| 2009-12-18 |
|
scone
| hello all | 00:00 |
|
| i've got a confusing issue | 00:00 |
|
| i have a repository which, on the remote server, returns no problems for git fsck | 00:01 |
|
kithpom
| hello dessert biscuit | 00:01 |
|
scone
| but when i try to pull, it errors out | 00:01 |
|
| even if i delete the local repository and try to pull a fresh one | 00:01 |
|
j416
| scone: you cannot clone? | 00:02 |
|
| what is the error message? | 00:02 |
|
scone
| well | 00:02 |
|
| the error message on a pull is something like "early EOF" | 00:03 |
|
| among a few other things | 00:03 |
|
j416
| I was asking about clone | 00:03 |
|
scone
| well | 00:03 |
|
| it'll take a bit | 00:03 |
|
| but hang on | 00:03 |
|
j416
| ok :) | 00:03 |
|
| (thought you tried already) | 00:03 |
|
scone
| no, i just init'd an empty repo and did a pull | 00:03 |
|
j416
| o_O | 00:03 |
|
scone
| it's the same thing, isn't it? | 00:03 |
| → ruediger joined | 00:04 |
|
j416
| I'm not sure. | 00:04 |
|
scone
| on pull the error happens after ~100M | 00:04 |
|
j416
| out of memory? | 00:04 |
|
scone
| nah | 00:04 |
|
j416
| how about the server? | 00:04 |
|
scone
| the machine has 4G, and i assume that it would say somethign to that effect | 00:05 |
|
j416
| ok | 00:05 |
|
scone
| it's a shared virtual server with 16G, but a pretty low limit | 00:05 |
|
j416
| how large is this repo? | 00:05 |
|
scone
| i can't run git fsck --full on the remote | 00:05 |
|
| umm | 00:05 |
|
j416
| does it have large binary files in it? | 00:05 |
|
scone
| compressed? | 00:05 |
|
| not too large | 00:05 |
|
| largest single files are like ~12M images | 00:05 |
|
j416
| ok | 00:05 |
|
scone
| it also has lots of little files | 00:05 |
| ← neoeinstein left | 00:06 |
|
j416
| shouldn't be much of a problem then | 00:06 |
| sdboyer → crapstain | 00:06 |
|
j416
| git often chokes on large files | 00:06 |
|
scone
| it's a game project, and the engine i'm using (Unity) is a real pain with its file structure | 00:06 |
| crapstain → sdboyer | 00:06 |
|
scone
| yeah i'm familiar with that heh | 00:06 |
| ← Paraselene_ left | 00:06 |
|
j416
| it sounds to me like the server has problems packing the repo | 00:08 |
| ← smotchkkiss left | 00:08 |
|
j416
| can you run git gc on the remote even? | 00:08 |
| → loffe joined | 00:09 |
| → Paraselene_ joined | 00:10 |
| ← hipe_ left | 00:10 |
| → hipe joined | 00:10 |
| ← drizzd_ left | 00:11 |
|
j416
| scone: if your repo is corrupt, Linus writes pretty detailed about how to fix it: http://kerneltrap.org/mailarchive/git/2007/7/31/253494 | 00:11 |
|
scone
| git gc? | 00:11 |
|
| here are the errors | 00:11 |
| ← necr0tik left | 00:12 |
|
scone
| remote: Counting objects: 3624, done. | 00:12 |
|
| remote: Compressing objects: 100% (1356/1356), done. | 00:12 |
|
| error: git-upload-pack: git-pack-objects died with error./s | 00:12 |
|
| fatal: git-upload-pack: aborting due to possible repository corruption on the re | 00:12 |
|
| mote side. | 00:12 |
|
| remote: aborting due to possible repository corruption on the remote side. | 00:12 |
|
| fatal: early EOF | 00:12 |
|
| fatal: index-pack failed | 00:12 |
| ← bugfux left | 00:12 |
|
scone
| ooh! yeah it looks like the server memory cap shuts down git | 00:13 |
|
| when i run git gc on the server | 00:13 |
|
| thx for the tip! | 00:13 |
|
j416
| :) | 00:14 |
|
| scone: when you clone it over the git protocol (or fetch), it will try to pack things up on the remote side first | 00:14 |
| ← JasonWoof left | 00:14 |
| → hummingkid joined | 00:15 |
|
j416
| if you clone over http for example, that won't happen | 00:15 |
|
scone
| yeah, i know that. i just didn't connect the dots that it was beign stopped while compressing | 00:15 |
|
| really... | 00:15 |
|
| how do i git over http? | 00:15 |
|
| i'm currently using ssh | 00:15 |
|
cbreak
| you can only pull over http | 00:15 |
|
| if that's enough | 00:15 |
|
scone
| hm | 00:15 |
|
| yeah it is actually | 00:15 |
|
cbreak
| just upload your repository somewhere | 00:15 |
|
| and pull from there | 00:15 |
|
scone
| although... | 00:15 |
|
| it's huge uncompressed | 00:15 |
|
doener
| run "git update-server-info" | 00:16 |
|
cbreak
| but make sure that you run before every upload | 00:16 |
|
| exactly that :) | 00:16 |
| ← Paraselene_ left | 00:16 |
|
cbreak
| there's a hook for it | 00:16 |
|
j416
| just put your repo on the server | 00:16 |
|
| oh | 00:16 |
|
| i was slow lol | 00:16 |
|
scone
| haha | 00:16 |
|
| yeah.. uncompressed it's like 600M | 00:16 |
|
j416
| wow. | 00:16 |
|
| huge | 00:16 |
| → JasonWoof joined | 00:16 |
| → bennyw joined | 00:17 |
|
j416
| scone: you should run git gc on the server from time to time | 00:17 |
|
doener
| IIRC there was also a plan to make git run that automatically when (IIRC) receiving a push. Not sure whether that was actually implemented though | 00:17 |
| ← danferranti left | 00:17 |
|
scone
| j416, oh, so when i pull, it'll still pull compressed? | 00:17 |
|
j416
| scone: not if you pull (fetch) via http | 00:18 |
|
doener
| scone: btw, there's the pack.windowMemory config value, which might help | 00:18 |
|
j416
| in that case it will just get what the server has | 00:18 |
|
| server will do nothing but send it | 00:18 |
|
| git will not be run on the server | 00:18 |
| ← icwiener left | 00:18 |
|
j416
| s/server/remote/ | 00:18 |
|
cbreak
| apache can compress it | 00:18 |
|
| I guess | 00:18 |
| → icwiener joined | 00:18 |
|
cbreak
| if git sents the http headers to support that, I have no idea if it does this | 00:19 |
|
j416
| yes, but that has nothing to do with git | 00:19 |
|
| me neither | 00:19 |
|
scone
| doener, where do i set that memory value? | 00:19 |
| ← d0ugal left | 00:19 |
|
j416
| scone: git config | 00:19 |
|
cbreak
| it doesn't | 00:20 |
|
scone
| cool | 00:20 |
|
| thanks guys | 00:20 |
|
cbreak
| hmm... "Accept: */*" maybe. | 00:20 |
|
scone
| i'll come back if i have more questions :) | 00:20 |
|
j416
| ;) | 00:20 |
| → nitrate joined | 00:23 |
|
Ilari
| HTTP compression doesn't help. The data is already compressed anyway. | 00:24 |
|
j416
| true | 00:24 |
| ← hipe left | 00:24 |
| → babilen joined | 00:24 |
| ← benjo2 left | 00:24 |
| bennyw → benjo2 | 00:24 |
| → hipe joined | 00:24 |
| ← tbuehlmann left | 00:24 |
| → timcharper joined | 00:25 |
| → tfarina_ joined | 00:26 |
| ← snearch_ left | 00:27 |
| ← loffe left | 00:28 |
| → jrmuizel joined | 00:29 |
| → nothingHappens joined | 00:29 |
|
sitaram
| j416: morning! | 00:29 |
|
j416
| hi :) | 00:29 |
|
sitaram
| Mpenz: I'm here now :) | 00:29 |
|
j416
| yeah, that was the reason I summoned you. :) | 00:30 |
|
| sorry | 00:30 |
| → akahn joined | 00:30 |
|
sitaram
| j416: np :) | 00:33 |
|
| Oxf13: hi | 00:33 |
| ← joce left | 00:34 |
| erlehmann → berlinbernd | 00:35 |
|
Mpenz
| sitaram: hi. | 00:35 |
|
| how are you | 00:35 |
| ← scone left | 00:36 |
| berlinbernd → erlehmann | 00:37 |
|
sitaram
| Mpenz: good! lets talk... you had a problem with upgrading? (I just woke up by the way!) | 00:37 |
|
Mpenz
| Yes. Can you give me 5 min. I'm eating dinner. | 00:38 |
|
| brb | 00:38 |
|
sitaram
| np | 00:39 |
| ← tfarina left | 00:40 |
| ← jrmuizel left | 00:42 |
| ← dsch04 left | 00:42 |
| ← dreiss left | 00:43 |
|
Mpenz
| ok sitaram i'm back. So, I'm moving my main repos to the branch with wildrepos. However, I had some issues earlier. | 00:43 |
| → jrmuizel joined | 00:43 |
| → aresnick joined | 00:45 |
|
Mpenz
| And the error I seem to get is: bad command: true. Make sure the repo name is exactly as in your config | 00:45 |
|
| pubkey access didn't work; please set it up using 'ssh-copy-id' or something | 00:45 |
|
| btw sorry about getting so late w/ the emails, I've been swamped with boring work like writing papers :( | 00:46 |
| ← hummingkid left | 00:48 |
|
sitaram
| no problem | 00:48 |
|
| I understand! | 00:48 |
|
| havent checked my email yet thought... if you had anything in there | 00:48 |
| → dreiss joined | 00:48 |
|
Mpenz
| Nothing about this issue I guess. | 00:49 |
|
| Just some thoughts I had while I was thinking about the wildrepos extensions you've made. | 00:49 |
|
sitaram
| so lets see: can you tell me what the exact command you're trying is? | 00:49 |
|
Mpenz
| ./src/gl-easy-install -q git host.server.ca | 00:49 |
|
sitaram
| what does "ssh [email@hidden.address] say? | 00:51 |
| ← ingo86 left | 00:51 |
| → warthog9 joined | 00:51 |
|
Mpenz
| it gives me my read/write permission on repos | 00:51 |
| → bennyw joined | 00:51 |
| → pinnacle joined | 00:52 |
|
sitaram
| Mpenz: in brief: you need a different key to do the install, a key that gets you a shell. can you paste "ls -al ~/.ssh"? | 00:52 |
|
Mpenz
| ah | 00:52 |
|
pinnacle
| I changed the name of a file and when I do a git status it says there is an untracked file. How to I roll this back? I tried git checkout -- . with no luck | 00:53 |
|
Mpenz
| -rw------- 1 hiren hiren 1671 2009-02-28 01:03 id_rsa | 00:53 |
|
| -rw-r--r-- 1 hiren hiren 396 2009-02-28 01:03 id_rsa.pub | 00:53 |
|
| -rw-r--r-- 1 hiren hiren 13298 2009-11-24 23:04 known_hosts | 00:53 |
|
RobertLaptop_
| sitaram what is big-config? You have branches come and go all the time :) | 00:53 |
|
wereHamster
| pinnacle: rename the file back | 00:53 |
|
pinnacle
| wereHamster: thanks :) | 00:53 |
| → struberg joined | 00:53 |
|
sitaram
| RobertLaptop_: 9000+ repos, 10000+ users makes the current "master" die with OOM | 00:54 |
|
RobertLaptop_
| Someone got that large I guess | 00:54 |
|
sitaram
| Mpenz: you only have one kkey here, and that key gets you gitolite access. Is this not the machine/account from which you did the previous easy-install? | 00:54 |
|
Mpenz
| sitaram: no this isn't the machine. | 00:54 |
|
pinnacle
| if i do a git log it display all my commits. Is there anyway I can delete the 3rd commit? | 00:55 |
|
kithpom
| git files are not visible to non-admin type when they create it? | 00:55 |
|
cbreak
| yes | 00:55 |
|
wereHamster
| pinnacle: rebase -i and drop the corresponding line | 00:55 |
|
cbreak
| but! | 00:55 |
|
sitaram
| Mpenz: then it won't work unless you do something specific. Want me to write this briefly right here or pastebin more detailed description? | 00:55 |
|
cbreak
| it rewrites history, so causes all kinds of trouble | 00:55 |
| → timcharper2 joined | 00:55 |
|
Mpenz
| sitaram: it | 00:56 |
|
| 's up to you, | 00:56 |
|
pinnacle
| wereHamster: what do you mean by dropping the line? | 00:56 |
|
timcharper2
| timcharper | 00:56 |
|
Mpenz
| might be easier if it was in pastebin sitaram | 00:56 |
|
wereHamster
| pinnacle: delete t | 00:56 |
| → futoase joined | 00:56 |
| ← timcharper2 left | 00:56 |
| ← timcharper left | 00:56 |
| → timcharper joined | 00:56 |
| → dsch04 joined | 00:57 |
|
wereHamster
| pinnacle: or easier: git rebase --onto <commit>{^,} | 00:57 |
|
pinnacle
| ahh | 00:57 |
|
| i'll try that | 00:57 |
|
| thanks | 00:57 |
|
wereHamster
| rebase --onto HEAD~3{^,} | 00:57 |
|
pinnacle
| what are all those symbols? | 00:58 |
|
wereHamster
| black magic! | 00:58 |
|
sitaram
| Mpenz: give me 5-10 min | 00:58 |
|
Mpenz
| Sure sitaram | 00:58 |
| ← benjo2 left | 01:00 |
| bennyw → benjo2 | 01:00 |
|
kithpom
| does it make sense that non-admin/wheel users can see or necessarily access git related files that they created? | 01:01 |
|
| can't* | 01:01 |
|
pinnacle
| omg, i i think that rebase just broke EVERYTHING on the repo | 01:01 |
|
| is there anyway i can undo the last 3 rebases i did? | 01:02 |
|
shd_
| pinnacle: look at reflog, git reset --hard oldid | 01:03 |
|
| pinnacle: see .git/logs/HEAD | 01:03 |
|
pinnacle
| shd_: reflog displays the branch i want to go back to | 01:03 |
|
cbreak
| pinnacle: told you | 01:04 |
|
| rebases cause all kinds of trouble :) | 01:04 |
|
| they rewrite history | 01:04 |
|
pinnacle
| cbreak: i thought you were talking to someone else :( | 01:04 |
|
| so from the reflog, if i want to go back to this version, what should i be typing? b87939c HEAD@{6} | 01:04 |
| → dduncan joined | 01:05 |
| ← dduncan left | 01:05 |
|
cbreak
| you can use the hash | 01:05 |
|
shd_
| pinnacle: on any branch, type git reset --hard b87939c and that branch will be set to that point in history | 01:05 |
|
| pinnacle: you can first try it with a test branch if you like.. | 01:06 |
|
pinnacle
| shd_: thanks :) | 01:06 |
|
shd_
| git checkout -b testingreset b87939c | 01:06 |
| ← patmaddox left | 01:06 |
| ← Dashkal left | 01:07 |
|
pinnacle
| shd_: thank you so much! | 01:08 |
|
| you saved my bacon! | 01:08 |
|
shd_
| pinnacle: you're welcome :) | 01:08 |
| ← reidab left | 01:08 |
|
cbreak
| in general, if you rewrite history | 01:08 |
|
mamund
| dang! msysgit's git-svn implementation is _dog-slow_ is this any better on unbuntu? | 01:09 |
|
cbreak
| you change every single commit that is a descendent of the commit you change | 01:09 |
|
| I don't know if it updates tags | 01:09 |
| ← futoase left | 01:09 |
|
cbreak
| but you can for example not push to a remote anymore, and if you do, other people can't pull from there anymore | 01:09 |
|
| (all with a "unless they know what to do") added | 01:10 |
| → WinterWeaver joined | 01:12 |
| → scarabx joined | 01:14 |
|
sitaram
| Mpenz: http://pastebin.com/m6b810014 | 01:14 |
|
pinnacle
| pretty much, dont rebase if you're a noob like me. get it hehe | 01:15 |
|
Mpenz
| ok sitaram | 01:15 |
|
kithpom
| should git be adding .gitignore to new files? | 01:18 |
|
alex87
| kithpom, adding it to new files? | 01:18 |
|
kithpom
| alex87: with git add . | 01:19 |
|
alex87
| kithpom, i don't understand what you mean, .gitignore is a file, how do you add it to new files? | 01:19 |
| ← SkyX left | 01:19 |
|
cbreak
| cat .gitignore >> newfile.txt | 01:19 |
| → shfx joined | 01:20 |
| → Dashkal joined | 01:20 |
| → drostin77 joined | 01:20 |
|
Mpenz
| sitaram: that worked! I was able to update it. | 01:20 |
|
kithpom
| alex87: nm, I'm guessing that it is supposed to be added as a new file to commit | 01:20 |
|
cbreak
| if you want | 01:20 |
|
| you can commit .gitignore, yes | 01:20 |
|
alex87
| kithpom, ah, yes. it's just a standard file that you want to commit so other people have the same ignores | 01:20 |
|
cbreak
| it's not mandatory that you add it though | 01:21 |
| ← WinterWeaver left | 01:21 |
| ← codeFiend left | 01:21 |
|
Mpenz
| great sitaram thanks! | 01:21 |
|
sitaram
| Mpenz: you're welcome! | 01:22 |
|
Mpenz
| But I have to get going. You'll probably be hearing from me via email. | 01:22 |
|
| Oh btw, I think the set operation things would be pretty neat. | 01:22 |
| ← Sonderblade left | 01:22 |
| → Tureba joined | 01:22 |
| → amirite joined | 01:24 |
|
amirite
| what should I try if I get "The authenticity of host 'github.com (207.97.227.239)' can't be established." | 01:24 |
|
cbreak
| it's probably a man in the middle attack :) | 01:24 |
|
| cbreak is paranoid currently | 01:24 |
| ← Stravy left | 01:25 |
| → Stravy joined | 01:26 |
| → aisipos joined | 01:26 |
| ← dreiss left | 01:28 |
| ← christophsturm left | 01:28 |
| → christophsturm joined | 01:28 |
| ← christophsturm left | 01:28 |
|
Ilari
| amirite: Try to ssh there and when it prompts, accept the hostkey? | 01:29 |
|
wereHamster
| win 23 | 01:29 |
| → Keiya joined | 01:31 |
| → jinzhu1 joined | 01:31 |
|
sitaram
| Mpenz: :) | 01:34 |
| → patmaddox joined | 01:35 |
| ← tekky left | 01:37 |
| ← Alagar left | 01:38 |
| → Jak_o_Shadows joined | 01:40 |
| ← timcharper left | 01:41 |
| ← drostin77 left | 01:43 |
| ← priidu left | 01:43 |
| ← patmaddox left | 01:44 |
| → dreiss joined | 01:44 |
| → jsquared joined | 01:45 |
|
jsquared
| so, I'm using gitolite, but as I'm going through the install process, I get "**** ABORTING ****; undefined group others" and then a fatal error from git about how gitolite-admin isn't a git repo | 01:46 |
|
| did I muck something up somewhere? | 01:47 |
| ← hipe left | 01:50 |
| → hipe joined | 01:51 |
| → tjc joined | 01:51 |
| ← agib left | 01:52 |
| ← Whtiger left | 01:54 |
| ← engla left | 01:54 |
| → bugfux joined | 01:56 |
| → threeve joined | 01:56 |
| ← tomoj left | 01:58 |
| → r11t joined | 01:58 |
|
Oxf13
| sitaram: hi, I sent you mail instead. | 01:58 |
| → drostin77 joined | 01:59 |
| → erlehmann_ joined | 01:59 |
| → steveire joined | 01:59 |
| ← pointlessjon left | 01:59 |
| ← drostin77 left | 02:00 |
| ← mugwump left | 02:00 |
| ← Tureba left | 02:00 |
| ← steveire_ left | 02:01 |
| ← Fullmoon left | 02:02 |
| → mugwump joined | 02:03 |
| → ehabkost joined | 02:06 |
| ← mmercer left | 02:08 |
| ← kithpom left | 02:08 |
| ← kmc_ left | 02:09 |
| → Whtiger joined | 02:10 |
| → lightcap joined | 02:10 |
| ← Keiya left | 02:12 |
| → Keiya joined | 02:12 |
| ← bugfux left | 02:14 |
| ← Yuuhi left | 02:16 |
| ← tjc left | 02:18 |
| → matten joined | 02:18 |
| ← lightcap left | 02:18 |
| ← matten left | 02:18 |
| → reprore_ joined | 02:22 |
| → jerbear joined | 02:24 |
| → reidab joined | 02:25 |
| → VVSiz_ joined | 02:25 |
| → Keiya_ joined | 02:25 |
| ← erlehmann left | 02:25 |
| ← reidab left | 02:26 |
| ← j416 left | 02:26 |
| → upul joined | 02:27 |
| ← TDJACR left | 02:29 |
| ← struberg left | 02:34 |
| → D[a]vey joined | 02:34 |
|
Oxf13
| when using git via ssh://, is there any way to avoid using the full absolute path to repos on the server? (makes it hard to do ssh://server/<repo>) | 02:39 |
|
petercooper
| ssh://server/~/... | 02:40 |
|
Oxf13
| hrm, yeah, not much more help there | 02:40 |
|
petercooper
| I have some at dreamhost and it's like server/~/git/whatever | 02:40 |
|
Oxf13
| damn, guess I'll have to introduce a top level dir | 02:40 |
| ← coolcat left | 02:41 |
| → TDJACR joined | 02:41 |
|
petercooper
| Since git is using openssh, and hence any .ssh/config options, I wonder if there's a way in openssh to define a "default" destination on a host | 02:42 |
|
| like you do with a port number, hostname, etc. | 02:42 |
|
Oxf13
| yeah, but that gets rough when you're talking about 1500 users | 02:43 |
| ← Keiya left | 02:43 |
|
petercooper
| This would be from your local machine, not the server | 02:43 |
| ← defectiv left | 02:43 |
| ← VVSiz left | 02:43 |
|
petercooper
| That said, the manpage for ssh_config isn't giving me much anyway | 02:44 |
| ← loincloth left | 02:46 |
|
bremner_
| Oxf13: maybe you would be happy with url insteadOf trick of making aliases for urls in .gitconfig | 02:48 |
|
Oxf13
| bremner_: but that only works after the initial clone wouldn't it? | 02:49 |
| ← gbacon left | 02:49 |
|
bremner_
| Oxf13: I but mine in ~/.gitconfig | 02:49 |
|
| s/b/p/ | 02:49 |
| ← cubuanic left | 02:50 |
|
Oxf13
| yeah, but do you really want 8K repos with that config in there? | 02:50 |
| → paltman joined | 02:50 |
|
bremner_
| I have one per host, not one per repo | 02:50 |
|
Oxf13
| I see. | 02:50 |
| ← godsflaw left | 02:50 |
|
bremner_
| [url "git+ssh://alioth.debian.org/git/collab-maint/] \n insteadOf = collab: | 02:51 |
| → CloudHopper joined | 02:51 |
|
bremner_
| git clone collab:foo.git | 02:51 |
| ← CloudHopper left | 02:52 |
| → CloudHopper joined | 02:52 |
| ← WALoeIII left | 02:52 |
| ← amirite left | 02:52 |
| ← CloudHopper left | 02:55 |
| ← johnw left | 02:56 |
| ← tfarina_ left | 02:58 |
| ← Haarg left | 03:00 |
| → cubuanic joined | 03:03 |
| ← paltman left | 03:09 |
| → reidab joined | 03:09 |
| ← aziz left | 03:12 |
| → dgarr joined | 03:16 |
| ← ereslibre left | 03:16 |
| → Tk_L joined | 03:17 |
| ← lresende left | 03:18 |
| → WALoeIII joined | 03:18 |
| ← troyt left | 03:18 |
| → troyt joined | 03:20 |
|
Mpenz
| so what'st he fuss about big-repos? | 03:20 |
| → cytrinox_ joined | 03:20 |
| → defectiv joined | 03:22 |
| → paltman joined | 03:22 |
| ← alex87 left | 03:28 |
| ← gaveen left | 03:29 |
| ← t0rc left | 03:31 |
| ← jerbear left | 03:31 |
| ← jmcantrell left | 03:32 |
| ← paltman left | 03:33 |
| → paltman joined | 03:33 |
| → jmcantrell joined | 03:33 |
| ← bry left | 03:34 |
| ← nitrate left | 03:35 |
| ← cytrinox left | 03:36 |
| cytrinox_ → cytrinox | 03:36 |
| → hipe_ joined | 03:40 |
| ← hipe left | 03:40 |
|
Mpenz
| sitaram: I had a quick question: After updating, will I be able to perform the setperms/getperms operations on repos that had been created already? | 03:41 |
| → Keiya joined | 03:42 |
