IRCloggy #git 2009-11-25

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.

2009-11-25

killerchicken meaning I have stuff in my working dir that's not added to the index yet00:00
← paltman left00:00
killerchicken git diff --cached is empty00:00
flex but what does git status tell you?00:00
← Haarg left00:00
killerchicken gives me a bunch of modified paths00:00
→ Haarg joined00:01
killerchicken and tells me no changes are staged00:01
dwmw2 → dwmw2_gone00:01
← Haarg left00:01
flex strange, that should work...00:02
→ hgimenez joined00:02
hebz0rl it should yeah00:02
dbrock what does git diff <commit> HEAD give?00:02
killerchicken shows me the correct diff between the two commits00:03
hebz0rl ah differenz between <commit> and HEAD?00:03
killerchicken yes00:03
dbrock but does it give the same as git diff <commit> ?00:03
hebz0rl yes00:03
killerchicken yes, that's the strange thing00:03
← Pewpewarrows left00:03
hebz0rl git diff <commit> <commit2> gets you the differenz between <commit> and <commit2> if <commit2> is NULL <commit2> = HEAD00:04
→ tfarina joined00:05
dbrock NULL?00:05
hebz0rl right?00:05
→ bugxek joined00:05
hebz0rl NULL is when you dont specifig command200:05
← pantsman left00:06
← vbgunz left00:06
hebz0rl like git diff <commit>00:06
→ Davey joined00:06
dbrock according to my 1.6.0.4 man page, it's not supposed to: when you say git diff <commit>, it's supposed to diff against your working tree00:06
→ dreiss joined00:07
hebz0rl and your working tree is?00:08
currently?00:08
dbrock killerchicken: maybe try creating a new simple repository and see if you get the same behavior?00:08
← Turias left00:08
dbrock hebz0rl: sorry, but you're not making much sense; killerchicken is having a problem and we're trying to help him/her00:09
killerchicken dbrock: yeah, will do tomorrow. Falling asleep now, I was hoping someone could tell my I'm stupid in 5 seconds and it would be over :)00:09
dbrock ah :)00:09
hebz0rl yeah thats what im trying to tell him :D00:10
→ shadfc joined00:11
← tbf left00:12
→ Determinist joined00:12
hebz0rl and you working tree is normally HEAD00:12
so yeah :D00:12
killerchicken it isn't00:13
doener hebz0rl: no00:13
← laj left00:13
doener hebz0rl: HEAD is a reference to what you have checked out, either a branch head or a commit00:13
hebz0rl woot?00:13
hmh yeah but if you commit it it is normally HEAD i maybe misunderstood that sry :/00:13
→ pantsman joined00:13
killerchicken hebz0rl: the thing is, I explicitly want to include the not yet committed stuff in the diff :)00:14
shadfc hey guys, im trying to apply a set of patches created by someone else. The patches were created against a commit 2 revs back, so I just did a `git checkout <old commit>`. Now when I try to apply the patches (`git apply <patchfile>`), I get to the 4th and it fails. Am I doing something wrong?00:14
← kiva left00:14
hebz0rl killerchicken, just commit it and then make a diff?00:14
doener killerchicken: hm, how did you check that "git diff <commit>" doesn't work as expected?00:14
killerchicken anyways, thanks for your help, I'll be back tomorrow or on the mailing list if I can't work it out :)00:15
hebz0rl i mean you dont have to push it dont you?00:15
→ blakemizerany joined00:15
killerchicken doener: I noticed that it shows the exact same thing as git diff <commit>..HEAD00:15
hebz0rl: sure, that works as a workaround. Shouldn't be necessary, though.00:16
hebz0rl hmh00:16
doener killerchicken: checked manually? (Just asking because just last week I read a diff again and again, and failed to spot a missing line)00:16
← DrNick left00:16
hebz0rl i just checked git diff and it gives me the diff of my working tree and HEAD00:17
→ awx joined00:17
hebz0rl maybe try that? :D00:17
doener hebz0rl: "git diff" without arguments gives a diff between the working tree and the index00:17
hebz0rl: and he wants a diff against a commit different than HEAD anyway00:17
killerchicken doener: yeah, the diffs are entirely different (the working tree undoes most of one of the commits)00:17
hebz0rl oh sry i should read the hole problem :/00:18
→ scarabx joined00:18
doener killerchicken: uhm, "are entirely different" or "should be entirely different"?00:18
killerchicken should be00:18
doener ok00:18
killerchicken it's clearly too late. I'll stop bothering you, and figure it out tomorrow. Sorry for the noise.00:18
← danferranti left00:19
killerchicken (I just checked with a different repo and couldn't reproduce, so something sure is strange)00:19
hebz0rl ah np00:19
doener killerchicken: I guess you ran the "git diff <commit>" after some "git status" call, too?00:19
← madewokherd left00:19
→ madewokherd joined00:20
doener killerchicken: maybe there's some optimization that stops it from actually reading the working tree file when the index is "seemingly" up to date. "git status" would fix such an out-of-sync index00:20
→ ppr joined00:21
← yashi left00:21
← tamm0r left00:21
killerchicken doener: hah. that might be it.00:22
← hebz0rl left00:22
killerchicken doener: is that a bug?00:23
← jksz_ left00:23
← scarabx left00:23
doener killerchicken: that the index can get out of sync? Not really. A bunch of things simply don't need it to be (completely) up to date.00:24
killerchicken no, I mean that diff doesn't force an update00:24
doener killerchicken: that a command that actually needs it up to date doesn't "fix" it: probably00:24
→ schmukkk joined00:24
killerchicken ok00:25
→ scarabx joined00:25
doener killerchicken: I'm not familiar with the rules used for updating the index, so I can't really tell whether that's diff's fault, or maybe some other command00:25
killerchicken: btw, which git version? (And did the "git status" thing actually help?)00:26
killerchicken I will try to create a reproducable test case, and send that to the ml00:26
→ jksz joined00:26
killerchicken git status helped, and 1.6.5.200:26
← jksz left00:26
doener ok, so we figured it out even before you went to bed :-)00:26
killerchicken yes :)00:26
doener and I'm once more confused as to how my brain decided to come up with that seemingly idiotic theory00:27
killerchicken The doener never fails to be brilliant00:27
← brettgo1 left00:27
doener I mean, I'd never come up with such an optimization myself, so why the heck could I guess that there is one in place?! (I'm somewhat convinced now that there is one)00:28
→ jksz joined00:29
← Determinist_ left00:29
← tetha left00:31
→ cbreak|evil joined00:31
→ tetha joined00:31
← dreiss left00:31
→ kpreid_ joined00:32
→ dalekurt_ joined00:33
← LiamH left00:34
← cbreak left00:35
cbreak|evil → cbreak00:35
← SRabbelier left00:37
→ ereslibre_laptop joined00:37
→ Pewpewarrows joined00:38
→ hummingkid joined00:39
→ threeve joined00:39
killerchicken if there is some caching going on, it's more subtle than my feeble attempts at a test case could show00:40
→ zachrose joined00:40
→ aisipos joined00:41
→ jrmuizel joined00:41
← mjf left00:41
← Pewpewarrows left00:42
← jrmuizel left00:43
→ TeckniX joined00:44
doener killerchicken: I guess changing the file without affecting the inode number nor the mtime could work00:44
killerchicken sure. But my editor should change the time.00:45
gnr00:46
I'm giving up on this for now.00:46
my bed is waiting :-)00:46
doener heh00:46
killerchicken I guess I better leave.00:46
← brain0 left00:46
doener git nite00:46
killerchicken :)00:46
← killerchicken left00:46
← WebDragon left00:47
← parasti left00:48
← dalekurt left00:50
→ paltman joined00:50
thenetduck is there an easy way to delete a lot of files in a git repo? I'm deleting some old migration files that I condensed for an initial deployment and it says no changes added to commit (use "git add" and/or "git commit -a")00:55
→ section31 joined00:55
← section31 left00:55
← mithro left00:57
selckin what00:57
thenetduck so I deleted some files00:57
and it asks me to manually delete them when I try to "commit -m "messsge" "00:57
selckin try git add -u00:58
thenetduck it tells me I need to do git rm <filename>00:58
what did that do00:58
selckin man git-add00:58
Gitbot selckin: the git-add manpage can be found at http://git.or.cz/man/git-add00:58
thenetduck oh ok that seemed to work. htanks00:59
thanks00:59
jedbrown Is anyone here familiar with git-send-email.perl?00:59
← pygi left01:00
→ danferranti joined01:02
MacGyverNL I know I'm a minority using Windows, but in the msysgit Git Bash, the command "$ git diff | grep -e warning: | wc -l" seems to only pipe from git diff to grep. The output from grep then gets printed to the terminal, and wc -l terminates with "0"01:02
Anybody else seen this behaviour?01:03
charon jedbrown: just ask. asking along the lines "can someone help" is less likely to attract help than an actual question.01:03
→ WebDragon joined01:03
→ ajonat joined01:03
doener MacGyverNL: stupid guess: Does the same happen without the colon?01:03
→ Jinzhu joined01:03
MacGyverNL Yeah.01:03
And with quotes, and everything.01:04
jedbrown charon: I have a new command for format-patch, you can give a query like --notmuch id:<MESSAGE-ID> and it will set up all the headers for you.01:04
selckin MacGyverNL: 2>&101:04
← meanburrito920_ left01:04
jedbrown charon: This works fine, including when called from git send-email, but it still prompts for the information that will be set once I have the response from format-patch.01:04
doener grep output going to stderr by default? Now that would be nice^Hhorrible01:05
MacGyverNL selckin: So $ git diff | grep -e warning 2>&1 | wc -l ?01:05
And about that, stderr by default?01:05
That would be weird.01:05
doener MacGyverNL: for that special case, I'd just pass -c to grep and drop the wc01:05
selckin MacGyverNL: one earlier01:05
← rettub_ left01:05
jedbrown charon: so I want to know how much change it would need just look at the headers produced by format-patch01:05
MacGyverNL doener: d'oh. Slap me.01:05
selckin MacGyverNL: the warning is probably on stderr not stdout01:06
Rolenun oooh,, MavGyver's using git......hey! get back on your Stargate program...the new one sucks! :D01:06
→ rettub_ joined01:06
MacGyverNL selckin: Oooh, no, no.01:06
The warning is simply a part of the diff output.01:06
charon jedbrown: well, that was general advice, you'll have to hope someone else knows about send-email specifics01:06
selckin MacGyverNL: then it should work.01:06
MacGyverNL But it doesn't.01:06
selckin call ghostbusters01:06
MacGyverNL That's what I'm trying to solve :P01:07
jedbrown charon: suggestions? Should I mail the list instead?01:07
MacGyverNL doener: Hmm. grep -c seems to be shorthand for piping it to wc -l...01:07
Because it has the *exact* same behaviour as | wc -l01:07
charon jedbrown: probably. though my gut feeling right now is, if you managed to patch format-patch, the general response to "how hard will doing X in send-email be" might be "well go have a look yourself"01:08
doener MacGyverNL: not here. strace shows only "git" and "grep" being execve'd01:08
MacGyverNL Hmm.01:09
jedbrown charon: Sure, but it's much less organized Perl, format-patch is well-structured C.01:09
MacGyverNL I think this might have something to do with diff using less to display output, or type, or something...01:09
doener MacGyverNL: git should detect the pipe and not invoke less01:09
MacGyverNL Because when I do git status | grep -e modified | wc -l it works fine.01:09
jedbrown charon: so I'm slightly lost because I don't see an elegant way to inject it (after a 10 minute look)01:10
doener MacGyverNL: just for fun, try with "git --no-pager diff ...."01:10
MacGyverNL: and if that still fails, maybe: git diff 2>&1 | ....01:10
MacGyverNL But the grep is working...01:10
So the stdout is what grep is receiving the output on.01:11
charon jedbrown: well, don't let me dissuade you :)01:11
doener MacGyverNL: are you looking for these? warning: CRLF will be replaced by LF in foo.01:11
MacGyverNL Yeah.01:11
← ereslibre left01:11
doener 00:23:30 <doener> MacGyverNL: and at least on the command line, it warns you when it does a crlf -> lf conversion IIRC01:12
MacGyverNL I know.01:12
doener MacGyverNL: and those warning go to stderr01:12
MacGyverNL But I want to count them.01:12
doener because they're not part of the diff output, but, well, warnings01:12
MacGyverNL Again... The *grep* is working.01:12
Or...01:12
selckin so why are you asking if you'r egoing to ignore all our suggestions thinking you know better01:12
MacGyverNL Not...01:12
Wait.01:12
What.01:12
charon jedbrown: maybe you can dig up and cc the guy who wrote the "send-email behaves like format-patch when needed" stuff, i can't remember off hand who did that01:12
doener MacGyverNL: the grep just kills _everything_ on stdin01:12
selckin people are so hopeless01:12
MacGyverNL doener: Yeah, I just realized that.01:12
ereslibre_laptop → ereslibre01:13
← aisipos left01:13
charon jedbrown: i think it was 5df9fcf01:13
Gitbot [git 5df9fcf]: http://tinyurl.com/yahu36p -- git send-email: interpret unknown files as revision lists01:13
MacGyverNL selckin: I'm not ignoring, nor planning to ignore, your suggestions. I'm vocal about my own opinions and thoughts, but so far I have tried every suggestion you've thrown my way.01:13
jedbrown charon: thanks01:14
← selckin left01:14
→ kergoth joined01:15
→ fcuk112 joined01:16
doener kergoth: git for-each-ref can translate the branch.<name>.{remote,merge} settings to a remote tracking branch01:16
kergoth: git for-each-ref --format='%(upstream)' $(git rev-parse --symbolic-full-name HEAD)01:16
← kergoth left01:16
charon MacGyverNL: such warnings definitely go to stderr, see usage.c, so as selckin pointed out several times you'll need git diff 2>&1 | grep -c warning:01:17
doener for once I recall that someone asked a certain question that he didn't get answered to, reply when he joins again, and he immediately leaves :-/01:17
MacGyverNL selckin: I've re-read your suggestions, I misunderstood what you said earlier about warning being on stderr. I thought it was on stdout because I thought I'd tried redirecting stderr for diff earlier. I was wrong, obviously.01:18
Oh, he's gone.01:18
charon also note that 2>&1 has weird semantics wrt its position01:20
doener charon: hm, which ones?01:20
→ mithro joined01:20
charon well, foo 2>&1 | bar redirects stderr to bar, but foo 2>&1 >bar does not, whereas foo >bar 2>&1 does.01:21
→ rchavik joined01:21
MacGyverNL Well, I guess this misunderstanding is one more sign I should go to sleep.01:21
If anyone feels like it, please transfer my sincerest apologies to selckin.01:22
doener charon: not weird if you think of it as a pointer. int *p1, *p2; "p1 = foo; p2 = p1;" vs. "p2 = p1; p1 = foo;"01:22
← fcuk112 left01:22
MacGyverNL And thanks.01:22
← danferranti left01:22
→ meanburrito920_ joined01:22
charon doener: i still think the foo 2>&1 | bar example is weird, and only works because it would be even stranger to have foo | 2>&1 bar (or whatever)01:22
← paltman left01:23
← Guest84405 left01:25
→ Determinist_ joined01:25
← meanburrito920_ left01:26
doener charon: hm, thinking about it, I never treated | as modifying stdout, but only as modifying the next process' stdin. Of course that's technically incorrect. I'll guess I'll now be confused about the redirections again :-/01:26
→ meanburrito920_ joined01:26
← pantsman left01:26
charon :)01:27
charon goes to bed, should have done that hours ago01:28
← meanburrito920_ left01:29
→ meanburrito920_ joined01:29
→ Pewpewarrows joined01:31
→ mre joined01:31
← ereslibre left01:32
← meanburrito920_ left01:32
→ capndiesel joined01:33
← sh1m left01:33
→ sh1mmer joined01:33
→ jinzhu1 joined01:35
← tfarina left01:36
← hummingkid left01:36
→ Zenopus_ joined01:36
← Zenopus left01:36
← cortana` left01:36
→ Pewpew joined01:36
← Pewpewarrows left01:37
← Determinist left01:41
← mre left01:43
← scarabx left01:46
→ Pew joined01:48
← Pewpew left01:48
← jaysern left01:49
← sh1mmer left01:50
← ivenkys left01:51
→ munichlinux joined01:58
← _6a6 left01:59
← munichlinux left01:59
→ _6a6 joined02:01
→ interferon joined02:01
interferon can i determine when a branch was created and what branch it was branched from?02:02
→ hummingkid joined02:03
Pieter no02:04
interferon so what should i do? mention it in a commit message?02:05
Pieter you should probably assume a workflow where it doesn't matter where the branch came from02:06
← ab3 left02:06
interferon hmmm. so i created a branch from master some time ago to implement a feature02:06
→ ab3 joined02:06
interferon now i want to see all the changes i made to implement that feature since i created the branch02:06
← \ask left02:07
Pieter git log master..branch02:07
shows all commits that are on the branch but not in the master02:07
interferon but master has moved forward significantly since the branch02:07
Pieter that doesn't matter02:07
interferon ah nice02:08
that works, thanks02:08
perfect! thanks, pielgrzym02:08
er, Pieter :)02:08
Pieter :)02:08
→ kergoth joined02:08
kergoth anyone know if its possible to get 'tig' to show a view like its default one when given rev-list output rather than a ref?02:09
hmm02:09
ah, nm02:10
→ tekky joined02:10
← mithro left02:11
→ vbgunz joined02:11
← blakemizerany left02:12
