnn left20:10
cmn how are you meant to know which hunk to stage if there is no way to ask you, though?20:10
eph3meral i don't need to be prompted in order for every one, i just click on the one I want and it goes in20:10
cmn, I have the output of git diff open in a different terminal20:10
cmn then you don't want anything from git20:10
having a diff somewhere else doesn't mean much20:11
eph3meral cmn, the point is that I want to write a plugin for my editor, to do what gitg does20:11
cmn git runs as a single process without context from whatever diff you ran before20:11
eph3meral where I can point and click on a single bit of change in a single file20:11
and stage just that change20:11
→ RaceCondition joined20:11
cmn gitg probably re-implements it itself20:11
eph3meral ahh20:12
so there's no command line way to do this then? by default anyway?20:12
→ LekeFly joined20:13
← v0n left20:13
→ pxdr joined20:13
Element9 I found a nice fork & pull workflow explanation for github: https://github.com/diaspora/diaspora/wiki/Git-Workflow But I don't understand what's 8th step for (fetching the upstream). What's that for?20:13
cmn the only way would be to show you the whole thing first, assume that absolutely nothing changes and then let you copy some magic number20:14
and that's not something that makes sense as UI20:14
← segher left20:14
→ jkremser joined20:15
← sq-one left20:15
← gigamonkey left20:15
→ segher joined20:16
→ lupo7 joined20:16
← aspotashev| left20:17
← theCole left20:17
← lottadot left20:18
eph3meral cmn, gitg makes plenty of sense as a UI20:23
→ Rajesh joined20:23
eph3meral cmn, i don't and it's nothing like what you described20:23
→ aspotashev| joined20:23
eph3meral nor would be my plugin20:23
cmn because I'm not describing either gitg or your plugin20:23
eph3meral ah, ok :)20:23
Rajesh → Guest9481720:23
cmn you asked about git20:24
mh` there are a number of decent libraries - high(er)-level bindings for libgit, and otherwise - in various scripting languages20:24
eph3meral well i think it makes perfect sense, it's just an advanced feature "use at your own risk" basically, i mean yes of course you would assume that nothing changes between the time you last looked at your OWN diff, and the time you ran the command to stage a single hunk - because that same problem exists with gitg, hence why there's a refresh command20:25
mh` if it's something you want to do programmatically it probably would not be hard to make something that would suit your needs form ost cases20:25
eph3meral a feature like this on CLI would just basically mean that you'd need to keep track of changes yourself, but it would be nice/useful for automation IMO20:25
← rehat left20:25
mh` eph3meral: i don't disagree, just saying how you could accomplish this :)20:26
eph3meral I wonder how gitg accomplishes it, because git status on CLI definitely knows that only part of the file has been staged20:27
like, i have the same file name listed under both staged and undstaged changes20:27
mh` that's totally possible, it's the selection process that's hard20:27
← FatDarrel left20:27
← AaronMT left20:27
mh` try it yourself: change two non-contiguous parts of a file, do git add -p, hit yes to the first and no to the 2nd20:27
now do git status20:27
and git diff --cached20:28
← milian left20:28
mh` then do git reset (to unstage those changes)20:28
do add -p again, but on the first one hit 'e' for edit20:28
and now you're editing the patch20:28
as you can see, you can stage any arbitrary thing (for better or worse)20:28
interactive mode stops you from staging a hunk that doesn't apply properly20:28
→ Chappo joined20:29
→ rtjure_ joined20:29
→ avpx joined20:31
eph3meral yes, ok I still don't believe I will be in danger of suffering from this problem of having hunks not apply properly20:31
← aspotashev| left20:31
eph3meral because, I don't anticipate my files changing while I'm not changing them20:31
and if I'm perusing through diffs, I'm not changing my files20:31
but that is a nice feature20:31
← rtjure left20:31
mh` no, it's not that20:31
→ specialGuest1 joined20:32
mh` i think that's where the disconnect is from what you guys were discussing before20:32
are you familiar with how patch(1) works?20:32
eph3meral ok20:32
mh` with regards to context20:32