| ← Keiya_ left | 03:43 |
| ← WALoeIII left | 03:45 |
| ← up_the_irons left | 03:45 |
| ← jayne left | 03:45 |
| ← jspiros left | 03:45 |
| ← selckin left | 03:45 |
|
sitaram
| Mpenz: no... unless you manually create the required file on the server side | 03:47 |
|
| Mpenz: it was getting too complicated already... | 03:48 |
| ← troyt left | 03:48 |
|
Mpenz
| Ah I see sitaram | 03:48 |
|
| Hm. I guess I can look into what file to add. | 03:49 |
| → jspiros joined | 03:50 |
| → up_the_irons joined | 03:50 |
| → jayne joined | 03:50 |
| → selckin joined | 03:50 |
|
sitaram
| Oxf13: you there? I didnt understand the ssh question. Yuo want to avoid the user@host part or the path/to/repo part? | 03:50 |
| → troyt joined | 03:51 |
| ← Solarion left | 03:51 |
|
Oxf13
| sitaram: I was looking for ways to avoid the path/to/repo part. I certainly can for users who hit gl-auth-command, but the few users who get a real shell on that box bypass that command | 03:51 |
|
sitaram
| Oxf13: with gitolite, the path/to/repo part is relative to whatever you had as $REPO_BASE, and, in the config you sent me the other day, it's all at the top level | 03:51 |
| → paltman_ joined | 03:51 |
|
Oxf13
| right | 03:51 |
|
sitaram
| Oxf13: you cannot use gitolite-served repos without going through gitoltie -- read (fetch/clone) will work but any kind of push will fail unless you trick it :) | 03:52 |
| → tjaway joined | 03:52 |
| → __iron joined | 03:52 |
|
sitaram
| Oxf13: because the hook will kick in (level 2 check) and it cant find sme basic sanity stuff it bombs out | 03:52 |
|
| some* | 03:52 |
|
Oxf13
| oh crap, right, stuff in the env. | 03:53 |
|
| hrm... | 03:53 |
| ← jayne left | 03:53 |
| ← jspiros left | 03:53 |
| ← up_the_irons left | 03:53 |
| ← selckin left | 03:53 |
|
Oxf13
| that's a wrinkle :/ | 03:53 |
|
Mpenz
| sitaram: is it a simple file to add on the server side? | 03:54 |
|
sitaram
| Oxf13: well without that the per-branch checks won't work | 03:54 |
|
| Oxf13: one simple solution: fix the hook so that, right at the top, it "exit 0"s under certain conditions, like the user has a $HOME/I-am-God file present or something | 03:55 |
|
Oxf13
| sitaram: yeah, I know. I'm trying to think through this. | 03:55 |
| ← petercooper left | 03:55 |
|
sitaram
| Mpenz: very simple; look at one of the existing files in a repo you created via wildrepos and see... just one line containing the owner user, in a file called "gl-creater" | 03:55 |
|
Mpenz
| I got it. | 03:56 |
|
| Just saw it on my test git install. | 03:56 |
|
sitaram
| Oxf13: how many people do you have who are like that? | 03:56 |
| ← Keiya left | 03:57 |
|
sitaram
| Oxf13: put them in a group and add a group test at the top ought to do it, or any such thing | 03:57 |
|
| s/top/top of the hook | 03:57 |
|
Oxf13
| about 15 people | 03:57 |
|
sitaram
| ok | 03:57 |
|
Oxf13
| won't it still need an ENV setting to find the conf file with the group definitions? | 03:58 |
| → Keiya joined | 03:58 |
| → psankar joined | 03:58 |
|
sitaram
| Oxf13: I meant unix group. Just check "is current user part of BYPASS group" | 03:58 |
|
Oxf13
| yeah, that could work | 03:59 |
|
sitaram
| but why do they need to do this? take a look at, for instance, gitolite's log file to see what sort of things won't be happening if you do this | 03:59 |
|
| I'm rather fond of that log file, awesome if someone rewound too far and force-pushed, for instance! | 03:59 |
|
Oxf13
| not allowing our admins to ssh into the box would be a pretty significant change | 03:59 |
|
sitaram
| Oxf13: they'd use a second key for gitolite, not their primary one | 03:59 |
|
Mpenz
| Ok sitaram I got it working. Thanks. | 04:00 |
|
Oxf13
| we'll have our own hook that logs everything, out through a message bus | 04:00 |
|
| sitaram: our system doesn't handle multiple keys :/ | 04:00 |
|
sitaram
| Oxf13: aaah! damn; didn't know that :) | 04:00 |
| ← rolfb left | 04:00 |
| ← paltman_ left | 04:01 |
|
sitaram
| Oxf13: well then this is the only sane way. You could do *anything* that is simple to administer and code that can be used to distinguish such users and have the hook bail on finding that condition met | 04:02 |
| ← tstclair left | 04:02 |
| → daviddo joined | 04:02 |
|
Oxf13
| yeah, that's what I'm going to do, exit 0 if using a real shell | 04:02 |
| → Alagar joined | 04:02 |
| ← Alagar left | 04:02 |
| → jspiros joined | 04:02 |
| → up_the_irons joined | 04:02 |
| → jayne joined | 04:02 |
| → selckin joined | 04:02 |
| ← troyt left | 04:03 |
| ← hipe_ left | 04:04 |
|
sitaram
| Oxf13: you mean if getpwent shows shell is something in /etc/shells? cool idea... | 04:04 |
|
Oxf13
| yeah | 04:04 |
|
sitaram
| Oxf13: let me add that in too... you'll have something by the end of the day my time | 04:04 |
|
Oxf13
| hrm, wait | 04:04 |
|
| that won't work, because our users have a valid shell listed, we just override that with the authorized_keys file | 04:05 |
|
| actually | 04:05 |
|
sitaram
| better to check if ~/.ssh/authorized_keys contains command= | 04:05 |
|
Oxf13
| if we get to the update hook, can we tell from there if one of the parent processes is gl-auth-command? | 04:05 |
| ← patcito left | 04:06 |
| ← paltman left | 04:06 |
|
sitaram
| Oxf13: that execs out to git-shell, so I dont' think so... I have 2 solutions: | 04:06 |
|
| Oxf13: (1) check ~/.ssh/auth..keys for 'command=' | 04:06 |
| ← Keiya left | 04:06 |
| ← ruediger left | 04:07 |
|
sitaram
| Oxf13: (2) treat the absence of "GL_RC" in ENV as indication that this did not come through gl-auth and exit 0 instead of dying <-- I have to think about this a little | 04:07 |
| → Keiya joined | 04:07 |
|
Oxf13
| k | 04:07 |
|
sitaram
| Oxf13: back in 10 min... quick meeting | 04:07 |
| → tstclair joined | 04:07 |
| ← tjafk left | 04:08 |
|
Oxf13
| sitaram: ok | 04:08 |
| → troyt joined | 04:10 |
| ← Twisted_Mentat left | 04:10 |
| ← akahn left | 04:12 |
| ← krawek left | 04:12 |
| → krawek joined | 04:12 |
|
Oxf13
| sitaram: question, any reason why one would copy the update hook into place rather than create symlinks back to a central copy (or hardlinks)? | 04:13 |
| → adamm joined | 04:13 |
| → hummingkid joined | 04:14 |
| ← agile left | 04:14 |
| → j416 joined | 04:17 |
| → patcito joined | 04:18 |
|
sitaram
| Oxf13: no real reason -- i think in the beginning I wasn't sure if it would work with symlinks (it does) and then it just stayed that way. | 04:18 |
|
Oxf13
| k | 04:19 |
|
| makes a lot of sense for us as we change the update hook it'd be nice not to have to copy it 9K times | 04:19 |
| → hipe joined | 04:20 |
| ← patcito left | 04:20 |
|
sitaram
| Oxf13: indeed; but you're bypassing all that anyway, no? | 04:20 |
|
Oxf13
| sitaram: I meant when we have to make chagnes to the hook itself | 04:20 |
| ← D[a]vey left | 04:20 |
|
Oxf13
| we're definitely going to have to have some local customization to it, to allow certain actions to be done locally like a mass branch | 04:21 |
|
sitaram
| Oxf13: ok let me think... you're creating the repos yourself, outside gitolite. Are you populating the hook also? | 04:21 |
|
| hooks/ I mean | 04:21 |
|
Oxf13
| (we're now considering running a second ssh server on a different port for admin shell work) | 04:21 |
|
| sitaram: yeah, that's the plan. Ideally we'd just symlink it back to a central version | 04:21 |
|
| I was just curious if you'd tried that and ran into problems with it | 04:22 |
|
sitaram
| Oxf13: I had not tried it in this context, but in some other context (testing for someone whether they can version control a hook!) I had tried symlinks and it worked | 04:22 |
|
Oxf13
| our hook would be version controlled, but installed via puppet most likely | 04:23 |
|
sitaram
| Oxf13: back to the bypassing hook for admin users issue: I'm trying to justify it from a security point of view -- just check through my reasoning if you could? | 04:25 |
|
| Oxf13: I reason that if the hook executed, then "git receive-pack" executed, which means he had some auth to do whatever he is trying to do. At that point, if GL_RC is not set, it means his auth bypassed gitolite, but an unauthorised person cannot get this far. Sound OK? | 04:25 |
| → patroy joined | 04:25 |
| → mark[oz] joined | 04:26 |
|
mark[oz]
| hey guys can i set up, for each branch its own origin path. | 04:26 |
|
| I seem to always switch to a branch, and git push origin master.. just out of habbit.. would be nice if i could git push.. and let git take care of where to go | 04:27 |
|
patroy
| Hi, I'm not sure if this is possible but is there a way to list all the diff commits between 2 branches? like: git diff my-branch other-branch ... but instead of the diff listing i would get a list of commits that are different between the 2 branches | 04:27 |
|
Oxf13
| sitaram: yeah, sounds pretty reasonable. | 04:27 |
|
| sitaram: basically file system ACLs would be allowing them the access, we'd just lose the per-branch ACL for this case | 04:28 |
|
sitaram
| yes | 04:28 |
|
Oxf13
| but I think we're going down the path of a different ssh server for shell users to shell in and do admin work | 04:28 |
|
sitaram
| Oxf13: and you dont care about the logging anyway | 04:28 |
|
Oxf13
| which would mean we're back to forcing all git access through gl-auth-command via authorized_keys | 04:28 |
|
sitaram
| Oxf13: well it's not a big change (I mean really, one line!) so let me know and I'll put it in when you need it. I'll keep it on hold for now | 04:29 |
|
Oxf13
| sitaram: I may still need it for repo wide admin tasks, such as mass branching | 04:30 |
|
sitaram
| ok I'll put it in | 04:30 |
|
bob2
| patroy: git log x...y? | 04:30 |
|
Oxf13
| but I could use the existance of a certain env variable or something | 04:30 |
| ← jmcantrell left | 04:30 |
|
sitaram
| Oxf13: I was going to just say "exit 0 unless exists $ENV{GL_RC};" right at the top -- that's really all you need | 04:31 |
|
Oxf13
| yeah, something like that | 04:31 |
|
sitaram
| Oxf13: I'll add it in big-repos, no problem | 04:31 |
|
| Oxf13: would you care if I refactored a bit and rewound it, by the way? | 04:32 |
|
Oxf13
| sitaram: nope, I can rebase | 04:33 |
|
sitaram
| ok; I usually avoid it unless the branch name starts with "pu", but if I do rewind a non-pu I will notify you by email too | 04:34 |
| ← patroy left | 04:35 |
| ← tedoc2000 left | 04:36 |
| ← AkumaStreak left | 04:37 |
| → Caelum joined | 04:38 |
|
Caelum
| where does Cygwin put commands like git-diff ? | 04:38 |
| ← whaley left | 04:38 |
| → secoif joined | 04:39 |
| → tedoc2000 joined | 04:39 |
| ← hummingkid left | 04:42 |
| ← tedoc2000 left | 04:44 |
| → r11t_ joined | 04:47 |
|
mark[oz]
| Caelum: do a "which git-diff" | 04:48 |
|
| and then go to linux, and live happily ever after | 04:48 |
| → ajonat joined | 04:49 |
|
Caelum
| yeah tried that :) | 04:49 |
|
mark[oz]
| git-diff is a virtual package i believe... you can get to it by typing "git diff" | 04:49 |
|
Caelum
| just trying to find the commands directory | 04:50 |
|
| which is /usr/lib/git-core on linux | 04:50 |
|
Oxf13
| or /usr/libexec/git-core/ | 04:50 |
| ← Bass10 left | 04:51 |
| ← r11t left | 04:51 |
| → ustunozgur joined | 04:51 |
| ← tstclair left | 04:52 |
| ← cubuanic left | 04:52 |
|
sitaram
| Oxf13: that username as argument thing: I'm making it conditional (if arg1 exists, use that, else use $USER). makes it easy to test ;-) OK with you? | 04:55 |
| → ba3 joined | 04:56 |
|
sitaram
| Oxf13: if user A's authkeys had "command=...gl-authcommand userB" (which it shouldn't...) then unpredictable things may happen. I'm going to ignore that possibility, ok? | 04:56 |
|
Oxf13
| sitaram: fine by me | 04:56 |
|
ba3
| hi? | 04:56 |
| → munichlinux joined | 04:57 |
|
ba3
| can some one talk 2 me???? | 04:57 |
|
munichlinux
| ba3, phone number :p | 04:57 |
|
ba3
| y? | 04:57 |
|
sitaram
| Oxf13: thanks | 04:58 |
|
ba3
| munichlinux whats up? | 04:58 |
|
munichlinux
| sitaram, y0! | 04:59 |
|
ba3
| can some body send me? | 05:00 |
| → babilen_ joined | 05:00 |
|
ba3
| '< bord @$ h3l | 05:01 |
|
| babilen whats up? | 05:01 |
| ← adamm left | 05:01 |
|
ba3
| h3llo? | 05:02 |
|
| :( | 05:02 |
| ba3 → lilmamabadash3ll | 05:04 |
| lilmamabadash3ll → cherry | 05:06 |
| cherry → Guest22572 | 05:06 |
| → Keiya_ joined | 05:07 |
|
sitaram
| munichlinux: hi! | 05:08 |
| Guest22572 → cherry | 05:09 |
| cherry → Guest47590 | 05:10 |
| → p0w3r3d joined | 05:11 |
| ← p0w3r3d left | 05:12 |
| → p0w3r3d joined | 05:12 |
| ← babilen left | 05:12 |
| ← jfkw left | 05:12 |
| Guest47590 → babyboop | 05:14 |
| ← r11t_ left | 05:14 |
| → r11t_ joined | 05:14 |
| ← scarabx left | 05:15 |
|
babyboop
| hello? | 05:15 |
| ← icwiener left | 05:15 |
| ← ustunozgur left | 05:15 |
| → scarabx joined | 05:16 |
| → davidddo joined | 05:17 |
| ← kukks left | 05:17 |
| ← daviddo left | 05:17 |
| ← cloudhead left | 05:18 |
| → help_jak_o joined | 05:25 |
| ← Keiya left | 05:26 |
| → hummingkid joined | 05:26 |
| ← Jak_o_Shadows left | 05:27 |
| ← help_jak_o left | 05:27 |
| → help_jak_o joined | 05:28 |
| ← help_jak_o left | 05:28 |
| → ankit9 joined | 05:28 |
| → help_jak_o joined | 05:28 |
| ← help_jak_o left | 05:28 |
| → Jak_o_Shadows joined | 05:29 |
| → p0w3r3d_ joined | 05:29 |
| ← p0w3r3d left | 05:31 |
| ← babyboop left | 05:31 |
| → EricInBNE joined | 05:32 |
| → eyzn joined | 05:33 |
| ← scarabx left | 05:35 |
| → scarabx joined | 05:36 |
|
eyzn
| hey there, i am trying to get more than one gitweb running on one system - but im struggling, i dont know how to tell apache to use different config-files (other than /etc/gitweb.conf) | 05:36 |
| → solofight joined | 05:37 |
| → rchavik joined | 05:38 |
| ← solofight left | 05:38 |
|
j416
| eyzn: so you are trying to run multiple web servers on the same system | 05:39 |
|
| ? | 05:39 |
|
eyzn
| no | 05:39 |
|
| one webserver | 05:39 |
| → solofight joined | 05:39 |
|
eyzn
| but i want to say, when you browse foo.bar/gitweb1 gitweb.cgi will be loaded with /etc/gitweb1.cfg, when you browse foo.bar/gitweb2 gitweb with gitweb2.cfg will be loaded - and so on | 05:40 |
| ← abadger1999 left | 05:42 |
|
eyzn
| actually i just want to list different repositories on each gitweb ($projects_list = projects1.list and projects2.list in /etc/gitweb.conf) | 05:42 |
|
j416
| ah... I see | 05:45 |
|
| I have never done that | 05:45 |
|
| sorry :/ | 05:45 |
|
eyzn
| np, ty for response anyway. perhaps someone else is able (and willing) to give me advise - i will idle some time :) | 05:47 |
| ← j416 left | 05:53 |
| → ycyuxin joined | 05:53 |
|
doener
| eyzn: the "local" config name is gitweb_config.perl, I have no idea in which directory it looks for it though | 05:55 |
|
| eyzn: that file overrides /etc/gitweb.conf | 05:55 |
|
| eyzn: and there are two env variables to override the filenames for those two | 05:55 |
|
eyzn
| mh yes, the README tells me that gitweb_config.perl is located by default in the same dir as gitweb.cgi - which isnt the case on my installation | 05:56 |
|
doener
| eyzn: GITWEB_CONFIG (default gitweb_config.perl) and GITWEB_CONFIG_SYSTEM (default /etc/gitweb.conf) | 05:56 |
|
| eyzn: so setting the env vars should do the trick I guess | 05:56 |
|
eyzn
| the env vars are global settings - how can i tell apache to use different ones? | 05:57 |
|
doener
| eyzn: http://httpd.apache.org/docs/2.0/mod/mod_env.html#setenv | 05:58 |
|
| eyzn: can be set per directory AFAIK | 05:58 |
|
eyzn
| oh cool! ty for the hint | 05:59 |
|
doener
| eyzn: there's also SetEnvIf that allows conditions, like checking the request URI | 05:59 |
|
eyzn
| ok | 05:59 |
|
| yes this works like a charm | 06:00 |
| ← p0w3r3d_ left | 06:00 |
| → abadger1999 joined | 06:02 |
| ← niki left | 06:04 |
| ← gambler left | 06:04 |
| → gambler joined | 06:05 |
| ← ehabkost left | 06:05 |
| ← r11t_ left | 06:07 |
| → dirk2 joined | 06:08 |
| ← reprore_ left | 06:08 |
| ← hummingkid left | 06:10 |
| ← defectiv left | 06:14 |
| → j416 joined | 06:15 |
| ← krawek left | 06:18 |
| → r11t_ joined | 06:18 |
| → thiago_home joined | 06:21 |
| → volodya joined | 06:21 |
| ← jrmuizel left | 06:23 |
| → jrmuizel joined | 06:23 |
| → meanburrito920_ joined | 06:23 |
| ← madewokherd left | 06:25 |
| → jceb joined | 06:28 |
| ← sergiodj left | 06:29 |
| ← mdel left | 06:31 |
| → roop joined | 06:31 |
| ← dgarr left | 06:31 |
| → tedoc2000 joined | 06:32 |
| ← wolog left | 06:35 |
| ← jrmuizel left | 06:35 |
| → bcardarella joined | 06:35 |
| ← ajonat left | 06:36 |
| ← sdboyer left | 06:37 |
| → Weasel[DK] joined | 06:40 |
| → jrmuizel joined | 06:41 |
| ← murbank left | 06:43 |
| → murbank joined | 06:43 |
| ← babilen_ left | 06:44 |
| → allenwei joined | 06:44 |
| → johnw joined | 06:52 |
| → thiana joined | 06:52 |
| ← robinr left | 06:52 |
| → vbgunz__ joined | 06:55 |
| → Ohan joined | 06:59 |
|
Ohan
| hi there guys | 06:59 |
|
| is there a way to check which remote branch the current local branch i'm on is tracking? | 07:00 |
| ← Jak_o_Shadows left | 07:00 |
| ← jrmuizel left | 07:00 |
|
rtyler
| Ohan: you can see all of them with `git remote show origin` | 07:02 |
|
Ohan
| oh, that is a neat command | 07:03 |
|
| had no clue it existed | 07:03 |
|
| thanks a ton rtyler ! | 07:03 |
| ← secoif left | 07:03 |
| → gaveen joined | 07:04 |
| → jrmuizel joined | 07:04 |
| ← dfr|work left | 07:05 |
| ← reidab left | 07:06 |
| → calin joined | 07:06 |
| → sde_antrix joined | 07:07 |
| ← jceb left | 07:08 |
| → engla joined | 07:08 |
| ← rchavik left | 07:08 |
| → hummingkid joined | 07:08 |
| ← SunilThaha left | 07:14 |
| ← dreiss left | 07:17 |
| → Sigma joined | 07:21 |
| ← jrmuizel left | 07:22 |
| → Zenopus_ joined | 07:23 |
| ← scarabx left | 07:25 |
| → dreiss joined | 07:27 |
| ← tedoc2000 left | 07:33 |
| → bentob0x joined | 07:37 |
| → ntoll joined | 07:37 |
| ← Zenopus left | 07:40 |
| Zenopus_ → Zenopus | 07:40 |
| L_R_N → LRN | 07:40 |
| → dermoth joined | 07:41 |
|
Rhonda
| When was git add patch editing added? In 1.6.6? | 07:43 |
| ← thiana left | 07:43 |
| → babilen joined | 07:45 |
| → loffe joined | 07:48 |
| → hebz0rl joined | 07:52 |
|
kristian42
| I still don't understand this: If I run git rerere while working, I can actually switch from merge commits to rebase at any given time in the future. But if I haven't run rerere, I will not be able to use conflict resolutions in merge conflicts when I try to rebase ? | 07:53 |
|
| Is there no-way to use the conflict resolutions from merges in rebase ? | 07:54 |
| ← munichlinux left | 07:54 |
| → st4rscr34m joined | 07:55 |
| → giallu joined | 07:55 |
|
st4rscr34m
| I am on the git site, and I'm trying to get Git for cygwin | 07:55 |
|
| But it just gives a link to cygwin | 07:55 |
|
kristian42
| st4rscr34m: You have to run cygwin setup | 07:55 |
|
zomg
| st4rscr34m: you may want to get msysgit instead | 07:56 |
|
| I had git on cygwin and for some reason git gui, gitk etc. worked awfully slow | 07:56 |
|
| ymmv | 07:56 |
| → d3xter joined | 07:56 |
|
godlygeek
| cygwin git worked much better than msysgit for me, personally | 07:56 |
|
st4rscr34m
| I want to try it this way as msys has pissed me off in the past | 07:56 |
|
kristian42
| st4rscr34m: I use cygwin git, but I tent to build from source, which is also really easy. just google it | 07:57 |
|
| st4rscr34m: The version in cygwin setup is old | 07:57 |
|
st4rscr34m
| k | 07:58 |
|
| thank you | 07:58 |
| → ph^ joined | 07:58 |
|
st4rscr34m
| is there docs or anything for building it? | 07:58 |
|
kristian42
| st4rscr34m: google "cygwin build git source" | 07:59 |
| ← JoeConyers left | 07:59 |
| ← hipe left | 08:00 |
| → babilen_ joined | 08:01 |
| → wolog joined | 08:02 |
| → Matjas joined | 08:02 |
|
st4rscr34m
| I know to google stuff... I'm not a complete idiot, everytime I do google something it is usually outdated information. | 08:02 |
|
| I'm checking though. | 08:02 |
| → sdboyer joined | 08:04 |
| → konr joined | 08:05 |
|
konr
| Is it a terribly bad idea to use git to backup my 10GB media collection? | 08:06 |
| → JoeConyers joined | 08:07 |
|
tetha
| I'm not sure if that is a good use case for a version control system... at least my music does not really change in small increments I would want to revert or investigate further | 08:08 |
|
st4rscr34m
| where should I put the git source? | 08:08 |
