IRCloggy #git 2012-05-05

Logs Search ←Prev date Next date→ Channels Documentation

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

2012-05-05

codeM0nK3Y it's a one time thing, I never want to use svn again in my life. I didn't realise I had so many projects using it in the first place00:00
← DofLeFou left00:01
→ jameswarren joined00:01
← kerrick left00:01
bremner there is a fast export based svn2git by the kde team00:01
it needs a local svn repo00:02
← szager left00:02
← LekeFly left00:03
bremner for small jobs git-svn is usually easier00:03
← Vert left00:04
codeM0nK3Y I was looking at this answer before I tried svn2git earlier so might try that. Seems to give more flexibility too which will be handy when it comes to making my tag names git-friendly00:05
← jfelchner left00:05
codeM0nK3Y http://stackoverflow.com/a/3972103/135934300:05
← jim0203 left00:06
→ Anchakor joined00:07
← txomon left00:08
Anchakor hi, how do I unstage files (both changed and new files) from commit? (not discarding the changes in them)00:08
codeM0nK3Y git reset --soft HEAD^00:08
that'll undo the last commit00:09
Anchakor well I didn't commit it yet00:09
I just staged lot of stuff with git add including stuff I didn't want to stage00:09
codeM0nK3Y Ahh, I can't remember that command but it tells you it after each git add00:09
Anchakor (use "git reset HEAD <file>..." to unstage)?00:09
codeM0nK3Y yeah that's the one00:10
← dreiss left00:10
Anchakor doesn't it discard the changes in the files?00:10
codeM0nK3Y shouldn't do, but backup the file if you don't trust it00:10
cmn reset only touches the worktree with --hard00:11
Anchakor ah ok, that is why I thought so00:11
thanks00:11
← tharkun left00:12
→ tharkun joined00:12
diegoviola stares at svn users: http://ompldr.org/vZG1rYg00:12
codeM0nK3Y what about the ones trying as hard as they can to get away from it :P00:12
→ pandeiro joined00:12
← bLN left00:13
diegoviola then we will help00:14
→ jfelchner joined00:17
← iwoj left00:18
← watsonian left00:19
← Anchakor left00:19
codeM0nK3Y think the main thing getting in my way right now is the fact that the repos are remote and require auth00:20
got one more method to try before i set up the repo locally and try moving that to git instead00:20
→ nutron joined00:20
→ iwoj joined00:20
← kpreid left00:21
→ kpreid joined00:21
← kiva left00:22
← wilmoore left00:23
← jfelchner left00:24
codeM0nK3Y ooo its working! man this shit's scary00:24
← tvw left00:25
← barrbrain left00:28
→ barrbrain_ joined00:28
frogonwheels_ → frogonwheels00:29
← apok left00:30
→ mister_m joined00:30
← AaronMT left00:31
← tharkun left00:31
→ edude03_ joined00:31
← edude03 left00:31
→ apok joined00:34
→ williamcotton joined00:36
→ tharkun joined00:36
← kermit left00:39
← gjoseph left00:39
→ jfelchner joined00:40
← mister_m left00:44
← osmosis left00:47
← fisted left00:47
→ cenuij joined00:50
← cenuij left00:50
→ cenuij joined00:50
← cenuij left00:50
→ cenuij joined00:51
← cenuij left00:51
→ cenuij joined00:51
← jfelchner left00:52
← mcluskydodallas left00:54
← williamcotton left00:54
→ fisted joined00:56
→ vitor-br joined00:57
← Eridius left00:58
← fayimora left00:58
← lonewulf` left00:58
→ Lord_DeathMatch joined01:00
← tharkun left01:00
→ HyperNerdV2 joined01:02
← cenuij left01:02
→ cenuij joined01:03
← cenuij left01:03
→ cenuij joined01:03
← cenuij left01:03
← spyro left01:03
← ColKurtz left01:03
→ spyro joined01:03
← spyro left01:03
→ spyro joined01:03
→ Goplat joined01:04
→ lonewulf` joined01:04
← mwrosen left01:05
→ tharkun joined01:06
→ ivand joined01:06
← ghvandoorn left01:06
→ kermit joined01:07
→ satyanash joined01:08
← xaka left01:09
← HyperNerdV2 left01:09
← SegFaultAX|work left01:10
← tharkun left01:11
→ Vampire0_ joined01:13
← Vampire0 left01:16
→ tharkun joined01:16
← guampa left01:17
→ gavin_huang joined01:19
→ Evanlec joined01:19
← Lgb left01:20
← cyphase left01:20
→ Lgb joined01:21
← akiress left01:24
← tharkun left01:25
→ tharkun joined01:25
← felgru left01:25
← ben225_ left01:26
→ pjn_oz joined01:29
→ Destos joined01:29
→ gerard0 joined01:30
→ stkrzysiak joined01:31
→ JohnyBaby joined01:33
Vinnie_win famous last words01:33
← friskd left01:33
← tharkun left01:34
→ guampa joined01:34
← Bass10 left01:34
← TomyLobo left01:34
← thelazydeveloper left01:35
codeM0nK3Y haha01:35
← calimocho left01:35
codeM0nK3Y seems I've sussed it, just redoing it so I don't add a bunch of old (previously deleted) branches that it pulled across01:36
→ cyphase joined01:36
→ BiggFREE joined01:36
← Chillance left01:36
→ calimocho joined01:39
← guampa left01:39
← jstout24 left01:39
→ guampa joined01:39
→ tharkun joined01:39
→ friskd joined01:40
→ neurodrone joined01:40
← maletor left01:42
← Lord_DeathMatch left01:43
→ Bass10 joined01:45
← lonewulf` left01:45
→ lonewulf` joined01:47
→ intripoon_ joined01:48
← friskd left01:50
← intripoon left01:51
← thebignoise left01:52
→ mister_m joined01:53
→ patrick99e99 joined01:53
patrick99e99 Hi git buddies.. I need some help...01:53
← tharkun left01:54
patrick99e99 So-- I forked rails.. I made a branch... made some changes.. made a pull request.. everything was cool..01:54
→ tharkun joined01:54
patrick99e99 then I made more changes and wanted everything to be in 1 commit.. so I did an interactive rebase and squashed my two commits01:54
and did git push -f mine01:54
....... Now my pull request shows TWO commits.. one by me, and one by someone else..01:55
which is wrong01:55
so..01:55
I am just wondering what I should do?01:55
codeM0nK3Y screaming often helps01:56
← Yuuhi`` left01:56
patrick99e99 so I see the hash before my rebase.01:57
do I just want to do git reset --hard <branch> <hash> ?01:57
Vinnie_win someone else commited to your fork?01:58
patrick99e99 no...01:58
something got screwed up with I did rebase -i and then push -f01:58
I don't quite know what..01:58
so I just wnat to undo my rebase01:58
*want01:58
Vinnie_win you should have backed up your work to another branch before the rebase01:58
patrick99e99 yes...01:59
but I still can revert back with the reflog hash01:59
I just don't remember what the proper way to do that is01:59
Vinnie_win Back up your local repo first01:59
patrick99e99 I did01:59
→ guampa|2 joined01:59
← guampa left01:59
Vinnie_win Get a clean master branch from rails, check it out to a new branch, then cherry pick your commits into it01:59
Then push that to your upstream02:00
If you have a backup of your local repo then you don't care what's in your remote, you can always re-fork and then apply your commits again02:00
guampa|2 → guampa02:00
→ flippo joined02:01
← harish left02:01
patrick99e99 thanks Vinnie_win02:02
→ ska-fan joined02:06
← smw left02:07
→ thebignoise joined02:08
← tharkun left02:09
→ krz joined02:09
→ harish joined02:11
→ tharkun joined02:15
→ xaka joined02:16
→ sdeboey joined02:16
→ wilmoore joined02:16
→ dreiss joined02:16
← apok left02:17
← oriba left02:17
← pjn_oz left02:17
→ apok joined02:17
← apok left02:18
← tharkun left02:20
→ com4 joined02:20
→ daethorian joined02:21
← Bass10 left02:22
→ jave joined02:22
→ Bass10 joined02:23
→ tharkun joined02:25
← thiago left02:26
→ rux joined02:26
← dvide left02:26
← guampa left02:31
→ guampa joined02:35
← todd_dsm left02:36
→ todd_dsm joined02:37
← BiggFREE left02:37
→ cybersphinx joined02:37
← jameswarren left02:39
← edude03_ left02:40
→ jameswarren joined02:40
← jameswarren left02:40
← the_cybersphinx left02:40
← orafu left02:40
← guampa left02:40
→ orafu joined02:40
→ bglusman joined02:40
← dreiss left02:41
→ edude03 joined02:42
→ igotnolegs joined02:44
← thearthur_ left02:45
→ dv310p3r joined02:46
→ arvind_khadri joined02:46
← arvind_khadri left02:46
→ arvind_khadri joined02:46
→ Orphis_ joined02:46
→ peppyheppy joined02:47
→ MestreLion joined02:49
MestreLion hum, this may sound eird, but... i want to rebase all changes introduced in commits from B to A (linear) on top of an un-related branch X, and I want branch X to move to the new rebased tip02:50
Vinnie_win what do you mean unrelated? the directory tree is completely different?02:50
MestreLion well, no... X is an ancestor of B02:51
like 20 commits before02:51
Vinnie_win so use rebase --interactive and rearrange everything how you'd like02:51
← fir31igh1 left02:51
→ tmcmahon joined02:52
Vinnie_win checkout X to a new branch (for testing) and just try it02:52
→ fir31ight joined02:52
MestreLion ok, but i'm not sure if I should use --onto or not... or how to make sure that branch X will move but branhces tipped at A won't02:52
Vinnie_win well just duplicate the X branch and have at it, if it doesn't do what you want, start over02:52
MestreLion i'll duplicate A too.. theres a branch tipped there, which i don't want to move02:53
Vinnie_win well A won't be changing02:53
← harish left02:53
← SkAzZ left02:53
Vinnie_win or did I misunderstand you?02:53
MestreLion so maybe git rebase --onto X B A ?02:54
Vinnie_win clone X first02:54
git checkout -b X2 (from X branch)02:54
MestreLion ok, I will... but then, what command should I use?02:54
Vinnie_win I dont know.02:54
MestreLion will git rebase -i X2 B A2 replay all changes from B to A on top of branch , and then move X2 to the newly-created tip?02:55
"on top of branch X2"02:56
Vinnie_win I don't know. I have never done such a thing, but what I am suggesting is you try it on a duplicate branch that you don't mind messing up. See if the result suits your taste, and if not then try again with some different option.02:56
Make a back up of your local repo first of course02:57
→ adamsilver joined02:58
MestreLion scenario is this: X is "stable" branch, and B-A are a series of commits from "testing" branch that are ready to be integrated in master02:58
Vinnie_win You could have done the command already and answered your own question!02:59
MestreLion i was cherry-picking the commits, but then I realized a rebase may be much faster...02:59
I won't answer anything by doing blind guesses03:00
Vinnie_win well with rebase you can just delete the lines in the editor for the commits you don't care about03:00
MestreLion i was trying to avoid trial and error03:00
Vinnie_win the branch passed as the argument to --onto would be the one you want to rewind the branch you're rebasing to03:01
So you want to extract commits B-A from X and commit them to master?03:02
MestreLion no, X is master in my example03:02
Vinnie_win where are B-A ?03:02
MestreLion stable branch03:02
B-A are in a testing branch03:02
Vinnie_win So, extract commits B-A from "testing" and commit them to master?03:03
MestreLion yes03:03
← adamsilver left03:03
← Bass10 left03:03
Vinnie_win let me think03:03
MestreLion and move master03:03
Vinnie_win well commiting to master will move it03:03
milki rebase with --onto would work03:04
MestreLion B-A are ahead of master03:04
milki: will it move master?03:04
milki no03:04
it will put you in a detached HEAD03:04
then you move master yourself03:04
Vinnie_win that sounds accident prone03:05
milki probably03:05
this is what i got from experiement with my stupid idea of changing ff merges into non-ff merges03:05
Vinnie_win can't he extract the desired commits into a new branch with rebase, and then merge the new branch to master?03:05
MestreLion there's no rebase syntax that move master after playing a series of commits on it?03:05
milki regular rebase does that03:06
Vinnie_win You could duplicate the testing branch, rebase it with respect to master and delete the undesired commits, then merge that when it looks right.03:07
MestreLion milki: and how to tell rebase to "play from B to A in master" ?03:07
→ davidbanham joined03:07
→ waveclaw joined03:07
milki rebase --onto master B~1 A03:07
MestreLion milki: I tested that.. but, as you said, it detached head03:08
milki right03:08
← bitkiller left03:08
milki then git branch -f master03:08
to move master03:08
just make sure its what you want first03:08
!03:08
Vinnie_win's way would be safer...03:08
MestreLion weird for rebase not to have an option for this.. I assumed that "picking some commits from a testing branch to a stable branch" was fairly common scenario03:09
milki o03:09
maybe cherry-pick will be good enough...03:09
→ dreiss joined03:09
milki you can cherry-pick a range of commits03:09
MestreLion yes, that's what I was doing so far03:09
Vinnie_win were you cherry picking one at a time?03:09
milki probably03:09
MestreLion but its boring... rebase -i gives me a nice screen with all commits03:10
milki lol03:10
indeed03:10
Vinnie_win I agree03:10
milki id want non-interactive rebase to show the commits being rebased too03:10
MestreLion i can choose all i want and not want in a single edit.. and the git do its magic03:10
Vinnie_win Like I said, clone "testing", then rebase -i master, do your edits, save, and then you can merge that to master and delete the cloned branch.03:10
MestreLion testing branch has like 40 commits (disorganized collaborator)03:11
Vinnie_win yeah so?03:11