kergoth well, that would still be nice, but i guess thisll do02:12
→ obiwan_chinobi joined02:13
← javatexan left02:15
← hgimenez left02:15
← hummingkid left02:16
← kkoehne left02:18
← orafu left02:20
→ orafu joined02:20
→ daviddo joined02:22
← davido left02:22
→ lunks_ joined02:23
← wagle left02:24
→ marc_m joined02:24
→ VVSiz_ joined02:24
← gaveen left02:25
← FunkeeMonk left02:26
→ alanhaggai joined02:26
← aresnick left02:26
→ Determinist joined02:28
← interferon left02:29
→ jaysern joined02:29
→ blakemizerany joined02:30
← tedoc2000 left02:30
→ hummingkid joined02:33
→ aresnick joined02:33
← schmukkk left02:34
← lunks left02:35
← Determinist_ left02:35
← marc_kdab left02:36
← flaguy48 left02:36
← agib left02:36
→ agenteo joined02:37
→ sh1mmer joined02:39
← imajes left02:39
← onigiri left02:40
→ MrCheetoDust joined02:40
← VVSiz left02:42
→ paltman joined02:43
← shadfc left02:44
→ imajes joined02:56
→ cubuanic joined02:58
← jedbrown left02:59
m-takagi_ → m-takagi03:01
→ me22 joined03:06
← tekky left03:06
← me22 left03:09
→ mdmkolbe joined03:09
→ me22 joined03:09
← aziz left03:10
← cubuanic left03:10
← Determinist left03:11
← godsflaw left03:12
→ Mizar-jp joined03:12
← jschoolcraft left03:12
← BgPerl-VB17 left03:15
← sh1mmer left03:17
← marc_m left03:17
← jinzhu1 left03:17
← zachrose left03:17
← awx left03:17
← Davey left03:17
← bugxek left03:17
← ustun left03:17
← Deradner left03:17
← kergoth left03:17
← mdmkolbe left03:17
← berkes left03:17
← ilteris_ left03:17
← schlortoo left03:17
← thijso left03:17
← kristian42 left03:17
← jkp left03:17
← cookiemon5ter left03:17
← JackWinter left03:17
← rotty left03:17
← pdusen left03:17
← univate left03:17
← ftsf left03:17
← [dmp] left03:17
← eshear left03:17
← spacebat_ left03:17
← pigeon left03:17
← hyperair left03:17
← rtyler left03:17
← marlun-w1rk left03:17
← Ingmar left03:17
← teknotus left03:17
← David-T left03:17
← arooni__ left03:17
← TheUni left03:17
← sbob left03:17
← tha left03:17
← s0ber_ left03:17
← sdboyer left03:17
← kadoban_ left03:17
← rektide left03:17
← Whtiger left03:17
← kmap left03:17
← Ratler left03:17
← ianmcorvidae left03:17
← morfic left03:17
← aptituz left03:17
← reaVer left03:17
← aussiegee left03:17
← lucs left03:17
← deavid left03:17
← madmoose left03:17
← piroko left03:17
← bartman left03:17
← jnl left03:17
← Khisanth left03:17
← gernot left03:17
← drfree left03:17
← dmlloyd left03:17
← doki_pen left03:17
← PerlJam left03:17
← trodrigues left03:17
← frsk left03:17
← lfranchi left03:17
← Tuomas left03:17
← sneakret_ left03:17
← LoRe left03:17
← flex left03:17
← hegge left03:17
← m0 left03:17
← jast left03:17
← argonel left03:17
← __bert__ left03:17
← haider left03:17
← bdowning left03:17
← plopix left03:17
← dwmw2_gone left03:17
← jpieper left03:17
← Natch| left03:17
← mstormo left03:17
← bs left03:17
← pfote left03:17
← crab left03:17
← zomg left03:17
← teuf left03:17
← tic left03:17
← russell_h left03:17
← spb left03:17
← ReinH left03:17
← asksol left03:17
← metze left03:17
← exitstate left03:17
← pilif|away left03:17
← merlin83 left03:17
← dsop left03:17
← jjuran left03:17
← Tronic left03:17
← Soliton left03:17
← wanders left03:17
← exg left03:17
← tk left03:17
← solsTiCe left03:17
← uau left03:17
← unreal left03:17
← mrtazz left03:17
← cota left03:17
← Solefald left03:17
← _rane left03:17
← dan_ left03:17
← faheem_ left03:17
← majoh left03:17
← mutex left03:17
← daed left03:17
← RurouniJones_ left03:17
← tmz left03:17
← loop left03:17
← dionoea left03:17
← omega_ left03:17
← Nafai left03:17
← ToxicFrog left03:17
← d1b left03:17
← bauruine left03:17
← drizzd_ left03:17
← jstemmer left03:17
← ^Einstein left03:17
← felipe left03:17
← emias left03:17
← pcc1 left03:17
← agenteo left03:17
← lunks_ left03:17
← vbgunz left03:17
← alanhaggai left03:17
← rettub_ left03:17
← Lisimba left03:17
← UncleCJ left03:17
← dbrock left03:17
← krawek left03:17
← qrush left03:17
← karihre_ left03:17
← maattd left03:17
← thiago_home left03:17
← stamina left03:17
← PataTwo left03:17
← segher left03:17
← ryoma left03:17
← flazz left03:17
← LotR left03:17
← Aikawa left03:17
← Pupeno left03:17
← rlan_ left03:17
← johan-s left03:17
← squentin left03:17
← LRN left03:17
← lamont left03:17
← Adlai left03:17
← rfay left03:17
← NetHawk left03:17
← mapreduce left03:17
← przemoc left03:17
← ferdy left03:17
← john268 left03:17
← eno__ left03:17
← ahf left03:17
← yhager left03:17
← Strogg left03:17
← stuffcorpse left03:17
← jas4711 left03:17
← amirite left03:17
← wagle_home left03:17
← jonesy_ left03:17
← pippijn left03:17
← jacobat left03:17
← jujule2_ left03:17
← ruphy left03:17
← imarcusthis left03:17
← maxb left03:17
← GodEater left03:17
← wereHamster left03:17
← x-demon left03:17
← shabble left03:17
← cYmen left03:17
← Cesario left03:17
← jrockway left03:17
← Hunner left03:17
← mikkom_ left03:17
← kusma_ left03:17
← jjore left03:17
← skyl left03:17
← neldoreth left03:17
← dfr|work left03:17
← Ilari left03:17
← bgerber left03:17
← agile left03:17
← _ikke_ left03:17
← VVSiz_ left03:17
← torben_ left03:17
← Woody2143 left03:17
← unixtippse left03:17
← simonb__ left03:17
← Tommy[D] left03:17
← Bass10 left03:17
← bremner left03:17
← nick_fn left03:17
← dermoth left03:17
← troyt left03:17
← tapple left03:17
← warthog9 left03:17
← levifig left03:17
← kloeri left03:17
← tomaw__ left03:17
← bambule left03:17
← impulze left03:17
← Thralas left03:17
← trochala left03:17
← tg left03:17
← ijuma left03:17
← barrym left03:17
← corecode left03:17
← hansent|away left03:17
← mikem23 left03:17
← fpcfan left03:17
← hatseflats left03:17
← kukks left03:18
→ agenteo joined03:18
→ alanhaggai joined03:18
→ VVSiz_ joined03:18
→ lunks_ joined03:18
→ vbgunz joined03:18
→ rettub_ joined03:18
→ dbrock joined03:18
→ Lisimba joined03:18
→ UncleCJ joined03:18
→ jpieper joined03:18
→ krawek joined03:18
→ qrush joined03:18
→ bauruine joined03:18
→ Natch| joined03:18
→ karihre_ joined03:18
→ maattd joined03:18
→ thiago_home joined03:18
→ merlin83 joined03:18
→ stamina joined03:18
→ PataTwo joined03:18
→ solsTiCe joined03:18
→ segher joined03:18
→ simonb__ joined03:18
→ drizzd_ joined03:18
→ ryoma joined03:18
→ flazz joined03:18
→ LotR joined03:18
→ m0 joined03:18
→ Aikawa joined03:18
→ Pupeno joined03:18
→ rlan_ joined03:18
→ Tommy[D] joined03:18
→ johan-s joined03:18
→ Bass10 joined03:18
→ squentin joined03:18
→ LRN joined03:18
→ rfay joined03:18
→ lamont joined03:18
→ Adlai joined03:18
→ NetHawk joined03:18
→ ReinH joined03:18
→ tic joined03:18
→ zomg joined03:18
→ teuf joined03:18
→ bs joined03:18
→ mstormo joined03:18
→ pfote joined03:18
→ spb joined03:18
→ plopix joined03:18
→ russell_h joined03:18
→ exitstate joined03:18
→ haider joined03:18
→ bdowning joined03:18
→ metze joined03:18
→ pilif|away joined03:18
→ crab joined03:18
→ asksol joined03:18
→ dwmw2_gone joined03:18
→ jjuran joined03:18
→ dsop joined03:18
→ Tronic joined03:18
→ Soliton joined03:18
→ tk joined03:18
→ wanders joined03:18
→ exg joined03:18
→ mapreduce joined03:18
→ przemoc joined03:18
→ ferdy joined03:18
→ uau joined03:18
→ jstemmer joined03:18
→ unreal joined03:18
→ mrtazz joined03:18
→ Solefald joined03:18
→ faheem_ joined03:18
→ RurouniJones_ joined03:18
→ majoh joined03:18
→ tmz joined03:18
→ dan_ joined03:18
→ omega_ joined03:18
→ mutex joined03:18
→ loop joined03:18
→ Nafai joined03:18
→ _rane joined03:18
→ ToxicFrog joined03:18
→ d1b joined03:18
→ dionoea joined03:18
→ daed joined03:18
→ cota joined03:18
→ john268 joined03:18
→ jast joined03:18
→ torben_ joined03:18
→ eno__ joined03:18
→ ahf joined03:18
→ yhager joined03:18
→ Strogg joined03:18
→ ^Einstein joined03:18
→ felipe joined03:18
→ stuffcorpse joined03:18
→ emias joined03:18
→ unixtippse joined03:18
→ Woody2143 joined03:18
→ pcc1 joined03:18
→ jas4711 joined03:18
→ amirite joined03:18
→ wagle_home joined03:18
→ jonesy_ joined03:18
→ pippijn joined03:18
→ skyl joined03:18
→ jacobat joined03:18
→ jujule2_ joined03:18
→ ruphy joined03:18
→ tapple joined03:18
→ imarcusthis joined03:18
→ bremner joined03:18
→ maxb joined03:18
→ GodEater joined03:18
→ wereHamster joined03:18
→ x-demon joined03:18
→ shabble joined03:18
→ cYmen joined03:18
→ Cesario joined03:18
→ jrockway joined03:18
→ Hunner joined03:18
→ jjore joined03:18
→ neldoreth joined03:18
→ mikkom_ joined03:18
→ Ilari joined03:18
→ dfr|work joined03:18
→ bgerber joined03:18
→ agile joined03:18
→ kusma_ joined03:18
→ _ikke_ joined03:18
→ argonel joined03:18
→ nick_fn joined03:18
→ __bert__ joined03:18
→ dermoth joined03:18
→ troyt joined03:18
→ warthog9 joined03:18
→ kloeri joined03:18
→ tomaw__ joined03:18
→ levifig joined03:18
← LRN left03:18
jonesy_ → jonesy03:18
← squentin left03:19
← obiwan_chinobi left03:19
← krawek left03:20
→ krawek joined03:21
← me22 left03:23
→ johnw joined03:23
← grumbel left03:24
← Jinzhu left03:24
daviddo → davido03:24
→ Jinzhu joined03:25
→ cytrinox_ joined03:26
→ bugxek joined03:27
→ tstclair joined03:29
→ mithro joined03:30
← bugxek left03:33
← engla left03:34
→ bugxek joined03:34
→ bcardarella joined03:35
← hummingkid left03:36
← cytrinox left03:37
cytrinox_ → cytrinox03:37
← ab3 left03:37
→ hummingkid joined03:38
→ Keiya_ joined03:44
← WebDragon left03:46
← Keiya left03:46
← agenteo left03:48
← Pew left03:50
→ dmlloyd joined03:51
→ volante joined03:54
volante → ashley_03:54
ashley_ im trying to get git-rerere to work.. the documentation is a little confusing: "Later, after you are done resolving the conflicts, running git-rerere again will record the resolved state of these files."03:55
does that mean i run rerere after i've fixed the files in my working dir, or after i stage them, or after i commit?03:55
doesnt seem to do much either way03:58
infact does nothing :|03:58
→ Khisanth joined04:01
← aresnick left04:02
→ g8tor joined04:06
g8tor hello all04:06
I have a repository with a project at version 1 and I need to start on version 2, but version 2 really shares very little with version 1 except that it is the next version. How can I accomplish this or am I thinking of this all wrong?04:08
← Twisted_Mentat left04:08
← tjaway left04:09
← hummingkid left04:10
← juan_arandaalvar left04:11
→ juan_arandaalvar joined04:11
→ mithro_ joined04:11
← mithro left04:11
→ hummingkid joined04:15
← TeckniX left04:17
→ Davey joined04:18
→ TheUni joined04:18
← tstclair left04:21
johnw g8tor: I would just tag version 1, and start commiting to what will become version 204:24
sitaram ashley_: long ago I did 'git config --global rerere.enabled 1' and after that never bothered to do anything more; it seems to just work04:25
g8tor johnw: Thanks I was thinking that but I wasn't sure.04:26
→ Pewpewarrows joined04:30
ashley_ sitaram: thanks, ill try that04:32
← _6a6 left04:33
← hummingkid left04:34
→ DrNick joined04:36
← bugxek left04:38
→ juan_arandaalva1 joined04:41
→ hummingkid joined04:41
← juan_arandaalva1 left04:41
→ tedoc2000 joined04:41
← jfkw left04:42
→ SLaYaH joined04:42
→ Haarg joined04:43
→ Analias joined04:44
← Haarg left04:44
← SLaYaH left04:45
← tedoc2000 left04:45
qrush → qrush|sleeeeep04:49
← juan_arandaalvar left04:50
← AkumaStreak left04:51
→ cehteh joined04:52
→ drayko joined04:58
← qrush|sleeeeep left04:58
← Pewpewarrows left05:05
← jksz left05:07
→ Twisted_Mentat joined05:07
← threeve left05:08
→ Zaphod joined05:11
Zaphod → Guest9716905:11
→ TheUni_ joined05:12
← TheUni left05:12
TheUni_ → TheUni05:14
nevyn_ → nevyn05:14
← Twisted_Mentat left05:17
→ laj joined05:19
← Dreamr_3 left05:23
← TheUni left05:24
→ psankar joined05:26
← johnw left05:28
→ roop joined05:29
→ TheUni joined05:29
→ munichlinux joined05:30
← TheUni left05:31
→ lunks__ joined05:33
→ TheUni joined05:33
← lunks_ left05:35
→ lightcap joined05:36
→ gnufied joined05:39
→ WALoeIII joined05:42
→ LRN joined05:42
← gnufied left05:43
→ gnufied joined05:43
← cloudhead left05:44
→ scarabx joined05:46
→ Mystikal joined05:49
← LRN left05:51
← vbgunz left05:55
→ luke` joined05:55
luke` How do I import only the HEAD of an svn repository into a local git repo?05:56
→ jceb joined06:03
← eletuchy left06:06
← WALoeIII left06:06
mithro_ → mithro06:08
→ WALoeIII joined06:11
→ johnw joined06:12
← hummingkid left06:13
→ univate joined06:17
← ashley_ left06:19
shuffle2 luke` i think cloning the svn repo will not take any history, not sure though...06:20
← WALoeIII left06:22
← g8tor left06:23
→ hummingkid joined06:24
← johan-s left06:25
→ captain_morgan joined06:26
← jonesy left06:27
→ sis joined06:27
→ AAA_awright joined06:27
→ wagle joined06:27
→ jksz joined06:27
→ solofight joined06:27
→ tjaway joined06:27
→ mdmkolbe joined06:27
→ marc_m joined06:27
→ kergoth joined06:27
→ jinzhu1 joined06:27
→ awx joined06:27
→ Deradner joined06:27
→ berkes joined06:27
→ schlortoo joined06:27
→ thijso joined06:27
→ kristian42 joined06:27
→ jkp joined06:27
→ cookiemon5ter joined06:27
→ pdusen joined06:27
→ JackWinter joined06:27
→ rotty joined06:27
→ ftsf joined06:27
→ [dmp] joined06:27
→ eshear joined06:27
→ spacebat_ joined06:27
→ pigeon joined06:27
→ hyperair joined06:27
→ rtyler joined06:27
→ marlun-w1rk joined06:27
→ Ingmar joined06:27
→ teknotus joined06:27
→ David-T joined06:27
→ arooni__ joined06:27
→ sbob joined06:27
→ tha joined06:27
→ s0ber_ joined06:27
→ sdboyer joined06:27
→ kadoban_ joined06:27
→ rektide joined06:27
→ Whtiger joined06:27
→ kmap joined06:27
→ morfic joined06:27
→ Ratler joined06:27
→ ianmcorvidae joined06:27
→ aptituz joined06:27
→ reaVer joined06:27
→ flex joined06:27
→ aussiegee joined06:27
→ lucs joined06:27
→ deavid joined06:27
→ madmoose joined06:27
→ piroko joined06:27
→ bartman joined06:27
→ jnl joined06:27
→ gernot joined06:27
→ drfree joined06:27
→ doki_pen joined06:27
→ PerlJam joined06:27
→ trodrigues joined06:27
→ frsk joined06:27
→ lfranchi joined06:27
→ Tuomas joined06:27
→ sneakret_ joined06:27
→ LoRe joined06:27
→ hegge joined06:27
→ eletuchy joined06:28
→ yatiohi joined06:28
→ bambule joined06:28
→ fpcfan joined06:28
→ hatseflats joined06:28
→ tg joined06:28
→ ijuma joined06:28
→ impulze joined06:28
→ hansent|away joined06:28
→ mikem23 joined06:28
→ barrym joined06:28
→ corecode joined06:28
→ Thralas joined06:28
→ jspiros joined06:28
→ chty joined06:28
→ eMBee joined06:28
→ stick joined06:28
→ osfameron joined06:28
→ dominikh joined06:28
→ f00li5h joined06:28
→ guerrilla joined06:28
→ dkr21 joined06:28
→ dirker joined06:28
→ up_the_irons joined06:28
→ tychoish joined06:28
→ halfline joined06:28
→ charon joined06:28
→ madduck joined06:28
→ vkroll joined06:28
→ lukesa joined06:28
→ jettero joined06:28
→ knittl joined06:28
→ shd joined06:28
→ roop_ joined06:32
← cao left06:32
← stuffcorpse left06:33
→ stuffcorpse joined06:33
← jceb left06:34
→ JSLint joined06:36
captain_morgan Can someone suggest a good way of searching backward in history for content?06:37
← bcardarella left06:38
→ cao joined06:38
→ girishr joined06:40
← munichlinux left06:41
→ cjs joined06:41
JSLint → munichlinux06:41
← psankar left06:42
cjs Is there a git command that will silently tell me whether the current directory is part of a checkout or not by setting the return status?06:42
Fissure cjs: git rev-parse --quiet HEAD, or something similar06:45
→ roop__ joined06:45
Fissure most of the really basic utility functions are in rev-parse06:45
← krawek left06:46
→ psankar joined06:47
cjs Great, thanks.06:47
→ sh1mmer joined06:47
→ Keiya joined06:48
Fissure although, a lot of commands will fail if not in a repository06:48
← madewokherd left06:48
← bigmack83__ left06:49
Fissure so you could probably use a bunch of others; rev-parse is probably the best choice since it won't fail for other reasons06:49
← roop left06:49
← Keiya_ left06:50
→ jbrennan_ joined06:51
→ karmi joined06:51
← adamm left06:51
→ Sho_ joined06:53
← dbrock` left06:54
→ kssddy joined06:56
← munichlinux left06:56
→ dizy joined06:57
← karmi left06:57
← jbrennan left06:58
dizy hi, i was wondering if someone could tell me how I can do a diff between 2 different commits, of a particular file06:58
← roop_ left06:59
roop__ → roop07:00
→ munichlinux joined07:00
_rane dizy: git diff <c1 sha> <c2 sha> path/to/file07:00
dizy _rane: thanks!07:00
→ eletuchy_ joined07:01
← eletuchy left07:01
← yhager left07:02
← patcito left07:03
→ siprbaum joined07:03
dizy what if i wanted to see the output in textmate ?07:05
_rane git diff ... | mate07:06
dizy ah, the pipe07:06
thanks again07:06
sitaram is in the habit of *always* using "--" before path names, in case there are conflicts07:07
cjs Fissure: well, it would fail if there were no HEAD. :-)07:07
sitaram so... git diff <c1 sha> <c2 sha> -- path/to/file07:07
← univate left07:09
_rane sitaram: what sort of conflicts? I've never had to use --07:11
← mithro left07:11
← hummingkid left07:12
jjuran _rane: Commit ids and branch names tend to be valid pathnames.07:12
wereHamster _rane: if you have both a branch and file named foo and you do 'git diff foo' then that's ambiguous07:12
→ WALoeIII joined07:15
← thiago_home left07:16
← roop left07:17
→ eletuchy joined07:18
→ anno^da joined07:19
→ eletuch__ joined07:21
← eletuchy left07:21
← patmaddox left07:22
← Bass10 left07:24
→ galderz joined07:24
→ univate joined07:25
→ Grahack joined07:25
→ boaz joined07:26
← bauruine left07:28