|
| within my cygwin folders? | 08:08 |
|
| It sounds good to me though tetha. | 08:09 |
|
tetha
| well, I'd rather use rsync or something like this | 08:09 |
| ← vbgunz__ left | 08:09 |
| VVSiz_ → VVSiz | 08:10 |
| → sdboyer_ joined | 08:10 |
|
tetha
| because the process of backing up media files is mostly copying binary blobs around | 08:11 |
|
konr
| tetha: is rsync easy to learn how to use? | 08:11 |
| ← sdboyer left | 08:12 |
| ← babilen left | 08:13 |
|
tetha
| konr: according to a quick google search and the manpage: yes. | 08:14 |
|
| in fact, articles for exactly your plan can be found | 08:15 |
| ← dreiss left | 08:16 |
|
zomg
| hey tetha, can you google something for me too? ;) | 08:16 |
|
konr
| tetha: haha, thanks for the clarification. Learning emacs modes is killing me - I can't stand digesting more technical information for now | 08:16 |
| → galderz joined | 08:17 |
| → cedricv joined | 08:17 |
| → ixti joined | 08:17 |
|
tetha
| zomg: http://www.google.de/search?q=yes+I+can!+%3AP | 08:18 |
| → codeFiend joined | 08:18 |
| ← st4rscr34m left | 08:18 |
| → pantsman joined | 08:18 |
|
zomg
| konr: learn Vim, it's easier and more fun ;) | 08:19 |
|
| Obviously if you go to #emacs people will say the opposite but... | 08:19 |
|
konr
| zomg: Haha, I'm originally a Vim user! | 08:19 |
|
| zomg: I've switched because of SLIME and org-mode | 08:19 |
| → reidab joined | 08:19 |
| → Jak_o_Shadows joined | 08:19 |
|
konr
| but I'm using VIPER, which emulates, sorta, vim | 08:20 |
|
tetha
| heh, like a year ago, I tried switching and eventually just reverted back | 08:20 |
| ← reidab left | 08:20 |
| ← codeFiend left | 08:20 |
| ← ph^ left | 08:22 |
|
d3xter
| konr: you should start using echo ;) | 08:24 |
|
| its easy to learn :D | 08:24 |
| ← pinnacle left | 08:26 |
| ← mbernstein left | 08:26 |
| ← hummingkid left | 08:27 |
| → sdboyer joined | 08:28 |
| ← sdboyer_ left | 08:29 |
| → johan-s joined | 08:30 |
| → zoniq_ joined | 08:32 |
| → psoo joined | 08:35 |
| → sdboyer_ joined | 08:36 |
| ← sdboyer left | 08:37 |
| → tbf joined | 08:38 |
| → munichlinux joined | 08:39 |
| → nitrate joined | 08:39 |
| → Chillance joined | 08:42 |
| ← bcardarella left | 08:42 |
| → tvw joined | 08:42 |
| ← fcuk112 left | 08:43 |
| ← Dashkal left | 08:44 |
| ← Chillance left | 08:45 |
| → Dashkal joined | 08:45 |
| → Chillance joined | 08:45 |
| → shruggar joined | 08:46 |
| ← Chillance left | 08:46 |
| → Chillance joined | 08:47 |
| ← Chillance left | 08:47 |
| → Chillance joined | 08:47 |
| ← Chillance left | 08:48 |
| → Chillance joined | 08:48 |
| ← Chillance left | 08:49 |
| → Chillance joined | 08:49 |
| ← mark[oz] left | 08:50 |
| ← Chillance left | 08:50 |
| → Chillance joined | 08:50 |
| ← Ohan left | 08:52 |
| → bousquet joined | 08:55 |
| → d0ugal joined | 08:59 |
| → ericboeh1 joined | 09:02 |
| → irc2samus joined | 09:05 |
|
irc2samus
| hi guys, how can I know from which commit a branch was started? | 09:06 |
|
j416
| irc2samus: git merge-base <branch> <otherbranch> | 09:09 |
| ← sdboyer_ left | 09:10 |
| ← ericboehs left | 09:13 |
| m-takagi → m-takagi_ | 09:14 |
|
doener
| irc2samus: git only has branch heads as first class entities. "Branches" as a series of commits aren't explicit, and thus there's no real starting point except for the root commits as far as git is concerned | 09:15 |
| ← zoniq_ left | 09:15 |
|
doener
| irc2samus: if you know from which branch head you forked, things like "git log upstream.." do the right thing | 09:15 |
| → zoniq_ joined | 09:16 |
|
doener
| irc2samus, j416: note that the merge-base will not necessarily be the commit you started your branch from | 09:16 |
|
j416
| doener: really? | 09:17 |
|
| aah of course | 09:17 |
|
| silly me | 09:17 |
|
| sorry | 09:17 |
|
doener
| j416: merges | 09:17 |
|
j416
| latest point of merge | 09:17 |
|
| yeah | 09:17 |
| → drizzd joined | 09:17 |
|
doener
| j416: "worst case": branch was merged upstream, then mergebase == branch tip | 09:18 |
| ← dirk2 left | 09:18 |
|
j416
| yeah | 09:18 |
|
doener
| which is pretty obviously not the commit you started from ;-) | 09:18 |
|
irc2samus
| hmm but from the branch perspective there is a point where the commits are the same as the branch where you started | 09:20 |
| → ArjenL joined | 09:20 |
|
irc2samus
| when you do git log <branch> --not <other> you gett all distinct | 09:21 |
|
| I need the opposite, the first common among those | 09:21 |
| ← psoo left | 09:21 |
|
doener
| irc2samus: > | 09:21 |
|
| > ciao | 09:21 |
|
| > Thies | 09:21 |
|
| > | 09:21 |
|
| > | 09:21 |
|
| ah crap... | 09:21 |
|
| http://git.pastebin.com/m1f62ab35 | 09:22 |
|
irc2samus
| what I'm trying to do with that is to diff the current commit from the point that branch started | 09:22 |
|
doener
| irc2samus: in that graph, can you tell which commits (from the "branch" perspective) belong to "branch" foo, and which to branch "bar"? | 09:22 |
|
irc2samus
| doener: no because you used all X :) | 09:23 |
| → Alfonzo joined | 09:23 |
|
doener
| irc2samus: would it make a difference if I had called them A,B,C...? | 09:23 |
| ← ycyuxin left | 09:23 |
|
doener
| irc2samus: but for your task at hand, you likely want: git diff A...B (three dots) | 09:23 |
|
| irc2samus: that gives a diff of B against the merge base of A and B | 09:24 |
|
irc2samus
| http://git.pastebin.com/d13b99398 I'm looking for B | 09:24 |
| → priidu joined | 09:24 |
|
doener
| irc2samus: which is like the diff of the changes a merge of B to A would introduce | 09:24 |
| → anno^da joined | 09:24 |
|
irc2samus
| ah that could be, let me try it | 09:24 |
|
j416
| I never understood that syntax | 09:24 |
|
| A...B v.s. A..B vs. A B | 09:25 |
|
| vs. ...are there more variants? | 09:25 |
| ← wanders left | 09:25 |
|
doener
| j416: for diff, "A B" and "A..B" are the same | 09:25 |
|
irc2samus
| A..B is B --not A if I'm not mistaken | 09:25 |
|
j416
| only for diff? | 09:25 |
|
doener
| j416: and ... is what I said above | 09:25 |
|
j416
| ok! | 09:25 |
|
doener
| j416: the range based commands (diff is end-to-end) work differently | 09:26 |
|
| j416: "git log A B" == everything reachable through either A or B or both | 09:26 |
|
j416
| ok! | 09:26 |
|
doener
| j416: "git log A..B" == "git log B ^A" == everything reachable through B, but not through A | 09:27 |
|
j416
| makes sense | 09:27 |
|
doener
| j416: and "git log A...B" == everything reachable through either A or B, but not both | 09:27 |
|
j416
| aah | 09:27 |
| → koke joined | 09:27 |
|
doener
| i.e. to show how they diverged | 09:27 |
|
j416
| so ... is like xor | 09:27 |
|
doener
| for log, yes | 09:27 |
|
j416
| ok | 09:27 |
|
| if we look at irc2samus's tree there | 09:28 |
|
doener
| j416: http://git.pastebin.com/m57a669be | 09:28 |
|
| j416: let's take that instead, a bit more interesting | 09:28 |
|
j416
| hm ok | 09:28 |
|
| git diff F...D | 09:28 |
|
| is the same as... | 09:29 |
|
| git diff F B ? | 09:29 |
|
irc2samus
| doener: I suppose it doesn't work if master has more commits right? | 09:29 |
|
doener
| j416: no, git diff B D | 09:29 |
|
j416
| ah | 09:29 |
|
doener
| j416: F...D == mergebase_of_F_and_D..D == B..D == B D | 09:29 |
|
irc2samus
| exactly like your last example | 09:30 |
|
j416
| doener: how do you remember this? | 09:30 |
|
irc2samus
| ah I was doing it reversed :P | 09:31 |
|
| so it master...foo | 09:31 |
|
j416
| so 'git diff diff F...D' I would use if I wanted to know the answer of what useful question? | 09:32 |
|
doener
| j416: 1) diff = end2end => differs from range commands, 2) A..B is pointless for diff, so "A..B" == "A B" (easy rule ;-)). 3) A...B for diff is like A..B for log (roughly speaking, but works to remember it) 4) just know how the DAG works, then it's obvious that A...B for diff uses the merge base | 09:32 |
|
johnw
| j416: which commits are in D that aren't in F | 09:32 |
|
j416
| aaaah | 09:32 |
|
doener
| johnw: that would be log, not diff | 09:32 |
|
j416
| hm | 09:32 |
|
johnw
| oh, you're right | 09:32 |
|
| i didn't read closely enough | 09:32 |
|
| j416 reading | 09:32 |
|
doener
| j416: "git diff A...B" ~~ Which changes would a merge of B to A try to introduce (roughly, of course it ignores everything that happened in A) | 09:33 |
|
irc2samus
| what I'm trying to get with this is a condensed diff of the changes on a branch from the point it started | 09:33 |
|
| I can't compare to master because there are more commits there | 09:33 |
|
| that's the useful question | 09:33 |
|
j416
| aha | 09:33 |
|
| good | 09:33 |
|
doener
| irc2samus: that's basically what A...B answers. | 09:33 |
|
j416
| I'm still figuring. :) | 09:33 |
| → girishr joined | 09:34 |
|
johnw
| doener: that's a good description of diff A...B | 09:34 |
|
doener
| irc2samus: just that it doesn't take the hypothetical starting point, but the merge base, because that gives an better answer | 09:34 |
| → ereslibre joined | 09:34 |
| ← konr left | 09:34 |
|
doener
| irc2samus: http://git.pastebin.com/m7f8103d5 | 09:35 |
|
irc2samus
| yes doener works perfectly, many thanks | 09:35 |
|
| this was what I was looking for | 09:35 |
|
j416
| doener: so in other words, git diff F...H can be "what have I done in this topic branch (whose head is F) since it last had contact with my master branch (whose head is H)?" | 09:35 |
|
doener
| irc2samus: with that example, if you used "diff against the starting point" (B), you'd also get the changes from the merge in H | 09:35 |
|
j416
| or was it H...F | 09:36 |
|
| arf | 09:36 |
|
doener
| irc2samus: and you don't really want them in your diff, they're not "your" stuff | 09:36 |
| → ngng joined | 09:36 |
|
j416
| H...F it is | 09:36 |
|
| sorry | 09:36 |
| ← ngng left | 09:36 |
|
doener
| irc2samus: so diff master...bar, diffs K against H (the merge base), and gets the right answer | 09:36 |
|
irc2samus
| ah true, in this case master wasn't merged into the branch yet | 09:36 |
|
| but if it would the result would've been wrong, you're right | 09:36 |
| → d3xter_ joined | 09:37 |
| ← koke left | 09:37 |
|
doener
| irc2samus: well, for topic branches that's a "bad" practise anyway, but for maintenance branches, it's common to be merged to master, and then master...maint is useful | 09:37 |
|
| irc2samus: I just took that as a random example for "why is the starting point wrong" | 09:37 |
|
| johnw: :-) | 09:38 |
|
j416
| git diff master...topic = "what changes that exist in my topic, are not yet merged into master" ? | 09:38 |
| ← d3xter left | 09:38 |
|
irc2samus
| we have a branch per task and merge master to them while working, once it finished the branch gets merged to master, is that ill advised? | 09:39 |
|
j416
| git diff master topic = "what's the difference between master and topic?" | 09:39 |
|
doener
| j416: yeah, that's closer (and basically what I said, because those changes are exactly the same that a merge would try to introduce) | 09:39 |
|
j416
| great! | 09:39 |
|
| thanks :) | 09:39 |
|
| then there is.. git log | 09:40 |
|
| j416 figures | 09:40 |
|
doener
| j416: yep, that last one is the plain difference, and quite often totally useless | 09:40 |
|
j416
| i use plain diff quite often | 09:40 |
|
| maybe it's because I never knew about ... :P | 09:40 |
|
doener
| irc2samus: it produces a rather cluttered history, because of the merges. I guess most of the merges, don't really add anything useful | 09:40 |
| → globalkeith joined | 09:40 |
|
doener
| irc2samus: e.g. if the topic branch works on X, and the changes on master dealt with A,B,C and D, the merge is totally uninteresting and pointless | 09:41 |
|
irc2samus
| they avoid merge conflicts | 09:41 |
|
doener
| irc2samus: uhm, how would they? | 09:41 |
|
| irc2samus: well, ok, they do, under certain conditions | 09:42 |
|
| irc2samus: namely: you happen to merge before you touch the line yourself. | 09:42 |
|
irc2samus
| yeah, whenever is possible | 09:42 |
|
| lesser conflicts == easier merge right? | 09:43 |
|
| it happens that on my project all work is quite related to each other, it rare to see non-conflicting branches | 09:44 |
|
| if they were totally unrelated I agree it has no advantage | 09:44 |
|
j416
| this is insane | 09:44 |
|
| I shouldn't have opened the man page on range specifications | 09:45 |
|
| lol | 09:45 |
|
| git log F^! D | 09:45 |
|
| o_O | 09:45 |
|
doener
| j416: F, but not its parents and D | 09:45 |
|
j416
| aah | 09:45 |
|
doener
| j416: it's similar to F..D, but includes F | 09:46 |
|
johnw
| we built doener out of pieces from the man pages, j416 | 09:46 |
|
j416
| "F^! D" is the same as "F D" | 09:46 |
|
| yeah | 09:46 |
|
| but with F | 09:46 |
|
| thanks .) | 09:46 |
|
| :) | 09:46 |
|
doener
| j416: "F D" for log, is a totally different thing | 09:46 |
| → jelmer_ joined | 09:46 |
| ← RobertMeta left | 09:46 |
|
j416
| "git log F! D" is the same as "git log F..D" ? | 09:46 |
| → rittyan joined | 09:46 |
|
doener
| j416: F^! D is "very similar" to F..D, but it includes F | 09:47 |
|
j416
| yeah | 09:47 |
|
| that's what I wrote | 09:47 |
|
| :) | 09:47 |
| ← rittyan left | 09:47 |
|
doener
| j416: no, you didn't have the dots | 09:47 |
|
j416
| oh... | 09:48 |
|
| right | 09:48 |
|
| i meant to write those | 09:48 |
|
| but the second one I wrote | 09:48 |
|
| F! D == F..D | 09:48 |
|
| is correct right? | 09:48 |
|
doener
| http://git.pastebin.com/m5c589453 | 09:49 |
| → psoo joined | 09:50 |
|
doener
| irc2samus: how high are the chances that you really happen to merge before you change a line? | 09:50 |
|
| irc2samus: In my experience, the distraction you get from merging over and over again, instead of focussing on your topic is worse than having to fix a few more conflicts | 09:51 |
| ← shuffle2 left | 09:51 |
|
j416
| thank you | 09:51 |
|
| hmm | 09:51 |
|
| the I^..L part confuses me though | 09:51 |
|
| would it have worked if I was not a merge commit? | 09:51 |
| → hummingkid joined | 09:52 |
|
doener
| j416: I^..L == F..L (I^ == first parent of I, which was F in this case) | 09:52 |
|
irc2samus
| well it's not mandatory but it really helps, as I've | 09:52 |
|
j416
| aaaaaaaah | 09:52 |
|
| first parent | 09:52 |
|
| right | 09:52 |
|
| makes sense :) | 09:53 |
|
| thank you | 09:53 |
| → NET||abuse joined | 09:53 |
|
irc2samus
| I require the branches passed to me to have master merged in, so I don't have to solve them myself | 09:53 |
| ← Keiya_ left | 09:53 |
|
irc2samus
| but if you receive a notice that something went to master while you were working | 09:54 |
| → [2]awnstudio joined | 09:54 |
|
irc2samus
| doing a merge there and solving few things is easier in my opinion that doing a merge at the end and solving lots | 09:54 |
|
[2]awnstudio
| git is pure awesome | 09:54 |
|
doener
| irc2samus: hm, so you "serialize" merges? (Because obviously when you merge A to master, then B won't have master merged anymore) | 09:54 |
|
johnw
| doener: what do you do for a living? | 09:55 |
|
j416
| johnw: he helps people with git on irc I think. | 09:55 |
|
| :D | 09:55 |
| → cenuij joined | 09:55 |
|
irc2samus
| each branch is a task that when finished is supposed to go to master, if the merge is a fast-forward I don't mind but if there's merge conflicts I require the person who sent it to solve those for me (because it's his code and I probably won't know who to anyway) | 09:56 |
|
doener
| johnw: at the moment: Nothing. I'm living on savings, working on my diploma thesis (due on Jan 12th) | 09:56 |
|
johnw
| ahhh | 09:56 |
|
irc2samus
| so if you're working on B and A gets merged to master then by the time you finish on B you'll also have to had A there | 09:57 |
| → [3]awnstudio joined | 09:57 |
| [3]awnstudio → awnstudio | 09:57 |
|
doener
| irc2samus: I dislike that, because it easily ruins the first parent history, which I learned to love, if crafted "carefully" | 09:58 |
|
| irc2samus: it's quite lovely how "git log --first-parent ..origin/master" get's you an high-level overview of what happened | 09:58 |
|
j416
| ooh | 09:58 |
|
doener
| IIRC I lately wrote some email about that, let's see if I can find it | 09:58 |
|
j416
| doener: what are you writing? | 09:59 |
|
irc2samus
| haven't used that | 09:59 |
|
j416
| is it for a PhD? | 09:59 |
|
doener
| j416: no, German "Diplom", comparable to a MSc | 09:59 |
|
j416
| ok! :) | 09:59 |
|
| so I take it you are studying CS? | 09:59 |
|
irc2samus
| I will take a look, but in any case it quite troublesome to make people commit to the right branch anyway I doubt we could do anything "carefully" | 09:59 |
| ← jelmer left | 09:59 |
|
j416
| what is the title of your dissertation? :) | 09:59 |
|
doener
| j416: connecting an ECA (embodied conversational agent) to Skype | 09:59 |
|
j416
| ph | 10:00 |
|
| oh* | 10:00 |
|
| interesting | 10:00 |
|
doener
| j416: "Skyping Max" | 10:00 |
|
j416
| hah | 10:00 |
|
doener
| (Max is the name of the ECA) | 10:00 |
| → SunilThaha joined | 10:00 |
|
doener
| The video part was "fun" | 10:00 |
| → Axius joined | 10:00 |
|
j416
| i took a course called Human Machine Interaction where we researched stuff like that | 10:00 |
| ← pantsman left | 10:00 |
| → nmouad joined | 10:01 |
|
j416
| it was... educational | 10:01 |
|
irc2samus
| (we'll soon find out that the real human is max and doener really came from the man pages :P) | 10:01 |
| → yhager joined | 10:01 |
|
j416
| :D | 10:01 |
| → void_0 joined | 10:05 |
|
irc2samus
| oh I see what you mean by the log --first-parent now | 10:05 |
|
johnw
| hehe | 10:05 |
|
doener
| irc2samus: hm, in this mail I mentioned something along those lines, but IIRC it's not the one I had in mind (I just can't find the right one): http://article.gmane.org/gmane.comp.version-control.git/132268 | 10:05 |
|
irc2samus
| but it's not aplicable to my case, I do small fixes directly on master | 10:06 |
|
| maybe I should tag each merge that integrates a branch | 10:06 |
|
doener
| irc2samus: I don't think that _everything_ needs to go to a branch, and master must be merge commits only in its first parent ancestry. | 10:06 |
|
| irc2samus: having small commits/fixes done directly on master is fine. Creating merge commits for those would just clutter things up. | 10:07 |
|
| irc2samus: what I try to avoid is a) "massive" fast-forwards, b) fast-forwards that "flip" the first parent | 10:08 |
| ← void_0 left | 10:08 |
|
doener
| irc2samus: e.g.: git checkout topic; git merge master; git checkout master; git merge topic | 10:08 |
|
| irc2samus: that would ff master to topic, so the previous "master" history becomes second parent | 10:09 |
| → wanders joined | 10:09 |
|
doener
| irc2samus: but I see how that can be very very hard to achieve if you're not quite as passionate about it ;-) | 10:10 |
|
irc2samus
| yes, there's a lot of way to work really, in my case I prefer that because I'm the only one merging to master | 10:10 |
| ← hummingkid left | 10:11 |
|
irc2samus
| also my team is not much experienced with git (I'm the most "advanced" and I know very little, so that I'm here asling) so I've tried to create a policy that eases the usage as much as possible | 10:11 |
|
doener
| yeah, and there's nothing "inherently" wrong with it. I just think that it makes the history uglier/harder to use (being able to really make use of the history is one of my favourite "features" of git, as you might have noticed *g*) | 10:12 |
|
| but it's totally up to taste, and maybe I'm just too dumb for complex histories ;-) | 10:12 |
| ← sitaram left | 10:13 |
|
irc2samus
| to put an example, recently I did some refactoring and split some files into several others; anyone working on those files will see a deleted/modifed merge conflict | 10:13 |
| → sitaram joined | 10:13 |
| ← Alfonzo left | 10:13 |
|
irc2samus
| so in that case the sonner they integrate the refactoring the less they have to solve | 10:13 |
| → curtana joined | 10:13 |
| ← [2]awnstudio left | 10:14 |
|
irc2samus
| in fact, my original question was because of this, I wanted to see the changes a guy did on his branch (which he has merged master yet) but without seeing my own refactoring | 10:14 |
|
| btw I use nomral diff a lot too, in this case it showed me the conflict so I knew I needed a different diff | 10:15 |
|
doener
