IRCloggy #git 2011-08-28

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.

2011-08-28

lisak kevlarman, indedd00:00
wtf00:00
why ?00:00
I'm missing here something :)00:00
important :)00:01
DrNick because they aren't a part of the branch you switched to00:01
kevlarman lisak: because git refuses to overwrite untracked files00:02
so if it didn't remove them, you wouldn't be able to switch back00:02
and leaving files around when they should no longer be there can cause problems00:02
← industrial left00:03
lisak got it, thank you00:04
← fread2281 left00:05
→ Spockz joined00:09
→ zemo joined00:11
← kennethreitz left00:11
zemo im not being able to use git archive: fatal: Not a valid object name00:11
any help?00:11
cirwin what are you passing to it?00:11
I think the first argument should be a revision, not just files00:12
i.e. git archive HEAD bar baz00:12
zemo you think?00:12
cirwin man git-archive00:12
jast *automatic message* the 'git-archive' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-archive.html00:12
zemo i never used that and so far worked00:12
cirwin there's a <tree-ish> before the files00:13
if you talk to the channel, you might find people know more than me can help00:14
→ towski joined00:14
← chessguy left00:14
zemo this has just worked: qtwebkit]$ git archive --prefix=qtwebkit-2.2/ qtwebkit-2.2 {autogen.sh,ChangeLog,configure.ac,GNUmakefile.am,Makefile,Source/,Tools/} | xz -9 > qt.tar.xz00:14
cirwin: i used the query to paste only00:14
→ Marquel joined00:15
cirwin I think qtwebkit-2.2 is a tag name?00:15
that is the <tree-ish>00:15
cirwin just guessing though00:15
→ p3rsist joined00:16
zemo cirwin: tag names arent output with `git branch .-r`?00:16
cirwin I don't think so, try git tag00:17
or git for-each-ref if you want allll of them00:17
zemo theres no tag name qt-webkit-2.200:17
cirwin shuts up then, I've no idea00:17
← cexsum left00:18
zemo cirwin: sorry, it exists, i wasnt in right repo00:18
→ cableray joined00:19
← patmaddo_ left00:19
← kpreid left00:20
→ kpreid joined00:21
→ caseymcg joined00:21
zemo SethRobertson: ping00:23
← kpreid left00:23
SethRobertson ?00:23
cirwin zemo: I don't know if you can use git-archive on uncommitted stuff00:23
→ patmaddo_ joined00:23
→ kpreid joined00:24
zemo SethRobertson: i have added konsole and kate into git local repo, but now fails to use git archieve00:24
even tried with gits archive00:24
SethRobertson: http://privatepaste.com/e15375686800:26
SethRobertson: isnt possible to use git archive when were added new fodlers into git locally?00:29
→ chrisf_ joined00:29
→ DofLeFou joined00:30
← w0bni left00:32
cbreak zemo: what?00:33
→ zamabe_ joined00:33
← chrisf_ left00:33
SethRobertson zemo: git-archive only works against files checked into git00:33
cbreak git archive archives commits00:34
you told it to archive HEAD00:34
do you have the things you want to archive in HEAD?00:34
if not, stop complaining that git can't do the impossible00:34
SethRobertson zemo: The problem is most likely that you cannot provide pathnames to archive when using gits, since not all pathnames will exist in all repos00:35
zemo: If you repeat the command using `git` and it fails, or without using pathnames, then we can talk00:36
← zamabe left00:36
zemo SethRobertson: yes it fails00:36
SethRobertson Can you pastebin that output00:37
cbreak how?00:37
SethRobertson also including `git status`00:37
cbreak I suspect user error :)00:37
→ Mac_Weber joined00:37
zemo http://privatepaste.com/3c4161677000:37
← zamabe_ left00:37
cbreak yeah, user error00:37
just leave away the path filters00:38
zemo http://privatepaste.com/38af602e3e00:38
Mac_Weber how to list only the files modified between branches?00:38
→ seivan joined00:38
→ zamabe_ joined00:38
← dreiss left00:38
zemo cbreak: user error?00:38
cbreak yes00:38
you are doing it wrong00:38
zemo did you saw the git status?00:39
cbreak are the things you want to archive in git ls-tree -r HEAD?00:39
→ cexsum joined00:39
SethRobertson zemo: just say: git(s) archive --prefix=kde-baseapps/ HEAD00:39
cbreak git status is irelevant00:39
SethRobertson I just wanted to make sure everything was committed00:39
cbreak git archive will archive a commit00:39
git status does not do anything to commits, it's for inspecting staging and work area00:40
zemo SethRobertson: it is00:40
cbreak if git ls-tree doesn't show the things, git archive can not archive them00:40
→ AaronMT joined00:40
SethRobertson zemo: Run the command I mentioned above, and let us know if that fails00:40
→ axl_ joined00:41
kevlarman Mac_Weber: git diff --name-only branch1...branch200:41
→ Adys joined00:41
← AaronMT left00:42
zemo with that command konsole abreaks...00:42
→ tureba joined00:43
→ kpreid_ joined00:44
→ berserkr joined00:44
← kpreid_ left00:45
← kpreid left00:45
→ kpreid joined00:45
zemo SethRobertson: it worked but doesnt archive folders kate and konsole like it would be expected00:46
← cexsum left00:46
SethRobertson zemo: Using `git` or `gits`00:46
zemo git00:46
ill try now with gits00:46
jast check that they exist in HEAD: git rev-parse HEAD:konsole00:46
error = doesn't exist00:46
← seivan left00:47
→ arunce joined00:49
zemo using gits creates the tar file but empty and continues in cli in a loop00:51
← churp_ left00:52
← saste left00:52
← shiba_yu36 left00:53
→ corburn joined00:56
← ehsan left00:57
← domino14 left00:57
→ volter joined00:58
zemo SethRobertson: seams using gits archive doesnt compress any00:59
SethRobertson Yes. it is because we have to update the tarfile as each archive runs. You can just gzip the resulting file afterwards.01:00
zemo SethRobertson: gits archive --prefix=kde-baseapps/ HEAD |xz -9 > example.tar.x01:01
SethRobertson: it simply created an empty file01:01
→ normanrichards joined01:01
← normanrichards left01:01
SethRobertson As stated in the gits archive man page: When you select the gits-tar option,01:02
you must supply an on-disk --output file and you cannot use any tar01:02
compression options01:02
Example: gits archive --format gits-tar -o /tmp/foo.tar master01:02
zemo ok thanks01:03
ill check that later, now i need to leave01:03
→ linusoleander joined01:04
← linusoleander left01:04
→ oc80z joined01:05
→ jeffery joined01:05
← brian_g left01:06
zamabe_ → zamabe01:13
← Mac_Weber left01:15
← __sorin__ left01:16
← volter left01:17
→ Araxia_ joined01:17
→ shurane joined01:19
← berserkr left01:20
→ cjuner_ joined01:20
← oc80z left01:21
→ oc80z joined01:22
→ rvsjoen joined01:23
← cjuner left01:23
← rvsjoen_ left01:25
→ jsjgruber joined01:27
← jsjgruber left01:27
→ simenbrekken joined01:30
← kpreid left01:30
→ kpreid joined01:30
← kpreid left01:34
← shurane left01:34
→ kpreid joined01:34
→ chrisf_ joined01:36
→ shurane joined01:37
→ artnez joined01:39
→ cypha joined01:40
cypha I'd like to see a diff of a .py file in my git repo, comparing today and 2 days ago01:41
can I do this?01:41
jast sure. find an old commit ID (e.g. using "git log"), then "git diff <old commit> HEAD <file>"01:42
this will compare against the most recently committed version of the file. if you want to compare to the working tree version, just omit the "HEAD"01:42
cypha working tree?01:43
jast the physical file01:43
as opposed to a version stored within one of the commits01:43
cypha ohh, gotcha01:43
what's git log?01:44
jast a command that shows previous commits on the current branch (or any other branch i fyou tell it to)01:44
it can also do a bajillion other things related to looking at your history01:44
← mlncn left01:44
kevlarman if you really wanted you could do something like git diff 'master@{2 days ago}'..HEAD -- file01:45
cbreak warning: this does not use the history01:47
it uses the reflog01:47
→ mlncn joined01:47
jast yeah, it's subtly different from actually using the version you had committed two days ago01:47
→ a_suenami joined01:47
cbreak using the history would give you the item as it was on that branch 2 days ago01:48
using the reflog would give you the item as it was in your version of that branch 2 days ago01:48
jast perhaps that's a topic for later :)01:48
cbreak for example, if you didn't commit at all in the last week, you'd get your 7 day old version with the reflog, but the newer version that you pulled with the history version.01:49
→ AaronMT joined01:49
→ jimubao joined01:50
← B|aster left01:50
← AaronMT left01:51
cypha git log01:53
→ Bass10 joined01:53
cypha I need to use the commit key?01:53
actually 'git log' is only showing me stuff from today01:54
changes from today01:54
cbreak it should give you the whole history01:54
everything that ever happened on that branch you are on01:54
← Bass10 left01:54
jast it opens a pager. try using the arrow keys or stuff like that. ;)01:54
→ patmaddox joined01:55
← gnoel left01:55
← gnoel_ left01:55
→ Bass10 joined01:55
← patmaddox left01:55
→ gnoel_ joined01:57
→ gnoel joined01:57
← patmaddo_ left01:58
→ B|aster joined01:58
← ceej left01:59
→ seivan joined02:00
→ ceej joined02:01
← p3rsist left02:01
→ redii joined02:02
← ceej left02:03
← kmc left02:03
← cjuner_ left02:05
← Andrew__ left02:06
→ VladGh joined02:07
← Octalot left02:07
→ SirFunk joined02:09
← artnez left02:10
→ broly_mcbroski joined02:13
← Khisanth left02:14
→ squiggy joined02:14
← broly_mcbroski left02:19
→ preflex joined02:25
→ ricky joined02:27
→ MUILTFN joined02:29
→ Khisanth joined02:30
→ frhodes joined02:30
← dv310p3r left02:33
→ CannedCorn joined02:34
→ dv310p3r joined02:37
→ ali1234 joined02:40
ali1234 if git diff -w hides whitespace changes, is there a way to do the opposite and show only whitespace changes?02:40
cirwin ali1234: not that i know of, what are you trying to do?02:42
ali1234 well02:42
i have some software in a git repo02:42
and a huge "here's everything i've changed" patch against it02:42
that patch includes all kinds of shit02:43
and i'm trying to clean it up into something sane02:43
cirwin hmm02:43
ali1234 by splitting out the whitespace changes etc into actual functional changes02:43
→ alnewkirk joined02:43
ali1234 so i have one commit with whitespace changes, one commit with the "find and replace all instances of the old name with my fork's name lol", and then a sequence for the interesting parts02:44
cirwin yeah, I can see why you'd want to do that02:44
but not sure how, sorry02:44
→ demods joined02:46
← jimubao left02:46
→ warthog9 joined02:51
→ Chib joined02:51
→ mino joined02:52
→ dstanek joined02:52
→ Vampire0 joined02:53
← dv310p3r left02:53
← alnewkirk left02:55
→ ericbarnes joined02:57
→ JakeyChan joined02:57
← bonhoffer left03:01
← MUILTFN left03:03
← demods left03:03
→ demods joined03:03
← demods left03:05
→ RobertLaptop joined03:08
→ bonhoffer joined03:08
→ demods joined03:08
← gnoel_ left03:08
← gnoel left03:08
← posciak left03:09
→ kvanderw joined03:09
← kvanderw left03:10
← B|aster left03:10
← cirwin left03:11
→ Orbitrix joined03:13
→ w09x joined03:13
→ doug joined03:15
doug i've got two branches that i want to keep only MOSTLY in sync..03:15
i.e. any commit to branch alpha i'd like to be able to easily merge into beta03:15
but there's a commit historically in alpha that i never want to merge over.03:16
(i.e. the file in the alpha branch has "alpha" in it, in beta it's "beta," but everything else is the same)03:16
i'm hoping git has a mechanism to ignore certain commits when doing merges...03:16
← simenbrekken left03:18
← demods left03:18
DrNick doug: nope03:18
doug boogers03:18
was hoping it didn't have to get messy03:19
→ marienz_ joined03:21
← JakeyChan left03:23
doug > Reverting a merge commit declares that you will never want the tree changes brought in by the merge. As a result, later merges will only bring in tree changes introduced by commits that are not ancestors of the previously reverted merge. This may or may not be what you want.03:25
hm, that kinda sounds like what i'd want03:25
← adamm left03:26
doug merge in the commit i want to avoid, then revert it...03:26
→ simenbrekken joined03:26
→ JakeyChan joined03:27
→ demods joined03:27
← ZenDualism left03:29
← bonhoffer left03:30
tango_ doug: you can merge, revert the commit you want to skip forever, and then squash the commit + revert together03:30
doug nice03:31
← simenbrekken left03:31
tango_ btw that only works if you're merging one-way only03:31
if you ever merge the other way too, that revert is going to hit the other branch too03:32
→ Slush- joined03:32
EugeneKaway That would still keep the commit in the history someplac, before it was reverted03:32
(and squashed)03:32
You may want to look at rebasing or cherry-picking03:32
→ roop joined03:32
EugeneKaway Or.... stop caring and just do the dang thang :-p03:32
tango_ having it in history is important03:32
