IRCloggy #git 2009-12-26

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.

2009-12-26

rhett joined00:00
ngwngw|away00:01
ngw|awayngw00:02
Paraselene_ left00:03
wurstlpraterpSyPoInt3r00:04
__iron joined00:05
maxter left00:06
Sky[x] left00:07
whaley left00:08
Paraselene_ joined00:09
__iron left00:11
pantsman left00:12
sshc left00:12
erikeldridge left00:12
wagle joined00:14
nessundorma_ left00:15
nessundorma_ joined00:17
d0k left00:17
shfx joined00:18
Adlai left00:21
Adlai joined00:22
__iron joined00:23
shfx left00:23
sanity joined00:24
sanity does git have an equivolent to svn's $LastChangedRevision$ mechanism?00:24
Turias joined00:24
Turias left00:24
sanity ie. a variable that can go in a source file which will be automatically replaced with the revision number?00:24
Arrowmaster no00:25
D[a]vey joined00:26
Arrowmaster although it can be done sorta using either filters or the replacements that get done when you use git archive00:26
Simoyd joined00:28
catalis joined00:28
catalis left00:28
catalis joined00:28
Simoyd when I use "git branch --track <blah>" (or the information shown in "git branch -vv"), where is the tracked branch name stored in the .git dir? (I want to change it)00:29
sanity Arrowmaster: ok, thx00:29
pkw_ left00:29
Simoyd how can I change what branch a specific branch of mine is tracking?00:31
nessundorma joined00:31
wereHamster Simoyd: set branch.<name>.merge to refs/heads/<whatever>00:32
AStorm left00:32
nessundorma_ left00:33
ness1d0rma left00:33
catalis left00:34
Sky[x] joined00:35
Simoyd sick00:35
thanks!00:35
wereHamster Simoyd: that information is stored in the config file (.git/config)00:35
catalis joined00:36
Simoyd oh... wow... I looked there but apparantly I'm blind. thankss!00:36
zackattack joined00:36
catalis left00:37
babilen_ joined00:38
babilen left00:38
babilen_babilen00:38
Yuuhi left00:39
Simoyd left00:45
__iron left00:45
erikvold joined00:47
dazman left00:48
project2501a left00:51
datechman joined00:51
JayM joined00:51
hipe joined00:52
__iron joined00:52
simcop2387 joined00:53
gambler left00:55
yhager_ joined00:55
yhager left00:57
__iron left00:58
kpreid_ joined00:58
Sky[x] left01:00
__iron joined01:03
erikvold left01:04
wagle_ joined01:06
wagle left01:06
rhett left01:07
hipe left01:09
defectiv joined01:11
__iron left01:12
pkw joined01:16
mithro joined01:18
brendon2020 left01:18
catalis joined01:20
catalis left01:21
imajes joined01:22
defectiv how can i refer to the parent of a commit?01:23
say i have the commit hash?01:23
wereHamster <hash>^101:23
defectiv hrm...tried that.01:23
wereHamster or ~101:23
defectiv i thought ~1 was a range.01:23
fb178ca10^1 would work?01:24
wereHamster it should01:24
^n is n-th parent of that commit (merge commits have more than one parent)01:24
~n is n times ^101:25
__iron joined01:25
defectiv wait what's the difference?01:25
wereHamster defectiv: man git-rev-parse explains it better than I ever could01:25
defectiv oh i see what you mean.01:25
Gitbot defectiv: the git-rev-parse manpage can be found at http://git.or.cz/man/git-rev-parse01:25
danopia left01:26
defectiv is there a good way to see when a file was introduced?01:27
ngwngw|away01:29
ngw|awayngw01:29
sgh left01:32
sgh joined01:32
Ilari defectiv: 'git log --diff-filter=C -- file'?01:33
godsflaw joined01:36
kpreid_ left01:39
erikeldridge joined01:40
erikeldridge left01:42
youwin joined01:43
babilen left01:43
defectiv left01:48
nessundorma left01:52
Determinist left01:55
wagle joined02:00
cloudhead left02:01
D[a]vey left02:01
schambers joined02:07
thiago_h1me joined02:08
wagle_ left02:08
sgh left02:10
aziz left02:11
sanity left02:12
s1kx- joined02:14
amirite joined02:16
tjafk left02:16
thiago_h1methiago02:17
S1kx left02:17
thiago left02:20
thiago_home left02:21
s1kx-S1kx02:21
ispeak joined02:21
shuffle2 joined02:22
j416 joined02:26
VVSiz_ joined02:26
khmarbaise left02:27
defectiv joined02:27
imajes left02:29
sanity joined02:31
kukks left02:33
VVSiz left02:33
shuffle2 left02:40
project2501a joined02:44
schambers left02:44
shuffle2 joined02:46
ngwngw|away02:48
ngw|awayngw02:48
j416 left02:49
loincloth left02:49
loincloth joined02:50
Paraselene_ left02:51
kpreid_ joined02:53
aconbere joined02:57
loincloth left03:01
jbrennan_jbrennan03:05
godsflaw left03:06
godsflaw joined03:06
FunkeeMonk joined03:08
project2501a left03:10
engla left03:10
ajonat joined03:12
ngwngw|away03:13
ngw|awayngw03:13
defectiv left03:17
gdrage joined03:18
FunkeeMonk left03:20
cytrinox_ joined03:20
parasti left03:22
ispeak left03:27
Pupeno left03:32
hipe joined03:32
gdrage left03:36
cytrinox left03:36
cytrinox_cytrinox03:36
loinclot_ joined03:38
ereslibre_laptop joined03:41
ron__ joined03:43
JayM left03:47
ajonat left03:51
dys` joined03:57
ispeak joined04:00
ispeak left04:00
loinclot_ left04:01
ron_ left04:02
ngw left04:03
neelam joined04:04
dys left04:09
neelam left04:10
godsflaw_ joined04:17
godsflaw_ left04:20
godsflaw left04:22
kpreid_ left04:25
hipe left04:27
erikeldridge joined04:30
bluebits joined04:30
erikeldridge left04:31
sanity left04:34
ereslibre left04:34
ereslibre joined04:34
bbits left04:35
JayM joined04:35
bbits joined04:35
LiamH left04:38
tommd joined04:42
tommd A simple git question: How do I combine two or more patches on my local repo into a single patch so I can format-patch and upstream just the one?04:43
wereHamster tommd: rebase -i and squash the two commits04:44
s/two/04:44
s/two//04:44
cbreak left04:46
cbreak joined04:46
mentor joined04:47
bluebits left04:47