eph3meral er, probably not - I mean I know what a patch is, basically, I know what a diff is basically, I'm not super familiar with manually creating or using either one20:32
i've run patch -p0 < foo.patch like a handful of times in my life20:32
mh` haha :)20:32
→ anonymuse joined20:33
← anonymuse left20:33
mh` k so patches that use "unified" diffs, instead of just specifying exact line numbers and adding/removing those, provide contextual lines around the ones you changed20:33
that way if the file changes you still have a chance of applying the patch20:33
(and it going to the right place)20:33
so the way you could make a hunk that won't apply is by doing something that causes it to not be able to make sense of the context20:34
← bubbles10101 left20:34
mh` which there are several ways to do, but it's not hard20:34
so that's how you could end up with a hunk that "doesn't apply", if you use this type of patching method20:34
← specialGuest left20:34
mh` by staging a diff that didn't bring in sufficient (or usable) context20:34
→ arand joined20:35
mh` eph3meral: take a look at this- http://nuclearsquid.com/writings/git-add/ ctrl-f down to the paragraph that begins with "Manually editing the hunk"20:36
← specialGuest1 left20:37
arand is it possible to get a diff --stat listing only whole-file add/removals? (both binary and text)20:37
eph3meral mh yes I understand editing20:37
i did it just now as you suggested20:37
mh` arand: --diff-filter i think?20:37
eph3meral er, a little bevore this20:37
mh` A and R are add/remove respectively20:37
eph3meral: yeah i was pointing you to the paragraph about the troubles of cleanly applyable hunks20:38
anyway, eph3meral20:38
for your case specifically20:38
eph3meral ahh20:38
→ bubbles10101 joined20:38
arand mh`: Ah, thanks :)20:39
← dfr|mac left20:39
mh` if you can be sure the source file doesn't change (you're only using this to stage changes, not generate patches).. it's probably possible to create very specific patches20:39
eph3meral: for that matter, have you just looked at the source of gitk or git-cola or etc to see how they do it?20:39
→ centipedefarmer joined20:39
eph3meral mh`, nope, not yet20:39
mh` what language is your editor in?20:39
the one you're building the plugin for20:40
or what language is your plugin in, i guess is a more appropriate question20:40
eph3meral i honestly don't feel like spending much more time on this at the moment, I have, for the moment decided to basically give up :)20:40
mh`, vimscript20:40
mh` that's no fun20:40
→ d3vic3 joined20:40
mh` oh20:40
→ aspotashev| joined20:40
mh` now i understand giving up! :)20:40
eph3meral mh`, well part of the problem, is that TPope can't code for shit20:40
mh` oh, you're using that20:40
eph3meral well, i was going to add to it, the couple of features that I wanted20:40
mh` fugitive, i assume?20:40
eph3meral yep20:41
mh` heh20:41
my dotfiles have a fork of that20:41
its the only vim plugin i forked20:41
out of like 3020:41
→ specialGuest joined20:41
mh` just to fix bugs20:41
eph3meral haha, his code is soooo fucking ugly i can't stand it20:41
← Davey left20:41
mh` i don't know vimscript though, so haven't contributed back.. they were just glaringly obvious20:41
eph3meral zero comments, fuck all for whitespace20:41
← lcc left20:41
mh` gimme a sec, im looking at git-cola to see how they do it20:42
eph3meral anyway, I'm actually going to use emacs for git stuff :P20:42
the emacs git plugin is pretty awesome - I was hoping it would be doable within a day to add similar functionality to fugitive, but hell no20:42
my only real complaint about the emacs plugin is that it's slow as fuck20:43
well, that's not a complaint about the plugin, that's actually just a complaint about emacs, but as a side effect, the plugin is slow as fuck also20:43
→ lcc joined20:43
petteyg_ → petteyg20:43
← cybersphinx left20:44
← aspotashev| left20:44
eph3meral quite honestly, all i really want is for gitg to handle my stashes20:45
so I can delete and apply them20:45
mh` https://github.com/git-cola/git-cola/blob/master/cola/cmds.py#L14020:45
eph3meral that's all20:45
→ cybersphinx joined20:45
→ fission6 joined20:45
eph3meral or something, to easily handle my stashes20:45
i don't really care that much about integrating vim with git, cuz gitg works basically fine20:45
mh` i just use git stash..20:46
eph3meral yeah but it's a huge pain to copy paste the stash hash no?20:46
