IRCloggy #git 2013-10-18

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.

2013-10-18

→ Algorithm joined00:00
→ caspingus joined00:00
→ cmpera joined00:00
→ vkz joined00:01
← maxorator left00:01
← hyperair left00:01
← ktosiek left00:02
→ frogonwheels_ joined00:02
→ Tok-A-Mak joined00:03
→ el3ctron joined00:03
→ ktosiek joined00:03
EugeneKay gdoteof - man git-format-patch00:04
gitinfo gdoteof: the git-format-patch manpage is available at http://jk.gs/git-format-patch.html00:04
← frogonwheels left00:04
← ZujkisNx left00:05
→ normanrichards joined00:05
← VictorCL left00:05
Tok-A-Mak is it possible to have a file in a repo, that doesn't keep any history, but is just the current version (it's a large jar and i don't want to bloat my repo with crap on every update)? if so, what should i read to understand what to do?00:06
← epicowl left00:06
← Zarthus left00:07
← kcj left00:08
→ bwreilly_ joined00:08
→ kcj joined00:08
← bwreilly left00:12
→ grilix joined00:13
→ R1KRDO1 joined00:13
→ valgrind joined00:13
← bwreilly_ left00:13
← imachuchu left00:14
→ hashpuppy joined00:15
← hashpuppy left00:16
← R1KRDO left00:16
→ Stygia joined00:16
← asQuirreL left00:16
imMute Tok-A-Mak: not directly, but have a look at git-annex00:17
Tok-A-Mak i will check it out. thanks00:17
← akurilin left00:18
→ rfoust joined00:19
← cenuij left00:20
→ Gamecubic joined00:20
→ akurilin joined00:20
← boblee left00:22
→ boblee joined00:22
→ reskp joined00:23
→ setmeaway joined00:24
→ tmcmahon joined00:24
→ thierryp joined00:26
← normanrichards left00:27
← jayrobot left00:28
→ hashpuppy joined00:29
→ normanrichards joined00:29
→ dyng joined00:30
← thierryp left00:30
← ehmatthes left00:30
← reskp left00:32
← chuckharmston left00:33
← jarr0dsz left00:34
← minim left00:34
← Fandekasp left00:36
← Stalkr^ left00:37
← AWinterman left00:39
→ Stalkr^ joined00:39
→ MavKen joined00:40
→ Fandekasp joined00:40
← Curtisbeef left00:41
→ cannonball joined00:41
→ BiggFREE joined00:42
→ ehmatthes joined00:42
→ delta6 joined00:42
→ ozette joined00:43
← simesy left00:43
← tmcmahon left00:44
→ simesy joined00:44
EugeneKay Tok-A-Mak - !annex00:46
gitinfo Tok-A-Mak: 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-media00:46
→ Curtisbeef joined00:46
Tok-A-Mak i will look into it. thank you00:46
ojacobson Really wish there was a log.defaulttoupstream option for git-log that automatically ran 'git log @{u}..HEAD'00:48
(yes, I can alias that.)00:48
← simesy left00:49
→ discord joined00:49
→ liquidmetal joined00:50
Tok-A-Mak !binary problem00:50
gitinfo Storing binary files in git causes repo balloon, because they do not compress/diff well. In other words, each time you change a file the repo will grow by the size of the file. See !annex for some solutions00:50
→ mmlac-bv joined00:55
→ rurufufuss joined00:55
← deuterium left00:56
← dorkmafia left00:56
← Mars` left00:57
← infynyxx left00:57
→ Mars` joined00:57
→ infynyxx joined00:57
← nicknisi left00:59
→ InvidFlower joined00:59
← devoldmx left01:00
→ nicknisi joined01:00
← normanrichards left01:00
→ dorkmafia joined01:01
← grilix left01:01
→ grilix joined01:02
← infynyxx left01:02
← DonRichie left01:02
← hashpuppy left01:02
→ weiyang joined01:04
← caspingus left01:04
→ Davey joined01:05
→ caspingus joined01:05
→ dhruvasagar joined01:06
← grilix left01:06
← cigarshark left01:07
← R1KRDO1 left01:08
← the_cybersphinx left01:09
← jfoy left01:09
→ mmitchell joined01:10
← PaulCapestany left01:10
← Gamecubic left01:10
← shinnya left01:10
← asedeno left01:10
← Mars` left01:11
→ asedeno joined01:11
→ the_cybersphinx joined01:11
→ Mars` joined01:11
fuzzybunny69y Hey guys I am working on this tutorial and I accidentally committed the completed file in one of my earlier commits. Is there anyway I can go back and change that file in the commit that I committed it in?01:12
→ chuckharmston joined01:12
frogonwheels_ fuzzybunny69y: sure, use git rebase -i01:13
fuzzybunny69y: there are a couple of ways you can do it, but if you can grok the 'edit' option on interactive rebase, that should do the trik.01:13
→ PaulCapestany joined01:14
← bartzy left01:14
← banisterfiend left01:15
fuzzybunny69y frogonwheels_: so if it was the first commit on my branch01:15
would I type git rebase -i ae022f201:15
zjason` → zjason01:15
→ jfoy joined01:16
← Mars` left01:16
→ banisterfiend joined01:16
← scroy left01:16
← v0n left01:18
← caspingus left01:19
→ normanrichards joined01:19
→ hyperair joined01:19
→ caspingus joined01:20
→ joshbetz joined01:20
← joshbetz left01:22
→ Swayz joined01:23
→ DrShoggoth joined01:23
→ dsdeiz joined01:23
→ rjhunter joined01:24
← caspingus left01:24
→ rchavik joined01:25
→ thierryp joined01:26
← richgilbank left01:26
→ ewnd9 joined01:27
← Stalkr^ left01:27
→ hashpuppy joined01:28
→ cesarkaw_ joined01:31
cesarkaw_ Is it possible to add a pattern to the .git/info/exclude of a submodule?01:31
← thierryp left01:31
← NicolasTarzia left01:32
→ HenryG joined01:32
→ caspingus joined01:32
← ewnd9 left01:32
→ ZujkisNx joined01:33
cesarkaw_ Ok, I'm an idiot. The .git file points to the correct .git dir to use01:33
← hahuang61 left01:35
← Algorithm left01:35
→ threesome joined01:35
→ lembacon joined01:37
→ bonhoeffer_ joined01:37
bonhoeffer_ is there a way to prevent a single file from committing (i copied the wrong file over another)01:38
bonhoeffer_ → bonhoffer01:38
→ mizu_no_oto joined01:38
→ minim joined01:38
← jfoy left01:40
bremner don't use commit -a?01:43
← aerlaut left01:43
→ v0n joined01:44
← rurufufuss left01:44
← discord left01:44
→ Sigma[Mtp] joined01:45
← Stygia left01:45
→ Vampire0_ joined01:47
← rchavik left01:49
→ mribeirodantas joined01:49
← Vampire0 left01:50
← harish left01:51
→ iamtakingiteasy joined01:53
← mizu_no_oto left01:54
→ m0ltar joined01:54
→ linelevel joined01:54
← mmlac-bv left01:54
← m0ltar left01:55
→ Gamecubic joined01:55
linelevel Hi, I have a repo that only me and one other person use. I had to do some rebasing on my local branch (I actually just edited commit message using `git rebase -i`), and when I `git push` it throws an error (and `git pull` corrupts my files so I have to `git reset --hard`). I tried to `git push --force`, but it still failed.01:56
What am I doing wrong?01:56
← Malukenho left01:57
→ Wolfspyre joined01:57
linelevel When I use`--force`, the error I get is "remote: error: denying non-fast-forward refs/heads/master (you should pull first)".01:57
← threesome left01:57
→ rurufufuss joined01:58
ojacobson your local branch and the shared branch on the server repo have diverged (see !lol for a visualization)01:58
gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all01:58
→ LuxuryMode joined01:58
→ Goplat joined01:58
ojacobson you will either need to rebase your branch onto the upstream branch, or merge them (which is what 'git pull' does by default; the "corruption" you're complaining about may well be conflict markers to show where the two sets of changes are incompatible).01:58
→ echo-area joined01:59
← normanrichards left01:59
← mribeirodantas left01:59
linelevel ojacobson: They do look like conflict markers, but it inserts them as text into my working tree.02:00
ojacobson That's the normal behaviour for merge conflicts, yes02:00
linelevel Okay, so how should I proceed?02:00
ojacobson You have a few options.02:00
→ neurodro1e joined02:01
ojacobson As I said, you can rebase your local work (assuming it's unpublished) and resolve the conflicts that arise; odds are good the conflicts will be finer grained, since rebase goes one commit at a time02:01
You can merge the upstream changes (or 'git pull', same thing), and resolve the conflicts all at once. man git merge has some documentation on how to mark files as resolved, and !conflict probably has some other advice02:01
gitinfo [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice!02:01
the git-merge manpage is available at http://jk.gs/git-merge.html02:01
ojacobson The resulting history is different in each case; if you don't care which one you get, use git-merge as it's history-preserving in some important ways02:02
→ daito joined02:02
← pigeon left02:02
→ zenpac joined02:02
linelevel `git pull` fails. The first run writes the conflict markers, and subsequent runs fail with "Pull is not possible because you have unmerged files."02:02
ojacobson If you want a sandbox to play around with merge conflicts in, have a look at https://github.com/ojacobson/conflicts :)02:02
zenpac If I revert to an older master (from devel), by doing "git checkout master" the newer devel files are still there too.02:03
← whunt left02:03
ojacobson Yep. After the first one, you need to either abort (with reset, or with git merge --abort) or complete the merge02:03
zenpac How do I clearn those out?02:03
ojacobson (by resolving conflicts and git-commiting the result)02:03
→ pigeon joined02:03
linelevel ojacobson: The "conflicts" seem to be my recent work. So, my origin (bare repo) is flat-- a single branch (master). All I did since my last sync with the origin is this: I made a few changes (which includes a few commits to my local branch), then I used `rebase -i` to reword some old commit messages. That's all I did.02:04
zenpac should I have rm -rf ?02:05
← joshh___ left02:06
← cmpera left02:06
zenpac git clean -f -d ????02:06
ojacobson zenpac: tracked files will be cleaned up automatically.02:07
linelevel ojacobson: In particular, there's nothing in the origin branch that's not in my local branch-- so I really just want to completely overwrite the origin (bare repo) commit history with my local commit history. Put another way, my local `master` branch is exactly how I want the origin `master` branch to look. How can I achieve that?02:07
ojacobson linelevel: your wording suggests you think you did something wrong, or that the other guy did something wrong.02:07
→ tbmiller joined02:07
← LuxuryMode left02:08
ojacobson Neither of these is true. Conflicts are part of working on the same code at the same time. (SCM is not a replacement for communication; it's a tool for structuring communication.)02:08
← Sigma[Mtp] left02:08
← jkremser left02:08
ojacobson You, and the other collaborator, have made changes to the same lines.02:08
→ xjiujiu joined02:08
ojacobson It's up to you two to figure our how to combine your changes into a sensible whole -- that's what conflict resolution is *for*.02:08
linelevel ojacobson: No, the other person hasn't made any commits to the origin since the last time I cloned it. I checked.02:08
This problem started when I changed commit messages with `rebase -i` locally.02:09
ojacobson Then your changes are based exactly on origin/master, and there are no conflicts.02:09
Do you have any other untrue statements you'd like to try?02:09
→ natechan joined02:09
ojacobson Have a look a 'git fetch --all --prune' followed by !lol and you will see the "other" commits02:09
gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all02:09
← eletuchy_ left02:09
linelevel ojacobson: Heh, I just checked though. :P So I don't know what to tell you. I cloned the origin in a new temp directory, looked at `git log` and all of the past 10 commits are from me.02:09
ojacobson Yours will be the chain ending with the one marked "master" (probably); the upstream ones you're conflicting with will be the ones marked 'origin/master' (probably)02:09
it's entirely possible you're conflicting with yoursel02:10
→ eletuchy joined02:10
→ nutron joined02:11
linelevel ojacobson: Okay, bare with me, I'm trying to understand. You keep saying I need to "resolve the conflict" but I'm not sure what that entails. My ideal resolution would be for origin/master to look exactly like my local master does right now -- same exact commit log.02:11
→ AWinterman joined02:11
→ LuxuryMode joined02:12
zenpac ojacobson, they were not, or I have made a big mistake.02:12
← nutron left02:13
→ mribeirodantas joined02:14
→ pr0ton joined02:14
← j416 left02:14
→ j416 joined02:15
→ nutron joined02:15
← boblee left02:15
→ boblee joined02:16
← maestrojed left02:16
← salamanderrake left02:16
→ petey joined02:17
linelevel ojacobson: I ran the !lol command you showed me, and I see, starting from the bottom, all the commits in origin/master, followed by all the commits in my local master. Most of them are the same just with an edited commit message, but I have a few new commits at the top of my local master.02:19
gitinfo ojacobson: A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all02:19
→ salamanderrake joined02:19
ojacobson linelevel: can you paste it somewhere?02:19
linelevel ojacobson: sure.02:19
→ joshh__ joined02:20
← MoSal left02:21
→ Mars` joined02:22
← caspingus left02:22
→ caspingus joined02:22
→ Algorithm joined02:23
linelevel ojacobson: http://pastebin.com/g5NaqS5E02:23
← bonhoffer left02:24
← LuxuryMode left02:24
ojacobson Okay, can you !repro the commands that are failing, and how they fail?02:24
gitinfo Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.02:24
linelevel As you can see, the bottom 5 commits are the same as the next 5 commits, the only differences are minor changes to the commit messages.02:24
← breck7_ left02:24
→ johnnyfuchs joined02:24
linelevel ojacobson: Okay, hold on.02:24
→ caspingu_ joined02:25
← pr0ton left02:25
→ breck7 joined02:25
← caspingus left02:25
← zenpac left02:25
→ Pyrogerg joined02:26
← Mars` left02:26
← MavKen left02:26
← cesarkaw_ left02:26
← neurodro1e left02:26
→ pr0ton joined02:27
← johnnyfuchs left02:27
→ thierryp joined02:27
→ LuxuryMode joined02:29
← breck7 left02:30
linelevel ojacobson: Here you go. I changed the server name and project name (mostly for readability, since they're long), and added some blank lines between terminal commands -- everything else is exactly as output: http://pastebin.com/iRAKA4x202:30
← thierryp left02:31
← BSaboia left02:32
← xjiujiu left02:32
ojacobson Is there a bunch of output missing after 'git pull'?02:32
linelevel ojacobson: Nope, I pasted everything.02:32
ojacobson Huh.02:32
What does 'git branch -vv' tell you about master? (fine to paste here)02:33
← SoftwareMechanic left02:33
linelevel * master 2640e36 [origin/master: ahead 8, behind 5] Compile stylesheets for production.02:33
← fuzzybunny69y left02:33
→ dsdeiz_ joined02:33
← ozette left02:33
MrDHat → MrDHat|offline02:33
← dsdeiz left02:34
ojacobson That's ... inconsistent with the graphlog output you pasted02:34
has the graphlog output changed? 'git pull' runs 'git fetch', which can download new commits + branches02:34
linelevel The graphlog output was before my last `reset --hard` (which you'll see at the top of my last pastebin)02:34
Let me do it again now.02:34
→ mribeiro1antas joined02:35
← valgrind left02:36
← mribeirodantas left02:36
linelevel ojacobson: The graphlog is exactly the same as what I pastebin'd earlier.02:36
ojacobson I'm clearly missing something. How much graph theory can you hold in your head?02:37
offby1 thiiiis much (holds thumb and finger pretty close together)02:37
ojacobson I need to get to bed, but I can tell you my operating theory before I do, if it'd help you diagnose this02:37
linelevel Sure.02:37
I was a math major, so I know some graph theory..02:37
ojacobson So the ahead 8, behind 5 counts refer to two ranges02:37
→ hahuang61 joined02:37
ojacobson origin/master and master identify two nodes in a directed acyclic graph formed by commits02:38
→ xjiujiu joined02:38
linelevel Okay.02:38
ojacobson the "ahead 8" count should be the number of nodes reachable from the commit identified by 'master', minus the number of nodes reachable from the commit identified by 'origin/master'02:38
that's consistent with your graphlog02:38
(lines 1-8 are all only reachable from 'master')02:38
the "behind 5" count is the number of nodes reachable from the commit identified by 'origin/origi... oh02:39
I see the problem02:39
→ gzg joined02:39
linelevel so, in this graph, edges point toward their parent/previous commits?02:39
ojacobson yep02:39
→ jfoy joined02:39
ojacobson !cs !gcs02:39
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 !bottomup02:39
[!concepts] "Git Concepts Simplified" explains the basic structures used by git, which is very helpful for understanding its concepts. http://gitolite.com/gcs/02:39
linelevel Awesome. :)02:39
I've read the latter.02:39
ojacobson The graphlog output is misleading me because normally vertical adjacency means there's an edge from the upper to the lower node02:39
but in your output, there's no edge from line 8 (4203570) to line 9 (ed4be05)02:40
you have two completely disjoint histories, probably because of the way you ran rebas02:40
rebase02:40
→ normanrichards joined02:40
ojacobson IF (and that's a big if) I'm right, you can get to a sensible state "easily" by transplanting the last three commits from your history to the "official" history as a new branch, replacing your local 'master' branch02:41
← mobileblue left02:41
linelevel I see. (I just followed a stackoverflow answer for how to edit old commit messages.)02:41
ojacobson (I'm assuming that 54d0825 and ed4be05 are identical bar rebase damage, here)02:41
linelevel That's correct.02:42
ojacobson OKAY. You're on 'master' (2640e36) right now? 'git rev-parse HEAD' will confirm.02:42
linelevel Confirmed.02:42
← glebihan_ left02:43
← yhager left02:43
ojacobson okay. Are you comfortable with the commit message of 9603904 being forevermore "wrong"? Mistakes are okay, IMO, but it's a project conventions thing, not a thing IRC can answer for you.02:43
I'm guessing that's the one you were trying to edit... :)02:43
→ mobileblue joined02:43
ojacobson eh, you know what02:43
linelevel I mean I'd rather use the newer commit messages obviously, but I'll take what I can get at this point. :)02:44
ojacobson I can give you what you actually asked for, if you're willing to put up with a lecture first and be responsible for fixing the other guy's clone after02:44
linelevel Deal.02:44
ojacobson Lecture: !rewrite02:44
gitinfo [!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://goo.gl/waqum02:44
→ mizu_no_oto joined02:44
ojacobson You've configured origin (the bare repo you're sharing) to reject force pushes.02:44
which is a great safety net against accidental changes of the above kind02:45
here, you're trying to do it intentionally02:45
← ZujkisNx left02:45
→ dsdeiz joined02:45
ojacobson if you're willing to help the other guy graft his local history onto the edited history (there are directions in man git-rebase, under "Recovering from upstream rebase"), then you can override the restriction02:45
gitinfo the git-rebase manpage is available at http://jk.gs/git-rebase.html02:45
← dsdeiz_ left02:45
ojacobson provided you can reconfigure the bare repo briefly02:45
← xjiujiu left02:46
→ fuzzybunny69y joined02:46
ojacobson Specifically, you'll need to configure it to allow force pushes: *in that repo*, 'git config --unset receive.denyNonFastForwards' will turn off the restriction, if it's enforced by git-config02:46
linelevel Sure. My guess is the other [girl]'s working tree hasn't changed recently, so she can probably just reclone and be done with it -- but if I'm wrong I will use the resource you gave me to help her with it.02:47
ojacobson Then 'git push --force origin master' to overwrite the repo's master branch with your fixed 'master' branch02:47
linelevel Okay, doing so now.02:47
ojacobson updates pronouns.02:47
ojacobson Then for gods' sakes set the config back with 'git config receive.denyNonFastForwards true' in the shared repo.02:47
Lest someone edit history accidentally (as you've seen, git doesn't do a lot to stop you from doing that) and overwrite your changes :)02:47
← tbmiller left02:48
← nicknisi left02:48
→ tbmiller joined02:48
linelevel Will the less explicit `git push --force` do the same thing (the `origin` and `master` are implied if not specified, right?)02:49
?*02:49