bluebits joined04:47
mentor I've got merge conflicts. I've been manually merging them (in the working state, not committed), but I've now released that I made a mistake in one file; how do I got back to the conflicted state?04:48
rchavik joined04:49
wereHamster mentor: git checkout --conflict -- file04:51
mentor Need to specify the merge style, but that's it; thanks.04:52
Sorry, I lie, that has not wroked04:54
wereHamster or checkout --merge -- file04:55
bbits left04:55
wereHamster I think when using --conflict you have to specify the style (--coflict=<style>)04:56
mentor :/04:57
:p04:57
aha, I've fucked myself by doing something else earlier04:58
shuffle2 left05:04
tommd wereHamster: Thank you, but I really don't understand git as well as I should. 'squash' two commits together doesn't seem to be slang here - it has a particular meaning? 'git rebase -i' needs more and doesn't accept something so trivial as a list of patches and a branch to base them on.05:06
wereHamster I assumed you knew how git rebase works05:08
tommd :-) Yeah, I'm going through man pages now. Thanks for the kick-start.05:09
kpreid_ joined05:10
wereHamster basically, git rebase -i <first commit you want to edit>^, that will pop up an editor, there you can reorder/edit/squash commits, save and exit the editor and git does its work05:11
gitready has an article about rebase, as does the git book05:11
tommd I'll hunt them down (or learn the hard way...)05:12
Thanks05:12
ereslibre_laptop left05:12
Fissure joined05:15
hyperair joined05:16
Jak_o_Shadows left05:19
numberGrey joined05:23
project2501a joined05:25
catalis joined05:29
patrikf left05:43
tedoc2000 joined05:46
catalis left05:46
piroko joined05:46
catalis joined05:46
catalis left05:47
crab joined05:50
bluebits left06:01
gebi joined06:04
bluebits joined06:04
madewokherd left06:04
tedoc2000 left06:04
munichlinux joined06:05
loincloth joined06:11
schmukkk joined06:13
schmukkk left06:14
schmukkk joined06:15
chalkbag joined06:31
sh1mmer joined06:31
meanburrito920_ joined06:34
cannonball left06:35
__iron left06:40
rudi_s joined06:46
ron__ron_06:47
rudi_s_ left06:48
meanburrito920_ left06:48
JakeSays left06:52
JakeSays joined06:52
sh1mmer left06:56
loinclot_ joined06:57
justin__ joined07:03
justin__ anyone here?07:05
bob2 lots o people07:06
justin__ why all silent ....07:06
bob2 that's how irc works07:07
sometimes there's lots of chatter, sometimes there's none07:07
justin__ lol07:07
datechman left07:08
datechman joined07:11
loincloth left07:14
gebi left07:16
datechman left07:21
datechman joined07:22
vbgunz joined07:24
incredible joined07:25
keystr0k joined07:31
datechman left07:32
mtkd joined07:32
datechman joined07:33
datechman left07:34
datechman joined07:36
nitrate joined07:37
quizme joined07:39
vbgunz_ joined07:43
vbgunz left07:44
exygen left07:49
nitrate left07:52
sis joined07:52
keystr0k left07:55
nitrate joined07:57
brizly joined07:59
nitrate_ joined08:02
snearch_ joined08:04
r11t joined08:08
up_the_irons left08:09
snearch_ left08:10
mbernstein left08:13
Sonderblade joined08:15
nitrate left08:16
Twisted_Mentat left08:16
nitrate__ joined08:18
nitrate_ left08:18
babilen joined08:23
up_the_irons joined08:23
hipe joined08:25
hyperair left08:26
droden joined08:26
hyperair joined08:29
Pupeno joined08:30
hohum joined08:32
hohum how do I give someone access to push to my repo?08:32
voker57 use ssh and maybe some script like gitosis for rights management08:34
hohum the user I want to give repo access to is local to the machine08:34
voker57 create user group with two of you, give repo to group, set up shared repo08:35
means "sharedrepository = 1" in config08:35
tekky joined08:39
hohum okay08:41
well08:41
I just did a "git push <path to original repo>" from the other user's directory08:42
and it didn't work08:42
I don't see any of the files I added from that user08:42
voker57 pastebin the error message08:42
hohum there is no error message08:42
[acorbe@apollo ~/danos]$ git push /usr/home/dcorbe/danos08:42
Everything up-to-date08:42
QshelTier left08:42
hohum but I don't see any of my recent adds or commits08:42
voker57 check if you are on right branch etc08:43
hohum I didn't make any branches08:43
bob2 push will not update the working copy08:43
(as the faq says, with rationale08:44
hohum all I did was init a new repo with group permissions08:44
bob2 )08:44
perhaps that's what is tripping you up?08:44
hohum make some changes as a different user08:44
it should update the copy I tell it to update, no?08:44
pull updates the working copy08:44
*sigh*08:45
drizzd joined08:45
hohum why is this so hard08:45
I'm not trying to update a working copy08:47
I'm trying to update the original repo08:47
I do a git clone /usr/home/dcorbe/danos08:47
edit some crap08:47
did an add08:47
git commit08:48
git push /usr/home/dcorbe/danos08:48
no errors08:48
I expect to see my changes in /usr/home/dcorbe/danos08:48
and I do not08:48
why?08:48
hmm08:49
cedric_ joined08:49
hohum ...08:49
what am I missing?08:49
I missing a step obviously08:49
hello?08:50
vivien joined08:51
cedricv left08:51
hohum well08:53
bob2 and you mean that "cd /usr/home/dcorbe/danos ; git log " does not show the changes?08:53
hohum cvs it is then08:53
er08:53
yes08:53
nitrate_ joined08:53
hohum I see the changes when I do a git log08:53
bob2 so, the problem IS that the working copy isn't being updated?08:53
http://git.or.cz/gitwiki/GitFaq#Whywon.27tIseechangesintheremoterepoafter.22gitpush.22.3F08:54