→ mike3993 joined07:29
→ azol joined07:32
→ giallu joined07:32
→ girishr_ joined07:32
mike3993 hello : )07:33
← jas4711 left07:33
→ bigmack83 joined07:34
→ bigmack83_ joined07:34
→ schmukkk joined07:36
_ikke_ _rane: sometimes git requires to use the --. It sometimes gives an error if you don't07:37
← drayko left07:37
← yann|work left07:37
→ yann|work joined07:38
← eletuchy_ left07:39
→ psoo joined07:39
← Davey left07:39
← bambule left07:41
→ ludde joined07:42
← ryoma left07:42
→ ixti joined07:43
→ sde_antrix joined07:43
← rlan_ left07:43
← vmiklos left07:43
→ rlan_ joined07:43
→ vmiklos joined07:44
← scarabx left07:44
← Grahack left07:45
← lightcap left07:47
← psoo left07:49
← dalekurt_ left07:49
← boaz left07:49
→ JasonWoof joined07:50
sitaram _rane: I should have said "ambiguity" not "conflict". But jjuran and wereHamster have explained it better07:51
crab hi.07:51
sitaram hi crab -- what's up?07:51
→ psoo joined07:51
crab suppose i want to merge (my own, not git's) next into master. i don't know if it's the best idea, so i git checkout -b tmp from master, then git merge next, fix all the conflicts, and examine the result. i determine that it's good. i want to do the same to the real master.07:52
what's the right way to do that?07:53
_rane sitaram: yep, I think I got it :P07:53
sitaram crab: if you did nothing else, just checkout master and merge temp -- it'll be a fast forward07:54
← girishr_ left07:55
sitaram crab: the only fly in the ointment will be the commit message will say "merge next into temp" instead of "merge next into master" but you can --amend that07:55
→ mastro joined07:55
→ Grahack joined07:55
crab sitaram: i'm sneaky. i changed the message to "into master" on tmp ;-)07:55
thanks.07:55
sitaram :)07:56
jacobat crab: You can always just do the merge and then "rollback" if you change your mind07:56
← maattd left07:56
jacobat crab: git reflog is a nice tool in that regard07:56
crab jacobat: (in this case, i wanted other people to also be able to comment on the merge results, hence the need to publish them somewhere other than master)07:57
← WALoeIII left07:57
crab i know, i could still have done it on master and pushed it elsewhere, but oh well.07:57
← Deradner left07:57
→ maattd joined07:58
→ ph^ joined07:59
← psoo left07:59
← azol left08:02
← bigmack83 left08:04
→ evl joined08:06
← mike3993 left08:08
→ bigmack83 joined08:09
→ priidu joined08:11
← JasonWoof left08:15
← bigmack83_ left08:16
← marc_m left08:18
← univate left08:19
→ univate joined08:20
→ ntoll joined08:20
→ miek4949 joined08:22
← ajonat left08:24
→ dizy_ joined08:25
← solsTiCe left08:26
→ yhager joined08:26
VVSiz_ → VVSiz08:27
← solofight left08:27
→ solofight joined08:28
→ karmi joined08:31
→ bigmack83_ joined08:33
→ petrux joined08:34
← dizy_ left08:34
→ hemmecke joined08:34
→ Stravy joined08:34
→ dizy_ joined08:36
→ dexter joined08:37
→ cilly joined08:38
→ tbf joined08:39
← marlun-w1rk left08:41
← dizy left08:41
dizy_ → dizy08:41
→ bauruine joined08:43
→ tamm0r joined08:44
→ snearch joined08:45
m-takagi → m-takagi_08:45
→ justin-george joined08:47
← Theravadan left08:47
← justin-george left08:47
→ justin-george joined08:47
← justin-george left08:47
→ erlehmann_ joined08:47
→ shruggar joined08:50
← JEEB left08:50
→ EugenMayer joined08:52
← cilly left08:53
→ cilly joined08:53
→ dpino joined08:54
EugenMayer hello. Iam using git clone, to create a "middleware" git repo. Goal is Pool <-> Staging (Clone)<-> Customer . Now when I create a other clone of staging(like the customer or a developer), then update staging, the "client"..lets say the cusomter, does not get any updates..as it tries to pull frm the "Pool" Version, which don't has the changes08:54
→ jschoolcraft joined08:55
← munichlinux left08:55
→ bigmack83__ joined08:55
EugenMayer how would I set it up, that every clone of Staging is not using the origin, but staging itself "on pull" ?08:56
→ froschi joined08:56
→ JEEB joined08:57
→ paulboone_ joined08:57
← dizy left08:58
→ LRN joined08:59
paulboone_ → paulboone08:59
wereHamster EugenMayer: why do you have the staging there in the first place if you never want to fetch from it?09:00
er, are you saying that when you clone staging and then pull it tries to fetch from pool?09:01
EugenMayer exactly09:02
→ psoo joined09:02
EugenMayer it pulls for the origin of Staging, therefor Pool09:02
bug staging is there to do some changes ( customer related changes )09:03
wereHamster which url does 'git remote show origin' show in the cloned repo (customer)?09:03
EugenMayer the fetch URL is the one from Staging, same for push09:03
→ johan-s joined09:04
→ lhunath joined09:04
EugenMayer http://paste.debian.net/5234609:04
→ trochala joined09:04
EugenMayer i have obfuscated the URLs, hope you undestand that09:04
→ _6a6 joined09:04
wereHamster well, I don't see why git fetch would fetch from pool instead of staging09:04
lhunath When I merge two branches, I get 'warning: too many files (created: 1085 deleted: 1340), skipping inexact rename detection' -- can someone explain what exactly it has decided to do (completely pass on any rename detection?) -- and why, and whether I can tell it to perform rename detection anyway?09:05
wereHamster are you sure the 'master' branch in staging has all the changes that you want?09:05
EugenMayer it is. Git Staging has arround 5 commits (with logs). The cusomter has none, I get told "it up to date"09:05
wereHamster: I will double chekt to not waste your time :)09:05
← paulboone left09:06
EugenMayer wereHamster: http://paste.debian.net/5234709:06
that's the Staging repo. As you see, 5 changes ahead of its master ( pool )09:06
→ malesca joined09:07
← psoo left09:07
EugenMayer commit 2e1c31b73a88d2d3950f807cea12d9bcf1f2128e09:07
← younger left09:07
EugenMayer that's the version of the Staging repo09:07
← erlehmann left09:07
EugenMayer --09:07
commit 5a5b64f199fa6980b7b6f4f96574cbcf434c533c09:08
wereHamster and git rev-parse origin/master in customer repo?09:08
← bigmack83_ left09:08
EugenMayer git rev-parse origin/master09:08
5a5b64f199fa6980b7b6f4f96574cbcf434c533c09:08
← bigmack83 left09:08
EugenMayer that is the first version of the staging repo, which IS actually the one I have imported from the Pol09:09
*Pool09:09
wereHamster did you run git update-server-info in the staging repo?09:09
← yatiohi left09:09
EugenMayer Wow!09:09
→ db__ joined09:09
wereHamster when you fetch over http you have to run git-update-server-info whenever you change any branches (commit/fetch etc)09:09
EugenMayer That fixed it. Could sou shortly explain?09:09
wereHamster: oh!09:10
db__ Hey is it possible to push also the svn/ref/remotes? I tried the following in .git/config push = svn/refs/remotes/tags/*09:10
EugenMayer well I have written a complete git + webdav turorial09:10
wereHamster db__: push to where? a git repo?09:12
→ bigmack83 joined09:12
db__ yes09:14
EugenMayer wereHamster: are there any docs on this git update-server-info? Do I need it also, when the Staging repo is bare?09:14
wereHamster db__: I don't think so09:14
← miek4949 left09:14
db__ hmm my svn repo consinsts of 80k changesets. it's really annoying to let people refetch everything09:15
wereHamster EugenMayer: man git-update-server-info, and yes, you need to run it always, even in bare repos09:15
Gitbot EugenMayer: the git-update-server-info manpage can be found at http://git.or.cz/man/git-update-server-info09:15
wereHamster db__: er, sorry. Of course it is possible to push tags into a git repo -.-09:15
EugenMayer wereHamster: just to be sure, as I include this information in my howto: Everytime the Stagin server recieves changes?09:15
db__ with the old layout you could do pushing because svn was under /refs09:15
but now it is in /svn09:16
wereHamster db__: but keep in mind that svnh tags aren't really tags, in git they are represented by branches09:16
db__ wereHamster: yeah i know09:16
EugenMayer wereHamster: it does not differ, if changes are done on the the Staging repo directly from the WC. It would be also the case if those changes have been pushed09:16
← luke` left09:17
EugenMayer ?09:17
it seems like it does some "head stracking" in info/refs09:18
wereHamster db__: well, after your peers fetch the svn 'tags' from the central git repository, they'd also have to update remote tracking branches (something like git update-ref svn/refs/remotes/tags/X refs/remotes/origin/tags/X)09:19
EugenMayer: whenever refs change in the staging repo you need to run git u-s-i09:19
← _6a6 left09:19
EugenMayer git u-s-i09:20
git: 'u-s-i' is not a git-command. See 'git --help'.09:20
wereHamster update-server-info09:20
EugenMayer ah sorry.09:20
and "when are refs" changes done. Maybe I should more readup what it is for09:20
wereHamster when you commit, pull etc09:21
EugenMayer seems like it tracks all origin / local versions of the branches09:21
← johnw left09:21
→ zoniq_ joined09:25
EugenMayer wereHamster: is there a reason not to automate git update-server-info in a hook?09:26
wereHamster EugenMayer: non-bare repos aren't usually serverd over http, and for bare repos take a look at hooks/post-update.sample09:28
→ tehbaut| joined09:28
EugenMayer wereHamster: yeah, the repo is not bare because I need to checking some submodules first09:29
← nighthwk1 left09:30
→ icwiener joined09:30
→ zonque joined09:30
← gnufied left09:31
← jaysern left09:31
← priidu left09:34
→ priidu__ joined09:34
← priidu__ left09:34
→ priidu joined09:36
→ ustun joined09:36
← tehbaut left09:44
→ Hendu joined09:49
← JackWinter left09:51
← lamont left09:51
← tehbaut| left09:52
→ JackWinter joined09:53
← dexter left09:55
← Teo` left09:57
→ dizy joined09:59
→ nevyn_ joined10:01
→ koltroll joined10:03
← solofight left10:06
→ solofight joined10:08
← maleknet left10:08
← karmi left10:08
← mugwump left10:09
← eletuch__ left10:11
→ eletuchy joined10:12
← dpino left10:12
→ psynaptic joined10:13
← nevyn left10:13
→ dpino joined10:14
_ikke_ what does patch mean in the context of interactive adding?10:14
→ fcuk112 joined10:14
Ilari _ikke_: IIRC, hunk-by-hunk adding.10:15
_ikke_: There's also 'git add -p <file>' which accesses patch mode directly.10:15
_ikke_ Ilari: so you can select what parts are being added?10:15
Ilari: ok.10:16
Ilari _ikke_: Yes.10:16
_ikke_ It's what i'm looking for10:16
thx10:16
← Guest97169 left10:17
→ maleknet joined10:17
→ tehbaut joined10:19
→ mugwump joined10:22
← bigmack83 left10:23
← bigmack83__ left10:24
→ Wazzzaaa joined10:30
→ gitte joined10:31
← ustun left10:31
→ psoo joined10:33
→ Yuuhi` joined10:33
Wazzzaaa I want to use versrion control for my hobby project. I had SVN but on my new PC I need to re-install SVN server. With git, I do not have this problem??10:33
spb the git client and server are pretty much the same thing10:34
zomg You don't actually need a server for git so yeah.. =)10:35
Wazzzaaa So it is easy to use it only local. And on another machine. just copy some dirs? of make a clone?10:35
zomg: thnx10:35
that's what I want to hear :)10:35
zomg On other machines you just clone and then push/pull for changes10:36
Ilari Wazzzaaa: Easiest way is to create directory, run 'git init' there and you have a repository. Just keep backups.10:36
Wazzzaaa oke10:37
Ilari Wazzzaaa: One thing to take into account if you have multiple repositories / machines involved: $faq non-bare10:37
Gitbot Wazzzaaa: Pushing to non-bare repositories is discouraged. See http://git.or.cz/gitwiki/GitFaq#non-bare10:37
← psoo left10:39
→ engla joined10:40
→ bigmack83 joined10:40
→ bigmack83_ joined10:41
Wazzzaaa ok, thnx for the info. I think I just do a tutorial first to get the feeling...10:41
→ Darkas_ joined10:42
Ilari Wazzzaaa: And one thing to remember about push and fetch (and thus about clone and pull): They don't transfer files. They transfer commits.10:46
→ dexter joined10:46
Wazzzaaa ok10:47
I'm reading now10:47
When I want to work from multiple machines, then I need to set up a public repository? i.e. a repo without a working copy which is available from every machine?10:48
← Akufen left10:49
Ilari Wazzzaaa: Its the easiest way. Not strictly mandatory, but...10:49
← Yuuhi left10:50
← cjs left10:50
← bigmack83 left10:50
Wazzzaaa yeah ok. just checking :)10:51
I think I'm gonna like git10:51
← lhunath left10:51
Ilari Wazzzaaa: Its useful to set 'git config --global receive.denyCurrentBranch refuse'. Makes push-to-HEAD error out instead mess up destination (if you push to non-bare repo).10:52
→ luke` joined10:52
Ilari Wazzzaaa: --global means default for user account.10:52
← koltroll left10:53
← bigmack83_ left10:54
→ NET||abuse joined10:54
→ bhierof joined10:55
Ilari Wazzzaaa: (that setting matters on remote side).10:55
→ bigmack83 joined10:56
→ jedbrown joined11:00
← Stravy left11:01
→ ffrank joined11:01
→ psoo joined11:02
ffrank hi. i noticed an oddity in git-am, but maybe i'm just doing it wrong11:02
i'm applying a larger series of patches by supplying them on the git am command line. some of them fail due to missing email addresses11:03
i need to fix that in .git/rebase-apply/info, right?11:04
Ilari ffrank: Maybe older git version that has header bug?11:04
ffrank Ilari: 1.6.5.2 ...11:04
anyway, fixing it manually works well enough. but then i need to git apply --index .git/rebase-apply/`cat .git/rebase-apply/next`11:05
→ __iron joined11:05
ffrank otherwise, git am -r will fail as nothing is in the index (or changed in the working tree for that matter)11:05
Ilari ffrank: Looking at the failing and working patches in editor, any differences in headers?11:06
ffrank: (even order of headers may matter)11:07
→ bigmack83_ joined11:07
ffrank the failing patches are almost certainly malformed. what i am wondering about is the recovery procedure required in such a case.11:07
Ilari ffrank: Try to fix them up with text editor before running am?11:08
ffrank Ilari: i see. so abort and restart is in order?11:08
→ hendry joined11:09
hendry git reset HEAD filename # Does not seem to un-add or unstage it on a commit. Any advice?11:09
Ilari ffrank: Perhaps the easiest way to deal with corrupt patches.11:09
← bauruine left11:09
→ bauruine joined11:10
Ilari hendry: It removes all staged changes for that file.11:10
ffrank Ilari: alright. in this case, i will stick to my "procedure", as my machine takes its time applying those 1700-odd patches ;/11:10
hendry Ilari: what i want to do, is not commit a certain build file Eclipse b0rked, which I need. Am I doing the right thing?11:11
Ilari ffrank: Its possible to save progress and continue where you left off...11:12
hendry: If the build file is supposed to be tracked, yes, it is correct way.11:12
ffrank Ilari: ah, that's another thing that had me confused. i tried fussing with the numbered patches in .git/rebase-apply/, but to no visible avail. git am -r was not impressed.11:13
← jinzhu1 left11:14
→ intripoon joined11:15
hendry Ilari: ah, the problem was that I was going a `git commit -a`11:16
← psoo left11:17
→ bigmack83__ joined11:17
Ilari hendry: Yeah, with -a it is less powerful than without.11:18
← __iron left11:18
← tamm0r left11:18
← vinse_ left11:18
→ ankit9 joined11:20
→ catphish joined11:20
→ lucsky joined11:20
← bigmack83 left11:21
→ __iron joined11:21
← __iron left11:23
→ rachman joined11:23
→ ankit9_ joined11:24
← ankit9_ left11:24
→ transplant joined11:24
← rchavik left11:25
rachman → rchavik11:25
→ omerj joined11:27
→ rachman joined11:27
→ jds joined11:28
← rchavik left11:29
rachman → rchavik11:30
← alanhaggai left11:30
→ meyering joined11:31
→ __iron joined11:31
→ work_ joined11:32
← cookiemon5ter left11:32
sykopomp what's the git command for getting the kind of information github spits out in its impact graph? (as in, get a list of committers and the number of lines changed for each committer?)11:33
Ilari sykopomp: Shortlog can be used to summarize who did what. But it doesn't include diffstat.11:33
→ cookiemon5ter joined11:34
work_ I created a new svn branch which created revision 1766, now if a fetch the previous revision 1765:HEAD it follows the branch point instead for 1766?11:34
← froschi left11:34
meyering howdy, what am I missing? I'd like to know just the names of the files affected by the latest commit. git log -1 --name-only is *almost* what I want, but includes that pesky log text at the top. I tried -z (would then extract all but first "record", but that doesn't insert a NUL byte between the log and the first name, so it'd be harder...11:34
← ia left11:34
sykopomp Ilari: yeah, I know how to spit out number of commits with shortlog. I'm trying to get a diffstat.11:34
← lunks__ left11:34
Ilari meyering: 'git diff --name-only HEAD^!'?11:35
→ ia joined11:35
Ilari sykopomp: I don't think there's builtin tool. Git is scriptable, so...11:35
meyering Ilari: bingo. Thanks!11:35
→ froschi joined11:37
→ name joined11:39
← Analias left11:39
← bigmack83__ left11:40
← bigmack83_ left11:41
← transplant left11:41
work_ What I am doing wrong I deleted a branch from an svn repository (r1768); I now get an "ambigous argument HEAD" error if I try to rebase on r1768?11:43
→ mgoetz[home] joined11:45
mgoetz[home] hello.11:45