| irc2samus: well, the diff is against the refactored version now ;-) | 10:15 |
|
irc2samus
| exactly, I needed to see that but I also needed to see the ... diff | 10:16 |
| → snearch joined | 10:16 |
| ← hebz0rl left | 10:16 |
|
doener
| irc2samus: hm? The ... diff is against the refactored version (I guess he merged that, right?) | 10:16 |
|
irc2samus
| no he didn't | 10:17 |
|
doener
| irc2samus: ah ok, then nevermind that comment ;) | 10:17 |
| ← johnw left | 10:17 |
|
irc2samus
| he'll have a hell of a merge :) and he's the one who complained the most about git | 10:18 |
|
doener
| irc2samus: but in fact, such a refactoring would probably make me rebase my branch (assuming that it's still private), to avoid a "logical" break in the patch series | 10:18 |
|
| irc2samus: I usually review by doing "git log --reverse -p ..topic" (i.e. patch by patch), and that refactoring would likely "disrupt" me | 10:19 |
| ← r11t_ left | 10:19 |
|
irc2samus
| I think having a clean history requires a very resposible use of git, which is great but this particualr guy is someone who used svn for +5 years and didn't even knew he could revert to a previous version | 10:19 |
|
| I really can't ask him to do anything fancy | 10:20 |
|
doener
| :-/ | 10:20 |
|
irc2samus
| but hey! I made him use git at least :D | 10:20 |
|
doener
| heh :-) | 10:20 |
|
| I guess he's in the "Blablabla I must write something" commit message camp? *g* | 10:20 |
|
| anyway, I guess we both should get some work done... At least I should... | 10:21 |
|
irc2samus
| oh man! just like that :) | 10:21 |
|
| yeah go ahead, many thanks for your help! | 10:21 |
|
doener
| you're welcome | 10:23 |
| → Alfonzo joined | 10:24 |
| ← allenwei left | 10:27 |
| → gitte joined | 10:29 |
| ← Axius left | 10:29 |
| ← Theravadan left | 10:32 |
| → intripoon joined | 10:32 |
| erlehmann_ → erlehmann | 10:34 |
| ← nmouad left | 10:35 |
| ← psoo left | 10:36 |
| → psoo joined | 10:37 |
| → pdhxni joined | 10:37 |
| → drizzd_ joined | 10:37 |
| → Haarg joined | 10:39 |
| → Hendu joined | 10:39 |
| → _Krieger_ joined | 10:42 |
|
_Krieger_
| what can i do if i want to initiate a repo (svn/git/etc) and to store existing git repo to it as a subdir, saving all its commits data? | 10:45 |
|
ludde
| can I get 'git gui' not to print the annoing 'The repository has 500 loose items, please compress it' | 10:45 |
|
wereHamster
| _Krieger_: man gitsubmodule | 10:45 |
|
| hm, git-submodule | 10:45 |
|
j416
| ludde: run git gc | 10:46 |
|
ludde
| j416: well I know, but i don't care about gcing, it just takes too much time | 10:46 |
|
j416
| ludde: then live with that message :) | 10:47 |
|
ludde
| there should be a way to turn off the message, and have it handled by git gc --auto instead. | 10:47 |
|
j416
| ludde: why don't you run git gc before you go to lunch or so? | 10:47 |
|
| you don't need to watch it. :) | 10:47 |
| → ehabkost joined | 10:47 |
|
j416
| your repo will work faster after you do it | 10:47 |
|
| ludde: well, git gc --auto is a bit "flawed" so to speak | 10:48 |
| → lmbois joined | 10:48 |
|
j416
| it works good for repos with many objects | 10:48 |
|
| but not as good for repos with few objects | 10:48 |
|
ludde
| it took just 10 seconds to do git gc.. | 10:48 |
|
j416
| because of the way it determines it it is to run gc or not | 10:48 |
|
| ludde: see.. | 10:48 |
|
| doesn't take that much time | 10:49 |
|
ludde
| the point is, i don't want the useless message to interrupt my workflow | 10:49 |
|
j416
| you don't need to run git gc that often | 10:49 |
|
ludde
| i lose focus | 10:49 |
|
j416
| if you are really annoyed by it and want automatic git gc to run every now and then (and don't mind waiting for that to finish), you can change its threshold value | 10:50 |
|
| I don't remember if there is some lowest-possible threshold, but I don't think so | 10:50 |
| ← Alfonzo left | 10:50 |
|
j416
| ludde: git config --global gc.auto 10 | 10:51 |
|
| that will run git gc when you have at least 10 loose objects | 10:51 |
|
| optimally | 10:51 |
|
ludde
| ok | 10:51 |
|
j416
| but with the way it determines the number of loose objects, it may well build up to many more objects | 10:51 |
|
| but at least it's better than nothing then, in your case. | 10:52 |
|
| (because git gc --auto is called in various places, it must be able to determine very fast if a gc is needed or not, so it can't spend time accurately counting all objects) | 10:52 |
|
| so it takes the object count in one bucket (happens to be the number of objects in the directory .git/objects/17/) and interpolates the total number of objects based on that | 10:53 |
|
| sha1 is distributed evenly, but for very few objects, there will obviously be variation | 10:54 |
|
| </end lecture> | 10:54 |
| → flatline joined | 10:55 |
|
flatline
| hello, just configured a server and I need to connect to git via a gui, isn't the adress required to connect: git://server_ip/example.git | 10:56 |
| ← drizzd left | 10:56 |
| ← aisipos left | 10:56 |
| → Alfonzo joined | 10:56 |
| ← lmbois left | 10:57 |
| → Geralt joined | 10:59 |
| → bry joined | 11:01 |
| ← Geralt left | 11:01 |
| ← pdhxni left | 11:02 |
|
bob2
| flatline: is that hwo you configured the git daemon? | 11:03 |
|
gitte
| flatline: if you want to push, no. | 11:04 |
| → pheaver joined | 11:06 |
| → struberg joined | 11:07 |
|
_Krieger_
| I need to check out upstream sources tree as explained here | 11:07 |
|
| http://www.kamailio.org/dokuwiki/doku.php/install:kamailio-3.0.x-from-git | 11:07 |
|
| , add my modules and store them in local repo, and pull updates from upstream | 11:07 |
|
| Could you direct me to right commands for having merged repo? Thanks | 11:07 |
| ← laj left | 11:08 |
| ← wolog left | 11:08 |
| → inswow joined | 11:09 |
| → nkuttler joined | 11:12 |
| ← froschi1 left | 11:12 |
| ← Dashkal left | 11:13 |
| → aleksei joined | 11:14 |
| → laj joined | 11:15 |
| → Keiya joined | 11:16 |
| → xbqutc joined | 11:17 |
| jelmer_ → jelmer | 11:18 |
| → niki joined | 11:20 |
| ← upul left | 11:20 |
| ← Alfonzo left | 11:21 |
| ← giallu left | 11:24 |
| → hummingkid joined | 11:28 |
| → christophsturm joined | 11:28 |
| ← inswow left | 11:29 |
| ← irc2samus left | 11:30 |
| aleksei → the_weasel | 11:32 |
| → vals_ joined | 11:36 |
| → vbgunz__ joined | 11:36 |
| → WinterWeaver joined | 11:37 |
| ← pheaver left | 11:38 |
| ← tango_ left | 11:38 |
| vals_ → tango_ | 11:38 |
| ← jinzhu1 left | 11:39 |
| ← d3xter_ left | 11:39 |
| ← anno^da left | 11:39 |
| → Alfonzo joined | 11:39 |
| ← hummingkid left | 11:41 |
| → cilly joined | 11:43 |
| ← drizzd_ left | 11:45 |
| → drizzd joined | 11:45 |
| → cedric_ joined | 11:46 |
| ← psankar left | 11:47 |
| ← snearch left | 11:47 |
| ← cedricv left | 11:49 |
|
jones--
| Hi. How do I enable bash completion for git? | 11:49 |
|
| Like "git br<TAB> foo<TAB>" | 11:49 |
| → hugoxrosa joined | 11:50 |
|
d0ugal
| jones--, what version are you using? its enabled by default (for me anyway) | 11:52 |
|
jones--
| d0ugal: 1.6.3.3 | 11:52 |
| ← xbqutc left | 11:52 |
| ← zoniq_ left | 11:52 |
| → noss joined | 11:52 |
|
d0ugal
| jones--, hmm well I'm not sure why you don't have it - I'm running 1.6.0.4 | 11:52 |
|
noss
| hey, if someone have been using git-svn and i forked that git, am i shit-out-of-luck if i want to continue git-svn on where my origin left off? | 11:53 |
| → zoniq_ joined | 11:53 |
|
Ilari
| d0ugal: For bash, you need to install it (or have OS package manager install it). | 11:53 |
|
jones--
| Ilari: what's the package name in debian/ubuntu? | 11:54 |
|
Ilari
| noss: You would need to know what settings the previous user used. Layouts, URLs, etc... | 11:54 |
|
d0ugal
| jones--, git-completion perhaps | 11:55 |
|
jones--
| Note, selecting git-core instead of git-completion | 11:55 |
|
| git-core is already the newest version. | 11:55 |
| → capndiesel joined | 11:55 |
|
Ilari
| jones--: Maybe it depends on bash completion package (but doesn't say it)? | 11:56 |
|
jones--
| Ilari: installing bash-completon-lib. Let's see if it helps. | 11:57 |
|
| Doesn't look like it. | 11:57 |
| ← _Krieger_ left | 11:59 |
| ← niki left | 12:01 |
| → Ryback_ joined | 12:02 |
| ← WinterWeaver left | 12:05 |
| ← zoniq_ left | 12:07 |
| ← bousquet left | 12:07 |
| ← tetha left | 12:07 |
| → tetha joined | 12:07 |
| → zoniq_ joined | 12:08 |
| → v0n joined | 12:09 |
| ← __iron left | 12:11 |
| ← andres left | 12:12 |
| ← Jak_o_Shadows left | 12:13 |
| ← LRN left | 12:16 |
| → Axius joined | 12:17 |
| → pheaver joined | 12:19 |
|
noss
| how do i get the list of git merge conflicts again? | 12:22 |
| ← godlygeek left | 12:22 |
| ← stuffcorpse left | 12:22 |
| → cedricv joined | 12:23 |
|
Ilari
| noss: 'git status' shows them. As does 'git ls-files -u'. | 12:24 |
|
noss
| thanks | 12:25 |
| ← calin left | 12:25 |
| ← cedric_ left | 12:26 |
| → LRN joined | 12:26 |
| → andres joined | 12:27 |
| → parigaudi joined | 12:29 |
| → giallu joined | 12:30 |
| → fcuk112 joined | 12:30 |
| ← cedricv left | 12:32 |
| → b0fh_ua joined | 12:35 |
| → WinterWeaver joined | 12:36 |
|
b0fh_ua
| Hi there! I want to find at which revision the file A (one customer had sent me) was committed to GIT | 12:36 |
| ← RobertLaptop_ left | 12:36 |
| ← ia left | 12:39 |
| ← awnstudio left | 12:39 |
| → ia joined | 12:40 |
| ← fcuk112 left | 12:40 |
| ← WinterWeaver left | 12:40 |
| → froschi joined | 12:40 |
|
Ilari
| b0fh_ua: That's pretty nasty to do. Search for history for commits that affect it and compare the hashes. | 12:46 |
|
bremner_
| How do I make a "Reviewed-By" line to a patch with git-send-email? by hand with --annotate? | 12:47 |
| ← pheaver left | 12:49 |
| ← dsch04 left | 12:49 |
| → hummingkid joined | 12:51 |
| → pheaver joined | 12:51 |
| → Yuuhi joined | 12:52 |
| ← selckin left | 12:53 |
| → selckin1 joined | 12:53 |
| → auslander joined | 12:54 |
| ← b0fh_ua left | 12:54 |
| → dsch04 joined | 12:54 |
| ← munichlinux left | 12:55 |
| → lucsky joined | 12:56 |
| → the_weas1l joined | 12:57 |
| ← struberg left | 12:58 |
| → Isenn joined | 12:58 |
|
Isenn
| Hello. I got a question. I am using git and can update my branch with git pull and push | 12:59 |
|
| but i have another branch called "test" that i do not know how to pull from the other host | 12:59 |
| ← Haarg left | 13:00 |
|
Isenn
| i am able to pull/push within my master branch from both hosts. But i do not know how to pull the "test" branch. Since you can not list it with git branch | 13:00 |
| → LeonBrussels joined | 13:00 |
| → irc2samus joined | 13:00 |
|
Grum
| you can list it with git branch -r | 13:01 |
|
Isenn
| Hmm, i got two different outputs from my hosts | 13:02 |
| → petercooper joined | 13:02 |
|
LeonBrussels
| Hi folks, maybe I am too stupid or I am missing something obvious, but I have a project here that will be split up in 2 (specifically I decided that my avr circuit board needs an auxiliary board which is very similar) I figured I could just create a new directory and move all the files in the with git mv file.c new_dir but all the history is lost, i.e. when I type git log file.c it doesn't show me the history... | 13:03 |
| → cedricv joined | 13:03 |
| → methanol joined | 13:03 |
|
Grum
| Isenn: yes the list can be different if you didnt fetch from the remote repo recently | 13:03 |
|
| or if you still have an old branch which no longer exists | 13:04 |
|
| Isenn: normally you make a local branch which 'tracks' the remote one (so git knows where to push/pull updates to/from) | 13:04 |
| ← cedricv left | 13:04 |
|
Isenn
| Ok | 13:04 |
|
Grum
| Isenn: git checkout -t origin/<branch> will create the local branch named the same as the remote and setup tracking | 13:04 |
|
| LeonBrussels: you didnt copy the .git dir | 13:05 |
|
| and erm, moving all files with git mv? | 13:05 |
| → berndj joined | 13:05 |
|
cbreak
| LeonBrussels: even with git log -C? | 13:05 |
|
LeonBrussels
| Grum: I was thinking I could have both boards in one repo since they are one unit... | 13:06 |
|
Isenn
| Grum: If i have created "local" branch with "git branch test" how to i make that branch avaiable on another computer that have issued "git clone ssh://host" | 13:06 |
| → cedricv joined | 13:06 |
|
Isenn
| Grum: the only branch we share right now is the master | 13:06 |
|
LeonBrussels
| cbreak: Not even with git log -C... | 13:06 |
|
cbreak
| weird | 13:07 |
|
| if you do a git log without mentioning any file | 13:07 |
|
| do you see the history? | 13:07 |
|
Grum
| Isenn: it depends on what ssh://host is, is that where you cloned from as well? | 13:07 |
|
| LeonBrussels: what move command did you do? did you accidently moved the files out of your repo? | 13:07 |
|
LeonBrussels
| cbreak: Yes, but not if I mention a specific file... | 13:08 |
| ← cedricv left | 13:08 |
|
cbreak
| what does git log --stat -C say? | 13:08 |
| → cedric_ joined | 13:08 |
|
cbreak
| the files you moved should be there as delete and recreate, or as move | 13:08 |
|
Grum
| Isenn: if so; then you first have to publish your local test branch; if your remote is called origin (which it will be if you cloned from it) then do: git push origin test (ea: push branch 'test' to remote 'origin') | 13:08 |
|
Isenn
| Grum: Both client have issued "git clone ssh://user@host/srv/git/project" | 13:08 |
| → maattd joined | 13:09 |
|
Isenn
| ./srv/git/project "is an "git clone --bare project" and then deleted the "orginal" project | 13:09 |
| ← pheaver left | 13:09 |
|
Grum
| Isenn: then from the other clone a: 'git fetch' will make the branch show up under 'git branch -r' after which you can do a 'git checkout -t origin/<branch>' | 13:09 |
|
LeonBrussels
| Grum: In the root of the repo: mkdir mainboard; git mv file.c mainboard; git commit -a; git log mainboard/file.c; I don't see the history of file.c | 13:09 |
|
Grum
| LeonBrussels: no ancient version of git i presume? | 13:10 |
|
LeonBrussels
| Grum: git version 1.6.5.2 | 13:10 |
| ← cedric_ left | 13:10 |
|
wereHamster
| LeonBrussels: git log --follow | 13:11 |
|
Isenn
| Grum: http://pastebin.com/m183dce | 13:11 |
|
| It feels like i have done something wrong | 13:11 |
| ← the_weasel left | 13:11 |
|
Isenn
| The only diff in .git/config is url (file:// and ssh://) | 13:12 |
| → flaguy48 joined | 13:13 |
|
Grum
| and wereHamster hits the nail on the head :D | 13:13 |
| → tbuehlmann joined | 13:13 |
|
Grum
| Isenn: oh that is quite normal and not a problem, no idea what causes it but it is the same ;) | 13:13 |
| ← Axius left | 13:14 |
| ← Adlai left | 13:14 |
|
Isenn
| Grum: do i need to do anything to commit the new branch? | 13:14 |
|
| I will google but thanks for your help | 13:14 |
| → Adlai joined | 13:14 |
|
divVerent
| looks I lost my own branch when git push --mirror from the git-svn tree to the "published" tree | 13:15 |
|
LeonBrussels
| wereHamster: It works! Let's see if I can also go back in history | 13:15 |
|
divVerent
| (not important data, though) | 13:15 |
|
| so, what should I do instead to push from the private git-svn tree to the published bare repo? | 13:16 |
|
| not using --mirror means the remote does not get the branches the git-svn repo has | 13:16 |
|
| which makes the bare repo kinds useless | 13:16 |
| ← Lisimba left | 13:18 |
|
Isenn
| The problem is i have local branches. How do i make them remote so other people can use them | 13:18 |
|
Grum
| Isenn: < Grum> Isenn: if so; then you first have to publish your local test branch; if your remote is called origin (which it will be if you cloned from it) then do: git push origin test (ea: push branch 'test' to remote 'origin') | 13:19 |
|
cbreak
| all local branches are remotes for people who pull from that local repository | 13:19 |
| → icwiener joined | 13:19 |
|
Grum
| except that Isenn cloned so he'll have to push it over first :) | 13:19 |
|
Isenn
| Grum: Thanks. No i have "commited" my local created branch :) | 13:20 |
| ← JasonWoof left | 13:20 |
|
LeonBrussels
| Grum: Example: http://gist.github.com/259489 git log --follow works, git checkout doesn't | 13:21 |
| → context joined | 13:21 |
|
Grum
| how does git checkout not work? | 13:22 |
| ← Wolfman2000 left | 13:22 |
|
Grum
| +LeonBrussels | 13:22 |
| ← loffe left | 13:23 |
|
LeonBrussels
| Grum: I appended it to the gist | 13:24 |
|
Grum
| LeonBrussels: i dont see the problem? :D | 13:25 |
| → _Krieger_ joined | 13:25 |
| → aronchi joined | 13:26 |
|
LeonBrussels
| Grum: I would like to be able to checkout an older version of file.c while it was still in / | 13:26 |
|
aronchi
| how can I specify the user on the bare server in config? | 13:26 |
|
| git pull uses the current user but I need to use another on ssh | 13:26 |
| ← ankit9 left | 13:27 |
|
aronchi
| I've found it | 13:27 |
|
Isenn
| Grum: if i checkout "git checkout origin/test" Do i need to create a local branch for it? Since now it is called (no branch) | 13:28 |
| ← hummingkid left | 13:30 |
|
LeonBrussels
| Grum: Maybe this is the wrong approach and I should just cp the repo and then start two parallel repos... | 13:30 |
| → fcuk112 joined | 13:30 |
| ← nothingHappens left | 13:31 |
| → tstclair joined | 13:31 |
| the_weas1l → the_weasel | 13:32 |
| ← jayne left | 13:33 |
| ← jspiros left | 13:33 |
| ← up_the_irons left | 13:33 |
| ← selckin1 left | 13:33 |
| → Matjas_ joined | 13:33 |
| → selckin1 joined | 13:34 |
| → jspiros joined | 13:34 |
| → up_the_irons joined | 13:34 |
| → jayne joined | 13:34 |
| ← vbgunz__ left | 13:34 |
|
Grum
| LeonBrussels: you want to see the contents or checkout the commit which had the file? because if you want to see the contents; git show will do | 13:35 |
|
| Isenn: you should really read what i've typed before; but here it is again: you should do: "git checkout -t origin/test" to make a local branch tracking the remote one | 13:36 |
|
| LeonBrussels: i have still not much of an idea of what you want to do | 13:36 |
|
| LeonBrussels: you are using mv and are talking about cp a line later | 13:37 |
|
LeonBrussels
| Grum: Bascially in the example I pasted I want to be able to revert back to an empty file (ca7eeae658329cf8544c9ad8f9746612466a8189, file.c was still in /) while keeping file.c in slac/ | 13:37 |
|
| sry | 13:37 |
|
| rofl/ | 13:37 |
|
Grum
| LeonBrussels: why do you want to revert? what could be the practical application for that? :D (oh and keep in mind 'git revert' is something completely different ;D) | 13:38 |
|
cbreak
| you can go back to any revision with git checkout | 13:38 |
|
Grum
| yeah but he tries to checkout a specific file | 13:38 |
|
cbreak
| I think that should be possible too | 13:38 |
|
| git checkout something -- somefile.h | 13:38 |
|
Grum
| yes but i'm guessing he wants something else frmo it :) | 13:38 |
|
LeonBrussels
| Well you are right the practical application is not a lot, I can see the contents of the file at any point in time anyway | 13:39 |
| selckin1 → selckin | 13:39 |
|
divVerent
| As I am obviously doing it wrong here... is there a website which explains how to make a public git mirror of a svn repo in which one can create one's own private branches? | 13:39 |
|
Grum
| yes; git show $SHA:<file> | 13:39 |
|
LeonBrussels
| And yes, I guess I meant checkout instead of revert... | 13:39 |
| ← selckin left | 13:39 |
|
divVerent
| and, also, what about a bidirectional setup that automatically will dcommit if someone pushes to the git repo? | 13:39 |
|
Grum
| divVerent: why not simply use git? :) | 13:39 |
|
divVerent
| because upstream is svn | 13:39 |
|
Grum
| then upgrade to git? ;) | 13:39 |
|
divVerent
| not that easy :P | 13:40 |
|
| maybe later, once I got this running | 13:40 |
|
Grum
| but easier than setting up what you want without having all kinds of edgecases :D | 13:40 |
|
LeonBrussels
| OK I think I will just do it like that, I still keep all my history... | 13:40 |
|
divVerent
| well | 13:40 |
| → selckin joined | 13:40 |
|
divVerent
| I can hardly convince the rest of the project of git | 13:40 |
| → bcardarella joined | 13:40 |
|
divVerent
| if I run into problems all the time myself | 13:40 |
|
| I first have to get it to work perfectly here | 13:40 |
|
Grum
| ok so what you want is different; you dont actually want that | 13:41 |
|
divVerent
| but well, an unidirectional mirror suffices for me | 13:41 |