if you do rebasing or cherry-picking, future merges are going to be a pain03:33
→ angelsl joined03:33
→ cyphase joined03:34
← shurane left03:35
→ dv310p3r joined03:36
← notjohn left03:36
← axl_ left03:45
→ Samoi joined03:45
→ bonhoffer joined03:47
→ kenperkins joined03:50
→ tiagobutzke joined03:52
→ dreiss joined03:57
→ shurane joined03:58
← Khisanth left04:01
← bonhoffer left04:02
← dv310p3r left04:03
← OOPMan left04:04
← seivan left04:06
→ BiggFREE joined04:08
→ kmc joined04:09
→ cilly joined04:13
→ peper joined04:13
→ Chaser joined04:15
← zamabe left04:16
← tiagobutzke left04:16
← Murr_ left04:16
→ Khisanth joined04:16
→ tiagobutzke joined04:18
→ cableray_ joined04:18
→ akosikeno joined04:18
← akosikeno left04:19
→ xiangfu joined04:20
← caseymcg left04:20
→ cirwin joined04:21
← cableray left04:22
cableray_ → cableray04:22
→ fread2281 joined04:23
→ OOPMan joined04:25
→ fread228_ joined04:26
← w09x left04:28
→ seivan joined04:28
← shurane left04:28
← fread2281 left04:28
← nullvariable left04:29
→ dr0id joined04:29
← Bass10 left04:33
← kmc left04:33
→ kmc joined04:34
← ericbarnes left04:37
→ maleknet joined04:38
← ISF left04:38
→ SimonNa joined04:39
← xiangfu left04:39
← maleknet left04:40
→ B|aster joined04:41
← chrisf_ left04:42
→ maleknet joined04:42
→ Gojohnnyboi joined04:43
← DofLeFou left04:43
→ xiangfu joined04:45
→ cytrinox_ joined04:45
← Textmode left04:46
← Khisanth left04:47
→ kvanderw joined04:47
← VladGh left04:48
cytrinox_ → cytrinox04:48
→ caseymcg joined04:49
← kvanderw left04:50
→ dannas joined04:53
← roop left05:01
→ Khisanth joined05:02
→ aminpy joined05:06
← madsy left05:06
→ Fudge joined05:07
→ NuckingFuts joined05:07
← Nuck left05:10
NuckingFuts → Nuck05:10
→ Civilize joined05:11
← Khisanth left05:12
← Civilize left05:13
→ FreezingCold joined05:13
→ Adaptee joined05:15
→ ehsan joined05:18
→ roop joined05:18
← caseymcg left05:21
→ pizza_ joined05:23
→ Khisanth joined05:26
← CannedCorn left05:29
→ tmpvar joined05:31
tmpvar hello, is there a way to get a `git tag -l` sorted by creation date?05:32
→ mrsrikanth joined05:35
→ robotmay joined05:35
← fread228_ left05:35
← kenperkins left05:35
← JakeyChan left05:38
→ shurane joined05:39
← jablko left05:40
cirwin git log --tags --no-walk --date-order --oneline ?05:43
--decorate05:44
→ akosikeno joined05:45
← towski left05:46
→ towski joined05:46
→ martinciu joined05:47
← ehsan left05:50
→ m4dc0d3r_ joined05:52
→ CannedCorn joined05:54
← xiangfu left05:54
→ JakeyChan joined05:56
→ chrisf_ joined06:02
→ helber joined06:07
← tiagobutzke left06:10
← kmc left06:11
← troyt left06:13
→ kmc joined06:15
→ caseymcg joined06:15
← JakeyChan left06:15
→ E0M joined06:15
← ricky left06:16
← julius_ left06:16
← kermit left06:18
→ ricky joined06:20
← helber left06:21
→ p3rsist joined06:21
→ xiangfu joined06:24
→ bgerber joined06:25
← himsin left06:25
← pizza_ left06:27
→ Myth17 joined06:28
Fudge hi i have a problem, my branch was setup to natty+fudge but i think along the line i copied my squeeze+fudge stuff to it and want to checkout squeeze+fudge but i cant, how can i discard and start again or checkout squeeze+fudge06:28
→ kermit joined06:30
→ julius_ joined06:30
← E0M left06:31
→ pen joined06:31
← pen left06:31
→ pen joined06:32
← pen left06:32
→ nitish joined06:33
← robotmay left06:36
← angelsl left06:37
← Myth17 left06:37
→ pen joined06:38
→ angelsl joined06:40
→ pizza_ joined06:41
cirwin Fudge: I'm not quite sure what you mean06:42
Fudge hi cirwin06:44
i was in natty branch, but was working on anothe rbox with squeeze branch and i think i dumped squeeze stuff into my dirs without checking out squeeze branc06:45
does tha tmake sense06:45
cirwin yup06:45
Fudge so now when i try to git checkout squeeze it says move or remove a load of stuff06:45
im very limitted with this06:45
cirwin git checkout -f06:45
should just ignore the warnings06:45
but it might also throw away any changes that you don't want it to :p06:46
→ Error404NotFound joined06:46
Fudge i can risk it06:46
← warthog9 left06:47
Fudge yay two chars06:47
:D06:47
thanks so much06:47
← pizza_ left06:47
cirwin no problem06:48
← robotarmy left06:48
Fudge :D06:48
→ dSebastien joined06:51
← dreiss left06:53
← kerrick left06:55
← p3rsist left06:55
→ tizzo-afk joined06:57
← jeffery left06:58
← Error404NotFound left06:59
← corburn left07:01
Fudge cirwin failed to push some refs07:04
cirwin that -f flag might come in handy again — but before you try do "git fetch"07:04
Fudge non fast forward updates07:04
cirwin once you've done a git-fetch, you can use git log master...origin/master to see what's changed on the server07:05
← mrsrikanth left07:07
→ pizza_ joined07:08
Fudge so if i want to push what i have locally overwriting the server?07:08
cirwin guess two characters ;)07:08
→ cesario joined07:09
Fudge i tried git push -f07:09
ooh maybe i need to commit07:09
cirwin that helps :)07:09
Fudge but it says nothing to commit07:10
i get so confused07:10
cirwin what does git status say?07:10
← ali1234 left07:10
Fudge # On branch squeeze+fudge07:10
nothing to commit (working directory clean)07:10
cirwin ok07:10
you need to pass the branch name to git push07:10
unless the branch already exists on the server07:10
git push origin squeeze+fudge07:10
← nitish left07:11
Fudge it does07:11
cirwin humm..07:11
cirwin suggests trying with the branch name and -f07:11
Fudge says everythign up to date07:11
cirwin then it's probably right :p07:11
→ jutaro joined07:11
cirwin what does git log -1 --decorate say?07:11
→ nitish joined07:11
→ Error404NotFound joined07:11
→ vmil86 joined07:12
Fudge http://paste.ubuntu.com/676412/07:14
oh thats the error when push sorry07:14
only says the commit i just tried to push07:14
cirwin did you mean to push the natty and squeeze branches as well as squeeze+fudge?07:14
Fudge no jus tsqueeze07:14
i didnt know i could do both07:14
are they merged?07:15
god i hope not07:15
cirwin they're not merged07:15
Fudge lol07:15
tango_ Fudge: so now squeeze+fudge is pushed, but the other are not because they wouldn't be fast-forward07:15
cirwin but your push tried to push them07:15
cirwin always uses "git push <remote> <branch>" it saves confusion :p07:15
Fudge ah i think iget it07:15
tango_ probably you have push set to push all branches07:15
Fudge sounds like a good idea07:15
what woudl actually be causing git push to try and be pushing all07:15
cirwin Fudge: it has bad default settings07:16
tango_ Fudge: the default setting07:16
Fudge yeah wow07:16
cirwin has two maxims for using git safely "don't use git pull" and "don't use git push with no arguments" :p07:16
Fudge can i change it to jsut the branch im in07:16
tha tworks07:16
so git push remote squeeze+fudge07:16
cirwin I think there's a config variable, but I always worry I'll change computer and forget to set it07:17
→ kerrick joined07:17
tango_ hm I think they changed the default recently07:17
cirwin git config --global push.default current07:17
← B|aster left07:18
tango_ my latest git only pushes the current branch by default07:18
Fudge: what git version is that?07:18
cirwin man git-push07:18
jast *automatic message* the 'git-push' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-push.html07:18
→ citizenkeys joined07:19
→ tkil joined07:19
Fudge Version: 1:1.7.4.1-307:19
→ stodan joined07:19
← Error404NotFound left07:19
cirwin tango_: the docs for the latest build of git seem to have the old config still :( — it's probably hard to change something so core like that07:19
tkil i've been using git-svn to work on a mostly-private svn repo. i'm looking at converting to git for the repo, and i'm curious if there's any recipe for switching my local repo back to the original git commits, not those created by "git svn dcommit"?07:19
← caseymcg left07:20
tkil (in particular, git-svn discards author-date info, and i'd *really* like to recover it...)07:20
→ xmfbs joined07:20
tkil the original commits are still on my workstation(s), so i hope i can get away with just some reflog swizzling...07:20
Fudge thanx thats great07:20
cirwin tkil: I'm sure google can help better than me, you might find man filter-branch useful; and "grafts" as keyword07:21
man git-filter-branch even07:21
jast *automatic message* the 'git-filter-branch' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-filter-branch.html07:21
tkil cirwin -- ok, i'll take a look. i'm familiar with the concept of filter branch, as i will prolly want to use it even if i weren't converting (original repo had some massive binary docs/SDK downloads checked in)07:21
was hoping for some automation, because SFAIU i'll have to graft at every burst of dcommits07:22
possibly at every commit created by dcommit. mumblefrotz.07:22
cirwin tkil: it sounds like a problem other people have had — if no-one here knows, the mailing list may be able to help07:22
← tizzo-afk left07:22
tkil anyway. was just curious if anyone else had dealt with this particular situation.07:22
cirwin -- i'll ask before i do anything rash. :) thanks!07:22
→ flavius joined07:22
← pizza_ left07:23
→ aristidesfl joined07:24
→ caseymcg joined07:25
→ isxek joined07:27
→ warthog9 joined07:29
→ urkud joined07:30
→ Error404NotFound joined07:31
→ Weiss joined07:33
← nitish left07:33
← shurane left07:33
← cableray left07:34
→ nitish joined07:36
← isxek left07:36
← tkil left07:37
→ pizza_ joined07:38
→ nitish_mythology joined07:40
← soulcake left07:40
← dr0id left07:42
→ arunce_ joined07:43
← nitish left07:43
← Khisanth left07:43
← arunce left07:44
← towski left07:46
→ arunce joined07:46
← OOPMan left07:47
← Gojohnnyboi left07:47
← aminpy left07:48
→ soulcake joined07:48
← arunce_ left07:49
→ aminpy joined07:49
← aristidesfl left07:49
→ aristidesfl joined07:51
← aminpy left07:53
← maleknet left07:54
→ shurane joined07:55
→ Khisanth joined07:57
→ maleknet joined07:58
→ thiago joined07:58
→ aminpy joined07:59
← sirdancealot left08:01
→ phretor_ joined08:02
→ MUILTFN joined08:03
← urkud left08:06
← madewokherd left08:06
← citizenkeys left08:08
→ jhoepken joined08:09
← jutaro left08:10
→ linusoleander joined08:11
← linusoleander left08:11
→ linusoleander joined08:11
flavius I have "git init", now I have the repo. How do I create a new branch?08:11
say, the master branch08:11
EugeneKawaway Make a commit.08:12
cirwin the master branch is already created08:12
→ nitish joined08:12
← kerrick left08:12
flavius git branch -a doesn't show anything08:12
cirwin hmm, that's actually not quite true08:13
← Khisanth left08:13
cirwin yeah, but HEAD points to master08:13
EugeneKay That's because you don't have a commit for it to show.08:13
cirwin so when you commit, the commit will be added to master08:13
← pizza_ left08:13
→ drizzd joined08:13
flavius Uhm I don't want to commit something, I want to go straight to a branch "test", so I don't pollute the master, but git co -b test says "you're on a branch yet to be born"08:15
← xmfbs left08:15
EugeneKay Commit something, then rename master to test. It's easier, really.08:15
← nitish_mythology left08:16
EugeneKay I'm sure there's an obscure command to change the first branch name, but the time spent looking it up is better spent browsing cat videos, or something08:16
← caseymcg left08:18
cirwin git symbolic-ref HEAD refs/heads/test will probably do it, but yes commit and then "git branch -m" might be less confusing08:18
EugeneKay And leads to more of that oh-so-precious cat-video-watching time.08:18
cirwin gives EugeneKay a cheezburger08:19
EugeneKay nom08:19
→ nitish_mythology joined08:19
← nitish_mythology left08:19
→ nitish_mythology joined08:19
← nitish left08:23
→ Khisanth joined08:26
← Matth1a31 left08:26
→ babilen joined08:29
→ nitish joined08:31
→ eijk joined08:31
← roop left08:32
← doug left08:32
← MagBo left08:33
← angelsl left08:33
← nitish_mythology left08:34
← shurane left08:34
← Adaptee left08:34
→ angelsl joined08:35
← angelsl left08:35
→ Phylock joined08:37
→ nitish_mythology joined08:38
→ tewecske joined08:38
→ angelsl joined08:40
→ Myth17 joined08:41
← nitish left08:41
← aristidesfl left08:41
← phretor_ left08:43
aminpy → aminpy|lunch08:44
← nitish_mythology left08:44
→ tatsuya joined08:44
→ phretor joined08:45
→ radarek joined08:46
→ saste joined08:46
→ pantsman joined08:47
← pantsman left08:47
→ pantsman joined08:47
→ angelsl` joined08:49
← angelsl left08:49
→ jutaro joined08:50
→ Saur joined08:52
← Saur left08:52
← angelsl` left08:54
← soulcake left08:54
← rvsjoen left08:56
← FreezingCold left08:56
→ rvsjoen joined08:56
→ nicoulaj joined08:59
← d0k left08:59
→ simenbrekken joined08:59
→ madsy joined09:01
← madsy left09:01
→ madsy joined09:01
← Myth17 left09:07
→ angelsl joined09:07
← dannas left09:08