ojacobson Given the config you're in, yes, but only by accident02:49
git 1.x's default mode is to match your local branches to the remote repo's branches by *name*, and push all matching pairs02:49
← hashpuppy left02:50
ojacobson you only have one local branch (master) and it matches the only remote branch (master).02:50
(That's what that warning on unqualified 'git push' is about, actually.)02:50
I find 'simple' mode easier, where unqualified 'git push' will (a) only push the current local branch, (b) only to a remote branch of the same name, and (c) only if the local branch tracks the remote branch02:50
which stops me from eg. accidentally pushing a feature branch over origin/master, while making it easy to push merges on master without pushing other branches too02:51
'simple' will be the default in git 2.x02:51
linelevel Ah, I see. I was planning to read more about those options.02:51
ojacobson (yes, I recognize the irony behind "simple" mode having three steps' worth of conditions. Welcome to git.)02:51
→ xjiujiu_ joined02:51
EugeneKay !cuz02:51
gitinfo [!why] Why? Because screw you, that's why.02:51
← InvidFlower left02:52
linelevel ojacobson: After doing that, is it safe to `git reset --hard HEAD` again (locally) to get rid of those conflict markers in my working tree?02:52
← tbmiller left02:53
linelevel and then `git pull`?02:53
← pandeiro left02:53
ojacobson you should probably get out of the 'incomplete merge' state, first02:53
I would generally use 'git merge --abort', not 'git reset --hard', though02:53
expressing intent means git will warn you if you're not in the state you expect02:53
← pr0ton left02:53
ojacobson git reset will blithely fuck you up02:53
("hey, you asked for it")02:53
mribeiro1antas → mribeirodantas02:54
← mribeirodantas left02:54
→ mribeirodantas joined02:54
fuzzybunny69y hey guys02:54
linelevel Okay, everything looks good now!02:54
fuzzybunny69y I am a trying to edit a file that is in the very first commit on my branch but I am not sure how to do it would someone be able to guide me through it02:55
I think I have to type git rebase -i --root02:55
← xjiujiu_ left02:56
→ glebihan_ joined02:56
linelevel ojacobson: Thank you! And I know you need to go to sleep, but one thing still confuses me: At what point did my local branch get rid of those first 5 "old" commits? They don't show up in a graphlog anymore, and I'm just wondering which step caused them to go away.02:56
ojacobson linelevel: since you seem reasonably maths-friendly, spend some time with git-show-ref, the graphlog stanza I showed you, 'git fetch', and 'git push' in some scratch repos02:56
← raonyguimaraes left02:56
ojacobson git's not complicated but it pushes a lot of complexity off to the user, and a basic grounding in its theory will help you in the long term02:56
linelevel I'm assuming it wasn't the `push`, so was it the `merge --abort`?02:56
ojacobson linelevel: the rebase did02:57
you created a whole new history, completely independent of the history of origin/master02:57
I assume you ran 'git rebase --root ...options...' somewhere02:57
(--root meaning "include this branch's root commits", which are the ones with no parents)02:57
→ xjiujiu joined02:57
linelevel ojacobson: I did, earlier, but it was still showing up in my graphlog until I just followed your instructions (the `push --force`, then the `merge --abort`).02:58
fuzzybunny69y ojacobson that is all I have so far02:58
linelevel ojacobson: The last time I rebased was before I entered this channel.02:58
ojacobson Anyways, that re-created all of the commits in the history of 'master'02:58
← jfoy left02:58
ojacobson for any commits you decided to touch, that will have changed their identities and the identities of all descendents02:58
the "behind 5" commits were exactly the original branch, before the rebase02:59
the "ahead 8" commits were exactly the re-created 5 commits from editing the first 5, plus your three local commits02:59
← mdiers left03:00
→ harish joined03:00
linelevel ojacobson: Right, I see that. For what it's worth, this was the stackoverflow post that seems to have caused this problenm for me: http://stackoverflow.com/questions/179123/how-do-i-edit-an-incorrect-commit-message-in-git/180085#18008503:00
→ DonRichie joined03:01
ojacobson That's my problem with SO03:01
its karma mechanic encourages "useful" over "thorough"03:02
that is in fact how you rewrite old commit messages03:02
it even mentions not wanting to touch published commits, but it doesn't do much to tell you how to identify them03:02
or why it sucks03:02
linelevel ojacobson: Oh, and I then used `git -i --root` (no other arguments) to edit the initial commit message. (I got that from a different SO question.) Someone said that on git 1.7.12+, that would suffice. That may have been what caused the problem.03:02
ojacobson heh03:03
and since that changes the identity of the root commit, and since commit identity is partially dependent on the identities of parent commits, that changes the identities of *every* commit03:03
← mizu_no_oto left03:03
→ mdiers joined03:03
ojacobson which is how you ended up with a completely duplicated copy of the branch, locally03:03
linelevel ojacobson: http://stackoverflow.com/a/17643063/57531103:03
For future reference, would this answer be a better way to achieve the same thing? http://stackoverflow.com/a/2119656/57531103:05
ojacobson Given that the commits are already part of the "shared" history, the better approach is to leave them alone, wonky commit messages and all03:05
and do better in the future03:05
→ jason237 joined03:06
ojacobson The approach in that post is equivalent to 'git rebase -i --root' with edits on the root commit, really03:06
only more complicated :)03:06
→ juvenal joined03:06
← joki left03:06
← dreiss_ left03:06
→ ej joined03:06
ojacobson git's history-sharing bits don't actually care how you author the history graph you're trying to share; they operate on a graph-theoretic level, not a what commands did I run level03:07
linelevel Fair enough, heh. But, say that there was no public repo yet. Is there a safe way (i.e., without duplicating the entire commit tree ) to edit an initial commit message?03:07
ojacobson It's safe, but needlessly complicated03:07
if it's a purely local history, 'git rebase -i --root' is equivalent and generally easier to use03:07
← banisterfiend left03:07
ojacobson Rebase only duplicated history because there was a remote that had the original history.03:08
← Gamecubic left03:08
← normanrichards left03:08
ojacobson You could just as easily have removed the remote repo (and possibly even deleted and re-created it) to solve the duplication. I picked a somewhat less destructive method, because I try not to lead you into destroying config/hooks/etc03:08
→ normanrichards joined03:09
linelevel Oh, got it. I think that just clicked for me. I thought the remote nodes were duplicated locally as well, but now I'm thinking it's the case that if nodes (vertices on the graph) become disconnected from any branch label (or HEAD), then they effectively cease to exist and won't show up in the graphlog.03:09
Is that right or am I still confused?03:09
ojacobson p much03:10
git operates by keeping labels for objects (usually, commits or tag objects) in the 'refs' system, which is where tags, branches, remote-tracking branches, and a few other things live03:10
→ MavKen joined03:10
ojacobson and then using graph-following to find history03:10
→ joki joined03:10
ojacobson eg. a branch is "just" a label for the most recent commit on that branch; there's a graph of parent commits, which are included implicitly03:11
→ bwreilly joined03:11
← rjhunter left03:11
ojacobson when you rebased refs/heads/master (the 'master' branch) the commits reachable from refs/remotes/origin/master (the 'origin/master' remote branch) are unaffected, but new commits are generated and refs/heads/master is adjusted to point to them03:11
← alvaro_o left03:12
ojacobson *generally* git-rebase is used to re-create commits that are already unique to a single branch, so this isn't an issue03:12
→ maestrojed joined03:12
linelevel So when objects become disconnected (i.e. there's no path in the DAG from any label to that object), then they'll stop showing up in logs and (I'm guessing) eventually be picked up by some sort of garbase collection?03:12
ojacobson (the branch is adjusted to point to the new commits, the old commits become unreferenced, and life goes on)03:12
yep03:13
linelevel garbage*03:13
Okay. Cool. :)03:13
← mribeirodantas left03:13
ojacobson git runs 'git gc --auto' periodically without your help, and you can run it yourself if you feel like it03:13
→ Sigma[Mtp] joined03:13
ojacobson and on that note03:14
← ojacobson left03:14
→ v0latil3 joined03:15
v0latil3 h03:15
← ej left03:15
→ ej joined03:16
← petey left03:17
→ m0ltar joined03:18
→ mizu_no_oto joined03:18
v0latil3 how can i use git with eclipse? do i just start the repo in my eclipse workspace?03:18
→ introom joined03:20
← ej left03:21
← introom left03:21
← fuzzybunny69y left03:21
→ Guest43791 joined03:21
← ianmcorvidae left03:23
→ bskywalker4 joined03:23
→ ianmcorvidae joined03:23
← g0bl1n left03:24
→ schemanic joined03:24
offby1 if I were you I'd just use it on the command line -- I don't trust IDEs that attempt to integrate with git, and I certainly don't trust Eclipse03:25
☁03:25
Guest43791 → diegoviola03:25
→ therealGent joined03:26
v0latil3 Yes.03:27
← Swayz left03:27
v0latil3 but how can i have a git repository, but use eclipse as an editor?03:27
→ thierryp joined03:27
← mizu_no_oto left03:27
→ Sjors joined03:27
thiago just do it03:28
← nardev left03:28
← caspingu_ left03:28
→ mizu_no_oto joined03:29
fairuz v0latil3: they are not related. You can use any editor that you like.03:29
← SamB left03:31
→ simesy joined03:32
← thierryp left03:32
→ r0bby_ joined03:33
→ btree joined03:35
→ nardev joined03:35
→ breck7 joined03:35
→ Swayz joined03:37
v0latil3 Okay thanks03:38
← revyd left03:38
→ revyd joined03:39
← breck7 left03:40
→ sqlnoob joined03:41
→ preflex_ joined03:43
← revyd left03:43
← preflex left03:43
preflex_ → preflex03:44
← the_cybersphinx left03:44
← Sigma[Mtp] left03:45
→ rchavik_ joined03:45
→ nafg joined03:46
→ yhager joined03:46
→ emlun joined03:47
→ Sigma[Mtp] joined03:48
→ fuzzybunny69y joined03:48
→ iamjarvo joined03:48
→ lng joined03:48
← whoami left03:48
emlun so I found the --depth flag on git fetch. Now, does git have a CLI front-end to fetch only a given tree object (and any child trees/blobs)?03:48
(that is, --depth lets me fetch only a single commit if I want. but what if I only want a specific subdirectory of the repo?)03:49
→ whoami joined03:51
← tirengarfio left03:52
→ jaseemabid joined03:53
← echo-area left03:53
← MavKen left03:53
fuzzybunny69y hey guys03:53
← AWinterman left03:53
fuzzybunny69y I just made a new branch and accidentally did a git rebase --root03:54
rchavik_ → rchavik03:54
fuzzybunny69y and I a just wondering what it did03:54
← dhruvasagar left03:54
← Roots47 left03:54
← harish left03:55
→ relipse joined03:55
→ rjhunter joined03:55
relipse why does "git status" take over 2 minutes?03:55
my repo isn't that big i dont think03:55
maybe a lot of directories?03:55
← joshsmith left03:56
← kukks left03:57
← natechan left03:58
emlun fuzzybunny69y: what about it? :)04:00
→ dhruvasagar joined04:00
fuzzybunny69y well it did some stuff but I don't know what it id04:01
did04:01
emlun relipse: uh, that doesn't sound right... how big is your repo and how fast is your computer?04:01
fuzzybunny69y I meant to do a git rebase -i04:01
relipse emlun, its on a shared hosting, probably over 100 other sites on it, and linux04:01
it only happens when i make changes to like a bunch of files > 1004:01
thiago so you have no idea how busy the machine is?04:01
← Norrin left04:02
relipse hmm not exactly no04:02
how do i check?04:02
emlun fuzzybunny69y: unless you've kept working a lot since then you should be able to see what04:02
thiago is that a virtual machine you're logged into?04:02
emlun ... has happened by using git reflog04:02
relipse no its shared hosting04:02
→ joshbetz joined04:02
relipse so we each have our own home dir04:02
thiago both things can happen04:02
fuzzybunny69y emlun well the reabse did something but I am not sure what it is04:02
did04:02
thiago run top and see if the machine is busy04:02
bskywalker4 shared hosting sometimes runs really slow, we used to run on a generic shared hosting site and randomly a simple script execution would go from 10 seconds to taking 5 minutes to run04:03
fuzzybunny69y my log seems to be the same04:03
emlun fuzzybunny69y: rebase --root is allowed to create a new "root" commit (that is, one with no parents)04:03
fuzzybunny69y oh well it didn't even let me do anything04:04
← japaric left04:04
emlun fuzzybunny69y: if you do a git log --graph --all you should be able to tell if it did or not04:04
or even git log --graph --all --oneline04:04
← mizu_no_oto left04:04
→ cads joined04:04
emlun fuzzybunny69y: if your command was only "git rebase --root" then it may very well have done exactly nothing :)04:04
fuzzybunny69y oh yeah that is what it was04:05
it did something though04:05
because it said it was doing some stuff04:05
but my history looks the same04:05
→ Norrin joined04:05
→ ej joined04:05
fuzzybunny69y I have been trying to modify a file in my first commit04:05
emlun ah04:05
fuzzybunny69y but I am not sure what I have to do with git rebase -i04:05
emlun if you've committed the changes you want to make, then you can do git rebase -i --root04:06
→ angela joined04:06
emlun and then rearrange the commits in the order you want, and/or squash them04:06
angela hi, quick question, I have a git rebase in progress04:06
→ tirengarfio joined04:06
angela a file has a conflict I fixed it04:06
← bwreilly left04:06
angela I added it using git add FILE04:06
→ Roots47 joined04:06
← rchavik left04:06
angela but when I run git rebase --continue04:07
fuzzybunny69y emlun it brought up a vim window04:07
← Roots47 left04:07
angela I get No changes - did you forget to use 'git add'?04:07
emlun fuzzybunny69y: vim :D04:07
fuzzybunny69y so do I just choose the commit I want to edit04:07
emlun angela: git status04:07
offby1 angela: you can do "git rebase --skip"04:07
angela git status shows '# rebase in progress; onto 00d24a4'04:07
fuzzybunny69y and replace pick with edit04:07
emlun fuzzybunny69y: have you used vim before?04:07
offby1 this happens when your fix matches one of the exiting versions, I think04:07
fuzzybunny69y emlun yeah I love it04:07
angela offby1: but wouldn't that be not applying the changes?04:07
offby1 angela: I think they already have been or will be. Try It And See™.04:08
→ blindingdawn joined04:08
emlun fuzzybunny69y: :D yeah, just rearrange them in the order you want. replacing "pick" with "edit" will make rebase pause and let you monkey around with the repo before continuing04:08
→ Roots47 joined04:09
fuzzybunny69y emlun ok I made the change I want to make04:09
→ japaric joined04:09
fuzzybunny69y now I am not sure what to do04:09
should I do a amend or a continue04:09
emlun fuzzybunny69y: just commit it in any way you like04:09
relipse thiago: http://pastebin.com/HXP656AP04:10
emlun then git rebase --continue will continue where it left off and base the remaining changes on whatever HEAD you're currently at04:10
fuzzybunny69y hmmm04:10
well I don't want the changes to show up in the latest commit04:10
emlun then an amend might be suitable04:10
fuzzybunny69y I want it to be as if they happened in the first commit04:10
emlun then again04:10
← lng left04:11
← diegoviola left04:11
emlun you could always just do it in its own commit, then git rebase --continue, then after that rebase again and squash the first two commits into one04:11
← ianmcorvidae left04:11
thiago relipse: high load average, 30% I/O,04:11
relipse: that's quite high04:11
emlun pick whichever you fancy :)04:11
← iamjarvo left04:11
← maestrojed left04:11
relipse well then i guess i thank God he even let me install git on it04:11
emlun fuzzybunny69y: oh, and if you04:12
gah04:12
relipse thiago: any suggestions from here?04:12
→ iamjarvo joined04:12
thiago relipse: yeah. Use a machine with lower use.04:13
emlun fuzzybunny69y: oh, and the commit you create now will immediately follow the current HEAD, which would in this case be your first commit. just as new commits always do.04:13
unless you commit --amend04:13
then the new commit will replace your first commit :)04:13
relipse thiago: what?04:13
fuzzybunny69y hmmm04:13
thiago relipse: use a different server04:13
fuzzybunny69y emlun so should I git add my change first04:13
relipse thiago, well ideallyy i shouldn't be developing on the same machine as my web server04:14
emlun fuzzybunny69y: yeah, otherwise you won't be able to commit it :P04:14
fuzzybunny69y or can I just do a git commit --amend or a git continue04:14
← m0viefreak left04:14
relipse ideally i should develop on my own machine and mimic the development and then rsync or something04:14
fuzzybunny69y ok I added it04:14
so if I do a git commit --ammend04:14
→ ianmcorvidae joined04:15
relipse you can change the commit message Fullma04:15
fuzzybunny69y,04:15
fuzzybunny69y it should replace my root commit with the change i just made04:15
?04:15
emlun fuzzybunny69y: there's no magic involved here - when git rebase hits an "edit" entry in the todo list it applies that commit, then exits and puts you back in your plain old shell04:15
fuzzybunny69y: yeah04:15
fuzzybunny69y hmmm04:15
← minim left04:15
fuzzybunny69y so what would it do if I did a git rebase --continue04:15
emlun then you would start a new rebase command with the todo list prepopulated with what was left to do when it exited last time04:16
fuzzybunny69y like if I git added this file04:16
but then just did git rebase --continue04:16
→ akiress joined04:17
← iamjarvo left04:17
emlun fuzzybunny69y: I'm actually not sure exactly what would happen. Perhaps your current index changes would be commited in their own commit, perhaps the first item in the rebase todo list would be applied before commiting04:17
but at any rate04:17
it would replay the changes saved in the rebase todo list on whatever state your repo is currently in04:18
fuzzybunny69y hmm ill try doing a git commit --amend04:18
→ germanoak joined04:18
emlun fuzzybunny69y: do you know how git models a repo?04:18
fuzzybunny69y no04:19
emlun the Directed Acyclic Graphs and stuff04:19
ok04:19
fuzzybunny69y hmmm04:19
this is why git always confuses me04:19
← ivan\ left04:19
emlun git may seem very complex at first, but it's actually exceedingly simple once you grasp how it works on a lower level04:19
fuzzybunny69y it brought up a window with the commit message of my latest commit04:19
why would it be asking me to change the commit message of my latest commit when I want to change my root commit04:20
→ hahuang65 joined04:20
← akiress left04:20
emlun huh?04:20
← totte left04:20
→ SoftwareMechanic joined04:21
emlun that doesn't sound right04:21
fuzzybunny69y yeah04:21
git is so confusing04:21
→ MoSal joined04:21
emlun try emptying the buffer (gg 100dd usually does the trick for me :P)04:21
fuzzybunny69y ecspecially rebasing04:21
emlun and then doing a git log04:21
fuzzybunny69y it never does what I think it is going to do04:21
yeah I just aborted it04:21
and the whole rebase04:22
→ kpreid joined04:22
← hahuang61 left04:22
fuzzybunny69y it seems like no matter what option I choose it always tries to edit my latest commit04:22
not my root one04:22
← RitchieFlick left04:23
→ irqq joined04:23
→ Guest3820 joined04:23
emlun hm04:23
← Sigma[Mtp] left04:23
emlun are you in a REBASING state now?04:24
→ maestrojed joined04:24
→ rchavik joined04:24
fuzzybunny69y no I just aborted it04:24
emlun ok04:24
fuzzybunny69y I have no idea what it was doing04:24