hohum I think I see why I fail at life ;)08:54
nitrate__ left08:54
sis left08:55
vbgunz_ left08:57
xlistener left08:58
d0k joined08:59
aconbere left09:02
__iron joined09:02
tekky left09:06
Axius joined09:08
cxreg i think this channel gets questions about pushing into non-bare repos like 5 times every day09:08
Axius left09:09
VVSiz_VVSiz09:09
gitster perhaps time to update /title to refer to a faq (perhaps mothership-satellite section)?09:10
dys`dys09:11
cxreg what ever happened to dscho's patch that would update the checkout on push?09:12
was it sufficiently scary to not merge?09:12
gitster if you have to ask, it probably wasn't ready.09:13
cxreg i wasn't sure if it simply wasnt ready, or if you rejected it on principle09:13
xsist10 joined09:15
univate_ left09:15
Axius joined09:16
univate_ joined09:16
maxter joined09:21
solofight joined09:25
r11t left09:27
r11t joined09:27
gitster No useful patch is rejected "on principle". If a patch is useful for some but hurts others (either "users with other workflows" or "limits our choice in future enhancement"), I may call that a "hack", and it may need polishing to round the edges that hurt others until it gets in. In that case, it is "not ready" until that polishing happens.09:27
bousquet joined09:28
cxreg Right, makes sense. I found the original post, and I think he was being snarky09:28
[DRY HUMOR PATCH] Add a few more values for receive.denyCurrentBranch09:29
where the value is 'updateInstead'09:29
though later, people mentioned that they were using it and found it genuinely interesting09:29
hyperair left09:29
gitster I don't recall I ever finding any of Dscho's patch "scary". Often his patches are not that intrusive, and sometimes I find them not sufficiently intrusive when a bit more intrusiveness would give them tighter and better integration.09:30
hyperair joined09:30
hobodave joined09:30
Twisted_Mentat joined09:30
cxreg he seems a bit self deprecating09:30
gitster Or being too busy.09:31
Nobody stops others including you to pick up the remains and help polishing them, though.09:31
agib left09:32
ph^_ joined09:32
cxreg I don't personally care much about the feature, but it comes up very often among newbies09:32
mbroeker joined09:32
munichpython joined09:35
metze left09:37
r11t left09:39
r11t joined09:40
r11t left09:40
munichlinux left09:40
Axius left09:42
amirite left09:44
_gpg_ joined09:44
_gpg_ hello09:45
merry christmas09:45
fcuk112 joined09:45
jjuran It's still Christmas in Hawaii, isn't it.09:46
nitrate__ joined09:47
jjuran And a happy December 26 to everyone else.09:47
xsist10 left09:47
ankit9 joined09:50
nitrate_ left09:52
vivien left09:59
Chillance joined10:00
jessep left10:02
vivien joined10:02
vivien left10:03
munichpython left10:04
munichpython joined10:04
owen1 left10:04
vivien joined10:05
sgh joined10:07
Ilari Whatever happened to that patch that fixed the couple of nits in last round of smart remote helpers? Just got lost somewhere?10:08
vivien left10:09
lucsky joined10:11
bousquet left10:11
metze joined10:13
khmarbaise joined10:16
ToxicFrog left10:17
Axius joined10:17
justin__ left10:24
nitrate__ left10:25
nitrate__ joined10:26
Sky[x] joined10:27
gebi joined10:28
Axius left10:29
nitrate_ joined10:30
quizme left10:35
vivien joined10:36
nitrate_ left10:37
vivien left10:37
engla joined10:38
vivien joined10:38
owen1 joined10:38
solofight left10:39
nitrate joined10:40
ngw joined10:41
fcuk112_ joined10:42
fcuk112_ left10:45
nitrate__ left10:45
nitrate_ joined10:45
brizly1 joined10:46
nitrate left10:47
ankit9 left10:47
nitrate__ joined10:49
ixti joined10:50
fcuk112_ joined10:50
schmukkk left10:51
nitrate joined10:51
fcuk112 left10:51
nitrate__ left10:53
intripoon joined10:54
nitrate_ left10:57
munichpythonmunichlinux10:57
Eren left10:57
calin joined10:59
Axius joined10:59
brizly left10:59
hummingkid joined11:00
pantsman joined11:01
nitrate_ joined11:03
nitrate left11:04
ngwngw|away11:05
mtkd left11:06
mtkd joined11:07
Twisted_Mentat left11:09
ngw|awayngw11:20
j416 joined11:20
nitrate__ joined11:22
nitrate_ left11:24
Sonderblade left11:24
nitrate_ joined11:25
drizzd left11:26
nitrate joined11:29
imajes joined11:30
icwiener joined11:31
gebi left11:31
nitrate_ left11:32
sgh left11:34
nitrate left11:34
pantsman left11:36
nitrate joined11:36
nitrate__ left11:38
nitrate_ joined11:38
Axius left11:39
ivan joined11:40
pantsman joined11:40
nitrate left11:41
j416 left11:44
nitrate_ left11:46
munichlinux left11:47
aziz joined11:52
nitrate_ joined11:55
gebi joined11:55
imajes left11:57
soreau joined11:58
soreau Hello, I was wondering if there's any way to see when the last time one single line of code has changed in the past, like when was the last time it changed11:59
or was initially added.. is that possible with git?11:59
nitrate joined12:00
pantsman left12:00
teuf soreau: git annotate toto.c12:01
nitrate__ joined12:01
nitrate left12:03
felipec teuf: it's git blame12:03
TDJACR left12:03
TDJACR joined12:04
teuf felipec: both give mostly the same output12:04
felipec teuf: yeah, because git annotate is an alias to git blame12:05
soreau Thanks teuf felipec12:08
soreau left12:08
kiko_ joined12:09
selckin left12:10
engla left12:11
nanothief left12:11
kiko_ hi, In .gitignore, how can I tell ignore .config/ but not .config/foo/ ? tried .config/ and !.config/foo/ but seems it still ignored.12:12
nevyn hrm12:12
felipec kiko_: how about .config$ ?12:13
nitrate_ left12:13
kiko_ felipec: seems whole .config is not ignored with .config$12:17