|
Grum
| but just a local something which works ;) | 13:41 |
|
divVerent
| well, a bit more | 13:41 |
|
| I want to make a branch that is managed in git only | 13:41 |
|
| and that others can pull from | 13:41 |
| ← Matjas left | 13:41 |
|
Grum
| but others are not using git? | 13:41 |
|
divVerent
| but the rest of the git repo should mirror the svn 1:1 | 13:41 |
|
| not yet | 13:41 |
|
| maybe will, once I have this running | 13:41 |
|
cbreak
| what will they pull with? | 13:41 |
|
Grum
| git apparently | 13:42 |
|
divVerent
| currently, svn, future, git | 13:42 |
|
cbreak
| svn can not pull from git | 13:42 |
|
divVerent
| I know you can commit to svn using git-svn, even if you had originally cloned a git mirror of the svn repo | 13:42 |
|
| so that would work for a transition too | 13:42 |
|
Grum
| divVerent: you will save yourself loads of trouble if you just convert to git without an intermediate state | 13:42 |
|
divVerent
| but how to convince everyone, if I have problems with git? | 13:42 |
|
cbreak
| problems(git) < problems(svn) | 13:43 |
|
divVerent
| for me? not yet | 13:43 |
|
| currently, svn works perfectly for me, and git makes trouble... | 13:43 |
|
| which is due to me not understanding it well yet | 13:43 |
|
Grum
| divVerent: i dont think that will ever work; afaik the metainformation needed to git-svn dcommit is not propagated when doing a clone | 13:43 |
|
divVerent
| okay, then let's forget that part | 13:43 |
|
| I then JUST want a git repo that mirrors the svn | 13:43 |
|
| and allows pusing into extra branches | 13:43 |
|
Grum
| git-svn clone svn:... | 13:44 |
|
divVerent
| which others can pull from | 13:44 |
|
| yes, that is how far I am :P | 13:44 |
|
| I already have this: | 13:44 |
|
Grum
| well then you are done ;) | 13:44 |
|
divVerent
| - git-svn clone'd repo | 13:44 |
|
| - a git clone'd repo from that | 13:44 |
|
| - latter is bare, and published | 13:44 |
|
Grum
| that repo has no branches | 13:44 |
|
| the last clone | 13:44 |
|
divVerent
| right, I found a script for that on the web | 13:44 |
|
| that copies the remotes to the heads | 13:44 |
| ← methanol left | 13:44 |
|
Grum
| hehe, do know that is major hackish? | 13:44 |
|
divVerent
| sure I do | 13:44 |
|
Grum
| good ;) | 13:45 |
|
divVerent
| but how else is it supposed to work? | 13:45 |
|
Grum
| not | 13:45 |
|
| git is supposed to work as git | 13:45 |
| ← jayne left | 13:45 |
| ← jspiros left | 13:45 |
| ← up_the_irons left | 13:45 |
| ← selckin left | 13:45 |
| → noes joined | 13:45 |
|
divVerent
| but well | 13:45 |
|
Grum
| not as git -> git-svn -> svn | 13:45 |
|
divVerent
| that part is already working well... just have ONE problem... | 13:45 |
|
Grum
| git-svn -> svn works; git -> git works | 13:45 |
|
| you dont want automatic dcommits | 13:45 |
|
divVerent
| okay, then I don't want that :P | 13:45 |
|
Grum
| think of what happens if you get a conflict because svn is simply not git | 13:45 |
|
divVerent
| read-only + my own branches suffices for now | 13:45 |
|
Grum
| you leave the intermediate repo in a conflict state .. which you cant solve without manual labour | 13:46 |
|
divVerent
| sure | 13:46 |
|
| but fine, don't need to commit from there | 13:46 |
|
| I want to maitnain a side branch in that git repo | 13:46 |
|
Grum
| and you prolly want to setup a cronjob to 'git svn fetch' | 13:46 |
|
divVerent
| which isn't going to be committed to svn ever | 13:46 |
|
| well | 13:46 |
|
| I read that I am not to publish a non-bare repo | 13:46 |
|
| and git-svn is always non-bare | 13:46 |
| → DavidKlein joined | 13:46 |
| → selckin joined | 13:46 |
| → jspiros joined | 13:46 |
| → up_the_irons joined | 13:46 |
| → jayne joined | 13:46 |
|
Grum
| doesnt matter much | 13:46 |
|
divVerent
| so, I have a bare clone of the git-svn repo and publish that | 13:46 |
|
Grum
| as long as you dont push to it | 13:46 |
|
divVerent
| well | 13:46 |
|
Grum
| (in the checked out branch) | 13:46 |
|
divVerent
| I want to push to the published repo | 13:47 |
|
| in extra branches | 13:47 |
|
| that aren't supposed to go to svn | 13:47 |
|
| and here comes my problem: | 13:47 |
|
| that pushing part worked well | 13:47 |
| ← nullvariable left | 13:47 |
|
divVerent
| on the next git-svn fetch, and git push --mirror barerepo | 13:47 |
|
Grum
| that will work; but it does mean that instead of cronning 'git svn fetch' you also need to cron 'git fetch' from the bare published repo | 13:47 |
|
divVerent
| it killed my branch | 13:47 |
|
| git-fsck shows it as dangling | 13:47 |
|
Grum
| yes; fetch from the bare repo :P | 13:47 |
|
divVerent
| I needed --mirror so the target repo gets the branches and stuff | 13:47 |
|
Grum
| dont push to it | 13:47 |
|
divVerent
| do I get the branches and tags from svn then? | 13:48 |
|
Grum
| i guess you might need to do the 'magic script' to change the refs again | 13:48 |
| ← jayne left | 13:48 |
| ← jspiros left | 13:48 |
| ← up_the_irons left | 13:48 |
| ← selckin left | 13:48 |
| → LiamH joined | 13:48 |
|
divVerent
| okay, will try that now | 13:48 |
| → selckin joined | 13:49 |
| → jspiros joined | 13:49 |
| → up_the_irons joined | 13:49 |
| → jayne joined | 13:49 |
|
divVerent
| the plan is | 13:49 |
|
| I manage a "stable" branch in git that way | 13:49 |
|
| as svn really sucks for that purpose | 13:49 |
|
Grum
| not just for that purpose lol | 13:49 |
|
divVerent
| and then, when people start using that branch read only | 13:49 |
|
| I can pitch the decision to switch to git totally | 13:49 |
|
Grum
| people will not use svn and then git next to it | 13:49 |
|
| they will blame git for being the cumbersome part of the setup | 13:49 |
|
| which is just what you dont want :P | 13:49 |
| → whaley joined | 13:50 |
|
divVerent
| well, it IS cumbersome... to me too, just like it unexpectedly killed the ref to the branch I created and pushed to the bare repo | 13:50 |
|
| sure, can create it again, but not going to, was just a test anyway | 13:50 |
|
| but well, I will try the fetch way now instead of the push way | 13:50 |
|
Grum
| well its not that unexpected to kill that ref | 13:51 |
|
divVerent
| why? | 13:51 |
|
Grum
| you basicly told it to 'make that other thing on the other end *exactly the same as i am' .. and then you were surprised it did it :D | 13:51 |
|
divVerent
| --mirror only says it kills stuff deleted upstream | 13:51 |
|
| now that it killes stuff created downstream | 13:51 |
|
Grum
| mirror means making it equal in my head, add new stuff, purge old/unknown stuff | 13:51 |
|
| oh and update existing stuff :P | 13:52 |
|
| but if you really want to show them the ease of git | 13:52 |
| ← orafu left | 13:52 |
| → orafu joined | 13:53 |
|
divVerent
| let's say the project is too large to suddenly switch to another vcs without already seeing the advantage | 13:53 |
|
Grum
| do you ever merge? | 13:53 |
|
divVerent
| 3GB of data, about 8000 revisions | 13:53 |
|
| not much | 13:53 |
|
| but I know, topic branches WILL be useful | 13:53 |
|
Grum
| that must be complete horror to work on | 13:53 |
|
divVerent
| no, not really | 13:53 |
|
Grum
| and 3gb of data? that doesnt appear to be sourcecode? | 13:53 |
|
divVerent
| yes | 13:53 |
| ← girishr left | 13:54 |
|
divVerent
| most of it is data files | 13:54 |
|
| the revs are mostly source code | 13:54 |
|
Grum
| if you dont have merges, you dont have branches, if you dont have branches, you dont have stable versions unless you halt all development | 13:54 |
|
divVerent
| I would say, over 6000 of the revs are source code changes | 13:54 |
|
| exactly that is our problem :P | 13:54 |
|
| before a release we halt all development and fix only bugs | 13:54 |
|
Grum
| the classical way of doing work, as said: must be complete horror to work on ;) | 13:54 |
| → mithro joined | 13:54 |
|
divVerent
| now, IF we were using git, we would do the "freeze" another way: | 13:55 |
|
| by only merging from topic branches that already had been accepted upstream | 13:55 |
|
| but not from newly created topic branches | 13:55 |
|
Grum
| but be aware, git is made for sourcecode, you will need to look into .gitattributes to handle the big datablobs properly (not create deltas etc) | 13:55 |
|
divVerent
| well, that'll be the smallest problem | 13:55 |
|
Grum
| it wont ;) | 13:55 |
|
divVerent
| just like with svn | 13:55 |
|
| all the crlf mess :P | 13:55 |
| → tstclair_ joined | 13:55 |
|
divVerent
| I hate it when a windows user edits a unix file, and then | 13:56 |
|
Grum
| bleh | 13:56 |
|
cbreak
| no windows, no crlf mess | 13:56 |
|
divVerent
| a) changes every single line of code to add a \r | 13:56 |
|
Grum
| no, use good editors on windows, no crlf mess | 13:56 |
|
divVerent
| b) his changes get \r, rest doesn't | 13:56 |
|
| we regularily see both, until we set svn to autoconvert all the source files | 13:56 |
|
Grum
| a good editor sees the line-endings being used and uses that | 13:56 |
|
divVerent
| good editors exist for windows? :P | 13:56 |
|
Grum
| yes? vim for example | 13:56 |
|
divVerent
| that doesn't count as windows | 13:57 |
| → babilen joined | 13:57 |
|
Grum
| gvim to be accurate | 13:57 |
|
| it does | 13:57 |
|
divVerent
| no windows user uses that | 13:57 |
|
Grum
| it has a window | 13:57 |
|
| you can click | 13:57 |
|
| i use it | 13:57 |
|
divVerent
| unless he primarily uses linux :P | 13:57 |
|
Grum
| busted | 13:57 |
| ← SunilThaha left | 13:57 |
|
divVerent
| there ARE users who primarily use windows, and STILL use gvim? | 13:57 |
|
spb
| yes | 13:57 |
|
divVerent
| do you have putty installed? | 13:57 |
|
spb
| yes | 13:57 |
|
Grum
| divVerent: yes and yes | 13:57 |
|
divVerent
| see, again a unix user :P | 13:57 |
|
Grum
| no, smart user | 13:58 |
|
divVerent
| same thing :P | 13:58 |
|
| seriously... | 13:58 |
|
| no "pure windows" user ever uses vim | 13:58 |
|
| if you don't know unix shell, you're unlikely to know vi commands | 13:58 |
| → rrossel joined | 13:58 |
|
Isenn
| Grum: Sorry for missing some of your posts | 13:58 |
|
Grum
| Isenn: no problemo :) | 13:59 |
|
| divVerent: that is not true | 13:59 |
|
divVerent
| I mean, I wish more people would be using vim | 13:59 |
|
Isenn
| Got another question i have deleted an remote branch with "git push origin :heads/test". But when i issue "git pull origin" from another hosts. The remote branches are still there. Why? | 13:59 |
|
divVerent
| that would e.g. get rid of all crlf problems | 13:59 |
|
Grum
| you just need someone saying: if you use vim you can be 10 times as efficient and then they might leaern it ;)) | 13:59 |
|
divVerent
| because vim DOES that right | 13:59 |
|
| we have people editing project files with visual studio :P | 14:00 |
|
Grum
| divVerent: no it doesnt; by default i think it makes .crlf files | 14:00 |
| ← noss left | 14:00 |
|
divVerent
| and they're not even in a VS supported language | 14:00 |
|
Grum
| divVerent: i'm 100% sure visualcrapio can do this properly | 14:00 |
| → ShadeHawk joined | 14:00 |
|
Grum
| Isenn: are you sure you are pulling from the right spot? what happens if you do git fetch origin ? | 14:00 |
|
| Isenn: i'm not 100% sure what *should* happen btw, im just guessing it would remove the remote entry but it might not .. if a branch you have is tracking it for example ... not sure again ;) | 14:01 |
|
ShadeHawk
| 1.6.5.7 | 1.6.6-rc3 <-- can anyone update #git channel description? | 14:01 |
|
Grum
| divVerent: google says: Try "Advanced save options" | 14:02 |
|
| divVerent: so even VS can be fixed | 14:02 |
| ← tstclair left | 14:02 |
|
Isenn
| Grum: I only have "one" centralised spot, origin. And developers "local" | 14:03 |
|
Grum
| as with most things, take 10 seconds of proper effort and fix it forever | 14:03 |
|
| Isenn: hmmm | 14:03 |
|
ShadeHawk
| warthog9: are you working on version 2 of patches adding caching support for gitweb, or do you not have time now? I'm wondering whether to work on them, or wait for the new version (for re-roll) | 14:03 |
| → p0w3r3d joined | 14:04 |
| → egonw joined | 14:04 |
| ← LRN left | 14:05 |
| → mbernstein joined | 14:05 |
|
_Krieger_
| why i can't clone shallow repo? how can i clone it without checking out full revision log? | 14:05 |
|
egonw
| hi all... in the past when I did 'git am' it applied parts of the patch for me to fix with 'git mergetool'... however, when I now apply a patch with 'git am' and it fails, it does not seem to do anything, let alone anything for me to fix ... how can I get 'git am' to apply as much as possible, and give those typical conflicts? | 14:06 |
| → notjohn joined | 14:06 |
| ← p0w3r3d left | 14:07 |
| → p0w3r3d joined | 14:07 |
| → ruediger joined | 14:07 |
|
egonw
| _Krieger_: you might check 'git clone --depth' ... | 14:07 |
| ← babilen_ left | 14:07 |
| → Asad_ joined | 14:08 |
| → LRN joined | 14:09 |
|
ShadeHawk
| egonw: git am -3 | 14:09 |
|
egonw
| ShadeHawk: agrhh... how could I have missed that option in the help :( thanx! | 14:10 |
| → Bass10 joined | 14:11 |
| → naeu joined | 14:11 |
|
egonw
| ShadeHawk: there is nothing I can do about "Cannot fall back to three-way merge.", I guess? | 14:12 |
| ← fcuk112 left | 14:13 |
|
Grum
| _Krieger_: well i guess theoretically at most you can get a shallow clone out of a shallow clone .. seems rather pointless though | 14:13 |
|
ShadeHawk
| no, it means that you don't have blobs (version of files) that the patch was based on | 14:13 |
|
_Krieger_
| thanks, egonw and Grum | 14:15 |
|
Grum
| _Krieger_: if you want to clone a shallow clone you might as well copy it or to a shallow clone from the original | 14:15 |
| → Sky[x] joined | 14:15 |
|
egonw
| ShadeHawk: OK, that I do not follow (sorry for that)... does it mean I do not have a path or file changed in the patch? but I don't get where the version comes in, as it's 'git am' which should not care about hashes, right? | 14:16 |
|
_Krieger_
| Grum, i just cloned full repo of public origin, making shallow repo was just pointed in software installation manual :) | 14:16 |
| → mbroeker joined | 14:17 |
|
Isenn
| I am in a "git checkout --track -b origin/testbranch" and issue "touch MARCUS ; git add MARCUS ; git commit -a ; git push origin". Why is not my changed pushed to the server? | 14:19 |
|
| changes | 14:19 |
| → AlainDelon joined | 14:19 |
|
Isenn
| The output of git push origin is "Everything up-to-date" | 14:19 |
|
AlainDelon
| LondonABC.co.uk is your exclusive guide to all that London has to offer. | 14:19 |
|
Isenn
| but there are one new file | 14:19 |
|
AlainDelon
| Whether you are looking for London Sightseeing Tour, London Eye, London Underground, Theatre Tickets, Nightlife, Hotels in London, London restaurants or even Shopping locations - LondonABC.co.uk is your one-stop guide to all that makes London a great city. | 14:19 |
| → aziz joined | 14:20 |
|
ShadeHawk
| you don't have the version the file was base on (diff is based on version A, and you don't have version A) | 14:20 |
| ← cilly left | 14:21 |
| ← AlainDelon left | 14:21 |
|
ShadeHawk
| Isenn: by default git pushes tracking branches. Is testbranch present on server? Does "git push origin testbranch" help? | 14:21 |
| ← roop left | 14:22 |
|
Isenn
| ShadeHawk: Both client can see origin/testbranch and both have local tracking of it | 14:23 |
|
| But i can not push or pull updates from it | 14:23 |
| → Paraselene_ joined | 14:23 |
|
Isenn
| ShadeHawk: outout is "error: src refsepc testbranch does not match any." | 14:23 |
| → nothingHappens joined | 14:24 |
| → Tureba joined | 14:24 |
| ← maattd left | 14:25 |
| → RobertLaptop joined | 14:25 |
| → fcuk112 joined | 14:25 |
| → tmountain joined | 14:25 |
|
tmountain
| any idea why this won't work? http://gist.github.com/259511 | 14:26 |
|
| already did this - git remote add -t trav-build trav-build [email@hidden.address] | 14:26 |
| → reto_ joined | 14:26 |
|
reto_
| heya | 14:26 |
|
| whats the easiest way to check if commit A is 'contained/reachble' by commit B | 14:27 |
| → blackxored joined | 14:27 |
|
blackxored
| how to remove remote tags? | 14:27 |
|
reto_
| blackxored: : | 14:27 |
| → jrmuizel joined | 14:27 |
|
reto_
| blackxored: :name-of-tag, but beware... removing a published tag is a pretty serious thing | 14:27 |
| → keystr0k joined | 14:28 |
| ← tstclair_ left | 14:28 |
|
reto_
| http://www.kernel.org/pub/software/scm/git/docs/git-tag.html | 14:28 |
|
| reas the discussion about 'Re-tagging' | 14:28 |
| noes → noss | 14:28 |
|
blackxored
| thanks | 14:28 |
|
reto_
| I guess removing it isn't as serious, as long as you dont retag | 14:29 |
|
ShadeHawk
| Isenn: just in case, check if you are not on detached HEAD (no branch in "git branch" output) | 14:29 |
| → tstclair joined | 14:29 |
|
reto_
| and if nobody has fetched it yet it isn't a problem anyway | 14:29 |
| ← JackWinter left | 14:29 |
| → JackWinter joined | 14:30 |
|
ShadeHawk
| reto_: "git merge-base A B" is A | 14:30 |
| → danferranti joined | 14:30 |
| → amirite joined | 14:31 |
|
Isenn
| ShadeHawk: What? | 14:31 |
|
reto_
| ShadeHawk: thats pluming :)? right? | 14:31 |
|
Isenn
| I am in origin/testbranch | 14:31 |
|
| And have uppdated and commited filed. But i can not push it | 14:32 |
| → ctusar joined | 14:32 |
|
Isenn
| I can push and pull updates from master branch. but not my own created branch | 14:32 |
|
reto_
| ShadeHawk: great thank you! | 14:33 |
|
Grum
| "git checkout --track -b origin/testbranch" <-- that might not do want if you have an old version i think | 14:34 |
|
ShadeHawk
| Isenn: did you try "git push -v <remote> <branch>" or with explicit refspec "git push -v <remote> refs/heads/<branch>:refs/heads/<branch>"? Any error message? | 14:34 |
|
Isenn
| error: src refspec refs/heads/testbranch does not match any | 14:36 |
|
| ShadeHawk: if i use wiehtout refs/heads, same message | 14:36 |
|
| "$ git branch" output: origin/testbranch | 14:37 |
|
Grum
| Isenn: yup that is expected when you call: git branch -b origin/testbranch | 14:37 |
|
ShadeHawk
| Isenn: it looks like you don't have testbranch. Check output of "git branch" (it looks like you created 'origin/testbranch' _local_ branch) | 14:38 |
| → __iron joined | 14:38 |
|
ShadeHawk
| "git branch -m origin/testbranch testbranch" | 14:38 |
|
reto_
| ShadeHawk: okay, assume I know the commit id, how can I determine what branches contain that commit? is there something for that? | 14:38 |
|
| doh, git branch --contains . | 14:39 |
|
Isenn
| When i created the local branch from origin/testbranch with "git checkout --track -b origin/testbranch" it creacted a local branch with same name | 14:39 |
|
egonw
| ShadeHawk: can I tell 'git am' to ignore the index, and to try apply the patch anyway? | 14:40 |
| ← mamund left | 14:40 |
|
Grum
| ShadeHawk: that will not work if he still needs to setup tracking | 14:41 |
|
| ShadeHawk: in "git checkout --track -b origin/testbranch" he just added the -b too much .. or forgot 'testbranch' at the end; its either or :P | 14:41 |
|
Isenn
| Eeh? | 14:42 |
|
| After i create "git checkout --track -b origin/branch" DO I NEED to rename it with "git branch -m origin/branch branch" so there is no nameing collision? | 14:42 |
|
_Krieger_
| i've renamed local git repo, how should i change workcopies pulling path? | 14:42 |
| ← __iron left | 14:43 |
|
ShadeHawk
| Isenn: you should | 14:44 |
| → corni joined | 14:44 |
|
ShadeHawk
| Isenn: the command you should use is "git checkout --track origin/testbranch" IIRC, or "git checkout -b testbranch origin/testbranch" | 14:45 |
|
| _Krieger_: edit .git/config | 14:45 |
|
_Krieger_
| thanks | 14:45 |
|
ShadeHawk
| Isenn: and check .git/config for '[branch "origin/testbranch"]' section, which after rename should be '[branch "testbranch"]' | 14:46 |
| ← Chillance left | 14:47 |
| → krynnota_ joined | 14:47 |
| ← tmountain left | 14:47 |
| → TeckniX_ joined | 14:49 |
| ← solofight left | 14:50 |
| → javatexan joined | 14:51 |
| → amirite_ joined | 14:51 |
| ← javatexan left | 14:52 |
| → inffcs00 joined | 14:52 |
| ← inffcs00 left | 14:52 |
| → _Mica_ joined | 14:53 |
| → Lisimba joined | 14:55 |
| → mengu joined | 14:56 |
| ← Sigma left | 14:57 |
| ← mengu left | 14:57 |
| → Sigma joined | 14:57 |
| → hgimenez joined | 14:57 |
| → mspo joined | 14:59 |
| → hummingkid joined | 14:59 |