just delete the commits you dont want03:11
MestreLion i'll split this branch in several feature branches03:11
Vinnie_win well now you're trying to do two things at once03:11
MestreLion some commits belong to i18n, some can go to stable, some will got to a new feature branch, and so on03:11
← davidbanham left03:11
milki sounds like you need to make branches anyways03:12
MestreLion so I'll first rebase to master to select the ones that go there03:12
then i'll rebase the same branch to another branch...03:12
In the end, I'll have basically the same commits, but neatly divided in logical branches03:12
logical and independent03:13
the "destination" branches already exist03:13
← metalfan_ left03:13
milki o.O03:13
MestreLion I want to update them with selected commits from that huge, linear teasting branch of his03:14
← ben_alman left03:14
→ angelsl joined03:14
← ska-fan left03:14
← herdingcat left03:14
← az left03:14
← EvanDotPro left03:14
→ ben_alman joined03:14
→ fuzzybunny69y joined03:15
MestreLion think of it this way: project already have several branches... one for translations, one for core engine, another for tools, etc03:15
that guy simply merged them all and started working from there, doing like 30 or 40 commits03:15
← tharkun left03:15
Vinnie_win just out of curiosity what language are the sources in?03:15
MestreLion all linear03:15
→ tharkun joined03:15
MestreLion Vinnie_win: python, why?03:15
Vinnie_win just wondering03:16
→ EvanDotPro joined03:16
→ ska-fan joined03:16
MestreLion I am able to idntify reasonably well which branch should each commit go03:16
milki so you want to rebase those select commits against master before putting them into their destination branches o.O03:16
MestreLion some of them can go directly to stable... others belong to some of our existing feature branches, etc03:16
Vinnie_win you should probably just clone the repo locally and play with it to have a dry run of what you want to do before you do it for real03:17
→ az joined03:17
Vinnie_win it sounds like you're going to be doing a lot more than just cherry picking a range03:17
milki so does cherry-pick range work then?03:17
MestreLion no milki: I want to split a 40+ series of linear commits in proper, independent branches03:17
milki cherry-pick for each branch then?03:18
← pandeiro left03:18
MestreLion milki: no, range doesnt... B - A was just his work...03:18
cherry pick each *commit* to one of our branches (master being just one) works03:18
but... thats a LOT of work03:18
milki theres more than B-A you are moving?03:18
MestreLion it's like picking fruits from a garden and placing ech one in the proper basket03:19
B - A is his feature branch, with 40 commits... some will go to branch X, some to Y, some to Z, etc03:19
← bondar left03:19
milki and the hard part is you want to automate this or find a command that does everything03:20
?03:20
→ Bass10 joined03:21
MestreLion if we number B = 0 and A = 30, it would be like: 1, 5, 14, 24 => master; 2, 4, 15-22, 29 => branch Y; 3, 18-20 => branch Z03:21
not, not everything... i was thinking about a rebase per branch03:21
Vinnie_win that's foolish03:21
MestreLion in 1 rebase I would pick all the selected commits that will go to master03:21
← mister_m left03:22
MestreLion then another rebase to pick commits for Y03:22
milki o, because interactive rebase can do it03:22
MestreLion etc etc03:22
Vinnie_win That's not the best way03:22
MestreLion exactly03:22
milki i dont think non-interactive can do that03:22
MestreLion it will be interactive03:22
f course03:22
yhager is there a way to determine where each commit goes (I mean mechanically, by a script)?03:22
MestreLion can't be otherwise03:22
Vinnie_win In 1 rebase you should first group the commits so that commits belonging to the same destination branch are sequential, then you can cherry pick ranges instead of going through the interactive routine for every new branch.03:22
→ smw joined03:22
MestreLion no yhager...03:23
milki i dont see how the rebase is necessary03:23
other than the interface >.>03:23
MestreLion yhager: I'll look in each commit message, the files changes, and make a fair choice03:23
but the interface IS the main reason for rebasing instead of copy and pasting like 20 commit shas' to cherry pick03:24
yhager then you can simply generate a list of commits (git log --pretty=oneline), put it in a file, and edit the file to contain only branch A commits, another copy will contain branch B commits etc. Then simply feed those files into the proper git cherry-pick commands03:24
→ herdingcat joined03:25
Vinnie_win Anyway, you could have done it by now with all the time spent talking.03:25
milki lol03:25
→ aarond_123a joined03:25
MestreLion ok guys, thanks for the ideas03:25
yhager Vinnie_win: it's always about having fun doing it.. :) (and talking about it)03:25
← dv310p3r left03:26
MestreLion this will be a long, several hours task... so I don't mind talking for like 20, 30 minutes if that helps designin a workflow that may ease my life later :)03:26
← tharkun left03:26
→ tharkun joined03:26
Vinnie_win what! several hours to reshuffle 40 commits?03:26
milki i took 2 months to reshuffle a couple hundred03:27
im not done yet too03:27
MestreLion I like Vinnie_win 's approach... I could create 4 clones of his testing... then rebase each, deleting commits... then cherry-picking a range... sounds like a good workflow03:27
← lacrymology left03:27
Vinnie_win thanks03:27
MestreLion it's not only about reshuffle.. there are commit messages, conflicts, etc etc03:28
and a *lot* of testing03:28
Vinnie_win oh03:28
MestreLion both to see that nothing is broken03:28
→ Karmaon joined03:28
Vinnie_win Hmmm03:28
MestreLion AND also to see if I actuallyabsorbed *all* of his work and don't leave any code line behind :)03:28
← rux left03:28
Vinnie_win I commit very frequently...every 5 to 10 minutes03:28
→ rux joined03:29
milki thinks you should just yell at the developer03:29
MestreLion he did that too, which is great... so basically each commit goes to a clear branch... not many "monster commits" that I would have to split into pieces to put each piece into a different branch03:29
Vinnie_win and I test my app for at least 10 seconds to make sure it still works03:29
milki lol 10 seconds03:30
Vinnie_win gotta have standards03:30
MestreLion but those also exists... i've seen some diffs... i will have to split some commits too03:30
Vinnie_win that truly sucks03:30
milki yup03:31
Vinnie_win a lot of time wasted on repo administration...taking away from development time03:31
MestreLion I just assumed that, if rebase can be seen as a "mass cherry-picker", I could use it to ease my work03:31
since a batch cherry pick is exactly what i need now :D03:31
Vinnie_win From what you are saying, it sounds like the cost of a single cherry pick is a lot less than the work of verifying the commit's correctness03:31
MestreLion Vinnie_win: copy and pasting sha's are tedious when git rebase -i already shows me the whole list and a "checkbox approach" to select them03:32
Vinnie_win you know...you can use the short hash..03:33
MestreLion Vinnie_win: and the commits are correct... I've extensively tested his branch tip.. everythin works03:33
so now it's time for surgery :)03:33
→ maletor joined03:33
MestreLion the end result is good... his work IS amazing03:33
and he commits often too... he just don't care much about branches and feature topics :P03:34
← yhager left03:35
→ khmarbaise_ joined03:35
milki compromises compromises -.-03:36
codeM0nK3Y woo! finally got everything converted :D03:36
milki yay?03:36
codeM0nK3Y had to retag everything, but small price to pay for one less project on svn :P03:36
milki aha03:36
→ HyperNerdV2 joined03:36
← herdingcat left03:36
→ thiago joined03:37
codeM0nK3Y this repo only had trunk, 1 branch and 3 tags... I'm dreading converting a project with 600+ commits across 6 branches and more tags :(03:37
← khmarbaise left03:37
khmarbaise_ → khmarbaise03:37
→ fayimora joined03:37
← fuzzybunny69y left03:38
← tharkun left03:40
← timoxley left03:43
← peppyheppy left03:45
← Destos left03:46
→ SegFaultAX|work joined03:46
→ tharkun joined03:46
→ dv310p3r joined03:46
→ ben225 joined03:46
← ben225 left03:47
← Twey left03:47
← fayimora left03:47
← maletor left03:48
→ Twey joined03:49
→ mister_m joined03:50
← aristidesfl left03:51
← dv310p3r left03:51
← bglusman left03:52
← Slix` left03:54
→ HyperNerdV2_ joined03:54
→ maletor joined03:55
→ dfr|sick joined03:56
← HyperNerdV2 left03:56
HyperNerdV2_ → HyperNerdV203:56
← thiago left03:57
→ hagabaka joined03:58
→ jameswarren joined04:00
→ harish joined04:00
← hagabaka left04:00
← BLZbubba left04:00
← HyperNerdV2 left04:00
→ jackh joined04:01
→ hagabaka joined04:02
→ BLZbubba joined04:02
→ peppyheppy joined04:02
← tharkun left04:03
→ aspotashev joined04:04
← maletor left04:06
← EvanDotPro left04:08
→ tharkun joined04:09
← edude03 left04:10
→ edude03 joined04:10
← parcs` left04:11
← jackh left04:13
← kukks left04:13
← peppyheppy left04:14
← edude03 left04:15
→ edude03 joined04:15
→ uu joined04:15
→ peppyheppy joined04:16
← sdeboey left04:17
← tharkun left04:18
← d3vic3 left04:18
→ tharkun joined04:18
→ EvanDotPro joined04:20
← stkrzysiak left04:22
→ d3vic3 joined04:22
← tmcmahon left04:23
← tharkun left04:23
← madewokherd left04:23
→ tharkun joined04:23
← SeySayux left04:24
← edude03 left04:24
→ Chib joined04:25
→ edude03 joined04:25
← arvind_khadri left04:25
← ISF_ left04:26
← Radium left04:27
← Bass10 left04:28
→ ISF joined04:29
← edude03 left04:29
→ skylamer` joined04:29
→ edude03 joined04:30
→ Bass10 joined04:30
→ Radium joined04:31
← Bass10 left04:32
→ Bass10 joined04:32
→ `Assassin joined04:33
→ SeySayux joined04:34
← tharkun left04:35
← jesse left04:36
→ helgikrs joined04:38
→ tharkun joined04:40
← dr_lepper left04:41
→ jesse joined04:42
← Bass10 left04:46
← chronos left04:48
variable what makes git svn clone so slow and large compared to svn checkout ?04:48
← BenL left04:51
← mister_m left04:51
offby1 variable: git clone pulls down the entire history of a repository, whereas svn checkout pulls down just one revision04:52
variable hm, true04:52
offby1 although I haven't found git repositories to be notably larger than svn checkouts.04:52
variable bit git svn clone is larger than the svn repo itself04:52
offby1 _despite_ containing vastly more information04:52
texasmynsted variable, also know that once down you have an entire repository local with git. (not the case with svn)04:53
offby1 that's surprising04:53
variable offby1: the entire repository is 14GB04:53
offby1 that the git svn clone is notably bigger than the svn repository.04:53
variable my current git svn clone of _just_ head04:53
offby1 can't explain that.04:53
variable is 17GB04:53
offby1 Perhaps it needs to be gc'd04:53
← dreiss left04:53
variable texasmynsted: I meant the svn repository, not svn checkout, sorry for the confusion04:54
texasmynsted oh04:54
variable texasmynsted: 14GB vs 17GB04:54
:'(04:54
offby1: texasmynsted: if I want to minimize size what can I do other than 'git gc' ?04:55
texasmynsted yeah, garbage collection might be needed?04:55
variable this is the initial checkout04:55
(took over 3.5 days)04:56
milki o.O04:56
variable hi milki04:56
milki 3.5 day checkout sounds painful04:56
texasmynsted oh, come on.04:57
variable milki: yeap04:57
texasmynsted: hm?04:57
milki whats in the repo?04:57
texasmynsted 3.5 days?04:57
variable milki: ports tree04:57
texasmynsted: yes04:57
milki fbsd?04:57
variable milki: yes04:57
texasmynsted you have a 2400 baud modem or something?04:57
milki go to #bsdports04:57
variable texasmynsted: nope04:57
milki and yes, the ports tree is that big no matter what04:58
texasmynsted ok, well I can see your interest in reducing the size04:58
variable milki: the entire svn repo is 14GB, the git svn checkout is 17GB :)04:58
→ BenL joined04:58
milki the svn repo was created with cvs2svn04:58
variable texasmynsted: I need to upload this to somewhere else for others to use04:58
fwiw, if I checkout out with my username04:58
if I upload it to the world04:58
milki and the svn repo is outdated04:59
variable will that matter?04:59
texasmynsted like github?04:59
variable texasmynsted: well, I was going to upload the .git folder as a seed for others to use to fetch from04:59
← gtrak` left04:59
variable so not cloning the entire repo, just the .git folder04:59
texasmynsted good luck. g'night all04:59
→ gtrak` joined05:00
teotwaki_ fyi, cbreak and others, last night I asked for information regarding some branch -- turns out I had added the wrong repo URL as suspected. Stupid me. Apologies for wasting your time.05:00
milki variable: fbsd commiters agree05:00
you are crazy05:00
:P05:00
← marcuy left05:00
milki so why are you doing this?05:01
variable offby1: texasmynsted: also, if the url changes later, will I be able to do what is essentially svn relocate ?05:01
milki: fun05:01
also, I'm testing the use git on the repo05:01