→ akiress joined04:24
emlun what was the rebase command you used to start it all?04:24
→ ivan\ joined04:25
→ axl_ joined04:25
emmajane → [emmajane]04:25
fuzzybunny69y ok04:25
← maestrojed left04:25
→ subhojit777 joined04:25
fuzzybunny69y so I just checked out my master branch04:25
and erased my test branch04:25
and now I am going to create another test branch04:25
← germanoak left04:25
fuzzybunny69y and check that out04:25
← johnf left04:25
fuzzybunny69y and then I will run git rebase -i --root04:25
emlun mm04:26
fuzzybunny69y and it brings up vim with a list of commits I can edit04:26
emlun mm04:26
← delta6 left04:26
juvenal → juvenal_away04:27
fuzzybunny69y and so I will change pick to edit04:27
for the commit I want to change04:27
→ johnf joined04:27
emlun mm04:27
which is the root commit?04:27
← Guest3820 left04:27
→ germanoak joined04:27
← axl_ left04:27
fuzzybunny69y it is the very first commit on the branch04:28
← MoSal left04:28
← el3ctron left04:28
fuzzybunny69y the initial commit04:28
→ thierryp joined04:28
emlun mm04:28
fuzzybunny69y ok so I changed that to edit04:28
then made my change04:28
→ Guest joined04:29
→ MoSal joined04:29
fuzzybunny69y and now i am going to do a git commit --amend04:29
emlun what does git show show now?04:29
← angela left04:29
fuzzybunny69y it just has changes to be committed with my change04:29
and I can do a git rebase continue04:29
or a git commit ammend04:29
← mmitchell left04:30
emlun in this case I think git rebase --continue will do04:30
but either should work04:30
they should be equivalent, even04:30
→ mmitchell joined04:30
→ Tommy[D]_ joined04:30
→ xxpor joined04:31
fuzzybunny69y hmmm04:31
i think it might of worked04:31
→ harish joined04:31
→ totte joined04:31
← thierryp left04:32
emlun fuzzybunny69y: Git from the bottom up is a great introduction to how git works conceptually: http://newartisans.com/2008/04/git-from-the-bottom-up/04:32
→ valgrind joined04:33
emlun fuzzybunny69y: having this understanding will make everything about git suddenly just make sense04:33
fuzzybunny69y I wonder why it worked this time04:33
→ h0bbit joined04:33
emlun fuzzybunny69y: :)04:33
fuzzybunny69y maybe I picked the wrong commit or something04:33
emlun maybe04:33
← Tommy[D] left04:34
← mmitchell left04:34
fuzzybunny69y thanks emlun04:35
← harish left04:35
fuzzybunny69y you are great04:35
emlun np :)04:36
fuzzybunny69y I will read that book04:36
emlun I really recommend you do04:36
it worked wonders for me04:36
git really is dead simple when you can see past the CLI commands and into what will be done to the DAG04:36
complexity through simplicity - such a beautiful thing04:36
frogonwheels_ fuzzybunny69y: !simple04:37
gitinfo fuzzybunny69y: At its heart git is made up of many concepts that are individually simple. Getting the whole picture right is often tricky, and it is usually about breaking up the complex concept into its simple, individual parts and grokking those. Both !bottomup and !cs will help with that.04:37
emlun simple rules, complex consequences :D04:37
frogonwheels_ emlun: ^^ you might like that trigger :)04:37
fuzzybunny69y I guess it is because I never understand the help04:37
emlun frogonwheels_: that04:37
fuzzybunny69y and what all the terminology they use is04:38
emlun frogonwheels_: yup, that's exactly my point ^^04:38
frogonwheels_ fuzzybunny69y: ah yeah. Getting the terminology right with its very specific meaning under git is linked closely with understanding how it works.04:39
fuzzybunny69y !cs04:39
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 !bottomup04:39
→ ewnd9 joined04:39
← LuxuryMode left04:39
fuzzybunny69y !concepts04:39
gitinfo "Git Concepts Simplified" explains the basic structures used by git, which is very helpful for understanding its concepts. http://gitolite.com/gcs/04:39
frogonwheels_ fuzzybunny69y: a nice tip, you can /msg gitinfo !cs to get gitinfo to reply to you personally04:39
gitinfo fuzzybunny69y: "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 !bottomup04:39
fuzzybunny69y oh lol04:40
→ lite_ joined04:40
frogonwheels_ fuzzybunny69y: one or two is ok, - a whole series starts to get annoying (unless you're directing it at somebody else)04:41
fuzzybunny69y oh ok04:42
well04:42
now that I made that change on my test branch04:42
← akurilin left04:42
fuzzybunny69y what is the best way to get it into my master branch04:43
or should I just rebase my master branch04:43
with the same procedure04:43
→ tbmiller joined04:43
← jaimef left04:44
→ jaimef joined04:44
→ gverri joined04:44
r0bby_ → robbyoconnor04:44
← rchavik left04:45
← Joshi3s` left04:45
→ minim joined04:46
← lite_ left04:46
emlun fuzzybunny69y: no04:46
→ lite_ joined04:47
emlun since you did rebase --root the test branch won't share any ancestors with the master branch04:47
so what you probably want to do is make sure the test branch is what you want, and then rename the test branch to master, overwriting the old master04:47
← Swayz left04:48
emlun of course, this is usually a very bad idea to do if you've pushed the master branch to a remote repository04:48
since then things will get messy if someone has cloned that and then tries to merge your new master04:48
← rjhunter left04:49
→ harish joined04:49
emlun the scenario is perfectly solvable, of course, but I'd say the gain (making the root commit prettier) isn't worth the trouble :P04:49
→ rjhunter joined04:50
← lite_ left04:51
→ cmpera joined04:52
← tpollard left04:52
← juvenal_away left04:54
← ej left04:54
→ juvenal joined04:54
← jaseemabid left04:55
← rjhunter left04:55
→ eaj joined04:55
← eaj left04:55
← germanoak left04:55
→ ej joined04:55
→ angela joined04:55
→ jaseemabid joined04:56
← jaseemabid left04:56
← fuzzybunny69y left04:57
→ jaseemabid joined04:57
angela hi I have a branch to work on a feature04:57
← johnf left04:57
angela and I pulled on the master branch04:57
but now when I try to rebase, I get conflicts, which I have fixed04:58
→ Swayz joined04:58
angela but after fixing the conflict and adding them using git add, I'm still unable to do a rebase continue, or skip04:58
or even checkout to another branch04:58
→ hoschi joined04:58
angela I get this when I try to checkout upload/catalog/view/theme/hitour/static/scripts/detail/main.js: needs merge error: you need to resolve your current index first04:58
→ johnf joined04:59
emlun angela: sounds like there are unstaged changes04:59
(aka red stuff in git status)04:59
→ jfoy joined05:00
→ lite_ joined05:00
← sig left05:01
→ Mars` joined05:02
→ sig joined05:03
← gverri left05:03
← fairuz left05:04
→ AWinterman joined05:04
ej → eaj05:06
← jfoy left05:07
← Swayz left05:07
← lite_ left05:07
→ lite_ joined05:08
→ fuzzybunny69y joined05:08
→ jfoy joined05:08
→ thierryp joined05:09
← PaulCapestany left05:10
← irqq left05:10
← AWinterman left05:10
fuzzybunny69y sorry guys I had a cock roach issue to take care of over here05:10
← daito left05:11
→ daito joined05:12
fuzzybunny69y and i got the boot05:12
← gzg left05:12
← lite_ left05:13
fuzzybunny69y ok05:13
I renamed my test branch to master05:13
and renamed my master branch05:13
→ epicowl joined05:14
raijin → raininja05:14
→ gverri joined05:15
→ cff joined05:16
→ Swayz joined05:17
→ beautifulmind joined05:18
→ jceb_ joined05:18
← beautifulmind left05:18
→ JustinMac joined05:18
← JustinMac left05:18
→ beautifulmind joined05:19
→ JustinMac joined05:19
← Tobbi left05:20
→ TomyLobo joined05:20
← yhager left05:20
→ gzg joined05:21
→ varlog joined05:21
← scorphus left05:22
← brisbin left05:22
← jpalmer_ left05:22
← tbmiller left05:22
→ jpalmer joined05:22
← normanrichards left05:23
→ tbmiller joined05:23
→ Tobbi joined05:23
← johnf left05:23
→ marthinal joined05:24
← beautifulmind left05:24
← akiva left05:24
→ johnf joined05:24
→ scorphus joined05:25
→ h12o joined05:25
← fuzzybunny69y left05:26
← vkz left05:26
← DrShoggoth left05:27
← tbmiller left05:27
→ beautifulmind joined05:27
← beautifulmind left05:27
← therealGent left05:27
← JustinMac left05:29
→ JustinMac joined05:30
juvenal → juvenal_away05:30
← yacks left05:30
← Algorithm left05:31
← jfoy left05:32
← gverri left05:33
→ glosoli joined05:33
← hoschi left05:33
→ gverri joined05:33
→ hoschi joined05:34
→ nickbelhomme joined05:34
← h12o left05:36
→ Xethron joined05:36
→ matanya joined05:36
→ h12o joined05:36
← eaj left05:37
→ fuzzybunny69y joined05:37
→ jfran joined05:38
→ ThomasLocke joined05:38
← johnf left05:38
→ jumperswitch joined05:40
→ mmitchell joined05:41
← Swayz left05:41
← jfran left05:42
← weiyang left05:43
← zjason left05:44
→ zjason joined05:44
→ lite_ joined05:44
← chepre` left05:45
← mmitchell left05:45
← lite_ left05:46
→ lite_ joined05:46
← Xethron left05:47
← nicedice left05:47
→ ciampix joined05:47
→ leeN joined05:47
← dorkmafia left05:47
← lite_ left05:47
→ lite_ joined05:47
→ Xethron joined05:48
← jdav_gone left05:51
← jaseemabid left05:52
→ beautifulmind joined05:52
emlun fuzzybunny69y: did you have some more questions? :)05:54
fuzzybunny69y ummm05:54
→ rath joined05:54
→ therealGent joined05:54
fuzzybunny69y I just renamed my branch lol05:54
I always have more questions05:55
I just don't know what they are yet lol05:55
← ewnd9 left05:55
← joshbetz left05:55
→ ezekielnoob joined05:55
← dhruvasagar left05:55
← Xethron left05:55
→ dhruvasagar joined05:57
→ Xethron joined05:57
← rue_XIV left05:57
→ rue joined05:58
→ pantsman joined05:58
← pantsman left05:58
→ pantsman joined05:58
→ reenignEesreveR joined05:58
← rue left05:58
→ rue joined05:59
→ codeFiend joined05:59
emlun fuzzybunny69y: ^^06:01
→ DrShoggoth joined06:01
emlun well as long as you keep references to (branches at) commits you want to keep, your data is always safe06:01
even if you don't, it is usually possible to salvage lost commits, but it may be more or less complicated depending on the circumstances06:02
← jceb_ left06:02
→ jceb joined06:03
← Xethron left06:03
jceb → Guest5160606:03
lite_ so..ive got a bunch fo changes in a feature branch that may need to get stashed for now. I would however like to have them in a local branch06:06
so: git stash - git checkout -b new_branch, git pop ?06:06
→ mishok13 joined06:06
← cmpera left06:06
← jumperswitch left06:07
← valgrind left06:07
← setmeaway left06:09
→ papegaaij joined06:10
→ fl0rian_ joined06:12
← Roots47 left06:14
→ dsdeiz_ joined06:14
← dsdeiz left06:14
← Guest51606 left06:14
→ cbreak-work joined06:15
← varlog left06:15
← pantsman left06:17
← Nicholas_Rishel left06:17
→ Roots47 joined06:17
→ saimaz joined06:17
→ SHyx0rmZ joined06:18
→ Raging_Hog joined06:21
← Goplat left06:23
← saimaz left06:23
→ saimaz joined06:23
→ Search12 joined06:23
→ fRAiLtY- joined06:23
← chuckharmston left06:24
emlun lite_: or just `git checkout -b stash_branch && git checkout other_branch` if I understand you correctly06:25
→ boo__ joined06:25
← boo__ left06:25
emlun that'll commit your staged changes on branch "stash_branch" and then check out "other_branch"06:25
← thierryp left06:27
→ dsdeiz joined06:27
→ thierryp joined06:27
← saimaz left06:28
← harish left06:28
→ thierryp_ joined06:28
← dsdeiz_ left06:29
lite_ cool, thanks06:29
← epicowl left06:29
→ jdav_gone joined06:30
← bskywalker4 left06:30
→ saimaz joined06:30
← glebihan_ left06:32
← thierryp left06:32
→ glebihan_ joined06:33
← SoftwareMechanic left06:33
→ reskp_ joined06:34
← oops left06:34
→ ewnd9 joined06:34
← therealGent left06:38
→ jargon joined06:38
→ echo-area joined06:38
→ dreiss_ joined06:39
→ netiul joined06:39
→ reskp__ joined06:40
→ rchavik joined06:43
← reskp_ left06:43
← reskp__ left06:45
← TomyLobo left06:45
← thiago left06:46
→ jceb joined06:47
jceb → Guest2589606:47
→ Xethron joined06:48
→ _nova joined06:48
← dreiss_ left06:50
← tjj left06:50
← hahuang65 left06:53
← dhruvasagar left06:54
→ oops joined06:54
← saimaz left06:55
→ saimaz joined06:56
→ dhruvasagar joined06:56
→ kbart joined06:57
→ JohnBat26 joined06:57
← Wolfspyre left06:58
→ Wolfspyre joined07:00
→ ThomasCle joined07:00
← saimaz left07:01
→ jarr0dsz joined07:01
→ saimaz joined07:01
→ sr78ger joined07:03
← Search12 left07:04
← igalic left07:05
→ txwikinger2 joined07:05
← txwikinger2 left07:05
→ txwikinger2 joined07:05
→ jefferai_gone joined07:05
← BlaDe^ left07:05
→ zizzfizzix joined07:05
← zizzfizzix left07:05
→ zizzfizzix joined07:05
← jefferai left07:05
← txwikinger left07:05
← zorzar left07:05
← solution left07:05
← BtbN left07:05
← TheBrayn left07:05
← ezekielnoob left07:05
→ solution joined07:06
→ tokkee joined07:06
→ mika joined07:06
← Luke-Jr left07:06
→ Kuukunen joined07:06
→ BlaDe^ joined07:06
→ igalic joined07:06
→ zorzar joined07:06
→ TheBrayn joined07:06
→ Will| joined07:07
→ Luke-Jr joined07:07
← SporkWitch left07:07
→ harobed_ joined07:07
← sr78ger left07:08
→ manitu joined07:08
→ BtbN joined07:08
← freestyl3r left07:09
→ reenignEesreveR- joined07:09
→ SporkWitch joined07:09
→ threesome joined07:10
→ sr78ger joined07:10
→ Search12 joined07:11
→ RitchieFlick joined07:12
← reenignEesreveR left07:12
← saimaz left07:12
→ saimaz joined07:12
→ ezekielnoob joined07:13
→ germanoak joined07:14
← defaultro left07:14
→ chuckharmston joined07:16
→ MafooUK joined07:17
← saimaz left07:17
→ DanielHollands joined07:17
← ezekielnoob left07:18
→ starlays joined07:18
← LimeBlast left07:19
← covi left07:19
starlays hello to all, how can I merge the content from a git fetch?07:19
→ harish joined07:20
← ewnd9 left07:20
→ lb1a joined07:21
← dhruvasagar left07:21
starlays refs are stored in .git/FETCH_HEAD07:21
fount it07:21
→ kewah joined07:23
→ brisbin joined07:23
→ gua_ould joined07:23
→ SergeyPenkin joined07:24
→ dhruvasagar joined07:24
frogonwheels_ starlays: that's not the normal way to do it - what exactly were you doing it?07:24
→ ezekielnoob joined07:24
← chuckharmston left07:24
frogonwheels_ starlays: oh, and you can just refer to that as FETCH_HEAD as a reference... but again, not normal to use it.07:25
frogonwheels_ → frogonwheels07:25
frogonwheels starlays: !xy07:25
gitinfo starlays: Woah, slow down for a bit. Are you sure that you need to jump to that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...07:25
← fl0rian_ left07:25
→ dreiss_ joined07:25
→ Balliad joined07:26
← jarr0dsz left07:27
→ saimaz joined07:27
→ Misan joined07:28
→ devoldmx joined07:28
← ezekielnoob left07:29
← Dreadnaught left07:30
→ GeertV joined07:30
→ h_kon joined07:30
← saimaz left07:32
← relipse left07:32
→ saimaz joined07:32
← dhruvasagar left07:33
← vigs left07:33
← gverri left07:33
→ ewnd9 joined07:35
← girishr left07:36
← beautifulmind left07:36
→ Naoe-Kanno joined07:36
→ dhruvasagar joined07:36
→ ezekielnoob joined07:36
← saimaz left07:37
→ cmpera joined07:37
← SHyx0rmZ left07:38
← hoschi left07:38
→ jarr0dsz joined07:38
→ intellix joined07:38
starlays frogonwheels: I've done a remote fetch with git fetch then done git diff and now i want to merge the fetched content in the current active branch07:40
← simesy left07:40
→ tatsuya_o joined07:41
→ simesy joined07:41
→ GT4066 joined07:41
frogonwheels starlays: When you do a normal fetch, it will fetch into the remote tracking branch.. which is something like remotes/origin/master07:41
starlays: if you've set things up properly, then you can git merge @{u}07:41
starlays: man git-parse-revs07:42
starlays: man git-parse-rev07:42
starlays: man git-rev-parse07:42
← ewnd9 left07:42
gitinfo starlays: the git-rev-parse manpage is available at http://jk.gs/git-rev-parse.html07:42
starlays ok07:42
frogonwheels bleh07:42
the last one :)07:42
← minim left07:42
starlays :)07:42
frogonwheels: ok, i'll read and see what is it about07:42
→ deuterium joined07:42
starlays frogonwheels: thank you07:42
← igalic left07:42
→ igalic joined07:42
frogonwheels starlays: no probs.07:43
→ untaken joined07:43
← JustinMac left07:44
→ Olea|2 joined07:44
→ hahuang65 joined07:44
← Naoe-Kanno left07:44
→ JustinMac joined07:45
untaken if I have a branch (B) and it has been rolled out on some servers, and I had roll the servers back to a older tag (A). If I then delete the B with 'git branch -D <branchName>; git push origin --delete <branchName>' would the servers work with this as it had B locally checked out from before.07:45
← cmpera left07:45
← simesy left07:46
← Tok-A-Mak left07:47
→ afuentes joined07:48
← reenignEesreveR- left07:49
→ jaseemabid joined07:49
← RitchieFlick left07:50
← mishok13 left07:51
→ hoschi joined07:52
← netiul left07:52
arand untaken: If non-fast-forwards are allowed, and if the repo is non-bare (which it sounds like, and that's a bit odd) if it allows updating the currently checked out branch...07:52
starlays frogonwheels: git branch -vv shows me that i have no @{u} for the current branch, i need to see how can i set remote reference for a branch that doesn't have one07:53
← angela left07:54
← dreiss_ left07:56
starlays hmm, git branch -u <remote>07:56
→ netiul joined07:57
← netiul left07:57
→ netiul joined07:57
→ cenuij joined07:57
→ reenignEesreveR joined07:57
→ alorence joined07:57
starlays frogonwheels: yey, problem solved, thank you once more, you saved my day07:57
← harish left07:58
← khushildep left07:58
→ samuel02 joined07:59
untaken arand: okay, so its down to the git configuration07:59
→ khushildep joined08:00
← ciampix left08:01
→ tobeplugged joined08:03
← Mars` left08:03
→ Mars` joined08:03
← Robert_L left08:04
← Roots47 left08:05
→ Robert_L joined08:05
← KillYourTV left08:08
→ Octalot joined08:08
← reenignEesreveR left08:08
← Mars` left08:08
→ reenignEesreveR joined08:09
→ saimaz joined08:13
→ avolz joined08:15
← saimaz left08:15
frogonwheels starlays: :) well done08:16
→ ciampix joined08:16
→ reskp_ joined08:16
→ dfanjul joined08:17
← cads left08:17
→ chuckharmston joined08:17
← reskp_ left08:17
→ g0bl1n joined08:17
→ reskp_ joined08:18
← rat_thing left08:20
→ mishok13 joined08:20
← h_kon left08:20
← germanoak left08:20
→ h_kon joined08:20
← chuckharmston left08:22
→ iderik joined08:22
← reskp_ left08:22
← Georg3 left08:23
→ juril joined08:23
← tchan left08:24
→ harish joined08:25
→ howlymowly joined08:25
← h_kon left08:25
← GodEater left08:27
→ chepre joined08:27
← reenignEesreveR left08:28
→ buibex joined08:29
→ m7w joined08:29
← zivester left08:30
← akitada left08:30
← beawesomeinstead left08:30
← raghunayyar left08:31
← scalability-junk left08:31
← jmcantrell left08:31
← _rgn left08:32
← Roch_ left08:32
← jab416171 left08:32
← arand left08:32
← deuterium left08:32
→ FabTG_ joined08:33
← Norrin left08:34
← wcpan left08:34
→ wcpan joined08:34
→ kini joined08:34
→ h_kon joined08:35
← Misan left08:35
← FabTG left08:36
→ Norrin joined08:36
→ Mars` joined08:36
← gzg left08:37
→ curvedmark joined08:39
← frgtn left08:39
→ frgtn joined08:40
→ reskp joined08:42
→ victorbjelkholm joined08:43
← frgtn left08:45
← victorbjelkholm left08:45
→ FrodoTheHobbit joined08:47
← JSharpe left08:47
→ zarubin joined08:48
→ asQuirreL joined08:50
→ simesy joined08:51
→ Stygia joined08:52
← kermit left08:53
→ nuclearkitten joined08:54
← echo-area left08:54
← JustinMac left08:55
→ arand joined08:55
→ ewnd9 joined08:55
← asQuirreL left08:56
← simesy left08:56
← rchavik left08:58
→ deuterium joined08:58
→ rchavik joined08:58
← rchavik left08:59
→ mateu joined08:59
→ nnyk_ joined08:59
→ ChrisAnn joined09:00
← glosoli left09:02
← zorzar left09:03
← xjiujiu left09:04
← leeN left09:04
← jaseemabid left09:05
→ xjiujiu joined09:05
→ irqq joined09:05
arand Is it possible to clone a git repo and store the result without any compression of objects? Intention is to adapt to an external non-git incremental backup, which would presumably be more effective if the underlying contents was not already compressed.09:06
moritz arand: I think you shouldn't bother09:07
→ ZujkisNx joined09:08
moritz arand: whether it's git or the backup that compresses the files is mostly irrelevant09:08
canton7 objects are *always* compressed09:08