felipec kiko_: all, right, only .config would be ignored, not .config/bar12:18
imajes joined12:18
felipec kiko_: then maybe .config/(^foo)12:18
wereHamster kiko_: why do you think .config/ and !.config/foo/ doesn't work?12:19
kiko_ wereHamster: I'm judging by git status not showing .config/foo/* as untracked file. Am I misunderstanding here already?12:21
hobodave left12:26
rchavik left12:27
thiago_home joined12:30
stamina_ joined12:32
engla joined12:39
cedric_ left12:41
partisan joined12:46
calin left12:52
partisan left12:53
Paraselene_ joined13:00
malumalu joined13:01
Octalot left13:02
kmap joined13:03
felipec left13:04
ngwngw|away13:07
Adlai I'm having trouble getting get-send-email to send a few patches. I found http://morefedora.blogspot.com/2009/02/configuring-git-send-email-to-use-gmail.html and followed that, but now I get: Can't locate Net/SMTP/SSL.pm in @INC (@INC contains: [snip]) at /usr/lib/git-core/git-send-email line 944, <FIN> line 3.13:07
munichlinux joined13:08
imajes left13:09
ngw|awayngw13:10
khmarbaise left13:12
wereHamster Adlai: how did you install git (or git send-email)?13:12
Adlai wereHamster, through my distro's package manager. I'm on Archlinux so that would be pacman.13:12
wereHamster hm, you should contact the packager that he has a missing dependency13:13
Adlai git-send-email was installed as part of the git package13:13
wereHamster but the packager didn't make sure that Net/SMTP/SSL.pm is installed before installing git13:14
see if you can install that perl package somehow13:15
Adlai I think I got it. I got prompted for a password, entered it, and then: Need MIME::Base64 and Authen::SASL todo auth at /usr/lib/git-core/git-send-email line 981, <STDIN> line 1.13:16
incredible left13:18
incredible joined13:19
khmarbaise joined13:20
keystr0k joined13:23
project2501a left13:24
malumalu left13:25
malumalu joined13:25
munichlinux left13:27
hummingkid left13:30
hyperair left13:30
hyperair joined13:30
zackattack left13:31
archis joined13:33
Pupeno left13:34
archis left13:34
kmap left13:36
Axius joined13:38
Adlai wereHamster, looks like I'm good. Had to dig around in the pacman packages get one more module (Authen::SASL) and now my patches are sent. :)13:39
hummingkid joined13:41
TDJACR left13:44
thiago_home left13:45
vbgunz_ joined13:47
khmarbaise left13:48
orafu left13:53
vivien left13:53
orafu joined13:53
vivien joined13:53
aresnick left13:57
Axius left13:58
khmarbaise joined13:59
mbernstein joined14:00
selckin joined14:01
klickverbot joined14:02
kain__ joined14:03
aresnick joined14:04
kain__ hi there. how to clone a repository 'stopping' at a specific commit/sha ?14:04
Axius joined14:04
ereslibre_laptop joined14:05
kain__ git checkout rev would suffice?14:06
catalis joined14:08
Axius left14:08
kain__ left14:09
klickverbot left14:10
Turias joined14:14
idletask joined14:17
idletask Hello14:17
shadeslayer joined14:17
idletask What exactly is the use of having a HEAD pointer on a remote repository? When you add a remote, there's no such pointer unless you git remote set-head14:17
What is this for exactly?14:18
shadeslayer hi suppose on gitorious.org i want to pull from the mainline branch and merge the latest source on my clone,whats the command?14:18
tibra joined14:18
idletask shadeslayer: do you already have a tracking branch for that mainline branch?14:18
tibra hi everyone14:18
shadeslayer idletask: tracking branch? like the url?14:19
idletask: i have 2 copies of the project,one on which i work on and the second which has the latest code14:19
idletask shadeslayer: no - a local branch that is set up to track a remote branch. When you have cloned the repository initially, the master branch has been set up as a tracking branch for origin/master14:20
shadeslayer: and you needn't have two copies14:20
tibra i've just installed a new git on a shared host. now trying to push a rails app over from local to that shared host, everytime that ssh asks me for "password", I can't get further, cause I am getting blocked by my server after this. is this a git issue or ssh?14:20
shadeslayer idletask: um for the merge requests i need just one copy?14:20
idletask Yes14:20
TDJACR joined14:20
idletask You simply add the necessary remotes14:21
shadeslayer idletask: ok i have my own clone copy on the HD,i havent set it up to follow any branch14:21
idletask shadeslayer: if you have cloned from gitorious, you _do_ have a tracking branch - master14:21
See above14:21
robinr joined14:21
shadeslayer idletask: yeah14:22
idletask shadeslayer: hold on, I'll pastebin an example14:23
shadeslayer idletask: ok sure :)14:23
roop joined14:24
shadeslayer im kinda new to revision systems... i set one up a few months ago.. forgot how i did it14:24
parasti joined14:26
ehabkost joined14:27
shadeslayer idletask: there?14:30
idletask shadeslayer: http://pastebin.com/m569d2f1a14:33
shadeslayer: in fact, the example I give is a rebase, not a merge, but it's easier for maintainers when you do a push request14:33
A pull request, sorry14:33
shadeslayer: it's a good idea to rebase regularly, too: if there are conflicts, they will be easier to solve14:35
It's the developers' work to resolve conflicts, not the maintainers' :p14:35
shadeslayer :D14:35
idletask shadeslayer: but as you can see, there is no need to have two copies, one is enough14:36
You just need to add the necessary remote repositories14:36
shadeslayer: also, you may want to have a different namespace, so adapt the commands (you can rename remotes, branches, etc)14:37