← nutron left05:02
variable milki: please ask the committer what exactly is wrong with git svn05:02
milki why dont you go to #bsdports and ask yourself05:02
variable milki: I don't do efnet05:03
milki and i dont relay messages05:03
variable :'(05:03
milki you do realize theres ongoing effort to optimize the repo for svn05:03
← tharkun left05:03
variable milki: oh really?05:03
milki and once thats complete, there will be an svn2git mirror05:03
variable right, but you can't commit with that mirror05:04
milki the current git mirror is suspended becaise of issues with cvs2git05:04
you cant commit if you arent a committer anyways05:04
variable milki: good to know05:04
milki if you are deving ports, theres no need to mirror the entire ports tree05:04
variable really? what if I want to touch a large number of files?05:04
milki you want to test out your changes05:05
checkout redports05:05
variable yaya05:05
I have a redports account05:05
milki thats good enough -.-05:05
you are using someone else's tinderbox essentially05:05
→ HyperNerdV2 joined05:06
variable yaya05:06
milki you yourself dont need the entire history of the cvs repo to dev ports05:06
>.>05:06
variable milki: for my work, I often do05:06
milki really?05:06
variable yeah05:06
milki whats in the history thats useful?05:06
variable infrastructure history05:06
and key ports05:06
milki doesnt understand05:07
variable milki: whenever I'm playing with the xorg ports or more recently the emulation@ ports05:07
it helps to know the history of the files05:07
I could just cvs co them05:07
or svn log them05:07
but git is faster ;-)05:07
milki: also, I'm doing this mostly as a test to make sure it works05:08
for a possible eventual git cutover (a few years from now?)05:08
milki interesting05:08
→ xsbeats joined05:09
→ tharkun joined05:09
← barrbrain_ left05:09
→ robbyoconnor joined05:11
milki shakes fist at variable05:12
variable milki: ohai05:12
variable spills milki05:13
← krz left05:13
→ apok joined05:14
milki ok then05:16
back to the original question05:16
on the svn2git then...05:16
variable milki: this is an experiment to make sure it works05:16
as there _will_ be people that want to use it05:16
milki i thought it was gonna be the de facto method once the tree was migrated to svn05:17
← tharkun left05:17
milki as in there would be the main svn repo and a git mirror of it that was synced05:17
→ tharkun joined05:17
← eno left05:18
variable milki: yes, this is a) for committers b) the seed for that mirror05:18
milki oo05:18
→ eno joined05:20
← eno left05:20
→ eno joined05:20
← Sahua left05:20
milki a git checkout is the smallest size of the repo05:21
i dont think theres anything else you can do...05:21
are there binaries in the repo?05:21
variable don't think so05:21
← tharkun left05:21
← teotwaki_ left05:22
→ tharkun joined05:22
milki maybe it would be easier to provide a zfs snapshot05:23
well, only the initial import would take 3.5 days05:23
everyone else wouldnt have that problem05:23
so the question is just trying to minimize the repo size huh05:24
→ irqq joined05:24
milki variable: how long does a git clone of that repo take? :P05:24
variable milki: have not yet tried05:25
milki hopefully its faster...05:25
milki doesnt particularly like svn2git05:26
variable likely05:26
milki maybe the history is just not very optimial in the conversion?05:26
variable no network connection :)05:26
← ISF left05:27
→ Sigma joined05:28
→ dreiss joined05:29
← MestreLion left05:32
← tharkun left05:33
→ tharkun joined05:33
→ Elotero joined05:34
→ waaron joined05:35
← Elotero left05:37
← SegFaultAX|work left05:40
→ aspotashev| joined05:41
← aspotashev left05:42
← Radium left05:43
← tharkun left05:45
→ maleknet joined05:46
→ _root_ joined05:47
→ Radium joined05:47
→ tharkun joined05:50
milki theres also ongoing work to reduce the index size05:52
that might help in reducing the size of the repo05:52
and using git directly on svn05:53
in gsoc201205:53
http://www.google-melange.com/gsoc/project/google/gsoc2012/flyingflo/1200205:53
http://www.google-melange.com/gsoc/project/google/gsoc2012/tgummerer/1200105:54
← tharkun left05:54
→ Sahua joined05:56
→ devincea_ joined05:57
→ guidov joined05:57
→ giallu joined05:59
milki http://git-scm.com/about/small-and-fast05:59
if you look at the bottom05:59
it shows a git clone larger than the svn checkout06:00
→ tharkun joined06:00
milki variable: how long does an svn checkout take?06:02
← JohnyBaby left06:03
→ JohnyBaby joined06:05
← dfr|sick left06:05
→ mishok13 joined06:06
→ Linkas joined06:07
← tharkun left06:07
→ tharkun joined06:07
← harish left06:08
→ Araxia joined06:11
← tharkun left06:12
→ tharkun joined06:13
← Goplat left06:16
→ mikepack joined06:16
← mikepack left06:18
→ mikepack joined06:18
→ fuzzybunny69y joined06:20
← thorbjornDX left06:21
→ beautifulmind joined06:22
→ magfv joined06:22
→ LongBeach joined06:22
← beautifulmind left06:23
← mikepack left06:23
← OOPMan left06:24
→ jackh joined06:25
← HyperNerdV2 left06:25
→ Kicer86 joined06:26
→ xiangfu joined06:26
← jackh left06:27
→ sattu94 joined06:28
→ marthinal joined06:29
← satyanash left06:29
sattu94 → satyanash06:29
← Davey left06:29
→ beautifulmind joined06:31
← tharkun left06:32
← mishok13 left06:32
→ sattu94 joined06:34
← satyanash left06:34
sattu94 → satyanash06:34
← waaron left06:36
← Kicer86 left06:36
→ Kicer86 joined06:36
→ jamesma joined06:37
→ tharkun joined06:37
← jamesma left06:38
→ iamntz joined06:39
← eno left06:39
← adamm left06:40
← Sigma left06:41
→ eno joined06:41
→ ehsan joined06:42
← iamntz left06:43
→ dasha joined06:44
→ iamntz joined06:44
← iamntz left06:44
→ iamntz joined06:44
← dasha left06:44
← fuzzybunny69y left06:45
← tharkun left06:47
→ tharkun joined06:47
← pppaul left06:47
→ pppaul joined06:48
→ harish joined06:48
→ umc joined06:48
→ Dave^| joined06:51
← Radium left06:53
← tharkun left06:54
→ tharkun joined06:54
← neurodrone left06:56
→ HyperNerdV2 joined06:56
→ Radium joined06:57
→ whitman joined06:59
→ Fullmoon joined07:00
← tharkun left07:01
→ mishok13 joined07:01
→ tharkun joined07:01
→ drizzd joined07:02
→ fling joined07:02
fling what does this mean? > fatal: ambiguous argument 'remote': unknown revision or path not in the working tree.07:03
Use '--' to separate paths from revisions07:03
frogonwheels fling: what command are you using07:04
← igotnolegs left07:06
← waveclaw left07:07
→ matzbitz joined07:08
← tharkun left07:09
← pppaul left07:09
→ pppaul joined07:10
fling frogonwheels: git show remote origin07:11
← subbyyy left07:11
frogonwheels fling: what did you mean that to do?07:11
→ ToBeFree joined07:12
→ dr_lepper joined07:12
← dr_lepper left07:12
→ dr_lepper joined07:12
→ vpdn joined07:13
fling frogonwheels: she is trying to list remote branches07:14
← pppaul left07:14
→ tharkun joined07:14
→ pppaul joined07:15
← beautifulmind left07:15
frogonwheels fling: you mean git branch -r or do you mean git ls-remote origin ?07:15
vpdn Anyone knows a way to add a non unique label to a commit? (Something like a gmail label)07:15
frogonwheels vpdn: what are you trying to achieve?07:16
← HyperNerdV2 left07:16
vpdn I'm running a CI and would like to upload a build based on some condition.07:18
Could be done with having a special keyword in the git commit and a hook, but would like to separate the two things.07:19
fling frogonwheels: idk, sorry, now she says something about her logic and the command is right one. thanks for help :p07:19
← Lgb left07:20
→ SegFaultAX|work joined07:20
vpdn My idea was something along the line: "If the commit has a label 'betabuild' then upload to the server for beta testing".07:20
→ Lgb joined07:21
teotwaki vpdn: if you're running CI, you shouldn't have to differentiate any kind of commit07:23
← dr_lepper left07:23
→ dr_lepper joined07:23
← dr_lepper left07:23
→ dr_lepper joined07:23
teotwaki vpdn: the way we do it here is that we build any commit (or top of any commit), regardless of status or tags07:23
vpdn: the only difference being that the branch/repository defines whether it's a dev or official release.07:24
← MissionCritical left07:24
vpdn teotwaki: we are building all commits. As a post commit step, there's an upload to the test servers for the beta testers to try out (we're building mobile apps).07:24
← w09x left07:24
teotwaki You don't have beta version numbers, in the source code?07:25
vpdn teotwaki: and the post commit step is the thing that I'm trying to optimize. For one we don't want testers to get irrelevant commit (e.g. TODO changes) and also the upload is quite bandwidth consuming.07:25
build versions are synchronized with the CI build numbers.07:27
→ w09x joined07:27
← tharkun left07:27
← diegoviola left07:28
→ adfsadf joined07:29
→ fyolnish joined07:32
→ tharkun joined07:33
vpdn teotwaki: Problem with adding version numbers to the code arises when you have a merge because two people increased the build number at the same time.07:33
teotwaki yeah sorry, meant a define flag, or at least some way to get the version number from the generated binaries.07:34
vpdn the merge will succeed but it's not obvious anymore by just looking at the build number that both were meant as two separate beta builds.07:34
teotwaki anyway, why would you want to multi tag?07:35
I mean, why not simply tag -d and retag?07:35
← adfsadf left07:35
teotwaki Another way to do things would be to require a manual promotion after the build07:36
For example, once the build has completed, you can use some kind of interface to promote the build to beta07:36
→ thorbjornDX joined07:36
teotwaki Rather than asking devs to tag something in advance, not knowing if it is actually going to build or not07:36
vpdn good point07:36
← tharkun left07:37
← khmarbaise left07:38
→ tharkun joined07:38
← Sahua left07:40
← thorbjornDX left07:42
← tharkun left07:42
→ thorbjornDX joined07:42
← brokenshakles left07:43
→ somya joined07:43
→ brokenshakles joined07:44
→ glcrazy joined07:44
→ MissionCritical joined07:45
neersighted → neersighted|AFK07:47
→ tharkun joined07:47
→ j416 joined07:48
vpdn just occurred to me that I could also track the set through another 'beta-tester' branch. The CI would build both branches but only upload changes in the 'beta-tester' branch.07:50
← tharkun left07:51
← aspotashev| left07:53
← dr_lepper left07:53
→ dr_leppe1 joined07:53
→ llaumgui joined07:54
→ francisfish joined07:54
→ JohnBat26 joined07:55
→ tharkun joined07:57
→ ferdna joined07:57
← Radium left07:57
→ raatiniemi joined07:58
← brokenshakles left07:58
→ timoxley joined07:59
→ brokenshakles joined08:00
→ Radium joined08:00
← mishok13 left08:00
← Mowsh left08:00
← brokenshakles left08:02
← llaumgui left08:03
← tharkun left08:03
← devincea_ left08:04
→ davidbanham joined08:05
→ mike201 joined08:06
← davidbanham left08:06
← kcj left08:06
← dr_leppe1 left08:07
← teotwaki left08:07
gitinfo set mode: +v08:08
→ tharkun joined08:08
mike201 Hey there, by using the no-ff flag with git merge and after deleting a branch, is it possible to lose any commit data when git runs it's cleaning function?08:08
→ syphar joined08:08
← ferdna left08:09
ndim mike201: "it's [sic] cleaning function" being what? "git clean"? "git purge"? "..."?08:10
mike201 ndim: i'm not entirely sure, i thought git runs some cleaning service on dangling commits every so often, and not sure if using the above method creates any08:12
ndim mike201: Ah, that.08:14
Yes, that will eventually delete the commits that were only on the deleted branch (and not (merge into) any other branch).08:14
← w09x left08:15
ndim Not immediately, though.08:15
→ davidbanham joined08:15
ndim (unless you force git to)08:15
← tharkun left08:15
← davidbanham left08:16
mike201 ndim: ah thanks.08:19
→ w09x joined08:19
→ arietis joined08:19
→ brokenshakles joined08:20
→ tharkun joined08:21
← w09x left08:22
← apok left08:22
← tharkun left08:25
→ RaceCondition joined08:25
→ tharkun joined08:26
← xaka left08:31
← tharkun left08:32
→ Sigma joined08:32
→ tiglog joined08:32
→ Element9 joined08:32
← Khisanth left08:33
→ Khisanth joined08:34
→ Xizor joined08:34
→ bglusman joined08:35
→ tharkun joined08:37
→ taipai joined08:37
→ Lord_DeathMatch joined08:38
← bglusman left08:40
← vpdn left08:40
→ ndngvr_ joined08:40
← tharkun left08:42
← Linkas left08:43
← rurufufuss left08:44
← harish left08:46
→ d0k joined08:47
→ tharkun joined08:48
→ Superlokkus joined08:51
→ shruggar joined08:53
← LongBeach left08:54
→ dr0id joined09:00
← magfv left09:01
← eno left09:02
← raatiniemi left09:02
← xiangfu left09:02
→ eno joined09:04
← bolt left09:04
← aarond_123a left09:06
→ piecyk joined09:07
← maleknet left09:07
→ davidbanham joined09:08
→ _iron joined09:08
← az left09:09
← somya left09:10
→ eijk joined09:13
← Superlokkus left09:13
→ az joined09:13
← tharkun left09:14
← SegFaultAX|work left09:14
→ fishguy joined09:14
← davidbanham left09:14
→ bolt joined09:17
← peppyheppy left09:18
→ tharkun joined09:19
← kadoban left09:21