moritz and note that if you don't git-repack very often, the old packfiles don't change much09:08
→ Misan joined09:09
_ikke_ When you clone, the objects already come packed09:09
→ tommyziegler joined09:09
→ raonyguimaraes joined09:11
→ PaddyPatPat joined09:11
← reskp left09:12
→ frgtn joined09:12
→ reskp_ joined09:12
arand Yeah, I was thinking you could make git unpack all packfiles after receive, though I guess you'd have to manually hack something like that.09:13
moritz arand: simply don't bother.09:14
→ szx joined09:14
_ikke_ Git is much better at packing the objects than the backup sollution can09:15
← Mars` left09:15
→ Mars` joined09:16
→ reskp___ joined09:17
← reskp_ left09:17
→ chuckharmston joined09:17
PaddyPatPat Hello, I have the "fatal: CRLF would be replaced by LF" issue for a silverlight .xap file that contains multiple CRLF files (and perhaps a binary?), "plupupload.silverlight.xap" I'm trying to get them to LF. I'm on OS X 10.8. Suggestions?09:18
_ikke_ !crlf09:18
gitinfo to fix problems with line endings on different platforms, check out http://line-endings.gitrecipes.de/09:18
canton7 also read core.safecrlf in man git-config09:19
gitinfo the git-config manpage is available at http://jk.gs/git-config.html09:19
→ rendar joined09:21
← Mars` left09:21
PaddyPatPat Thank you _ikke_ and canton709:21
← chuckharmston left09:22
→ cilly joined09:22
→ marhaban joined09:23
→ PeterBooker joined09:24
→ Ozch joined09:25
arand _ikke_: In a single repo at a single point in time, yes. With for example multiple repos, there's no guarantee that similar objects will have similar packfiles, right?09:25
(Though I guess at that point I should be looking at leveraging .git/objects/info/alternates as well...)09:26
→ reenignEesreveR joined09:26
← fRAiLtY- left09:27
→ dsdeiz_ joined09:27
→ chuckharmston joined09:27
← dsdeiz left09:29
← khushildep left09:30
→ jkremser joined09:30
← jkremser left09:30
→ jkremser joined09:30
→ the_cybersphinx joined09:30
→ rat_thing joined09:31
_ikke_ To get any advantage, you'd probably have to uncompress every object09:32
← chuckharmston left09:33
→ VictorCL joined09:33
→ Chryyz joined09:36
← rat_thing left09:36
→ banisterfiend joined09:37
→ stayarrr joined09:40
← dyng left09:40
→ dyng joined09:41
→ beautifulmind joined09:41
→ caseymcg joined09:41
← mdiers left09:42
→ rchavik joined09:43
← Davey left09:43
← xjiujiu left09:43
→ dyng_ joined09:43
→ bartzy joined09:43
← h_kon left09:44
→ xjiujiu joined09:44
← dyng left09:44
← harobed_ left09:48
← nnyk_ left09:48
← cilly left09:48
→ reenignEesreveR- joined09:50
← reenignEesreveR- left09:51
← reenignEesreveR left09:52
→ b4n1 joined09:55
→ fRAiLtY- joined09:55
→ jumperswitch joined09:57
→ saimaz joined09:57
← sword left09:57
← Baluse left09:58
← nafg left09:58
→ Baluse joined09:58
← Baluse left09:59
→ nnyk_ joined09:59
← maroloccio left10:00
← Chryyz left10:01
← m0ltar left10:02
→ bonhoeffer_ joined10:02
→ mklappstuhl joined10:03
→ jaseemabid joined10:03
→ maroloccio joined10:03
← saimaz left10:03
→ stamina joined10:03
← farciarz84 left10:04
← mitz_ left10:04
→ m0ltar joined10:05
→ farciarz84 joined10:05
← ezekielnoob left10:05
← Guest25896 left10:06
→ germanoak joined10:06
→ ezekielnoob joined10:06
→ rednaks joined10:08
→ jceb joined10:08
← jceb left10:08
→ jceb joined10:09
jceb → Guest7557910:09
→ mikecmpbll joined10:10
← ezekielnoob left10:11
← nickbelhomme left10:11
← ZujkisNx left10:11
martisj what is the difference between clone and fork?10:13
is fork a term from github?10:13
iveqy martisj: that's a github term yes10:13
selckin they kinda redefined existing usage of the word fork10:13
iveqy martisj: and a fork on github have some references to the original repo10:13
← curvedmark left10:15
→ rat_thing joined10:16
→ ezekielnoob joined10:18
← tommyziegler left10:19
← lite_ left10:19
→ lite_ joined10:20
martisj iveqy: so if i wanted to submit a pull request to a repo on github i could either clone it or fork it, make some changes and submit a pull request?10:20
← glebihan_ left10:20
iveqy martisj: yes, but you can only use the github pull-request interface if you do a fork10:20
usually you fork a project and then clone your fork10:20
→ glebihan_ joined10:21
→ lorenx_ joined10:22
← reskp___ left10:22
→ Roch_ joined10:22
← ping__ left10:22
→ reskp joined10:22
→ _goTAN joined10:22
lorenx_ excuse me, why after checking out a previous commit git sometimes shows me some untracked files/dirs? i would expect a clean enviroment...10:23
← ezekielnoob left10:23
← dcpc007 left10:23
← harish left10:23
→ caraconan joined10:24
→ nickbelhomme joined10:24
← lite_ left10:24
→ lite_ joined10:24
caraconan Hi there. It is possible to revert local changes done in a file which is not even added for commit? Thanks10:24
→ brick joined10:24
→ dcpc007 joined10:25
→ h_kon joined10:25
iveqy lorenx_: that would mean that git would delete untracked stuff and since they are untracked you wouldn't get them back10:25
caraconan: man git checkout10:25
gitinfo caraconan: the git-checkout manpage is available at http://jk.gs/git-checkout.html10:25
← glebihan_ left10:26
iveqy lorenx_: so that would be stupid. However if you still want to, there's man git clean10:26
gitinfo lorenx_: the git-clean manpage is available at http://jk.gs/git-clean.html10:26
→ spidererrol joined10:26
→ ezekielnoob joined10:26
M-ou-se can i get git svn to name all remotes for svn branches like 'svn/mysvnbranch' instead of just 'svnbranch'? that would avoid a lot of confusion when using local branches with those names10:26
caraconan The situation is that I modified let's say 5 files, but just 1 should be reverted to its initial status10:26
→ glebihan_ joined10:26
caraconan And then, I will do the commit (with the 4 files added)10:26
→ dbuch joined10:26
caraconan How to proceed with this 1 file?10:26
lorenx_ iveqy: yeah, i expect that git deletes those files cause they are not in the commit i'm checking out. there are not lost actually, cause they'll be checked out with a later commit, what am i missing?10:27
← reskp left10:27
← HicksD left10:27
lorenx_ iveqy: i know "git clean -fd" but i would expect git checkout to do that10:27
← spidererrol left10:27
→ ph88 joined10:27
ph88 how can i let git diff print 2 spaces for a tab ?10:27
lorenx_ i mean, i'm checking aout e previos situation where those files where not there10:28
iveqy caraconan: just checkout that file you want10:28
lorenx_ s/aout/out/10:28
→ raghunayyar joined10:28
→ jmcantrell joined10:28
iveqy lorenx_: no, git doesn't care about untracked files10:28
caraconan iveqy: ok, I'll try this10:28
→ scalability-junk joined10:28
← Fullma left10:28
caraconan I'm coming from svn and checkout still sounds somehow extrange for me10:29
iveqy lorenx_: I guess it's a matter of taste10:29
→ d8bhatta joined10:29
lorenx_ iveqy: wait, they are untracked cause they didn't exists in that previous commit, git should know this10:29
iveqy caraconan: but in this case the flow is the same as in svn? Checkout the specific file you want and not more10:29
d8bhatta Hello everyone10:29
← h0bbit left10:29
caraconan svn revert I would say, maybe I'm wrong...10:29
iveqy lorenx_: yes that git knows, however have you edit them so that you've untracked changes in them?10:30
→ ismlages joined10:30
→ akitada joined10:30
iveqy caraconan: I guess that would also work with svn, but for git revert is something else10:30
→ h0bbit joined10:30
caraconan Anyway I understand (now) the checkout command, and it worked as expected10:31
ivegy: many thanks10:31
→ xeqution joined10:31
→ beawesomeinstead joined10:31
iveqy caraconan: you're welcome10:31
→ _rgn joined10:32
lorenx_ iveqy: no. let's say i have commit A. than i create a new file "newfile" and commit it in commit B. if i then do "git checkout A", i would expect no newfile and not an untracked one...10:32
d8bhatta I have release branch for my live site...develop branch for dev or staging site...multiple people works on develop branch and commit there.....i can merge develop branch to release branch but the problem merging i.e. git pull also pulls all those commit made on dev branch...I really need to pull only few commits in release branch10:32
← caraconan left10:32
← b4n1 left10:32
d8bhatta can you please let me know how can i do this without conflicts and all10:32
→ jab416171 joined10:32
d8bhatta there is cherry pick but not sure how can i use this ideally10:32
iveqy d8bhatta: yes you would need to do a cherrypick for that10:33
however your workflow seems broken and a cherrypick isn't great10:33
d8bhatta: !flow10:33
gitinfo d8bhatta: [!gitflow] The description of the gitflow branch workflow model is at http://nvie.com/posts/a-successful-git-branching-model/ while a tool to help implement this workflow is at https://github.com/petervanderdoes/gitflow See http://sethrobertson.github.com/GitBestPractices/#workflow for other workflow suggestions/references10:33
iveqy lorenx_: that's correct, and that's the behavior of git10:34
lorenx_ so the actual command to restore a previous clean situation is: "git commit <commit-sha> && git clean -fd"?10:34
iveqy: ^10:34
iveqy lorenx_: no10:35
lorenx_ iveqy: what do you mean10:35
iveqy lorenx_: that's not the correct way to use git commit10:35
see man git commit10:35
gitinfo the git-commit manpage is available at http://jk.gs/git-commit.html10:35
iveqy lorenx_: Here's how I understand your problem and git does "clean tracked files": http://pastebin.com/YkrQnqYz10:36
lorenx_ excuse me, i mean "git checkout <commit-sha> && git clean -fd" :P10:36
← ezekielnoob left10:37
d8bhatta iveqy: what i am thinking to do is: merge particular commit into release branch from dev branch and push it through my local sys. And just use git pull origin release in my live site10:37
any thoguth?10:37
→ jalvarez joined10:37
d8bhatta thought10:37
← Misan left10:38
lorenx_ iveqy: thanks for the paste but what if you do a git status? if file "b" untracked?10:38
→ reskp_ joined10:39
← jalvarez left10:39
iveqy d8bhatta: don't merge on a live site... see !deploy10:39
gitinfo d8bhatta: Git is not a deployment tool, but you can build one around it(in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/the-list-and-irc/deploy.html10:39
iveqy lorenx_: as you can see from the ls the file b isn't there10:39
so no, it won't show up as utracked since it doesn't exists10:39
d8bhatta iveqy: I am not merging it in live site..i am merging the stuffs in my local...and push it to repo...from live site i am just using git pull10:40
iveqy d8bhatta: !pull10:40
gitinfo d8bhatta: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/10:40
iveqy you pull, hence you merge10:40
thevishy I have 2 branches .... which is forked from the same master ... and the base is same ... I do the same change and commit it ... will the commit hash be the same ditto ?10:41
→ raf joined10:41
iveqy thevishy: git doesn't have something called a fork10:42
moritz thevishy: if absolutely everytihng in those two commits is identical, including the timestamps, then yet10:42
iveqy thevishy: but no, they won't since they'll have different timestamps10:42
thevishy time stamp comes into picture , thats intersting10:42
moritz s/yet/yes/10:42
thevishy thank you moritz iveqy intersiting10:43
interesting*10:43
iveqy thevishy: not at all, the commit-id is the sha1 of all the data in a commit and a commit is containing a timestamp10:43
→ Droolio joined10:43
thevishy however the file name hash will be same10:43
because its same content10:43
← BiggFREE left10:43
iveqy thevishy: yes, but a commit is not a file, but a delta10:43
moritz no, a snapshot, not a delta :-)10:44
← lembacon left10:44
thevishy correct , thank you ... I will read up more ....10:44
→ lpapp joined10:44
iveqy moritz: are you sure about that? I'm pretty sure git doesn't store snapshots but deltas10:44
lpapp hi, is this a proper place for asking about gitweb10:45
or this channel is purely for git itself.10:45
iveqy lpapp: you can try10:45
lpapp iveqy: fantastic10:45
thevishy ipapp if the contents has relevance to git , everyone would be happy to help10:45
lpapp http://gerrit.googlecode.com/svn/documentation/2.1/config-gerrit.html#commentlink -> do you know if this is supported by gitweb ?10:45
thevishy git is a service oriented thing these days :)10:45
→ ezekielnoob joined10:45
iveqy lpapp: no idea, see the gitweb manual?10:46
lpapp basically a dedicated match to jump to the relevant bugtracker entry.10:46
iveqy lpapp: I don't think so...10:46
lpapp gitphp supports it fwiw10:46
and gitweb seems more widely used, so I would assume so.l10:46
so.*10:46
→ yacks joined10:47
iveqy lpapp: sounds like a fair assumptions10:47
s/s//10:47
→ sgambino joined10:48
→ dyng joined10:49
← lorenx_ left10:49
← dyng_ left10:50
← GT4066 left10:50
← tobeplugged left10:50
→ curvedmark joined10:50
← curvedmark left10:51
← _nova left10:52
dwmw2_gone → dwmw210:52
← h_kon left10:52
→ tobeplugged joined10:52
→ _nova joined10:55
← mklappstuhl left10:55
→ h_kon_ joined10:56
→ mdiers joined11:00
catsup i have a strange problem. i can't push because "non-fast-forward" but it seems like it should be a fast-forward: the remote side's head is 52c0814503096f8301f33b6e9177e9548de24d8b and i have 52c0814503096f8301f33b6e9177e9548de24d8b as a parent in my log11:00
git log|grep "$(git show remotes/joe/master|head -1)"11:01
commit 52c0814503096f8301f33b6e9177e9548de24d8b11:01
← grahamsavage__ left11:01
→ grahamsavage__ joined11:01
catsup i did do some rebasing stuff, but i don't see how it could cause this11:01
→ kukks joined11:03
← paulweb515 left11:03
→ Misan joined11:03
→ mizu_no_oto joined11:03
← Vampire0_ left11:03
← Olea|2 left11:03
← matanya left11:04
← donmichelangelo left11:04
← mdiers left11:05
→ rnsanchez joined11:05
→ donmichelangelo joined11:05
→ mdiers joined11:06
catsup oh wow11:06
YoungFrog perhaps you have a merge commit and the other parent is older ? not sure what git does in this case11:06
catsup even after i do a pull, it won't let me push11:06
→ m3ga joined11:07
YoungFrog capitulates11:07
catsup er, oh, nevermind that11:07
← kewah left11:07
m3ga "git branch" says i'm on branch 'prepare'. "git commit ..." says "Unknown branch 'prepare'." why?11:07
catsup i didn't merge with the pull, it said i needed to specify a branch. then it created a merge commit when i did11:07
← marhaban left11:07
→ marhaban joined11:08
← lite_ left11:08
catsup omg11:08
← marhaban left11:08
catsup there was acommit there, it wasn't a fastforward11:08
→ lite_ joined11:08
→ cmpera joined11:08
catsup it wasn't in remotes/joe/master after i did a fetch11:08
← DanielHollands left11:08
catsup probably the same branch name issue and i didn't read the output :/11:09
sorry11:09
→ LimeBlast joined11:09
→ dimitar joined11:09
→ Vampire0_ joined11:09
← kaendfinger[BNC] left11:09
→ spaceonline joined11:10
dimitar → boichev11:11
→ harish joined11:11
boichev How to clone a repo inside an existing folder and then to add all the existing files in that folder to the project I cloned ?11:12
→ kaendfinger joined11:12
iveqy boichev: man git clone11:12
gitinfo boichev: the git-clone manpage is available at http://jk.gs/git-clone.html11:12
← lite_ left11:13
← rednaks left11:13
iveqy m3ga: !repro11:13
gitinfo m3ga: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.11:13
→ jbnicolai joined11:13
→ asQuirreL joined11:14
jbnicolai Hi guys, I'm looking for a way to create a branch, remotely. So something like $ git create-remote-branch origin <branchname>.11:14
→ kolarski joined11:15
boichev iveqy: I saw the man but there it says "Cloning into an existing directory is only allowed if the directory is empty."11:15
jbnicolai This to avoid having to clone the repository locally, create a branch, push the new branch to origin.11:15
iveqy jbnicolai: you can't. But you can push a branch to a new remote branch, see man git push11:15
gitinfo jbnicolai: the git-push manpage is available at http://jk.gs/git-push.html11:15
iveqy boichev: then you've your answer =)11:15
jbnicolai iveqy: Yeah, I know I can push a branch, but I'm looking for a way to push a new branch without having to clone first.11:15
iveqy jbnicolai: you can't11:16
jbnicolai Could I do it with a hook on the server somehow?11:16
iveqy jbnicolai: or of course you can, just use git on the server11:16
boichev iveqy: what will happen if I clone it someware and move the .git folder inside my folder, then add the files ?11:16
moritz iveqy: (re snapshots vs. diffs) internally git stores commits as diffs, but from a user-facing point of view, it often makes much more sense to think of them as snapshots11:16
_ikke_ jbnicolai: where should that new branch be based on?11:16
iveqy jbnicolai: how would you execute a hook if you don't have a repo?11:16
← subhojit777 left11:16
moritz iveqy: for example 'git checkout $commit' and 'git diff $c1 $c2' both make much more sense if you think of commits as snapshots11:16
iveqy boichev: git might be confused but it should work. But why would you do that? :S11:17
← reskp_ left11:17
→ reskp joined11:17
iveqy moritz: yes of course, but are the sha1 calculated as the snapshot or as the diff?11:17
_ikke_ jbnicolai: why do you want to create a new branch on the remote?11:17
boichev iveqy: because I have a 2 Gb project that I want ot add to an empty GIT repo11:17
moritz iveqy: both, in some sense11:18
m3ga iveqy: https://gist.github.com/erikd/704010811:18
→ Olea|2 joined11:18
_ikke_ moritzz, git stores snaoshoits11:18
snaphots11:18
← jbnicolai left11:18
_ikke_ moritz: Each commit contains the full state of the tree11:19
← reskp left11:19
moritz ... compressed by doing diffs11:19
_ikke_ delta's11:19
moritz right11:19
iveqy boichev: then just cd in to the path and to a git init11:19
_ikke_ Those are different things11:19
iveqy moritz: both? (and btw. it's the deltas and not the diffs that's tracked).11:19
boichev iveqy: and then git remote add origin ?11:19
_ikke_ the delta's are only an implementation detail in the pack format11:20
git itself works with snapshots, and git calculates diffs on the fly11:20
iveqy boichev: correct11:20
boichev thanks11:20
← raonyguimaraes left11:20
→ matanya joined11:21
iveqy m3ga: what's with all the strange characters?11:21
→ reskp__ joined11:21
→ lite_ joined11:21
moritz iveqy: well, by having the parent commit's sha1 inside the data that is hashed, in some sense it uniquely identifies both the diff and the snapshot11:21
m3ga thats the normal colour highlighting git uses11:21
← hyperair left11:21
iveqy _ikke_: so the commit sha1 is of the snapshot even if it's stored as deltas?11:21
← kolarski left11:22
_ikke_ Yes, gits object model is about snapshots11:22
← glebihan_ left11:22
→ kolarski joined11:22
→ magu|q joined11:22
_ikke_ the delta is only a storage format, and is a low level implementation detail11:22
→ glebihan_ joined11:22
iveqy _ikke_: yes, but it would make sense to not needing the complete worktree to calculate a sha111:23
m3ga: very strange, I don't know actually11:24
_ikke_ if you create a new repo, the objects are stored separately11:24
→ mklappstuhl joined11:24
_ikke_ no delta's involved11:24
← bgerber left11:24
← bonhoeffer_ left11:24
→ dbr joined11:24
_ikke_ iveqy: the sha *is* based on the whole tree11:24
← magu|m left11:24
iveqy _ikke_: yes but if've a worktree of 2 GB and do a commit I need a new sha1 calculated. That's expensive, if I instead just calc the sha1 of the deltas and the parents sha1, I will have the whole tree inside that sha1 anyway11:25
← sgambino left11:25
m3ga iveqy: damn, upstream project installed a pre-commit hook.11:26
j416 iveqy: you don't have a worktree of 2 GB11:26
iveqy: tree ≠ blob11:26
← matanya left11:27
iveqy m3ga: :)11:27
_ikke_: never mind, you would get all files sha1 already from the objects...11:27
→ glosoli joined11:27
kolarski if you have 3 directories & change 1 file in 1 directory the sha of the other directories will be the same & will be ignored11:27