"ld.so.1: git-index-pack: fatal: relocation error: file git-index-pack: symbol deflateBound: referenced symbol not found" on Solaris. What does that suggest? Re-compiling git?11:45
← fero left11:46
→ alanhaggai joined11:46
wereHamster mgoetz[home]: you need libz.so in the ld search path11:46
→ bigmack83 joined11:47
→ bigmack83_ joined11:47
ffrank work_: is the r1768 tag still present? does show-ref work on it?11:47
← Grahack left11:50
work_ show-ref doesn't produce any output11:50
← dexter left11:50
work_ Here's a link to the repository http://code.google.com/p/eurocarb/source/list11:51
ffrank work_: then your repo probably has no notion of r1768 anymore11:51
try and check git reflog for the sha1 and use that instead11:51
→ gordonh joined11:51
← icwiener left11:51
→ icwiener joined11:52
→ ehabkost joined11:52
mgoetz[home] wereHamster: hm, ldd says that it finds the proper libz.so11:53
gordonh hi, which git-cvsimport is there a way to import older tags/branches into the git repository?11:54
wereHamster mgoetz[home]: maybe you have two libz.so (one from sun and one from gnu), IIRC deflateBound is not available in all versions11:54
mgoetz[home]: did you compile git yourself?11:55
→ lunks joined11:55
← bigmack83_ left11:56
→ dexter joined11:57
← loop left11:57
→ laj_ joined11:57
dwmw2_gone → dwmw211:58
→ Ilari_ joined11:58
→ parigaudi joined11:58
→ JEEBcz joined11:58
→ loop joined11:59
← Ilari left11:59
← bigmack83 left12:00
gordonh Basically if I import a CVS repository (basically I want to mirror it) if I have a commit on 1 import and the tag is created on the next the tag is lost, how to I avoid that?12:00
Ilari_ → Ilari12:00
← kpreid_ left12:02
mgoetz[home] wereHamster: thanks, i will ask the owner of the machine12:02
← doener left12:02
→ doener joined12:02
← jksz left12:02
→ d0k joined12:03
← JEEB left12:04
← __iron left12:07
→ bigmack83 joined12:08
→ bigmack83_ joined12:08
← laj left12:10
→ petar joined12:10
→ til joined12:10
→ parasti joined12:11
← hendry left12:11
petar when doing a `find .git/refs/tags' and a `git tag' on a particular oss project (glibc), the output varies (find displays just a few tags, while git tags displays hundreds of tags). why?12:11
← intripoon left12:14
ffrank petar: git tag lists remote tags, apparently12:14
JEEBcz → JEEB12:15
← work_ left12:16
petar ffrank: so you imply that `find .git/refs/tags' will only display local tags? that can't be, since i've never tagged anything in that repo..12:16
← catphish left12:16
Ilari petar: Packed refs.12:16
petar packed refs?12:17
Ilari petar: Take a look inside .git/packed-refs12:17
← lunks left12:17
→ lunks joined12:18
petar Ilari: ah, ok, i see them. but whats the difference between packed refs and .git/refs/tags?12:18
← bigmack83_ left12:23
Ilari petar: Two different ways to store refs.12:23
petar Ilari: ok, thanks, i'm reading git-pack-refs(1) right now12:24
→ __iron joined12:24
← NET||abuse left12:27
← meyering left12:28
→ NET||abuse joined12:29
→ munichlinux joined12:29
← name left12:31
← bigmack83 left12:33
→ tvw joined12:33
→ bigmack83_ joined12:34
→ younger joined12:35
younger hi, how does Git handle concurrent access on the same repository?12:35
if two process do "git clone" from the same repository, will there be a problem?12:36
such as resource competing?12:36
hyperair nope12:36
it's perfectly safe12:36
Ilari younger: fetch is read-only from remote repo POV.12:37
hyperair even pushing while someone is cloning is perfectly safe12:37
younger great, thank you!12:37
→ bigmack83 joined12:37
hyperair iirc there are no weird locks like bzr and svn have, because git just pushes the objects/commits first, and then updates the HEAD when all's done12:37
→ GarethAdams|Work joined12:37
← Wazzzaaa left12:38
Ilari Two pushes on same time can cause push to fail with error message but no real harm.12:38
→ jonesy_ joined12:39
← NET||abuse left12:40
← jmatthews left12:40
→ mcella joined12:42
→ jmatthews joined12:43
→ Captain_Haddock joined12:45
karihre_ → karihre12:45
→ bschindler joined12:45
→ karmi joined12:45
→ vbgunz joined12:46
← jonesy_ left12:47
bschindler Hi - I just wanted to do git commit -a, but I saw this: copied: vtkFlowmapNode.h -> BinGridReader/vtkBinaryUnstructuredGridReader.h12:47
these files have nothing in common basically - how can I prevent git to mark this copy?12:47
the rest was detected nicely12:47
Captain_Haddock Hi, this is the first time that I'm playing with git.. trying to setup a rpeository on unfuddle... getting the following error during my initial commit12:48
Ilari bschindler: Just ignore it.12:48
Captain_Haddock "fatal: 'git@example.unfuddle.com/example/example.git' does not appear to be a git repository12:48
fatal: The remote end hung up unexpectedly"12:48
Ilari Captain_Haddock: Wrong syntax.12:48
Captain_Haddock I guess that would be "push" not "commit" :|12:48
Ilari: I've removed my info and replaced it with "example".12:49
→ NET||abuse joined12:49
Captain_Haddock if that's what you are referring to12:49
oh, this was with "git push unfuddle master"12:49
Ilari Captain_Haddock: Any git URL that points to another system _ALWAYS_ has at least one ':' in it.12:49
Captain_Haddock: Maybe try 'git@example.unfuddle.com:example/example.git'?12:50
Captain_Haddock Ilari: sorry, my bad.. I was not informative enough.. Previous to that I did a 'git remote add unfuddle [email@hidden.address] and then 'git config remote.unfuddle.push refs/heads/master:refs/heads/master', then I added files, did a local commit and then tried the above 'git push unfuddle master'12:52
Is that the correct syntax?12:52
ankit9 → ankit9|afk12:52
← solofight left12:53
ankit9|afk → ankit912:53
wereHamster Captain_Haddock: that would be the correct syntax, but your 'fatal' errors have a different URL (without the colon)12:53
Captain_Haddock brb12:53
Ilari Captain_Haddock: And remote push lines are not used if any respecs are present in push command. So to use that push line, you would have to do 'git push unfuddle'.12:54
→ GDG joined12:55
→ hgimenez joined12:55
GDG how can I make a pull and ignore local changes to a specific file?12:55
→ work joined12:56
work Is it possible to get git-svn to clone a branch rather than a specific revision12:56
work → drd2012:56
Ilari Captain_Haddock: And "fatal: 'git@example.unfuddle.com/example/example.git' does not appear to be a git repository" means its trying to look up the repository on local side.12:57
Captain_Haddock Ilari: I'm getting the same error even with just a plain 'git push unfuddle'12:57
hmm12:57
← przemoc left12:58
Ilari Captain_Haddock: Open .git/config and edit?12:58
→ przemoc86 joined12:58
milli GDG: try "man git-stash"12:59
Gitbot GDG: the git-stash manpage can be found at http://git.or.cz/man/git-stash12:59
doener drd20: git-svn always tries to get a branch with all of its history.12:59
← bigmack83_ left12:59
← bigmack83 left12:59
→ bigmack83 joined12:59
→ squentin joined13:00
Captain_Haddock Ilari: http://gist.github.com/242688 is what I have .. so the url is incorrect?13:01
← froschi left13:04
Captain_Haddock maybe it's an authentication issue?13:04
→ nevyn joined13:05
→ Ryback_ joined13:05
doener kergoth: regarding a question of yours from yesterday: git for-each-ref can translate the branch.<name>.{remote,merge} settings to a remote tracking branch13:05
kergoth: git for-each-ref --format='%(upstream)' $(git rev-parse --symbolic-full-name HEAD)13:05
← drd20 left13:06
m-takagi_ → m-takagi13:07
Ilari Captain_Haddock: Change the first '/' in URL to ':'.13:07
← nevyn_ left13:08
Captain_Haddock Ilari: looks promising :)13:08
Ilari: that did it, thanks.13:09
so, pebkac eh?13:09
← luke` left13:09
Ilari Captain_Haddock: Yeah. Incorrect URL.13:10
→ peteinlux joined13:10
→ peteinlux_ joined13:13
EugenMayer in the docs "If the key submodule.$name.update is set to merge, this option is implicit." ... where do I configure that?13:13
← peteinlux_ left13:14
← hipe left13:16
ffrank EugenMayer: git config?13:17
→ bigmack83_ joined13:17
EugenMayer ~/.gitconfig or gitrepo/.git/config?13:17
← gordonh left13:18
EugenMayer does someone work with git submodule and using --merge?13:18
or --rebase13:18
ffrank "git config -l" - "git config submodule.$name.update merge"13:18
EugenMayer ah cool13:18
← maleknet left13:18
→ rolfb joined13:18
← peteinlux left13:19
→ hipe joined13:19
→ flaguy48 joined13:19
← munichlinux left13:20
→ aeiou joined13:20
sitaram orion: ping13:20
EugenMayer ffrank: when I clone a git repo, which has submodules, and in this new git repo, I call13:20
git submodule init13:20
→ tamm0r joined13:21
EugenMayer git submodule update --merge13:21
aeiou hi, what version of git is required for redmine 0.8.7 ?13:21
EugenMayer I should get all submodules without detaching them from head, or?13:21
ffrank EugenMayer: i don't know about submodules, sorry13:21
← Adlai left13:21
aeiou im running git 1.6.3.3 but im just getting "13:21
The entry or revision was not found in the repository."13:21
orion sitaram: pong!13:21
← karmi left13:22
← Keiya left13:23
→ Adlai joined13:23
aeiou ive jsut upgraded from 0.8.2 > 0.8.7 and now git doesnt seem to want to work!13:23
ffrank aeiou: don't the redmine release notes say?13:23
← rchavik left13:25
aeiou ffrank, nope - cant see anything13:25
← omerj left13:26
← parigaudi left13:27
→ tstclair joined13:27
→ botanicus joined13:29
← kssddy left13:30
← bigmack83_ left13:31
← bigmack83 left13:31
→ yatiohi joined13:31
→ bigmack83 joined13:32
→ Fullmoon joined13:32
← rlan_ left13:32
→ rlan_ joined13:32
← Lisimba left13:33
→ Stravy joined13:34
→ fero joined13:34
aeiou bah redmine cant find my git binary13:34
→ Akufen joined13:34
→ Aku joined13:34
← tvw left13:35
← hipe left13:36
→ hipe joined13:36
→ peteinlux joined13:39
→ mm_work joined13:39
→ Teo` joined13:41
→ reprore joined13:41
→ froschi joined13:43
→ mithro joined13:43
← xenoterracide left13:44
← malesca left13:46
→ malesca joined13:47
→ d0k_ joined13:47
← edenc left13:47
→ Bass10 joined13:47
← flaguy48 left13:48
← Teo` left13:49
← trochala left13:49
→ Teo` joined13:49
→ bigmack83_ joined13:49
← berkes left13:51
→ berkes joined13:52
→ bentob0x joined13:52
→ jfkw joined13:52
→ LiamH joined13:55
← snearch left13:55
→ bdiego joined13:56
← cota left13:56
← d0k left13:59
d0k_ → d0k13:59
← psynaptic left14:00
→ edenc joined14:01
← girishr left14:01
← mithro left14:02
← malesca left14:03
← ringlej left14:03
← haider left14:04
kergoth doener: ah! great, thanks much14:04
← Captain_Haddock left14:04
EugenMayer could it be that submodule is currently totally waste?14:05
← kergoth left14:05
→ Sigma joined14:08
→ jreed joined14:09
→ malesca joined14:09
→ TeckniX joined14:10
← ftsf left14:10
→ omerj joined14:11
← anno^da left14:12
jreed Git newbie question: I have two branches, I want to remove files from one branch, keep in the other branch, but be able to continue to merge from the branch where the files were deleted. Any easy way to do this? Right now, I'm copying into /tmp, git rm'ing, change branches, git merge, then git adding back in. Can I merge but just skip one changeset or something?14:12
→ hugoxrosa joined14:13
→ Lisimba joined14:13
→ hummingkid joined14:13
← lukesa left14:13
ffrank jreed: maybe you want to cherry-pick from the other branch rather than merging?14:14
shruggar just merge and when the conflict comes up saying "deleted by X but changed by Y", resolve it by deleting14:14
← hgimenez left14:14
→ corni joined14:15
→ yashi joined14:15
jreed yeah, maybe cherry pick... never used that. i just basically want to always merge but occasionally skip a commit.14:15
not sure i understand shruggar, but i can try that.14:16
oh wait, maybe i do understand shruggar. only problem is I never want to merge the other direction.14:17
← jjuran left14:17
jreed branch A is core platform, branch B is customer customizations. trying to move stuff that was in core to "B" branch basically.14:18
ffrank "always merge"?14:18
jreed: you want to rebase B to A occasionally14:18
shruggar cherry-pick is _very_ rarely what is desired14:18
jreed so everything in "A" i want in B, but never the other way14:18
ok, i'll study up on rebase14:18
→ ctusar joined14:19
jreed and unfortunately i've got some stuff in A that needs to only be in B. thus my question.14:19
← dominikh left14:20
shruggar merge it, and as part of the merge-commit, remove the stuff you don't want, is the "general" answer14:20
ffrank jreed: there are changes in A that should be in B, but not in A anymore?14:20
→ dominikh joined14:20
jreed using --no-commit perhaps?14:20
← yashi left14:20
jreed right ffrank14:20
→ yashi joined14:21
Weiss wants something like "git reset -p" (unstage or forget changes, patch by patch)14:21
→ karmi joined14:21
ffrank jreed: then here's what i'd do: cherry pick those changes to B (yes, i know), then revert them in A (or rebase -i them out, if the branch isn't public)14:21
← paltman left14:21
jreed ok14:22
thank you very much for the help, both of you!14:22
→ threeve joined14:23
jreed unless you disagree violently with shruggar, i may try his suggestion first, as it uses commands i know better.14:23
shruggar I suppose full-disclosure might be to mention that I use workflows which almost never "merge"14:24
ffrank i don't understand why a merge was suggested, if only a limited set of changes is desired, but my understanding of merges isn't quite deep14:24
jreed i will report back momentarily on how well this works...14:25
← reprore left14:28
jreed ok, so i merged and got:14:29
→ fualo joined14:29
jreed # deleted: grails-app/controllers/JobController.groovy14:29
# modified: grails-app/controllers/TimeSheetController.groovy14:29
# new file: grails-app/domain/TimeSheetEntry.groovy14:29
# deleted: grails-app/views/job/create.gsp14:29
# deleted: grails-app/views/job/edit.gsp14:29
# deleted: grails-app/views/job/grid.gsp14:29
# deleted: grails-app/views/job/portlet.gsp14:29
# modified: grails-app/views/timeSheet/create.gsp14:29
# modified: grails-app/views/timeSheet/edit.gsp14:29
# modified: grails-app/views/timeSheet/print.gsp14:29
now i want to undo all the deletes.14:29
but14:29
(~/Source/web/netcenter) git checkout -- grails-app/views/job/create.gsp14:29
error: pathspec 'grails-app/views/job/create.gsp' did not match any file(s) known to git.14:29
forgot i was getting that last time i tried this. drat.14:30
shruggar git checkout HEAD^ -- grails-app/views/job/create.gsp may do it14:30
jreed (~/Source/web/netcenter) git checkout HEAD^ -- grails-app/views/job/create.gsp14:30
error: pathspec 'grails-app/views/job/create.gsp' did not match any file(s) known to git.14:30
shruggar git checkout HEAD^:grails-app/views/job/create.gsp ?14:31
jreed that might work, now i get:14:31
(~/Source/web/netcenter) git checkout HEAD^:grails-app/views/job/create.gsp14:31
grails-app/domain/Job.groovy: needs merge14:31
error: you need to resolve your current index first14:31
shruggar #the "manual way" git show HEAD^:grails-app/views/job/create.gsp > grails-app/views/job/create.gsp14:31
jreed let me resolve that and see what happens14:31
ok14:32
shruggar don't "just resolve", since getting rid of the deletions is part of the conflict14:32
→ NETabuse joined14:32
shruggar (if I'm reading correctly)14:32
← bigmack83 left14:34
← bigmack83_ left14:34
jreed now it just says:14:34
(~/Source/web/netcenter) git checkout HEAD^:grails-app/views/job/create.gsp14:34
D grails-app/controllers/JobController.groovy14:34
M grails-app/controllers/TimeSheetController.groovy14:34
A grails-app/domain/TimeSheetEntry.groovy14:34
D grails-app/views/job/create.gsp14:34
D grails-app/views/job/edit.gsp14:34
D grails-app/views/job/grid.gsp14:34
D grails-app/views/job/portlet.gsp14:34
M grails-app/views/timeSheet/create.gsp14:34
M grails-app/views/timeSheet/edit.gsp14:34
M grails-app/views/timeSheet/print.gsp14:34
but doesn't seem to do anything. weird.14:34
sitaram all pastebins dead?14:35
jreed maybe i'll just do the manual thing this time around :-)14:35
← Aku left14:35
→ sde_antrix2 joined14:35
sitaram jreed: could you put long pastes like that into a pastebin please?14:35
jreed ok sure sorry'14:35
shruggar yeah, manual sounds safe, I don't actually know about git checkout HEAD^:grails-app/views/job/create.gsp in terms of whether it even knows what is meant14:35
← NETabuse left14:35
← sde_antrix left14:36
← dpino left14:36
EugenMayer someone using braid in here?14:36
← NET||abuse left14:37
offby1 It wasn't me, I swear14:37
← offby1 left14:37
← botanicus left14:39
→ Analias joined14:39
→ botanicus joined14:39
jreed oh well, maybe i'll try ffrank's suggestion next time.14:40
→ dpino joined14:43
→ godog joined14:43
→ mbroeker joined14:44
← EugenMayer left14:45
→ psoo joined14:45
← godog left14:45
→ kukks joined14:46
→ muthu joined14:50
→ loupgaroublond joined14:50
loupgaroublond hi all14:50
→ markoa joined14:51
loupgaroublond granted i shouldn't be mixing git-svn with submodules, but is it possible to have 'git svn dcommit' commit the contents of the submodule instead of just the git submodule metadata stored in the parent repo?14:51
→ jbrennan joined14:51
← psoo left14:52
← Lisimba left14:54
← mbroeker left14:55
ffrank jreed: unstaging deletes? wouldn't git add <file> do that? :)14:56
→ psoo joined14:56