→ dannas joined09:09
→ orafu joined09:10
aminpy|lunch → aminpy09:16
→ caseymcg joined09:16
← dnyy left09:18
→ isxek joined09:18
← Pupeno left09:18
← stodan left09:19
→ Murr_ joined09:21
→ churp joined09:21
→ dnyy joined09:22
← dannas left09:23
→ matrixiumn joined09:23
matrixiumn hello09:23
using gitgui how do I make a first commit?09:24
→ dannas joined09:25
← madsy left09:27
← tatsuya left09:27
→ lisak joined09:31
← a_suenami left09:31
lisak what does it mean when $git log file prints nothing ? I cloned a repository, then created a file09:33
now I run status09:33
and the file is not listed anywher09:34
→ moshef joined09:34
lisak and I dont' know if I have added and committed it or nothing or what just happened09:35
← caseymcg left09:35
lisak it should be either untracked or not staged, staged or committed09:36
cirwin lisak: git status would show a file if you haven't committed it, and "git show" will show it if it was in the most recent commit09:36
unless it matches a .gitignore file?09:37
→ LotR joined09:37
lisak git show shows nothing09:37
omg sorry they really have this in .gitignore09:38
cirwin is the file on the filesystem?09:38
ah :)09:38
git add <file>09:38
lisak git add <file> is forcing add despite its presence in .gitignore ?09:39
→ dr0id joined09:39
cirwin well, any new file needs to be git added the first time09:39
but a git add will cause it to show up in git status09:39
← dr0id left09:39
→ dr0id joined09:39
lisak ok, thank you man09:40
→ brian_g joined09:44
← linusoleander left09:44
→ marinosi joined09:46
→ linusoleander_ joined09:48
→ kumbayo joined09:48
← frhodes left09:48
← pen left09:49
→ ThomasLocke joined09:52
→ linusoleander joined09:54
← linusoleander_ left09:56
moshef how do i view the log of a specific branch and pushes of a specific author ?09:57
cirwin you can't see who pushed what, but git log --author=John <branchname> will show you the stuff that John did09:59
moshef well here is the thing, im trying to look at a branch who hasn't been updated for quite some time10:00
when i get into the branch and ask for log i get lots of other commits that im not interested in, made by other authors10:01
\10:01
→ davzie joined10:01
moshef so figured this will help me navigate more10:01
im open for other solutions10:01
→ squeese joined10:02
← saste left10:02
→ blast_hardcheese joined10:04
→ Anubisss joined10:04
→ pen joined10:04
moshef cirwin: i still get the commits of this author from other branches as well, how do i define to get commits from this specific branch only?10:05
\10:05
→ entourl00p joined10:05
← entourl00p left10:05
← davzie left10:05
cirwin you probably want to exclude things on master10:05
try adding --not master10:05
→ entourl00p joined10:06
→ SRabbelier joined10:06
→ davzie joined10:06
moshef git log --not master --author= ?10:06
cirwin yeah10:06
and give it the branchname you do want10:06
moshef not getting anything one10:06
← entourl00p left10:06
moshef im already inside that branch10:06
its not enough?10:06
ive been told that it is10:06
cirwin if you specify --not, you need to specify HEAD explicitly10:06
moshef seeing how i still get commits from everywhere, i started to question that tho10:07
cirwin git log HEAD --not master10:07
moshef oh10:07
well list is smaller now but its still not enough. there is no way to ask for commits on this branch only ?10:08
→ rendar joined10:08
moshef as in branch = x or anything like that, cuz the list is still pretty huge10:08
cirwin if you know the starting point of the branch, sure10:08
← Khisanth left10:09
moshef its like finding a needle in...10:09
← isxek left10:09
moshef what do you mean the starting point10:09
sorry, im pretty new to git as you can see :)10:09
cirwin there must be one commit that's not on this branch10:09
but which is the parent of a commit that is10:09
→ rostayob joined10:09
cirwin git for computer scientists might help if you have some reading time10:09
the basic problem is that git doesn't know which branch a commit is on, as branches can be moved around on a whim10:10
they just point to commits10:10
so you need a way to identify the set of commits you're interested in10:10
moshef thanks for the reference, i will make time :)10:10
cirwin git merge-base can help with that10:10
moshef git merge-base and branch name?10:12
→ pizza_ joined10:15
← matrixiumn left10:15
← kmc left10:17
← pen left10:18
→ pen_ joined10:18
→ Dave^| joined10:19
→ Pupeno joined10:19
← Pupeno left10:19
→ kmc joined10:20
moshef cirwin ?10:20
← cirwin left10:21
marienz_ → marienz10:22
→ Agos joined10:24
→ apocalyptiq joined10:25
→ Khisanth joined10:26
→ bonhoffer joined10:26
→ phretor_ joined10:29
→ shruggar joined10:29
→ jargon- joined10:31
→ tobago joined10:31
→ jceb_ joined10:32
→ LongBeach joined10:32
← phretor left10:32
← martinciu left10:32
→ Hallin_ joined10:35
→ Gitzilla joined10:36
← Hallin_ left10:36
→ DPav joined10:36
← LongBeach left10:37
← jceb_ left10:37
← dstanek left10:37
← Araxia_ left10:37
← Spockz left10:37
← SillyWabit left10:37
← MUILTFN left10:37
← julius_ left10:37
← Chib left10:37
← Sonderblade left10:37
← mino left10:37
← nyuszika7h left10:37
← plopix left10:37
← flavius left10:38
→ LongBeach joined10:38
→ jceb_ joined10:38
→ MUILTFN joined10:38
→ julius_ joined10:38
→ dstanek joined10:38
→ mino joined10:38
→ Chib joined10:38
→ Araxia joined10:38
→ Spockz joined10:38
→ Sonderblade joined10:38
→ SillyWabit joined10:38
→ nyuszika7h joined10:38
→ CyberPony joined10:38
→ bauruine joined10:38
→ Spyro joined10:38
→ koollman joined10:38
→ crass joined10:38
→ f00li5h joined10:38
→ MetaCosm joined10:38
→ daleharvey joined10:38
→ serverhorror joined10:38
→ kylebk joined10:38
→ guide joined10:38
→ nb joined10:38
→ tuxx- joined10:38
→ s0undt3ch_ joined10:38
→ ronny joined10:38
→ SJS joined10:38
→ GodEater_ joined10:38
→ Bombe joined10:38
→ elyob joined10:38
→ ben_ir joined10:38
→ Lisimba joined10:38
→ milli joined10:38
→ nazgul101 joined10:38
→ rue joined10:38
→ dwmw2 joined10:38
→ segher joined10:38
→ makz- joined10:38
→ lamont joined10:38
→ alexs joined10:38
→ Tianon joined10:38
→ texel joined10:38
→ Silex joined10:38
→ Kindari joined10:38
→ Cerise joined10:38
→ masterkorp joined10:38
→ cmyers joined10:38
→ SethRobertson joined10:38
→ jdav_gone joined10:38
→ hsb joined10:38
→ bottiger joined10:38
→ hohum joined10:38
→ JP| joined10:38
→ Zenopus joined10:38
→ chrisf joined10:38
→ ayust joined10:38
→ ramsey joined10:38
→ kloeri joined10:38
→ wcarss_ joined10:38
→ Nibato joined10:38
→ joelpet joined10:38
→ BearPerson joined10:38
→ Ingmar joined10:38
→ imarcusthis joined10:38
→ kanru joined10:38
→ Zygo joined10:38
→ iivvoo joined10:38
→ Kicer86 joined10:38
→ Stravy joined10:38
→ mw44118 joined10:38
→ jakob joined10:38
→ gabiruh joined10:38
→ speakman joined10:38
→ clever joined10:38
→ jhford joined10:38
→ heidar joined10:38
→ esc joined10:38
→ NfNitLoop joined10:38
→ d1b joined10:38
→ Lgb joined10:38
→ felipe joined10:38
→ wanders joined10:38
→ fr joined10:38
→ avtobiff joined10:38
→ przemoc joined10:38
→ soma` joined10:38
→ eli joined10:38
→ tazle_ joined10:38
→ antix_ joined10:38
→ bwidmer joined10:38
→ Ratler joined10:38
→ bluesceada joined10:38
→ sp3000 joined10:38
→ jjore joined10:38
→ up_the_irons joined10:38
→ Zeelot joined10:38
→ haarg joined10:38
→ ezkl joined10:38
→ Pieter joined10:38
→ metze joined10:38
→ Utoxin joined10:38
→ Fiouz joined10:38
→ liberum joined10:38
→ _aeris_ joined10:38
→ dsop joined10:38
→ heinrich5991 joined10:38
→ bartman joined10:38
→ zomg joined10:38
→ plopix joined10:38
← phretor_ left10:38
← marinosi left10:38
← dr0id left10:38
← Phylock left10:38
← demods left10:38
← Orbitrix left10:38
← frogonwheels left10:38
← tsyd left10:38
← farktronix left10:38
→ phretor_ joined10:38
→ marinosi joined10:38
→ dr0id joined10:38
→ Phylock joined10:38
→ demods joined10:38
→ Orbitrix joined10:38
→ frogonwheels joined10:38
→ tsyd joined10:38
→ farktronix joined10:38
→ fahadsadah joined10:38
→ yeryry joined10:38
→ gerard0 joined10:38
→ Drakevr joined10:38
→ bond` joined10:38
→ koki joined10:38
→ Davey joined10:38
→ cbz joined10:38
→ Nickeeh joined10:38
→ MacGyver1L joined10:38
→ EugeneKay joined10:38
→ jspiros joined10:38
→ tansell joined10:38
→ prs joined10:38
→ malesca joined10:38
→ ping-pong joined10:38
→ tinmrn joined10:38
→ j0ran joined10:38
→ IcePic joined10:38
→ flamingspinach joined10:38
→ durarara joined10:38
→ jamxNL joined10:38
→ JonasF joined10:38
→ skoop joined10:38
→ acidjazz joined10:38
→ Kiall joined10:38
→ Utkarsh joined10:38
→ electrogeek joined10:38
→ lu_zero joined10:38
→ tomaw joined10:38
→ aperlstein joined10:38
→ osfameron joined10:38
→ myme joined10:38
hubbard.freenode.net set mode: +v10:38
→ flavius joined10:38
← MUILTFN left10:38
← nyuszika7h left10:38
← pen_ left10:38
← rendar left10:38
← radarek left10:38
← thiago left10:38
← ricky left10:38
← cilly left10:38
← Vampire0 left10:38
← squiggy left10:38
← cypha left10:38
← nd___ left10:38
→ MUILTFN joined10:39
→ Yuuhi joined10:39
→ pen_ joined10:39
→ rendar joined10:39
→ radarek joined10:39
→ thiago joined10:39
→ ricky joined10:39
→ cilly joined10:39
→ Vampire0 joined10:39
→ squiggy joined10:39
→ cypha joined10:39
→ nd___ joined10:39
→ poppiez joined10:39
→ ianmcorvidae joined10:39
→ hyperair joined10:39
→ Kaimei_ joined10:39
→ paissad joined10:39
→ toobluesc joined10:39
→ andres joined10:39
→ mtkd joined10:39
← simenbrekken left10:39
← drizzd left10:39
← vmil86 left10:39
← Marquel left10:39
← kmc left10:39
→ simenbrekken joined10:39
→ drizzd joined10:39
→ vmil86 joined10:39
→ Marquel joined10:39
→ selckin joined10:39
→ frerich2 joined10:39
→ compengi joined10:39
→ mmattice joined10:39
→ kanzure joined10:39
→ Orphis joined10:39
→ Soliton joined10:39
→ bentkus joined10:39
→ yorick joined10:39
→ fluido joined10:39
→ egs joined10:39
→ linuxNewb joined10:39
→ wolog joined10:39
→ phlebas joined10:39
→ deavidsedice joined10:39
→ elhadr joined10:39
→ SubZero_ joined10:39
→ biomilch joined10:39
→ markatto joined10:39
→ kokx joined10:39
→ Blackshark joined10:39
→ architekt joined10:39
→ biz joined10:39
← Anubisss left10:39
← Weiss left10:39
← bgerber left10:39
← peper left10:39
← Slush- left10:39
← preflex left10:39
← redii left10:39
← zemo left10:39
← flaguy48 left10:39
← JEEB left10:39
← markatto left10:40
← Agos left10:40
← pizza_ left10:40
← SRabbelier left10:40
← moshef left10:40
← cytrinox left10:40
← Samoi left10:40
← lanthan_ left10:40
← d4rk1ink left10:40
→ Anubisss joined10:40
→ Weiss joined10:40
→ bgerber joined10:40
→ peper joined10:40
→ Slush- joined10:40
→ preflex joined10:40
→ redii joined10:40
→ zemo joined10:40
→ flaguy48 joined10:40
→ JEEB joined10:40
→ else joined10:40
→ friskd joined10:40
→ sitaktif joined10:40
→ floppyears joined10:40
→ f0i joined10:40
→ Bucciarati joined10:40
→ pax- joined10:40
→ makubi_ joined10:40
→ eightyeight joined10:40
→ intchanter joined10:40
→ andyl joined10:40
→ Martz joined10:40
→ milki joined10:40
→ nick_h_ joined10:40
→ coldboot joined10:40
→ fairuz joined10:40
→ adri joined10:40
→ wagle joined10:40
→ shachaf joined10:40
→ cwillu_at_work joined10:40
→ bremner joined10:40
→ rhl joined10:40
→ piti joined10:40
→ shad joined10:40
→ numbah joined10:40
→ awx joined10:40
→ bburhans joined10:40
→ Agos joined10:40
→ pizza_ joined10:40
→ SRabbelier joined10:40
→ moshef joined10:40
→ cytrinox joined10:40
→ Samoi joined10:40
→ lanthan_ joined10:40
→ d4rk1ink joined10:40
→ sivy joined10:40
→ alex__c2022 joined10:40
→ rfx_ joined10:40
→ sordina2 joined10:40
→ chde joined10:40
→ tro joined10:40
→ surfdue joined10:40
→ Gunni joined10:40
→ Rejinka joined10:40
hubbard.freenode.net set mode: +v10:40
← tobago left10:40
← squeese left10:40
← linusoleander left10:40
← LotR left10:40
← tewecske left10:40
← aminpy left10:40
← fission6 left10:40
← Pacman_Home left10:40
← lucas left10:40
→ nyuszika7h joined10:40
← Gunni left10:41
← kumbayo left10:41
← orafu left10:41
← lisak left10:41
← Chaser left10:41
→ markatto joined10:41
→ kumbayo joined10:42
→ lisak joined10:42
→ orafu joined10:42
→ Chaser joined10:42
→ ducch joined10:42
→ mpenzi joined10:42
→ xxthink joined10:42
→ sbell joined10:42
→ seanius joined10:42
→ tyteen4a03 joined10:42
→ kelsin joined10:42
→ ndim joined10:42
→ ruskie joined10:42
→ Sebastian_ joined10:42
→ unmanbearpig joined10:42
→ Grum joined10:42
→ xrogaan joined10:42
→ MichealH joined10:42
→ qwertz_ joined10:42
→ rudi_s joined10:42
→ an0maly_ joined10:42
→ stockholm joined10:42
→ twem2 joined10:42
→ s0ber joined10:42
→ Orangy joined10:42
→ logankoester joined10:42
→ _W_ joined10:42
→ snafoo_ joined10:42
→ csgeek joined10:42
→ ilmari joined10:42
→ Ksg89 joined10:42
→ dotCOMmie joined10:42
→ GunArm joined10:42