iveqy j416: yeah, I realised that, thanks11:27
m3ga iveqy: do i look like i'm smiling?11:27
j416 m3ga: :)11:28
→ valtido joined11:28
iveqy kolarski: git doesn't track directories, git doesn't track files, git does track changes. So yes, only your changes will be commited11:28
m3ga: commit hooks are funny =)11:28
→ Stalkr_ joined11:29
← WakiMiko left11:29
_ikke_ git tracks content to be more precise11:29
It's object model is based on contents of files11:29
iveqy m3ga: or wait, you won't be affected if your remote has a pre-commit hook. It won't run for you since you commit locally11:30
_ikke_ two files with the same contents are stored as one object11:30
→ matanya joined11:30
→ menace joined11:31
← rat_thing left11:31
→ JoshuaA joined11:31
← Ozch left11:32
JoshuaA Hi guys, is there a way to trigger a manual/custom git hook (or basically any script) remotely through git?11:32
_ikke_ no11:32
← _Vi left11:32
→ HicksD joined11:32
→ tommyziegler joined11:32
JoshuaA Something like $ git trigger-hook origin my-custom-hook11:32
_ikke_ the git protocol is only about transfering history11:32
→ subhojit777 joined11:33
← lhunath left11:33
JoshuaA Fair answer, was hoping to be able to piggyback on the already established git credentials11:33
← matanya left11:33
← subhojit777 left11:33
← AAA_awright left11:33
← grahamsavage__ left11:33
→ bgerber joined11:33
→ subhojit777 joined11:34
← JoshuaA left11:34
← subhojit777 left11:34
j416 JoshuaA: you could install the hook on the remote and trigger it on some git event, of course. But perhaps you already found that not sufficient.11:34
_ikke_ What protocol are you using?11:35
→ subhojit777 joined11:35
← subhojit777 left11:35
← jaeckel left11:36
→ subhojit777 joined11:36
← kolarski left11:37
← gua_ould left11:38
→ kolarski joined11:38
→ khushildep joined11:38
→ mmitchell joined11:38
→ kewah joined11:38
→ AAA_awright joined11:39
→ gua_ould joined11:40
← dbuch left11:40
← mmitchell left11:41
← mizu_no_oto left11:41
→ mmitchell joined11:42
→ Ozch joined11:44
→ sgambino joined11:44
← threesome left11:45
→ grahamsavage__ joined11:45
→ grilix joined11:46
← chepre left11:47
← markovh left11:47
← mmitchell left11:47
← JohnBat26 left11:47
→ markovh joined11:47
→ jaeckel joined11:47
← jaeckel left11:47
→ jaeckel joined11:47
→ matanya joined11:48
→ threesome joined11:48
← kcj left11:50
→ InvidFlower joined11:50
→ m0viefreak joined11:51
← tobeplugged left11:53
← mobileblue left11:53
→ mobileblue joined11:54
→ svector joined11:56
← tirengarfio left11:56
→ babilen joined11:56
→ Entomo joined11:56
→ hashpuppy joined11:56
← menace left11:57
← Entomo left11:57
→ psoo joined11:57
→ saurabh joined11:57
→ tirengarfio joined11:58
→ gzg joined11:58
→ mizu_no_oto joined11:58
← _goTAN left12:01
→ pfFredd joined12:01
→ harobed_ joined12:01
→ chepre joined12:01
← juvenal_away left12:03
← stayarrr left12:04
→ juvenal joined12:04
← banisterfiend left12:05
← asQuirreL left12:05
→ milojaden joined12:08
→ banisterfiend joined12:08
← MafooUK left12:08
→ rednaks joined12:09
← mikecmpbll left12:10
← robbyoconnor left12:10
→ tobeplugged joined12:10
← milojaden left12:11
← ewnd9 left12:11
→ normanrichards joined12:13
→ WakiMiko joined12:13
← rednaks left12:15
← SicoWork left12:16
→ lattera_ joined12:16
← lattera left12:16
→ starlays1 joined12:16
← perlstein left12:17
← schemanic left12:17
← nioncode left12:17
→ kalleth_ joined12:17
← h34d3r left12:17
← ikanobori left12:17
→ SicoWork joined12:18
→ breakingthings joined12:18
→ averell joined12:18
→ ikanobori joined12:18
← starlays left12:18
→ nioncode joined12:18
→ AlexanderS joined12:18
→ Seveas joined12:18
→ btree_ joined12:18
→ Fuxx joined12:18
→ kenyon joined12:18
→ SegFaultAX joined12:19
→ SleePy joined12:19
← Stalkr_ left12:19
→ Stalkr_ joined12:19
→ MafooUK joined12:19
→ h34d3r joined12:20
→ decotii joined12:20
← btree left12:20
→ freeed joined12:21
→ mmitchell joined12:21
juvenal → juvenal_away12:22
← matanya left12:22
← nnyk_ left12:22
← chepre left12:24
juvenal_away → juvenal12:25
← sgambino left12:27
→ keller joined12:28
← Orbitrix left12:30
→ nnyk_ joined12:31
→ momomomomo joined12:32
→ paulweb515 joined12:32
← grilix left12:32
← keller left12:33
→ grilix joined12:33
→ IrishPrime joined12:33
← dyng left12:33
→ keller joined12:34
→ dyng joined12:34
← Stalkr_ left12:34
→ Stalkr_ joined12:35
→ dyng_ joined12:35
← dyng left12:35
← keller left12:35
← MoSal left12:36
→ keller joined12:36
→ wiherek joined12:36
wiherek hi, i have a repo and I added another repo as submodule inside a directory12:37
← lpapp left12:37
wiherek but that submodule is not pushed into my main repo. When i do a new clone, the directory is empty12:37
← grilix left12:37
iveqy wiherek: !submodule12:37
gitinfo wiherek: git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://book.git-scm.com/5_submodules.html12:37
iveqy wiherek: seems correct, you need to read up about submodules, see the link12:38
← dyng_ left12:38
← paulweb515 left12:38
→ dyng joined12:38
→ sword joined12:38
→ xjiujiu_ joined12:38
→ _Vi joined12:39
← sqlnoob left12:40
→ milojaden joined12:40
→ Orbitrix joined12:40
← Orbitrix left12:40
→ TheBrainWork joined12:40
→ grilix joined12:40
← mizu_no_oto left12:40
→ Orbitrix joined12:40
← dlan left12:41
← xjiujiu left12:41
← boichev left12:41
→ rat_thing joined12:42
→ dlan joined12:42
→ Guest1779 joined12:42
→ luxgeek joined12:42
← dyng left12:43
→ chepre joined12:43
← flaguy48 left12:44
→ intripoon joined12:45
→ m7w_ joined12:45
← intripoon_ left12:45
→ simesy joined12:45
← gzg left12:46
← m7w left12:46
→ R1KRDO joined12:46
m7w_ → m7w12:46
← milojaden left12:46
→ rat_thin_ joined12:46
← rat_thing left12:46
→ schemanic joined12:47
← normanrichards left12:47
→ el3ctron_ joined12:47
→ normanrichards joined12:48
→ shuaib joined12:49
→ tchan joined12:49
← Stalkr_ left12:49
→ Streward joined12:49
← nnyk_ left12:49
→ paulweb515 joined12:50
→ Stalkr_ joined12:50
→ nnyk_ joined12:51
← schemanic left12:52
→ ewnd9 joined12:54
← pfFredd left12:54
→ rjhunter joined12:54
m3ga i'm using 'git send-email', how do i get the v2 part of the '[PATCH v2 X/Y]' subject line?12:54
→ flaguy48 joined12:55
→ btree joined12:55
m3ga i've read the man page12:55
→ Sigma[Mtp] joined12:56
← btree_ left12:57
→ jerius joined12:58
← normanrichards left12:59
[emmajane] → emmajane13:00
→ kcj joined13:00
→ dsdeiz joined13:00
← dsdeiz_ left13:01
→ iamjarvo joined13:01
← dlan left13:01
→ dlan joined13:02
← R1KRDO left13:02
← Wolfspyre left13:02
→ psynor joined13:04
→ R1KRDO joined13:04
← _ikke_ left13:05
→ dsdeiz_ joined13:06
← dsdeiz left13:07
→ stephCoue joined13:08
→ gregsmith joined13:08
→ ozette joined13:08
→ Burer joined13:09
← Burer left13:09
→ Burer joined13:09
→ sgambino joined13:09
→ angelsl joined13:09
← Burer left13:10
← breakingthings left13:10
→ Burer joined13:11
← Burer left13:11
→ Burer joined13:11
→ stayarrr joined13:11
← tobeplugged left13:11
← starlays1 left13:13
← Raging_Hog left13:13
← tommyziegler left13:15
← MafooUK left13:15
→ normanrichards joined13:17
← reskp__ left13:17
→ reskp_ joined13:18
← cotton left13:19
← iamjarvo left13:20
→ dserodio joined13:20
← dsdeiz_ left13:20
← Burer left13:20
→ cotton joined13:20
← keller left13:20
→ iamjarvo joined13:21
→ devoldmx3 joined13:21
→ Burer joined13:21
→ waaadim joined13:21
waaadim hello everybody13:22
gitinfo waaadim: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.13:22
→ reskp__ joined13:22
→ rideh joined13:22
← Burer left13:22
→ zorzar joined13:22
← devoldmx left13:22
→ tobeplugged joined13:22
→ iamjarvo_ joined13:22
→ breakingthings joined13:22
→ Burer joined13:23
← Burer left13:23
→ Burer joined13:23
← reskp_ left13:23
waaadim so I have 2 branches, and I want to create a third one from this 213:23
← _nova left13:23
← xjiujiu_ left13:23
← gdrc left13:23
waaadim what I did is:13:23
→ xjiujiu joined13:23
waaadim checkout branch113:23
git checkout -b branch313:24
git rebase branch213:24
← psoo left13:24
← kolarski left13:24
→ kolarski joined13:24
→ Zarthus joined13:24
waaadim and the problem is that now some changes were made on branch113:24
what is the best way to add those changes to my branch3 (branch1 + branch2)13:25
← alorence left13:25
j416 waaadim: rebase or merge13:25
waaadim: possibly cherry-pick13:25
waaadim: read about those three and decide13:25
← iamjarvo left13:26
← jaseemabid left13:26
← juril left13:26
← mklappstuhl left13:26
← iamjarvo_ left13:26
→ iamjarvo joined13:27
waaadim j416: ok I'll take a look at cherry pick. because rebase gave me some conflict in a few files witch I didn't change in branch113:27
← lite_ left13:27
j416 waaadim: good luck. come back if you have more questions.13:28
MrDHat|offline → MrDHat13:28
→ lite_ joined13:28
→ jaseemabid joined13:28
← jaseemabid left13:28
→ Roots47 joined13:28
→ asQuirreL joined13:28
← Roots47 left13:29
→ jaseemabid joined13:29
→ Roots47 joined13:29
→ tommyziegler joined13:30
→ mizu_no_oto joined13:31
→ MrV joined13:31
Roots47 i just did a "git pull branchname" while i was on the wrong branch and now i have merge conflicts. How can I undo the pull and just keep my local changes from prior to the pull?13:31
← subhojit777 left13:31
lattera_ → lattera13:32
← lite_ left13:32
← ezekielnoob left13:32
→ herdingcat joined13:32
← kcj left13:33
→ kcj_ joined13:33
arand Roots47: git merge --abort ?13:33
→ Dougie187 joined13:34
← waaadim left13:34
← reskp__ left13:34
→ reskp_ joined13:35
→ devoldmx joined13:35
→ apipkin joined13:35
← v0n left13:35
← reskp_ left13:35
→ mikecmpbll joined13:35
→ tham joined13:36
→ DoniRS joined13:36
→ user258467 joined13:37
← h_kon_ left13:38
← devoldmx3 left13:38
→ Chryyz joined13:38
← iamjarvo left13:38
→ iamjarvo joined13:39
→ echo-area joined13:39
→ ezekielnoob joined13:40
→ tbmiller joined13:41
→ aChatir joined13:41
← kcj_ left13:42
→ MafooUK joined13:42
← Guest75579 left13:42
→ kcj joined13:42
← iamjarvo left13:43
← freeed left13:44
← apipkin left13:44
→ freeed joined13:44
← tommyziegler left13:44
→ adac joined13:45
← devoldmx left13:45
→ richgilbank joined13:45
→ Baluse joined13:45
→ devoldmx joined13:47
adac I was wondering: If I have a local repository and I wanted to merge it wit a second repository from github: how to do that? My local repository should stay the default remote13:48
→ MyTallest joined13:49
→ reskp_ joined13:49
→ valgrind joined13:50
← MyTallest left13:50
→ phile joined13:50
lov git remote add secondRepo git://path/to/repo; git fetch secondRepo; git merge secondRepo/master;13:50
this may or may not be what you actually want, however.13:50
do you actually want a !submodule ?13:50
gitinfo git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://book.git-scm.com/5_submodules.html13:50
← InvidFlower left13:50
→ Gooder joined13:52
→ freestyl3r joined13:52
← momomomomo left13:52
← marthinal left13:54
→ AaronMT joined13:54
→ _nova joined13:56
adac lov, actually i have my local git repo. I could add a new branch and call it "original" and then clone the master of the github repository into my branch "orignal"13:56
donÄt know if taht is possible13:56
← thierryp_ left13:56
→ thierryp joined13:56
lov er...13:56
adac the remote of the branch then would point to the github repo13:56
→ joshbetz joined13:56
lov ok, so I'm not exactly sure what you're trying to do.13:56
Are you just trying to add a remote?13:57
← devoldmx left13:57
lov if so, just use git remote add and add the github repo as a remote. Create a local branch as you see fit.13:57
(don't forget to fetch)13:57
→ devoldmx joined13:57
→ v0n joined13:58
adac lov my "original" branch should have the github remote13:58
my other branches should have the remote of my server13:59
→ mklappstuhl joined13:59
lov ok13:59
← japaric left13:59
← s0ber left13:59
← glosoli left13:59
lov git remote add github git://path/to/github/repo13:59
git fetch github13:59
git branch original github/master13:59
→ s0ber joined13:59
lov actually, you should probably do git branch --track original github/master14:00
← donmichelangelo left14:00
→ devoldmx3 joined14:00
lov (--track is the default unless you changed something)14:00
← tham left14:00
adac lov, that sounds good14:00
:)14:01
→ juvenal_ joined14:01
→ donmichelangelo joined14:01
← herdingcat left14:01
→ tham joined14:01
← juvenal_ left14:02
← devoldmx left14:02
← Search12 left14:02
← juvenal left14:02
→ juvenal joined14:03
← el3ctron_ left14:03
→ japaric joined14:05
→ el3ctron_ joined14:06
→ devoldmx joined14:07
← kbart left14:07
← normanrichards left14:08
→ marthinal joined14:09
← MrV left14:09
← g0bl1n left14:10
← netiul left14:10
→ epicowl joined14:10
← devoldmx3 left14:11
← ezekielnoob left14:11
← simesy left14:12
→ simesy joined14:13
→ thannoy_ joined14:13
→ LuxuryMode joined14:13
← jaseemabid left14:13
← Guest left14:14
← chepre left14:14
→ perlstein joined14:15
← h0bbit left14:15
→ herdingcat joined14:17
← marccollin left14:17
→ hyperair joined14:17
→ keryell joined14:17
← simesy left14:17
→ ezekielnoob joined14:18
← mdiers left14:19
← rainbyte left14:19
→ cagedwisdom_ joined14:20
→ _ikke_ joined14:22
← cagedwisdom left14:22
→ fostertheweb joined14:23
← fostertheweb left14:23
← japaric left14:23
← epicowl left14:26
→ Wolfspyre joined14:26
← kewah left14:26
← R1KRDO left14:26
→ Malukenho joined14:27
← raf left14:27
→ notcharon joined14:28
→ tplaner joined14:29
← nickbelhomme left14:29
→ kewah joined14:29
→ mastro joined14:31
→ rainbyte joined14:32
→ PaulCapestany joined14:32
→ iamjarvo joined14:34
← xjiujiu left14:35
← Streward left14:35
→ chuckharmston joined14:35
← xobb left14:35
← iamjarvo left14:36
→ xobb joined14:36
← modInfo3 left14:36
→ iamjarvo joined14:36
→ bwreilly joined14:37
← notcharon left14:37
← bwreilly left14:37
← daito left14:37
→ marhinal_friday joined14:37
← thierryp left14:38
→ daito joined14:38
← marthinal left14:38
← marhinal_friday left14:39
→ acharlieh joined14:39
→ Stalkr^ joined14:39
← richgilbank left14:40
→ gzg joined14:40
→ davividal joined14:42
← beautifulmind left14:42
← Baluse left14:42
← jefferai_gone left14:43
→ Baluse joined14:43
← daito left14:43
← ezekielnoob left14:43
→ iamjarvo_ joined14:43
→ marhinal_friday_ joined14:44
→ benkay joined14:44
→ richgilbank joined14:45
benkay if I commit a thing14:45
→ matteo joined14:45
benkay checkout an old commit14:45
and then reset HEAD14:45
how fucked am I?14:45
← gregsmith left14:45
← blindingdawn left14:45
dr_lepper benkay: it won't do anything at all14:46
← iamjarvo left14:47
marhinal_friday_ → marhinal_friday14:47
→ SHyx0rmZ joined14:47
marhinal_friday → marthinal14:47
benkay mkay. i was briefly concerned because git log did not show commits forward in time from my detached head state14:47
but then discovered git reflog14:48
valtido has anyone had an experience with Dev, Test, Production/live environment work flow?14:48
← tobeplugged left14:48
dr_lepper git log never shows commits forward in time14:48
← kewah left14:48
→ Jackneill joined14:48
→ thiago joined14:48
jeek i have dev/qc/test/prod experience, with an additional ops environment where the ops team gets to break and fix things14:48
← wiherek left14:49
← PaddyPatPat left14:49
→ devoldmx3 joined14:49
valtido is there something I can read on how to achieve that ?14:49
→ Swayz joined14:49
→ normanrichards joined14:50
benkay dr_lepper: i realize this now :)14:50
valtido coz my team and I are a little all over the place, with GIT workflow, using a single (master branch) but I know it should be different14:50
@jeek14:50
benkay valtido: http://nvie.com/posts/a-successful-git-branching-model/14:51
lov valtido: note that it's important to understand that git is not a deployment mechanism, just a source control mechanism.14:51
← devoldmx left14:51
→ ezekielnoob joined14:51
← richgilbank left14:51
lov having dev, qa, and prod endpoints is orthogonal to your source control choice14:51
benkay on the other hand, i like deploying on merges to master14:52
← saurabh left14:53
← anli left14:54
→ richgilbank joined14:56
valtido @benkay I have read that, but it's unclear if there is a person controlling the multiple branches or a number of people, what happened with us once when we had multiple branches is, that eventually they started to diverge14:56
← xxpor left14:56
valtido it was becoming a hussle14:56
→ dorkmafia joined14:56
→ RobW_ joined14:57
→ leeN joined14:58
← dcpc007 left14:58
← herdingcat left14:58
→ bwreilly joined14:58
← marthinal left14:58
RobW_ Is there a way to respect only the gitignore rules of the gitignore in the cu rrent directory?14:58
→ natechan joined14:58
← Misan left14:58
RobW_ FWIU gitignore settings cascade from top directories down.14:59
top/parent14:59
← rath left14:59
← zjason left14:59
milki ignore anything with a / in them?14:59
unignore*14:59
← rjhunter left14:59
→ herdingcat joined15:00
← VictorCL left15:00
benkay valtido: well, you do need to do some work to keep your branches up to date.15:00
→ tommyziegler joined15:00
→ zjason joined15:00
benkay valtido: you can't all work in a vacuum, that don't make no damn sense15:00
→ apipkin joined15:00
lov valtido: realistically, you don't want to have branches except for possibly your prod, where you'll potentially need to deploy cherry-picked commits rather than your whole upstream.15:01
→ King_Hual joined15:01
→ Davey joined15:01
lov or rather, while you do have branches, you want them to basically just be pointers, except for when you need to exclude certain commits.15:01
dwmw2 → dwmw2_gone15:02
← ThomasLocke left15:02
valtido hmmmm, so how many branches do you have typically ?15:03
→ lmat joined15:03
benkay does ops just go off and do crazy things for weeks at a time?15:03
→ shinnya joined15:03
benkay on my local machine or the remote repository i maintain?15:03
on my local machine i branch freely, and mostly always from master.15:04
RobW_ milki: So just unignore everything at the top of the local gitignore then add back in?15:04
valtido ok I think I may have tapped into something :) you have different branches on local, and different on remote ?15:04
← jjasinski left15:04
benkay yes.15:04
that is a particularly compelling feature of git.15:04
→ diegoviola joined15:04
valtido so on remote (to start with) how many u keep ?15:05
MrDHat → MrDHat|offline15:05
← dhruvasagar left15:05
benkay branch locally, fuck everything up, cherry pick to a nice clean branch, merge to master and deploy.15:05
→ ojacobson joined15:05
← reskp_ left15:05
benkay daaah, 5?15:05
maybe 715:05
master and test15:05
plus various features in the works15:05
→ reskp joined15:05
→ marthinal joined15:05
diegoviola hi15:06
→ tang^ joined15:06
valtido oh I see, that is enlightening ...15:07
→ chepre joined15:07
valtido how often do you do cherry picking?15:07
nightly ?15:07
benkay as infrequently as possible.15:07
devs are responsible for pushing clean branches.15:07
devs *must* be responsible for pushing clean branches.15:07
← ph88 left15:07
benkay ain't nobody want to clean up your stinking pile of mess in the middle of the conference room :(15:07
→ Advocation joined15:08
benkay the ideal process is for a dev to have responsiblity for a feature, work that feature in a branch, and then push that branch up to the remote either when they're ready to share it with people or deploy it15:08
→ dreiss_ joined15:09
benkay in either event, the branch quality is the developers responsibility, not the repository maintainers.15:09
valtido what if for example I and another dev have both done clean pushes, but somehow when the two are there may pose a bug...15:09
→ tobeplugged joined15:09
benkay your branches won't conflict until you merge them both into a different branch15:09
unless you named them the same thing, which, maybe don't do that.15:09
→ reskp___ joined15:10
benkay developers are responsible for keeping their local feature branches up to date with test and master15:10
← reskp left15:10