Ilari 'git reset <file>'?14:57
shuffle2 if i have a few commits on a local branch, can i dcommit just one?14:57
← jbrennan_ left14:58
→ Turias joined14:58
← Turias left14:58
Ilari shuffle2: Use interactive rebase to squash them first.14:58
_ikke_ shuffle2: you can't commit commits ;)14:58
shuffle2 boo!14:59
so eh, wtf then ...14:59
good thing it's only a few lines...14:59
_ikke_ you want to push a single commit?14:59
shuffle2 yes15:00
← psankar left15:00
_ikke_ that doesnt't really makes sense for git15:00
← schmukkk left15:00
shuffle2 actually it's not a commit yet, so i was just going to make a branch and do a commit there -_-15:00
← sh1mmer left15:01
_ikke_ the changes are still in te staging area?15:01
shuffle2 yeah15:01
but i have commits too15:01
of other changes15:01
Ilari shuffle2: Ah, git svn also has set-tree to commit to SVN.15:01
_ikke_ well, if you just checkout the new branch, those changes are switched over15:01
← psoo left15:02
← imajes left15:02
shuffle2 set-tree looks good15:03
→ Lisimba joined15:04
← aeiou left15:05
→ WebDragon joined15:06
← Pupeno left15:07
younger hi, has anyone used -o or --output option of git archive? did it work for you?15:07
for me, it was error: unknown switch `o'15:07
← corni left15:11
← fero left15:11
← tstclair left15:12
Ilari younger: Works fine for me. Old git version?15:12
← __iron left15:12
← ankit9 left15:12
→ tstclair joined15:13
younger 1.6.0.415:13
→ Pupeno joined15:13
Ilari younger: That's probably too old...15:13
younger llari: can you paste the command you used?15:13
Ilari younger: "git archive -o foo.tar HEAD"15:14
→ mbroeker joined15:14
→ paltman joined15:15
younger the example does not work for me. probably should try a different version15:15
thank you!15:15
→ offby1 joined15:16
← microm left15:16
Ilari younger: I use git version 1.6.6.rc0.256.g0f4f215:17
younger thank you llari, i will try the latest stable15:18
Ilari younger: Latest master...15:18
younger nono, i am lazy and i use pre-built packages whenever possible...15:21
→ madewokherd joined15:21
→ girishr joined15:21
← bauruine left15:21
→ Hronom joined15:21
→ ankit9 joined15:21
→ Haarg joined15:22
→ suman joined15:23
rettub_ → rettub15:24
Ilari younger: Those are often out of date.15:25
← loupgaroublond left15:26
younger yes, i know, but it is impossible to keep track of all latest software, so it is much easier to handle that with a package tool such as apt15:27
unless i really need the latest features15:27
but thank you pointing that out!15:27
Ilari younger: The way I do it: Have git.git clone and periodically fetch it and then rebase and install.15:28
ffrank apt-cache policy git-core ----> Candidate: 1:1.6.5.3-115:28
doesn't look out of date to me15:28
erlehmann_ → erlehmann15:28
← kbingham left15:28
younger i use 9.04, are you using 9.10?15:28
ffrank this is debian unstable15:29
← jedbrown left15:29
Ilari ffrank: 1.6.5.3 is already bit out of date... :-)15:29
ffrank /topic15:29
younger .............15:29
Ilari ffrank: (yes, it is newest non-rc, but)...15:30
→ bcardarella joined15:30
younger llari: i will follow your advice. trying to get the packages from two repositories seems to be very confusing15:30
ffrank bah. RCs everywhere. you torvalds people are all the same ;P15:31
Ilari ffrank: As primary version, I follow next. Also have testing version being pu + some patches.15:31
← GDG left15:32
← arooni left15:32
ffrank Ilari: are you a git developer?15:33
dexter Ilari: any new cool features will arrive with 1.6.6? ;)15:33
← bdiego left15:34
Ilari ffrank: No.15:34
← nothingHappens left15:35
Ilari dexter: Smart HTTP.15:35
dexter Ilari: ?!15:36
Ilari dexter: "git rev-list v1.5.6.3..origin/master | wc -l" => 543315:36
Oops.15:36
388 if one uses proper version.15:36
dexter: Oh, and also 'notes'.15:37
<HEAD:Documentation/RelNotes-1.6.6.txt>15:38
Gitbot [git RelNotes-1.6.6.txt]: http://tinyurl.com/yk27jdu [blob]15:38
→ nothingHappens joined15:38
→ fero joined15:38
← przemoc86 left15:38
→ vital joined15:39
← markoa left15:40
→ przemoc joined15:42
→ jeffxl joined15:43
gbacon given a branch with merge commits, how can I seed git-rerere with the conflict resolutions?15:44
← sergiodj left15:45
← eletuchy left15:45
→ tfarina joined15:45
→ dexter__ joined15:48
→ scarabx joined15:48
← girishr left15:48
← dexter left15:49
dexter__ → dexter15:50
← tswicegood left15:50
→ voker57 joined15:52
← kloeri left15:53
← johan-s left15:54
charon gbacon: briefly put, redo the merge and take the resolution from the original merge. see <HEAD:contrib/rerereimport/rerereimport.sh> and <HEAD:contrib/rerere-train.sh>15:56
Gitbot charon: I'm sorry, there's no such object: HEAD:contrib/rerereimport/rerereimport.sh.15:56
→ thiago_home joined15:56
← magcius left15:57
gbacon charon: thanks!15:57
→ magcius joined15:57
charon hmm, either i'm blind or the bot is broken again15:57
<HEAD:README>15:57
Gitbot [git README]: http://tinyurl.com/yaf2wxw [blob]15:57
charon shrugs15:58
→ javatexan joined15:58
→ nighthwk1 joined15:59
Pieter <HEAD:contrib/rerere-train.sh>15:59
Gitbot [git rerere-train.sh]: http://tinyurl.com/yh8gous [blob]15:59
Pieter the first one you mention doesn't exist15:59
→ agib joined15:59
Pieter Gitbot only looks for the first thing15:59
charon hmm. i wonder why my tree has it then.16:00
perhaps it's a case of "a month of research can frequently save a few hours in the library" on my part16:00
Pieter perhaps you created it?16:00
:)16:00
→ psoo joined16:03
→ rchavik joined16:03
→ dazman joined16:04
→ AkumaStreak joined16:04
← voker57 left16:08
→ keystr0k joined16:08
shuffle2 blah!16:10
git diff only shows the unstaged files?16:10
how about the...reverse of that? :p16:10
→ eletuchy joined16:11
charon git diff --cached?16:11
← Kudos left16:12
→ eletuchy_ joined16:12
→ DavidKlein joined16:12
→ tedoc2000 joined16:13
→ adamm joined16:14
← DavidKlein left16:14
→ DavidKlein joined16:14
← psoo left16:14
← evl left16:19
→ Kudos joined16:19
→ jedbrown joined16:24
→ imajes joined16:25
→ bob765 joined16:27
bob765 what's the cleanest way to rebase to the repo I forked after that repo has been recreated and my fork would not rebase cleanly?16:27
← eletuchy left16:28
→ snearch_ joined16:29
← petar left16:29
→ jaysern joined16:29
→ patcito joined16:30
← dexter left16:31
← bob765 left16:32
← imajes left16:32
→ intripoon joined16:32
← karmi left16:35
← bschindler left16:39
→ mikepence joined16:39
→ matthewmcculloug joined16:39
← ffrank left16:40
← tamm0r left16:41
mikepence hello16:41
I am looking to work on a Rails project and to have its plugins checked out from my own github forks of them, concurrently16:42
would git submodules be the best approach for this?16:42
← edenc left16:42
→ edenc joined16:42
mikepence this being, project and project subkdirectories coming from various git sources16:42
← galderz left16:45
→ thom_ joined16:47
→ jceb joined16:48
→ sh1mmer joined16:48
thom_ I did a "git checkout *blob-id*" to look at the first version of my checked in file, but now when I do a git log it doesn't show any of my changes. how do I get back to my most recent version????16:49
Ilari mikepence: Rails projects use submodules a lot.16:49
← gbacon left16:49
Ilari thom_: 'git checkout master'?16:49
mikepence Ilari, thanks, that seems to be the way to go16:49
← db__ left16:50
← lunks left16:50
← Jinzhu left16:50
→ lunks joined16:51
thom_ hmmm, so "git checkout" moves the head pointer???16:51
← DrNick left16:51
← scarabx left16:51
→ aziz joined16:53
← jaysern left16:54
→ jtrindade joined16:54
→ bigmack83 joined16:55
cbreak it also changes the working directory16:55
← matthewmcculloug left16:55
→ Grahack joined16:55
→ dreiss joined16:56
← malesca left16:57
charon thom_: note that in git, HEAD is the current commit that your worktree is looking at. it is _not_ the "newest" commit in any sense.16:59
(it's quite hard to define what "the" newest commit should be in a DVCS, anyway.)16:59
and yes, because checkout changes what you're looking at, it changes HEAD.16:59
→ khmarbaise joined17:00
→ matthewmcculloug joined17:00
← capndiesel left17:02
→ imajes joined17:02
→ Theravadan joined17:03
→ Aikawa_ joined17:04
← mcella left17:06
→ Keiya joined17:07
← drry left17:07
→ sergiodj joined17:07
→ malesca joined17:08
→ drry joined17:08
→ ehabkost_ joined17:10
← froschi left17:10
← Teo` left17:12
← Aikawa left17:12
→ Teo` joined17:12
→ bigmack83_ joined17:13
→ Tk_L joined17:14
← ehabkost left17:14
→ Sake joined17:17
Sake how come when I clone a github repo only one of the branches exist in my local copy?17:17
teuf the others must be here as remote branches17:18
git branch -a ?17:18
→ DrNick joined17:18
→ zonque_ joined17:18
Sake so git branch -a gives "master \n origin/HEAD \n origin/development \n origin/master"17:19
teuf so here is the other branch ?17:19
Sake it's origin/development17:20
so how do I work on that branch locally?17:20
teuf git checkout -t origin/development17:20
← captain_morgan left17:20
← matthewmcculloug left17:20
Sake why -t?17:20
→ vinse joined17:20
→ matthewmccull joined17:21
rolfb --track ?17:21
teuf Sake: mainly so that it guesses the name of the local branch using the name of the remote one17:21
Sake so how come when you clone it only puts master locally?17:22
I thought a clone was supposed to be a full replica of the remote repo17:22
teuf it is17:22
thiago_home when you clone, it creates one branch only17:23
teuf origin/development is available locally17:23
thiago_home with the same name as the remote's HEAD17:23
if you need more, you create them17:23
teuf there is just no corresponding local branch to hack on it17:23
← rchavik left17:23
← malesca left17:23
← MrCheetoDust left17:24
Sake oh, so when I git checkout -t origin/development17:24
← mastro left17:24
Sake it actually takes the origin/development branch stored locally and puts it in a local branch called 'development'?17:25
teuf it just changes a few "pointers" in local files17:25
→ malesca joined17:25
← zonque left17:26
eno__ → eno17:26
← engla left17:27
Sake teuf: I follow now17:27
→ razvand joined17:27
← Stravy left17:28
→ Alien_Freak joined17:28
→ evl joined17:28
← Hendu left17:28
→ tbuehlmann joined17:29
→ jforman joined17:30
← jtrindade left17:30
jforman i recently adding a submodule to one of my git repos, but i notice that everytime i clone this repo somewhere new, i have to do the git submodule init/update dance. is there a way to do this init/update automatically when i clone my repo?17:30
← giallu left17:31
wereHamster jforman: nope17:31
← matthewmccull left17:32
→ matthewmccull joined17:32
Tk_L where do I specify the port when I clone a project, if the default (22) is changed17:32
jforman wereHamster: thanks17:33
← malesca left17:34
← jforman left17:34
← orion left17:34
rolfb Tk_L: git://host.xz[:port]/~user/path/to/repo.git/17:36
→ j416 joined17:36
Tk_L rolfb: when I do this git clone git://ip_address:30000/home/git/project117:40
I get an error17:41
Initialized empty Git repository in /Users/hg/rails/project1/.git/17:41
cworth What's the standard way to make git complain about trailing whitespace at pre-commit time?17:41
Tk_L fatal: protocol error: bad line length character17:41
← shruggar left17:41
cworth (I used to use "chmod a+x .git/hooks/pre-commit" for this, but it looks like that functionality is missing from the current pre-commit.sample now.)17:41
Is there a configuration option that does this? (I notice that "git diff" knows about the problems with the core.whitespace option and displays them in red at least.)17:42
← javatexan left17:47
→ jlilly joined17:47
rolfb Tk_L: do you put a space in there?17:48
Tk_L rolfb: its underscore ip_address17:49
rolfb :)17:49
try without git://17:50
→ malesca joined17:50
rolfb or with ssh:// instead of git://17:50
→ ryoma joined17:51
→ mederEsite joined17:51
→ FunkeeMonk joined17:51
mederEsite when you guyse use difftool and vimdiff does it always prompt you "hit return to launch", for every subsequent file?17:51
Tk_L rolfb: with ssh:// I get asked a password . ..17:53
I dont remember my password . .17:53
rolfb try putting username@ before the hostname17:54
if you have keys17:54
→ dizy_ joined17:55
Alien_Freak would a git push origin pull my tags too? or do I have to do something special?17:55
Pieter you'd have to pull, at least17:55
→ t0rc joined17:56
← intripoon left17:56
→ johan-s joined17:56
→ jtrindade joined17:57
Alien_Freak no.. wait.. I want to write my local tags to remote why would pull matter?17:57
Pieter you were talking about pulling, not me17:57
← Creap left17:57
← rolfb left17:57
→ corni joined17:57
→ crazychenz joined17:57
→ Creap joined17:58
→ WebDragon|away joined17:59
Alien_Freak bad choice of verb. I meant once I create a locale tag, I want to push it out to origin, would a git push origin be enough?17:59
← sh1mmer left18:00
mederEsite wow18:00
git diff is great18:00
i just realized how great it is18:00
heh18:00
→ krawek joined18:00
→ maattd_ joined18:02
← alanhaggai left18:03
Rolenun git has a lot of very nice features :)18:04
← jceb left18:04
Alien_Freak never mind.. I got it18:04
mederEsite I was using difftool and it opened every file and I had to manually close them18:04
that was annoying as hell18:05
Alien_Freak git push --tags origin18:05
Rolenun lol18:05
mederEsite difftool by itself is great18:05
← JakeSays left18:05
mederEsite with vim-like controls18:05
Rolenun I like gitk too :)18:05
→ JakeSays joined18:05
Nugget I use `tig` mostly because I hardly ever run git locally. it's always on the other side of an ssh session.18:06
http://jonas.nitro.dk/tig/18:06
← WebDragon left18:06
mederEsite i wish my eyes werent so diff virginized18:06
Nugget it's like a curses gitk18:06
→ loincloth joined18:07
loincloth heyyy18:07
is there a way to change the user.email on existing commits ?18:07
→ gbacon joined18:08
loincloth i rebuilt my machine recently and forgot to set that global config before i did a bunch of work18:08
thiago_home loincloth: via git-filter-branch. That is, by rewriting history.18:08
loincloth hrmm18:08
i don't like the sound of that18:08
thiago_home have you pushed those commits to other people?18:08
→ Shyke joined18:08
Shyke I seem to somehow have changed the branch to "t". This of course has all of my local chances as a difference. How do I switch back to local?18:09
loincloth thiago_home: no18:09
→ lhz joined18:10
thiago_home loincloth: then you can use git filter-branch on them18:10
loincloth hrmm18:11
→ Sigma[Mtp] joined18:11
loincloth ok i'll think about it :) thx18:11
← dizy left18:11
dizy_ → dizy18:11
← Theravadan left18:12
→ dalekurt joined18:12
Shyke Do any of you know how?18:14
tstclair → tstclair_afk18:14
→ radarek joined18:15
→ voker57 joined18:16
← Keiya left18:16
← matthewmccull left18:16
← hemmecke left18:16
← gitte left18:16
← t0rc left18:16
→ Keiya joined18:16
← johan-s left18:17
→ juan_arandaalvar joined18:18
← bigmack83_ left18:18
← Shyke left18:19
← bigmack83 left18:19
→ bigmack83 joined18:19
← bigmack83 left18:19
→ maattd__ joined18:20
Rolenun Shyke: git checkout local18:22
← Sigma left18:22
← priidu left18:22
→ priidu joined18:22
→ bigmack83 joined18:23
→ WALoeIII joined18:23
← maattd left18:23
→ dominikh1 joined18:23
→ priidu__ joined18:23
← maattd_ left18:24
maattd__ → maattd18:26
← khmarbaise left18:29
→ drizzd joined18:29
← cookiemon5ter left18:30
← priidu__ left18:31
← dpino left18:31
← priidu left18:32
← Tk_L left18:33
← loincloth left18:33
→ psoo joined18:34
← Keiya left18:35
→ Keiya joined18:35
→ priidu_ joined18:35
→ godsflaw joined18:36
← botanicus left18:36
→ dereine joined18:38
dereine hi. how can i execute svn hooks when i commit stuff via git svn dcommit?18:38
← j416 left18:40
→ MetaCosm joined18:41
→ Determinist joined18:41
← Determinist left18:41
→ bambule joined18:42
← petrux left18:42
→ ablemike joined18:43
← priidu_ left18:45
← drizzd_ left18:45
→ priidu_ joined18:46
← priidu_ left18:46
← zonque_ left18:46
→ priidu joined18:47
→ jas4711 joined18:47
→ t0rc joined18:48
← jstemmer left18:49
→ vital_ joined18:49
← vital left18:50
← bentob0x left18:58
→ loincloth joined19:00
→ HektoR_ joined19:00
HektoR_ hi all... guys how to resolve conflicts ?? what ways are there ?19:00
Ilari dereine: Don't they execute anyway?19:00
→ \ask joined19:01
← MetaCosm_ left19:01
Ilari HektoR_: The basic way is to look up the files that have them (git status) then open each in text editor, search for <<< and fix the conflicts. Then add.19:01
dereine Ilari: they do if yo commit with svn commit19:01
Ilari dereine: Client side hooks???19:01
dereine HektoR_: you can also use git mergetool19:01
← iT|Free left19:01
dereine Ilari: no repo-side-hooks19:02
Ilari dereine: Wonder how dcommit manages to bypass those?19:02
dereine Ilari: good question :) i gues it is not using svn commit internally19:02
HektoR_ Ilari dereine thanks guys ... i'll try19:02
Ilari dereine: Its the same protocol on the wire.19:02
wereHamster dereine: git-svn uses the perl bindings..19:03
→ iT|Free joined19:03
← FunkeeMonk left19:03
→ jstemmer joined19:04
→ tamm0r joined19:05
← zoniq_ left19:06
← tango_ left19:08
← vital_ left19:08
→ vital_ joined19:08