|
mspo
| is there a good doc on tagging vs branching in git? | 15:00 |
|
Ilari
| mspo: Tags are mostly used to mark stable points. Branching is used for paralel development. | 15:01 |
|
mspo
| Ilari: can you keep developing in a tag, or is it set in stone? | 15:02 |
|
Ilari
| mspo: Tags shouldn't be changed. | 15:02 |
| ← krynnotaur left | 15:02 |
|
mspo
| cvs style, got it | 15:02 |
|
Ilari
| mspo: Git isn't CVS. | 15:02 |
|
divVerent
| best way to insult git is to compare to cvs :P | 15:02 |
|
| but regarding tags, it's svn which is weird there, and has no tag or branch support at all | 15:02 |
|
| (it's all copies instead) | 15:02 |
| ← amirite left | 15:03 |
|
divVerent
| or, "copies that behave like branches in that they are more like copy-on-write symbolic links" | 15:03 |
| → squentin joined | 15:03 |
|
Ilari
| mspo: In fact, tags are more like read-only branches in Git. | 15:03 |
|
mspo
| a clone will download tags as long as they were pushed with -t or whatever the flag is, right? | 15:03 |
|
ShadeHawk
| you can't develop on tag. you can rewrite it (retag with the same name) but you can't develop on tag. as it should be | 15:03 |
|
divVerent
| it's one of the few things where cvs was right :P | 15:04 |
|
ShadeHawk
| fetch would by default autofollow tags, i.e. get all tags that point at dowloadec commits | 15:04 |
|
Ilari
| mspo: Clone grabs all tags from remote repo that exist there. | 15:04 |
|
mspo
| great, thanks | 15:04 |
|
divVerent
| never used tags in svn, because I never found out how to move a tag to a later revision other by destroying it and recreating from scratch | 15:05 |
|
Ilari
| mspo: -t meaning --tags? That only means "push all tags". Tags can be pushed with 'git push <remote> <tagname>' too. | 15:05 |
| → altrux joined | 15:05 |
|
ShadeHawk
| 'git push <remote> tag <tagname>' | 15:06 |
|
mspo
| Ilari: apparently tags are not committed by default | 15:06 |
|
Ilari
| mspo: Tags can't be committed. | 15:06 |
|
mspo
| Ilari: I find that hard to believe | 15:07 |
| ← radii left | 15:07 |
|
Ilari
| mspo: Whatever can't be added to the index (staging area) can't be committed. | 15:07 |
|
ShadeHawk
| errr... tags are outside DAG of revisions, they point into it. | 15:08 |
|
| they are created with 'git tag' | 15:08 |
|
mspo
| git push origin tag name-of-tag | 15:08 |
|
Ilari
| mspo: That just pushes the tag. And if you don't have anything else named 'name-of-tag', you can leave that 'tag' keyword out. | 15:08 |
|
ShadeHawk
| of course you can configure remote to always push all tags | 15:08 |
|
notjohn
| If anyone can help with a tangled git repo: http://drop.io/h0peoek/asset/gitlatest2-png | 15:09 |
|
| i'd like to know how to fix it a | 15:10 |
|
Ilari
| mspo: Tags can't be committed just like branches can't be committed. They have their own namespace outside working directory / snapshot namespace. | 15:10 |
|
mspo
| Ilari: until you push them to the origin? | 15:10 |
|
Ilari
| mspo: Push is push. Push isn't commit. | 15:10 |
|
notjohn
| or well...really I think i merged or rebased on something i shouldn't have so i'd like to revert to the branch i had before i made chagnes | 15:10 |
| ← Asad_ left | 15:11 |
|
notjohn
| or merges | 15:11 |
|
Ilari
| mspo: And push just copies ref (tag or branch) to another repository. | 15:11 |
| → loincloth joined | 15:11 |
| → Chillance joined | 15:12 |
|
threeve
| Ilari: if you make an annotated tag then an object is created in the repo though, right? | 15:12 |
|
egonw
| if I do 'git rebase', how do I know which branches LOCAL and REMOTE refer to? | 15:12 |
| → malumalu joined | 15:13 |
|
Ilari
| threeve: Yes, But it can be pushed. | 15:13 |
| ← ixti left | 15:13 |
| → pheaver joined | 15:13 |
|
mspo
| anyway, thanks | 15:15 |
| ← mspo left | 15:15 |
| → Turias joined | 15:16 |
|
notjohn
| what does it mean if i see "merge branch 'blah' of blah.url.com/path into 'blah'" | 15:17 |
|
| it's as if i merged a local branch onto it's remote/origin branch | 15:18 |
| ← _Mica_ left | 15:18 |
| → stuffcorpse joined | 15:18 |
| → maattd joined | 15:19 |
| → kukks joined | 15:19 |
| ← camonz left | 15:20 |
| → munichlinux joined | 15:21 |
| ← munichlinux left | 15:21 |
| ← johan-s left | 15:21 |
| → __iron joined | 15:21 |
| ← giallu left | 15:22 |
| ← tstclair left | 15:23 |
|
Ilari
| notjohn: It says merge into blah, so what its merge to is local. | 15:24 |
|
drizzd
| notjohn: not sure where your confusion comes from. the first 'blah' is the remote branch, the url is the remote url, and the second 'blah' is your local branch | 15:25 |
|
notjohn
| did you see my screenshot ? | 15:25 |
| → giallu joined | 15:27 |
| ← hummingkid left | 15:28 |
| ← jrmuizel left | 15:28 |
|
notjohn
| maybe it's the step before that.... | 15:28 |
| → kennyp_ joined | 15:28 |
| → Keiya_ joined | 15:28 |
|
notjohn
| we were trying to keep the git tree kind of trim by rebasing | 15:28 |
|
| and i think i rebased at the wrong time or on the wrong branch | 15:29 |
| ← engla left | 15:29 |
| → jrmuizel joined | 15:30 |
| ← jrmuizel left | 15:30 |
| → noes joined | 15:30 |
| → engla joined | 15:30 |
| ← noss left | 15:30 |
| ← Keiya left | 15:32 |
| → Simoyd joined | 15:33 |
| ← EricInBNE left | 15:34 |
|
Simoyd
| does anyone know an easy way to display *only* the date from the most recent commit that a specified file was changed? | 15:34 |
|
drizzd
| git log -1 --pretty="format:%d" <file>, something like that? | 15:35 |
|
Simoyd
| %d doesn't appear to work, but %ai does what I want | 15:37 |
|
| I wasn't aware of the -1 parameter | 15:37 |
|
| thank you! | 15:37 |
| → patmaddox joined | 15:38 |
| → tedoc2000 joined | 15:38 |
| → khmarbaise joined | 15:38 |
| ← __iron left | 15:39 |
| → Mezner joined | 15:41 |
|
Simoyd
| what's the difference between committer date and author date (in pretty format for the log command) | 15:41 |
|
j416
| Simoyd: author is the one who wrote the contents (code) | 15:42 |
| ← lu_zero left | 15:42 |
|
j416
| committer is the one who committed it into the repo | 15:42 |
|
| often this is the same person | 15:42 |
|
| but say you send a patch by mail | 15:42 |
|
| then someone else will be committing that into a tree | 15:42 |
| ← aronchi left | 15:42 |
|
j416
| s/a tree/a repo/ | 15:42 |
|
| so you will be the author, and the one who commits it will be the committer | 15:43 |
| ← gaveen left | 15:43 |
| ← RobertLaptop left | 15:43 |
|
lack
| j416: Does it also change the "committer" (but no the author) if I rebase or cherry-pick someone else's commit? | 15:43 |
|
Simoyd
| j416: what if, I'm not patching but pushing a commit to a remote repo. will the commit date be different on the remote repository? | 15:44 |
|
j416
| pushing doesn't change anything | 15:44 |
|
Simoyd
| ok | 15:44 |
|
drizzd
| is the committer date guaranteed to be ordered in the same order as the commit ancestry? | 15:44 |
| ← Tk_L left | 15:44 |
|
thiago
| drizzd: no | 15:44 |
| → allenwei joined | 15:44 |
|
thiago
| dates are not ordered at all | 15:44 |
|
| SVN imposes that dates must be monotonic | 15:44 |
|
| Git doesn't | 15:45 |
|
drizzd
| interesting, 'cause I thought transport kind of relies on that | 15:45 |
|
lack
| drizzd: If the committer has a broken clock... | 15:45 |
|
thiago
| what transport? | 15:45 |
|
j416
| lack: not sure about that | 15:45 |
|
drizzd
| the git transport protocol keeps track of the oldest commit it's transferring | 15:45 |
|
lack
| j416: I *think* cherry-pick... Let me test! | 15:45 |
|
thiago
| drizzd: indeed | 15:45 |
|
| but why would it need to use the date for that? | 15:45 |
|
j416
| lack: trying is the best :) | 15:46 |
| ← intripoon left | 15:46 |
|
thiago
| the oldest commit is the last one in the commit ancestry :-) | 15:46 |
|
drizzd
| thiago: that's not what I mean, give me a minute | 15:46 |
|
| thiago: see the oldest_have variable in upload-pack.c | 15:47 |
|
lack
| j416: Yes, cherry-pick put me as committer, left author alone. | 15:47 |
|
cbreak
| git rebase -i can reorder commits | 15:47 |
|
lack
| cbreak: But when it does so, doesn't it re-do the committer name,email,and date? | 15:47 |
|
j416
| lack: ok :) | 15:47 |
|
cbreak
| I doubt that it changes the name or e-mail | 15:47 |
| → what joined | 15:47 |
|
cbreak
| hmm. | 15:47 |
| ← Simoyd left | 15:47 |
|
cbreak
| no idea, I work solo | 15:47 |
|
j416
| rebase does change the date though | 15:47 |
| what → simoyd | 15:47 |
|
j416
| committer date | 15:47 |
|
| iirc. | 15:48 |
|
| so it would make sense if it changed the committer info as well | 15:48 |
|
| lack: please report if you try it :) | 15:48 |
| → konr joined | 15:48 |
|
drizzd
| thiago: essentially, git stops looking for ancestors older than the oldest confirmed common commit | 15:48 |
|
simoyd
| is there an easy way to hash a blob file (or get the hash of one) that is build into git somehow? | 15:48 |
|
| built* | 15:49 |
|
drizzd
| thiago: but it could still have a common ancestor, if dates are not strictly ordered | 15:49 |
|
j416
| hash a blob file? | 15:49 |
|
| hash a file? | 15:49 |
|
| try git hash-object | 15:49 |
|
simoyd
| sorry if my terminology is wrong :S I'll take a look at that commant, it looks like it might be what I want | 15:49 |
|
| command* | 15:49 |
|
| apparantly my Ts and Ds are all backwards today lol | 15:50 |
|
j416
| git hash-object works even outside of a git repo | 15:50 |
| ← BC-bd left | 15:50 |
| → selckin1 joined | 15:50 |
|
j416
| it'll create an object for the for the specified file and return its hash | 15:50 |
| ← selckin left | 15:50 |
|
j416
| if you specify -w it will write the object into the repo object database | 15:50 |
|
simoyd
| cool | 15:51 |
|
| will it let me hash a file from a specific refspec? | 15:51 |
|
j416
| without -w it does nothing but print the sha1 | 15:51 |
|
| what do you mean? | 15:51 |
|
simoyd
| like | 15:51 |
|
| I have a commit that's 2 years old | 15:51 |
|
j416
| if the file is already in git, it's already hashed | 15:51 |
|
simoyd
| and I want the hash of a file in that commit | 15:51 |
|
| how do I get it? | 15:51 |
|
j416
| ah | 15:51 |
|
| not git hash-object then | 15:51 |
|
simoyd
| ok | 15:51 |
|
j416
| I think... git ls-tree | 15:52 |
|
| hm | 15:52 |
|
| yeah | 15:52 |
|
| git ls-tree | 15:52 |
| → neoeinstein joined | 15:52 |
|
thiago
| drizzd: you don't need dates | 15:52 |
|
| drizzd: you only need parent child relationship | 15:52 |
|
lack
| j416: I just did a 'rebase -i' as a test on someone else's commits, and I poached them out-of-order. Final result: The commit name and email are me, and the committer dates are properly ascending. The other person is still the author, and the author date is now out-of-order. | 15:53 |
|
j416
| ok :) | 15:54 |
|
| good to know | 15:54 |
|
| thank you for verifying | 15:54 |
|
| this makes sense. | 15:54 |
|
lack
| Which makes sense! So maybe committer dates *are* always in order? | 15:54 |
|
| Except for people with mis-set system times? | 15:54 |
| ← Sigma left | 15:54 |
|
j416
| they don't have to be | 15:54 |
|
| if you alter history with other tools they won't be | 15:54 |
|
lack
| Yes, that makes sense. filter-branch, for example. | 15:55 |
|
j416
| you can use git filter-branch for instance | 15:55 |
|
| yeah | 15:55 |
|
| and/or grafts | 15:55 |
|
Grum
| or grafting/wrong clocks | 15:55 |
|
simoyd
| ok I have no idea how to use ls-tree.. just keeps saying invalid object name.. I must be specifying something wrong :S | 15:55 |
|
j416
| simoyd: easy way: checkout the commit you are interested in | 15:55 |
|
| go to the tree where the file is | 15:56 |
|
| then do: git ls-tree HEAD | 15:56 |
| → cannonball joined | 15:56 |
|
j416
| and it will show you the sha1 of everything in that directory | 15:56 |
|
drizzd
| thiago: during git transport, we are trying to figure which commits we have in common. we don't know the full ancestry of the other end, but since we are transferring haves in date-order this would tell us which commits we already know about. But if date order is not guaranteed it doesn't really tell us anything. | 15:56 |
|
j416
| s/tree/directory/ | 15:56 |
|
| sorry | 15:56 |
| ← bcardarella left | 15:56 |
|
drizzd
| *if date order is not the same order imposed by parent-child relationship | 15:57 |
|
simoyd
| j416: cool | 15:57 |
|
loincloth
| hey all | 15:57 |
|
j416
| more info: man git-ls-tree | 15:57 |
|
| :) | 15:57 |
|
Gitbot
| j416: the git-ls-tree manpage can be found at http://git.or.cz/man/git-ls-tree | 15:57 |
|
loincloth
| anyone know why this happens | 15:57 |
|
| git co -b 'query-reviewer/upgrade/profiling-fix' | 15:57 |
|
| error: unable to resolve reference refs/heads/query-reviewer/upgrade/profiling-fix: Not a directory | 15:57 |
|
| fatal: Failed to lock ref for update: Not a directory | 15:57 |
|
| i can do branch names with one slash just fine.. two breaks it.. or..? | 15:57 |
|
| hmm no two slashes is fine | 15:58 |
|
| maybe the dashes have something to do with it | 15:58 |
|
| weirddd | 15:58 |
|
selckin1
| what is co? | 15:58 |
|
simoyd
| j416: I was doing "<branch>/<file>" instead of "<branch> <file>" (colon vs space) cause git cat-file uses colon. teehee =D works great now thx =D | 15:58 |
|
j416
| selckin1: checkout I suppose | 15:59 |
|
| simoyd: :) | 15:59 |
|
selckin1
| who knows what else he aliased in it | 15:59 |
|
j416
| who knows... | 15:59 |
| ← sde_antrix left | 16:00 |
| → the_weas1l joined | 16:01 |
| → krawek joined | 16:01 |
| → tstclair joined | 16:02 |
| → catalis joined | 16:03 |
|
| sitaram aliased "git" to "g". 66% savings :) | 16:03 |
| noes → noss | 16:04 |
| → godsflaw joined | 16:05 |
|
PerlJam
| sitaram: and do you also alias your git subcommands to one letter? | 16:06 |
|
lericson
| oo, g c | 16:06 |
|
PerlJam
| "git status" -> "g s" | 16:06 |
| ← DemO_ left | 16:06 |
| ← globalkeith left | 16:07 |
| ← flaguy48 left | 16:07 |
| → jrmuizel joined | 16:07 |
| ← patmaddox left | 16:08 |
| ← parigaudi left | 16:08 |
| → mulicheng joined | 16:08 |
|
sitaram
| PerlJam: some, like s=status, b=branch, d=diff | 16:08 |
|
| PerlJam: yeah, "g s", and sometimes ghostscript starts up if my space bar doesnt "take" :) | 16:09 |
| → reprore joined | 16:09 |
|
PerlJam
| heh | 16:09 |
| → the_weas2l joined | 16:10 |
| ← j416 left | 16:10 |
|
mulicheng
| suppose I 1) do work on branch 2) merge master 3) find out my merge has a minor error, 4) do a commit. How is one to use rebase to squash my fix into the merge commit? | 16:10 |
| ← godsflaw left | 16:11 |
| ← stuffcorpse left | 16:11 |
|
mulicheng
| Seems there is a way but It's not standing out to me. | 16:11 |
|
jacobat
| Why do you need to rebase or squash? | 16:12 |
| ← Bass10 left | 16:12 |
|
mulicheng
| Because I don't want to push extra commits with trivial changes. | 16:12 |
|
keystr0k
| When git commits a file that just has a mode change 10644 => 10075 myfile.tar.gz does git re-store the whole file? | 16:12 |
|
mulicheng
| I want the change to be part of the merge commit. | 16:12 |
| → Bass10 joined | 16:12 |
| → madewokherd joined | 16:13 |
|
jacobat
| mulicheng: Well, if you're aware of the issues with rewriting history the easier way imho is to undo the merge commit, commit to the branch and merge again | 16:13 |
|
mulicheng
| The merge is hard though, that is my concern. | 16:13 |
|
jacobat
| mulicheng: So if you're on master right after having merged and figure out you need to add something do: | 16:13 |
|
mulicheng
| It has a number of conflicts and takes a while to resolve/commit. | 16:14 |
|
jacobat
| git reset --hard HEAD^ (undo the merge); git checkout branch; edit/fix/commit; git checkout master; git merge branch | 16:14 |
|
| mulicheng: You can't have everything :) | 16:14 |
|
mulicheng
| But it seems there must be a way. | 16:14 |
|
jacobat
| Yeah, merge again | 16:15 |
| ← ShadeHawk left | 16:15 |
|
mulicheng
| It's easy to squash a commit with rebase -i for instance, but is this not the case for a merge commit. | 16:15 |
|
jacobat
| True | 16:15 |
|
mulicheng
| rebase seems to need a parent commit that has only one parent, not two to start. | 16:15 |
|
| So in effect, there is not way to squash a commit onto a merge commit then right? | 16:15 |
| amirite_ → amirite | 16:16 |
|
jacobat
| I don't know, there might be | 16:16 |
|
| How about just "git commit --amend" ? | 16:16 |
|
mulicheng
| does commit --amend allow you to add additional files etc? I thought that only let you change the commit message. | 16:16 |
| ← hgimenez left | 16:17 |
| → NfNitLoop joined | 16:17 |
|
jacobat
| No, it allows you to do a new commit with content and all | 16:18 |
|
NfNitLoop
| Does git-svn not use $GIT_DIR/info/exclude ? I'm putting file patterns in there and 'git status' doesn't seem to be updating. | 16:18 |
| → _dermoth joined | 16:18 |
|
mulicheng
| Sounds like that is the way to go then. | 16:18 |
| ← d0ugal left | 16:18 |
|
mulicheng
| Perhaps reset --soft will also work. | 16:18 |
|
NfNitLoop
| (assmuing that on Linux $GIT_DIR should be ~/.git) | 16:18 |
| ← the_weasel left | 16:18 |
|
jacobat
| NfNitLoop: Usually you just put stuff in .gitignore | 16:18 |
|
mulicheng
| Will test a bit, Thx. | 16:18 |
|
NfNitLoop
| jacobat: well, upstream doesn't use git so dropping a .gitignore in their repo might be considered rude. | 16:18 |
|
jacobat
| NfNitLoop: How's about a "git config --global core.exludesfile /path/to/excludes" then? | 16:19 |
|
NfNitLoop
| Hrmm, trying. :) | 16:19 |
| → RobertLaptop joined | 16:20 |
|
NfNitLoop
| still no luck, even with a pattern of * | 16:21 |
|
mulicheng
| jacobat: ci --amend works. | 16:21 |
|
jacobat
| mulicheng: cool | 16:21 |
| ← _Krieger_ left | 16:22 |
|
jacobat
| NfNitLoop: What does ~/.gitconfig show? | 16:22 |
|
mulicheng
| I didn't realize that amend would add/change files in addition to changing the status message. | 16:22 |
|
| learn something new every day! | 16:22 |
| ← Analias left | 16:22 |
| → nullvariable joined | 16:23 |
| → gbacon joined | 16:23 |
| ← the_weas1l left | 16:23 |
| ← christophsturm left | 16:23 |
|
NfNitLoop
| jacobat: http://gist.github.com/259593 | 16:23 |
|
jacobat
| NfNitLoop: Looks about right | 16:24 |
|
NfNitLoop
| Yeah. *shrug* I can live with a bit too verboase of a 'git status', I guess. | 16:25 |
|
jacobat
| NfNitLoop: And if you have a file unknown to git it still shows up with git status? | 16:25 |
|
NfNitLoop
| yep. | 16:25 |
|
simoyd
| what is some good ALM software that I can use with git? | 16:25 |
|
jacobat
| Interesting | 16:25 |
|
NfNitLoop
| is it trying to honor svn:ignore properties at all? | 16:26 |
|
| OH. | 16:26 |
|
| it seems ot have created a .gitignore in my project. *facepalm* | 16:26 |
|
jacobat
| Huh? | 16:27 |
|
lack
| loincloth: If you have a branch called 'query-reviewer/upgrade' you cannot also create one called 'query-reviewer/upgrade/profiling-fix' | 16:27 |
|
jacobat
| NfNitLoop: The global exclude and .gitignore are not mutually exclusive | 16:27 |
|
lack
| simoyd: Automated Lion Management? I prefer to tame my lions by hand. | 16:27 |
|
loincloth
| lack: obviously | 16:28 |
|
| lack: care to elaborate on why? | 16:28 |
|
NfNitLoop
| jacobat: It seems to be acting that way. there are things in /.gitignore that it is ignoring. It looks like one of my coworkers added .gitignore. I was stupidly assuming nobody else here would've done that. (But, yay!) | 16:28 |
| ← noss left | 16:28 |
|
simoyd
| lack: lol | 16:29 |
|
| does "CruiseControl.NET" work with git? | 16:29 |
| → agile joined | 16:29 |
| ← brendon_work left | 16:29 |
|
lack
| Because branches are stored as files. When you have a branch query-reviewer/upgrade this is represented as a file 'upgrade' in a directory called "query-reviewer". If you try to create 'query-reviewer/upgrade/profiling-fix', that would mean there would have to be a directory called 'upgrade' inside a directory called "query-reviewer", but there's already a file of the same name there, which is why you get the "Not a directory" error. "upgrade" is not a d | 16:30 |
|
NfNitLoop
| jacobat: and adding my pattern to the proper .gitignore worked. *shrug* Thanks for the help, sorry I'm braindead. :p | 16:30 |
|
jacobat
| NfNitLoop: Weird, I have both a global and a local worked fine for me | 16:30 |
|
loincloth
| lack: i see.. thx | 16:31 |
|