benkay a good continuous integration tool can help a lot with this: if a new merge hits master it can test that merge against all branches and auto-merge master into your developers branches without them even noticing15:11
ojacobson also SCM tools are not a replacement for communication :)15:11
benkay yeah that15:11
it actually sounds like you don't have solid engineering practices15:11
sorry that came off more harshly than i meant :(15:11
→ dhruvasagar joined15:11
→ kewah joined15:11
→ Mattias joined15:12
benkay another common pattern is "okay guys everbody pull"15:12
valtido it depends if I get offended easy, but no I don't15:12
→ jefferai joined15:12
benkay (that's nice. i try to err on the side of not pissing people off.)15:12
valtido we use jenkins for CI15:12
← adac left15:12
← cbreak-work left15:13
benkay anyways15:13
valtido it's better to ask stupid questions, than to not ask and be ignorant forever... or something along those lines...15:13
benkay yeah totally15:13
once you smash through the barriers of ignorance into fluency with the toolchain you can do some pretty awesome stuff15:13
but it takes the humility to say "i don't know" in public15:14
→ DABLGSG joined15:14
← ezekielnoob left15:14
← deuterium left15:14
← apipkin left15:14
benkay that said, the best knowledge i've ever gotten was from saying "hey so i have no idea what i'm doing here but this is what i think is going on and can someone please explain where my paradigms are broken?"15:15
← Xethron left15:15
← cagedwisdom_ left15:15
← banisterfiend left15:15
→ deuterium joined15:16
valtido I'll copy that :) and paste it (as Im sure I would be handy)15:16
you have been very helpful, many thanks.15:16
benkay any time!15:17
← mishok13 left15:17
benkay give back to the community when you reach that point in your career :)15:17
→ Search112 joined15:17
valtido will do :)15:18
← rat_thin_ left15:19
← Mattias left15:20
→ ezekielnoob joined15:22
→ ngladitz joined15:22
valtido @benkay your CI is it something you share with others or have on your local machine or both ?15:22
→ rethus joined15:22
benkay i'm a fan of circleci15:22
it watches the repository, runs tests and shows the results to everyone15:23
rethus how can i local with 2 git repos? I want to have one master. And create a clone of it for development.15:23
benkay but we also keep a pretty transparent development flow15:23
lots of code sharing and peer education etc15:23
← deuterium left15:23
→ simesy joined15:24
benkay rethus do branches not work for your use case?15:24
← svector left15:24
← Stalkr^ left15:24
← mklappstuhl left15:24
← keryell left15:25
valtido Well, I myself am a newbie on git, just know the basics... other team members came from *caugh caugh* SVN, so they use git as an SVN * terrible but hey bad habbits die hard*15:26
lmat I use git svn find-rev "r"revnum to find a git commit name for an svn rev. The problem is that often there is no git hash for a given svn rev (say it's on a branch I don't have or something). What I would like to do is what svn does, go back to the latest rev before the one I gave.15:26
Do I need to write a special script for this ? Or does git do it for me ?15:27
→ looper joined15:27
benkay branch agressively valtido15:27
maybe use gitbox15:27
← tbmiller left15:28
← simesy left15:28
rethus benkay: I use eclipse with egit. I have git-main repo in /git, and my projekt in /projekts/Pro1. Than i fetch the "master" from my main git, but no files in my project-folder are visible15:28
← lb1a left15:30
benkay eclipse and egit are out of my domain, friend :(15:30
← mizu_no_oto left15:31
← TheBrainWork left15:31
← grilix left15:31
→ TheBrainWork joined15:31
→ grilix_ joined15:31
← _Vi left15:31
← arand left15:32
← PaulCapestany left15:32
← MafooUK left15:32
→ DeltaHeavy joined15:32
DeltaHeavy Is there any reason BitBucket would be giving me "No repository at given URL". I'm 100% sure I have a repo there, but idk how it'd even access it without a password so that's why I'm wondering. Anybody nkow why? git://example.com/home/decord/public_html/staging15:33
ojacobson !repro15:34
gitinfo Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.15:34
juvenal → juvenal_away15:34
ojacobson Your URL there is useless, being example.com-ish15:34
DeltaHeavy ojacobson: I don't really want to provide my whole URL but w/e, here it is - git://decorgroupinc.com/home/decord/public_html/staging15:34
→ deuterium joined15:34
ojacobson How does bitbucket get involved here?15:34
← hoschi left15:34
DeltaHeavy It's asking me for the repo URL15:34
Trying to get this repo on here15:35
ojacobson If there's a git-daemon server on decorgroupinc.com, it's not listening to the internet15:35
attempts to contact it stall, then time out15:35
← mikecmpbll left15:35
juvenal_away → juvenal15:35
DeltaHeavy I don't have a git-daemon on there and can't really get one on there right now.15:35
→ mikecmpbll joined15:36
ojacobson Does that URL work for you if you clone it yourself?15:36
DeltaHeavy Yeah, but I'm using SSH15:36
ojacobson So "no", then15:36
(a different URL does work, but not the one you're trying to use.)15:36
DeltaHeavy Thing is bitbucket doesn't have 'ssh' under protocols I can use =/15:37
→ defaultro joined15:37
ojacobson Back up about eight steps and describe what goals you're trying to achieve here15:37
When all of this "works", what will you have accomplished?15:38
← szx left15:38
ojacobson I need more context to make sense of what you're saying. :)15:38
DeltaHeavy Honestly I really just want a GUI to browser file changes.15:38
That's it pretty much15:38
→ vkz joined15:38
DeltaHeavy Way to frustrating to do it via CLI15:38
ojacobson !gui15:38
gitinfo Graphical user interfaces are not supported here. If you want to get support, it needs to be through the git CLI. Reasons: 1) Because very few people here use the graphical interface. 2) Because giving instructions for GUI's is difficult. 3) The command line gives you a history of what commands you have executed.15:38
ojacobson wrong factoid15:38
There are a few decent *local* GUIs you can run, on whatever repos you have cloned/mirrored locally - have you tried them?15:38
DeltaHeavy Nope15:39
ojacobson SourceTree, Tower, Tortoise, hell even git-gui and gitk :)15:39
DeltaHeavy What would you suggest?15:39
→ dyng joined15:39
ojacobson GitX is okay but completely stagnant15:39
→ erichynds joined15:39
benkay gitbox is pretty excellent15:39
→ Mattias joined15:40
→ TomyLobo joined15:40
DeltaHeavy Checkout out sourcetree right now15:40
← dreiss_ left15:41
dwmw2_gone → dwmw215:43
→ dreiss_ joined15:43
← dhruvasagar left15:43
← dyng left15:44
← howlymowly left15:44
← irqq left15:44
← benkay left15:44
← Woet left15:45
→ irqq joined15:45
→ SoftwareMechanic joined15:45
← Dudalus left15:46
→ apipkin joined15:47
← valgrind left15:47
← gua_ould left15:48
→ glosoli joined15:48
← kewah left15:49
→ hoschi joined15:50
← ciampix left15:50
→ tyoc213 joined15:51
← aspiers left15:51
→ epicowl joined15:51
tyoc213 suppose that the same project, was "published two times" but they both started on a diferent set of programmer changes... so there is A and B repos of the "same source" with different branches, now I want to "move" one branch of A -> B it is possible... I mean, I don think they started exactly with the same set of files....15:52
→ yaymukund joined15:52
← rainbyte left15:52
tyoc213 think as they have no "version system" before... and they started their own each one15:52
← ThomasCle left15:53
→ Woet joined15:53
→ maestrojed joined15:53
lmat Okay, I wrote the script http://sprunge.us/XXPI15:53
yaymukund I just did git rebase -i [60 commits ago] and tried to squash everything into one commit. how can that produce a conflict while merging?15:53
← papegaaij left15:53
← kcj left15:53
→ ThomasCle joined15:53
qsusOFFcs → qsuscs15:53
yaymukund I thought it wasn't possible to conflict with squash, since it's just applying the commits in order15:53
→ infynyxx joined15:54
← glosoli left15:54
← ehmatthes left15:54
masak yaymukund: not with squash per se, I think. but if you're also rebasing to some other branch or commit, then there might be a conflict, as usual.15:55
yaymukund specifically, I'm doing `git rebase -i $(git merge-base master HEAD)`15:55
that makes sense15:55
→ geekbri joined15:55
← bwreilly left15:56
→ Norm__ joined15:56
juvenal → juvenal_away15:57
← Boohbah left15:57
→ mixed joined15:57
→ akiva joined15:57
→ glosoli joined15:57
← ThomasCle left15:57
mixed a few days ago i typed git checkout VERSION in order to checkout an older version; how can i know which version i've used?15:57
← user258467 left15:58
masak yaymukund: yes, then you're also rebasing on master, besides squashing.15:58
← mikecmpbll left15:59
← _Norm_ left15:59
yaymukund masak: ahhh, need to reread docs for rebase. thank you, that makes sense15:59
→ primitive joined16:00
← Balliad left16:00
esc http://dpaste.com/1421354/16:01
wtf16:01
→ gverri joined16:01
← Wolfspyre left16:01
← dreiss_ left16:01
→ Guest51606 joined16:01
← rethus left16:02
DeltaHeavy ojacobson: Working awesome. Thanks!16:02
← rchavik left16:02
← SergeyPenkin left16:02
→ MoHaX joined16:02
→ Wolfspyre joined16:02
← maestrojed left16:03
→ ehmatthes joined16:04
← hashpuppy left16:04
← Mattias left16:04
→ Mattias joined16:05
→ maestrojed joined16:06
→ krik joined16:06
← tommyziegler left16:06
→ h0bbit joined16:07
← ezekielnoob left16:07
← h0bbit left16:08
← Mattias left16:08
→ Mattias joined16:08
→ Destos joined16:09
← rurufufuss left16:10
← alip left16:10
← Ingmar left16:10
→ keryell joined16:10
→ GodEater joined16:11
← iamjarvo_ left16:11
← jargon left16:11
→ iamjarvo joined16:11
→ imachuchu joined16:12
→ mmlac-bv joined16:12
← deuterium left16:12
← GeertV left16:13
→ rath joined16:13
← Wolfspyre left16:13
→ GeertV joined16:13
→ johnnyfuchs joined16:14
→ juniorsilva7 joined16:15
→ yhager_ joined16:15
← iamjarvo left16:15
→ ezekielnoob joined16:15
→ Wolfspyre joined16:16
← Wolfspyre left16:16
← reskp___ left16:17
→ reskp_ joined16:17
← looper left16:17
→ yhager joined16:18
→ looper joined16:18
← tatsuya_o left16:18
← yhager left16:18
← looper left16:18
yhager_ → yhager16:18
← GeertV left16:18
← JokerxHarley left16:19
← howitdo left16:19
← bartman left16:19
← RobW_ left16:20
← Gooder left16:20
← cenuij left16:21
→ bwreilly joined16:21
← bwreilly left16:21
→ bartman joined16:21
juvenal_away → juvenal16:21
← reskp_ left16:21
→ bwreilly joined16:21
→ deuterium joined16:22
→ Misan joined16:22
← h12o left16:23
← deuterium left16:23
→ buibex_ joined16:23
← ChrisAnn left16:23
→ dtauzell joined16:24
← buibex left16:24
← keryell left16:25
dtauzell I'm switching from svn to git. Some of my files need to always have specific line endings (some are just LF ,some are just a CR) . Is there a way to force certain files to always have a specific line ending when checked out?16:25
ojacobson dtauzell: make sure automatic line conversion is turned off16:27
← chepre left16:27
ojacobson then check them in with the endings you want16:27
→ marhaban joined16:27
moritz and you an editor that picks up the line ending from the file16:27
(all decent editors do)16:27
← thannoy_ left16:27
moritz s/you/use/16:27
dtauzell so users should not have the "auto convert" setting then?16:27
→ JokerxHarley joined16:28
→ tommyziegler joined16:28
← buibex_ left16:29
← germanoak left16:29
← donmichelangelo left16:30
→ h12o joined16:30
→ donmichelangelo joined16:31
← mattp_ left16:31
→ howitdo joined16:32
→ dhruvasagar joined16:32
→ boombatower joined16:32
→ mattp_ joined16:33
← howitdo left16:33
← D9_ left16:33
→ howitdo joined16:34
← cmpera left16:35
rath dtauzell: read section "End-of-line conversion" in !man gitattributes16:35
gitinfo dtauzell: the gitattributes manpage is available at http://jk.gs/gitattributes.html16:35
Nevik .version16:36
gitinfo changed the topic to: Welcome to #git, the place for git-related help and empty cardboard boxes | Current stable version: 1.8.4.1 | Start here: http://jk.gs/git | Seeing "Cannot send to channel" or unable to change nick? /msg gitinfo .voice | May the forks be with you16:36
Nevik \o/16:36
← fRAiLtY- left16:36
← MoHaX left16:36
← infynyxx left16:37
← m0viefreak left16:37
→ AWinterman joined16:37
iveqy dtauzell: !crlf16:37
gitinfo dtauzell: to fix problems with line endings on different platforms, check out http://line-endings.gitrecipes.de/16:37
→ infynyxx joined16:37
→ bamj0rb joined16:37
gitinfo set mode: +v16:37
→ Mayumi joined16:38
Mayumi i have a c sharp mvc4 project that i want to put on github, are there any special things i should ignore?16:38
iveqy dtauzell: but IMHO this isn't something that git should be bothered with.16:38
Mayumi: all autogenerated stuff that you can generate from the information you've in your repo16:38
→ shoshin joined16:38
→ Georg3 joined16:38
Mayumi iveqy: i know, but i'm not sure about some files.16:39
→ goshawk joined16:39
Mayumi i'm not super experienced in c sharp yet16:39
and aren't sure about the things i can leave out16:39
← xobb left16:39
iveqy Mayumi: then you'll probably get a better answer in a c-sharp forum16:39
moritz Mayumi: it's pretty easy: if you do a build and then 'git status', it should be clean16:39
Mayumi good idea16:39
thanks16:39
=)16:40
iveqy moritz: no! :(16:40
moritz iveqy: no?16:40
iveqy moritz: that would mean he track binaries, right?16:40
moritz: he should not track compiled code16:40
moritz iveqy: no, that .gitignore's binaries16:40
*that he16:40
→ cmpera joined16:41
iveqy moritz: where does he say so?16:41
moritz iveqy: he doesn't, but he should :-)16:41
→ RobW_ joined16:41
← infynyxx left16:41
→ travisstaton joined16:43
← krik left16:43
← stayarrr left16:43
← maqr left16:43
→ Shusshu joined16:44
← Shusshu left16:44
→ germanoak joined16:44
← germanoak left16:44
← tobeplugged left16:44
→ germanoak joined16:44
← germanoak left16:45
→ germanoak joined16:45
← germanoak left16:45
Sigma[Mtp] → Sigma16:45
→ germanoak joined16:46
← devoldmx3 left16:47
→ devoldmx joined16:48
← bobdobbs left16:48
→ bobdobbs joined16:49
→ cmpera_ joined16:49
← cmpera left16:50
cmpera_ → cmpera16:50
→ banisterfiend joined16:52
→ whunt joined16:52
→ PaulCapestany joined16:53
→ iamjarvo joined16:53
→ keryell joined16:53
← Chryyz left16:53
→ rudak joined16:54
← btree left16:55
← v0latil3 left16:55
← richgilbank left16:56
← khushildep left16:57
← nuclearkitten left16:58
← mixed left16:58
← sword left16:58
→ ThomasCle joined17:00
Vampire0_ → Vampire017:00
← Swayz left17:00
← hahuang65 left17:01
← jkyle left17:01
→ maqr joined17:02
← irqq left17:02
← dtauzell left17:03
← Advocation left17:04
→ cenuij joined17:04
← koki left17:04
← harobed_ left17:05
LuxuryMode Someone asked me to change my commit msg for a pull request on github, how do i do that?17:05
← juniorsilva7 left17:05
PerlJam They don't like your commit message?17:05
→ breck7 joined17:06
← aChatir left17:06
→ Swayz joined17:08
→ infynyxx joined17:09
← iamjarvo left17:09
→ Mars` joined17:09
← infynyxx left17:09
← ezekielnoob left17:11
← asQuirreL left17:11
→ jasom joined17:12
← normanrichards left17:12
→ grilix joined17:12
← maestrojed left17:12
← DoniRS left17:13
jasom Is there a config option to set pull to use --ff-only?17:13
there's a merge.ff but not a pull.ff option17:13
→ hobodave joined17:14
iveqy jasom: !pull17:14
gitinfo jasom: pull=fetch+merge (or with flags/config also fetch+rebase). It is thus *not* the opposite of push in any sense. A good article that explains the difference between fetch and pull: http://longair.net/blog/2009/04/16/git-fetch-and-merge/17:14
→ aberrant joined17:14
aberrant hi all17:14
iveqy LuxuryMode: man git rebase17:14
gitinfo LuxuryMode: the git-rebase manpage is available at http://jk.gs/git-rebase.html17:14
→ dreiss_ joined17:15
iveqy LuxuryMode: look at -i17:15
← grilix_ left17:15
LuxuryMode PerlJam, nope. they have a very specific format bc it's a large multi-project thing17:15
← ewnd9 left17:16
jasom iveqy: I know the difference between fetch and pull. I prefer pull to do a fetch and a ff-only merge (which I can do with git pull --ff-only). I'm asking if I can make that the default behavior of git pull17:16
→ ewnd9 joined17:16
kellysaurusx → ELLIOTTCABLE17:16
← ewnd9 left17:17
→ juniorsilva7 joined17:17
→ ewnd9 joined17:17
→ maestrojed joined17:18
iveqy jasom: so set the option for merge and be happy?17:18
→ ezekielnoob joined17:18
→ deuterium joined17:19
← Forkk left17:19
← Baluse left17:19
← zarubin left17:20
jasom iveqy: and then I have to pass different options to merge... I'll stick with what I'm doing now, which is an alias for pull --ff-only17:20
→ Forkk joined17:21
← DeltaHeavy left17:21
tang^ you can set the pull to use rebase by default instead of merge17:21
jasom tang^: I don't want it to rebase by default17:21
← Davey left17:22
← ewnd9 left17:22
tang^ jasom: merge --ff-only and rebase are effectively the same, no?17:22
jasom tang^: not even close17:22
→ a-l-e joined17:22
→ ewnd9 joined17:22
tang^ okay17:22
still, your merge.ff settings should affect your pull, since a pull is fetch+merge by default17:23
← Octalot left17:23
← glebihan_ left17:24
← marhaban left17:24
← ezekielnoob left17:24
→ Davey joined17:24
← GodEater left17:24
→ GodEater joined17:24
jasom In the simple case (where a ff is possible) they are the same, since there are no local changes to rebase; but if you have any changes not in the branch you pull from, then the ff-only will fail and the rebase will roll back your branch, do a ff-merge and then re-apply all your local changes17:24
→ glebihan_ joined17:24
→ hholtmann joined17:24
← joshbetz left17:25
← Droolio left17:25
← deuterium left17:25
→ deuterium joined17:27
iveqy jasom: do I miss something here? Since a pull is a fetch+merge the options set for merge will apply even if you pull17:27
→ revyd joined17:28
→ iamjarvo joined17:29
→ ezekielnoob joined17:30
→ koki joined17:31
→ hahuang65 joined17:31
← thomas_b left17:31
← Mars` left17:32
← nnyk_ left17:34
← stephCoue left17:34
← mastro left17:35
← avolz left17:36
← ezekielnoob left17:36
→ BiggFREE joined17:36
← dhruvasagar left17:36
→ minim joined17:37
→ Mars` joined17:38
→ infynyxx joined17:38
← tommyziegler left17:38
→ tommyziegler joined17:39
→ asQuirreL joined17:40
→ svector joined17:41
← Asenar left17:41
→ ezekielnoob joined17:43
→ raonyguimaraes joined17:44
→ zarubin joined17:44
← Search112 left17:45
← fling left17:46
→ Search112 joined17:46
← gdoteof left17:46
← Stygia left17:46
← devoldmx left17:46
→ jfoy joined17:47
← BiggFREE left17:47
→ richgilbank joined17:47
→ Advocation joined17:49
← nwertman_ left17:50
← shoshin left17:50
→ mdiers joined17:51
← Advocation left17:51
→ Advocation joined17:51
→ k4rtik joined17:51
← keryell left17:54
← ezekielnoob left17:54
→ danman joined17:55
← iamjarvo left17:55
→ iamjarvo joined17:56
← hyperair left17:56
← jfoy left17:56
→ iamjarvo_ joined17:58
→ bwreilly_ joined17:58
→ nicedice joined17:59
← Kronuz left17:59
← iamjarvo left17:59
← Advocation left18:00
← gverri left18:00
→ iamjarvo joined18:01
← bwreilly left18:02
← Guest51606 left18:03
← herdingcat left18:03
← bwreilly_ left18:03
→ Kronuz joined18:04
← maqr left18:04
← dfanjul left18:04
→ maqr joined18:05
← iamjarvo_ left18:05
→ gverri joined18:05
← iamjarvo left18:05
← RobW_ left18:05
← gverri left18:05
→ iamjarvo joined18:06
→ hashpuppy joined18:07
→ gverri joined18:09
→ pantsman joined18:09
← revyd left18:09
→ revyd joined18:09
→ keryell joined18:10
← svector left18:10
← luxgeek left18:10
← sr78ger left18:11
emmajane → emmajane|lunch18:11
→ Prazanna joined18:11
→ dhruvasagar joined18:11
← angelsl left18:12
← minim left18:12
→ snql joined18:12
→ pr0ton joined18:13
← hashpuppy left18:13
→ byprdct joined18:13
→ koki__ joined18:13
← byprdct left18:14
← revyd left18:14
→ scelis joined18:14
← glosoli left18:14
← scelis left18:14
→ eephillip joined18:15
→ bwreilly joined18:15