→ tic joined10:42
→ meder joined10:42
→ cehteh joined10:42
→ Will| joined10:42
→ bjornbjorn joined10:42
→ bittin joined10:42
→ deni joined10:42
→ dhoss joined10:42
→ nevyn joined10:42
→ dErFuX joined10:42
→ tychoish joined10:42
→ vuliev joined10:42
→ helgikrs joined10:42
→ ajpiano joined10:42
hubbard.freenode.net set mode: +vv10:42
→ murr4y joined10:42
→ nimrod10 joined10:42
→ Caius joined10:42
→ hal1on joined10:42
→ bitslip joined10:42
→ bball- joined10:42
→ dahl joined10:42
→ callumacrae joined10:42
→ Juri_Agarin joined10:42
→ Gunni joined10:42
→ cjuner joined10:42
→ tobago joined10:42
→ squeese joined10:42
→ linusoleander joined10:42
→ LotR joined10:42
→ tewecske joined10:42
→ aminpy joined10:42
→ fission6 joined10:42
→ Pacman_Home joined10:42
→ atdiehm_home joined10:42
→ peterwald joined10:42
→ c00kiemon5ter joined10:42
→ aboudreault joined10:42
→ ericdc joined10:42
→ alip joined10:42
→ Gizmokid2005|AFK joined10:42
→ jstemmer joined10:42
→ cats4gold joined10:42
→ aLeSD joined10:42
→ divVerent joined10:42
→ mirage335 joined10:42
→ Cap_J_L_Picard joined10:42
→ blop joined10:42
→ dhruvasagar joined10:42
→ raydeo joined10:42
→ lucas joined10:42
→ sven^_ joined10:42
→ BinGOs joined10:42
→ EdwardIII joined10:42
→ mquin joined10:42
→ demi` joined10:42
→ Confiscative joined10:42
→ kmc joined10:42
→ TrineLocke joined10:42
← Gunni left10:42
→ Gunni joined10:42
← alip left10:42
← kpreid left10:43
← Khisanth left10:43
→ Khisanth joined10:43
← stepnem left10:43
← antix_ left10:44
→ alip joined10:44
→ Tabmow joined10:44
← jargon- left10:45
← apocalyptiq left10:45
← davzie left10:45
← churp left10:45
← pantsman left10:45
← arunce left10:45
← Fudge left10:45
← mainrain left10:45
← dyek left10:45
← bburhans left10:45
→ jargon- joined10:45
→ apocalyptiq joined10:45
→ davzie joined10:45
→ churp joined10:45
→ pantsman joined10:45
→ arunce joined10:45
→ Fudge joined10:45
→ mainrain joined10:45
→ dyek joined10:45
→ __iron joined10:45
→ unreal joined10:45
→ sonnym joined10:45
→ sejo joined10:45
→ bronson joined10:45
→ phantomcircuit joined10:45
→ pygmalion joined10:45
→ ferrousw1eel joined10:45
→ kimico joined10:45
→ chrizz_ joined10:45
→ Samual joined10:45
→ toofar joined10:45
→ yaotti_aw joined10:45
→ pielgrzym joined10:45
→ bburhans joined10:45
← sejo left10:45
→ antix_ joined10:45
→ sejo joined10:46
← Nuck left10:46
→ phretor joined10:47
← Vile left10:47
← kloeri left10:47
→ kloeri joined10:47
jargon- how can i view the diffs between the most recent commit and a currently uncommitted edit?10:48
on linux debian10:48
→ kpreid joined10:48
nevyn git diff10:48
→ casbah joined10:49
casbah hi, how do i check what my git --global values are?10:49
← phretor_ left10:50
nevyn git config --global --list10:51
jargon- nevyn: i'm new with this,could u walk me thru it? i'm basically trying to open them side by side like with vimdiff and see what changes i've made10:51
← jceb_ left10:51
→ jceb joined10:51
→ frantk joined10:52
nevyn jargon-: so the output of that is a unified diff10:52
A is what you have now. B is what is in HEAD (your last commit)10:52
→ saste joined10:55
jargon- nevyn: it just says they differ,but it doesn't open both files to show the diffs10:55
← ThomasLocke left10:55
jargon- nevyn: aah wait. i found it. it's git diff HEAD10:56
thanks :)10:56
→ MissionCritical joined10:57
← TrineLocke left10:57
casbah is it necassary for our github email to match the email in our github globals?10:59
thanks btw nevyn10:59
nevyn casbah: umm dunno.11:00
casbah actually it doesn't apparently11:00
your github.token in the globals just has to match11:01
sweet11:01
and your rsa key of course11:01
thanks again nevyn11:01
selckin noting needs to match11:01
casbah how do branches work in git? what are they used for?11:01
selckin if github knows your ssh keys you're all good11:01
casbah are branches just used for branching out different ideas and stuff? and different contributors to a project?11:02
which then get merged into the 'official' branch like svn called 'master'?11:02
or am i wrong/11:02
selckin branches are the same concept in svn or git or any other vcs11:03
→ OOPMan joined11:04
→ Vile joined11:04
→ arnoldB joined11:05
nevyn except they're unusable and avoided in most vcs's11:05
and standard and the prefered way of working in git11:05
← arnoldB left11:06
casbah i haven't used branches ever before in any vcs11:06
new to git11:06
used svn for one project once11:06
let's say i'm the only person working on a project in git11:06
nevyn yep11:06
casbah would it be necassary even or recommended for me to use branches? why/when?11:07
nevyn so necessary... no.11:07
reccomended ys11:07
cbreak use branches when ever you want to make something new that has the possibility of taking more than a few hours to program, or will be more than one commit11:10
obviously, you should commit every few hours, even if it's just an amend or a temporary commit, or a combination thereof11:10
casbah ok, gotcha cbreak, thanks11:10
nevyn use branches whenever you're writing something that takes more than 10 minutes to write11:11
casbah i generally commit every ~30 min or so, and try to do only enough that would fit in a commit message11:11
i haven't worked on a super intensive stuff yet though, just in CS 101 right now11:11
nevyn making a branch in git is really cheap.11:12
cbreak I commit a few times per day, depending on what I do11:13
at least once, at most 10 or 20 times though11:13
nevyn commits a lot more that that on his local tree.11:13
cbreak and if I commit a lot, many of those are amends to temporaries11:13
nevyn that's what rebase -i is for11:13
cbreak I don't commit if it doesn't compile11:14
with some exceptions11:14
like not having committed for an hour or so11:14
but every finalized commit must compile11:14
nevyn works on production interpreted code that's rolling release.11:14
cbreak this rule allows me to do bisect testing11:15
so, do you commit stuff that has syntax errors? :)11:15
or runtime errors?11:15
nevyn nah.11:15
cbreak last week I did something I didn't do before: Write a camera driver API binding for a camera driver that I did not have11:16
so I could not compile at all11:16
I could just write code according to documentation11:16
took three days, and I still don't know if it even works11:16
that was one of the first times when I committed untested code11:16
→ Pacman` joined11:18
jargon- hmm. how do i edit a commit message,when i forgot to add some info?11:18
← chrisf_ left11:19
casbah !help11:19
jast *automatic message* the whole extent of documentation on jast's bot features is the script's source code. You can always view the current version here: http://j.mp/gitircsrc11:19
casbah !rebase11:19
← Pacman_Home left11:19
casbah !git-rebase11:19
!manpages git-rebase11:20
← casbah left11:20
→ frhodes joined11:24
← linusoleander left11:27
cbreak have you considered using google?11:27
← phretor left11:28
→ jelmer joined11:28
← lanthan_ left11:28
← Anubisss left11:29
← tmpvar left11:29
tango_ jargon-: commit --amend11:30
jargon- tango_: do i rewrite the whole comment or just the bit that needs to be added?11:31
← Murr_ left11:31
cbreak a commit is atomic11:32
tango_ an editor pops up11:32
cbreak you can not change just a small bit11:32
you can only change it completely11:32
← frhodes left11:32
→ frhodes joined11:32
→ phretor joined11:32
→ gebi joined11:32
→ n1ck3 joined11:32
→ eddiefullmetal joined11:34
jargon- tango_: cbreak: thank u11:35
← radarek left11:35
← babilen left11:38
← Tabmow left11:41
→ Tabmow joined11:41
← DrNick left11:42
→ tansell-laptop joined11:42
→ eletuchy joined11:44
→ berserkr joined11:44
→ DofLeFou joined11:45
← eletuchy left11:45
→ SeySayux joined11:45
← Spockz left11:48
→ radarek joined11:49
→ rchavik joined11:52
→ B|aster joined11:52
← n1ck3 left11:52
→ Spockz joined11:53
→ himsin joined11:54
← pantsman left11:56
← frhodes left11:58
→ linusoleander joined11:59
← linusoleander left11:59
→ gry joined12:03
gry https://dev.freenode.net/redmine/projects/eir/repository <-- how do I clone that? (Am I missing something, or some redmine default perhaps?)12:04
← radarek left12:04
cbreak gry: git clone repourl12:05
gry k12:05
← gry left12:05
← bonhoffer left12:07
← pizza_ left12:07
→ mrsrikanth joined12:08
← Khisanth left12:09
SeySayux Does git ignore .exe files automatically?12:10
cbreak no12:10
the only thing git ignores automatic are .git folders12:11
SeySayux Hmm, I guess it's the eclipse git plugin12:11
← aminpy left12:11
SeySayux command line git does it just fine12:11
cbreak in general it's stupid to commit .exe files anyway12:11
same as any build product12:12
so ignoring them is a good ting12:12
→ aminpy joined12:12
SeySayux The exe is part of the build process, so you can't really start the build without it...12:12
(Well, you could, but you'd have to compile it manually)12:12
← GunArm left12:13
← dSebastien left12:13
→ GunArm joined12:14
→ dSebastien joined12:14
_ikke_ It still should not be part of a repository12:14
cbreak then don't commit it12:14
commiting build products bloat up the repository and make merges much more complicated12:15
SeySayux Why not?12:15
cbreak aren't those not enough? :)12:15
SeySayux Erm, it's not a build product, it's a build *tool*.12:15
cbreak it's a build product12:15
if it can be built12:15
SeySayux It can be compiled. I mean, next you're going to tell me I can't put shell scripts in my repo.12:16
cbreak in general, I don't like committing libraries and toolchains either12:16
sure you can12:16
but you would not commit makefiles if you use automake12:16
or visual studio projects if you use qmake12:16
SeySayux No, of course not12:16
cbreak because those are all intermediate results12:16
← arunce left12:17
cbreak in general, only neccessary things should be committed, build products, even if intermediate such as a Makefile, should not12:17
SeySayux It's just something that gets launched while cmake runs.12:17
On *nix, a shell script gets launched.12:17
cbreak then cmake should build it12:17
if it's buildable12:17
SeySayux The problem is that we need it before we can determine if there is a suitable compiler...12:18
cbreak you need an exe for that?12:18
doesn't cmake have facilities to find compilers?12:18
that thing is worse than I thought12:18
SeySayux Yea, but I put high demands on what my compiler does12:19
cbreak anyway, I recommend you to not commit it :)12:19
→ radarek joined12:19
cbreak you can just do it anyway and suffer the consequences12:19
→ StupidMop joined12:19
← SillyWabit left12:19
SeySayux And what are the consequences? It's just a binary file that sits there and isn't ever going to change12:19
I could just ask all Windows users to install a real shell, though.12:20
← dnyy left12:21
cbreak "commiting build products bloat up the repository and make merges much more complicated"12:21
if it never changes, the bloating up is a fixed value12:22
→ isxek joined12:22
cbreak and merges won't be affected much.12:22
→ ereslibre joined12:22
← ereslibre left12:22
→ ereslibre joined12:22
cbreak but be aware that git stores a full copy of every single file you ever committed12:22
kmc i'm a fan of committing build products in a separate repository when you ship them12:22
cbreak so if you update it, you double the required storage12:22
kmc: A coleague of mine worked for nvidia, and they committed their whole compiler!12:23
they weren't using git though :)12:23
→ Khisanth joined12:23
→ bonhoffer joined12:23
→ w0bni joined12:24
SeySayux 24 kb is not something I call major bloat.12:26
My makefile is larger than that12:27
(Cmakefile, that is)12:27
And I commit that, don't i :P12:27
← rchavik left12:28
→ aristidesfl joined12:30
cbreak human created text files are easily delta compressible12:30
← bonhoffer left12:30
cbreak build products usually are not12:30
that's why I said what I said above12:30
SeySayux I don't usually commit binaries either.12:31
But I don't see much difference between this and, say, a 24 kb png file.12:31
And I think it's easier just to commit this binary rather than a) add another buildstep people have to run or b) ask that everyone installs cygwin/msys that wants to build my project.12:33
→ frhodes joined12:33
patrikf sounds like something to include in a release tarball but not in the repo12:34
← Tabmow left12:34
→ Tabmow joined12:34
DPav indeed, what patrikf said12:34
← aminpy left12:34
← aristidesfl left12:34
SeySayux never mind... *sigh*12:36
← himsin left12:36
← phretor left12:37
← DPav left12:39
← tobago left12:39
→ heftig joined12:39
→ phretor joined12:39
← moshef left12:40