shadeslayer idletask: ok suppose i have my clone in gitorious,i cant create a branch on the mainline repo14:37
idletask You very probably can't, unless you're a maintainer on the main repo14:37
shadeslayer how do i update my clone on git with the lastest source from mainline then?14:37
idletask: no im not the maintainer of the main repo,i have a seprate clone14:38
idletask: like this is the main repo : http://gitorious.org/rekonq : and my clone is : http://gitorious.org/~shadeslayer/rekonq/shadeslayers-clone-docs14:38
idletask Add the mainline repo as a remote, fetch from it, create a tracking branch for the master branch and merge it regularly with your clone's master branch14:38
zackattack joined14:39
tibra_ joined14:39
shadeslayer idletask: lines 1-10?14:39
idletask shadeslayer: OK, I'll adapt the pastebin to show you, but this will take a little time :p14:39
ereslibre_laptop left14:39
idletask And I need a coffee14:40
shadeslayer idletask: um can you mail it to me? im running out of time :P14:40
tibra_ left14:40
idletask shadeslayer: I can, just /privmsg me your mail14:40
pantsman joined14:41
shadeslayer bye all14:41
thanks :D14:41
shadeslayer left14:41
tibra_ joined14:41
zackattack left14:42
kukks joined14:43
pantsman left14:47
robinr left14:48
LiamH joined14:48
tibra left14:48
tibra_tibra14:48
RaceCondition joined14:49
gambler joined14:51
robinr joined14:52
__iron left14:56
__iron joined15:03
robinr left15:04
yaotti_aw joined15:05
madewokherd joined15:06
tibra_ joined15:07
tibra left15:07
tibra_tibra15:07
whaley joined15:12
developernotes joined15:13
developernotes left15:14
roop left15:16
incredible left15:16
incredible joined15:16
tvw joined15:18
ereslibre_laptop joined15:19
ereslibre left15:21
catalis left15:22
reprore joined15:27
_gpg_ left15:28
pantsman joined15:28
Xeross left15:35
Xeross_ joined15:35
tibra left15:35
d0k left15:37
sanity joined15:37
sgh joined15:37
ngwngw|away15:39
vals_ joined15:42
catalis joined15:43
hyperair left15:44
pantsman left15:44
hyperair joined15:46
tango_ left15:47
Turias1 joined15:50
Turias left15:50
tibra joined15:52
whaley left15:53
mithro left15:54
aspotashev_ joined15:54
vals_tango_15:55
tibra left15:56
aziz left15:56
tibra joined15:57
stamina_ left15:57
thiago_home joined15:58
tibra left15:58
tedoc2000 joined15:59
partisan joined16:00
aziz joined16:01
Alagar joined16:02
ngw|awayngw16:04
catalis left16:05
godsflaw joined16:05
gebi left16:05
xenoterracide left16:06
aspotashev left16:07
xenoterracide joined16:08
Adlai` joined16:10
Adlai left16:10
Adlai`Adlai16:11
incredib1e joined16:11
ngwngw|away16:15
incredib1eincredible_test16:16
ngw|awayngw16:24
godsflaw left16:26
datechman left16:27
ngwngw|away16:27
AAA_awright_ joined16:28
nothingHappens joined16:30
cloudhead joined16:32
tvw left16:32
mtkd left16:33
datechman joined16:34
AAA_awright left16:34
developernotes joined16:34
developernotes left16:34
tbuehlmann joined16:35
idle_task joined16:40
Turias1 left16:40
engla left16:41
loinclot_loincloth16:42
sgh left16:42
Yuuhi joined16:43
voker57 left16:45
ereslibre_laptopereslibre16:45
catalis joined16:50
catalis left16:51
TML joined16:52
idletask left16:53
idle_taskidletask16:58
schacon joined16:59
kpreid_ left17:01
keystr0k left17:01
MrC joined17:05
jceb joined17:09
schacon left17:11
tibra joined17:20
nothingHappens left17:22
nothingHappens joined17:22
tedoc2000 left17:23
aspotashev_ left17:24
idletask If anyone is alive, I'd like an explanation... What is the origin/HEAD for when you list remote branches, and why, on added remotes, isn't there such a HEAD defined?17:25
thiago_home origin/HEAD is origin's HEAD17:28
it exists if the remote has it17:29
though I think pull/fetch doesn't add it17:29
idletask thiago_home: and what purpose does it serve? Is this for the clone operation?17:30
idletask needs to check his assumption that such a HEAD never appears on other defined remotes even after fetching from them17:30
idletask Well, I can check that right now17:31
ereslibre left17:31
idletask And I confirm17:31
No HEAD in subsequent added remotes17:31
subsequently, sorry17:33
keystr0k joined17:34
keystr0k left17:34
vbgunz_ left17:35
thiago_home when you clone, you check out the same branch that the remote has as HEAD17:36
idletask thiago_home: OK, that's what I figured17:38
That probably explains why HEAD is not defined on other remotes you add17:38
thiago_home yeah17:39
malumalu left17:39
idletask Hmm, I have noticed what I think is a bug, probably - here, I'm on a non tracking branch and try to git pull: http://pastebin.com/m52627af917:40
Look at line 2-417:40
It tries to contact the remote without checking the configuration first17:41
Fissure pull runs fetch, which defaults to origin17:41
idletask Fissure: ... but does so without even checking whether the branch you are currently trying to pull is a tracking branch at all17:42
As witnessed here17:42
ereslibre joined17:42
Fissure pull always updates all tracking refs for the remote, not just the one you're pulling17:43
idletask Yes I know17:43
But that's not an excuse :p17:43
wereHamster except if you use 'git pull <remote> <branch>17:43
idletask wereHamster: in this particular case, I didn't17:43
Fissure wereHamster: for certain definitions of "always" ;)17:44
idletask imho, git should have detected the missing configuration before doing the fetch17:44
Fissure idletask: questionable design decision, maybe, but i'm not sure it's unintended17:45
RaceCondition left17:48
