Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-11-23

| Channels | #git index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:01 Raed joined #git
00:03 Vampire0 joined #git
00:09 Gsham joined #git
00:10 SpeakerToMeat Question, if anybody here uses meld as a mergetool. I'm trying out meld as a mergetool, in mode Diff3, I see somewhere, the middle file (the BASE) is the end thing.... so, When i check stuff that differs with my local, and I want to keep the base version, do I need to hit the change so that the base "change" gets applied to local? what happens whith insertions that are only on the LOCAL side?
00:13 tcurdt joined #git
00:13 xocket joined #git
00:15 Charliechin joined #git
00:16 Rodya_ joined #git
00:17 creppe joined #git
00:17 b1tchcakes joined #git
00:21 SpeakerToMeat davidfetter_ge: Hello
00:22 rbr joined #git
00:23 gfixler joined #git
00:26 charles` joined #git
00:27 GavinMagnus joined #git
00:28 rbr joined #git
00:28 rchavik joined #git
00:29 neca joined #git
00:30 charles` I have a repo that I want to share with another user on my machine; but I didn't create it with --shared=group
00:30 charles` how can I convert it?
00:31 charles` is it just setting the sticky bit?
00:35 mellernoia joined #git
00:36 perrier-jouet joined #git
00:37 perrier-jouet hi all
00:38 perrier-jouet is it possible to add more than one worktree ???
00:39 perrier-jouet for managing my dotfiles I have $HOME as worktree and wanted to add another worktree
00:41 mischat joined #git
00:42 los1 joined #git
00:42 mischat joined #git
00:43 mischat joined #git
00:43 perrier-jouet because if I add a file from another directory I get /directory not a repo
00:44 mischat joined #git
00:44 perrier-jouet nm I found worktree add
00:44 mischat joined #git
00:45 mischat joined #git
00:46 mischat joined #git
00:46 Orion3k joined #git
00:47 mischat joined #git
00:48 mischat joined #git
00:48 clemf joined #git
00:49 tristanp joined #git
00:49 fragMental charles`: I would recommend cloning that repo with --bare to some folder where you both can access it. That way you can allow the other user to push to that repo, without any risk of him destroying work in your own copy.
00:50 johnny56 joined #git
00:54 terminal_echo joined #git
00:55 madewokherd joined #git
00:55 woodruffw joined #git
00:55 woodruffw joined #git
00:57 railssmith joined #git
01:05 JamesBaxter joined #git
01:06 davidfetter_ge SpeakerToMeat, sorry, I was in a meeting.
01:08 ruskie joined #git
01:08 woodruffw joined #git
01:13 xock_stream joined #git
01:18 Rodya_ joined #git
01:22 thiago joined #git
01:24 m___ joined #git
01:25 Gobbert joined #git
01:25 polyrob joined #git
01:32 pwnz0r joined #git
01:32 johnny56 joined #git
01:33 Sasazuka joined #git
01:36 raijin joined #git
01:37 manfredn joined #git
01:40 qpdb joined #git
01:44 Garen joined #git
01:44 giraffe joined #git
01:46 frivol joined #git
01:48 GavinMagnus left #git
01:48 mischat joined #git
01:51 pwnz0r joined #git
01:54 pwnz0r joined #git
01:54 Orion3k joined #git
01:56 jaguarmagenta joined #git
02:03 ResidentBiscuit joined #git
02:03 cqi joined #git
02:05 legumbre joined #git
02:08 d^sh joined #git
02:08 morenoh149 joined #git
02:12 floatingman joined #git
02:14 finalbeta joined #git
02:15 arescorpio joined #git
02:15 anonymuse joined #git
02:19 ahmed_elgabri joined #git
02:19 Rodya_ joined #git
02:20 fmcgeough joined #git
02:21 Noldorin joined #git
02:28 jaguarmagenta joined #git
02:31 dodobrain joined #git
02:31 NeverDie joined #git
02:33 LaptopBrain joined #git
02:33 pks joined #git
02:35 mizu_no_oto joined #git
02:36 xo_odin joined #git
02:36 howdoi joined #git
02:38 _leb joined #git
02:48 ilbot3 joined #git
02:48 Topic for #git is now Welcome to #git, the place for git help and serious business | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.10.2 | Getting "cannot send to channel"? /msg gitinfo .voice | The git-jokes project has been suspended for lack of contribution. Help revive it! https://madeitwor.se/git-jokes
02:49 mischat joined #git
02:50 cdg joined #git
02:51 dodobrain joined #git
02:51 steelsw0rd joined #git
02:53 enitiz joined #git
02:53 Vampire0_ joined #git
02:55 xo_odin joined #git
02:57 sgtbigman joined #git
03:00 A5101 joined #git
03:01 aavrug joined #git
03:02 ThunderChicken joined #git
03:02 ThunderChicken joined #git
03:03 griffindy joined #git
03:04 netj joined #git
03:07 jimi_ joined #git
03:07 pigeon joined #git
03:08 sunri5e joined #git
03:09 xo_odin joined #git
03:10 duarteper joined #git
03:14 davimore_ joined #git
03:14 chachasmooth joined #git
03:14 pikajude joined #git
03:15 Rodya_ joined #git
03:15 gmn8719 joined #git
03:20 ianp joined #git
03:21 barajasfab joined #git
03:21 cdg_ joined #git
03:22 hexagoxel joined #git
03:23 Random832 joined #git
03:28 diogenese joined #git
03:30 jimi_sanchez joined #git
03:35 hendry how do I avoid that "Merge branch 'master' of..." commit when I'm git pulling again ?