dropdrive Sorry, newbie question, is there a short-hand for HEAD~1..HEAD, or $sha~1..$sha?18:16
← koki left18:17
EugeneKay To see the diff between the last two commits?18:17
`git show HEAD` will do most of what you want18:17
← Mars` left18:17
← infynyxx left18:17
→ Mars` joined18:18
→ infynyxx joined18:18
EugeneKay Actually, just `git show` will assume HEAD, so that's really short.18:18
→ GeertV joined18:18
→ scelis joined18:18
dropdrive EugeneKay: I actually want something to pass to git-diff, git-log, etc., so I want something explicit18:19
← jarr0dsz left18:19
EugeneKay Then you've already got the most shortened version ;-)18:19
→ jarr0dsz joined18:19
← maqr left18:19
← keryell left18:20
grawity there's "HEAD^.." (or in future versions "@^..")18:20
and there's `git log -n 1` or just `git log -1`18:20
→ infynyxx_ joined18:21
→ maqr joined18:21
→ pbujok joined18:22
← Malukenho left18:22
← Mars` left18:22
→ gdoteof joined18:22
← twoface88 left18:22
← infynyxx left18:22
→ Malukenho joined18:23
← magu|q left18:24
→ BSaboia joined18:24
← Olea|2 left18:24
→ nwertman joined18:26
→ hyperair joined18:26
→ zacts joined18:27
EugeneKay Does that work for "$SHA^.." too? Or just assumes HEAD at the end?18:28
General case, yay18:28
← ewnd9 left18:29
→ darinm joined18:30
← babilen left18:31
darinm → darinm_away18:32
→ MrV_ joined18:32
← Norm__ left18:34
→ Nicholas_Rishel joined18:34
→ tvw joined18:35
← Nicholas_Rishel left18:36
← GuySoft left18:38
← maqr left18:38
← przemoc left18:39
← Sigma left18:39
→ przemoc joined18:40
→ marcosc joined18:40
→ nessib joined18:40
← richgilbank left18:40
← bgerber left18:41
marcosc Hi, I'm having troubles trying to squash commmits... I tried to squash a set of commits, then someone checked in some code on upstream, so now I ended up with this big mess :( https://github.com/mozilla/servo/pull/1069/commits18:41
→ Norm__ joined18:41
← infynyxx_ left18:41
marcosc Now every time I try to squash, things just get worst :(18:41
→ richgilbank joined18:41
marcosc I should have a single "Implemented HTMLMainElement"18:42
← dougsland left18:42
→ Sigma joined18:42
→ mike_af joined18:43
← MrV_ left18:43
milki marcosc: how do the upstream commits play into your squash?18:44
marcosc milki: hopefully they shouldn't18:45
milki marcosc: they why did it make a mess?18:45
← mmlac-bv left18:46
marcosc milki: I don't think the upstream made a mess... those are all my commits (except 1), but they are just not squashing18:46
ThomasCle → Bumblebee18:46
Bumblebee → ThomasCle18:46
marcosc I just seems to end up with a new "Implement HTMLMainElement" every time I try to squash18:46
instead of actually squashing18:47
→ infynyxx joined18:47
milki sounds like you should start over from a fresh branch, rebase your original changes, squash, and force-push to the pull branch again18:47
→ jackdempsey joined18:48
iveqy milki: force a puch is a bad idea usually18:48
marcosc: !public18:48
gitinfo marcosc: [!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://goo.gl/waqum18:48
jackdempsey → jackdempsey|afk18:48
milki iveqy: its a pull request, thats how you clean up a pull reuqest18:48
→ Olea|2 joined18:48
milki you can either start a new pull request with a new branch, or fix the branch associated with the pull request18:48
→ irqq joined18:48
iveqy milki: sorry, missed the link18:48
← raonyguimaraes left18:48
→ kesha joined18:49
→ Andrevan joined18:49
marcosc gitinfo: they people who own the repo don't like having long commit histories because it makes it hard for them to find where bugs were introduced (that's what they tell me)18:49
→ bgerber joined18:49
→ raonyguimaraes joined18:50
iveqy marcosc: !bot18:50
gitinfo marcosc: [!gitinfo] I am an IRC bot which responds to certain keywords to provide helpful(?) information to humans. Please see http://jk.gs/git/bot for more information about how to use me.18:50
→ rat_thing joined18:50
← nardev left18:51
→ fobius joined18:51
→ maqr joined18:52
Dougie187 marcosc: They are full of crap. :P Tell them to look at `git bisect`18:52
fobius I'm trying to merge in new files from a subtree. I did a pull on the subtree and if I check the subtree out the new files show up. Then I tried to do git merge -s subtree --squash --no-commit my_subtree and it said everything went okay but the new files aren't showing up.18:52
→ nardev joined18:52
← GeertV left18:53
marcosc Dougie187: the reviewer said " The majority of these commits won't build independently, which would make bisecting frustrating."18:54
→ GeertV joined18:54
Dougie187 marcosc: That is true. It does make bisecting rather useless.18:54
If you can't test it.... you can't debug it.. :P18:54
marcosc still a bit unsure what to do :(18:54
marcosc should I try rebasing -i and then push -f ?18:55
iveqy marcosc: squash and push like milki said?18:55
→ krik joined18:55
← epta left18:55
marcosc iveqy: ok, will try18:55
→ epta joined18:55
← lmat left18:55
← rath left18:56
← jarr0dsz left18:58
← GeertV left18:58
→ dougsland joined18:58
→ mmlac-bv joined18:59
marcosc milki: when you say start over from a fresh branch, do you mean without any changes at all?18:59
or branch from my current messy branch?18:59
← matteo left19:00
→ MoHaX joined19:00
→ alvaro_o joined19:01
→ rat_thin_ joined19:01
← k4rtik left19:01
← rat_thing left19:01
→ revyd joined19:02
← pr0ton left19:02
→ Mars` joined19:04
JesseC Does anyone know how you would select a specific value such as the title, from this json response? All of the items I'm finding on stackexchange iterate through the entire response with $.each, but there has to be a better way right?19:06
http://gdata.youtube.com/feeds/api/videos/gzDS-Kfd5XQ?v=2&prettyprint=true&alt=jsonc19:06
← tham left19:07
sonOfRa JesseC, what does this have to do with git, exactly?19:07
JesseC sonOfRa, oops! wrong room, my bad19:07
sonOfRa (if using javascript, just treat that as json, and select response->title)19:08
← jerius left19:10
JesseC sonOfRa: yeah, I tried various things I found without success, gonna drop it in the correct room and write up a fiddle for an example.19:10
← revyd left19:11
← bwreilly left19:11
→ revyd joined19:11
→ vurma joined19:12
→ jerius joined19:12
← Pyrogerg left19:12
← mmlac-bv left19:12
← juniorsilva7 left19:13
fobius Does git merge -s subtree not add new files? I don't understand why the files exist in the subtree but aren't being added in the merge19:13
← germanoak left19:14
→ _Vi joined19:15
→ keryell joined19:16
emmajane|lunch → emmajane19:16
← revyd left19:16
→ minim joined19:17
→ joshsmith joined19:18
← banisterfiend left19:19
→ robbyoconnor joined19:19
→ banisterfiend joined19:20
→ pfFredd joined19:25
→ relipse joined19:25
→ VictorCL joined19:26
bamj0rb result.data.title, from the looks of it.19:28
← infynyxx left19:28
← rendar left19:29
← banisterfiend left19:29
← MoHaX left19:30
bamj0rb fobius: I've never attempted to merge a subtree before. Are the new files in the current branch or the merged branch?19:30
→ petey_ joined19:30
→ banisterfiend joined19:31
→ Kismet010 joined19:31
→ darnir joined19:32
bamj0rb Sounds like subtree is a rather odd merge strategy. I assume its for when the repository is moved higher up the file system...19:32
Kismet010 it's a good workflow to create a branch for each member in a team of two? something like: member1 -> master <- member219:32
← keryell left19:33
yaymukund Kismet010: why? I would just stick to feature branches19:33
← Olea|2 left19:34
→ keryell joined19:34
← asQuirreL left19:34
kalleth_ → kalleth19:35
Kismet010 to avoid conflicts when working on the same code lines at very separate dates19:35
← zjason left19:35
→ zjason joined19:35
→ tommyziegle joined19:36
← vurma left19:36
← Guest1779 left19:37
→ stayarrr joined19:37
→ jfoy joined19:38
← joki left19:38
← Search112 left19:38
→ Olea|2 joined19:39
← jfoy left19:39
← FrodoTheHobbit left19:40
← cff left19:40
← asakura left19:40
← pbujok left19:40
← dorkmafia left19:40
← tommyziegler left19:40
bamj0rb Kismet010: Don't try to avoid conflicts. That is what version control is for. The only way to avoid conflicts is communication, but generally it's unavoidable if you want software to get better (and want more than one person actually working on it).19:41
→ FrodoTheHobbit joined19:41
bamj0rb Kismet010: Use "topic" or "feature" branches instead. Make them temporary. Don't share them with each other unless the other person needs to look at/work on it.19:42
Kismet010: Well, there are many different workflows to choose from, but that's what I like to do.19:42
→ MoHaX joined19:42
iveqy Kismet010: !panic19:42
gitinfo Kismet010: [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice!19:42
← darinm_away left19:42
iveqy Kismet010: you might also be interested in !flow19:43
gitinfo Kismet010: [!gitflow] The description of the gitflow branch workflow model is at http://nvie.com/posts/a-successful-git-branching-model/ while a tool to help implement this workflow is at https://github.com/petervanderdoes/gitflow See http://sethrobertson.github.com/GitBestPractices/#workflow for other workflow suggestions/references19:43
→ shoshin joined19:43
← shoshin left19:43
iveqy Kismet010: personal branches sounds stupid from a history point. Since branching is so simple in git you usually want more than one branch per user19:43
→ shoshin joined19:43
← ozette left19:43
→ mmlac-bv joined19:44
→ darinm joined19:44
darinm → darinm_away19:44
← tyoc213 left19:45
→ bwreilly joined19:45
← shoshin left19:46
Kismet010 bamj0rb: it's a project for learning a framework, so it will be many conflicts/commit reset to do for any mistake we do and we'll work separately in time, so I thought it will be a good idea merge to master when one has any code ready, we'll not work on features, we'll change a lot of code with no purpose but learn19:46
it's still a bad idea?19:47
iveqy19:47
→ germanoak joined19:48
← germanoak left19:48
→ germanoak joined19:48
→ luke_ joined19:49
bamj0rb In that case your "person1" and "person2" branches are really just your topic branches.19:50
Note that if you do that it will be hard to merge.19:50
If you split changes up into logic thoughts and keep them small it is much easier to merge that.19:50
→ pbujok joined19:50
← jackdempsey|afk left19:51
bamj0rb The problem with doing 50 unrelated things and trying to merge is that the conflicts are compounded by all of the unrelated work. Instead of figuring out how to merge one idea, you have to figure out how to merge all of the ideas at once. Often this is overwhelming and you'll make mistakes.19:51
← MoHaX left19:52
→ pr0ton joined19:52
→ devbug joined19:52
← nardev left19:52
fobius bamj0rb: The new files are in the subtree, i.e. I pulled them into the branch that I have for the subtree. But when I go to merge them into the main project i.e. master branch, nothing happens19:52
bamj0rb fobius: I'm not really the best person to help with the subtree merge strategy since I'm not sure how it works. I'm guessing. Maybe it would help to pastebin the tree structure of both branches?19:53
→ dvide joined19:53
→ infynyxx joined19:53
← gdoteof left19:53
→ g0bl1n joined19:53
← zacts left19:54
← maestrojed left19:54
← Mars` left19:54
→ Mars` joined19:55
Kismet010 bamj0rb: I'm already inmersed in a mistake, this is way I'm reconsidering another options ^^19:56
problem is the other member doesn't know about git, I created his branch to just tell him: do checkout and commit19:57
→ sq-one joined19:57
Kismet010 If I tell him to do branches, we'll run in very big mistakes I think19:58
fobius bamj0rb: Eh, I'd rather not because the log messages are a bit confidential. I'm sure this is something stupid. It literally does nothing, even if I delete the subdirectory before merging. It just says Automatic merge successful.19:58
← Mars` left19:59
→ ReachingFarr joined19:59
← snql left20:00
bamj0rb fobius: I wasn't asking for the logs. I was asking for the tree structure. i.e., the file system structure. Just so we're on the same page about what you're trying to do. Fair enough if that's confidential too... You could mock up an example with dummy names though.20:00
→ dofunc joined20:00
bamj0rb Kismet010: Ah, so the burden will be on you to merge. You can always let your collaborator do whatever he wants to do on master and then keep your own branch(es) separate.20:01
fobius bamj0rb: Do you just want like the output of ls -R? Sorry, a bit of a newb with git.20:01
bamj0rb Kismet010: Ultimately, the problem is the same in the end. It's going to be messy to merge. At least if your changes are organized nicely it will be easier, even if his are a nightmarae.20:01
→ infynyxx_ joined20:01
bamj0rb fobius: Sure. Just to give an idea of the structures.20:02
→ tmcmahon joined20:02
← paulweb515 left20:03
→ lmat joined20:03
lmat How do I compare the index to HEAD ?20:03
moritz git diff --cached20:03
fobius bamj0rb: http://pastebin.com/d8gZrUcr20:03
← relipse left20:04
fobius Basically, lib has been added as a remote branch but it's on the local filesystem. I read it in with git read-tree and now I'm trying to pull in some new files that weren't originally tracked in lib.20:04
← diegoviola left20:04
← gverri left20:04
→ relipse joined20:04
← infynyxx left20:04
Kismet010 bamj0rb: he usually forgive to checkout to his branch and change master, I was considering master as the 'production' branch... but your idea sounds better: master branch for him, develop branch for me and production branch when code is ready20:04
← Luke-Jr left20:05
lmat moritz: super :)20:05
→ b_rare joined20:05
→ Luke-Jr joined20:05
← mmlac-bv left20:06
lmat moritz: I should have seen this. I did git help diff and searched for "index" (since the top line of DESCRIPTION says, "Show ... changes between the index and a tree...".20:06
Kismet010 it's easier that creating a hook for locking master as I did...20:06
lmat moritz: Then, now I see the second synopsis explained as "This form is to view the changes you staged for the next commit..." they didn't say index!20:06
← g0bl1n left20:06
lmat moritz: Would you make sure the terminology is ironed out...soon ? :) thanks ^_^20:06
ReachingFarr I have used the `git svn clone` command to import the trunk of an SVN project into Git. I would now like to import two SVN branches, with history. After they are imported, I don't want to push changes to these branches back up to the SVN repo. What should my next step be?20:07
← relipse left20:07
← iamjarvo left20:07
→ nardev joined20:07
→ relipse joined20:07
bamj0rb fobius: I'm not sure I understand where the subtree part comes in. There doesn't seem to be any similarities between lib/* and ./*.20:07
EugeneKay !beer20:07
gitinfo Beer! It's what's for dinner!20:07
→ joki joined20:07
moritz lmat: I won't make sure, because I'm not invovled with the git docs or anything20:07
bamj0rb EugeneKay: \o/20:07
moritz lmat: if you want it improved, improve it yourself20:07
→ iamjarvo joined20:07
← zarubin left20:07
bamj0rb fobius: Honestly I don't know how to use --strategy=subtree so you should ask again in a bit. :)20:07
→ mmlac-bv joined20:08
← iamjarvo left20:08
← relipse left20:08
EugeneKay I'm not sure what the syntax is to fiddle with a svn repo's branches. My git-svn experience is limited to --stdlayout20:08
But I'm sure alcohol will help.20:08
→ relipse joined20:08
→ iamjarvo joined20:08
_ikke_ !beer20:08
gitinfo Beer! It's what's for dinner!20:08
fobius bamj0rb: No, there aren't - they are separate. The merge should go automatically. I'm following instructions from http://git-scm.com/book/ch6-7.html. Basically the idea is that lib has it's own branch and it just has an existence in the main directory. It's tracked by both its own branch and the master branch.20:08
bamj0rb EugeneKay: In my experience, it is easy to make a mess of things. I wish I understood it better so I could clean it up manually. I generally end up having to clone again and again after a while. :P20:08
← bwreilly left20:09
→ zarubin joined20:09
← relipse left20:09
fobius bamj0rb: I pulled to lib_branch successfully and if I check it out the files are there. But for some reason I can't merge the files into ./lib.20:09
← richgilbank left20:09
→ relipse joined20:09
juvenal → juvenal_away20:10
← Tommy[D]_ left20:10
→ ZujkisNx joined20:12
→ gdoteof joined20:12
bamj0rb fobius: Oh neat. I never knew about that before. Thanks. :D20:13
fobius: I think it will be hard to help without seeing what steps you took. And I imagine by now your command history is a mess from trying and retrying.20:13
fobius: Perhaps you should try again, and then time try to keep the steps straight-forward and then pastebin the whole command history...20:13
← LuxuryMode left20:13
fobius bamj0rb: Okay, I'll try that.20:14
→ Tommy[D] joined20:15
→ GT4066 joined20:15
← pr0ton left20:16
← marthinal left20:16
bamj0rb Maybe you'll get lucky and the second time will be the charm. :D20:17
juvenal_away → juvenal20:17
→ pr0ton joined20:19
← pr0ton left20:19
← DrNick left20:19
← IrishPrime left20:20
fobius bamj0rb: How do I delete the remote branch? It says branch lib/master not found but git checkout lib/master works fine.20:20
← geekbri left20:20
← zarubin left20:20
← iamjarvo left20:22
→ iamjarvo joined20:23
← tmcmahon left20:24
→ iamjarvo_ joined20:24
← iamjarvo_ left20:24
← AaronMT left20:25
_ikke_ git branch -rd lib/master20:25
→ iamjarvo_ joined20:25
← Roots47 left20:26
← jkremser left20:27
← iamjarvo left20:27
← renatow_ left20:28
→ AaronMT joined20:29
← iamjarvo_ left20:29
dofunc → subfunc20:29
← AaronMT left20:30
→ iamjarvo joined20:30
← subfunc left20:30
→ babilen joined20:30
→ subfunc joined20:30
← subfunc left20:31
→ subfunc joined20:31
→ DrNick joined20:31
← subfunc left20:32
→ marhaban joined20:32
← spaceonline left20:32
→ subfunc joined20:32
bamj0rb fobius: Hopefully you're trying this on a fresh clone of your repo too. That will eliminate any bleeding over.20:32
← hoschi left20:32
bamj0rb fobius: Never hurts to have a backup in case you royally f'something up too. >:)20:33
← infynyxx_ left20:33
← maroloccio left20:34
← koki__ left20:34
→ richgilbank joined20:34
← iamjarvo left20:34
→ infynyxx_ joined20:34
→ rurufufuss joined20:35
← marhaban left20:37
→ marhaban joined20:37
→ lavaman1 joined20:37
← lavaman left20:37
← Jackneill left20:40
→ Mars` joined20:40
→ user_ joined20:41
→ luxgeek joined20:41
→ sword joined20:42
→ DolphinDream joined20:42
→ jkremser joined20:43
← jkremser left20:43
→ jkremser joined20:43
← user_ left20:43
→ finisherr joined20:43
← keryell left20:43
← erichynds left20:44
fobius bamj0rb: I did everything over and it seems to be working okay now. I still don't know what was going on before. =(20:44
← petey_ left20:44
→ petey_ joined20:45
← gdoteof left20:45
→ gdoteof joined20:45
finisherr So, I have a bunch of local git repositories on my MacPro. I've also got a MacMini that I want to set up as a remote server. I've created a git user and set up ssh keys and that works fine. But, I'm not exactly sure how to get all of my repositories set up on the server. I followed some instruction from the documentation by created a bare version of one of my git repos, then copying that over to the server, then running git init --shared --bare or20:46
something like that. Now, how on earth do I update the remote?20:46
iveqy man git remote20:46
gitinfo the git-remote manpage is available at http://jk.gs/git-remote.html20:46
← Kicer86 left20:46
→ Oishikatta joined20:46
← ehmatthes left20:46
iveqy finisherr: however, you're probably better off with !gitolite20:46
gitinfo finisherr: Gitolite is a tool to host git repos on a server. It features fine-grained access control, custom hooks, and can be installed without root. Download: https://github.com/sitaramc/gitolite Docs: http://gitolite.com/gitolite/20:46
→ petey__ joined20:46
← petey_ left20:46
← richgilbank left20:47
finisherr git push?20:47