mh` copy paste it to where?20:46
eph3meral to operate on a stash20:46
mh` they have names, optionally20:46
eph3meral mh` are you aware that you have multiple stashes?20:46
ahh, i don't usually name mine20:46
mh` yes, of course20:46
eph3meral i just git stash20:46
mh` git stash save somename20:47
git stash list20:47
git stash pop stashname20:47
etc20:47
← darren left20:48
mh` if you have tab completion its trivial20:48
just name it, then in stash list you'll see it20:48
→ ries joined20:48
mh` then git stash pop <tab>20:48
ries Hello! How do I 'undo' a commit?20:48
eph3meral mh`, oh, hmm, ok that's easy enough, and to delete?20:48
mh` 'drop'20:49
eph3meral ok, thanks20:49
mh` man git stash20:49
gitinfo the git-stash manpage is available at http://git-scm.com/docs/git-stash.html20:49
mh` its one of the few manpages thats simple :P20:49
← _iron left20:49
eph3meral cool thanks20:49
mh` np20:49
Element9: the upstream thing is so you can pull in changes that have occurred in the upstream20:50
since you forked it20:50
Element9 mh`: why do I need to fetch. couldn't I just pull?20:51
mh` Element9: fetch downloads, pull merges/applies/what-have-you20:52
eph3meral oh, fugg... lol I can do this from within gitg already anyway :)20:52
mh` downloads into the index20:52
→ jmcantrell joined20:52
mh` but in their howto there, they are rebasing, not pulling20:53
so that's why20:53
fetch downloads it into the local repo's index, without affecting your working tree20:53
Element9 mh`: could they have just skipped the fetching step? pull would fetch, would it?20:55
eph3meral hah, ok well, all this time, ... now I can scratch this off the list :)20:55
← longbeach left20:55
→ cwang joined20:56
mh` Element9: it would, but then it would also do the pull part :) they are rebase'ing, which is not the same20:56
they link to an explanation- https://github.com/diaspora/diaspora/wiki/Git-Workflow#wiki-gitrebase120:56
← specialGuest left20:57
→ specialGuest joined20:57
← whitman left20:58
→ dr_lepper joined20:59
← chuckf left20:59
Element9 but they do pull in step 9 :) ah man, seems like I'm really missing something here20:59
← varg left21:00
mh` step 9?21:00
oh21:00
i see21:00
← rtjure_ left21:00
← shreya left21:00
mh` yeah, you are missing something :)21:00
you're working on your own feature branch that you created in step 521:00
and thats what you're rebasing (off of master)21:01
when they do the pull21:01
they switched branches back to master first21:01
see it now?21:01
→ carloshh joined21:01
← eph3meral left21:01
← lcc left21:01
Element9 yes, but pull will do fetch... which is for all branches, isn't it?21:02
mh` fetch downloads them but it doesn't apply it to your working tree21:03
→ oakdog8 joined21:04
→ Davey joined21:04
mh` (it just updates what's inside the .git dir)21:04
← cantoma left21:05
← d3vic3 left21:05
mh` pull brings in those new changes (merging) to your working tree (your actual files in your local repo)21:05
Element9 so is the step redundant then as pull will do the fetch too?21:05
mh` i'm not sure if pull, by default, will fetch all the branches21:07
but if master was the only branch of concern then yes it would be redundant in some sense21:07
however... it's a good habit to get into (separating fetch/pull), which is probably why they are doing it here21:07
there are many times where it's what you'd want to do21:07
← _Vi left21:07
→ chrisdev joined21:07
mh` s/many/most21:07
→ _Vi joined21:08
→ linoj joined21:09
Element9 that's what I wanted to know. thanks for the help! :)21:09
mh` np!21:09
← dv310p3r left21:11
→ _oakdog8 joined21:12
→ kcj joined21:12
← BadDesign left21:12
← skorgon left21:13
→ oakdog joined21:13
→ FatDarrel joined21:13
← oakdog8 left21:14
← FatDarrel left21:14
← Araxia left21:15
← pppaul left21:16
← _oakdog8 left21:17
← oakdog left21:17
→ oakdog joined21:17
← carloshh left21:17
oakdog → oakdog821:17
→ pppaul joined21:18
→ tshauck joined21:19
← timthelion left21:19
→ rehat joined21:19
→ dougmencken joined21:23
dougmencken hi guys; dev asked me about "git format patches are very easy to apply for me", but:21:24
git: 'format' is not a git command. See 'git --help'.21:24
thiago format-patch21:24
dougmencken I'm using git diff, and he is not satisfied21:24
thiago: git format-patch?21:24
trying it21:24
git format-patch prints nothing21:25
thiago give it some arguments21:25