incredible Hi! I try to run "git gui" but it says "git: 'gui' is not a git-command. See 'git --help'." I am using Ubuntu 9.10 (on VirtualBox from Win7) if that matters (but I think it doesn't).17:48
gitk works splendid, but I wanted to try other interfaces as well...17:49
DrNick you probably have to install more git packages17:50
incredible I installed git-core (on Ubuntu)17:50
Do I need something else?17:50
DrNick probably17:53
incredible DrNick: ok, thanks... I'll search (the ubuntu repositories)...17:54
wereHamster incredible: if you are on ubuntu, probably 'apt-get install git-gui'17:55
mjf joined17:58
incredible wereHamster: thank you vey much... now works like a charm :D thanks a lot :D17:58
doener idletask: origin/HEAD also works as a shortcut. It's a symref, to e.g. origin/master. And then you can say "git log origin/master", or "git log origin/HEAD", or (and that's the shortcut) just: git log origin17:59
idletask: i.e. it kind of defines the "default" remote tracking branch for that remote18:00
incredible I think git gui won't work on Windows or am I wrong?18:00
doener idletask: and you can create/update such <remote>/HEAD symrefs using "git remote set-head ..."18:00
idletask doener: I have read about the set-head operation indeed18:01
exygen joined18:01
idletask doener: I need to reread the manpage in the light of what I have just learned18:01
doener idletask: so if you're mostly working with foo/topicX, you can say "git remote set-head foo topicX", and get foo/HEAD as a symref to foo/topicX. And then you can start using e.g. "git log foo" to save some typing18:02
idletask doener: I suppose there's a way to push that HEAD if you have push power on a remote?18:02
incredible_test left18:02
idletask Well, "power"18:03
You see what I mean18:03
ngw|awayngw18:03
doener idletask: you mean to update the "real" HEAD of the remote repo? No.18:05
idletask: The name origin/HEAD (or whatever) is (to some degree) misleading18:06
ubitux joined18:06
ubitux hi18:06
idletask Ah18:06
So, I indeed have other things to understand18:06
doener idletask: clone happens to set origin/HEAD to e.g. origin/master, when the remote's HEAD references the remote's refs/heads/master18:06
idletask: but that's really the only "link" between origin/HEAD and origin's HEAD18:07
RaceCondition joined18:07
aconbere joined18:07
Fissure incredible: git-gui works fine on windows... i think tk looks better on windows than it does on linux, even18:07
ngwngw|away18:07
idletask opens the steam valve - his brains are overheating18:07
ia left18:07
ubitux I've an issue with git bisect, it seems that the reset option does not work as I expected: http://pastie.org/75712118:08
hobodave joined18:09
doener idletask: basically, just think of origin/HEAD as a convenience thing, it allows you to say "origin" instead of "origin/foo". But it has no connection to origin's HEAD, it's purely local18:09
ubitux can anyone explains this to me ? :/18:09
wereHamster ubitux: what does 'git rev-parse master' say?18:09
ubitux 122b5696805a9a94cae16334ffe125dacb12a19b18:10
wereHamster ubitux: 'git bisect bad' == 'git bisect bad HEAD' and when you are on master it translates to 'git bisect bad 122...'18:10
so you said that 122b.. is both good and bad18:11
incredible Fissure: thanks18:11
idletask doener: a side question maybe, but what happens on a clone when HEAD points to a branch with another name than master?18:11
Say, foo18:11
ubitux oh I made a mistake, I though it was an old revision18:11
sorry, thanks wereHamster18:11
idletask Will you have origin/HEAD pointing to origin/foo, and the initially checked branch being foo?18:11
icwiener left18:12
doener idletask: right (in general, there are a few 'tricky' things, when the protocol can't transfer symrefs as symrefs [e.g. git://])18:12
ngw|awayngw18:13
Archiloque joined18:14
aconbere left18:18
idle_task joined18:18
idle_task Disconnected again18:18
I "love" my Internet connection18:18
ngwngw|away18:20
ia joined18:23
schmukkk joined18:24
ngw|awayngw18:24
tibra left18:27
yhager_yhager18:29
idletask left18:31
idle_task http://qt.gitorious.org/qt-jambi/pages/GitGuide <-- anyone care to read up to "Old stuff below – don’t read, please…" and comment? :p18:31
voker57 joined18:34
brett_h joined18:35
Axius joined18:35
doener left18:36
project2501a joined18:37
doener joined18:37
ius left18:39
babilen left18:39
ngwngw|away18:39
babilen joined18:40
froschi1 joined18:41
teukka fyi: a simple vim syntax highlighting for gitolite.conf: http://www.vim.org/scripts/script.php?script_id=290018:41
MrC I've got a bare repository which I'm pushing my local commits to via SSH. The location is accessible with a webserver, so that others can clone the contents via HTTP. The setup worked great for several months, but now cloning over HTTP stopped working (curl_result=22, http_code=403). The Apache log shows some 404 errors. Any ideas how I could fix this?18:45
Fissure 403 is permission denied, yes? what happens if you try to navigate to the repo via a browser?18:49
MrC It shows the contents.18:50
And yes, 403 is permission denied.18:50
ius joined18:50
AAA_awright_AAA_awright18:50
selckin well figure out the 404s18:51
some probably probably not readable anymore by apache18:52
+files18:52
MrC It says it can't find objects/15/be897b73bd52d08cbc02f0716c680b26a16ec518:52
Indeed, that file isn't there.18:52
d-snp joined18:52
MrC But it also isn't existing on my local branch.18:52
tg left18:53