StupidMop I'm using git and gitolite. I have a project i've been working on locally that doesn't have a repo yet. How do I add that to the server so others can push/pull/contribute?12:41
→ jason237 joined12:42
m1sc StupidMop: just create a new repo as described in the gitolite docs, add it as remote locally, push12:42
→ sikao_lfs1 joined12:43
StupidMop m1sc: It's that middle step. How do i add it as a remote locally?12:45
m1sc StupidMop: man git remote12:46
jast StupidMop: *automatic message* the 'git-remote' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-remote.html12:46
StupidMop Excellent! Thanks for the point in the direction m1sc :-)12:47
→ braiam joined12:49
braiam there is easy way to export a git changelog into a debian one????12:49
ndim braiam: The Debian people probably have some script to do that12:51
cbreak git doesn't have a changelog12:51
braiam I was searching, but I only found about Debian-native packages, ndim12:52
→ bonhoffer joined12:53
→ robottinosino joined12:53
ndim braiam: I thought they now supported git for maintaining their packages, and I think I have seen something to keep git commit logs and debian/changelog like files in sync somehow. But I might remember incorrectly.12:56
→ kukks joined12:56
→ frhodes_ joined12:58
braiam ndim: yeah, but they do in the way that i don't need, migrating from a deb package to a git repo12:58
I want to do the opposite, a git repo from the upstream (without a deb native package or scrips) to a deb packages12:59
→ pantsman joined13:01
← pantsman left13:01
→ pantsman joined13:01
← frhodes_ left13:01
← robottinosino left13:02
→ d3vic3 joined13:03
→ robottinosino joined13:03
→ spaceonline joined13:04
→ canton7 joined13:04
← frhodes left13:05
bremner braiam: you want the command "git-dch" and the channel #debian-mentors on irc.oftc.net13:06
← bonhoffer left13:07
← mlncn left13:08
→ bonhoffer joined13:10
← d3vic3 left13:13
→ d3vic3 joined13:13
→ JakeyChan joined13:13
← maleknet left13:14
← pantsman left13:14
→ pantsman joined13:15
→ mlncn joined13:15
braiam bremner: i've tried that, but I just get that I don't have the changelog, bla, bla, bla13:16
← canton7 left13:20
bremner I doubt anyone can help without the exact error, message, and in any case it isn't a git issue per se, but a debian packaging issue, hence my suggestion of the other channel.13:20
← tyteen4a03 left13:22
→ angerman joined13:27
angerman ok, I have a git repository. And I'd like to create a new --bare repository in my Dropbox as remote. How would I go about doing that?13:29
← kpreid left13:30
← envi left13:30
← Gitzilla left13:31
→ kmc_ joined13:33
→ hakunin joined13:33
→ kpreid joined13:34
cmn angerman: you do what man git init says (git init --bare) and then add it as a remote like man git remote says (git add remote some-name /some/path)13:34
jast angerman: *automatic message* the 'git-init' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-init.html13:34
← JakeyChan left13:35
← kmc left13:35
← kpreid left13:35
→ kpreid joined13:36
angerman cmn: thanks13:36
→ fread2281 joined13:37
→ paul_andrew joined13:38
← jceb left13:39
← fread2281 left13:39
← braiam left13:40
← chrizz_ left13:44
→ dcarper joined13:46
← robottinosino left13:46
← jhoepken left13:48
← deni left13:49
→ deni joined13:49
← apocalyptiq left13:50
→ jhoepken joined13:50
→ dv310p3r joined13:51
← nyuszika7h left13:51
→ posciak joined13:52
→ cexsum joined13:54
→ eijk_ joined13:54
→ frhodes joined13:55
→ chrizz joined13:56
→ ISF joined13:57
← eijk left13:57
→ jamesw joined13:58
→ mtbre joined13:58
ChanServ set mode: +v13:58
jamesw can anyone help me with a messed up repo?13:59
did a git pull this morning with no problems13:59
made some changes and went to add them13:59
and I got "fatal: index file is smaller than expected"13:59
the only advice I could find on that was to manually remove the index file so I moved it out of the .git folder13:59
now when I try to commit I get messages like "fatal: cannot lock HEAD ref"14:00
→ Adaptee joined14:00
jamesw and "error: unable to resolve reference HEAD: no such file or directory"14:00
← ISF left14:00
→ ISF joined14:00
jamesw and something about a "loose object LONGHASHHERE (stored in .git/objects/blah/blah) is corrupt"14:01
→ Cromulent joined14:01
→ Andrew_ joined14:02
→ ben_alman joined14:02
← OOPMan left14:03
→ __sorin__ joined14:03
SimonNa jamesw: corrupt objects? You should make a manual backup copy of your working directory before proceeding. I don't know the best idea from here though. Maybe recreating the file HEAD will alraedy work.14:04
echo "ref: refs/heads/master" > .git/HEAD14:05
→ asakura joined14:05
← kmc_ left14:06
→ Glycan joined14:06
Glycan Hello14:06
How do you remove all files that match .gitignore?14:07
→ madewokherd joined14:07
rudi_s jamesw: git fsck --full --strict might also give you more information (though I have no experience with corrupted git repos)14:07
→ kmc_ joined14:07
← jutaro left14:07
rudi_s Glycan: git clean -x should work, try it with git -clean -nx to see what is going to be removed.14:07
→ kenperkins joined14:07
← cypha left14:09
Glycan Uh... doesn't work.14:09
Oh14:09
Bad gitignore14:09
rudi_s Glycan: Ah, use -X.14:09
-x removes also untracked but not ignored files.14:09
Glycan I know14:09
I noticed14:10
→ cypha joined14:10
Glycan What would remove all local untracked changes and sync with the last remote commit?14:10
← jamesw left14:10
rudi_s Glycan: git reset --hard origin/master (if the remote branch is master); destroys all local changes!14:11
Glycan Yes, that's the whole idea14:11
rudi_s (And removes local commits!)14:11
← thews left14:11
Glycan Um14:11
I guess14:12
How do you check where you are compared to remote repo?14:12
Like, ahead by one commit or something14:12
rudi_s Glycan: To remove only uncommitted changes: git reset; git checkout .14:12
Glycan git checkout . inculuding the .?14:12
rudi_s Glycan: I normally just use tig/gitk to have a look at the commits.14:13
Yes.14:13
(tig/gitk --all)14:13
Glycan gitk, huh14:13
← kenperkins left14:13
Glycan GUIs scare me. :p14:14
rudi_s Glycan: Then use tig, no gui.14:14
SimonNa "git status" outputs a single line telling you how far you're ahead.14:14
Glycan I kid14:14
← dcarper left14:14
← kpreid left14:15
→ p3rsist joined14:15
→ dcarper joined14:16
→ kpreid joined14:16
← dcarper left14:16
← kpreid left14:16
← xiangfu left14:16
→ kpreid joined14:16
→ roop joined14:17
← sikao_lfs1 left14:17
Glycan How do you remove a head?14:17
← kpreid left14:18
→ kpreid joined14:18
← kpreid left14:18
rudi_s Glycan: You mean a branch? HEAD is just the currently checkout branch.14:19
→ kpreid joined14:19
rudi_s (s/branch/commit/)14:19
← pen_ left14:19
Glycan Not HEAD, head14:19
A pointer to a commit14:20
But branch is another word for that14:20
Oh, and clean didn't work14:20
rudi_s Glycan: AFAIK the only pointers are branches and tags.14:20
"didn't work"?14:20
→ nyuszika7h joined14:21
← tewecske left14:21
← dannas left14:21
→ keller_ joined14:21
Glycan clean -Xn = no output14:22
clean -X = clean failed becouse of.. indecsion about dryclean/force?14:22
kmc_ → kmc14:22
rudi_s Glycan: To be honest I've never used -X but clean -Xf should work.14:25
← hakunin left14:26
→ hakunin joined14:27
← dr0id left14:27
← kpreid left14:28
→ kpreid joined14:28
← radarek left14:30
→ ChrisLE joined14:31
← roop left14:31
← julius_ left14:32
← mrsrikanth left14:34
→ xiangfu joined14:35
← seivan left14:35
→ gusnan joined14:37
→ dannas joined14:38
→ madsy joined14:38
← madsy left14:38
→ madsy joined14:38
← frhodes left14:39
→ tyteen4a03 joined14:43
← p3rsist left14:44
← paul_andrew left14:48
→ mklappstuhl joined14:49
→ troyt joined14:51
← nyuszika7h left14:53
← EricInBNE left14:54
← linuxNewb left14:54
← eijk_ left14:55
→ nyuszika7h joined14:55
→ a_suenami joined14:55
← cyphase left14:58
→ texel` joined15:02
← texel left15:03
← texel` left15:03
→ dcarper joined15:05
→ linuxNewb joined15:06
← flaguy48 left15:07
→ phretor_ joined15:07
← MUILTFN left15:08
→ frsk joined15:08
← keller_ left15:10
← phretor left15:10
lisak is it possible to create a patch with binary files ? so that after applying that patch the binary files will be there?15:11
or something like that15:11
→ radarek joined15:11
bremner yes, git-format-patch should do that. But the patches need git to apply.15:11
→ tewecske joined15:11
← Error404NotFound left15:11
→ Murr_ joined15:11
lisak so that usual git format-patch master --stdout > fix_empty_poster.patch15:12
git apply --stat fix_empty_poster.patch15:12
→ clifton joined15:12
bremner well, you could alway check man git-format-patch for some binary option15:13
jast well: *automatic message* the 'git-format-patch' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-format-patch.html15:13
← clifton left15:14
→ VladGh joined15:14
→ steinrs joined15:14
steinrs Hi - got a nasty problem here where a mass rename dodges the git heuristics - how can i force git into groking i've been renaming files, or NOT been renaming files?15:14
→ herdingcat joined15:14
← ChrisLE left15:14
← dv310p3r left15:15
→ jutaro joined15:16
Xeross is there any way to do a 'git commit --amend' whilst skipping editing the commit message?15:18
← ISF left15:18
bremner probably with -C ?15:19
→ ISF joined15:20
lisak bremner, if I use --binary option together with --stdout15:20
it doesn't do anything else15:20
if I do git format-patch master --binary > ~/fix_empty_poster3.patch15:20
the patch contains a few lines like 0001-custom-maven-deploy-changes.patch 0002-Modeshape-integration.patch15:21
Xeross bremner: Thanks, is working ^w^15:21
lisak don't know15:21
← xiangfu left15:22
→ tobias__ joined15:22
→ p3rsist joined15:25
← StupidMop left15:25
← __iron left15:26
→ bel_airman joined15:26
← ben_alman left15:28
→ flaguy48 joined15:28
→ zamabe joined15:28
→ roop joined15:28
→ tatsuya joined15:29
→ drev1 joined15:30
← bel_airman left15:31
bremner lisak: that isn't the patch15:31
look at those files15:31
→ babilen joined15:31
→ dv310p3r joined15:31
lisak yup the destination was .15:31
instead of the output destination15:31
→ fread2281 joined15:32
← Murr_ left15:32
← roop left15:33
lisak then I git apply --stat fix_empty_poster3.patch15:34
← fread2281 left15:34
→ bonhoffer_ joined15:34
lisak having all those 000* files there15:34
in the same dir15:35
→ rvsjoen_ joined15:36
← bonhoffer left15:36
bonhoffer_ → bonhoffer15:36
← Internet13 left15:37
→ Murr_ joined15:37
← zamabe left15:37
hohum is it possible to take a git branch and import it (including all of the history) into a new bare repo15:37
IE, I want to fork a specific branch into a brand new project15:37
← rvsjoen left15:38
PerlJam hohum: why just the branch and not all history?15:38
→ AaronMT joined15:38
selckin just add it as a remote, fetch it, and branch like normal15:39
← lisak left15:39
← ereslibre left15:39
→ Internet13 joined15:40
← Internet13 left15:41
hohum PerlJam: because of the way trac has historically worked I've got a single repo for all my projects. Each project has its own "master" branch.15:42
one in particular needs a dedicated trac instance15:42
← linuxNewb left15:43
→ Internet13 joined15:43
→ cylence joined15:44
cbreak hohum: just git remote add the repository15:44
and git fetch just that branch15:44
hohum coolio15:44
thanks15:44
cbreak or do it the other way15:45
add the bare as a remote from your source15:45
and just git push that single branch15:45
(that's probably easier)15:45
← jelmer left15:46
→ Raging_Hog joined15:47
→ seivan joined15:47
→ mlncn-agaric joined15:48
← redii left15:49
← dcarper left15:49
← mlncn left15:50
→ slant joined15:50
→ benizi joined15:51
← d3vic3 left15:52
← bonhoffer left15:53
← cylence left15:53
← herdingcat left15:54
→ ThomasLocke joined15:55
→ jogla joined15:55
→ linuxNewb joined15:56
← benizi left15:57
→ aminpy joined15:57
← aminpy left15:58
→ aminpy joined15:58
→ herdingcat joined15:59
← B|aster left16:00
← slant left16:00
→ cylence joined16:01
→ bonhoffer joined16:01
← isxek left16:01
→ TrineLocke joined16:02
← shruggar left16:02
→ grekko joined16:02
→ OOPMan joined16:05
→ bonhoffer_ joined16:06
← cylence left16:07
→ zamabe joined16:07
← tewecske left16:07
← bonhoffer left16:08
bonhoffer_ → bonhoffer16:08
→ cylence joined16:08
→ tewecske joined16:09
→ soulcake joined16:10
→ kpreid_ joined16:10
← kpreid_ left16:11
→ nd__ joined16:11
→ SirFunk_ joined16:11
← kpreid left16:11
→ ceej joined16:12
← tewecske left16:13
steinrs [reask]Hi - got a nasty problem here where a mass rename dodges the git heuristics - how can i force git into groking i've been renaming files, or NOT been renaming files?16:13
→ benizi joined16:14
← nd___ left16:15
→ tewecske joined16:15
cbreak steinrs: just ignore it16:15
steinrs: or read man git-log16:15