Sake I have a tmp/logs folder in my repo that contains a folder called "empty". I want the file "empty" in the repo but I want to .gitignore everything else. What do I put in my .gitignore to make this happeN?19:10
← priidu left19:10
← tedoc2000 left19:11
Thumper_ Sake: tmp/logs/empty is a directory (folder) or a file?19:12
tamm0r do you have a file called empty or a folder called empty?19:12
← wagle left19:12
→ t0rc_ joined19:13
Sake it's a file19:13
sorry.. should've specified19:13
I thought I needed the file to be able to have the folder in my repo19:13
Thumper_ Sake: then git add the file, and add a .gitignore in that directory with * in it - I think that'll do what you want19:13
→ Stravy joined19:13
Thumper_ Sake: you can add an empty .gitignore to a directory if you want the directory to exist19:13
tstclair_afk → tstclair19:14
Thumper_ Sake: or in the top-level .gitignore you can add tmp/logs after you've git added tmp/logs/.gitignore (which is empty)19:15
Sake ah ok, but an empty .gitignore file will cause any files created in that directory to be ignored?19:16
that's much better than 'empty' files19:16
Thumper_ no that will just ensure the directory is created19:16
git won't create empty directories - there has to be a file (you're using the empty .gitignore) in the directory for git to create the dir on checkout19:17
Sake oh ok, I thought it actually caused the contents to later be ignored19:17
so essentially, an empty .gitignore file is the same thing as an empty file named 'empty', no?19:17
→ voker57_ joined19:17
Thumper_ any globs you have in .gitignore will be ignored - if it's empty it doesn't ignore anything19:17
sure19:17
except .gitignore won't show up in normal directory listings due to the leading .19:18
and it's special for git19:18
Trey can i get git to store a symlink? i've tried, and it lets me commit it, but when i do a new checkout all i get is a plain text file whose contents is the symlink's target.19:18
Thumper_ yes19:18
← mrtazz left19:19
← _rane left19:19
← Solefald left19:19
← dan_ left19:19
← majoh left19:19
← mutex left19:19
← daed left19:19
← tmz left19:19
← dionoea left19:19
← omega_ left19:19
← Nafai left19:19
← unreal left19:19
← uau left19:19
← ToxicFrog left19:19
← d1b left19:19
← RurouniJones_ left19:19
← faheem_ left19:19
→ dionoea_ joined19:19
→ omega_ joined19:19
→ daed joined19:19
→ mrtazz joined19:19
→ d1b joined19:19
→ axiom joined19:19
→ faheem joined19:19
→ _rane_ joined19:19
axiom → majoh19:19
← sde_antrix2 left19:19
→ unreal joined19:19
Thumper_ Trey: hmm I thought it worked normally with symlinks - saving the symlink only19:19
→ Nafai` joined19:19
Thumper_ Trey: but I haven't verified it recently19:19
→ ToxicFrog joined19:19
← HektoR_ left19:19
→ RurouniJones joined19:19
→ cubuanic joined19:20
→ Solefald joined19:20
→ uau joined19:20
→ sde_antrix joined19:20
→ hemmecke joined19:20
→ dan_ joined19:20
← Stravy left19:20
→ tmz joined19:20
→ mutex joined19:20
Nafai` → Nafai19:21
→ Stravy joined19:21
Trey Thumper_: so creating the symlink, doing a 'git add', and a commit, *should* work? that's all i did, but when i checkout i get a plain file. if i try to repair the situation with a git rm, followed by a new ln -s and git add, git status tells me there's nothing to commit. git seems to see a symlink and a file containing the target of the symlink as identical19:21
Thumper_ Trey: works for me19:22
tamm0r I have a question regarding git svn... I usually work on a branch, lets call it A, and merge it back to master. Then I dcommit master. Occasionally I want to continue working on that branch A and merge it again into master at a later time. Is that possible or is there any hassle to expect?19:22
← daifuku left19:22
Thumper_ to test it I created a branch after that, removed the symlink and made it a regular file19:22
then committed that and went back to master19:22
and the symlink reappears and points at teh right thing19:22
tamm0r I forgot to say that I want to dcommit again after every merge19:23
Trey Thumper_: i'm not sure what i've done wrong then. it's git version 1.6.4.419:23
Thumper_ Trey: 1.6.6.rc0.54.gb073b19:23
here19:23
← voker57 left19:24
Trey ok, maybe i need to try 1.6.6. the symlink points to an executable script, so obviously it all breaks when the symlink gets replaced by a plain text file19:24
Thumper_ Trey: just try it in a junk repo in /tmp and see if it acts as you expect19:24
Trey Thumper_: alright.19:24
← stuffcorpse left19:24
→ Wolfman2000 joined19:25
Thumper_ Trey: is it possible that you're checking out a point in history before the symlink was commited?19:25
← jreed left19:25
→ FunkeeMonk joined19:26
← chuck left19:27
Trey Thumper_: i'm not sure what you mean. i just noticed that my symlink had been converted to a text file (and in fact, no version in history was correct, it appears the svn->git conversion was when the symlink got mangled) and was trying to repair the damage by removing the plain files and replacing them with symlinks. if i try to do it as a single commit with git rm followed by ln -s followed by git add followed by git commit, i'm told there's nothing to ...19:28
... commit. if i do it in two commits, one with the removes and a second with the adds, it stores two commits, but when i do a new checkout that requires the symlink to be recreated in the working tree, it creates them as plain files19:28
tamm0r any ideas anyone?19:28
Thumper_ Trey: oh - I wasn't aware git-svn was involved and that you didn't originally commit it as a symlink19:29
Trey Thumper_: git-svn is no longer involved, but it was for the initial import19:29
we've switched entirely over to git now, the subversion stanza in the git-config is gone19:30
Thumper_ Trey: right but it might have been wrong at conversion time (ie the symlink was never created)19:30
← thom_ left19:30
Trey Thumper_: this is possible, but i need to figure out how to repair it19:30
→ Theravadan joined19:31
← sde_antrix left19:31
→ sde_antrix joined19:32
→ __iron joined19:32
→ jjuran joined19:33
→ rudi_s joined19:34
Trey Thumper_: in a new repo in /tmp, it works as expected: http://gist.github.com/24295919:34
← shuffle2 left19:35
Trey the "typechange" line is the bit i would expect to see in my real repo, and never do19:36
→ ideogon joined19:36
← WebDragon|away left19:36
→ exygen joined19:36
→ WebDragon|away joined19:36
← dominikh1 left19:37
→ bigmack83_ joined19:38
← snearch_ left19:39
Trey Thumper_: I have a paste that shows what happens in the other repo, but gist.github.com is not loading for me all of a sudden19:39
→ Administrator__ joined19:39
→ bigmack83__ joined19:40
ideogon I just attempted an octopus merge of two ancestor branches into Master (0.2 and Admin_panel). There was a merge conflict in the process with one file. I fixed the conflict in that file and git-added it, then committed the merge. However, it appears that git only merged Admin_panel into Master, and didn't actually merge any of 0.2 into Master yet. Is that what happens if an octopus merge fails--all further branch merging is ceased until a conflict is resol19:40
Trey Thumper_: I have pasted it in another pastebin I can get to: http://paste.lopsa.org/6319:41
→ Keiya_ joined19:42
→ shuffle2 joined19:42
→ r11t joined19:44
← DavidKlein left19:44
← Stravy left19:44
← Keiya_ left19:44
→ Stravy joined19:45
← dizy left19:45
← Stravy left19:45
→ Stravy joined19:46
→ DavidKlein joined19:47
→ minaguib joined19:47
→ Keiya_ joined19:47
minaguib Hi. I discovered "git log --graph", but I'm having a hard time seeing where the branch heads are like in gitx/gitx, any ideas ?19:47
→ psynaptic joined19:48
wereHamster minaguib: git log --graph --decorate19:49
→ ded joined19:49
ded I want to delete a remote branch on github. I'm using 'git push origin :master' but am getting /usr/local/bin/ssh-git-wrapper: line 3: 13202 Segmentation fault /bin/sh -c "$SSH_ORIGINAL_COMMAND"19:50
minaguib wereHamster: Perfect. Thank you19:50
← Keiya left19:50
ded fatal: The remote end hung up unexpectedly19:50
← radarek left19:50
Ilari ded: What is 'ssh-git-wrapper'?19:51
wereHamster ded: what is the ssh-git-wrapper for?19:51
→ Paraselene__ joined19:51
ded Ilari, don't know it must be on github19:51
Ilari, it isn't present on my machine (ubuntu)19:52
thiago_home ded: it means it's a problem on the remote side19:52
Trey Thumper_: d'oh! it was fine in the repo, but core.symlinks was set to false for some reason. maybe the initial git-svn thing monkeyed with that19:52
thiago_home ded: unless you're a github admin, there's nothing you can do19:52
wereHamster ded: sure you can do something: send an email to github support ;)19:52
Ilari ded: 'echo $GIT_SSH' outputs nothing, right?19:52
ded Ilari, right.19:53
Paraselene__ Question: say you have a project versioned by git, you make a single branch at a very early stage and then you basically remake your entire project within that branch... should you merge this new branch into the original branch, or checkout the old branch and hard --reset to the tip of the new branch?19:53
Ilari ded: And .ssh/config doesn't contain anything funky like ProxyCommand?19:53
ded Ilari, don't think so. It was working fine until recently.19:54
→ Twisted_Mentat joined19:54
Ilari ded: Yeah. Looks like the segfault occurs on remote side.19:54
ded What is the more long-winded syntax for deletion of a remote branch? I.e. alternative to :remote_branch?19:54
thiago_home ded: that's the shortest and longest way :-)19:55
ded: unless you have shell access19:55
← godsflaw left19:55
← dreiss left19:55
wereHamster I think recent versions have also --delete (or it's a planed feature for next releases)19:56
ded thiago_home, are you sure? I thought there was something like 'git push origin X:remote_branch' where X is 'refs' or 'heads' or something. As you can see I'm not really clear on what the syntax means.19:56
thiago_home ded: no, sorry. X must be empty there19:56
it means push "nothing" into that ref19:56
Ilari ded: And looking from error, the mere push attempt is making it segfault.19:57
wereHamster ded: the longer version would be :refs/heads/remote_branch19:57
→ jrmuizel joined19:57
← omerj left19:57
Ilari ded: That is, the remote side has not seen what you are trying to create/modify/delete (it only knows the repository).19:58
ded OK, thanks all. I'll just work in a different branch and wait for github to sort it out.19:58
← bigmack83 left19:59
minaguib Paraselene__: Is the old branch still useful ?20:00
← mikepence left20:00
minaguib Paraselene__: You could just rename the branches and avoid merging altogether if you're abandoning the old one20:01
Paraselene__ minaguib: Nope - like I say, it's been completely recreated20:01
minaguib Paraselene__: Say A -> B, then rename A to OLD_A and B to A20:01
tamm0r wereHamster: once again I have a question regarding git svn... I usually work on a branch, lets call it A, and merge it back to master. Then I dcommit master. Occasionally I want to continue working on that branch A and merge it again into master at a later time (. Is that possible or is there any hassle to expect?20:02
→ JasonWoof joined20:02
→ mikepence joined20:02
← ded left20:03
← kristian42 left20:04
← Administrator__ left20:04
→ kristian42 joined20:05
→ icwiener_ joined20:05
→ EqS joined20:05
→ omerj joined20:05
→ JackWinter2 joined20:06
→ yhager_ joined20:06
wereHamster tamm0r: from what I have heard, it should be possible. But I've never used git-svn myself20:06
→ pcgod_ joined20:06
→ dionoea joined20:07
→ Blub\0_ joined20:07
→ Fatal_ joined20:07
→ docgnome_ joined20:07
→ ybit_ joined20:07
→ Borbus_ joined20:07
→ Spark__ joined20:07
→ MadCoder_ joined20:07
→ pom_ joined20:07
→ dngor_ joined20:07
→ d1b_ joined20:07
→ omega__ joined20:07
→ pielgrzy1 joined20:07
→ r0bby_ joined20:07
→ Rh0nda joined20:07
← docgnome left20:07
← JackWinter left20:07
← icwiener left20:07
→ peper joined20:07
← Spark left20:07
← EgS left20:07
← psynaptic left20:07
← ybit left20:07
← Borbus left20:07
← pielgrzym left20:07
← dionoea_ left20:07
← Blub\0 left20:07
← MadCoder left20:07
← omega_ left20:07
← d1b left20:07
← pcgod left20:07
← ppr left20:07
← r0bby left20:07
← pom left20:07
← Rhonda left20:07
← Fatal left20:07
pcgod_ → pcgod20:07
→ ded joined20:08
← yhager left20:08
brainbox1100 Hi. Does git have anything like hg's queue? For patch management?20:08
← asksol left20:09
brainbox1100 Just looking for a feature name - I'll go RTFM after.20:09
→ Paperface joined20:09
Paperface does anybody know if it's possible to include a file within your global .gitconfig?20:10
ded I'm trying to add a directory with "git add cmdline/". But git status afterwards does not indicate that the files in the dir have been staged. How do I investigate this. git add -v also has no output.20:10
← jtrindade left20:11
← bhierof left20:12
← hemmecke left20:13
→ patrikf joined20:13
← yhager_ left20:13
→ yhager_ joined20:13
← maattd left20:14
wereHamster ded: maybe the files are already tracked, or they are ignored with a .gitignore rule20:16
← Spark__ left20:16
→ Spark joined20:16
→ maattd joined20:17
→ bigmack83 joined20:17
← bigmack83__ left20:17
wereHamster brainbox1100: people usually use man git-rebase to manage patch queues20:17
Gitbot brainbox1100: the git-rebase manpage can be found at http://git.or.cz/man/git-rebase20:17
← bigmack83_ left20:17
→ Geralt joined20:18
brainbox1100 wereHamster: I'll check it out. Thanks.20:18
← AkumaStreak left20:19
← exygen left20:19
→ lamont joined20:20
ded wereHamster, not .gitignore. What does "already tracked" mean? Neither "git status" not "git ls-files" shows these files.20:21
→ bigmack83_ joined20:21
wereHamster ded: what about git rev-parse :path/to/directory?20:22
ded wereHamster: ~/src/shellfish[z]> git rev-parse :./cmdline20:23
:./cmdline20:23
fatal: ambiguous argument ':./cmdline': unknown revision or path not in the working tree.20:23
Use '--' to separate paths from revisions20:23
20:23
← sergiodj left20:23
← lucs left20:23
← razvand left20:24
→ uroboros joined20:25
→ cookiemon5ter joined20:25
← muthu left20:25
→ MrCheetoDust joined20:26
wereHamster ded: git rev-parse :cmdline20:26
← univate left20:26
Thumper_ Trey: ah! glad you figured it out :)20:26
ded wereHamster: That gave same output. As did with and without trailing /.20:27
← robinsmidsrod left20:27
wereHamster ded: sorry, no idea then20:27
ded: what about git add -f ?20:27
← bigmack83 left20:28
uroboros → mjf20:28
→ bigmack83 joined20:28
← bigmack83_ left20:28
ded wereHamster: I did rm -r and then re-obtained the dir from elsewhere and it worked. Can't try -f now.20:28
← Geralt left20:28
→ malumalu joined20:29
← Sho_ left20:29
← jspiros left20:30
→ jspiros joined20:30
← MrCheetoDust left20:30
← bigmack83 left20:32
→ bigmack83 joined20:32
→ stuffcorpse joined20:32
← icwiener_ left20:33
→ giallu joined20:33
→ muthu joined20:35
→ univate joined20:37
→ mgmart joined20:39
→ tedoc2000 joined20:40
← Fullmoon left20:41
→ lucs joined20:42
← lhz left20:44
docgnome_ → docgnome20:45
← maattd left20:45
→ sergiodj joined20:45
→ exygen joined20:45
→ Ibw joined20:47
Ibw Is git seen as an alternative to svn, or do they fill different roles? (i.e. different types of projects)20:48
→ captain_morgan joined20:49
dereine Ibw: you can do the same kind of development with git like with svn20:49
but git support even other types20:49
→ _grepper joined20:49
dereine like distributed development20:49
with many different branches, etc.20:50
Ibw ya, I was doing some reading. git looks pretty cool20:50
← blakemizerany left20:51
← cilly left20:51
← Paperface left20:51
→ viiitor joined20:52
mapreduce Should you [mistakenly] decide that moving to git was wrong, is it doable to revert to svn? Of course it'd be lossy in terms of history, but is there support for it?20:53
MacGyverNL *wavy hand thing* moving to git is never wrong.20:53
viiitor Hello! Does someone has experience with svn2git? I am trying to import an existing svn repository but, when I run svn2git, it only creates a .git directory. It don't downloads the code or anything, maybe because the repository is private (and it never asks my password)...20:53
MacGyverNL But doesn't gitsvn support that?20:53
← Twisted_Mentat left20:54
MacGyverNL But I think some developers do that all the time? Check out an svn repo as git, do development on git, then check it back in into the svn repo?20:54
→ b_jonas joined20:54
mapreduce viiitor: git svn clone -s <repository>20:54
-s to treat trunk/ tags/ and branches/ as they normally are in svn repos.20:55
viiitor mapreduce: Isn't svn2git better?20:55
mapreduce In what way?20:55
b_jonas is there something like git ls-files -t that shows the files that are modified or deleted etc in my working directory as compared to the index?20:55
mapreduce I tried it recently and couldn't get it to work properly.20:55
b_jonas I'd like something like git status but formatted differently20:56
viiitor mapreduce: At least in its website, it says so...20:56
mapreduce viiitor: Does it say how it is better?20:56
viiitor mapreduce: Yes20:56
http://github.com/nirvdrum/svn2git20:56
Take a look20:56
← stuffcorpse left20:56
→ sylr joined20:56
sylr hello20:56
viiitor hum20:57
kanzure hi all20:57
viiitor mapreduce: Even with your command, it don't works. I think.20:57