dougmencken does it have man page?21:25
thiago yes21:27
mh` did you try man git format-patch21:27
gitinfo the git-format-patch manpage is available at http://git-scm.com/docs/git-format-patch.html21:27
→ r0bby_ joined21:27
← Internet13 left21:27
← ahayweh left21:28
dougmencken thanks guys!21:28
← robbyoconnor left21:29
dougmencken see you21:29
← dougmencken left21:29
→ flippo joined21:29
→ robbyoconnor joined21:30
← jey left21:31
← Guest94817 left21:31
← jkremser left21:31
← r0bby_ left21:31
← tshauck left21:32
← linoj left21:33
→ pngl joined21:33
→ jey joined21:37
← pppaul left21:37
→ pototo joined21:39
→ w41_ joined21:39
← jceb left21:40
← joey77 left21:41
→ Internet13 joined21:45
← jey left21:46
← dzonder left21:48
← afuentes left21:50
← lupo7 left21:51
← optixx left21:51
← nifty_ left21:55
→ Araxia joined21:56
→ jey joined21:56
→ reactormonk joined21:56
→ davidbanham joined21:56
→ Speed joined21:57
← Speed left21:57
→ Speed joined21:57
← davidbanham left21:57
→ psoo joined21:58
→ alanp_ joined22:01
→ reds joined22:03
reds → Guest817122:03
← irqq left22:04
← alanp left22:05
→ milian joined22:05
← j416 left22:07
← Guest8171 left22:07
← jey left22:07
← _Vi left22:08
→ Tommy[D]_ joined22:08
→ mizu_no_oto_ joined22:08
→ btree joined22:08
→ _Vi joined22:09
← Tommy[D] left22:11
→ tshauck joined22:11
← fayimora left22:11
← NimeshNeema left22:11
← TheJH left22:11
fission6 if i do git merge somebranch, how do i undo the whole git merge22:11
or do i have to revert each indivudal commit22:11
→ obiwahn joined22:11
→ i7c joined22:12
→ canapp joined22:12
→ TheJH joined22:12
thiago fission6: have you pushed the merge somewhere?22:13
fission6 yes, to origin22:13
but i know now one else has pulled the changes22:13
→ NimeshNeema_ joined22:13
→ Fretta joined22:13
fission6 thiago: ?22:16
← giallu left22:17
← noyb left22:19
← TheJH left22:19
← Sonderblade left22:20
← Tommy[D]_ left22:21
← nor5trom left22:22
→ Tommy[D] joined22:22
erynofwales fission6: you could just reset to the commit before the merge, unless it's a fast-forward, then it might be slightly more complicated because you have to move both the current branch and some branch back to their previous commits.22:24
→ nor5trom joined22:24
← pelao left22:25
erynofwales some branch = somebranch22:25
fission6 what do you mean by the last line22:25
→ nifty_ joined22:25
← tjadc left22:26
erynofwales which part?22:26
fission6 unless its a fast forward22:26
→ Sebo joined22:26
fission6 it has several commits, the merge, because there were no conflicts22:26
→ ku joined22:26
ku How badly would Git perform with a bunch (say 100) 500MB (approx) files with dozens of changes each?22:27
← Chappo left22:27
FauxFaux ku: Poorly. !big22:27
gitinfo ku: [!large_files] Git isn't yet great at large files(larger than RAM). Look at !annex for solutions. You can find them (after gc) with: git verify-pack -v .git/objects/pack/pack-*.idx | grep blob | sort -k3nr | head | while read s x b x; do git rev-list --all --objects | grep $s | awk '{print "'"$b"'",$0;}'; done22:27
→ Chappo joined22:28
tizzo-afk → tizzo22:29
ku FauxFaux: 500MB isn't larger than my RAM?22:29
thiago Git isn't designed to handle large, binary files22:30
it's designed to handle lots of small, text files22:30
FauxFaux ku: I'm reasonably sure it'll go horribly wrong if the repo size, or the history of a single file in total, is over RAM.22:30
thiago it _works_, in the sense that it keeps those files perfectly and revisions them22:30
FauxFaux There've been lots of improvements recently, though, so maybe it's okay!22:31
ku Will future improvements be applicable backwards?22:31
s/backwards/retroactively22:31
thiago that's just performance improvements22:31
it doesn't change the storage format22:32
if you have 100 * 500 MB files * 10 changes each, if they are binary, they compress and delta very little22:32
erynofwales fission6, so a fast-forward merge is where the branch you're merging (somebranch in this case) is a child of the branch you're currently on. So, you have something like this: A B C [currentbranch] D E F [somebranch]. In this case, git can just move currentbranch up to F so you have this: A B C D E F [currentbranch, somebranch]. Does that make sense?22:32
fission6 how can i show files changes for a specific file across many commits22:32
SethRobertson Some (bupish ones for instance) might change storage22:32