MrC After that, it looks for two files in objects/info: alternates and http-alternates.18:53
Both not there.18:53
selckin do you run git-update-server-info after pushing?18:53
tg joined18:53
MrC Yes, I've got the hook enabled, but running it manually didn't solve the problem, either.18:53
froschi left18:56
Archiloque left18:56
Ilari MrC: Try 'git rev-list --all --objects >/dev/null' in remote bare repository and see if it complains about anything?18:57
vivien left18:57
Ilari MrC: Better yet, instead of redirecting to /dev/null, grep the output for 15be897b73bd52d08c18:58
MrC No output18:59
After grep, that is, of course.18:59
Oh, no, wait....18:59
Typo19:00
It found 15be897b73bd52d08cbc02f0716c680b26a16ec5 description19:00
Ilari MrC: 'git cat-file -t 15be897b73bd52d08cbc02f0716c68'?19:00
project2501b joined19:01
MrC Ilari: "blob"19:02
Ilari MrC: So since that ins't loose (it would show up with the filename you looked for) its packed... Wonder why that pack doesn't get fetched...19:03
MrC: And 403 but apache logs show 404? 403's should show up in logs.19:03
bambule joined19:03
MrC Ilari: Just checked again, no single 403 in the Apache logs.19:05
ramlev joined19:07
ramlev left19:07
ramlev joined19:07
ramlev left19:07
project2501a left19:08
HG` joined19:09
ankit9 joined19:09
FunkeeMonk joined19:10
icwiener joined19:10
idle_task So, no comments on my guide yet, it seems :p19:14
munichlinux joined19:15
mjf left19:17
ngw|awayngw19:19
kuber left19:19
pantsman joined19:19
kpreid_ joined19:20
Lisimba left19:20
calin joined19:25
_gpg_ joined19:28
Xeross_Xeross19:32
hobodave left19:35
ehabkost_ joined19:36
jelmer_ joined19:37
jelmer_ left19:37
pantsman left19:41
Axius_ joined19:42
brett_h left19:42
Twisted_Mentat joined19:43
jelmer_ joined19:43
Axius left19:43
jelmer left19:44
jelmer_jelmer19:44
jelmer_ joined19:44
nullvariable joined19:46
malumalu joined19:46
DaveIngram joined19:47
rhett joined19:48
dreiss_ joined19:50
ralf_e left19:50
sgh joined19:52
ehabkost left19:53
gitster idle_task: given the audience and the disclaimer at the beginning of the document, I wouldn't expect you to get much feedback here on #git and it seems that comments on technical accuracy is discouraged, so I would refrain from commenting myself.19:54
wolog left19:57
Axius_ left19:57
wolog joined19:57
Axius joined19:58
idle_taskidletask19:58
hyperair left19:58
bugfux joined20:01
ToxicFrog joined20:02
idletask gitster: that makes sense20:03
But I was rather asking for blatant errors20:03
khmarbaise left20:06
gitster Other than an unqualified "tracking branch" that is a cause of confusion, I didn't spot any (I stopped reading at "please don't read" line).20:07
Often people refer to refs under "refs/remotes" as "tracking branches" (dropping "remote" from "remote tracking branches") but that is not what you are describing there.20:09
thiago_home that's very confusing20:10
local branches that track remote branches, the local copy of remote branches, the branches on the remote repository20:10
gitster Other people refer to local branches that builds on top of (meaning "pulls from and sometimes pushes to") branches they obtain from the remote (aka "remote tracking branches") "tracking branches". That is confusing as well.20:10
idletask Hmm, yes20:10
I came up with "tracking branches" for lack of a better way of explaining things20:11
gitster If anything, they are "trailing" branches perhaps, but don't use that word (I've never seen anybody use that word to describe them).20:11
idletask But thiago_home has come up with something20:11
"the local copy of remote branches"20:11
thiago_home do like the FSF: come up with new terms for everything20:11
idletask That's a more accurate description20:11
thiago_home lawyers love the "convey" in GPLv3 :-P20:11
HG` left20:11
gitster I'd try not to say "tracking branches" to refer to either. For remote stuff I say "remote tracking branches" and for local "your own fork" is how I call them to avoid confusing.20:11
reprore left20:12
hyperair joined20:12
idletask OK, I'll rework this part, but I need to gather my thoughts first20:12
And of course, thanks for the feedback!20:12
thiago_home note that you can track another local branch20:14
gitster s/track/build on/20:14
thiago_home yeah20:16
bluebug joined20:21
khmarbaise joined20:25
FunkeeMonk left20:25
FunkeeMonk_ joined20:25
FunkeeMonk_FunkeeMonk20:25
kpreid_ left20:26
ngw left20:27
DaveIngram left20:27
bluebug left20:29
Skolem joined20:29
Skolem Hey, I have a .git directory - how do I check out the latest version of all the files? I did "git checkout" but it just printed a bunch of lines beginning with 'D'20:30
Ilari Skolem: Sounds like $faq non-bare20:32
Gitbot Skolem: Pushing to non-bare repositories is discouraged. See http://git.or.cz/gitwiki/GitFaq#non-bare20:32
charon admires Ilari's crystal ball20:33
Skolem Ilari, I'm taking that to mean I should search for the terms [non-bare repository] on Google. Thanks!20:34
numberGrey left20:35
Ilari Skolem: Basically, don't push to checked out branch...20:35
Skolem I basically have the .git directory, but I deleted all the files outside of .git... I want to get them back20:36
(deleted with /bin/rm, not with git)20:36
Ilari Skolem: 'git checkout -- .'?20:36
Skolem Ah, the '.' made all the difference. I'll have to re-read "git help checkout". Thanks a bunch, I'm all set now.20:37
Axius left20:37
FunkeeMonk left20:39
sgh left20:40
yhager left20:40
idletask Argh, this is not easy :/20:42
Axius joined20:46
jelmer left20:46