lack
| loincloth: This is one of the *very* few restrictions on branch naming in git. | 16:31 |
|
loincloth
| that kind of sucks but oh well | 16:31 |
|
| thanks for the explanation | 16:31 |
|
lack
| Well, use something other than a '/' then... query-reviewer/upgrade_profiling-fix | 16:32 |
| ← NfNitLoop left | 16:34 |
| → botanicus joined | 16:35 |
| ← engla left | 16:36 |
| ← krawek left | 16:37 |
| → krawek joined | 16:37 |
| ← froschi left | 16:39 |
| ← psoo left | 16:40 |
| ← allenwei left | 16:42 |
|
raz
| what's the best way to check what a git pull would modify before actually doing it? | 16:45 |
| → the_weasel joined | 16:46 |
|
sitaram
| git fetch; git diff --stat HEAD..FETCH_HEAD maybe? | 16:46 |
| → patmaddox joined | 16:47 |
|
raz
| awesome, thx | 16:47 |
| ← Iff left | 16:48 |
| → niki joined | 16:49 |
| ← Alfonzo left | 16:49 |
| → Alfonzo joined | 16:49 |
| ← LeonBrussels left | 16:50 |
| ← icwiener left | 16:50 |
| → mm_work joined | 16:50 |
| → bigmack83 joined | 16:51 |
| → Iff joined | 16:52 |
| ← Keiya_ left | 16:52 |
| ← patmaddox left | 16:52 |
| ← petrux left | 16:52 |
| → ph^ joined | 16:53 |
| ← amirite left | 16:55 |
| → engla joined | 16:55 |
| → [particle] joined | 16:56 |
|
[particle]
| hola! i'm migrating the head of one branch to a new repo. can i lock down the old branch so it's read-only? | 16:57 |
|
selckin1
| chmod -w .git | 16:57 |
|
| selckin1 hides | 16:57 |
| → amirite joined | 16:59 |
| ← jksz left | 16:59 |
| → jksz joined | 17:00 |
| → raichoo joined | 17:01 |
| ← p0w3r3d left | 17:01 |
| → Haarg joined | 17:01 |
|
raz
| hmmm any idea what could be wrong when git diff simply shows no changes for modified files? (they are recognized by git status as "modified") | 17:02 |
|
sitaram
| [particle]: a branch may move, but a commit (SHA) is readonly always. Or make a tag perhaps? | 17:02 |
|
[particle]
| sitaram: i want to lock down the old branch so no new commits can be made to it | 17:03 |
|
| hrmm... perhaps i need a pre-commit hook? | 17:03 |
| ← the_weas2l left | 17:03 |
| → spearce joined | 17:03 |
|
sitaram
| [particle]: lock down from whom? | 17:03 |
| ← raichoo left | 17:04 |
| → p0w3r3d joined | 17:04 |
|
[particle]
| i'm taking a piece of in-house developed software open source. | 17:04 |
|
sitaram
| [particle]: is this linux? what selckin1 said is not a bad idea, but I'd change it to something like "chattr +i .git/refs/heads/lock-this" | 17:05 |
|
[particle]
| the code is currently in a git repo internally. i'm moving the head of master to an external repo | 17:05 |
|
| it's linux, yes | 17:05 |
| ← khmarbaise left | 17:05 |
|
[particle]
| i want master on the internal repo to be read-only | 17:05 |
|
| there are branches for older versions that we need to keep read-write | 17:06 |
|
sitaram
| [particle]: lock down from whom? Who might "write" to it? | 17:06 |
|
[particle]
| anyone who has access to the internal git repo | 17:06 |
|
sitaram
| [particle]: how do they currently have access? ssh? | 17:06 |
| → hgtdjh joined | 17:06 |
| ← yhager left | 17:07 |
|
[particle]
| does it matter? they still need access to the other branches | 17:07 |
|
lack
| Write a small pre-receive hook that rejects any updates to that branch. | 17:07 |
|
| Or maybe an update hook would be more appropriate. | 17:08 |
| → alexcabrera joined | 17:08 |
|
sitaram
| [particle]: it does matter... could you answer please how they currently have access? :) | 17:08 |
|
[particle]
| git+ssh | 17:09 |
| → rudi_s joined | 17:09 |
| → lhz joined | 17:10 |
|
sitaram
| [particle]: try gitolite? | 17:10 |
|
lack
| What about an update hook like this: #!/bin/bash if [[$1 == "lockedbranch"]]; then echo "This branch is locked"; exit 1; fi; exit 0 | 17:11 |
| ← lhz left | 17:11 |
| → lhz joined | 17:11 |
|
sitaram
| lack++ | 17:11 |
|
| sitaram hides his over-engineered (for this specific case) solution... | 17:12 |
| → xfdjpc joined | 17:12 |
|
Ilari
| Heh. | 17:12 |
|
lack
| ;) | 17:12 |
|
alexcabrera
| hi, i'm having an issue with git branching that I can't quite figure out. I created a branch in my local repo, wanted to push it to my gitosis-server which seemed to work fine. Then on my production server I want to pull it into a new directory and switch to the new branch (a prototype of our new site) | 17:14 |
| → hekaldama joined | 17:14 |
| → defectiv joined | 17:14 |
|
alexcabrera
| but the new branch doesn't show up on the freshly cloned repo | 17:14 |
|
sitaram
| alexcabrera: how did you push from local repo to gitosis server? just "git push" or "git push remotename branchname"? | 17:15 |
|
alexcabrera
| sitaram, git push origin branchname | 17:15 |
|
hekaldama
| the latter is usual, the previous can be configured to have defaults afaik | 17:16 |
|
| so when you say git push it will default to a location and branch | 17:16 |
|
sitaram
| alexcabrera: on the local repo, when you do "git ls-remote origin" do you see the new branch, with the correct SHA? | 17:16 |
|
alexcabrera
| sitaram: I see the new repo with the SHA | 17:17 |
|
| it says refs/heads/branchname | 17:17 |
|
sitaram
| alexcabrera: on the freshly cloned repo, try the same command -- does it also show it? | 17:18 |
|
alexcabrera
| sitaram: yes, exactly the same | 17:18 |
|
| sitaram facepalms | 17:18 |
|
alexcabrera
| but when I do a git branch -l it only shows the master branch | 17:18 |
|
| Ok, I figured I was doing stupid | 17:19 |
|
sitaram
| alexcabrera: did you mean "git branch" on the new clone does not show it or "git branch -r" does not show it? | 17:19 |
|
alexcabrera
| I was doing git branch -l | 17:19 |
|
| because I'm apparently fucking retarded | 17:19 |
|
sitaram
| alexcabrera: why "-l"? anyway try "-r". Then to make it local, "git branch -t -b branch origin/branch" | 17:20 |
| → kithpom joined | 17:20 |
| ← unreal left | 17:20 |
|
| sitaram thinks if you have to be retarded, you may as well have fun ;-) | 17:20 |
| ← Alfonzo left | 17:21 |
|
alexcabrera
| sitaram: that did it, except needed to do git checkout instead of git branch | 17:21 |
|
| thank you so much | 17:21 |
|
sitaram
| alexcabrera: yeah it was my turn to be fscking retarded I guess :) | 17:21 |
|
alexcabrera
| at least I'm trying to not pull a flickr :\ | 17:21 |
|
| thanks a lot | 17:21 |
|
sitaram
| ?? | 17:21 |
|
| whats that? | 17:21 |
| ← catalis left | 17:22 |
|
alexcabrera
| their pride in not using a scm solution and instead hardcoding flags in their code | 17:22 |
|
sitaram
| oh wow... didnt know this! | 17:22 |
| → Alfonzo joined | 17:23 |
|
alexcabrera
| sitaram: http://code.flickr.com/blog/2009/12/02/flipping-out/ | 17:23 |
|
| it'll make you cry | 17:23 |
| ← loincloth left | 17:23 |
|
sitaram
| looking... /me is a masochist I guess | 17:23 |
| ← ruediger left | 17:24 |
|
sitaram
| wow... I'm willing to bet thedailywtf.com will be getting some submissions soon :) | 17:26 |
| → patrikf joined | 17:27 |
| ← hgtdjh left | 17:27 |
| ← mulicheng left | 17:27 |
| → johan-s joined | 17:28 |
| ← nkuttler left | 17:28 |
| → nkuttler_ joined | 17:28 |
| nkuttler_ → nkuttler | 17:28 |
| ← ntoll left | 17:29 |
| → warlock_mza joined | 17:30 |
|
bremner_
| alexcabrera: that is because only old people use scm | 17:31 |
| → dreiss joined | 17:33 |
| → ihoka joined | 17:34 |
| → GyrosGeier joined | 17:34 |
|
GyrosGeier
| hi | 17:34 |
|
| GyrosGeier needs to rewrite the email addresses in a bunch of commits | 17:34 |
|
GyrosGeier
| not pushed yet, so safe | 17:34 |
|
| I remember there being a way to do that | 17:35 |
|
lack
| GyrosGeier: Probably git format-patch | 17:35 |
|
| No, wait, brain did wrong tab-completion. | 17:35 |
| → khmarbaise joined | 17:35 |
| ← xfdjpc left | 17:36 |
|
lack
| GyrosGeier: filter-branch The *other* long one that starts with f and has a dash | 17:36 |
|
GyrosGeier
| ah | 17:36 |
|
| GyrosGeier reads up | 17:36 |
| → pxikeg joined | 17:36 |
| → unreal joined | 17:36 |
| ← reprore left | 17:37 |
| ← srcerer left | 17:37 |
|
lack
| GyrosGeier: As I'm sure you're reading, the --env-filter is probably what you want. | 17:37 |
| ← curtana left | 17:38 |
| ← Chillance left | 17:38 |
| → radii joined | 17:38 |
| → bdiego joined | 17:38 |
|
| GyrosGeier tries | 17:38 |
| ← thiago left | 17:39 |
| ← shruggar left | 17:41 |
| → clsnnr joined | 17:41 |
| ← botanicus left | 17:42 |
| ← Matjas_ left | 17:42 |
| → hipe joined | 17:43 |
| ← pxikeg left | 17:43 |
|
GyrosGeier
| \o/ | 17:44 |
|
| worked | 17:44 |
|
| thanks | 17:44 |
|
| lack++ | 17:44 |
| → patmaddox joined | 17:46 |
| → Tk_L joined | 17:47 |
| ← Alfonzo left | 17:49 |
| ← krynnota_ left | 17:50 |
| ← priidu left | 17:52 |
| ← nothingHappens left | 17:54 |
| ← Fissure left | 17:54 |
| → akahn joined | 17:54 |
| → krynnotaur joined | 17:55 |
|
akahn
| If I git commit --amend'ed to change the commit message, after i had already pushed to a remote, and now i can't push to that remote, how can I just ditch the ammended commit and go back to the one that i had originally? | 17:55 |
| ← patmaddox left | 17:55 |
| ← kennyp_ left | 17:57 |
|
kmap
| akahn: How about resetting to a previous commit and rebasing to the remote one? | 17:57 |
| → Alfonzo joined | 17:58 |
| → kennyp joined | 17:58 |
|
akahn
| this is actually the last commit | 17:58 |
|
| so i guess i could just reset | 17:58 |
|
kmap
| Right. So I'd do git reset --hard HEAD | 17:58 |
|
| Rather, HEAD^ | 17:59 |
|
| And then, rebase to origin/master, or the like | 17:59 |
| ← naeu left | 17:59 |
|
akahn
| cool, thank you | 18:00 |
|
kmap
| No problems. Hope it worked. | 18:00 |
|
defectiv
| okay, i think i'm going crazy. i create a branch. i add a file. i switch branches back to master... and the file is now in master!! i delete it in master, and it's deleted in the feature branch. what. the. heck. | 18:05 |
| → dfr|work joined | 18:05 |
| → mkultras joined | 18:05 |
| → ablemike joined | 18:06 |
| ← selckin1 left | 18:06 |
|
cbreak
| defectiv: did you commit it? | 18:07 |
|
defectiv
| no. | 18:08 |
|
| why? | 18:08 |
|
| do you have to commit all staged stuff before switching bracnhes? | 18:08 |
| ← fcuk112 left | 18:09 |
| ← clsnnr left | 18:10 |
|
PerlJam
| defectiv: you don't *have* to, but if you expect git to keep track of it, you must commit it. | 18:10 |
| → srcerer joined | 18:10 |
|
defectiv
| oh i thought you just had to stage it. | 18:12 |
|
| interesting. | 18:12 |
|
drizzd
| defectiv: you can use -m to make git merge it in-place (dangerous) | 18:12 |
|
defectiv
| i don't like dangerous. | 18:12 |
|
| thanks for clearing that up. | 18:12 |
| → methanol joined | 18:13 |
| → jelmer_ joined | 18:13 |
| ← engla left | 18:15 |
| ← zoniq_ left | 18:15 |
| → zoniq_ joined | 18:16 |
| → iaccwf joined | 18:18 |
| ← Hendu left | 18:19 |
| ← lucsky left | 18:20 |
| → maattd_ joined | 18:21 |
| ← rrossel left | 18:23 |
| ← bigmack83 left | 18:23 |
| → pigdude joined | 18:24 |
| → konr` joined | 18:24 |
| ← srcerer left | 18:24 |
|
pigdude
| How can I troubleshoot an unpacker error? | 18:24 |
|
| I'm pushing to a branch that I've never pushed to before | 18:24 |
| → srcerer joined | 18:25 |
| ← jelmer left | 18:26 |
| ← maattd left | 18:28 |
| maattd_ → maattd | 18:28 |
| → engla joined | 18:28 |
| ← JakeSays left | 18:29 |
| ← vinse left | 18:29 |
| → JakeSays joined | 18:29 |
| → vinse joined | 18:29 |
| ← Keith-VB91 left | 18:29 |
| ← krawek left | 18:29 |
| → pkw_ joined | 18:29 |
| → godsflaw joined | 18:29 |
| ← Alfonzo left | 18:29 |
| ← pkw left | 18:30 |
| → krawek joined | 18:30 |
|
Ilari
| pigdude: Can you pastebin full output? | 18:30 |
|
| pigdude: Can be for example permissions error. Full output would probably tell more... | 18:30 |
| → r11t joined | 18:30 |
| → Keith-VB92 joined | 18:30 |
|
pigdude
| sure, in an hour :) | 18:31 |
|
| gotta go for now | 18:31 |
| ← galderz left | 18:32 |
| ← ablemike left | 18:32 |
| → froschi joined | 18:32 |
| → tomoj joined | 18:32 |
| ← konr left | 18:33 |
| ← iaccwf left | 18:34 |
| ← doener left | 18:36 |
| → doener joined | 18:36 |
| ← srcerer left | 18:36 |
| ← NetHawk left | 18:37 |
| → coolcat joined | 18:37 |
| → froschi1 joined | 18:37 |
| → srcerer joined | 18:37 |
| → NetHawk joined | 18:38 |
| → OtherJakeSays joined | 18:40 |
| → Administrator__ joined | 18:41 |
| → tehbaut joined | 18:41 |
| → sdboyer joined | 18:41 |
| → juan_arandaalvar joined | 18:43 |
| → nothingHappens joined | 18:43 |
| ← srcerer left | 18:43 |
| ← sdboyer left | 18:43 |
| ← egonw left | 18:44 |
| ← Adlai left | 18:45 |
| → srcerer joined | 18:45 |
| → Adlai joined | 18:45 |
| ← giallu left | 18:45 |
| → ixti joined | 18:47 |
| → sieni_ joined | 18:47 |
| ← Administrator__ left | 18:48 |
| → psoo joined | 18:50 |
| → DavidKlein^ joined | 18:50 |
| ← froschi left | 18:52 |
| → AkumaStreak joined | 18:54 |
| → loincloth joined | 18:54 |
| → flaguy48 joined | 18:54 |
| → fcuk112 joined | 18:55 |
| ← altrux left | 18:56 |
| → reprore_ joined | 18:56 |
| ← tehbaut| left | 18:57 |
| ← Adlai left | 18:57 |
| ← JakeSays left | 18:57 |
| → Adlai joined | 18:58 |
| ← DavidKlein left | 18:59 |
| → hummingkid joined | 19:00 |
| ← sieni left | 19:00 |
| ← NET||abuse left | 19:00 |
| ← unreal left | 19:04 |
| → Theravadan joined | 19:06 |
| ← GyrosGeier left | 19:07 |
| ← jayne left | 19:07 |
| ← jspiros left | 19:07 |
| ← up_the_irons left | 19:07 |
| → sockmonk joined | 19:07 |
|
magcius
| how do I abort a conflicted merge? | 19:07 |
|
Ilari
| magcius: 'git reset --hard'. | 19:08 |
| ← babilen left | 19:10 |
| ← hipe left | 19:11 |
| → hipe joined | 19:11 |
| → unreal joined | 19:11 |
| → dan_ joined | 19:13 |
| → botanicus joined | 19:13 |
| → tcurdt joined | 19:13 |
| → priidu joined | 19:15 |
|
tcurdt
| can I have a global ~/.gitattributes just like there is a ~/.gitconfig? | 19:15 |
|
Ilari
| tcurdt: IIRC, no. Global exceludes is possible. | 19:16 |
| ← hummingkid left | 19:17 |
| ← pheaver left | 19:18 |
| ← madewokherd left | 19:19 |
|
tcurdt
| Ilari: what about thing like "*.pbxproj -crlf -diff" ? | 19:19 |
| ← tbuehlmann left | 19:19 |
|
Ilari
| tcurdt: IIRC, templates exist. And are pbxproj's project files of some windows IDE? | 19:21 |
| → GyrosGeier joined | 19:21 |
| → jspiros joined | 19:21 |
| → up_the_irons joined | 19:21 |
| → jayne joined | 19:21 |
| → patmaddox joined | 19:21 |
| → cubuanic joined | 19:21 |
|
tcurdt
| Ilari: just an example from the docs :) ... templates? | 19:22 |
| → madewokherd joined | 19:23 |
|
Ilari
| tcurdt: Templates are used to initially populate repository. Those cause stuff like that obsolete 'branches' directory. | 19:23 |
| ← reprore_ left | 19:23 |
|
tcurdt
| ah | 19:24 |
|
| hm | 19:24 |
|
| still means I have to have it in every project | 19:24 |
|
Ilari
| tcurdt: But that would only affect new projects. | 19:24 |
|
tcurdt
| sure | 19:25 |
| → bauruine joined | 19:25 |
| → babilen joined | 19:26 |
| ← jayne left | 19:26 |
| → jayne joined | 19:26 |
| ← methanol left | 19:28 |
| ← jelmer_ left | 19:29 |
| → FunkeeMonk joined | 19:31 |
| ← mbroeker left | 19:33 |
| ← ph^ left | 19:33 |
| ← DavidKlein^ left | 19:34 |
| ← neoeinstein left | 19:34 |
| ← maattd left | 19:34 |
| → aisipos joined | 19:34 |
| → ph^ joined | 19:37 |
| ← aresnick left | 19:38 |
| ← hipe left | 19:40 |
| → hipe_ joined | 19:40 |
| → ruediger joined | 19:44 |
| → reprore joined | 19:47 |
| xenoterracide_ → xenoterracide | 19:48 |
| ← flatline left | 19:49 |
| → pheaver joined | 19:50 |
| ← LRN left | 19:53 |
| → gpkvt joined | 19:54 |
| → methanol joined | 19:55 |
| ← dan_ left | 19:55 |
| → dan_ joined | 19:55 |
| ← dan_ left | 19:57 |
| ← gpkvt left | 20:02 |
| ← konr` left | 20:04 |
| → sdboyer joined | 20:06 |
| → DavidKlein joined | 20:08 |
| ← methanol left | 20:09 |
| → binjured_ joined | 20:13 |
| ← binjured_ left | 20:13 |
| → bcardarella joined | 20:14 |
| ← binjured left | 20:14 |
| ← patmaddox left | 20:14 |
| → patmaddox joined | 20:16 |
| ← priidu left | 20:18 |
| ← blackxored left | 20:18 |
| ← patmaddox left | 20:18 |
| ← v0n left | 20:19 |
| ← tehbaut left | 20:19 |
| → tehbaut joined | 20:19 |
| ← reprore left | 20:19 |
|
pigdude
| Ilari, oh, I figured it out...silly mistake, SSH'ing as the wrong user. | 20:23 |
| → giallu joined | 20:24 |
| ← gitte left | 20:26 |
| ← Adlai left | 20:26 |
| → Adlai joined | 20:27 |
| → selckin joined | 20:31 |
| ← ixti left | 20:32 |
| → paltman joined | 20:34 |
| ← aisipos left | 20:35 |
| ← dreiss left | 20:39 |
| ← krunk- left | 20:40 |
| ← volodya left | 20:41 |
| ← bantic left | 20:43 |
| → binjured joined | 20:46 |
| ← binjured left | 20:47 |
| → binjured joined | 20:47 |
| ← troyt left | 20:50 |
| ← irc2samus left | 20:51 |
| → ehabkost_ joined | 20:53 |
| → juan_arandaalva1 joined | 20:54 |
| → jelmer joined | 20:55 |
| → Dvyjones joined | 20:55 |
|
Dvyjones
| How would I add a tag in git? I want to add a version commit (for 0.2.0). | 20:56 |
| → troyt joined | 20:56 |
| → beni- joined | 20:57 |
| ← godsflaw left | 20:57 |
| ← paltman left | 20:58 |
| → robinr joined | 20:58 |
| ← cannonball left | 20:59 |
| → cannonball joined | 20:59 |
| → cousine joined | 20:59 |
|
lack
| Dvyjones: Well, for an annotated tag, 'git tag -a <tagname>' For a non-annotated tag 'git tag <tagname>' For more details, see man git-tag | 21:00 |
|
Gitbot
| Dvyjones: the git-tag manpage can be found at http://git.or.cz/man/git-tag | 21:00 |
|
Dvyjones
| lack: Does an annotated tag look like a commit (for instance something like "tagged v0.2.0")? | 21:01 |
| → godsflaw joined | 21:02 |
| → ShadeHawk joined | 21:04 |
| ← beni- left | 21:04 |
| → beni- joined | 21:04 |
| ← zoniq_ left | 21:04 |
| ← lhz left | 21:04 |
|
ShadeHawk
| warthog9: improved/revised versions of patche 1-3 of 6 sent (based in gitweb-ml-v2 branch) | 21:05 |
|
lack
| Dvyjones: You should read the man page. A non-annotated tag is really just like a branch, it's a simple pointer to a commit object. An annotated tag is kind of like a commit plus a tag, it is an object that has a log message, but points at a specific commit. | 21:05 |
|
warthog9
| ShadeHawk: yeah saw those | 21:05 |
|
| ShadeHawk: need to finish the upgrades on these last two korg machines then I can flip back around on gitweb | 21:05 |
| ← ehabkost left | 21:08 |
|
Dvyjones
| lack: Ok, thanks. | 21:08 |
| → juan_arandaalva2 joined | 21:09 |
| → vbgunz joined | 21:10 |
| ← capndiesel left | 21:10 |
|
auslander
| question: i want to convert my svn repo to git, locally first. do i need to worry about tags if all i care about is branches and trunk? can i just specify -b branches -T trunk and leave out -t? | 21:11 |
|
| i want to avoid tags become remote branches (from what i am reading, thats what happens?) | 21:11 |
| → ixti joined | 21:11 |
| ← juan_arandaalvar left | 21:13 |
| ← tbf left | 21:13 |
|
lack
| Dvyjones: Glad to ;) | 21:14 |
|
ShadeHawk
| O.K. I'd rather send patches before you stat new round | 21:17 |
|
| start | 21:19 |
| ← akahn left | 21:19 |
| ← juan_arandaalva1 left | 21:19 |
| ← simoyd left | 21:21 |
| → godsflaw_ joined | 21:21 |
| ← fcuk112 left | 21:21 |
| → Cromulent joined | 21:22 |
| → __iron joined | 21:23 |
| → smotchkkiss joined | 21:24 |
| ← godsflaw_ left | 21:24 |
|
raz
| is there a way to include another file in the gitconfig? | 21:25 |
|
| i want to override it but still reference the repository one | 21:25 |
| ← warlock_mza left | 21:25 |
|
raz
| (override via GIT_CONFIG or GIT_CONFIG_LOCAL0 | 21:25 |
| ← godsflaw left | 21:28 |
| → robbiet480 joined | 21:29 |
|
robbiet480
| hey is there a git IRC bot that isnt cia.vc for post-commits? | 21:29 |
| → godsflaw_ joined | 21:30 |
| ← notjohn left | 21:31 |
| ← krynnotaur left | 21:32 |
| ← ihoka left | 21:32 |
|
ShadeHawk
| raz: no, there are three levels of git config (repo, user/global and system), but no include directive | 21:32 |
| ← keystr0k left | 21:32 |
| → keystr0k joined | 21:32 |
| ← keystr0k left | 21:32 |
|
raz
| ShadeHawk: yea, i worked around it now by using GIT_CONFIG_LOCAL and moving all repository specific stuff to ~/.gitconfig | 21:32 |
| ← kennyp left | 21:33 |
|
hekaldama
| I have only seen the github one in action | 21:35 |
|
robbiet480
| hekaldama: was that for me? | 21:35 |
|
hekaldama
| yeah, sorry | 21:35 |
|
robbiet480
| np | 21:36 |
|
| well that sucks | 21:36 |
|
| i think ill just have to write one! | 21:36 |