← Infiltrator left09:24
← tharkun left09:27
→ tharkun joined09:27
→ blacktulip joined09:31
→ JohnyBaby_ joined09:32
← JohnyBaby left09:32
→ resmo_ joined09:36
→ sattu94 joined09:37
→ TomyLobo joined09:39
← uu left09:40
← satyanash left09:40
neersighted|AFK → neersighted09:41
→ wbednarski joined09:42
→ Shusshu joined09:45
neersighted → neersighted|AFK09:45
→ psoo joined09:46
← taipai left09:48
← codeM0nK3Y left09:48
→ uu joined09:50
→ Ripp__ joined09:50
← uu left09:50
→ uu joined09:51
→ jceb joined09:52
→ nailora1 joined09:53
gitinfo set mode: +v09:53
→ alexukf joined09:53
nailora1 congratulations on the new site! is there any public information who paid/drove the the redesign?09:54
→ mbroeker joined09:54
← Shusshu left09:56
→ jackh joined09:57
→ khmarbaise joined09:57
cmn it was made by github09:57
← Tinti left09:58
cmn all you have to do to figure it out is click on the source link09:58
→ flavius joined10:00
← mike201 left10:00
← timoxley left10:01
← tharkun left10:01
← Radium left10:02
→ glennpratt joined10:02
→ tharkun joined10:02
← khmarbaise left10:03
→ harish joined10:04
→ Radium joined10:05
→ devinceartas joined10:06
→ hachque joined10:07
← tharkun left10:09
→ tharkun joined10:09
→ arvind_khadri joined10:14
← umc left10:15
→ thelazydeveloper joined10:16
← tharkun left10:16
→ tharkun joined10:16
→ FreeArtMan joined10:17
→ gjoseph joined10:17
← tiglog left10:18
→ kcj joined10:19
← tharkun left10:20
← Ripp__ left10:20
← ping-pong left10:20
← psoo left10:21
→ theCole joined10:22
← rux left10:24
→ tharkun joined10:26
← devinceartas left10:28
← gjoseph left10:28
← arvind_khadri left10:31
← ipalaus_ left10:32
→ raatiniemi joined10:33
→ ivan12345 joined10:33
→ ivan12346 joined10:33
→ crash\ joined10:34
→ ping-pong joined10:34
→ Tinti joined10:35
← ivan12348 left10:36
← ivan12347 left10:36
← pppaul left10:38
→ ipalaus_ joined10:38
→ adamcooke joined10:38
→ pppaul joined10:38
← dreiss left10:41
← francisfish left10:41
← theCole left10:43
→ devinceartas joined10:45
→ gjoseph joined10:45
← jceb left10:46
→ gusnan joined10:47
→ jceb joined10:47
← tharkun left10:49
→ tharkun joined10:49
← lausi left10:50
→ lausi joined10:51
← RaceCondition left10:53
← gjoseph left10:53
← jceb left10:54
← jackh left10:57
← drizzd left10:57
← wbednarski left10:57
→ wbednarski joined10:58
→ n1x joined10:58
← devinceartas left11:00
← n1x left11:02
→ OOPMan joined11:03
← ipalaus_ left11:04
→ n1x joined11:05
→ Reznov_ joined11:07
→ Russell^^ joined11:08
← Reznov_ left11:08
← n1x left11:09
→ mklappstuhl joined11:11
→ n1x joined11:11
→ cenuij joined11:17
← cenuij left11:17
→ cenuij joined11:17
← tharkun left11:18
← JohnBat26 left11:18
→ txomon joined11:21
← txomon left11:21
→ txomon joined11:21
→ francisfish joined11:22
→ theCole_ joined11:23
→ tharkun joined11:23
← txdv left11:24
→ txdv joined11:24
→ syphar_ joined11:25
← syphar_ left11:25
← n1x left11:25
→ devinceartas joined11:26
→ theCole__ joined11:27
← OOPMan left11:29
← theCole_ left11:29
← syphar left11:29
→ ColKurtz joined11:32
← eightyeight left11:32
→ Sky[x] joined11:32
ColKurtz How should I approach setting up my VPS so that deploying code to it is as simple as "git push production"11:32
Sky[x] hi11:32
gitinfo Sky[x]: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.11:32
ColKurtz ?11:32
→ eightyeight joined11:32
→ arvind_khadri joined11:33
ColKurtz I know that you're not supposed to push to non-bare repositories11:33
← arvind_khadri left11:33
→ arvind_khadri joined11:33
← dmilith left11:33
← tharkun left11:33
→ dmilith joined11:33
bremner ColKurtz: there is !web11:34
gitinfo ColKurtz: [!deploy] Git is not a deployment tool, but you can build one around it for simple environments. Here is an example hook to get you started: https://github.com/EugeneKay/scripts/blob/master/bash/git-deploy-hook.sh11:34
ColKurtz ty for link11:34
→ SkAzZ joined11:34
→ SkAzZ1 joined11:36
→ bglusman joined11:36
← shruggar left11:36
← flavius left11:38
→ tharkun joined11:39
← SkAzZ left11:40
→ khmarbaise joined11:40
← angelsl left11:40
← bglusman left11:41
→ shruggar joined11:41
← glennpratt left11:43
← devinceartas left11:43
← ToBeFree left11:44
→ Vile joined11:46
← arvind_khadri left11:46
← eijk left11:47
← tharkun left11:48
→ tharkun joined11:48
→ timoxley joined11:49
← timoxley left11:50
← Vile left11:52
← lausi left11:53
→ lausi joined11:53
← shruggar left11:55
← tharkun left11:56
← hachque left11:57
← adamcooke left11:57
← thebignoise left11:59
→ arvind_khadri joined12:01
← Xizor left12:01
→ tharkun joined12:02
← babilen left12:04
→ Chillance joined12:05
→ jargon- joined12:05
→ dvide joined12:07
→ aristidesfl joined12:07
→ parcs` joined12:09
→ davidbanham joined12:09
→ madewokherd joined12:11
← arvind_khadri left12:12
← tharkun left12:17
→ chris2 joined12:19
→ devinceartas joined12:21
→ TTSDA joined12:22
→ tharkun joined12:22
← TTSDA left12:24
→ theCole joined12:25
→ TTSDA joined12:25
→ syphar joined12:26
→ polysics joined12:26
polysics hi12:26
gitinfo polysics: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.12:26
← tharkun left12:26
polysics what's the proper way to update a branch i am working on with the contentfrom another i know has been updated remotely, but is not updated locally?12:27
← raatiniemi left12:27
← theCole__ left12:27
← TTSDA left12:27
→ tharkun joined12:27
→ mishok13 joined12:28
raek polysics: first you can run `git fetch origin` (where origin is the name of the remote repo with the new stuff)12:29
← Russell^^ left12:29
← BenL left12:29
→ xiangfu joined12:29
→ jceb joined12:30
→ BenL joined12:30
→ Russell^^ joined12:30
polysics eventuall ispecifying a branch, correct?12:30
→ TTSDA joined12:30
raek then you can compare the branches using log and diff12:30
polysics if i know i want what's in "develop", for example?12:30
→ arvind_khadri joined12:31
← arvind_khadri left12:31
→ arvind_khadri joined12:31
polysics say it's some new model, thus not conflicting with anything i have12:31
git merge develop while on my branch?12:31
→ raatiniemi joined12:31
raek origin/develop12:31
polysics great, thanks!12:31
raek unless you have a local develop branch that tracks origin/develop12:31
← pppaul left12:32
→ intripoon joined12:32
→ pppaul joined12:32
raek before merging you might want to do `git log --left-right my-local-branch...origin/develop`12:34
→ airborn joined12:34
raek this will give you an idea of what you are going to merge12:34
← xiangfu left12:34
→ babilen joined12:35
← gusnan left12:35
← intripoon_ left12:35
→ traatiniemi joined12:36
→ gusnan joined12:36
← TTSDA left12:36
→ TTSDA joined12:37
← pppaul left12:38
→ pppaul joined12:40
← arvind_khadri left12:40
→ Bass10 joined12:41
← tharkun left12:44
← marthinal left12:44
→ xiangfu joined12:47
→ acl_ joined12:49
→ tharkun joined12:49
→ joepie91 joined12:50
← Russell^^ left12:50
← pppaul left12:52
→ pppaul joined12:52
← acl_ left12:53
← fisted left12:53
→ patp joined12:54
→ fisted joined12:54
← `Assassin left12:54
← alexukf left12:55
← patp left12:56
→ fgh joined12:58
← Bass10 left12:59
← devinceartas left13:00
→ sdeboey joined13:02
→ Myhro joined13:04
→ Destos joined13:06
← kcj left13:06
→ angelsl joined13:08
← skylamer` left13:14
← traatiniemi left13:15
→ izissise joined13:17
→ jpieper joined13:19
← Lgb left13:20
← cagedwisdom left13:20
→ Lgb joined13:21
← davidbanham left13:21
← tharkun left13:22
→ alexukf joined13:25
← fgh left13:26
→ davidbanham joined13:26
← juvenal left13:26
← hagabaka left13:28
← ehsan left13:28
→ aspiers joined13:28
→ tharkun joined13:28
← Myhro left13:28
aspiers anyone know how can I git svn clone this repo with a rather unconventional layout of only branches at the top level? http://themes.svn.wordpress.org/panorama/13:29
→ stkrzysiak joined13:29
→ mgpcoe joined13:29
→ ehsan joined13:31
← tharkun left13:34
→ tharkun joined13:34
← theCole left13:35
→ sdaone joined13:36
← iamntz left13:37
→ iamntz joined13:37
→ munichlinux joined13:38
← tharkun left13:39
→ Yuuhi joined13:42
← Tinti left13:43
→ hwrdprkns joined13:44
→ apok joined13:44
→ tharkun joined13:45
→ Bass10 joined13:47
→ ronniy joined13:47
← matzbitz left13:48
← d0k left13:48
← Bass10 left13:48
→ Bass10 joined13:49
→ d0k joined13:49
→ marcuy joined13:49
← mklappstuhl left13:49
→ ben225 joined13:52
← munichlinux left13:53
→ neurodrone joined13:55
→ munichlinux joined13:56
→ Fiek joined13:56
gitinfo set mode: +v13:57
Fiek → Icyberg13:57
← ben225 left13:59
→ rsewill joined13:59
Icyberg Hey, been getting a "rejected due to non-fast-forward" or something, anyone have any idea what could be wrong?13:59
tried git pull13:59
and git push -force (all with no avail)13:59
aspiers it's --force I think14:00
or -f14:00
cbreak you have to make it fast forward by merging or similar14:00
← rsewill left14:00
cbreak or kill remote history by forcing for the branch that is non-fast-forward14:00
aspiers you'll lose history if you force14:00
Icyberg Sorry, --force14:00
aspiers you probably shouldn't use --force unless you really understand what's going on14:01
Icyberg cbreak: so "git merge"?14:01
cbreak if you want more detailed advice, pastebin everything you typed and the output14:01
Icyberg: no...14:01
Icyberg err, I'm kinda new to git14:01
xD14:01
aspiers definitely don't --force then14:01
pastebin14:01
Icyberg I'm using Zend Studio + EGit so I'm not exactly what commands were executed. Neither pushing in Zend or from the CLI works.14:02
cbreak not interesting.14:02
aspiers probably best to learn with the CLI first14:02
Icyberg cbreak, aspiers: what exactly would you like me to pastebin?14:02
:P14:02
← Nibato left14:02
cbreak everything you typed and everything git replied14:03
Icyberg ah, okay14:03
cbreak for example the git push that complained about non-ff14:03
ikrabbe hehe, once you learned CLI you will hate the other interfaces14:03
aspiers ikrabbe: I disagree :)14:03
← ronniy left14:04
Icyberg cbreak, aspiers http://pastie.org/386386214:04
← txomon left14:04
flexd Uh, is anyone else experiencing git being slow on windows?14:05
cbreak that makes no sense.14:05
aspiers Icyberg: why did you --force?14:05
flexd Just installed git for the first time on a window machine (at work), I usually use osx14:05
cbreak flexd: that's normal. Windows is that way.14:05
jast git has a strong tendency to be slow on windows14:05
aspiers cbreak: :)14:05
Icyberg aspiers: that was output from earlier14:05
jast windows doesn't lend itself very well to the optimizations git uses14:05
flexd git status is taking time from the time I press enter to something happening :(14:05
Icyberg before i joined this channel14:05
cbreak Icyberg: git remote -vv and git branch -a, pastebin14:05
aspiers IcePic: but you said --force didn't work14:06
flexd cbreak: Glad I do not use windows normally then14:06
aspiers Icyberg: : but you said --force didn't work14:06
IcePic: sorry ignore14:06
Icyberg aspiers: it didn't. it reported success but nothing was changed14:06
aspiers Icyberg: the remote was changed14:06
cbreak Icyberg: git branch -avv even14:06
Icyberg aspiers: take a look at the repo (URL is in the output)14:06
aspiers Icyberg: that's the whole point of a push14:06
→ Myhro joined14:07
→ ronniy joined14:07
cbreak Icyberg's push ref spec seems broken14:07
aspiers yeah14:07
Icyberg: what's that refs/for/refs/heads/master nonsense?14:07
jast Icyberg: refs/for/refs/... doesn't show in github's branch list14:07
proper branches are of the form refs/heads/foo14:08
Icyberg O_o14:08
again; Zend Studio/EGit is doing all that for me14:08
jast the /for/ thing reminds me of gerrit14:08
→ Nibato joined14:08
jast it's not something git does normally14:08
Icyberg jast: germit?14:08
cbreak maybe that egit/zend thing is garbage then.14:08
jast gerrit is a code review system that sits on top of git14:08
Icyberg oh..14:09
jast my one personal experience with egit wasn't so good ;)14:09
Icyberg well I guess I don't *need* an IDE14:09
could stick to the CLI14:09
aspiers Icyberg: bad idea to mix IDE/CLI when you don't understand everything the IDE is doing w.r.t. git14:09
Icyberg and a text editor14:09
cbreak anyway, give me the output I requested14:10
aspiers yeah, that's a much better way to learn git14:10
cbreak git branch -avv and git remote -vv14:10
aspiers Icyberg: once you have a solid understanding, then move to the higher level tools14:10
flexd Icyberg: You can still use the IDE for code14:10
← blacktulip left14:10
jast anyway, it might be possible to fix the config on the command line and have the IDE thing work14:10
flexd Just do git stuff in CLI14:10
← Lord_DeathMatch left14:10
Icyberg cbreak: http://pastie.org/386388814:10
→ gjoseph joined14:10
→ glennpratt joined14:10
← francisfish left14:10
cbreak original? definitely weird14:10
and no mention of this refs/for thing14:11
wonder why push wanted to screw with that.14:11
check your .git/config file14:11
aspiers cbreak: original is probably from filter-branch14:11
Icyberg I'll just recreate everything14:11
cbreak he said he'd be a git newbie14:11