jast steinrs: *automatic message* the 'git-log' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-log.html16:15
bremner or, git config merge.renameLimit 1000016:15
← tewecske left16:15
bremner for merging.16:15
steinrs cbreak: i can't ignore it - got non-renames that are taken as renames16:16
cbreak ignore it anyway16:16
→ jonnyATroot joined16:16
cbreak it's just a visual problem16:16
steinrs uh, no16:17
cbreak you can tell git log to not detect renames16:17
then they won't show up anymore16:17
it's just visual16:17
steinrs this is a merging problem - got another branch where the files are recorded as del/add16:18
cbreak git always records files as delete/add16:19
git does not record renames16:19
it always detects them when it needs them16:20
for example, when looking at the history with git log16:20
bremner and when doing a merge16:20
← aboudreault left16:21
→ tewecske joined16:21
→ muneeb joined16:22
← Cromulent left16:23
← AaronMT left16:26
steinrs git is being a git16:27
← tatsuya left16:28
→ fread2281 joined16:28
← Chib left16:28
→ xmfbs joined16:30
← seivan left16:30
→ slant joined16:36
→ corni joined16:38
← cylence left16:38
→ shruggar joined16:40
← fread2281 left16:41
→ stickyboy joined16:42
stickyboy I have some files I want to move, but also edit. Is there any way I can do it in one commit so it's as "atomic" as possible?16:43
bremner sure. just move them. then edit, then commit16:44
eeanm → eean16:45
→ hakunin_ joined16:45
→ kerrick joined16:45
← herdingcat left16:45
→ SebastianFlyte joined16:46
nyuszika7h → eir16:46
→ jablko joined16:47
stickyboy bremner: Ok... so git will follow the file.16:47
← eir left16:47
stickyboy Lemme try.16:47
← Murr_ left16:47
cbreak stickyboy: depends16:47
the automatic rename detection is not perfect16:48
← hakunin left16:48
→ DrNick2 joined16:48
SimonNa I have unpushed linear history in master, and off master, there's several other branches. I want to squash some commits that lead to master, and it shall affect all the numerous branches. Is there a way to do it without "rebase --onto newmaster oldmaster somebranch" for each somebranch?16:49
→ Pupeno joined16:49
→ nyuszika7h joined16:49
→ kumbayo_ joined16:49
SimonNa (i.e. automate this rebasing of the subbranches without using a shell loop)16:50
kmc i would use a shell loop for that16:51
SimonNa alright, thanks.16:51
← kumbayo left16:52
→ Renegade001 joined16:52
→ mandric joined16:53
Renegade001 I've been trying to setup a remote git repository, its /var/www to easily upload updates16:54
← stickyboy left16:54
Renegade001 but I can't seem to push to the repository16:54
bremner !fag non-bare16:54
err, faq non-bare16:54
jast err: *automatic message* Pushing to non-bare repositories is discouraged: please see https://git.wiki.kernel.org/index.php/GitFaq#non-bare16:54
→ stickyboy joined16:54
bremner Renegade001: ^16:54
→ Murr_ joined16:55
Renegade001 bremner: Sorry, I'm really quite new to Git16:56
How do I check if a git repository is bare or not?16:56
bremner Renegade001: !bare16:56
jast Renegade001: *automatic message* an explanation of bare and non-bare repositories can be found here: http://bare-vs-nonbare.gitrecipes.de/16:56
bremner finally, see http://toroid.org/ams/git-website-howto for one way to do it.16:56
or just use rsync to deploy16:57
→ lut4rp joined16:58
Renegade001 Thank you bremner16:58
← jutaro left16:59
→ ziggles joined17:04
→ Juo joined17:04
← jablko left17:06
→ jablko joined17:06
← Spockz left17:06
← SeySayux left17:07
→ towski joined17:07
→ Spockz joined17:08
← pantsman left17:10
→ mdpatrick joined17:11
mdpatrick What's the difference between "git log master..github/master" and "git log github/master ^master", if any?17:11
→ redii joined17:11
← angelsl left17:11
→ srji joined17:12
→ thews joined17:12
→ pantsman joined17:13
← pantsman left17:13
→ pantsman joined17:13
← slant left17:14
→ nadavoid joined17:15
→ DPav joined17:15
← p3rsist left17:17
← Spockz left17:18
→ adamm joined17:18
cbreak mdpatrick: don't think there's a difference in the semantics17:18
→ canton7 joined17:19
← Murr_ left17:20
← mandric left17:21
→ Spockz joined17:21
← cexsum left17:23
mdpatrick cbreak: So for all intents and purposes its going to show the same results?17:23
→ MaikB joined17:25
MaikB Hello. I have a question but I need to put some context around it to make sense, so here we go17:26
FauxFaux MaikB: Please use a pastebin or type it all on one line; multi-=line questions are so annoying.17:26
← zamabe left17:26
MaikB FauxFaux, good call17:26
I do a pastebin17:26
FauxFaux Aaaaaaaaaaaaaaa17:26
→ zamabe joined17:26
MaikB there it is: http://pastebin.com/MiM86Qyz17:27
Thanks FauxFaux ^^17:27
ChanServ set mode: +v17:28
→ sirdancealot joined17:28
xmfbs how can I, knowing a tag name, get the name of the guy that tagged it?17:28
← __sorin__ left17:28
bremner xmfbs: if it is an annotated tag, git show, otherwise, not possible17:29
FauxFaux git log --format="%an" -1 tagname17:29
→ jceb joined17:29
FauxFaux Oh, uh, never mind. ¬_¬17:29
xmfbs yeah, that only shows the guy who did the commit...17:30
bremner probably a lightweight tag17:30
← Renegade001 left17:30
xmfbs that was what i initially tried17:30
→ Murr_ joined17:30
→ Cromulent joined17:30
FauxFaux xmfbs: I second bremner's "not possible"; the information is not stored.17:31
xmfbs okay, thanks. so only annotated tags will work?17:31
← dannas left17:31
← steinrs left17:32
bremner in the sense of tracking tag author info, yes17:32
→ dcarper joined17:33
xmfbs okay17:33
mdpatrick Are there any differences between using "git log master..github/master" and "git log github/master ^master" to determine which commits haven't been merged in to my master branch or is this approximately going to give the exact same results?17:36
← Pupeno left17:36
cbreak how about using git merge-base?17:37
→ dannas joined17:38
← stickyboy left17:39
← ziggles left17:40
→ __sorin__ joined17:40
DrNick2 → DrNick17:44
→ balau joined17:46
→ urkud joined17:49
← mlncn-agaric left17:49
→ Belliash joined17:50
← zamabe left17:51
→ zamabe joined17:51
Belliash hi17:52
← srji left17:53
Belliash i have a repo and i have made some changes to it (added few files, few removed, few changed etc...). Now i dont want to commit them and i want to do 'git pull'. Whats the easiest way to revert all changes and proceed with 'git pull' w/o problems?17:54
→ Araxia__ joined17:54
mdpatrick Belliash: git reset --hard HEAD17:55
→ cyphase joined17:56
Belliash mdpatrick: i tried that but new files still exists in tree17:56
mdpatrick Belliash: --hard tells it to delete all subsequent commits after whatever commit SHA you give it, and then check it out and flush what's currently in the working directory. So in this case it pretty much means "flush the crap in my working directory" since you're specifying HEAD as the commit you're resetting to (which, of course, is the very last commit that you've got)17:56
→ surikator joined17:57
mdpatrick Belliash: Use git clean --force17:57
That's because they're not tracked17:57
If they're not tracked, reset won't touch them17:57
Belliash so i should use both these commands?17:57
← Araxia left17:58
← corni left17:58
← radarek left17:58
← jogla left18:00
→ jogla joined18:00
mdpatrick Belliash: yes18:01
reset will take your files back to the way they were in the last commit (all of the tracked ones), then git clean removes the untracked ones, Belliash... per the man page for "git clean": git-clean - Remove untracked files from the working tree18:01
And I think the git clean syntax is git clean -f, not --force, sorry.18:02
→ cableray joined18:02
→ d0k joined18:03
Belliash thanks18:03
btw according to man, both -f & --force are correct :)18:03
→ apocalyptiq joined18:04
← dannas left18:04
← xmfbs left18:05
← davzie left18:05
← MaikB left18:06
← demods left18:07
→ mndoci joined18:07
→ mlncn joined18:08
→ cableray_ joined18:10
→ Cam joined18:10
mdpatrick Belliash: Good to know. Thanks18:11
Belliash thank you too18:11
← apocalyptiq left18:12
← DrNick left18:13
← cableray left18:13
cableray_ → cableray18:13
→ p3rsist joined18:14
← shruggar left18:18
← bonhoffer left18:18
← akosikeno left18:19
← __sorin__ left18:19
← f0i left18:19
→ f0i joined18:20
→ Pupeno joined18:21
→ Gojohnnyboi joined18:24
→ mandric joined18:25
→ chrisf_ joined18:25
→ jelmer joined18:28
→ unguest joined18:31
← kermit left18:31
→ DrNick joined18:31
→ zamabe_ joined18:32
← zamabe left18:32
→ radarek joined18:35
← Gojohnnyboi left18:37
→ alnewkirk joined18:38
← Cromulent left18:39
→ intmed joined18:39
intmed No refs in common and none specified; doing nothing.18:40
Perhaps you should specify a branch such as 'master'.18:40
fatal: The remote end hung up unexpectedly18:40
error: failed to push some refs to 'ssh://git@ugit/~/git/groupchat007'18:40
cbreak intmed: kind of self explaining, isn't it? :)18:41
intmed cbreak: i ran the command "git push orgin"18:42
cbreak yeah.18:42
which part of that message is unclear?18:42
intmed Perhaps you should specify a branch such as 'master'.18:44
ran "git branch -r" but no branches18:44
→ zimbabeway joined18:44
cbreak of course not18:44
← zamabe_ left18:44
cbreak you have to tell it which of your branches you want to push18:44
so, for example, git push origin master18:44
EugeneKay git shove18:44
← yorick left18:44
← aminpy left18:45
← dcarper left18:45
← surikator left18:46
→ surikator_ joined18:46
surikator_ → surikator18:46
intmed EugeneKay: ok, got tip. i tried "git push orgin master"18:46
← DrNick left18:47
← alnewkirk left18:47
→ belkiss joined18:48
intmed Error: http://pastebin.com/s2LTqi6D18:48
← canton7 left18:48
kevlarman intmed: see faq non-bare18:48
jast intmed: *automatic message* Pushing to non-bare repositories is discouraged: please see https://git.wiki.kernel.org/index.php/GitFaq#non-bare18:48
→ v0n joined18:50
→ kermit joined18:50
→ jogla_ joined18:52
cbreak intmed: that repository was not made to be pushed to18:52
make a bare one18:52
or don't push to it18:52
or push to an other branch than the master branch18:53
← rostayob left18:53
intmed cbreak: ok18:53
← glennpratt left18:53
intmed can i make it bare18:53
cbreak just git init --bare a new one18:54
← jogla left18:54
jogla_ → jogla18:54
← Araxia__ left18:55
intmed cbreak: created a bare repo and pushed successfully18:55
cbreak good :)18:56
intmed thank u guys18:56
← dSebastien left18:56
cbreak if you want to set it as default18:56
← unreal left18:56
intmed :-)18:56
cbreak you should git push -u origin master18:56
after that you can then do git push18:56
without further arguments18:56
intmed cbreak: ok18:56
cbreak: that is much easier18:56
→ unreal joined18:57
intmed bye geeks :-)18:57
← intmed left18:57
→ dcarper joined18:58
→ kvanderw joined18:58
← kerrick left18:59
→ the_metalgamer joined19:00
→ mlncn-agaric joined19:00
→ kerrick joined19:01
← mlncn left19:02
← dcarper left19:02
→ alnewkirk joined19:02
← tobias__ left19:02
→ Fullmoon joined19:04
zimbabeway → zamabe19:06
← v0n left19:06
→ dunpeal joined19:06
dunpeal Yo. What's the best to undo a merge immediately after performing it?19:07
→ robotarmy joined19:07
→ v0n joined19:07
← nadavoid left19:08
→ B|aster joined19:08
dunpeal Is there anything (better/more specific) than `git reset --hard HEAD@{1}`?19:08
→ glennpratt joined19:09
→ paul_andrew joined19:10
→ Araxia_ joined19:12
← kvanderw left19:12
EugeneKay That's the way.19:13
→ apocalyptiq joined19:14
→ nadavoid joined19:14
← nadavoid left19:14
dunpeal EugeneKay: thanks, only problem is it doesn't work if the merge was done by rebase.19:17
Since then HEAD@{1} is just the commit before last that got rebased.19:17
EugeneKay That's cuz you rewrite history with a rebase19:17
→ Cromulent joined19:17
→ kvanderw joined19:18
→ danpker joined19:20
→ mastro joined19:20
dunpeal Indeed.19:20
← surikator left19:20
← Raging_Hog left19:22
→ mikedfunk joined19:23
→ zivester_ joined19:23
→ JamesMowery joined19:24
ChanServ set mode: +v19:24
mikedfunk hey i have a question19:24
i'm having a hell of a time setting up a post-receive hook with github19:24
→ HG` joined19:25
→ mndoci___ joined19:25
← v0n left19:25
mikedfunk i have a php script that just does shell_exec('git pull origin develop'); but it fails19:26
→ dSebastien joined19:26
→ v0n joined19:26
→ Matth1a3 joined19:26
FauxFaux mikedfunk: Why does it fail?19:27
← mndoci left19:27
mikedfunk the rsa key is not working because php runs under the apache user. the rsa key isn't generated as the apache user. how do i generate an rsa key for apache if i can't ssh in as apache?19:28
→ rjs1 joined19:28
← DPav left19:29
FauxFaux mikedfunk: Sounds like a problem you need your host to deal with, nothing to do with git.19:31