|
crankharder
| if I rm foo.txt, and it's unstaged -- how do I bring it back? | 21:36 |
| ← pigdude left | 21:36 |
|
hekaldama
| robbiet480: rbot would probably be quite easy to get going | 21:37 |
|
robbiet480
| hekaldama: thanks for the tip | 21:37 |
|
| ill look into it | 21:37 |
|
cxreg
| crankharder: debugfs? | 21:37 |
|
hekaldama
| np | 21:37 |
| ← bremner_ left | 21:37 |
| → bremner joined | 21:38 |
|
crankharder
| cxreg: sorry, the file is tracked, but the rm is unstaged | 21:39 |
|
| so i just need to bring it back from the repo | 21:39 |
| → mjf joined | 21:40 |
| → scotty42 joined | 21:40 |
|
scotty42
| How do I drop some commits off the tip of a remote svn branch in my local repo? I think I have a recent svn commit that's corrupt in my local repo, and I want to refetch it from the server. | 21:42 |
|
lack
| crankharder: Did you see what 'git status' tells you to do? | 21:42 |
|
| crankharder: git checkout -- <file> | 21:42 |
|
crankharder
| eh, i was trying something with reset...ty | 21:43 |
|
robinr
| spearce: ping | 21:44 |
| ← [particle] left | 21:45 |
| → krynnotaur joined | 21:45 |
|
spearce
| pong | 21:46 |
| → LRN joined | 21:47 |
| ← tcurdt left | 21:48 |
| ← ShadeHawk left | 21:49 |
| ← defectiv left | 21:49 |
| → fcuk112 joined | 21:50 |
| → defectiv joined | 21:50 |
| ← fcuk112 left | 21:50 |
| → fcuk112 joined | 21:50 |
|
raz
| hmm i'm confused after futzing with origins. can someone look at my config and tell my why "git remote" complains "warning: Remote origin has more than one URL"? | 21:50 |
|
| http://dpaste.com/135172/ | 21:50 |
|
Ilari
| raz: Via global gitconfig (there shouldn't be any remote there)? | 21:52 |
| ← beni- left | 21:52 |
|
raz
| Ilari: .git/config is empty, i put it all in .gitconfig (so the stuff i pasted should be the complete config) | 21:52 |
| → beni- joined | 21:52 |
|
Ilari
| raz: There should not be any remote settings in .gitconfig | 21:53 |
| → dreiss joined | 21:53 |
| → timcharper joined | 21:53 |
|
raz
| why not? | 21:54 |
| ← Cromulent left | 21:54 |
|
robbiet480
| hekaldama: found a supereasy one: http://irccat.rubyforge.org/ | 21:54 |
| → gdrage joined | 21:54 |
|
Ilari
| raz: Because they are global to multiple repositories and global settings do not make sense. | 21:56 |
| ← mm_work left | 21:56 |
|
raz
| well, in my case i need that as a workaround because git doesn't give me a way to read a custom config *and* the repository config | 21:56 |
|
| it shouldn't be a problem here because i'm dealing only with one git repo under that user | 21:56 |
| ← pheaver left | 21:57 |
|
raz
| ..except i don't understand why it complains now, does it make up some implicit stuff when the repo config is missing? | 21:57 |
|
scotty42
| how do I change which revision is the tip of a remote branch? I can do this with reset for local branches, but that doesn't seem to work for remote branches (understandably) | 21:57 |
|
Ilari
| scotty42: push | 21:58 |
| → pheaver joined | 21:58 |
|
scotty42
| Ilari: can I use push to drop revisions off the tip of the branch? | 21:58 |
|
Ilari
| scotty42: Yes, that needs forcing and is nasty to others. | 21:58 |
| ← corni left | 21:58 |
| → juan_arandaalvar joined | 21:59 |
|
scotty42
| yeah, this is a remote branch created with git-svn, so others don't see it | 21:59 |
|
| hrm, I don't see how I can do what I want to do with git-push. | 22:00 |
|
| what I really want to do is literally change what rev the tip of the remote branch is pointing at, without doing ANYTHING else | 22:01 |
|
| is there a lower level command that does just that? | 22:01 |
| ← khmarbaise left | 22:01 |
| ← bdiego left | 22:02 |
| ← kithpom left | 22:02 |
| ← Mezner left | 22:02 |
|
cbreak
| you can reset | 22:02 |
|
| on the remote | 22:02 |
|
| or you can use vim and change it yourself | 22:02 |
|
patrikf
| scotty42: git push -f [remote] +<commit-ish>:refs/heads/<branch> | 22:02 |
|
| er, no + i think | 22:03 |
| → altrux joined | 22:03 |
|
scotty42
| patrikf: there's no remote though... | 22:03 |
|
patrikf
| scotty42: what do you mean there's no remote? | 22:03 |
|
scotty42
| patrikf: It's a remote branch created by git-svn | 22:03 |
|
patrikf
| scotty42: you mean a remote-tracking branch= | 22:04 |
|
scotty42
| err, yes:) | 22:04 |
|
| cbreak: what file would I edit directly? | 22:04 |
|
patrikf
| scotty42: use git update-ref | 22:04 |
|
cbreak
| the branch head file I guess | 22:04 |
|
patrikf
| scotty42: assuming you know what you're doing, of course! | 22:04 |
|
cbreak
| I don't think this is a good idea though | 22:04 |
|
| should be in .git/refs/heads/branchname or something like that | 22:05 |
|
patrikf
| cbreak: its not as simple as that, it might be a packed ref | 22:05 |
|
| update-ref will handle that | 22:05 |
|
scotty42
| patrikf: Yeah, I definitely don't know what I'm doing:) | 22:05 |
|
patrikf
| scotty42: then at least make backups ;D | 22:05 |
|
cbreak
| ... I'd make a backup first anyway :) | 22:06 |
|
scotty42
| yeah, I've got one | 22:06 |
|
| I've got a corrupt svn commit near the tip of the remote-tracking branch | 22:06 |
| → ruediger_ joined | 22:06 |
|
scotty42
| my hope is to reset the tracking branch to before that commit, and refetch it | 22:06 |
|
| lemme look at update-ref | 22:06 |
| ← tstclair left | 22:06 |
|
scotty42
| yeah, I think this is what I want | 22:07 |
|
patrikf
| scotty42: actually, i think git svn reset would be more appropriate | 22:07 |
|
| (which you can pass a revision to reset to) | 22:08 |
|
scotty42
| patrikf: wow, you're absolutely right | 22:08 |
|
| completely missed that | 22:08 |
|
| thanks! | 22:08 |
|
patrikf
| :) | 22:08 |
| ← p0w3r3d left | 22:08 |
|
scotty42
| giving that a shot now | 22:08 |
| ← AkumaStreak left | 22:10 |
| → AkumaStreak joined | 22:11 |
| ← the_weasel left | 22:11 |
| → jackdempsey joined | 22:12 |
| thiago_home → thiago | 22:12 |
| ← ashH left | 22:13 |
| → cesario_ joined | 22:13 |
| ← smotchkkiss left | 22:13 |
| ← juan_arandaalva2 left | 22:14 |
|
scotty42
| that did what I expected, but didn't solve my problem:( | 22:14 |
| → Cromulent joined | 22:14 |
|
scotty42
| I have two files that are "stuck" - they show as modified, despite not being modified. Any attempt to reset them fails | 22:14 |
|
| ie, after a "git reset --hard HEAD", they still show as modified in the working copy | 22:14 |
|
selckin
| rm -f file; git checkout -- file | 22:15 |
| → AkumaStreak_ joined | 22:15 |
|
scotty42
| same situation | 22:15 |
|
thiago
| scotty42: Windows? | 22:15 |
|
scotty42
| os x | 22:15 |
|
thiago
| I don't know then | 22:16 |
|
scotty42
| yeah, I've never seen anything like this | 22:16 |
| → tstclair joined | 22:16 |
| ← LiamH left | 22:16 |
| → t0rc joined | 22:16 |
| ← AkumaStreak left | 22:17 |
| → AkumaStreak joined | 22:17 |
| ← patrikf left | 22:17 |
| ← beni- left | 22:18 |
|
lack
| scotty42: If they're showing as 'modified', git diff should say way, I would think... | 22:19 |
| ← ixti left | 22:19 |
|
lack
| Or, why, rather. | 22:19 |
|
scotty42
| it does | 22:19 |
| → AkumaStreak__ joined | 22:19 |
|
scotty42
| I haven't tried hand modifying them to make them match what git diff says | 22:19 |
|
lack
| So, what exactly does git diff say? | 22:19 |
|
scotty42
| however, any attempts to reset/recreate them completely fails | 22:20 |
|
| it shows major changes to the file | 22:20 |
|
| from where, I'm not quite sure | 22:20 |
|
lack
| Whitespace? | 22:20 |
|
| (auto line-ending?) | 22:20 |
|
scotty42
| no, code changes | 22:20 |
|
lack
| Cool. | 22:20 |
|
| That's entirely bizarre. | 22:20 |
|
scotty42
| yeah, it really is | 22:20 |
| ← AkumaStreak__ left | 22:21 |
| ← jackdempsey left | 22:21 |
| ← AkumaStreak_ left | 22:21 |
| ← ruediger left | 22:22 |
|
scotty42
| ok, I think I may have tracked it down. I have two commits which both claim to be adding the same two files | 22:22 |
|
lack
| Curious :) Some sort of graft-gone-wrong? | 22:23 |
| → voxpelli-laptop joined | 22:24 |
|
scotty42
| unclear at this point. Looks like it's an issue in the svn repo, so it's not a git issue. Thanks for the help, all! | 22:25 |
| ← kristian42 left | 22:25 |
| ← AkumaStreak left | 22:26 |
| → babyboop joined | 22:26 |
| ← babyboop left | 22:27 |
| ← jstemmer left | 22:27 |
| ← froschi1 left | 22:28 |
| → unreal_ joined | 22:28 |
| → JasonWoof joined | 22:28 |
| → kristian42 joined | 22:28 |
| → EricInBNE joined | 22:28 |
|
hipe_
| how can i see a diff of only the things i have added to be comitted ? | 22:30 |
|
yath
| git diff --cached, i think | 22:31 |
| ← hugoxrosa left | 22:32 |
|
hipe_
| yath: yup that's it thanks ;) | 22:32 |
|
yath
| np | 22:32 |
| ← auslander left | 22:32 |
| ← timcharper left | 22:33 |
| → JEEBcz joined | 22:33 |
| → axiom joined | 22:33 |
| → edenc_ joined | 22:33 |
| → Tronic_ joined | 22:33 |
| → Risar` joined | 22:33 |
| → Rh0nda joined | 22:34 |
| ← yann left | 22:34 |
| → teknotus_ joined | 22:34 |
| → ramus`_ joined | 22:34 |
| → shtylman__ joined | 22:34 |
| → KiBi_ joined | 22:34 |
| → ArjenL_ joined | 22:34 |
| → Will___ joined | 22:34 |
| → logankoe1ter joined | 22:34 |
| → eagleeye joined | 22:34 |
| → crankhar1er joined | 22:34 |
| ← vmiklos left | 22:34 |
| → Mpenz_ joined | 22:34 |
| → damog_ joined | 22:34 |
| ← SRabbelier left | 22:34 |
| → beppu_ joined | 22:34 |
| → balboah joined | 22:34 |
| → ppr joined | 22:34 |
| → StarkD_ joined | 22:34 |
| → imarcust1is joined | 22:34 |
| → jettero_ joined | 22:34 |
| → SRabbelier1 joined | 22:34 |
| → seanius_ joined | 22:34 |
| → nothingmuch_ joined | 22:34 |
| → gdsx_ joined | 22:34 |
| → alip` joined | 22:34 |
| → wifs_ joined | 22:34 |
| → Thumper__ joined | 22:34 |
| → geeKISSexy joined | 22:34 |
| → vmiklos joined | 22:34 |
| → awx_ joined | 22:34 |
| ← scrooloose left | 22:34 |
| ← r0bby left | 22:34 |
| ← logankoester left | 22:34 |
| ← crankharder left | 22:34 |
| ← elliottcable left | 22:34 |
| ← Tronic left | 22:34 |
| ← comps left | 22:34 |
| ← seanius left | 22:34 |
| ← kmap left | 22:34 |
| ← edenc left | 22:34 |
| ← jrockway left | 22:34 |
| ← shtylman_ left | 22:34 |
| ← alip left | 22:34 |
| ← ToxicFrog left | 22:34 |
| ← ianmcorvidae left | 22:34 |
| ← awx left | 22:34 |
| ← Woody2143 left | 22:34 |
| ← Rhonda left | 22:34 |
| ← MadCoder left | 22:34 |
| ← divVerent left | 22:34 |
| ← Will left | 22:34 |
| ← djpig left | 22:34 |
| ← imarcusthis left | 22:34 |
| ← ThaDon left | 22:34 |
| ← ramus` left | 22:34 |
| ← chrizz left | 22:34 |
| ← Acro left | 22:34 |
| ← rtyler left | 22:34 |
| ← pmcnabb left | 22:34 |
| ← beppu left | 22:34 |
| ← cbreak left | 22:34 |
| ← Risar left | 22:34 |
| ← fujin left | 22:34 |
| ← wifs left | 22:34 |
| ← juice_d left | 22:34 |
| ← cota left | 22:34 |
| ← halfline left | 22:34 |
| ← jettero left | 22:34 |
| ← ArjenL left | 22:34 |
| ← Mpenz left | 22:34 |
| ← murr4y left | 22:34 |
| ← bs left | 22:34 |
| ← majoh left | 22:34 |
| ← Epiglottis left | 22:34 |
| ← balboah_ left | 22:34 |
| ← teknotus left | 22:34 |
| ← schlortoo left | 22:34 |
| ← mariusSO left | 22:34 |
| ← pigeon left | 22:34 |
| ← jpieper left | 22:34 |
| ← reuss left | 22:34 |
| ← lack left | 22:34 |
| ← rotty left | 22:34 |
| ← BeerSerc_ left | 22:34 |
| ← electrogeek left | 22:34 |
| ← deavid left | 22:34 |
| ← d1b left | 22:34 |
| ← zomg left | 22:34 |
| ← pcgod left | 22:34 |
| ← RickXy left | 22:34 |
| ← redondos left | 22:34 |
| ← berndj left | 22:34 |
| ← trodrigues left | 22:34 |
| ← Hunner left | 22:34 |
| ← damog left | 22:34 |
| ← madmoose left | 22:34 |
| ← ice799 left | 22:34 |
| ← KiBi left | 22:34 |
| ← Soliton left | 22:34 |
| ← JEEB left | 22:34 |
| ← StarkD left | 22:34 |
| ← torben_ left | 22:34 |
| ← piroko left | 22:34 |
| ← Thumper_ left | 22:34 |
| ← nothingmuch left | 22:34 |
| ← cworth left | 22:34 |
| ← docgnome left | 22:34 |
| ← wildfire left | 22:34 |
| ← Voker57 left | 22:34 |
| ← IRSeekBot left | 22:34 |
| ← dominikh left | 22:34 |
| ← peper left | 22:34 |
| ← gdsx left | 22:34 |
| ← gernot left | 22:34 |
| → BeerSerc joined | 22:34 |
| → rtyler_ joined | 22:34 |
| → kmap_ joined | 22:34 |
| → r0bby_ joined | 22:34 |
| → reuss_ joined | 22:34 |
| → murr5y joined | 22:34 |
| → yann joined | 22:34 |
| → scrooloose_ joined | 22:34 |
| → comps_ joined | 22:34 |
| → bernd__ joined | 22:34 |
| ← Keith-VB92 left | 22:34 |
| → juice__d joined | 22:34 |
| axiom → majoh | 22:34 |
| nothingmuch_ → nothingmuch | 22:34 |
| → jpieper joined | 22:34 |
| → ThaDon_ joined | 22:34 |
| Tronic_ → Tronic | 22:34 |
| eagleeye → zomg | 22:34 |
| → chrizz joined | 22:34 |
| ← loincloth left | 22:34 |
| → madmoose joined | 22:34 |
| → rotty joined | 22:34 |
| → trodrigues joined | 22:34 |
| → docgnome joined | 22:34 |
| → pmcnabb joined | 22:34 |
| → schlortoo joined | 22:34 |
| → jrockway joined | 22:34 |
| → Soliton joined | 22:34 |
| → wildfire joined | 22:34 |
| → cworth joined | 22:34 |
| → MadCoder joined | 22:34 |
| → piroko joined | 22:34 |
| → lack joined | 22:34 |
| → ianmcorvidae joined | 22:34 |
| → ToxicFrog joined | 22:34 |
| → halfline joined | 22:34 |
| → torben joined | 22:34 |
| → cbreak joined | 22:34 |
| → Acro joined | 22:34 |
| → ice799 joined | 22:35 |
| → Woody2143 joined | 22:35 |
| → djpig joined | 22:35 |
| → Hunner joined | 22:35 |
| → deavid joined | 22:35 |
| → dominikh joined | 22:35 |
| → pigeon joined | 22:35 |
| → redondos joined | 22:35 |
| → maattd joined | 22:35 |
| → fujin joined | 22:35 |
| ← robbiet480 left | 22:35 |
| → mstormo joined | 22:35 |
| ← ph^ left | 22:35 |
| ramus`_ → ramus` | 22:35 |
| → divVerent joined | 22:35 |
| → Voker57 joined | 22:35 |
| → gernot joined | 22:35 |
| → pcgod joined | 22:36 |
| → electrogeek joined | 22:37 |
| → Epiglottis joined | 22:37 |
| ← nkuttler left | 22:37 |
| → IRSeekBot joined | 22:37 |
| ← alip` left | 22:37 |
| ← pheaver left | 22:38 |
| → alip joined | 22:38 |
| → bs joined | 22:39 |
| ← unreal left | 22:39 |
| unreal_ → unreal | 22:39 |
|
magcius
| is there a way to git diff for more than one file? | 22:39 |
| ← krawek left | 22:40 |
| ← kristian42 left | 22:41 |
| → Dashkal joined | 22:42 |
| → schacon joined | 22:42 |
|
_rane
| just ´git diff´ for all files | 22:42 |
| → pantsman joined | 22:42 |
| ← bentob0x left | 22:43 |
|
scotty42
| mystery solved; Two files of the same name in two different commits, but with different casing. OSX isn't case sensitive for filenames, and so git blows up | 22:43 |
| → kristian42 joined | 22:44 |
| ← malumalu left | 22:44 |
| ← tstclair left | 22:44 |
| → milli joined | 22:45 |
| → AkumaStreak joined | 22:45 |
| ← kristian42 left | 22:47 |
| wifs_ → wifs | 22:48 |
| murr5y → murr4y | 22:49 |
| → johnw joined | 22:50 |
| → kristian42 joined | 22:50 |
| ← gdrage left | 22:52 |
| ← Lisimba left | 22:53 |
| → tfarina joined | 22:54 |
| ← sockmonk left | 22:54 |
| ← AkumaStreak left | 22:55 |
| → krunk- joined | 22:56 |
| ← scotty42 left | 22:56 |
| → psil joined | 22:57 |
| → Keith-VB93 joined | 22:58 |
| ← voxpelli-laptop left | 23:01 |
| ← johnw left | 23:01 |
| → cota joined | 23:02 |
| → dgarr joined | 23:02 |
| ← GyrosGeier left | 23:06 |
| ← jspiros left | 23:06 |
| ← up_the_irons left | 23:06 |
| ← bs left | 23:06 |
| ← Woody2143 left | 23:06 |
| ← torben left | 23:06 |
| ← Will___ left | 23:06 |
| ← yann left | 23:06 |
| ← unreal left | 23:06 |
| ← selckin left | 23:06 |
| ← Keith-VB93 left | 23:06 |
| → bs joined | 23:07 |
| → Woody2143 joined | 23:07 |
| → torben joined | 23:07 |
| → yann joined | 23:07 |
| → Will___ joined | 23:07 |
| → unreal joined | 23:07 |
| → selckin joined | 23:07 |
| → GyrosGeier joined | 23:07 |
| → jspiros joined | 23:07 |
| → up_the_irons joined | 23:07 |
| → bigmack83 joined | 23:07 |
| ← JoeConyers left | 23:09 |
| → WALoeIII joined | 23:09 |
| → Jak_o_Shadows joined | 23:10 |
| ← GyrosGeier left | 23:11 |
| ← jspiros left | 23:11 |
| ← up_the_irons left | 23:11 |
| ← Woody2143 left | 23:11 |
| ← unreal left | 23:11 |
| ← yann left | 23:11 |
| ← Will___ left | 23:11 |
| ← torben left | 23:11 |
| ← selckin left | 23:11 |
| ← bs left | 23:11 |
| → bs joined | 23:11 |
| → Woody2143 joined | 23:11 |
| → torben joined | 23:11 |
| → yann joined | 23:11 |
| → Will___ joined | 23:11 |
| → unreal joined | 23:11 |
| → selckin joined | 23:11 |
| → GyrosGeier joined | 23:11 |
| → jspiros joined | 23:11 |
| → up_the_irons joined | 23:11 |
| → JoeConyers joined | 23:12 |
| ← GyrosGeier left | 23:13 |
| ← jspiros left | 23:13 |
| ← up_the_irons left | 23:13 |
| ← Woody2143 left | 23:13 |
| ← unreal left | 23:13 |
| ← yann left | 23:13 |
| ← Will___ left | 23:13 |
| ← torben left | 23:13 |
| ← selckin left | 23:13 |
| ← bs left | 23:13 |
| → bs joined | 23:13 |
| → Woody2143 joined | 23:13 |
| → torben joined | 23:13 |
| → yann joined | 23:13 |
| → Will___ joined | 23:13 |
| → unreal joined | 23:13 |
| → selckin joined | 23:13 |
| → GyrosGeier joined | 23:13 |
| → jspiros joined | 23:13 |
| → up_the_irons joined | 23:13 |
| ← GyrosGeier left | 23:15 |
| ← jspiros left | 23:15 |
| ← up_the_irons left | 23:15 |
| ← Woody2143 left | 23:15 |
| ← unreal left | 23:15 |
| ← yann left | 23:15 |
| ← Will___ left | 23:15 |
| ← torben left | 23:15 |
| ← selckin left | 23:15 |
| ← bs left | 23:15 |
| → D[a]vey joined | 23:15 |
| ← tedoc2000 left | 23:15 |
| → bs joined | 23:15 |
| → Woody2143 joined | 23:15 |
| → torben joined | 23:15 |
| → yann joined | 23:15 |
| → Will___ joined | 23:15 |
| → unreal joined | 23:15 |
| → selckin joined | 23:15 |
| → GyrosGeier joined | 23:15 |
| → jspiros joined | 23:15 |
| → up_the_irons joined | 23:15 |
| ← thiago left | 23:15 |
| kmap_ → kmap | 23:17 |
| → pastie joined | 23:18 |
| → Lisimba joined | 23:20 |
| rtyler_ → rtyler | 23:24 |
| → loincloth joined | 23:25 |
| ← agile left | 23:28 |
| → scarabx joined | 23:31 |
| → agliodbs joined | 23:33 |
| ← Dashkal left | 23:33 |
|
agliodbs
| is there any way to get the "pull" history of a local repository? | 23:33 |
| → AkumaStreak joined | 23:34 |
|
agliodbs
| that is, when pulls were made from a remote repository (github)? | 23:35 |
| Will___ → Will | 23:36 |
| → tjc joined | 23:36 |
| → godsflaw joined | 23:36 |
|
DrNick
| agliodbs: git log -g | 23:37 |
| ← bcardarella left | 23:38 |
| ← botanicus left | 23:39 |
|
agliodbs
| thanks! | 23:40 |
|
| that did it | 23:40 |
| ← AkumaStreak left | 23:41 |
| → gmarquezWRS joined | 23:42 |
| ← godsflaw_ left | 23:43 |
| ericboeh1 → ericboehs | 23:43 |
|
gmarquezWRS
| d | 23:45 |
| ← gmarquezWRS left | 23:45 |
| ← Cromulent left | 23:45 |
| → giMarq joined | 23:45 |
| → Dashkal joined | 23:45 |
| ← Ryback_ left | 23:45 |
| ← vbgunz left | 23:46 |
|
giMarq
| does anyone know how to get these Perl Modules: File::Find & File::Basename on RHEL5? | 23:46 |
|
| I installed some others using yum install but can't find these ones for gitweb... | 23:47 |
|
agliodbs
| perl -MCPAN -e shell | 23:47 |
|
| use cpan | 23:47 |
| ← tg left | 23:47 |
| → tg joined | 23:47 |
|
giMarq
| that will not overwrite any other existing yum installed Perl mods correct? | 23:47 |
|
agliodbs
| well, you'll need to check what dependencies it wants to install | 23:47 |
|
| because it may want to | 23:48 |
| ← jrmuizel left | 23:48 |
|
giMarq
| okay I've used cpan before but was sort of keeping everything same with yum, but I'll try that since I can't locate with yum... | 23:48 |
|
| thnx | 23:48 |
|
agliodbs
| RHEL5 Yum is pretty old. You'll ahve difficulty getting any git tools from it | 23:49 |
| ← dermoth left | 23:49 |
| ← tvw left | 23:49 |
| _dermoth → dermoth_ | 23:49 |
| → yhager joined | 23:50 |
| → Geralt joined | 23:50 |
|
giMarq
| yea yum is limited.... | 23:50 |
|
| i'm trying to get gitweb-caching installed, then going to try to get it to work with gitolite for web UI ACLs as well... | 23:51 |
| dermoth_ → dermoth | 23:51 |
| ← defectiv left | 23:53 |
| → hummingkid joined | 23:53 |
| ← Turias left | 23:54 |
| → anno^da joined | 23:54 |
| → defectiv joined | 23:55 |
| → ploch joined | 23:55 |
| ← niki left | 23:57 |
|
agliodbs
| giMarq: the basic issue is that RHEL5 is over 2 years old | 23:57 |
|
| giMarq: RHEL6 will fix this, for a while | 23:57 |
| → gmarquez_ joined | 23:58 |
| ← JackWinter left | 23:58 |
| ← psoo left | 23:59 |
| → jessep joined | 23:59 |
| → JackWinter joined | 23:59 |