Icyberg and refork the original repo14:11
:P14:11
cbreak I don't think he filterbranched14:11
Icyberg and not use an IDE14:12
aspiers cbreak: the IDE might have14:12
Icyberg: good plan :)14:12
cbreak an IDE that filter-branches?14:12
→ berserkr joined14:12
jast an IDE that filter-branches deserves a slow and painful death14:12
cbreak what's next? An IDE that resets --hard when you save files?14:12
aspiers haha14:12
jast not that eclipse doesn't for other reasons :}14:12
→ francisfish joined14:12
aspiers well, come up with a better explanation then :) I know filter-branch creates original/* for sure14:12
cbreak it's a remote14:13
jast this is about a *remote* called 'original'14:13
not about refs/original/14:13
aspiers ah14:13
ah well, that's almost certainly the IDE then14:13
cbreak I'd be interested to see the .git/config14:13
aspiers it's own idea of 'origin'14:13
s/it's/its/14:13
cbreak maybe it has some screwed up push ref spec14:13
aspiers anyone here use emerge with merge-tool?14:15
Icyberg okay, just recreated the repo, guess i'll try set it up with git14:15
→ freakabcd joined14:15
freakabcd hi all.. What is the equivalent of "svn revert myfile.txt" ?14:16
cbreak Icyberg: recreate?14:16
freakabcd i accidentally did: git rm myfile.txt14:16
cbreak just git clone again...14:16
Icyberg cbreak: i meant reforked14:16
cbreak freakabcd: git checkout HEAD -- filename14:16
reforked? Stupid...14:16
freakabcd how do i get it back on disk? i tried git reset HEAD --14:16
← gusnan left14:16
cbreak your problem was just client side14:16
freakabcd thanks cbreak14:17
aspiers cbreak: no need to call a newbie stupid ...14:17
Icyberg wow14:17
Freenode hasn't changed a bit14:17
cbreak freakabcd: checkout in file mode is for index manipulation, not file system manipulation14:17
aspiers: no need to not call stupid acts stupid14:17
I don't blame newbies for doing stupid things14:18
but they have to know that it is stupid so they learn :)14:18
← eno left14:18
cbreak never heard that people learn from mistakes?14:18
they don't learn from mistakes if they don't realize them...14:18
aspiers cbreak: it's not stupidity, it's ignorance14:18
cbreak: there's a difference14:18
cbreak nope.14:18
The act is stupid.14:18
jast at any rate, 'stupid' isn't the most diplomatic word there is :)14:18
Icyberg The big issue I had with Freenode was the majority of active users flamed noobies for not understanding something, or because they were "stupid". I left Freenode years back just because of it.14:19
cbreak I don't call you stupid.14:19
aspiers jast: exactly. there are more tactful ways of putting it14:19
cbreak I call what you did stupid14:19
jast but there is a difference between calling a person stupid and calling an action stupid14:19
cbreak if you really want you can refork all day14:19
but it won't fix your problem14:19
aspiers cbreak: actually it was entirely ambigious what/who you were calling stupid14:19
→ eno joined14:19
cbreak it's client side.14:19
I am quite sure it's an issue in your repository's .git/config file14:20
→ amigojapan joined14:20
jast reforking is a windows-flavoured solution: have you tried turning it off and on again? :}14:20
cbreak reminds me of reinstalling the operating system14:21
jast if it works, fine, but there's no enlightenment down that road14:21
Icyberg cbreak: I didn't/don't even understand what the problem is, you could've just left out the "stupid" bit and told me exactly what was happening14:21
jast and enlightenment does tend to make things easier later on14:21
cbreak I told you what the problem is14:21
← mgpcoe left14:22
jast the problem is that some piece of software, presumably z14:22
(grrr)14:22
← uu left14:22
jast ... "zend studio/egit", messed up the local repository config14:22
Icyberg I know that, I mean what it was doing wrong14:22
jast and that can be fixed by changing the config14:22
cbreak "I'd be interested to see the .git/config", "maybe it has some screwed up push ref spec"14:22
that sounds fairly unambiguous to me.14:23
jast man, different keyboard layouts just kill my productivity14:23
cbreak jast: yeah.14:23
jast at least I got a USB keyboard at work now14:23
but the insert/delete/home/end/pageup/down block is not where it's supposed to be14:23
cbreak I bought a new US keyboard for home use because switching back and forth from work was too time consuming14:24
← munichlinux left14:24
jast the manufacturer decided to cram some ACPI keys in there, and now the insert/... block is jammed right up to the cursor keys14:24
Icyberg cbreak: Well, since you wanted to see it I'll pastie it.14:24
jast I keep using the wrong keys and it's annoying, but at least less so than the notebook's keyboard that I had to use before14:24
and of course *this* notebook has yet another different layout14:24
Icyberg cbreak: http://pastie.org/386393714:25
jast I use German keyboards but set to the US layout... works great for me14:25
cbreak Icyberg: the push part in remote "original"14:25
→ subbyyy joined14:26
jast wow, that's so screwed up it's not even funny14:26
→ bitkiller joined14:26
→ shruggar joined14:26
cbreak jast: I used swiss german keyboard with US layout, it was similar but not the same as the real US keyboards14:26
small differences, but still14:26
jast git config --unset remote.original.push14:26
also maybe set push.default to upstream14:26
cbreak you can also edit the file directly14:27
if you know what you're doing :)14:27
→ Lord_DeathMatch joined14:27
jast sure, but easier to tell someone to copy&paste a command :)14:27
Icyberg jast: what's the command to set push.default to upstream?14:27
cbreak it's more educative to tell people what happens though.14:27
jast git config push.default upstream14:27
:)14:27
cbreak I'd do it --global14:28
(in fact, I did it global ages ago)14:28
Icyberg jast: should I try to push it now?14:28
jast what it does is this: "git push" will figure out the right remote and branch to push to for the currently checked out branch14:28
and if it can't, it'll ask you to tell it once where to look14:28
that way minimises potential for errors14:28
sure14:28
cbreak Icyberg: upstream info is set for you14:28
so you can do that14:28
otherwise, for new branches, you have to do a git push -u original branchname first14:29
Icyberg Pushed :)14:29
jast what your broken config said is this: whatever branch is checked out, write its contents to refs/for/... on the server14:29
not a particularly useful thing to do, obviously14:29
cbreak it was stupid...14:29
wonder why the IDE or what ever did that14:29
jast we may never know14:30
I kind of doubt that it's egit's fault14:30
Icyberg Zend built onto EGit14:30
so its probably something to do with that14:30
jast despite all my complaints about eclipse and java, at least shawn knows his git14:30
cbreak also, the + in the spec should have forced it anyway14:30
jast yeah, let's blame zend... that's a good heuristic in my opinion14:30
cbreak what's zend?14:31
Icyberg ^ :)14:31
jast makers of PHP14:31
cbreak eeew14:31
probably their fault.14:31
jast my point exactly :)14:31
Icyberg jast: how do I tag releases?14:32
cbreak git tag namehere14:33
jast with 'git tag', with -a or -s14:33
cbreak Icyberg: man git-tag14:33
gitinfo Icyberg: the git-tag manpage is available at http://jk.gs/git-tag.html14:33
jast look at the manpage for details14:33
of course, if anything isn't clear, feel free to ask14:33
← marcuy left14:36
→ Nathandim joined14:37
← davidbanham left14:37
→ bglusman joined14:37
→ txomon joined14:38
→ marcuy joined14:39
→ sq-one joined14:40
← drev1_ left14:42
← bglusman left14:42
← syphar left14:42
← harish left14:43
← gavin_huang left14:44
← twisted` left14:47
← aristidesfl left14:47
← raatiniemi left14:48
← glennpratt left14:49
→ harish joined14:50
→ raatiniemi joined14:50
→ aristidesfl joined14:50
→ adamcooke joined14:51
→ OOPMan joined14:51
→ Superlokkus joined14:51
← Vinnie_win left14:56
→ DofLeFou joined14:56
→ JohnyBaby joined14:57
← JohnyBaby_ left14:57
← shruggar left14:58
→ Vinnie_win joined14:58
→ torukawahata joined15:00
← angelsl left15:03
→ glennpratt joined15:04
→ dv310p3r joined15:05
→ angelsl joined15:05
→ Jackneill joined15:06
→ munichlinux joined15:08
→ nessundorma joined15:10
← JohnyBaby left15:11
→ JohnyBaby joined15:11
→ ZooLer joined15:18
→ pandeiro joined15:18
→ Vile joined15:18
← gjoseph left15:20
→ resmo__ joined15:22
sattu94 → satyanash15:22
← mbroeker left15:24
→ HyperNerdV2 joined15:24
→ drizzd joined15:24
← resmo_ left15:25
→ arvind_khadri joined15:25
← arvind_khadri left15:25
→ arvind_khadri joined15:25
→ pfFredd joined15:27
→ incog joined15:28
← amigojapan left15:30
→ gjoseph joined15:31
← aspiers left15:31
← dark_element left15:31
← francisfish left15:31
← wilmoore left15:32
← stkrzysiak left15:32
← alexukf left15:33
← arvind_khadri left15:34
→ arvind_khadri joined15:34
← arvind_khadri left15:34
→ arvind_khadri joined15:34
← munichlinux left15:34
← fyolnish left15:35
← flaguy48 left15:36
→ fyolnish joined15:36
← h0bbit left15:36
→ Per4u3e joined15:37
← airborn left15:38
→ wilmoore joined15:38
← arvind_khadri left15:38
← xiangfu left15:40
← Per4u3e left15:41
→ pe joined15:41
← sq-one left15:41
→ francisfish joined15:41
→ dreiss joined15:42
→ noqqe__ joined15:42
→ pfFred joined15:44
← pfFredd left15:44
← iamntz left15:45
→ QuantumQuanta joined15:45
← crhan left15:46
→ iamntz joined15:46
QuantumQuanta This is probably more of a workflow question, but I'm working on a little website that uses an API key to fetch some info. My key shouldn't be publicly revealed...is there a good way to keep the key around while I'm working but not accidentally include it in the repo?15:47
← fyolnish left15:47
QuantumQuanta I'm using Python, if that helps...15:47
→ fyolnish joined15:47
→ traatiniemi joined15:47
variable if I want to optimize repo space what should I run? 'git gc' exactly? With any special options?15:48
← khmarbaise left15:48
et so you want to stop yourself from 'git add <filethatincludeskey>'? you should be able to do that with hooks15:48
QuantumQuanta et: so maybe move the key to a different file on it's own and then import it into the main code file15:49
?15:49
variable QuantumQuanta: .gitignore as well15:49
or move the key into a file above the repo15:50
→ Sahua joined15:50
QuantumQuanta And then maybe include the file once with a dummy key just to demonstrate the format?15:50
→ JohnyBaby_ joined15:50
← francisfish left15:50
cmn read it from a configuration file or the environment15:50
→ wolog joined15:50
et QuantumQuanta: yes, but if you accidently 'git add' the key file that won't help, that's why i suggested a hook15:50
← JohnyBaby left15:52
→ williamcotton joined15:53
← aristidesfl left15:54
→ alexukf joined15:55
→ pantsman joined15:55
← pantsman left15:55
→ pantsman joined15:55
→ macmartine joined15:56
→ aristidesfl joined15:57
← nessundorma left15:57
← fyolnish left15:59
← _iron left15:59
EugeneKay cmn - regarding ! deploy from the other day; my apologies. I forgot that the bot was pointing to the link. One of these days I'll finish writing git-deploy properly....16:02
cmn no biggie, just bad timing16:03
→ Dragooon joined16:03
EugeneKay raek - same, but consider that my script's approach doesn't even need the deploy script to be on the same machine as the web server. All it cares about is rsync access to the DocRoot ;-)16:03
→ dr_diamond joined16:04
→ devinceartas joined16:04
← jceb left16:05
→ xaka joined16:08
→ yoklov joined16:09
← adamcooke left16:09
← patrick99e99 left16:10
← Element9 left16:10
← pe left16:10
← Radium left16:11
← polysics left16:12
→ ScottNYC joined16:15
→ RaceCondition joined16:17
→ crhan joined16:19
→ dr_lepper joined16:19
← dr_lepper left16:19
→ dr_lepper joined16:19
→ kadoban joined16:19
← osX64 left16:19
→ osX64 joined16:19
← JohnyBaby_ left16:20
← gjoseph left16:20
→ KoRaH joined16:21
← KoRaH left16:21
← dr0id left16:22
← freakabcd left16:22
→ khmarbaise joined16:24
→ flaguy48 joined16:25
→ mikepack joined16:26
→ ology joined16:26
→ Radium_ joined16:27
← devinceartas left16:28
← wolog left16:28
→ wolog joined16:29
← Destos left16:29
→ jmaya joined16:30
← jmaya left16:31
→ JohnyBaby joined16:31
← wilmoore left16:32
← glennpratt left16:33
→ airborn joined16:33
← BlindOracle left16:35
← xaka left16:35
→ gjoseph joined16:36
→ nadavoid joined16:38
→ adamm joined16:40
← Radium_ left16:42
← FreeArtMan left16:42
← guidov left16:47
→ twisted` joined16:48
← kadoban left16:48
← ScottNYC left16:51
← ben_alman left16:52
→ ben_alman joined16:53
← gjoseph left16:53
→ Radium joined16:53
→ AnkA_maviyesil joined16:55
→ gjoseph joined16:55
← nadavoid left16:56
← AnkA_maviyesil left16:56
→ dark_element joined16:57
→ BlindOracle joined16:59
← dark_element left16:59
← traatiniemi left17:00
→ markus_ joined17:00
← kerrick_ left17:00
→ glennpratt joined17:04
← Superlokkus left17:05
→ ph^ joined17:05
← BlindOracle left17:05
← harish left17:07
→ mklappstuhl joined17:07