← mainrain left19:32
kevlarman mikedfunk: keys aren't tied to a user, you just need it in apache's ~/.ssh/ with the proper permissions19:32
→ kenperkins joined19:32
mikedfunk oh thank you kevlarman19:33
i thought it was for users specifically19:33
→ mainrain joined19:34
→ NuckingFuts joined19:34
← mlncn-agaric left19:36
← LongBeach left19:36
→ jham joined19:38
← churp left19:38
→ churp joined19:38
jham hi. is there a way to make git-show to print to stdout?19:38
→ Gojohnnyboi joined19:39
→ tmz joined19:39
kevlarman jham: git show|cat or GIT_PAGER= git show19:39
→ kvanderw_ joined19:40
jham ah.. thank you kevlarman19:40
kevlarman if you're just trying to pipe it somewhere for a script you don't need to do anything19:40
← kvanderw left19:41
→ indietrash joined19:42
← kenperkins left19:43
indietrash hi. git newbie here. I set up a project on github, and am writing a sort of tutorial for a friend, on how he can register there and set up his SSH-key and so on. but how exactly does he set up the project on his computer, and pull?19:44
NuckingFuts → Nuck19:45
← mndoci___ left19:45
← kermit left19:45
FauxFaux indietrash: man git clone19:46
jast indietrash: *automatic message* the 'git-clone' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-clone.html19:46
FauxFaux indietrash: man everyday is probably better, actually.19:46
jast indietrash: *automatic message* the 'everyday' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/everyday.html19:46
← Spockz left19:46
indietrash thanks19:46
→ Spockz joined19:47
← chrisf_ left19:47
dunpeal I have a repository with file foo.txt, I'd like to rewrite history as if foo.txt has always been bar/foo.txt.19:48
How do I do that?19:48
← urkud left19:48
→ chrisf_ joined19:49
kevlarman dunpeal: see man git-filter-branch19:51
jast dunpeal: *automatic message* the 'git-filter-branch' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-filter-branch.html19:51
→ beiter joined19:52
← towski left19:52
indietrash FauxFaux: so git clone will give him the current commit?19:52
FauxFaux: i.e. he doesn't need to pull after making a repository clone?19:53
FauxFaux indietrash: Try it and find out.19:53
← paissad left19:53
indietrash FauxFaux: the reason I asked was to save myself those few commands that creating and removing the clone would take... but fair enough I guess, heh. thanks.19:54
→ phretor joined19:55
← OOPMan left19:55
← zamabe left19:55
→ swombat joined19:57
→ DPav joined19:57
← phretor_ left19:58
dunpeal kevlarman: thanks.19:58
← mastro left19:58
← Matth1a3 left19:58
← CannedCorn left19:58
→ shruggar joined19:59
→ CannedCorn joined19:59
← dunpeal left20:00
→ darwinbits joined20:02
← kvanderw_ left20:03
← churp left20:04
← zemo left20:04
← swombat left20:04
→ mlncn joined20:04
← redii left20:06
← Murr_ left20:06
→ Murr_ joined20:09
→ OOPMan joined20:09
→ mastro joined20:11
← chrisf_ left20:11
→ kermit joined20:11
← eddiefullmetal left20:15
← s0ber left20:16
← kerrick left20:17
→ kennethreitz joined20:17
→ xmfbs joined20:17
→ s0ber_ joined20:18
s0ber_ → s0ber20:18
← xmfbs left20:18
→ tty234 joined20:20
→ towski joined20:20
→ ereslibre joined20:20
← ereslibre left20:20
→ ereslibre joined20:20
← jason237 left20:21
→ mudge joined20:21
← Murr_ left20:22
mudge hello, i want to use git for deploying my website, i have a version of the website on my local machine and I want to push it to a repository on my remote server which is the live site20:23
→ ningia joined20:23
mudge i am wondering what is the best way to do this20:23
→ ziggles joined20:23
mudge so i make changes to my local version and then push it to the server to make the local changes live on the remote server20:23
→ Textmode joined20:23
← pantsman left20:23
→ pantsman joined20:24
← pantsman left20:24
→ pantsman joined20:24
→ archis joined20:24
m1sc http://toroid.org/ams/git-website-howto20:24
cypha i deleted some stuff from my git directory locally20:25
but when I push to master, it doesn't remove those files20:25
how do I make it match what I currently have locally?20:25
← GunArm left20:25
← danpker left20:25
m1sc cypha: "push to master" what exactly co you do?20:25
← archis left20:26
← m4dc0d3r_ left20:26
→ GunArm joined20:26
→ danpker joined20:26
cypha co?20:26
git add .20:26
git commit -m "updates"20:26
m1sc co = do20:26
cypha git push -u origin master20:26
mudge thansk mlsc20:27
← kennethreitz left20:27
→ guns joined20:28
m1sc cypha: how did you verify your changes haven't been applied?20:28
← lut4rp left20:28
cypha i got on my repo and see if the files I removed were still there20:29
and they are20:29
m1sc cypha: so, you're pushing into a non-bare repo? !faq non-bare20:29
jast cypha: *automatic message* Pushing to non-bare repositories is discouraged: please see https://git.wiki.kernel.org/index.php/GitFaq#non-bare20:30
→ Murr_ joined20:30
→ kmc_ joined20:31
→ lut4rp joined20:31
cypha i don't get it20:31
what project doesn't have a worktree "attached"?20:31
m1sc cypha: bare repos don't have a worktree. a bare repo only contains the stuff you find under .git in your copy..20:32
← towski left20:32
→ DrNick joined20:33
m1sc cypha: as the wiki says: "..and never touches the working tree files"20:33
cypha it doesn't touch the working tree file20:33
s20:33
i'm uploading my working tree to github20:33
→ towski joined20:33
← Glycan left20:33
cypha but I'd like for the online repo to match it20:34
m1sc cypha: it should be a bare repo20:34
← kmc left20:34
→ thebignoise joined20:34
cypha and you're saying that's bad20:34
← DofLeFou left20:35
m1sc cypha: pushing to a non-bare repo is bad for the reasons in the faq..20:35
cypha k20:35
so I won't do that20:35
but how can I get it to remove the files then?20:35
(althought I don't really see why it's an issue)20:36
kevlarman cypha: because the working copy gets trashed if you push into a checked out branch20:36
cypha checked out?20:36
← jhoepken left20:38
← apocalyptiq left20:38
← angerman left20:39
← balau left20:40
← mandric left20:40
→ Slush-_ joined20:40
← paul_andrew left20:40
cypha so how do I remove the files I want removed in the repo?20:41
← the_metalgamer left20:41
cypha is there a way to do this on the website?20:41
kevlarman cypha: you stop trying to push into a nonbare repo20:42
cypha so how do I do that?20:42
← danpker left20:42
← nicoulaj left20:42
→ the_metalgamer joined20:42
cypha https://github.com/cypha/cypha20:42
that's non-bare?20:42
← dSebastien left20:42
kevlarman wait did you just rm the files?20:44
→ dnyy joined20:44
kevlarman you need to tell git that they're gone20:44
cypha locally20:44
yeah, how?20:44
kevlarman just like you need to tell it about any changes20:44
git rm20:45
EugeneKay git rm <path>20:45
cypha it can't just figure out what's gone?20:45
it was like 8 files20:45
→ shurane joined20:45
kevlarman cypha: git doesn't "figure out" any changes20:45
← Araxia_ left20:45
kevlarman you have to explicitly tell it that you want it to keep track of a change20:45
← guns left20:46
cypha cool20:46
automatic sounds easier20:46
kevlarman it's also wrong20:46
cypha how do I do that?20:46
oh20:47
← JamesMowery left20:47
cypha , why?20:47
← Phylock left20:47
cypha is that the non-bare-repo thing?20:47
kevlarman no20:47
→ jason237 joined20:47
kevlarman often people will have several things in their working copy20:47
and want to generate a commit from only some of them20:47
that's one of the reasons commit -a is discouraged20:47
cypha i always want everything just matched up20:48
FauxFaux kevlarman: Well, they're wrong. >.>20:48
cypha since i'm learning, and using it to get help20:48
← Fullmoon left20:48
cypha using git, that is20:48
and I want people to see what I'm looking at20:48
kevlarman FauxFaux: splitting a commit into smaller logical commits isn't wrong20:49
← shurane left20:50
FauxFaux Unless you want your history to pass tests at all time.20:50
jason237 sometimes I want to use -a, sometimes -p20:50
cypha i don't even know hwo to check my history20:50
EugeneKay Unit test builds at every commit can be problematic with a commit-often workflow20:50
cypha i thought it would be similar to dropbox's straightforward interface20:50
but it isn't20:50
mdpatrick dropbox?20:50
dropbox is a filesharing service20:50
whats that got do with git20:51
or revision control in general20:51
cypha i don't see any website capabilities, like adding or removing files from folders20:51
dropbox has revision control20:51
m1sc cypha: plz do some reading in !book20:52
jast cypha: *automatic message* 'Pro Git' is a complete book about git that is available online for free: http://progit.org/book/20:52
→ envi joined20:53
cypha can someone just tell me how to just make git match my local directory?20:53
i really don't wanna have to remove the repo and recreate it again20:54
→ paul_andrew joined20:54
m1sc cypha: it's already matching, apparently you're just ignoring what we're saying all the time..20:55
← Spockz left20:57
mdpatrick cypha: You can do "git reset --hard HEAD" and "git clean -f" if you want to revert all of the changes in your current directory to the last commit you made (HEAD). Otherwise "git commit -a" will stage all already tracked files and commit them, if you've added new files that aren't tracked you'll need to track them with "git add" before running "git commit -a" though.20:58
And yeah. Read the book. Several times. :/ I also recommend gitref.org.20:58
← v0n left21:00
→ Anubisss joined21:00
← jelmer left21:03
← d4rk1ink left21:07
→ jelmer joined21:07
→ v0n joined21:08
← jelmer left21:08
→ [2]jer joined21:09
→ fread2281 joined21:10
→ caseymcg joined21:11
← fread2281 left21:11
→ Joefish joined21:12
← sivy left21:13
offby1 if you're gonna run "git clean", I recommend doing "git clean -dxn" first (a dry run), and then "git clean -dxf"21:15
→ kenperkins joined21:15
← jogla left21:15
← beiter left21:16
← thebignoise left21:18
cypha mdpatrick, too many options!21:19
just tell me one :)21:19
→ n1ck3 joined21:19
→ redii joined21:22
← grekko left21:23
← saste left21:23
→ thebignoise joined21:23
← ericdc left21:23
← jonnyATroot left21:24
← redii left21:25
offby1 cypha: to make (the current commit on the currently-checked-out branch of) your git repository match your work tree, I _think_ you'd want "git add . ; git commit"21:26
haven't tried it, but it sounds right21:26
→ zamabe joined21:26
cypha offby1, i did that21:27
but it didn't remove the files I removed21:27
→ kerrick joined21:27
offby1 hmm21:28
← DPav left21:28
→ steinrs joined21:29
offby1 cypha: maybe try the "-u" option21:29
steinrs hi21:29
cypha push -u?21:29
FauxFaux add -u21:29
steinrs so this puzzled me: I do "git add test.txt" on master.21:29
when i switch to another branch, "git status" shows 'test.txt' as a new file21:29
cypha perfect21:29
steinrs it should be untracked in other branches, no?21:30
cypha thank you offby1, FauxFaux21:30
← w0bni left21:30
offby1 bows21:30
→ ehsan_ joined21:30
← kumbayo_ left21:30
offby1 steinrs: nope21:30
FauxFaux steinrs: checkout maintains uncommitted working copy state; if you commit it first then it'll vanish when you switch branches.21:30
offby1 steinrs: what you're seeing is indeed the correct behavior21:30
cypha offby1, what exactly is -u?21:30
offby1 what FauxFaux said :)21:30
FauxFaux cypha: man git add21:30
jast cypha: *automatic message* the 'git-add' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-add.html21:30
offby1 cypha: man git-add21:30
jast cypha: *automatic message* the 'git-add' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-add.html21:30
offby1 FauxFaux: jinx21:30
steinrs that's just idiotic21:30
cypha man git add21:30
jast *automatic message* the 'git-add' manpage can be found at http://www.kernel.org/pub/software/scm/git/docs/git-add.html21:30
FauxFaux Totally almost called jinx.21:30
steinrs: Oh no it's not!21:31
offby1 hits FauxFaux on the upper arm21:31
steinrs it is21:31
totally unexpected21:31
offby1 steinrs: OK, it's idiotic. Next question.21:31
steinrs no more questions21:31
→ shiba_yu36 joined21:31
← jceb left21:33
← darwinbits left21:34
← mastro left21:34
→ jelmer joined21:34
→ WolfTail joined21:35
steinrs are there other crazy "branchless" commands, other than git add?21:37
offby1 dunno what you mean by that21:37
bremner that was a question21:37
well, more of a troll21:37
steinrs you don't understand the definition of troll21:37
bremner steinrs: yes, there are.21:37
steinrs list'em21:37
offby1 I will charitably interpret it as "are there other commands whose behavior is likely to surprise me"21:37
← envi left21:37
steinrs yup21:37
offby1 steinrs: all of 'em :)21:38
steinrs nope21:38