thiago your repository will be around 500 GB22:32
you REALLY don't want that22:32
fission6 erynofwales: yea.22:32
→ ksksal joined22:32
SethRobertson fission6: I suggest using: gitk --all --date-order -- filename22:33
→ davidbanham joined22:33
ku thiago: wouldn't that still be better than pure snapshots?22:34
← ksksal left22:34
thiago given even a little compression, yes, it would be better than pure snapshots22:35
in terms of storage22:35
it doesn't mean Git will be fast, though22:35
ku How about other solutions like boar? Would that be better?22:36
I think git-annex will support some kind of versioning soon, but not yet, as part of its Kickstarter campaign22:36
thiago I can't speak about it, since I've never heard of them22:36
SethRobertson Um, git-annex supports versioning now. It is part of git22:36
well, used with git22:36
ku SethRobertson: link?22:37
SethRobertson !annex?22:37
gitinfo git-annex and git-media are two solutions to the !binary problem. They work by keeping the blobs outside of the repo, storing a reference to the blob in the repo instead. See http://git-annex.branchable.com/ and https://github.com/schacon/git-media22:37
← david1010 left22:38
→ tjholowaychuk joined22:38
→ ISF joined22:38
← ology left22:38
→ \ology joined22:39
ku SethRobertson: no, I mean for versioning. I don't see any information on annex versioning22:39
→ eph3meral joined22:40
← cwang left22:40
SethRobertson Why do you need to version the annex? You version the git repo and git points you at whichever version of the files you need22:40
eph3meral how can I answer the question of "at what point did my current branch diverge from master"22:40
er, rather "show me the commit at the point when my current branch diverged from master"22:40
so that I can do this:22:40
SethRobertson eph3meral: git merge-base master22:40
eph3meral git diff ORIGINALHASH HEAD22:40
← monsieurlu left22:40
ku SethRobertson: I'm not sure I understand. so annex supports versioning binaries?22:41
eph3meral to see the "full diff" of my branch as compared to wherever it diverged from master22:41
\ology → ology22:41
SethRobertson eph3meral: `git diff master..`22:41
eph3meral SethRobertson, ok thanks, what does that show specifically?22:41
er, the merge-base command that is, what does that show/22:41
SethRobertson man git-merge-base22:41
gitinfo the git-merge-base manpage is available at http://git-scm.com/docs/git-merge-base.html22:41
→ noyb joined22:41
eph3meral gotcha, thanks22:41
ku SethRobertson: would something like git -> annex -> bup be best?22:42
SethRobertson But of course if all you are doing is running diff, you can do the master.. thing I said22:42
eph3meral ok cool22:42
SethRobertson ku: !xy22:42
gitinfo ku: This sounds like an "XY Problem" http://mywiki.wooledge.org/XyProblem So let's step back for a minute. What are you actually trying to achieve? Why are you doing it this way?22:42
eph3meral is there a decent way to aggregate all my commits in to a single commit before submitting a pull request for example?22:42
SethRobertson, I'm looking to help make the actual git history atomic, while being able to still use git as a "state save" machine on a regular/working basis22:43
SethRobertson eph3meral: many. One is !perfect22:43
gitinfo eph3meral: [!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.22:43
eph3meral ooh, nifty thanks22:43
i use rebase regularly already22:43
ku SethRobertson: I am trying to track big, binary files using a Git-style decentralized workflow22:43
SethRobertson Yes. That doesn't require bup22:44
ku Isn't bup meant for packing binaries?22:45
← leeN left22:45
ku which would save more space than simply git + annex?22:45
SethRobertson bup is for backups22:45
→ leeN joined22:45
ku ok22:45
but wouldn't annex be a little inefficient? if used by itself22:46
also, do you have a guide on using annex with versioning?22:46
SethRobertson git does the versioning. annex is just saving the files22:47
annex is trading bandwidth/disk space for speed of local operations. Is this good for you? Only you can decide.22:47
ku I know, but I don't understand what commands to use for that22:48
I think I'll try annex first22:48
← Chappo left22:48
→ Chappo joined22:49
← Anubisss left22:49
← wereHamster left22:49
→ Anubisss joined22:49