→ BlindOracle joined17:07
→ yacc joined17:07
yacc Any standard way to locate the "root" of a git repository from inside a subdirectory?17:08
Basically the .git directory.17:08
← Radium left17:08
← khmarbaise left17:09
cmn man git rev-parse has a lot of flags17:09
gitinfo the git-rev-parse manpage is available at http://jk.gs/git-rev-parse.html17:09
cmn I believe --top-level does what you want17:09
EugeneKay yacc - man git-rev-parse, see --git-dir17:09
gitinfo yacc: the git-rev-parse manpage is available at http://jk.gs/git-rev-parse.html17:09
→ cakehero_ joined17:10
EugeneKay cmn - I think you mean --show-toplevel, and that's for the worktree, not .git17:10
cmn yeah, but when you say "root", you usually mean the top of the tree17:11
EugeneKay Except when you say "basically the .git directory"17:11
cmn ah, I stopped parsing too soon17:12
EugeneKay I pride myself on being technically correct. The BEST kind of correct.17:12
← incog left17:14
← albel727 left17:14
→ Vert joined17:14
→ thearthur joined17:15
← williamcotton left17:16
← thearthur left17:18
→ glimpsenirvana joined17:18
← babilen left17:19
offby1 ◎_◎ ಠ_ಠ17:20
← cakehero_ left17:20
milki yet we still dont know how to find .git17:20
← markus_ left17:21
→ thearthur joined17:21
→ Radium joined17:23
→ gusnan joined17:24
→ Goplat joined17:24
→ veyres joined17:25
→ chaten joined17:27
→ mister_m joined17:28
→ stkrzysiak joined17:28
→ albel727 joined17:28
cbreak milki: ?17:30
← Chib left17:30
→ blacktulip joined17:30
← pandeiro left17:31
Sky[x] hi i have a problem with git clone [email@hidden.address] ssh required password but i have added in /srv/gitosis/.ssh/authorized_keys cat from my .pub file any idea ?17:32
roflin you need to put your pubkey on the remote not on your local machine's authorized_keys17:33
Myhro this is not a gitosis channel, but i'll give a try17:33
Sky[x] the key in keydir/ have the '.pub' in file name?17:33
cbreak you should switch to gitolite17:34
gitorious is dead since ages17:34
← forrestv left17:34
Sky[x] Myhro: have. pub and without .pub both files17:35
cmn gitorious is till alive :p17:35
bremner cbreak: I think you confuse gitorious and gitosis17:35
← drizzd left17:35
→ arvind_khadri joined17:35
← arietis left17:35
Myhro Sky[x] you don't have to ever edit the 'authorized_keys' by hand with gitosis. it's a new setup? or you are just adding a new user?17:35
Sky[x] new setup on clean ubuntu install17:36
→ forrestv joined17:36
bremner new setup of !gitosis makes no sense. u17:36
gitinfo gitosis is no longer maintained and supported by the author; we usually recommend gitolite instead which has much better documentation and more features: http://github.com/sitaramc/gitolite17:36
Myhro you did a 'gitosis-init < my_pub_key.pub'?17:37
Sky[x] Myhro: yes17:37
cbreak bremner: right. Gitosis :)17:37
← dr_diamond left17:37
Myhro Sky[x] it should work after that, without editing the 'authorized_keys' file later17:38
→ bglusman joined17:38
Myhro did you run the command on the correct user?17:38
→ SegFaultAX|work joined17:38
Myhro 'gitosis' under 'remotehost.example.com'17:38
← Lord_DeathMatch left17:39
Sky[x] Myhro: yes sudo -H -u gitosis gitosis-init < urban_gitserver_virtualbox_ubuntu.pub17:39
← berserkr left17:40
Myhro Sky[x] that maybe silly to ask, but are you using 'openssh-server'? gitosis doesn't work with other ssh server like dropbear17:40
Sky[x] yes i run openssh-server17:41
← az left17:42
Myhro you could could try running ssh with '-v' (or -vv/-vvv) argument against the server to see if you can find what is going wrong17:43
Sky[x] i can17:43
← angelsl left17:43
Myhro therefore, I suggest you to read the chapter about gitosis on the 'Pro Git' book17:43
← bglusman left17:43
→ harish joined17:43
→ az joined17:43
→ Guest78533 joined17:44
← arvind_khadri left17:44
→ cakehero joined17:44
← stkrzysiak left17:45
→ twitchy joined17:45
← SegFaultAX|work left17:49
← glimpsenirvana left17:50
← fisted left17:52
← torukawahata left17:53
→ h0bbit joined17:53
→ arvind_khadri joined17:53
← arvind_khadri left17:53
→ arvind_khadri joined17:53
← chaten left17:54
← pfFred left17:54
Sky[x] Myhro: can be the problem on my side on local from where i try to connect on gitosis ? i'm using osx17:55
Myhro Sky[x] only if your ssh client is trying to authenticate with the wrong key17:56
I've seen a gitosis server with multiple cliente-OSes doing quite well17:57
→ rurufufuss joined17:57
→ pfFredd joined17:57
→ JPT joined17:57
← twitchy left17:58
Sky[x] Myhro: the problem was on my side :)17:59
Myhro Sky[x] solved?17:59
Sky[x] Myhro: ˜/.ssh/config i have to set 3 line which file use, port and user with host17:59
Myhro ssh is running on a non-standard port?18:00
Sky[x] the problem was on my side :)18:00
no18:00
txomon is there any way to make git have all the commit history (hashes) but not all the data?18:00
I mean, not all the files18:00
Myhro Sky[x] anyway, I'm glad you solved it18:01
:)18:01
Sky[x] i have to add this lineS http://pastebin.com/CSqPdKxd18:01
← Sahua left18:01
Myhro so the problem was the private key file name?18:02
:P18:02
txomon Sky[x], why do you use gitosis?18:02
Myhro sometimes gitosis 'just work'18:03
and it's preaty easy to setup18:03
txomon Myhro, ... gitolite is too18:03
much more easier to make it work18:03
and pretty simple to setup18:04
Myhro I'll try gitolite someday18:04
→ marcogmonteiro joined18:04
cmn and supported18:04
Myhro I've been using gitosis for near a year18:04
txomon Myhro, you should18:04
cbreak gitosis is dead for several18:04
→ mgpcoe joined18:04
Myhro I know...18:04
:P18:04
→ rryoumaa joined18:04
gitinfo set mode: +v18:04
txomon I started with gitosis, but in 3 hours, I changed to gitolite in the very first versions18:05
← JohnyBaby left18:05
txomon now its awesome what you can get to be done18:05
cbreak, any idea on my question?18:05
cbreak nope. what was it?18:06
ah, no, that's not possible18:06
EugeneKay txomon - make a git repository containing the hashes but not the files? Not really. You can parse the output of git log to get such a thing, but.... ergh.18:06
→ lapistano joined18:06
→ Sahua joined18:06
cbreak maybe git rev-list > file18:06
EugeneKay I think !xy applies here18:06
gitinfo [!doinitrong] It sounds like you're approaching this problem in the wrong manner. Let's step back for a minute - What are you actually trying to achieve? Why are you trying to do it this way?18:06
rryoumaa are the timestamps of the files and directories in .git ignored by all git commands?18:07
txomon EugeneKay, I am trying to do a dropbox with git18:07
EugeneKay rryoumaa - git really doesn't care about mtime, only the contents(sha1)18:07
txomon - !dropbox18:07
gitinfo txomon: Storing a bare repository on DropBox is a Very Bad Idea (tm). Everybody who has tried it has reported repo corruption, slowness in synchronizing, and general pain. Use a free code-hosting service such as Bitbucket or GitHub instead.18:07
txomon I cant have all the history with me18:07
rryoumaa EugeneKay: thanks18:07
txomon EugeneKay, dont start spamming all the words I say18:07
I said making a dropbox with git, not using dropbox with git18:08
EugeneKay Don't tell me what to do. You're not my mother.... and if you were, well, she's a bit of a bitch anyway.18:08
cmn txomon: it's caled experience18:08
and either way, it only works for limited use-cases18:08
txomon ...18:08
EugeneKay So, you want to make git behave as a file storage mechanism.18:08
cmn take a look at sparklshare18:08
EugeneKay This is not what git's object model is oriented towards. Projects such as SparkleShare attempt to build such a thing, but it doesn't always work well18:08
cbreak txomon: check out git-annex18:09
txomon I want to do it on myself, and I thought about git18:09
cbreak txomon: also maybe bup18:09
EugeneKay !annex18:09
gitinfo Storing lots of binary files in git causes repo balloon. git-annex and git-media are two solutions to work around this by keeping the blobs out of the repo. http://git-annex.branchable.com/ and https://github.com/schacon/git-media18:09
txomon cbreak, good idea!18:09
EugeneKay I'm not sure if annex lets you purge old blobs, but it ought to.18:09
txomon but gitolite support for annex is not ready yet... Will have to stay on g2 for the server18:10
EugeneKay You could add it18:10
txomon add support you mean?18:10
I refered to g3 gitolite, in g2 branch it exists18:11
→ berserkr joined18:12
← mgpcoe left18:12
→ arietis joined18:12
EugeneKay Indeed.18:13
txomon cbreak, annex idea was _REALLY_ good idea! Its perfect. I will have to have some tracked archive though18:13
→ hollyw00d joined18:13
cbreak it's a hack.18:13
EugeneKay All the best scripts are18:14
← hollyw00d left18:15
→ hollyw00d joined18:16
← Dragooon left18:16
txomon cbreak, it is also usefull for the problem on the other day (when my team wanted to have access to different branches at the same time)18:17
cbreak nah18:17
for that you'd use the git new-worktree script from contrib18:18
or just clone locally18:18
annex can't help you with that problem18:18
but being on two branches at the same time isn't something people normally do18:18
txomon I meant, that I could store both copies in tarballs and have them available as something "extra"18:18
← sdeboey left18:19
cbreak sounds stupid... both copies of what?18:19
← Nathandim left18:20
← thearthur left18:21
txomon cbreak, of all the kernel...18:21
cbreak then it makes no sense at all18:22
txomon why? something like distribution tarball18:22
cbreak because the sources are already in the repository18:22
why store a tarball?18:22
git archive can generate those on the fly18:22
txomon didn't know about git archive...18:22
oki18:22
→ thearthur joined18:23
txomon good to know about it18:23
milki cbreak: will they always have the same checksum?18:23
← arvind_khadri left18:23
← anildigital left18:23
texasmynsted The on the fly tar ball is a nice, useful feature for git.18:24
cbreak milian: who?18:24
txomon milki, it would be stupid if they didn't (I suppose)18:24
cbreak sorry18:24
milki: who?18:24
milki git archive18:24
cbreak the archives?18:24
milki ya18:24
cbreak no idea.18:24
they will have the same data content18:24
(if you archive the same treeish)18:24
txomon the problem would come with the date18:25
← gjoseph left18:25
rryoumaa if you archive a treeish the date will change18:25
if you archive a ref it will not18:25
(according to man page)18:25
→ gjoseph joined18:26
milki i redistribute gitolite as a tarball because 1) ensure consistent checksum, 2) requires one additional step18:26
→ anildigital joined18:26
rryoumaa do people use bup these days or gibak?18:27
→ Per4u3e joined18:28
milki havent heard of either..18:28
← glennpratt left18:31
← blacktulip left18:32
→ mwrosen joined18:32
← JPT left18:33
→ blacktulip joined18:33
variable if I want to optimize repo space what should I run? 'git gc' exactly? With any special options?18:34
← iamntz left18:34
← jargon- left18:34
→ JPT joined18:35
cbreak variable: --aggressive or so18:36
→ ThunderChicken joined18:36
→ drizzd joined18:37
← hollyw00d left18:37
milki cbreak: will it do anything on a fresh clone?18:38
variable cbreak: what about http://gcc.gnu.org/ml/gcc/2007-12/msg00165.html ?18:38
→ sdeboey joined18:39
← Per4u3e left18:39
cbreak milki: no18:39
milki variable: sounds like it could just be shorthand for the repack commandat the end18:40
cbreak variable: what about it?18:40
milki variable: tailor the window/depth to the repo perhaps?18:40
variable cbreak: if I'm understanding the email correctly linus is saying that --aggressive will make things worse?18:40
milki: I have no clue what numbers I should pick18:41
milki from what i understand, it just says it takes a whole lot of cpu time18:41
!tias18:41
gitinfo Try it and seeâ„¢. You learn much more by experimentation than by asking without having even tried. http://sitaramc.github.com/1-basic-usage/tias.html may help with git-specific TIAS.18:41
milki :P18:41
variable milki: tias doesn't work when it takes hours to complete18:41
milki thats where :P comes in18:41
sounds like it should match the longest delta chain to any single file18:42
variable punches milki18:42
variable but understood. how do I determine that ?18:42
milki lots of grepping commits18:42
variable :'(18:42
← ronniy left18:43
cbreak variable: maybe. It'll make things compact.18:43
milki variable: hit up the mailing list about delta-chains18:43
maybe someone has a script to examine delta-chains18:43
variable cbreak: understood18:44
cbreak the email seems to say that it's a waste of time18:44
not that it does bad stuff18:44
I think18:44
only flew over18:45
milki its a waste of time, if git gc has been doing its job over time18:45
variable milki: this is the initial checkout18:45
cbreak: understood.18:45
thanks18:45
milki yes, he mentioned that import tools might not be efficient18:45
so it should do something if svn2git isnt super smart18:46
← mklappstuhl left18:47