Skolem left20:46
bugfux left20:46
munichlinux left20:47
Axius left20:48
nanothief joined20:51
incredible left20:51
Axius joined20:52
becomingGuru joined20:53
_Vi joined20:57
Axius left20:59
tg left21:00
project2501b left21:00
tg` joined21:00
Chillance left21:00
Axius joined21:02
s0ber left21:05
AkumaStreak joined21:07
Yuuhi left21:07
Karajan joined21:09
D[a]vey joined21:11
jceb left21:11
yhager joined21:12
s0ber joined21:13
ixti left21:13
D[a]veyDavey21:15
shadowmasterblahblah21:18
Karajan left21:20
_gpg_ left21:20
blahblahshadowm_laptop21:20
shadowm_laptoppatchpatch21:21
Axius left21:21
patchpatchshadowmaster21:23
thiago_home left21:24
Lisimba joined21:25
AkumaStreak left21:26
AkumaStreak joined21:26
Mathnerd314 joined21:27
xlistener joined21:27
bugfux joined21:27
aspotashev joined21:29
Axius joined21:32
_Vi left21:35
Axius left21:35
iron joined21:35
__iron left21:36
bugfux left21:36
jelmer joined21:40
gdrage joined21:41
jeanrussou joined21:41
jeanrussou left21:41
jelmer left21:41
Rolenun joined21:42
Axius joined21:42
bambule left21:43
khmarbaise_ joined21:43
Axius left21:43
kristian42 left21:44
kristian42 joined21:45
Mathnerd314 left21:45
Davey left21:45
neun_ joined21:47
_Vi joined21:48
jelmer joined21:54
nullvariable left21:57
nullvariable1 joined21:57
nullvariable1nullvariable21:57
Sky[x] left21:57
skyl joined21:57
mbroeker left21:57
skyl I want to go back to what I had in commit 7f01e83... for a file, foo/bar.baz21:58
jtbandes I think you use checkout for that21:59
khmarbaise left21:59
khmarbaise_khmarbaise21:59
gitster check the file out of the commit (i.e. "git checkout 7f01e83 -- foo/bar.baz")22:00
dreiss_ left22:01
icwiener left22:02
calin left22:03
calin joined22:04
CM__ joined22:04
CM__ Hi -- i've a git repo which originally was done from a git cvsimport -- occasionally what i would like to do is clone this git repo, and then use git cvsimport to update this repository from CVS -- what's the best way of doing this?22:05
khmarbaise left22:09
skyl thanks all, see you next time22:10
skyl left22:10
japhb joined22:10
japhb I have a topic branch off of master, and that topic branch is getting ready to merge back; I've rebased it against the current HEAD of master and resolved all conflicts there. However, the topic branch is somewhat dirty (a few bad or mixed-quality commits, a few commits that don't belong in master yet), and I'd like to copy the good bits to a clean branch of master, ready for final test and merge back. What's the best way to go about this?22:15
From reading the docs, it looks like 'git rebase -i' would do the right thing if I wanted to edit the topic branch directly -- but I want to do my cleanup in a new branch instead.22:15
kpreid_ joined22:18
yhager left22:20
yhager_ joined22:20
gambler left22:20
ehabkost_ left22:20
gitster "git checkout -b new-branch topic && git rebase -i on-top-of-whatever"?22:22
gambler joined22:24
MrC left22:24
tekky joined22:26
kpreid_ left22:31
bcardarella joined22:32
stamina_ joined22:32
cannonball joined22:34
thiago_home joined22:37
nitrate__ left22:37
d0k joined22:38
FunkeeMonk joined22:39
d0k left22:42
tg` left22:45
japhb gitster, ah, interesting idea. Thanks, will give that a try.22:45
tg` joined22:45
d0k joined22:48
hipe left22:50
jelmer left22:53
tekky left22:53
gitster actually wants to know why "checkout -b && do it" was not obvious and instead was "interesting" for "I want to do my cleanup in a new breanch instead". It is a sad indication that something is wrong in the documentation and tutorial...22:57
cannonball left22:58
thiago_home left23:03
japhb gitster, well, unless I understood you, sadly it's interesting in the sense that it doesn't actually work.23:03
lucsky left23:03
japhb er, "misunderstood"23:04
I can't seem to figure out an incantation of 'git rebase -i --onto master [<something here>]' that produces anything other than an error message.23:06
bcardarella left23:06
neun_ left23:07
maxter left23:12
idletask Question: what are (non-object) tags wrt branches? Just symbolic refs that you cannot commit on?23:15
Ilari idletask: symbolic refs are something different (they are like symbolic links).23:16
idletask: Tags in git are pretty much like branches one can't directly check out...23:17
idletask OK, forget about "symbolic" then23:17
Ilari japhb: What's the first commit on current you don't want to rebase?23:18
idletask Ilari: uh, really?23:18
Ilari idletask: Trying to check one out results detached HEAD.23:19
idletask Ilari: ah, OK - but then it doesn't prevent you from bisecting using one of these, or creating a branch from one of these, AFAICS - correct?23:20
Ilari idletask: Right.23:20
idletask: In fact, for such purposes, any commit ID can be used.23:21
idletask Ilari: same for git archive, I gather?23:21
And by "commit ID", I gather you mean the SHA1?23:21
stamina_ left23:22
rhett left23:23
Ilari idletask: commit ID is special case of object ID. And both are also called SHA1s.23:24
cannonball joined23:24
idletask Ilari: well yes, I didn't mean to generalize that far, but I am partially correct, amn't I? :p23:26
Ilari idletask: Both are used...23:27
tibra joined23:29
fcuk112_ left23:29
fcuk112__ left23:30
hobodave joined23:30
drizzd joined23:35
segher_ joined23:36
tibra left23:36
stamina_ joined23:42
segher left23:43
Jak_o_Shadows joined23:45
intripoon left23:52
tbuehlmann left23:57
gdrage left23:57
jelmer joined23:57
aleray joined23:59

Logs Search ←Prev date Next date→ Channels Documentation