offby1 srsly, git is very different from other systems; I'm not surprised that you're surprised.21:38
steinrs diff is branch aeare21:38
aware21:38
add is not21:38
→ darwinbits joined21:38
offby1 fair enough21:38
bremner steinrs: there is only one index21:38
steinrs offby1: well, yeah, muwhaha21:38
offby1 "git add" tweaks the inde21:38
x21:38
steinrs bremner: i figured that much :D21:38
fucking weird21:38
→ jonnyATroot joined21:39
bremner so commands that work with the index...21:39
steinrs i work on a branch, so i expect all shit to happen on that branch21:39
offby1 I think there's an option to "git diff" that compares the index to the work tree, so that is probably "branchless"21:39
← n1ck3 left21:39
← kerrick left21:39
offby1 steinrs: well, your expectations might not match reality. It happes.21:39
happens.21:39
steinrs bremner: what's the rationale for not having per-branch indices?21:39
→ shurane joined21:40
bremner cost-benefit thing, I guess21:40
branches are currently super cheap in git21:41
also, it isn't obvious to me what is more useful21:41
offby1 I got a six-pack of branches yesterday for $3.75, but they were a month old21:41
steinrs can't see how a global index contributes to that21:41
← Anubisss left21:41
bremner well, a per branch index would make a branch into something with more state than a single pointer21:42
DrNick and since the whole point of the index is to serve as an intermediate state between what's in the branch and what's in the working tree, making the index a part of the branch would be just weird21:42
bremner people complain about one index. Imagine the multiplication factor ;)21:43
← TrineLocke left21:43
DrNick don't forget that the index can actually hold multiple copies of the same file when you're doing a merge21:43
steinrs i don't want the index, no intermediate state needed21:43
where do i turn it off?21:43
steinrs dodges the troll bullets21:43
offby1 steinrs: you can't, exactly, but there are a couple of options to common commands that force you to think about it less.21:44
I think "-a" for "git add" is one such21:44
steinrs righto21:44
offby1 steinrs: google around; there are surely a bunch of "how to use git more like svn" tutorials.21:45
← dnyy left21:45
offby1 Naturally I recommend that you actually learn to love the index, but ...21:45
steinrs i'm not coming from svn, but monotone21:45
← caseymcg left21:45
steinrs which is slightly saner21:45
to my simple-minded self21:45
← phretor left21:45
steinrs alas, have to use git in next-project-shit21:45
avar Git or: How I Learned to Stop Worrying and Love the Index21:45
steinrs reads up on index21:45
offby1 monotone is much simpler than git, if I recall correctly.21:46
avar The idea of multiple indexes is actually quite nice, I often wanted that feature.21:46
offby1 and yet git was inspired by it.21:46
avar: :-)21:46
steinrs offby1: it's totally intuitive and does all i want21:46
avar offby1: Sorry to derail your train :)21:46
steinrs except i can't use it this time around21:46
bremner I have noticed a pattern that everyone loves the system they start with.21:47
← jablko left21:47
m1sc steinrs: so for your own projects you still would choose monotone?21:47
bremner well, unless they start with CVS21:47
steinrs bremner: i started with perforce and hate it21:47
m1sc: totally21:47
bremner: haha, yea21:47
avar But basically git keeps stuff in the object store, tags, branches, then it keeps a bunch of ad-hoc metadata around about the current state. That's the index, .git/rebase* .git/*HEAD* etc.21:47
→ phretor joined21:47
steinrs i tried to tell the dude git was overkill for this project, but apparently it's the latest craze21:48
offby1 bremner: I started with CVS, I guess. Used RCS. Wasn't crazy about either. Learned Subversion and thought it was the bee's knees for a long time.21:48
avar Now, the index is a special non-tree datastructure but I wonder if it couldn't just be stored in a blob along with the rest of that metadata and then you could just swap out the blob pointer.21:48
offby1 bremner: I think I grew disillusioned with svn when they introduced merge tracking21:48
→ dajhorn joined21:48
← Joefish left21:48
avar But the main argument against something like "the index as a normal structure" is that you don't want to have to sha1 the index eath time it changes.21:48
steinrs offby1: my goodness, had to use non-distributed rcs's in the past too, combined with heavy travelling (and thus mostly offline) - crappy21:49
bremner remembers trying to use svk21:49
steinrs crazzy21:49
and what the fuck is up with the billion .svn directories everywhere?21:50
someone was lazy in the design phase21:50
bremner svk doesn't have those ;)21:50
steinrs or just wanted cvs-ism21:50
→ JamesMowery joined21:50
steinrs bremner: +1 to that21:50
FauxFaux Or they thought that gzipping and packaging files into large pack files, or inventing your own storage framework when the filesystem already has one, would be horribly complex for people trying to implement things, and suck complete cocks when any file anywhere in your working copy would be over a few mb, and wanted partial checkouts, and..21:51
← radarek left21:51
m1sc steinrs: svn is going away from the .svn madness fortunately iirc21:52
avar The .svn directories are actually quite interesting, in that if they'd actually coalesced them into one directory they'd probably have to change enough to have implemented a better VCS :)21:52
← spaceonline left21:52
steinrs i think i'll write my own, in Java for speed o.O21:52
← rhl left21:52
offby1 bremner: oooh I remember svk too. Never could even build it.21:52
*shudder*21:53
avar svk was quite nice if you had to use svn before git-svn21:53
→ rhl joined21:53
offby1 and if you could get it working. I barely did.21:53
bremner I'm not sure if I was unfair, but I remember describing svk as "fragile as a bag of glass hammers"21:53
cmn avar: they are moving to just having one top-level .svn dir, and instead of a lot of .svn dirs, they're going to use SQLite21:53
← Agos left21:54
DrNick sqlite to store what, exactly?21:54
steinrs metadata21:54
cmn everything21:54
steinrs pristine copies21:54
avar The .svn dirs also store a complete copy of every original upstream revision.21:54
steinrs crazy21:54
cmn as in, what you've changed, where you've moved, the pristines21:54
avar So what if you clone a 2TB repository, you'll have a single 2TB SQLite file?21:54
cmn I guess21:55
steinrs SQLHeavy21:55
cmn though maybe the pristines are out of the DB, I'm not positive on that21:55
what I am sure they store there is the revision each file/dir is at, and whether it's been moved21:56
steinrs always wanted to rewrite history with SQL21:56
← d0k left21:56
→ d0k joined21:56
← ThomasLocke left21:56
→ Niedar joined21:57
→ urkud joined21:58
→ nuhya joined21:59
← simenbrekken left22:00
← belkiss left22:01
← shiba_yu36 left22:02
kmc_ → kmc22:03
→ eijk joined22:04
→ d4rk1ink joined22:07
→ sivy joined22:07
→ jimubao joined22:07
← drizzd left22:09
→ redii joined22:10
← HG` left22:10
← redii left22:11
← jonnyATroot left22:11
← squeese left22:12
→ jeffery joined22:12
← SirFunk_ left22:13
← ehsan_ left22:15
→ mike_perdide joined22:15
← cypha left22:16
← steinrs left22:18
← urkud left22:18
← a_suenami left22:18
← Slush- left22:19
← Morphous left22:19
→ poppiez_ joined22:19
→ shiba_yu36 joined22:19
← poppiez left22:20
→ eijk_ joined22:20
← Cromulent left22:20
← mikedfunk left22:20
← kst left22:22
← [2]jer left22:23
← eijk left22:24
→ himsin joined22:24
→ EricInBNE joined22:24
← thebignoise left22:24
→ dnyy joined22:29
→ reuss joined22:31
→ danny_ joined22:31
→ mandric joined22:32
← danny_ left22:32
← nuhya left22:32
→ HG` joined22:32
→ shiba_yu_ joined22:33
← shiba_yu36 left22:33
→ Morphous joined22:35
← marienz left22:38
→ lisak joined22:38
← dv310p3r left22:40
← p3rsist left22:40
→ Error404NotFound joined22:42
← Error404NotFound left22:42
→ Error404NotFound joined22:42
← mklappstuhl left22:43
← mtbre left22:44
→ marienz joined22:45
← JamesMowery left22:46
→ jmowery joined22:46
→ urkud joined22:46
← shiba_yu_ left22:46
← shruggar left22:47
← urkud left22:47
← jmowery left22:47
→ JamesMowery joined22:47
lisak I run $git format-patch master --binary > fix_empty_poster3.patch22:48
→ kst joined22:49
lisak but when applying it22:49
http://pastebin.com/R0XJRU9R22:49
it doesn't source those 000* patches22:49
do I have to manually apply them all, or I do something wrong22:51
← sivy left22:51
avar you probably want the --stdout option22:52
→ sivy joined22:52
lisak instead of --binary22:52
or in addition to22:52
avar in addition to22:52
lisak both args is the same as only stdout22:53
kevlarman lisak: also you want git am for git-format-patch's output, not git apply22:53
lisak am says "it's emtpy" too22:54
this sourcing of patches in a file .... I don't know what it is about22:54
avar you have to source actual patches, not a list of files you wrote out22:55
hence --stdout22:55
→ vanmik joined22:55
avar or git am *.patch22:55
← Gojohnnyboi left22:55
← ziggles left22:56
lisak I thought that the --binary option includes binary files into the patch22:56
but it is nonsense, isn't it22:57
man : In addition to --full-index, output a binary diff that can be applied with git-apply.22:57
FauxFaux LOOK AT ONE OF THE FILES INSTEAD OF RANDOMLY SPEAKING22:57
← posciak left22:59
→ envi joined23:00
lisak I did, 0005 is a binary patch, the rest textual patches23:00
without --binary option, the binary patch wouldn;t be part of it ?23:01
← the_metalgamer left23:01
→ nazgul102 joined23:01
← eijk_ left23:02
lisak I see, it would, everything in one file...23:02
← nazgul101 left23:04
→ kerrick joined23:04
→ tcurdt joined23:05
avar The program by default writes out patches in the cwd and a list of those files23:05
you piped the list of files to a file23:05
then you tried to apply it, this obviously doesn't work.23:05
instead you want --stdout, which'll put all the patches in one big file23:05
or tar up the individual patches23:06
← BiggFREE left23:07
→ gtar joined23:07
← gtar left23:07
← tcurdt left23:09
→ dabd joined23:10
← thiago left23:13
← rendar left23:15
→ kpreid joined23:15
→ caseymcg joined23:19
→ Cromulent joined23:20
→ maestrojed joined23:20
← ningia left23:21
lisak avar, I wasn't sure that it includes the binary files23:23
btw the patch is messed up :-) http://pastebin.com/qk3Ag9Va23:23
if I look at those lines I don't see anything wrong23:24
avar the patch is fine, that's just pedantry in git-am23:25
(Which I think is beyond moronic to have as the default behavior)23:25
← Andrew_ left23:25
← glennpratt left23:25
→ sh_sh joined23:26
kevlarman avar: default is to print a warning, and it's seriously irritating to deal with code like that23:27
← tewecske left23:27
kevlarman i've been tempted on multiple occasions to install update hooks that reject code containing what git considers whitespace errors23:28
→ macmartine joined23:31
avar It's seriously irritating that git by default has warnings that mostly only apply for code with lkml-like coding styles.23:32
E.g at work there's a bunch of designers using Git for HTML templating where nobody cares about whitespace and that error really confuses them23:33
macmartine Anyone know what would cause this error? I'm not sure if the gem is to blame or not https://github.com/gunark/rubycas-server/issues/6223:33
← shurane left23:35
← mandric left23:35
← maestrojed left23:36
← vmil86 left23:38
← lisak left23:40
→ jfelchner joined23:40
← alnewkirk left23:42
→ caseymcg1 joined23:42
frogonwheels avar: pretty sure you can turn off the whitespace warnings.23:43
→ urkud joined23:43
frogonwheels avar: the problem is that adding/removing trailing whitespace can cause unwanted merge issues.23:43
← caseymcg1 left23:43
→ xpot-mobile joined23:44
→ alnewkirk joined23:46
avar The real reason that warning is there is because Linus is a line-ending whitespace pedant23:46
But actually the biggest problem surface for it is that git-rebase will invoke git-am (or equivalent) and print out those warnings23:46
But "you can configure it" is really not a good answer if the problem is "it confuses non-technical users"23:47
E.g. I've had designers tell me that they don't use rebase because "they always get errors", which turns out to be this stuff.23:47
kevlarman avar: git confuses non-technical users23:49
→ yorick joined23:50
avar Not all of its features confuse them equally23:50
← madewokherd left23:51
→ simplechat joined23:51
← brian_g left23:56
→ SirFunk_ joined23:56
← dnyy left23:56
→ dcarper joined23:57
→ p3rsist joined23:57
→ shurane joined23:58
← alnewkirk left23:59
→ bigkm joined23:59

Logs Search ←Prev date Next date→ Channels Documentation