charon variable: --aggressive simply discards delta chains that exist already. this may occasionally help if the original pack is terrible, but usually is just a massive waste of time. see e.g. the numbers in http://article.gmane.org/gmane.comp.version-control.git/196506 (note that gc --aggressive implies repack -f)18:47
← Jackneill left18:48
→ webguynow joined18:48
→ g0bl1n joined18:48
← g0bl1n left18:48
milki variable: just set windowsize to over 900018:50
→ pfFred joined18:51
← mishok13 left18:53
→ sq-one joined18:53
← pfFredd left18:54
← marcogmonteiro left18:54
→ marcogmonteiro joined18:56
→ SkAzZ2 joined18:59
← SkAzZ1 left19:01
← ZooLer left19:04
→ williamcotton joined19:04
→ SegFaultAX|work joined19:05
← Guest78533 left19:05
→ glennpratt joined19:06
→ johtso joined19:10
→ Guest78533 joined19:12
→ ZooLer joined19:12
→ francisfish joined19:13
rryoumaa my experience on a medium size repo is that aggressive works well at reducing size but turns the computer into a decorative heat source for hours19:14
← txomon left19:14
→ benellix joined19:14
← benellix left19:14
← gusnan left19:16
→ Dragooon joined19:16
→ Logxen joined19:16
← tharkun left19:17
→ tharkun joined19:17
offby1 for some value of "decorative"19:17
milki oo19:17
sounds like a good heater19:17
maybe i should do that next time when its too cold19:17
cbreak just don't use all cores...19:18
milki rryoumaa: medium would be? variable is working with a 19GB repo19:18
variable milki: haha19:18
rryoumaa: yeah, how big is 'medium' ?19:19
rryoumaa about 100mb19:19
milki >.<19:19
variable rryoumaa: :]19:19
milki variable: did you see the links i posted last night?19:19
variable milki: no19:19
milki hmmm19:19
← Lgb left19:20
milki 22:53 < milki> http://www.google-melange.com/gsoc/project/google/gsoc2012/flyingflo/1200219:20
22:54 < milki> http://www.google-melange.com/gsoc/project/google/gsoc2012/tgummerer/1200119:20
22:59 < milki> http://git-scm.com/about/small-and-fast19:20
22:59 < milki> if you look at the bottom19:20
23:00 < milki> it shows a git clone larger than the svn checkout19:20
the first two are current gsoc projects that might be related19:20
index format19:20
and better svn tool19:20
variable ah, cool19:20
→ q4brk joined19:21
milki last shows git repo will probably be larger19:21
← gjoseph left19:21
→ HyperNerdV2_ joined19:21
→ Lgb joined19:21
← HyperNerdV2 left19:24
← QuantumQuanta left19:24
← izissise left19:25
→ nadavoid joined19:25
← HyperNerdV2_ left19:26
← thearthur left19:26
← joepie91 left19:27
→ joepie91 joined19:28
← Parhs left19:28
→ Parhs joined19:28
← parcs` left19:29
← SkAzZ2 left19:31
→ Parhs8 joined19:33
← Parhs left19:33
← rryoumaa left19:34
← aristidesfl left19:34
→ thearthur joined19:34
→ SkAzZ joined19:36
← LucasTizma left19:36
← DofLeFou left19:37
→ aristidesfl joined19:37
→ SkAzZ1 joined19:37
← Bass10 left19:39
← piecyk left19:39
→ mgpcoe joined19:40
← Sigma left19:40
← SkAzZ left19:41
→ sixteneighty joined19:41
→ kfe joined19:41
→ aspotashev joined19:41
→ mishok13 joined19:42
← nadavoid left19:42
← glennpratt left19:42
→ rryoumaa joined19:43
→ friskd joined19:45
← friskd left19:45
→ glennpratt joined19:46
→ ToBeFree joined19:46
← irqq left19:48
← drizzd left19:48
← marcogmonteiro left19:49
→ drizzd joined19:49
charon thought that *highly* depends on contents19:50
← resmo__ left19:50
← Logxen left19:50
→ shruggar joined19:51
→ urbanmonk joined19:51
charon milki, variable: i had one case where the worktree was ~300MB, the SVN backing store took more than 600 and the git conversion ended up at 35 after liberal application of --window-size19:53
so it's really *highly* content dependent19:53
(35 for the pack, i should say)19:53
→ thiago joined19:54
variable charon: understood.19:56
← dr_lepper left19:57
gitinfo set mode: +v19:58
rryoumaa what does window size affect? only aggressive? is it slower?19:58
← Parhs8 left20:00
charon the window refers to the set of objects that are considered as a delta base for the current object20:00
← urbanmonk left20:00
charon the rules according to which the window is chosen are somewhat complicated20:00
but yeah, runtime goes through the roof if you set it higher20:01
← ehsan left20:01
→ Parhs joined20:01
charon whether it makes a difference largely depends on the content... if a higher window size means it can suddenly find a good delta base, then great. if not, you just wasted a lot of cpu searching in vain20:01
← h0bbit left20:01
→ LekeFly joined20:02
← berserkr left20:03
cmn peff posted some statistics on the ML some days ago20:04
going beyond 40 or so just adds CPU time20:04
charon yeah i linked that earlier20:04
→ berserkr joined20:04
charon http://article.gmane.org/gmane.comp.version-control.git/19650620:04
cmn ah, right; it had scrolled off my window20:05
charon note that he also had an example where it made a massive impact20:05
← twisted` left20:06
charon kind of makes you wonder whether a randomized probing approach might see what would be a good number20:06
→ smw_ joined20:07
→ voidx joined20:08
← j416 left20:08
→ twisted` joined20:09
← Parhs left20:10
→ Parhs joined20:11
→ jfelchner joined20:14
← thearthur left20:15
→ wsc joined20:15
→ connormai joined20:17
← connormai left20:17
cmn a completely randomized algo should make some things slower20:18
at least that's what git's packing assumes20:18
→ Slix` joined20:18
→ connormai joined20:18
← connormai left20:18
cmn you could end up with HEAD being behind a few delta chains20:18
charon huh?20:18
cmn git stores the latest objects undeltified20:19
assuming you'll use them more often20:19
or maybe you were going in a completely different direction20:19
charon no i meant something else... what if for 2^-i fraction of the objects, you try a 2^i window instead of a smaller one20:19
then if that works out well, start tuning it upwards20:20
or some such.20:20
or pick a random subset of the objects, and see what a larger window would do *before* packing anything20:20
just something that makes it content-aware instead of a fixed "experience-based" approach that peff clearly showed does not apply very well in some cases20:20
cmn yeah, dynamically changing the window size could produce interesting results20:21
neersighted|AFK → neersighted20:22
→ diegoviola joined20:23
→ ferdna joined20:23
→ connormai joined20:25
← thiago left20:25
← connormai left20:26
← sdeboey left20:26
← dreiss left20:30
→ kukks joined20:31
← veyres left20:32
← Icyberg left20:33
← mishok13 left20:34
→ Sigma joined20:36
← voidx left20:37
→ voidx joined20:37
← MichaelRpdx left20:38
→ bglusman joined20:39
← voidx left20:39
→ voidx joined20:40
→ connormai joined20:41
← connormai left20:41
← bglusman left20:44
→ resmo_ joined20:45
← apetrescu left20:46
→ Mowsh joined20:47
← subbyyy left20:47
← Dave^| left20:47
← Sigma left20:48
→ cagedwisdom joined20:53
milki charon: what was the content of your repo?20:53
variable's is really a lot of tiny text files20:53
→ DofLeFou joined20:53
milki makefiles, diffs, lists of iles20:53
f20:53
← fairuz left20:54
ori shruggar: what is the git annex alternative ?20:56
→ dvide_ joined20:56
← dvide left20:57
← sixteneighty left20:57
→ n2diy joined20:58
← alexukf left21:00
→ peppyheppy joined21:01
← aristidesfl left21:02
→ fairuz joined21:02
→ subbyyy joined21:03
→ eijk joined21:03
← q4brk left21:04
← wsc left21:05
→ boombatower joined21:05
shruggar ori: what's that?21:07
ori: are you asking what git annex is, or an alternative to git annex?21:07
ori i read about annex, says it just keep tracks of the file info and not content21:08
→ ipalaus_ joined21:08
shruggar (btw: I'm not usually responsive on the weekend, this window just auto-loads)21:08
→ juvenal joined21:08
ori but earlier you said there is another alternative21:08
i want to read on it21:08
shruggar I'll try !annex21:08
gitinfo Storing lots of binary files in git causes repo balloon. git-annex and git-media are two solutions to work around this by keeping the blobs out of the repo. http://git-annex.branchable.com/ and https://github.com/schacon/git-media21:08
shruggar git-media, there it is21:08
← DofLeFou left21:09
shruggar the point its: they don't store the data in git, just "pointers to" the data, as git doesn't handle large data well21:09
→ ipalaus__ joined21:09
shruggar I would expect annex to include solutions for fetching the data itself21:09
ori yeah but it has to be *safe*21:10
not what i want21:10
shruggar store it somewhere safe. I like amazon's s321:11
cbreak you think that's safe?21:12
shruggar safer than anything I have locally, yeah21:12
cbreak but not safer than storing it in multiple places21:12
cmn safe != secure21:12
← ipalaus_ left21:12
cbreak I've heard of quite a few cloud storage failures21:12
EugeneKay s/safe/reliable/21:12
cbreak there were some with data loss at amazon too21:13
something with a router issue or so21:13
cmn reminds me of the guy who had some bitcoin purse on EC2 and shut down the machine, deleting all the data21:13
cbreak could have been azure, not sure21:13
EugeneKay Azure had a massive SSL failure stemming from Feb 29th21:13
cmn MS managed to loose (or almost loose) a shitload of data from that phone they only have in the US21:13
EugeneKay Shouldn't happen again until 201621:13
← d0k left21:14
ori any of you guys use dropbox to host the remote ?21:14
cmn !dropbox21:14
gitinfo Storing a bare repository on DropBox is a Very Bad Idea (tm). Everybody who has tried it has reported repo corruption, slowness in synchronizing, and general pain. Use a free code-hosting service such as Bitbucket or GitHub instead.21:14
ori who is everybody21:14
→ are_kim joined21:15
← xsbeats left21:16
← are_kim left21:16
cmn people who have a) used dropbox to host their git repos and b) reported that it corrupts objects21:16
I realise it looks like it should work, but it doesn't21:16
dropbox sees two files with the same name and gets horribly confused21:16
cbreak cmn: danger sidekick thing? They lost everything afaik21:17
← Dragooon left21:17
ori why, all files are sha1 based21:17
right ?21:17
→ xsbeats joined21:17
cbreak ori: don't use dropbox, it's garbage21:17
← jfelchner left21:18
cmn cbreak: yeah, that one21:18
ori cbreak: did you use it ?21:18
EugeneKay ori - try it, cry, don't try it again. ;-)21:18
Or, skip straight to: don't do that21:18
cmn if you want to do it, do it21:18
← iwoj left21:18
ori its going to be my 4th repo21:18
cbreak I heard of someone in here who had massive problems with the daemon in the background messing up his repository21:18
cmn but don't say we didn't warn you21:18
rryoumaa EugeneKay: got a better, gratis, open-protocol alternative?21:18
cbreak it didn't work until he stopped the daemon21:18
ori so if it dies its an experience, not a terrible outcome21:18
→ iwoj joined21:18
cmn then do it21:19
← ology left21:19
→ eddiee joined21:19
rryoumaa i'd like to have a simple command line interface to gdrive or similar21:19
(curl, rsync, etc.)21:19
cbreak cam21:19
can't you mount it?21:19
your drive?21:19
→ aristidesfl joined21:20
EugeneKay rryoumaa - !gitolite ;-)21:21
gitinfo rryoumaa: Want to host as many git repos (and users!) as you like, on your own server, with fine-grained access control? You want gitolite: https://github.com/sitaramc/gitolite - Documentation: http://sitaramc.github.com/gitolite/master-toc.html21:21
rryoumaa EugeneKay: i meant alternative to dropbox21:21
gratis backup stuff (don't even care about git or sync)21:21
EugeneKay My only usecase for dropbox at present is hosting images from my phone is imgur. I do all of my backup via rsync to my server farm.21:22
s/something in that sentence/imgur21:22
← eddiee left21:22
→ Mjiig joined21:22
cbreak you have a server farm?21:23
cbreak wants one too21:23
EugeneKay Corn-raised21:23
← mister_m left21:23
→ thebignoise joined21:23
ori i am pushing the same commit to two repos21:24
one reports that its 284.33mb the other 284.26mb... each repo on a different drive, ideas?21:24
EugeneKay The commit object? Or the whole repo size?21:24
cmn one what reports?21:24
EugeneKay Because that sounds like it's within the margin of error of packfiles on different repos21:25
ori write object..... and then total 428 delta 118 vs write obj ..... total 428 delta 12021:25
cmn you're pushing to different repos21:25
→ igotnolegs joined21:25
cmn they're different21:26
ori physically different, but identical in every count21:26
i push to 3 repos on each commit21:26
EugeneKay Talk to git-gc about it21:26
ori i have no idea what you mean21:27
→ BiggFREE joined21:27
EugeneKay A repo is a collection of objects. These objects are usually packed together to save space. Differnet repos can come up with different packfiles for (mostly) the same contents.21:27
It's a non-deterministic algorithm21:27
← thebignoise left21:28
ori its the same objects, same commit pushed to all 3 repos sequentially21:28
how can a compression algorithm be non deterministic21:29
→ catsup joined21:29