mapreduce I think most if not all of that is covered by git svn clone -s20:57
kanzure does anyone know of a git-based/flatfile issue/bug tracker that i can drop into one of my projects?20:57
viiitor mapreduce: It only creates the .git directory20:57
← ideogon left20:57
sylr is there a way to pull/push a file from/to another branch ?20:57
→ Twisted_Mentat joined20:57
mapreduce viiitor: Is the svn repo public (e.g., an open source project)?20:58
viiitor mapreduce: Again, it don't asks for my password.20:58
wereHamster sylr: no, push/fetch operate on changesets, not single files20:58
viiitor mapreduce: No. That's the problem, I think.20:58
wereHamster sylr: what you can do is checkout a single file from another branch and copy it into your current working tree20:58
viiitor mapreduce: It needs a password. Also, it uses SSL.20:58
mapreduce viiitor: Can you use svn to check it out on the command line?20:59
→ ereslibre joined20:59
mapreduce (i.e., not using tortoise or some IDE)20:59
viiitor mapreduce: Yes20:59
Just tried that, just to be sure...20:59
sylr wereHamster: thanks, i'm going to search in that way20:59
mapreduce viiitor: If you also have local file access somehow to the repo you could do it via that.21:00
kanzure hm "bugseverywhere" looks about right. http://bugseverywhere.org/21:00
mapreduce I did that by sshing into our repo and cloning it to git on that machine.21:00
← mm_work left21:00
viiitor hum21:01
kanzure wereHamster: if you copy it will the revision history stay?21:01
wereHamster kanzure: of course not21:01
viiitor mapreduce: Using localhost/my-repo instead of mysite.com/myrepo?21:01
mapreduce: Or using /var/myrepo?21:01
kanzure wereHamster: so i guess you better never put something in a branch that you expect to go to another branch then?21:01
mapreduce file:///var/myrepo I think.21:01
viiitor That way it should work?21:02
I'll have some headaches to get access, but could be done...21:02
wereHamster kanzure: never put commit something as part of a bigger commit that you expect to go into another branch21:02
.. would be more accurate21:02
→ stuffcorpse joined21:02
kanzure i see21:02
mapreduce viiitor: I cannot say as I am a newbie, but it could be worth a go. Obviously make sure you cannot accidentally alter the svn repo.21:03
viiitor hmmm21:03
weird21:04
mapreduce: If I do git svn clone https://my-repo, it works...21:04
mapreduce Instead of what?21:04
← krawek left21:04
→ krawek__ joined21:04
mapreduce [don't forget the -s if you have trunk/ tags/ and branches/]21:04
viiitor mapreduce: clone -s21:04
mapreduce Ahh.21:04
viiitor If I try clone -s, it don't works21:05
mapreduce Are you cloning the repo or a project?21:05
viiitor hmmm21:05
← tbf left21:06
viiitor A project (which has trunk/branches/tags) inside a bigger repo...21:06
(The repo has many projects)21:06
hmmm... this line maybe could give a clue of what's happening...21:07
"Using higher level of URL: https://svn.lavid.ufpb.br/svnroot/gingacdn/architecture => https://svn.lavid.ufpb.br/svnroot/gingacdn"21:07
The URL of the project I want to clone is the first (/gingacdn/architecture)21:07
But maybe it's trying to get from /gingacdn?21:07
← suman left21:08
viiitor (Which don't have branches/trunk/tags)21:08
mapreduce No idea. I'd google the generic parts of that message, and if that doesn't help, panic blindly.21:08
← mdmkolbe left21:08
→ dahl joined21:09
viiitor mapreduce: Thanks, I'll try :P21:10
→ krawek joined21:10
← rfay left21:10
← kristian42 left21:11
← d0k left21:11
→ kristian42 joined21:12
← keystr0k left21:12
← gbacon left21:12
← sde_antrix left21:12
← loincloth left21:13
→ gbacon joined21:13
← hipe left21:13
← mikepence left21:15
← til left21:15
viiitor mapreduce: I think I found the problem/solution21:15
mapreduce: http://stackoverflow.com/questions/1453416/git-svn-clone-checkouts-wrong-repo21:15
→ mikepence joined21:15
← malumalu left21:16
← corni left21:16
viiitor mapreduce: Can't use it right now (old version of git), but I think it'll work21:17
→ blakemizerany joined21:17
→ Fullmoon joined21:19
→ rfay joined21:20
← rettub left21:20
→ rettub joined21:20
← viiitor left21:22
← krawek left21:22
→ krawek_ joined21:22
← bcardarella left21:23
← sis left21:23
← krawek__ left21:24
← Sigma[Mtp] left21:26
← krawek_ left21:26
→ Keiya joined21:27
→ krawek joined21:27
→ Number5 joined21:27
krawek → Guest4953221:27
Number5 hello guys, I'm a bit new to git. I made my own branch from the master. But the master is updated from a remote repository and my branch not. How can I get the latest update from my master?21:30
I hope you understand what I mean.21:30
← DrNick left21:31
wereHamster Number5: git pull origin master21:31
← Pupeno left21:32
wereHamster or git merge master if you only want to merge changes in your local master branch21:32
Number5 wereHamster, what does origin exactly mean in this context?21:32
← Stravy left21:32
wereHamster the name of the remote repository (see also 'git remote show origin')21:33
Number5 wereHamster, yes ok, I did that and saw the output of the git://repository21:33
but21:33
I rather want it from the local, so you're second suggestion is a more proper one. Cause my changes should be with repsect to my local branch.21:34
I meant to my local master branch.21:35
→ Pupeno joined21:35
Number5 Last question, how can I see differences between the HEAD of my local master branch and my own new branch?21:36
← Pupeno left21:36
Number5 git diff HEAD master/HEAD?21:36
← b_jonas left21:36
wereHamster git diff HEAD master21:37
→ krawek joined21:37
wereHamster HEAD is your currently checked out branch. You have only one HEAD in a git repository21:37
so master/HEAD doesn't make sense21:37
← yhager_ left21:38
← lamont left21:38
Number5 wereHamster, what I mean is the difference between HEAD of the current branch (which is not the master) and HEAD of the master21:38
wereHamster there is no 'HEAD of the master'. HEAD is your currently checked out branch21:39
→ loincloth joined21:39
Number5 hmmm, actually I could do simply log or diff between HEAD and previous HEADs21:39
wereHamster that would be 'git diff HEAD HEAD@{1}'21:39
but if you want the diff between your currently checked out branch and the master branch, git diff HEAD master is it21:40
Number5 aaah, wereHamster, I reall y want to thank you for your patient because of my "stupid" questions. But my teacher once said, you are really stupid if you don't ask at all.21:40
← rfay left21:40
→ cxvii joined21:41
EqS → EgS21:41
→ MrMagic joined21:42
← deavid left21:43
→ deavid joined21:43
← Keiya_ left21:44
→ hgimenez joined21:45
← dereine left21:45
→ m104 joined21:46
→ lamont joined21:46
← Grahack left21:47
→ tango__ joined21:47
← hyperair left21:47
← fcuk112 left21:47
ankit9 → ankit9|zzz21:47
← Keiya left21:47
→ javatexan joined21:47
→ Keiya joined21:47
← mgmart left21:49
← captain_morgan left21:49
→ karmi joined21:50
→ radarek joined21:50
→ yhager joined21:50
← yhager left21:51
← Guest49532 left21:51
→ s0ber joined21:51
→ rfay joined21:52
→ kpreid_ joined21:54
← drizzd left21:54
→ bcardarella joined21:54
← Hronom left21:55
← krawek left21:56
→ smotchkkiss joined21:58
← wolog_ left21:58
← flazz left21:58
→ krawek joined21:59
→ wolog_ joined21:59
krawek → Guest1288921:59
→ capndiesel joined21:59
← cxvii left21:59
← bcardarella left22:01
cehteh anyone here happen to be a asciidoc guru?22:01
← JasonWoof left22:03
→ krawek joined22:03
← s0ber_ left22:04
→ bcardarella joined22:05
→ RobertLaptop_ joined22:05
← crazychenz left22:06
RobertLaptop_ → RobertM22:06
RobertM → Guest2550922:07
← RobertLaptop left22:07
← lucsky left22:07
Guest25509 → RobertLaptop22:07
← RobertLaptop left22:07
← smotchkkiss left22:07
← Guest12889 left22:07
→ RobertLaptop joined22:08
← Darkas_ left22:09
→ gaveen joined22:11
← krawek left22:11
→ krawek joined22:11
krawek → Guest7938822:12
loincloth is there something like ls-remote that i can use to check for files in a local branch?22:12
without switching to that branch?22:12
← muthu left22:12
cbreak maybe git show?22:12
wereHamster loincloth: check for files == see if a file exists?22:13
loincloth wereHamster: yeah i think so22:13
→ onigiri joined22:14
wereHamster loincloth: git rev-parse branch:path/to/file 1>/dev/null 2>&1; $? == 0 if file exists22:14
$? == 128 if not22:14
Gitbot wereHamster: Type '<plugin name>?' to get brief help and a command list, where <plugin name> is one of: gitfaq, gitman, gitweb, irc, plugin, test, user. Also try '<plugin name> <command name>?'.22:14
← joeytwiddle left22:14
loincloth wereHamster: interesting, lemme try22:15
hmm22:15
can i use globs with that approach?22:15
wereHamster no22:15
loincloth yeah22:16
hmm22:16
wereHamster or you can git ls-tree or git ls-files22:16
← karmi left22:16
loincloth ok lemme check those out22:16
wereHamster forget about git ls-files, use git ls-tree [-r]22:16
loincloth hmm ls-tree doesn't seem to support globs22:17
wereHamster what exactly is it that you are trying to do?22:17
loincloth what i want to do is check for README.* in, say, master22:17
and whatever README.* is there, i want to checkout in the current branch22:17
wereHamster whether any file which starts with README. existst in master?22:17
loincloth yeah.. if there is any file match README.* i wants it22:18
wereHamster if there are two?22:18
loincloth i'm gonna assume there is one, if any, for now ;)22:18
→ CarlFK joined22:19
wereHamster git ls-tree master | grep README | head -n1 | while read mode type hash path; do git checkout master -- $path; done # untested22:19
← sylr left22:19
← threeve left22:20
→ Keiya_ joined22:20
loincloth wereHamster: haha, thx for your help22:21
→ joeytwiddle joined22:21
→ krawek joined22:21
loincloth wereHamster: works! :)22:21
← Hunner left22:22
→ ab3 joined22:22
loincloth wereHamster: big thanks, sir or ma'am22:22
→ Pupeno joined22:23
→ bigmack83_ joined22:24
← peteinlux left22:24
→ Shyke joined22:24
Shyke Willing to pay $50USD via PayPal for anyone that can resolve what should be a fairly simple issue with GiT. I basically messed up my repository and I can't get it back to how it was. Private chat for more info.22:25
→ Hunner joined22:25
→ Keiya__ joined22:25
cbreak it's probably cheaper if you just ask in here, most people help for free :)22:26
← krawek left22:26
→ krawek joined22:27
→ Determinist joined22:27
→ erikvvold joined22:27
Shyke I have a main local repository. "master". I tried to pull another patch into it and someone it got switched to the pulled patch "t" and all my changes were displayed as a conflicted difference. I still have "master" but if I switch to it, pulling from remote causes MASSIVE conflicted issues.22:28
How do I reasolve this?22:28
I'll get the list of commands that I used. One moment.22:28
→ DrNick joined22:28
Shyke (basically, I screwed around when I shouldn't have)22:28
cbreak you can do git reset --hard @{1} to go back to what you had previously22:29
but you'll lose all local uncommited changes22:29
Shyke I've got more money then time.22:29
Two and a half hours, to be exact.22:29
loincloth i once saw shell history from somebody who royally fucked up a repository.. it was scary.. dude was using git commands i have never heard of and applying --force to just about everything..22:29
Shyke I don't have any uncommited changes.22:29
cbreak loincloth: :/22:30
Shyke I have multiple copies of the folder. I'll try what you said, cbreak.22:30
← ntoll left22:30
← hugoxrosa left22:30
Shyke Due to my limited time, I don't think I've panicked this hard in a while.22:31
Everything appears to be in order after that command. I am still in "t" though.22:32
cbreak you can git checkout master to go to master22:32
← Keiya__ left22:32
← patrikf left22:32
cbreak if you mean a branch with "t"22:32
← DavidKlein left22:32
← LiamH left22:33
→ Keiya__ joined22:33
Shyke I try to checkout master and a get a bunch of error then: "Staying on branch: 't'."22:33
← Keiya left22:33
cbreak hmm. weird. can you pastebin the errors?22:34
loincloth so, the errors.. those would be the important bits of information22:34
← ludde left22:34
← voker57_ left22:34
wereHamster Shyke: 'staying on branch 't'' is from git gui. Try to issue the commands on the commandline instead22:35
→ voker57_ joined22:35
Shyke Alright. Deleting my folder again and reloading from the backup (made right after I messed it up...)22:35
cbreak you should be able to continue with this one :)22:36
doener Shyke: is this on windows?22:37
Shyke Yes22:37
command lining git reset --hard @{1}22:37
tedoc2000 I have a disk that went bad that was serving up a git repository22:37
I have fsck-ed it22:37
and not git fsck is reporting corrupt objects22:38
I can git show <objectid> on another machine22:38
← mbroeker left22:38
tedoc2000 how do I fix that one file on my git server?22:38
← Yuuhi` left22:38
→ Yuuhi` joined22:38
cbreak tedoc2000: have you considered creating a new clone from an other repository?22:39
tedoc2000 I have :)22:39
Shyke http://pastebin.com/m710832ba <-- The response from command lining git reset22:39
wereHamster tedoc2000: gitfaw Howtofixabrokenrepository22:39
gitfaq Howtofixabrokenrepository22:39
tedoc2000 thanks22:39
doener Shyke: symptoms sounds like you have some crlf mess. You can confirm by doing "git diff" after that reset. If that shows some large diff, try again with "git diff -w" which should show an empty diff in that case (shows just filenames)22:39
wereHamster tedoc2000: http://git.or.cz/gitwiki/GitFaq#Howtofixabrokenrepository.3F22:39
Shyke git diff had no result. Just a bunch of blank lines.22:40
loincloth wereHamster: hey.. how come that shell magic you gave me adds the checked out README to the index?22:40
Shyke git diff -w had just one line22:40
wereHamster loincloth: because that's what checkout does..22:40
Shyke doener: How do I go about going back to Master now?22:40
← Guest79388 left22:40
← Keiya__ left22:40
loincloth wereHamster: SO SORRY22:40
→ Keiya joined22:40
loincloth thx anyway22:41
doener Shyke: as you don't have any uncommitted changes anyway, try: git checkout -f master22:41
Shyke doener: I'm terrified... lol22:41
wereHamster loincloth: you could replace checkout with show master:$path > $path, but that won't run any filter you might have configured22:41
Shyke doener: From doing that, I lost a bunch of patches that I've made.22:41
doener Shyke: I still suspect that to be some bad "file with crlf in repo but core.autocrlf set" stuff. That makes git think that files are modified, the -f makes it ignore that22:42
Shyke: "patches" == commits? You can't lose them that easily22:42
Shyke doener: Mind if I private message you with what I did to cause this mess?22:42
← Keiya_ left22:43
doener Shyke: pastebin preferred, url in private ok22:43
Shyke: reading stuff in an irc window is a pain22:43
at least such stuff22:43
cbreak is it very private?22:43
Shyke I'll just pastebin22:44
No, what I did is fail.22:44
http://pastebin.com/d3975c8d322:44
As I said, fail.22:44
cbreak what differentiates humans from animals is that they can learn from other people's mistakes :)22:44
← mikepence left22:44
← rfay left22:45
Shyke Something like that22:45
I can't remember what I originally tried. That's probably really close though.22:45
→ Keiya_ joined22:45
doener Shyke: hm, that doesn't look particularly interesting. The first checkout probably failed, the second created a branch called "anti" that references the same commit as what you had checked out (likely "master"), and the last line would have created an empty file antipatch.patch22:46
Shyke I was trying to make a difference patch.22:46
doener (the empty file thing assumes that you had "master" checked out initially)22:47
Shyke I had "master" initially.22:47
doener Shyke: at least, there's nothing destructive in that sequence in any way22:47
Shyke I was trying to get a patch file of t.git on what I had locally ("master") and instead this crap happened.22:47
← ctusar left22:48
← Keiya_ left22:48
→ wagle joined22:48
doener isn't actually sure what "this crap" is anymore22:48
→ eno__ joined22:48
cbreak some conflict thingie22:48
← bambule left22:48
→ cortana` joined22:49
doener Shyke: basically, I take it that you want to checkout "master" again, and go back to some old state of it, right?22:49
Shyke Umm.. My branch is now anti. It has the anti stuff in it from t.git and my local changes.22:49
From doing those commands.22:49
← Ibw left22:49
Shyke Or something very simmilar to it.22:49
← TeckniX left22:50
Shyke So22:50
cbreak you can show what branches you have with git branch22:50
and you can switch branches with git checkout22:50
at least you should :)22:50
Shyke Anti now = local repository + what I checked out originally + anti.22:50
→ Keiya_ joined22:50
← paltman left22:51
← eno left22:51
Shyke Basically. I need "master" (local) to be the current branch and it to be EXACTLY the same as it was before I messed all of this up.22:51
doener Shyke: http://git.pastebin.com/m14b1576c22:51
← cubuanic left22:52
Shyke Every time I try, I fail22:52
What will that do?22:52