→ paulweb515 joined20:48
Oishikatta Is it possible with git/github to have a branch that applies a small patch, but otherwise stays in sync with master?20:48
← a-l-e left20:48
← banisterfiend left20:49
→ Cervantes joined20:49
→ keryell joined20:49
→ bwreilly joined20:51
→ corni joined20:52
finisherr oh, git --set-upstream20:53
i'm all set20:53
git push --set-upstream that is20:53
← Mars` left20:53
→ Algorithm joined20:53
→ jackdempsey joined20:53
← Algorithm left20:53
→ Mars` joined20:53
← subfunc left20:53
→ tagava joined20:54
→ tagava|2 joined20:54
→ Roots47 joined20:54
→ carbohydrate joined20:55
← Roots47 left20:55
← robbyoconnor left20:55
← tang^ left20:55
← howitdo left20:56
← tvw left20:56
← Mayumi left20:56
← cmpera left20:56
→ peroht joined20:56
← bwreilly left20:57
→ Reacto joined20:57
← minim left20:58
→ therealGent joined20:58
← freeed left20:58
→ ehmatthes joined20:59
→ avolz joined20:59
Oishikatta Or, say there's a branch with some changes, and master has other changes. How can I copy the changes from master to the other branch without replacing it's changes (assume there are no conflicts)?21:00
→ ubuntu__3 joined21:00
← mmlac-bv left21:00
← krik left21:01
→ richgilbank joined21:01
← Spyro left21:02
→ tjj joined21:02
EugeneKay Oishikatta - man git-cherry-pick21:02
gitinfo Oishikatta: the git-cherry-pick manpage is available at http://jk.gs/git-cherry-pick.html21:02
← Dougie187 left21:02
→ pr0ton joined21:02
Oishikatta Thanks21:03
EugeneKay It's not really a merge, you're just copy-pasting a commit onto another branch. But it sounds like that's what you want ;-)21:03
→ Spyro joined21:03
→ minim joined21:03
← mdiers left21:03
→ mdiers joined21:03
Oishikatta what I actually want is 'all commits since <branch> was created', but looks like that would work21:04
skorgon Oishikatta, probably rather man git rebase?21:04
gitinfo Oishikatta: the git-rebase manpage is available at http://jk.gs/git-rebase.html21:04
skorgon well, or merge, depending on your workflow21:05
→ zarubin joined21:05
→ dorkmafia joined21:05
← TheBrainWork left21:06
← onethfour3 left21:09
→ onethfour joined21:09
← therealGent left21:10
← vkz left21:10
EugeneKay merge will combine the two branches, if you want all changes from both21:11
If you have conflicts you'll need to sort them out yourself(or use -Xours/-Xtheirs option to try resolving it automagically by favoring one side)21:12
→ robbyoconnor joined21:12
← breakingthings left21:13
← stayarrr left21:13
qsuscs → qsusOFFcs21:14
← psynor left21:15
← rudak left21:15
→ rudak joined21:15
→ Bosox20051 joined21:15
→ maestrojed joined21:15
← ojacobson left21:17
bamj0rb fobius: It happens. :( The good news is that it's working. Sometimes doing it once helps you to better understand it the second time and helps you not miss or screw up steps. Sometimes gremlins.21:17
fobius bamj0rb: yeah, I think there was something funny about the way that it was attaching the subtree to that folder.21:17
← carbohydrate left21:18
← robbyoconnor left21:19
→ normanrichards joined21:19
→ MILLIONAIRE joined21:20
MILLIONAIRE can somene help me upload a project to github21:20
Oishikatta merge did what I was looking for, thanks skorgon / EugeneKay21:21
← chuckharmston left21:22
bremner MILLIONAIRE: #github can.21:22
→ robbyoconnor joined21:22
skorgon github has that documented, AFAIK21:22
← rudak left21:23
→ rudak joined21:23
jamesd i send a link to such documentation for a small fee, about $500 usd should do it21:23
→ r0bby_ joined21:23
bremner MILLIONAIRE: do you have a more specific question?21:23
→ marcuy joined21:27
← robbyoconnor left21:27
→ Lancien joined21:28
← Xiti left21:28
← SHyx0rmZ left21:29
← m0ltar left21:29
← donmichelangelo left21:30
← bamj0rb left21:30
→ donmichelangelo joined21:31
→ bonhoeffer_ joined21:32
→ pfFred joined21:33
juvenal → juvenal_away21:33
← richgilbank left21:34
← cenuij left21:36
→ dbuch joined21:36
→ cenuij joined21:36
← tjj left21:37
← pfFredd left21:37
← peroht left21:38
← dvide left21:38
← ReachingFarr left21:40
← joshsmith left21:40
→ Roots47 joined21:41
← sq-one left21:41
→ ReachingFarr joined21:41
lmat So I guess "Lines starting with '#' will be ignored..." is not for git commit -F - < commitmessage ...21:42
→ Pyrogerg joined21:43
→ richwestcoast joined21:44
Kismet010 so, instead of this http://nvie.com/img/2009/12/Screen-shot-2009-12-24-at-11.32.03.png21:44
← kesha left21:45
← johnnyfuchs left21:45
← PaulCapestany left21:45
← davividal left21:45
→ kesha joined21:45
← rfoust left21:46
→ dvide joined21:46
← Olea|2 left21:47
→ PaulCapestany joined21:47
Kismet010 it's a good idea to have another branch for 'production' instead of master, and use master as 'development' branch? just for simplicity: master is defaulting, no need to change to develop every time and avoid mistake form others member in the production-ready branch21:49
← pr0ton left21:49
← zarubin left21:50
→ cff joined21:50
iveqy Kismet010: I thought the same and was "furious" when they changed it at my company21:50
← decotii left21:50
→ groglogic joined21:51
iveqy master is default and should be development IMHO however most projects have master as their "production" and it's less confusing to do as "everybody else" rather than doing "the right way"21:51
← TomyLobo left21:51
← kesha left21:52
Kismet010 so master is standard just because everybody (as me) usually does?21:52
→ pr0ton joined21:52
← pr0ton left21:53
← irqq left21:53
iveqy Kismet010: master is standard for the production branch, as showed in the picture you linked21:54
→ pr0ton joined21:54
← pr0ton left21:54
← jerius left21:54
Kismet010 and any project you'd see in github21:55
iveqy Kismet010: yes.21:55
→ pr0ton joined21:55
→ erichynds joined21:56
← pr0ton left21:56
Kismet010 i'll look for something to change default branch21:56
bremner tbh, I don't think projects on github are a good place to get workflow ideas21:56
→ Xiti joined21:56
iveqy Kismet010: that's just to change the HEAD of your remote21:57
→ mdiers1 joined21:57
iveqy Kismet010: but I'm not sure you should bother, since noone else seem to do it. Better invest more in education for your users21:57
→ banisterfiend joined21:58
← phile left21:58
→ jumperswitch_ joined21:59
Kismet010 you just said the more difficult way to solve this ^^21:59
most*21:59
→ howitdo joined21:59
← kylef left21:59
← jumperswitch left22:00
→ jnewt joined22:00
← mdiers left22:01
→ pr0ton joined22:01
← pr0ton left22:02
iveqy Kismet010: yes, but the correct one.22:02
← mdiers1 left22:02
iveqy Kismet010: why give developers a tool they can't use? How costly isn't it to have developers not knowing their tools?22:02
→ mdiers joined22:02
← infynyxx_ left22:03
→ jumperswitch joined22:03
← apipkin left22:04
→ pr0ton joined22:04
← v0n left22:05
← Ozch left22:05
← jumperswitch_ left22:05
← pr0ton left22:05
← idoru left22:05
→ jumperswitch_ joined22:05
← jackdempsey left22:06
← marhaban left22:06
← mmattice left22:06
← bonhoeffer_ left22:06
← s00pcan left22:06
← petey__ left22:06
← mmitchell left22:07
→ s00pcan joined22:07
→ petey_ joined22:07
Kismet010 in a ideal world maybe, I have to prevent changes on master because it's default -with a hook I think- or change the usual workflow master as 'dev' branch22:07
→ mmitchell joined22:07
→ pr0ton joined22:07
← jumperswitch left22:08
→ JustinMac joined22:08
← eephillip left22:08
→ groglogi_ joined22:09
→ eephillip joined22:09
iveqy Kismet010: the thing is that nobody should ever work on the dev branch either, but in topic branches22:10
→ cagedwisdom joined22:10
← devbug left22:10
iveqy Kismet010: this ideal world thing actually exists, but I also know that it's not something that can exists in all corporate enviroments, I'm sure you're the better judge of what fits your enviroment22:10
← pbujok left22:11
→ mizu_no_oto joined22:11
← groglogic left22:11
iveqy Kismet010: you're on the right track, you can prevent pushes with hooks. If you're using gitolite you even can do access control on branch level22:11
← petey_ left22:12
Kismet010 I found an option on bitbucket22:12
← mmitchell left22:12
jnewt i want to setup a directory on the server where a few people can push their repo's i don't want to have to create the bare repo for them to push to everytime they decide to create a repo, and i don't want them manually ssh to the server and poking around with all the repos.22:12
_ikke_ jnewt: gitolite has support for that22:13
jnewt am i asking for the impossible?22:13
_ikke_ look for wild repositories22:13
http://gitolite.com/gitolite/wild.html22:13
← eephillip left22:14
→ kermit joined22:14
← Roots47 left22:14
→ rednaks joined22:14
← yaymukund left22:15
← marcosc left22:15
→ gdrc joined22:15
gdrc this git is driving me crazy!22:15
→ beastd joined22:15
→ jumperswitch joined22:16
→ epicowl_ joined22:16
iveqy gdrc: !refund22:16
gitinfo gdrc: If you are not satisfied with git, or the support provided by the volunteers in #git, you are entitled to a full refund of the purchase price, and are invited to use another VCS. Elsewhere.22:16
→ tjj joined22:16
gdrc :)22:16
iveqy :)22:16
_ikke_ :)22:16
iveqy gdrc: git usually drives you crazy until you understand it, then it's really simple and powerful22:16
jnewt iveqy, until you want to do something new, then you start all over22:17
gdrc I'm trying to use it without studied that's becouse I can't use it ;)22:17
← epicowl left22:17
epicowl_ → epicowl22:17
iveqy jnewt: nah, when you understand the underlying model, it's all simple =)22:17
→ jumperswitch__ joined22:18
iveqy gdrc: !book !talks, I specially recommend Jessicas22:18
gitinfo gdrc: 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 !parable22:18
gdrc: Some good video talks about Git: [yt] http://goo.gl/z72s (Linus Torvalds: History&Concepts); [yt] http://goo.gl/R9H2q (Scott Chacon: Git basics, live examples); http://vimeo.com/35778382 (Randal Schwartz: Git basics, descriptional); http://vimeo.com/46010208 (Jesica Kerr: Git basics, descriptional)22:18
← jumperswitch_ left22:18
← grilix left22:18
gdrc oh nice tomorrow I will see torvalds video22:19
→ grilix joined22:19
iveqy gdrc: that video won't help you learn git, at all22:19
jnewt iveqy: BS, i've been using git for a year at least, every day. then you want to do something different, and you're back to the man, searching for something that looks similar to what you want for hours22:19
gdrc btw, want I want to do and fail it that is: roll back to commit #122:19
iveqy gdrc: it's very entertaining though22:19
gdrc: !revert22:19
gitinfo gdrc: That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?22:19
skorgon iveqy, but it'll motivate you to throw all alternatives over board22:19
iveqy skorgon: it sure will =)22:20
← Cervantes left22:20
→ jumperswitch_ joined22:20
← tjj left22:20
iveqy jnewt: I guess it depends on how you're using git. Using it frequently doesn't make you good at it.22:20
← jumperswitch left22:20
← homo_vitruvius left22:20
← hobodave left22:21
← Swayz left22:21
gdrc iveqy: with git log I get fatal: bad default revision 'HEAD'22:21
← Bosox20051 left22:21
iveqy gdrc: !repro22:22
gitinfo gdrc: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session -- or, even better for complex issues, design a minimal case in which your problem can be reproduced, and share it with us. This will help immensely with troubleshooting.22:22
iveqy gdrc: have you broken your repo?22:22
gdrc I think so22:22
jnewt iveqy: agreed, and if you're using it the same way, every day, you'll obviously never have any problems, and will be happy in understanding your workflow, which could be only a fraction of what others use22:22
← jumperswitch__ left22:23
gdrc it broke my balls haha22:23
skorgon jnewt, usually google + stackoverflow have good answers for everything you could possibly try to do with git22:23
← defaultro left22:23
← hholtmann left22:23
→ v0n joined22:23
← grilix left22:23
← normanrichards left22:23
iveqy jnewt: yes, and in many cases git isn't the right tool for you then, because most people that uses git that way uses it in a svn way. (of course, there's people that uses branches and then they are right about using git)22:24
← pantsman left22:24
dwmw2 → dwmw2_gone22:24
→ iamjarvo joined22:24
→ jackdempsey joined22:25
← rnsanchez left22:25
→ chrob joined22:25
← Stalkr_ left22:25
← rideh left22:25
→ pbujok joined22:25
← Sigma left22:27
← bartzy left22:28
→ jfoy joined22:28
→ rnsanchez joined22:29
gdrc iveqy: ok. I hard-reset the branch and I see all my commits (10 commits). I want to roll back to the first one.22:29
→ Roots47 joined22:29
← lmat left22:29
← ws2k3 left22:29
← danman left22:30
→ smlgs joined22:30
→ ws2k3 joined22:30
← corni left22:32
← pr0ton left22:32
iveqy gdrc: then hard-reset your branch22:33
→ tjj joined22:33
← sgambino left22:33
← iderik left22:34
→ v0latil3 joined22:34
← _Vi left22:35
← MILLIONAIRE left22:35
← dbuch left22:35
gdrc iveqy: get some issues :s22:35
← samuel02 left22:36
← Roots47 left22:37
→ samuel02 joined22:37
→ adlez joined22:37
→ Swayz joined22:38
→ mmitchell joined22:38
← rudak left22:38
← tjj left22:38
← jfoy left22:39
← Andrevan left22:40
→ tjj joined22:40
→ hashpuppy joined22:40
← natechan left22:41
→ yaymukund joined22:41
← mike_af left22:41
← cff left22:41
← samuel02 left22:42
→ reskp_ joined22:42
→ Roots47 joined22:42
← Burer left22:43
m3ga what's the secret of getting the v2 part of the '[PATCH v2 X/Y]' subject line when using git-send-email? i've read the man page but that doesn't help. do i really need to trawl through the perl code?22:43
skorgon man git format-patch22:44
gitinfo the git-format-patch manpage is available at http://jk.gs/git-format-patch.html22:44
skorgon look for -v (reroll count)22:44
m3ga, or, if you git is too old, --subject-prefix and manually setting it to 'PATCH v2'22:44
→ madewokherd joined22:45
m3ga no i've got 1.8.4rc3. i didn't realise the patch had to be added during format-patch rather than send-emai.22:46
← mmitchell left22:46
m3ga skorgon: thanks22:46
→ pr0ton joined22:46
skorgon m3ga, i think send-email accepts most of the switches for format-patch and passes them on, but i never tried that22:46
← VictorCL left22:47
m3ga skorgon: thanks! working now.22:47
juvenal_away → juvenal22:48
skorgon cool22:48
← jasom left22:48
← r0bby_ left22:49
→ bwreilly joined22:49
← yaymukund left22:49
← mizu_no_oto left22:49
→ thomas_b joined22:50
← reskp_ left22:51
→ reskp joined22:51
→ simesy joined22:52
→ Guest___ joined22:52
← ngladitz left22:52
← ReachingFarr left22:53
← jumperswitch_ left22:55
← Davey left22:56
← Prazanna left22:56
← falconfox left22:56
→ erbist joined22:56
→ epicowl_ joined22:56
← reskp left22:56
← darnir left22:57
← Pyrogerg left22:57
→ falconfox joined22:58
→ bonhoeffer_ joined22:59
Smirnov_ → Smirnov22:59
→ Joshi3s` joined23:00
iveqy gdrc: what issues?23:00
← King_Hual left23:00
gdrc I solved with git push -f origin master23:00
← epicowl left23:00
epicowl_ → epicowl23:00
← pbujok left23:01
← ThomasCle left23:04
← bonhoeffer_ left23:05
iveqy gdrc: correct23:05
gdrc: however !publice23:05
← grahamsavage__ left23:05
← breck7 left23:08
→ breck7 joined23:09
← minim left23:09
← tplaner left23:09
← germanoak left23:10
→ therealGent joined23:11
← jnewt left23:11
← finisherr left23:12
gdrc iveqy: thx23:13
← darinm_away left23:13
← breck7 left23:13
← erbist left23:14
→ kcj joined23:14
← acharlieh left23:16
← m7w left23:16
← bwreilly left23:19
← Mars` left23:20
→ Mars` joined23:20
← pr0ton left23:20
groglogi_ → groglogic23:20
← jackdempsey left23:21
→ pr0ton joined23:21
→ grahamsavage__ joined23:22
← _nova left23:25
← pr0ton left23:25
→ pr0ton joined23:26
← pr0ton left23:26
← boombatower left23:26
→ AWinterm_ joined23:27
→ asQuirreL joined23:27
→ jackdempsey joined23:27
← simesy left23:28
→ pr0ton joined23:28
→ simesy joined23:28
← cenuij left23:29
→ minim joined23:29
Guest___ → Alc23:31
← AWinterm_ left23:31
→ Gamecubic joined23:31
← AWinterman left23:31
→ AWinterman joined23:31
← lavaman1 left23:32
← primitive left23:32
→ tengelic joined23:32
← simesy left23:33
← asQuirreL left23:33
iveqy gdrc: !public23:34
gitinfo gdrc: [!rewriting_public_history] Rewriting public history is usually bad. Everyone who has pulled the old history have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push -f` to force (and the remote may reject that, anyway). See http://goo.gl/waqum23:34
tengelic hello! I messed up my local repo. is there a way to remove all references, fetch the remote references to the unreferenced local files, and redo the last commit?23:35
← marcuy left23:35
← knigitz left23:36
→ knigitz joined23:36
← AWinterman left23:36
← leeN left23:36
grawity not entirely sure what you mean by "fetch remote references to the unreferenced local files"23:39
← bdavenport left23:39
→ devoldmx joined23:39
grawity (or why you think removing everything is necessary)23:40
tengelic my terminology might be wrong, but I try to explain23:42
← epicowl left23:42
→ drasko_ joined23:42
drasko_ hi all. How to clean history from removed file?23:42
I want to make .git directory smaller23:42
→ bdavenport joined23:43
drasko_ by removing unnecessary snapshots of the history23:43
← travisstaton left23:43
drasko_ of the binaries that I removed from repository23:43
→ _Norm_ joined23:44
→ Prazanna joined23:44
grawity https://help.github.com/articles/remove-sensitive-data#cleanup-and-reclaiming-space23:45
→ pfFreddd joined23:46
→ richgilbank joined23:46
→ psynor joined23:46
bremner man git-filter-branch is pretty decent23:46
gitinfo the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html23:46
← Norm__ left23:47
tengelic grawity, I made a commit and changed the author name on several old commits (while this wasn't really necessary), and now I can't push my repository to the remote one. I would like the redo everything with Git (get back the old history), and make that last commit again.23:47
← Prazanna left23:47
← pfFred left23:47
grawity "redo everything with Git" and "get back the old history" aren't synonyms23:48
you still have the old history, you can just go back to it23:48
← richgilbank left23:48
grawity e.g. if you're on the 'master' branch, you can just `git reset --hard origin/master`23:48
→ adlez_ joined23:48
← erichynds left23:48
grawity (and use `git cherry-pick` to re-apply that commit)23:49
← deuterium left23:49
→ bwreilly joined23:50
← _Norm_ left23:50
→ Norm__ joined23:50
tengelic grawity, so git rebase doesn't rewrite the history?23:51
grawity it does23:51
but more precisely, it creates new rewritten history23:51
it doesn't immediately discard old one, just removes references to it23:51
and lets git garbage-collect it sometime later23:51
and 'later' can take a while, since the old history remains in the branch's reflog for about two weeks23:52
(+ HEAD's reflog)23:52
← adlez left23:52
← Alc left23:53
→ AWinterman joined23:54
← jaimef left23:55
→ jaimef joined23:55
melee i pushed a commit with a bad message (referenced the wrong ticket) - is there an easy way to amend this or would that involve rewriting history?23:55
← iamjarvo left23:55
→ johnnyfuchs joined23:55
grawity amending *is* rewriting history23:55
→ alip joined23:55
→ iamjarvo joined23:56
melee that's what i thought :( suggestions for fixing?23:56
grawity but if you do `git commit --amend`, fix it, and `git push --force` REALLY QUICK, maybe nobody will notice :P23:56
melee i'm a solo developer, no one will notice23:56
:P23:56
BUT i do like to do things correctly haha23:56
grawity the largest issue with rewriting history is that it becomes annoying to other developers23:56
← JustinMac left23:57
grawity it's not really a technical problem – git does have the most extensive history rewriting commands out of most VCSes, after all23:57
→ ozette joined23:58
melee i just went ahead and forced it, haha23:58
linus hates me23:58

Logs Search ←Prev date Next date→ Channels Documentation