cmn by producing different results21:29
ori thats the outcome, not the reason21:29
cmn it tries some combinations and has to decide21:29
it's not an outcome21:30
← shruggar left21:30
← giallu left21:30
cmn the reason is that Linus wrote it that way way back when21:30
EugeneKay Unless you can 100% guarantee that the same files went to each repo at the same exact time in the same exact order, then it is perfectly understandable that you get different results.21:30
Like I said, <EugeneKay> Talk to git-gc about it21:30
ori yes same files went to each repo, not at the same time, but sequentially. meaning that after pushing to repo1 it pushed to 2, then 3.21:32
← pantsman left21:33
← tharkun left21:33
← mikepack left21:33
→ mikepack joined21:34
→ sdeboey joined21:34
← mikepack left21:34
→ mikepack joined21:35
→ vervic joined21:35
← iwoj left21:35
← sq-one left21:37
→ thiago joined21:38
← thiago left21:38
→ thiago joined21:38
→ zaiste joined21:39
Mjiig ori: if they went at different times then the trees would have different timestamps in them, so the files would be marginally different21:39
← Chillance left21:39
Mjiig which makes a difference when you go onto compress21:40
← todd_dsm left21:40
→ Chillance joined21:40
cbreak trees don't store timestamps21:40
CareBear\ ori : my guess is that you have not always pushed exactly the same changes to these three repos at the same time21:41
ori : why is the small difference in size a problem for you?21:41
← BiggFREE left21:41
ori not a problem, wanted to know21:41
yes so if the time is a factor. that must be it21:42
speaking of time, is there a way to store meta data some how. so i get the original files creation date and time21:42
Mjiig cbreak: i'm not that great on my vim internals, something does though :) might be commits...21:42
→ DofLeFou joined21:42
Mjiig git not vim obviously21:42
cbreak commits store time stamps.21:43
but one commit has exactly two time stamps21:43
unchangeably21:43
embedded in history21:43
nothing can change it21:43
← milki left21:43
ori yeah i got that21:44
i was talking about an addon or something that also keeps the original creation time of files21:44
→ tharkun joined21:44
Mjiig cbreak: but if you make the commit twice to 2 different repos, identical in every way except when you make them, then the two repos would be (slightly) different because they would have different timestamps on the commit21:44
cbreak and a different commit21:45
which means diverging history21:45
ori i commit once21:45
cbreak which means you have to merge later on21:45
ori push 3 times21:45
not commit 3 times21:45
cmn there is no space in the tree to store timestamps21:45
ori thats why i have asked about an extension to git21:46
cmn that would make it completely incompatible with git21:46
ori i read something earlier, but dont remember if its real or not21:46
cbreak I don't see a point for storing that21:46
← francisfish left21:46
cmn people keep tyring to use git as a filesystem21:46
rryoumaa ori: it exists21:46
cmn bup might care21:46
rryoumaa i don't know what it is called, but it stores a file with metadata21:46
ori but bup relies on annex21:47
→ psoo joined21:47
cmn are you sure?21:47
cbreak it does not.21:48
→ drev1 joined21:48
cbreak it's an unrelated project.21:48
ori http://git-annex.branchable.com/special_remotes/bup/21:48
maybe i am wrong21:48
cbreak you are wrong.21:48
bremner annex can optionally use bup21:48
← peppyheppy left21:49
ori looks like its only for big files21:49
cbreak bup is a backup system21:49
it's for any kind of files21:49
← tharkun left21:50
→ maletor joined21:50
→ marcuy_ joined21:51
← Mjiig left21:53
→ tharkun joined21:54
← marcuy left21:54
→ dr_lepper joined21:54
← dr_lepper left21:54
→ dr_lepper joined21:54
← alfplayer left21:55
→ alfplayer joined21:56
← berserkr left21:58
→ Element9 joined21:59
joepie91 where can I find documentation on the exact format of git (v2) packfiles?22:00
EugeneKay The source code.22:01
cmn Documentation/pack-format.txt22:01
or similar22:01
joepie91 cmn: doesn't that just cover v1?22:01
← raatiniemi left22:01
joepie91 several articles I found that refered to pack-format.txt said it was about v1 only :/22:02
→ avinashhm joined22:02
cmn have you checked personally?22:03
also, are you sure you're talking about packfile format and not pack index format?22:03
joepie91 whoops, I meant pack index format indeed22:03
sorry22:04
cmn as we say here, !blog22:04
gitinfo Blog posts, while helpful and informative, are quite often outdated, give bad advice, or are just plain wrong. Please don't rely solely upon them, or treat them as authoritative.22:04
← ToBeFree left22:05
← Radium left22:05
→ felher joined22:06
cmn and I know for a fact that it's described there because that's what I used to implement a packfile indexer22:06
← sdaone left22:07
joepie91 right, thanks :P22:07
← tharkun left22:08
→ tharkun joined22:08
→ stkrzysiak joined22:08
→ kcj joined22:13
→ Radium_ joined22:15
→ Noobus joined22:15
← smw_ left22:15
← glennpratt left22:16
→ _W_ joined22:16
← ipalaus__ left22:16
← thelazydeveloper left22:17
← blacktulip left22:18
→ milki joined22:18
→ patrick99e99 joined22:18
← mikepack left22:18
→ mikepack joined22:19
→ TehDeviance joined22:19
← TehDeviance left22:19
← kcj left22:19
→ jim0203 joined22:20
→ thelazydeveloper joined22:21
← drizzd left22:21
→ kcj joined22:22
← kfe left22:23
← mikepack left22:23
← whitman left22:24
→ fisted joined22:24
→ mister_m joined22:28
← pppaul left22:31
← ph^ left22:31
→ shruggar joined22:32
→ pppaul joined22:32
← psoo left22:34
→ pythonsnake joined22:36
pythonsnake hi22:36
gitinfo pythonsnake: welcome to #git, a place full of helpful gits. If you have a question, just ask it–somebody should answer shortly. For more information about git and this channel, see the links in the topic. It can't hurt to do a backup (type !backup for help) before trying advice you receive here.22:36
pythonsnake i have git checkout'ed22:36
now how can I reset to when i just checked out22:36
SethRobertson !fixup22:36
gitinfo So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!22:36
pythonsnake without what i added/removed22:37
or modified22:37
charon milki: tex sources for lecture notes, mostly. some figures in various source and/or ps/pdf formats.22:37
← keltvek left22:37
pythonsnake i didn't git commit22:37
SethRobertson pythonsnake: !fixup22:37
gitinfo pythonsnake: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!22:37
pythonsnake i didn't git commit22:37
SethRobertson do it anyway22:37
(follow the link I mean)22:38
pythonsnake oh thanks22:38
thanks!!22:38
been searching for hours22:38
minutes*22:39
→ ivan` joined22:39
pythonsnake :P22:39
← apok left22:39
→ keltvek joined22:39
← Salash left22:40
→ Salash joined22:42
← Salash left22:43
→ ISF joined22:44
← tharkun left22:44
→ tharkun joined22:44
→ Salash joined22:45
milki charon: so possibly the gc with window size will indeed help a lot for variable22:45
← Salash left22:45
← patrick99e99 left22:45
→ Salash joined22:46
→ apok joined22:47
← saurik left22:47
← fishguy left22:48
← Fullmoon left22:49
← apok left22:49
← Salash left22:50
→ kadoban joined22:52
← mister_m left22:52
← DofLeFou left22:53
← mwrosen left22:57
→ tr-808 joined22:59
Parhs omg that sucks... http://www.newsit.gr/default.php?pname=Article&art_id=137886&catid=3 (irrilevant )23:01
→ mister_m joined23:01
← aristidesfl left23:02
offby1 what really sucks is that Google Translate won't translate it into English for me23:02
→ Lulzon joined23:03
→ codeM0nK3Y joined23:04
Parhs why that offby1?23:04
← tr-808 left23:04
offby1 I don't know why.23:04
→ aristidesfl joined23:05
Parhs http://translate.google.gr/translate?sl=el&tl=en&js=n&prev=_t&hl=el&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.newsit.gr%2Fdefault.php%3Fpname%3DArticle%26art_id%3D137886%26catid%3D3&act=url23:05
that works23:05
→ glennpratt joined23:06
→ ErsatzRyan joined23:06
← glennpratt left23:06
← kadoban left23:06
→ kadoban_ joined23:07
→ LucasTizma joined23:08
→ Bass10 joined23:13
← robbyoconnor left23:13
← brokenshakles left23:14
→ tr-808 joined23:14
← tharkun left23:16
← Noobus left23:16
→ tharkun joined23:16
→ erichynds joined23:18
→ khmarbaise joined23:18
← khmarbaise left23:19
← tharkun left23:20
→ nutron joined23:20
→ BBBThunda joined23:21
← teweWORK left23:21
→ Prime1309 joined23:21
→ apok joined23:22
→ brokenshakles joined23:22
← cehteh left23:22
→ GNUcifer joined23:25
← ErsatzRyan left23:25
← aspotashev left23:25
← diegoviola left23:25
cmn is that article about something in particular?23:26
BBBThunda when I do a "git remote set-url origin ssh://git@hostname" I get "ssh: hostname:: no address associated with name" .... but when I ping hostname or ssh to it, dns resolves it and it works fine. Any ideas? (I'm using Git Bash with msysgit on Windows 7)23:26
cmn make sure you haven't overriden the hostname in the ssh config23:27
and that you haven't made a typo23:27
BBBThunda sorry... forgot to mention I get that error when doing a git push or git pull etc.23:28
← airborn left23:28
cmn I assumed23:28
← Lulzon left23:29
→ Nugget__ joined23:29
→ Lulzon joined23:29
cmn also check that you're using ssh instead of plink or whatever to connect23:29
BBBThunda cmn, would that be under ~/.ssh ? if so, all I have there are key files and known hosts23:30
cmn there are a fwe places, but yeah, you'd have configured in ~/.ssh/config23:31
→ tharkun joined23:31
→ ehsan joined23:31
← glcrazy left23:32
BBBThunda I tried deleting the known hosts entry and doing an ssh again, which worked... but still can't push/pull with hostname23:32
cmn that doesn't change anyhting23:32
other than opening you up to a MITM attack23:32
← Soliton left23:32
← GNUcifer left23:33
→ zendeavor joined23:34
← arietis left23:34
← Radium_ left23:34
BBBThunda i don't know where else I would have any ssh config overriding anything... does git maybe use the "host" command to resolve? if so I don't think that command exists in git bash23:34
cmn you mean the host utility/23:34
git uses the systems' DNS functions23:35
BBBThunda I assume ping does also... and that one resolves23:35
→ SamDiablo joined23:37
← SamDiablo left23:37
cmn yes, we've establishes that the hostname exits23:38
what about the other stuff?23:38
BBBThunda i used ssh command in git bash, not plink/putty23:38
cmn fine, s/you're/git is/23:40
offby1 BBBThunda: !repro might help us help you23:40
gitinfo BBBThunda: Please paste (using https://gist.github.com/ or similar) a transcript (https://gist.github.com/2415442) of your terminal session, or at least explain exactly what you did that led up to the problem. This will help immensely with troubleshooting.23:40
→ robbyoconnor joined23:40
← stkrzysiak left23:40
→ BiggFREE joined23:41
→ petisnnake joined23:42
← eijk left23:43
← rryoumaa left23:47
→ stkrzysiak joined23:47
→ GNUcifer joined23:48
BBBThunda Can't get a transcript because the git bash buffer doesn't go back that far. I honestly don't think it would help anyway... but I'll re-type and put it in a gist if you think that will help. Note that I'm not using github. I'm using my own server for the remote repo.23:48
→ Soliton joined23:48
← Mowsh left23:50
cmn huh? did they disable the scrollback?23:52
→ Radium joined23:52
← Lulzon left23:52
← tharkun left23:52
cmn either way, there's no need to retype anything23:52
→ airborn joined23:53
cmn use the history23:53
FauxFaux cmn: It's cmd hosted, so probably has the default 200 lines or whatever nonsense.23:53
→ apetrescu joined23:53
offby1 BBBThunda: that's annoying, if the terminal has no scrollback. (I myself generally disable scrollback because I'm a cowboy, but I'd expect it to be enabled by default)23:54
FauxFaux offby1: Don't worry, screen(1)'ll maintain it for you anyway.23:54
→ diegoviola joined23:54
cmn ugh, I get annoyed when I have less than a few screenfulls of scrollback23:54
BBBThunda there's a scrollback, but it only goes back, maybe 350 lines or so23:54
cmn then use the history to redo it23:55
offby1 FauxFaux: screen doesn't work on Windows.23:55
cmn they can't have fucked that up as well23:55
offby1 I assume you're pulling my leg.23:55
← lonewulf` left23:55
FauxFaux Surely someone's implemented screen by now. >.>23:55
cmn offby1: that's why you ssh into a real computer23:55
offby1 BBBThunda: ah, and you've done lots of stuff since then. Gotcha23:55
cmn: well, yes, it is :) But I doubt that will help BBBThunda at this moment.23:56
← petisnnake left23:56
cmn maybe not, but there's no need for that anyway23:56
← Radium left23:56
cmn there is no problem; bash keeps a history of the executed commands23:57
BBBThunda oh you mean bash history?23:57
i thought you waned a git log23:57
cmn whatever for?23:57
your problem isn't with the git history23:58
← cakehero left23:58
BBBThunda i didn't think it was... but figured you knew something I didn't :)23:58
← apok left23:58

Logs Search ←Prev date Next date→ Channels Documentation