ku SethRobertson: what commands should I use for versioning annex? annex seems to have a different interface from plain git22:49
← Eridius left22:49
← Anubisss left22:49
→ Anubisss joined22:50
← tshauck left22:51
SethRobertson ku: When a file is annexed, its content is moved into a key-value store, and a symlink is made that points to the content. These symlinks are checked into git and versioned like regular files. You can move them around, delete them, and so on. Pushing to another git repository will make git-annex there aware of the annexed file, and it can be used to retrieve its content from the key-value store.22:51
← blami left22:52
→ AaronMT joined22:52
SethRobertson You can arrange for the files in the annex to have nonce names and point the symlink file to the correct name.22:52
ku nonce?22:52
noncense?22:52
* sense22:52
← Sebo left22:53
ku SethRobertson: how *exactly* would I do that though? I don't see any options from fetching a specific version in annex22:53
SethRobertson git-media looks like it might be a little more direct in that respect22:54
← noyb left22:54
SethRobertson That's the point. You store a file, named perhaps 000000001, into the annex. You get a symlink in the local repo. Rename the symlink to be the actual desired name of the file. Commit22:55
ku SethRobertson: but that has to be done manually?22:56
SethRobertson At that point, you have a link between this git commit SHA with the filename to a specific file in the annex22:56
Try it out. Perhaps it handles versions automagically. I've never used it. However, from reading the text I pasted in, it very much sounds like the technique I listed will work. Other techniques might also work.22:57
→ thirdknife joined22:57
→ Lord_DeathMatch joined22:57
→ blami joined22:57
CareBear\ does anyone know if the same repo added as submodule in more than one places means multiple copies of the submodule repo, or just multiple workdirs of it?22:58
ku SethRobertson: thanks, I'll see what I can do22:58
SethRobertson CareBear\: Separate repos unless you take lots of special care22:59
CareBear\ as in engineer the tree manually?22:59
→ v0n joined22:59
CareBear\ or what do you mean?22:59
← LekeFly left22:59
SethRobertson Yes, doing manual work to cause the bits and the pointers to consolidate23:00
CareBear\ yeah ok23:00
mumble mumble23:00
SethRobertson disk space is cheap23:00
CareBear\ mh annoying23:01
ku SethRobertson: native versioning is coming to git-annex assistant: http://git-annex.branchable.com/design/assistant/syncing/23:01
← leeN left23:01
gitinfo set mode: +v23:02
SethRobertson That page doesn't talk about versioning. It is talking about scheduling and managing transfers23:02
ku SethRobertson: if you read the blog, you'll see that syncing includes versioning23:03
SeoZ-work[AWAY] → SeoZ23:03
CareBear\ SehtRobertson : thanks for the answer though!23:03
tizzo → tizzo-afk23:03
gitinfo set mode: +v23:04
→ simesy joined23:06
← kewah left23:06
← Speed left23:08
← _Vi left23:08
SethRobertson ku: It appears to work now: https://github.com/joeyh/git-annex/blob/55c512ff2cdd420bda5da2a0cdc222ed40376849/doc/walkthrough/modifying_annexed_files.mdwn23:09
→ _Vi joined23:09
→ brandure1 joined23:10
← brandure1 left23:11
← maletor left23:12
→ dvide joined23:12
← nifty_ left23:12
→ nifty_ joined23:13
← thirdknife left23:16
← davidbanham left23:17
→ brandure1 joined23:17
→ bulletmark joined23:17
← nifty_ left23:18
← arietis left23:22
← drougge left23:24
← brandure1 left23:24
→ jobin joined23:24
→ brandure1 joined23:24
← thiago left23:25
→ thirdknife joined23:25
← pngl left23:25
v0n isn't possible to view the patch with git show when using --name-status option?23:26
← psoo left23:27
hyperair --name-status? what's that?23:27
the git show manpage doesn't show anything about it23:28
v0n hyperair, it inherits it from git diff23:28
← tjholowaychuk left23:31
→ jjang joined23:36
← Element9 left23:38
→ ehsan joined23:38
← bulletmark left23:39
← ehsan left23:41
← FernandoBasso left23:42
→ Psycho_pr joined23:43
← Psycho_pr1 left23:43
← converge left23:43
hyperair okay, so why do you want to use --name-status and view the patch?23:44
sounds like you want --stat instead23:44
← thirdknife left23:45
← shruggar left23:47
→ pngl joined23:48
→ maletor joined23:49
← hyperair left23:54
← specialGuest left23:58
→ darren joined23:59

Logs Search ←Prev date Next date→ Channels Documentation