doener Shyke: I a) checked out master again (forced), b) looked at the reflog to see the previous states of "master", c) resetted master back to the state I want it to be in (master@{1} in this case, the "Do something more" commit, before the "Mess up" commit)22:52
Shyke: of course, master@{1} might be some other entry, like master@{5} or whatever in your case22:53
Shyke Copying my backup back over to try to restore it again... lol22:54
Two hours before I'm screwed.22:54
cbreak I begin to think that git hates you22:54
I can't remember having similar problems :/22:54
Shyke Git does hate me. Git has always hated me. I could never "GiT" it at all. Never caught on to how to properly use it.22:55
I screw up massivly like this, about twice a year.22:55
Pieter that's gotta hurt22:55
doener cbreak: My gut feeling is that I'd consider the "clone to get a diff" thing (at least that's what the command sequence looked like) a, uhm, suboptimal use of git. And honestly windows with the whole crlf mess doesn't make things easier22:55
Shyke doing what you have done in the pastebin.22:56
doener cbreak: If I had a bit more time and a smart idea, autocrlf would be _the_ thing I'd work on to improve it22:56
cbreak is it unreasonable to expect windows users to use unix style lineendings?22:57
→ godsflaw joined22:57
→ JasonWoof joined22:57
→ dominikh1 joined22:57
← thiago_home left22:58
Shyke Looks like HEAD@{1} was the cause. So...22:58
What now?22:58
git reset --hard master@{1}22:58
?22:58
wereHamster if master@{1} is the commit just before you screwed up..22:58
then yes22:58
← tstclair left22:58
Shyke {2} then22:59
← psoo left22:59
← Keiya left22:59
Shyke fatal: Log .git/logs/refs/heads/master is empty.22:59
...23:00
doener Shyke: hu? How did you get the listing from "git reflog show master" then?23:00
Shyke: ... and determined that master@{1} was the bad one23:00
Shyke doener: Doing it all again from scratch. lol.23:00
← jlilly left23:01
cbreak maybe you don't have per-branch reflogs?23:01
← Keiya_ left23:01
→ Keiya joined23:02
doener cbreak: hm, that'd be weird though. They're enabled by default since 1.5.023:02
cbreak when I do git reflog here (locally) it spits out HEAD@{blah}23:04
← ereslibre left23:04
fpcfan → fpcfan-away23:04
wereHamster that's because 'git reflog show' displays the reflog of HEAD23:04
doener cbreak: hm?23:04
← madewokherd left23:04
wereHamster if you run 'git reflog show master' it would display the reflog of master (with entries like master@{0} etc)23:05
cbreak interesting. I have to try that with a repo where I actually have any other branch :)23:05
→ jlilly joined23:05
→ ereslibre joined23:05
doener cbreak: if you ever did a rebase, the reflogs for HEAD and master will differ23:05
→ shade joined23:06
← radarek left23:06
doener cbreak: HEAD's reflog gets an entry for every rebased commit, master's reflog gets only a single entry for the completed rebase23:06
← shade left23:06
Shyke Doener, cbreak: http://git.pastebin.com/d51915a2223:06
teuf doener: wow, really interesting, I knew I could learn things by idling here ;)23:06
Shyke I don't know why "$ git reflog show master" is in there twice. I may have pasted it twice.23:07
doener Shyke: uhm, wow, that looks totally broken to me23:07
Shyke doener: Looks to me that HEAD@{1}: is the cause.23:07
What do you think caused it?23:08
I cut off everything after that23:08
As they were all good.23:08
doener Shyke: given those three lines, you detached HEAD sometime before that23:08
Shyke Wonderful.23:08
How do I reattach it?23:08
Do I want it detached?23:09
← ixti left23:09
→ psynaptic joined23:09
doener Shyke: you did so at HEAD@{1} (the line must have ended with " to anticheat", which is missing from the pasted line)23:09
cbreak usually you can reattach heads by just checking out a branch head23:09
like git checkout master23:09
but in your case I have no idea what's going on anymore23:10
Shyke See. The GiT gods laugh at me.23:10
cbreak git status would show you if you're on a branch23:10
I think23:10
doener Shyke: HEAD@{1} was also an action that didn't really change anything. And the merge in HEAD@{2} was on the detached HEAD, so it didn't affect any branch23:10
wereHamster Git - Capital 'G', 'it' in small letters!23:10
→ UncleCJ_ joined23:10
Shyke Thank-you for correcting me, wereHamster.23:10
doener Shyke: could you paste a longer part of the output of just "git reflog"? Harder to use, but the reflog for master seems messed up23:11
doener wishes he could just ask for a shared screen session23:11
Shyke http://git.pastebin.com/d679f30cf23:11
cbreak :)23:11
→ bauruine joined23:12
← UncleCJ left23:12
Shyke I did 2-5 personally.23:12
cbreak windows probably doesn't do screen23:12
doener Shyke: that's the whole output?23:12
← loincloth left23:12
Shyke No, I'll get you the entire thing if you want it though23:12
dngor_ → dngor23:12
doener Shyke: look for "moving from master" or so, then paste from the start to that one23:13
Shyke there isn't one.23:13
just 1 and 223:13
have moving from master23:13
err23:13
0 and 123:14
doener Shyke: hm, ok, let's just try 20 lines or so, maybe that tells something23:14
Shyke After 5 it's just 10 lines or so of my commits.23:14
doener Shyke: and "git rev-parse refs/heads/master" please23:14
cbreak would try git cloning the buggy repository just out of desperation23:15
Shyke I can't seem to resize this window23:15
cbreak you can send stuff to a file with >23:15
doener Shyke: ok, then let's start with just the rev-parse23:15
cbreak git reflog > file.txt23:15
doener Shyke: if we're lucky, that's already good enough23:15
Shyke Cutting it off after 20 or so23:17
http://git.pastebin.com/d313a3f6e23:17
Prepare to blow your mind23:17
lol23:17
doener actually, I'm stupid, the result of the rev-parse is already in the reflog23:17
so nevermind that23:18
→ flazz joined23:18
Shyke Was your mind blown?23:18
lol23:18
doener not at all23:18
Shyke I wonder how I'd just go back to23:18
fd2bbce0e19be680c89dbe9eb451dc9bdf6b5b5623:18
Switch to master and go to fd2bbce0e19be680c89dbe9eb451dc9bdf6b5b5623:18
doener Shyke: ok, seems like you detached HEAD looong ago23:19
Shyke Seems so23:19
doener Shyke: if fd2bbce0e19 is the "good state" that you want master to be at, just reset --hard to that23:19
Shyke: but just to make sure, I'd do "gitk fd2bbce0e19 master"23:19
→ ajonat joined23:19
doener Shyke: or maybe "gitk fd2bbce0e19...master" (three dots)23:20
← jlilly left23:20
doener Shyke: to visualize the history and make sure that fd2bbce0e19 is really what you want23:20
→ Cap_J_L_Picard joined23:20
cbreak where can I read up what the difference between two and three dots is?23:21
wereHamster cbreak: the git-rev-parse man page :)23:21
d3xter_ guys, mind if i ask an offtopic question? ;)23:21
cbreak thanks :)23:21
Shyke 5151306ea49eb3bc1d4feee4d574edc30c096cb0 appears to be the last and best commit, right before the screw up.23:21
doener Shyke: hm, that's not even in the reflog, i.e. it's very old23:22
Shyke no, it's as of today.23:22
5 minutes before the screw up23:22
fd2bbce0e19...master that is ancient23:23
doener Shyke: "very old" as in "quite a few things happened since then"23:23
Shyke fd2bbce0e19...master = very, very, very, VERY old.23:23
fd2bbce0e19...master = very old as in 7 months ago.23:23
← Lisimba left23:24
doener wow. That would likely mean that you detached HEAD 7 months ago, without noticing or switching to a branch23:24
Shyke I told you I'm pretty fail with Git23:25
→ flazz_ joined23:25
cbreak can you even commit to a detached head? :/23:25
wereHamster sure23:25
Shyke I can commit to local.23:25
I've been commiting to local forever,.23:25
← Keiya left23:25
Shyke I rarely commit to remote.23:25
cbreak local doesn't mean detached head23:26
wereHamster commits are always local...23:26
→ scarabx joined23:26
Shyke I give my code to others and they commit.23:26
doener Shyke: anyway, if you say that 5151306 is better than fd2bbce0e, then reset to that commit. I can't tell, I don't know your stuff. Just noticed that that doesn't appear in the reflog23:26
→ Keiya joined23:26
cbreak detached head means not on any branch23:26
Shyke How to I revert to 5151306ea49eb3bc1d4feee4d574edc30c096cb0?23:26
doener Shyke: assuming that you have "master" checked out: git reset --hard 5151306ea49eb3bc1d4feee4d574edc30c096cb023:26
← evl left23:26
cbreak I wonder if you can attach your head by git checkout -b somename HEAD23:26
no, that doesn't make sense23:27
doener cbreak: just 'git checkout -b somename" is enouhg23:27
s/enouhg/enough/23:27
cbreak: it defaults to HEAD23:27
cbreak: of course, "somename" may not exist yet23:27
Shyke doener: You magnificent bastard! It seems to be all in tact.23:28
cbreak I'd hope no one calls their branches like that :)23:28
doener Shyke: good, so you have two hours left? ;-)23:28
Shyke Damn. I'm still missing some stuff.23:28
An hour and a half.23:28
But I'm missing some major patches23:28
doener Shyke: if noone else complains, donate the $50 to some charity23:28
teuf cbreak: of course not, everyone knows branches should be called foo and bar23:29
doener Shyke: again, "patches" == "commits"?23:29
Shyke: i.e. things you committed?23:29
Shyke Yes.23:29
cbreak if you know the hashes, you can cherry pick them23:29
Shyke Err. Shit. Umm. It reverted it to the remote "master" and not the local master.23:29
I'm reloading my backup, lol23:29
doener Shyke: hm? Your 'master' was out of date as you said23:30
← flazz left23:30
doener Shyke: and you adjusted that to 5151306ea49eb3bc1d4feee4d574edc30c096cb0, because you said that that is the "good" commit23:30
Shyke: maybe after all fd2bbce0e was the good one instead?23:30
Shyke: you can just do "git reset --hard fd2bbce0e" in this repo then, no need to restore the backup again23:31
← onigiri left23:31
Shyke Already restoring it =/23:31
doener Shyke: the backup would just be needed for major f.. ups, not minor stuff like this23:31
← Number5 left23:31
doener Shyke: oh well, then it's "git checkout -f master; git reset --hard fd2bbce0e"23:32
d3xter_ → d3xter23:32
Shyke doener: will do23:32
→ bentob0x joined23:33
→ jeffxl_ joined23:33
Shyke doener: I don't even know why I ever bother to argue with you, lol. Every time you're right.23:34
← ablemike left23:35
cbreak I heard rumors that he was wrong once23:35
→ gordonh joined23:35
cbreak nothing specific though.23:35
← gaveen left23:35
→ gaveen_ joined23:35
← ehabkost_ left23:36
Shyke Testing to see if the code compiles23:36
I'll be back in 15 minutes or so... (damn c++)23:36
cbreak when it compiles, it's ready to ship :)23:37
Shyke So you think this will have resolved all of my issues?23:37
← tbuehlmann left23:38
doener Shyke: honestly: I'm not sure. You (initially) said that fd2bbce is good. But you didn't actually change anything since that commit.23:39
← malesca left23:39
doener Shyke: and the reflog doesn't go back to 5151... (the commit you later declared as being the good one)23:40
Shyke ========== Build: 13 succeeded, 0 failed, 0 up-to-date, 0 skipped ==========23:40
Perfect.23:40
→ xenoterracide joined23:40
Shyke Time to see if the file functions23:40
← _grepper left23:41
cbreak Shyke: if I was you, I'd from now on make sure to check if HEAD is attached23:41
you can detach it by checking out specific sha1 or reflog entries I think23:42
→ pantsman joined23:42
doener cbreak: passing anything that's not a branch head without the refs/heads/ part detaches HEAD23:42
cbreak: i.e. you get a symbolic ref for HEAD _only_ for "git checkout $foo" when refs/heads/$foo exists23:43
Shyke What's so important with having the HEAD attached?23:43
cbreak I feel like making a bad joke.23:43
doener Shyke: git checkout master; git commit ==> master gets updated to reference the new commit23:43
Shyke: git checkout HEAD~3; git commit ==> HEAD got detached by the checkout, the commit only updates HEAD, not any branch head23:44
→ flazz joined23:45
← Keiya left23:45
Shyke doener: I don't follow and...23:45
→ malesca joined23:45
Shyke cbreak: I know what you were going to say, lol23:45
doener faq detached23:45
Gitbot doener: You are on a detached HEAD and might lose commits.. See http://git.or.cz/gitwiki/GitFaq#detached23:45
→ Keiya joined23:46
Shyke Usually death would follow, no? (cbreak :P)23:46
doener cbreak: and committing on a detached HEAD can be very useful when you know what you're doing23:46
cbreak I thought about mentioning the lack of blood flow in a detached head's brain23:46
but I guess death would be the high level result23:46
doener: so useless to me :)23:46
doener cbreak: git checkout HEAD^0; git merge topicA topicB topicC; *test*; git checkouter master # Throw away the test integration merge23:46
cbreak anything you couldn't do with a temporary branch?23:46
← rudi_s left23:47
← jmatthews left23:47
doener cbreak: of course not, a detached HEAD is like an implicit temporary branch. So you don't have to do the clean up23:47
cbreak that does sound promising :)23:48
Shyke Now morbid jokes are making me giggle. Damnit.23:48
→ Lisimba joined23:48
Shyke doener: Everything seems to work. Which charity would you like me to donate the $50 to (they have to accept paypal).23:48
?23:49
→ EugenMayer joined23:49
EugenMayer with git branch, do I branch the current WC?23:49
doener Shyke: let's take this one http://www.vim.org/sponsor/index.php23:49
← jeffxl left23:49
Shyke doener: Will do.23:50
EugenMayer I mean, lets say I have " master and *foo" .. while iam currently in the foo branch. No I added some fiel files in foo and writing "git branch bar". What is bar branched from23:50
doener I'll trust Bram on that23:50
wereHamster EugenMayer: current HEAD. But keep in mind that git branch merely creates a branch, it doesn't switch to it23:50
EugenMayer what is HEAD in the git terminology?23:50
→ madewokherd joined23:50
doener EugenMayer: what you have checked out23:50
EugenMayer head?23:50
wereHamster EugenMayer: currently checked out branch (foo in your case)23:50
Shyke $33EUR incoming to that charity.23:50
EugenMayer so the Working directory, using the svn terminology23:50
doener EugenMayer: if you have "foo" checked out, HEAD is a symbolic reference to refs/heads/foo23:51
EugenMayer ok, thank you guys23:51
doener EugenMayer: in svn terms, I guess it would be BASE or COMMITTED (I never understood those...)23:52
cbreak doener: Vim is a charity? :)23:52
Shyke doener: Donation of $33 EUR ($50 USD) sent to [email@hidden.address] What is VIM?23:52
EugenMayer doener: well never mind. Iam hell of glad having git up and running now :)23:52
cbreak Vim's a text editor :)23:52
← jeffxl_ left23:52
doener Shyke: the best editor in the world23:52
cbreak should you ever feel like trying it23:52
← jedbrown left23:53
cbreak remember the following23:53
":q!"23:53
took me half an hour to find out how to quit it again... :)23:53
doener cbreak: Bram asks for donations for the ICCF which helps children in Uganda23:53
EugenMayer "how to get out maann..i guess i have to shutdown"23:53
cbreak nice :)23:54
← ereslibre left23:54
doener Shyke: thanks :-)23:54
EugenMayer doener: when I branch of from foo to bar . Then foo gets updated, how would I merge the changed to bar?23:55
I guess that's fast forward merge or?23:56
cbreak you can use git merge or git rebase23:56
doener EugenMayer: git checkout bar; git merge foo23:56
→ lunks_ joined23:56
doener EugenMayer: note that uncommitted changes are not on _any_ branch. They're uncommitted and exist only in your working tree and/or index23:56
EugenMayer thank you guys. Sorry to ask you such questions. I was reading the docs and head a hell of a job trying to get submodule doing what I want (I removed it later, its simply crap)23:56
doener EugenMayer: so if you do "git checkout other_branch" git will try to keep those uncommitted changes, so you don't lose them23:57
EugenMayer ah, ok23:57
uncommitted changes == stash23:57
doener EugenMayer: same deal as with "svn switch svn://host/repo/project/branches/other_branch"23:57
EugenMayer stash is the git terminology, or?23:57
→ patmaddox joined23:57
doener EugenMayer: the "stash" is kind of a temporary storage area23:57
cbreak it's a functionality23:57
normal uncommitted changes are either local to your working directory or in your index23:58
→ mithro joined23:58
doener EugenMayer: if you do "git stash", then git saves your uncommitted changes and removes them from the working tree/index, so those are clean again23:58
EugenMayer so wait, when I changed a existing file, and I did not run "git add file". What is that file called then?23:58
→ yarddog joined23:58
cbreak it's in your working directory. git status will show uncommited changes23:58
if you have colors enabled, it'll be red23:58
yarddog → Guest2200623:59
doener EugenMayer: I don't know of any name for that file, but you then have uncommited (and unstaged) changes to "file"23:59
EugenMayer how to enable colors? ( aim using zsh )23:59
doener EugenMayer: assuming that "file" is already tracked23:59
EugenMayer: git config --global color.ui auto23:59
EugenMayer ok, so uncommitted == unstaged and would be "modiified" somehow in SVN terms23:59
cbreak I have this in my .gitconfig http://paste.the-color-black.net/288023:59

Logs Search ←Prev date Next date→ Channels Documentation