03:40 rbr joined #git
03:44 Peetz0r joined #git
03:45 Random832 hendry, git pull --rebase
03:45 Random832 or fetch and then decide what to do manually by comparing to the remote tracking branch
03:47 railssmith joined #git
03:48 done joined #git
03:48 rbr joined #git
03:50 govg joined #git
03:50 mischat joined #git
03:51 Mowgli joined #git
03:51 MrWoohoo joined #git
03:54 svm_invictvs joined #git
03:55 phanimahesh joined #git
03:57 morenoh149 joined #git
03:59 a_thakur joined #git
04:01 diogenese joined #git
04:06 gmn8719 joined #git
04:07 JamesBaxter joined #git
04:11 mycyber joined #git
04:15 Vortex34 joined #git
04:15 KnownSyntax joined #git
04:19 romerocesar joined #git
04:23 NeverDie joined #git
04:23 Gsham joined #git
04:23 tjs joined #git
04:27 sangy joined #git
04:32 Masber joined #git
04:32 _leb joined #git
04:34 Goplat joined #git
04:40 subhojit777 joined #git
04:42 Shentino joined #git
04:42 synx joined #git
04:45 dermoth joined #git
04:48 Rodya_ joined #git
04:49 romerocesar joined #git
04:49 gmn8719 joined #git
04:49 romerocesar joined #git
04:49 donmichelangelo joined #git
04:50 fstd_ joined #git
04:51 mischat joined #git
04:52 danielsh left #git
04:53 azerus joined #git
04:59 azerus joined #git
05:01 Rodya_ joined #git
05:02 ayogi joined #git
05:04 dreiss joined #git
05:06 gmn8719 joined #git
05:06 venmx joined #git
05:07 azerus joined #git
05:08 spykins joined #git
05:13 Cabanossi joined #git
05:14 NimeshNeema joined #git
05:14 Anupkumar joined #git
05:16 ismithers joined #git
05:19 Samual joined #git
05:24 Gsham joined #git
05:25 nivag joined #git
05:30 blackwind_123 joined #git
05:30 dreiss joined #git
05:32 _leb joined #git
05:35 bocaneri joined #git
05:35 Andrew_K joined #git
05:37 darkbit joined #git
05:44 romerocesar joined #git
05:44 quesker joined #git
05:45 quesker how can I see the newest file change in this dir?  everything has been committed and pushed some time ago
05:45 morenoh149 joined #git
05:49 cqi joined #git
05:52 ThunderChicken quesker, "ls -lt | head" ?
05:52 mischat joined #git
05:52 quesker ok I assumed the date was from the last clone or whatever
05:53 spykins joined #git
05:53 hhee joined #git
05:54 hhee joined #git
05:54 kadoban quesker: 'git log -- thedir' ?
05:54 quesker ls shows wildly different dates between clones on two machines for me
05:55 kadoban git doesn't track modification date, or try to replicate it on clone or anything
05:56 quesker < ThunderChicken> quesker, "ls -lt | head" ?
05:56 snthsnth joined #git
05:57 quesker anyway that git log -- . seems to be in reverse date order which is just right thanks
05:57 kadoban 'welcome
06:02 Macaveli joined #git
06:04 dreiss joined #git
06:05 WayToDoor joined #git
06:07 spykins joined #git
06:10 svm_invictvs joined #git
06:14 zq joined #git
06:15 zq left #git
06:15 zq joined #git
06:15 zq is there a direct way to retrieve the commit message from inside a --commit-filter script?
06:16 _ikke_ zq: man git filter-branch says it gets the commit message on STDIN
06:16 gitinfo zq: the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html
06:18 sdothum joined #git
06:19 netj joined #git
06:19 zq _ikke_: uh, isn't that only for --msg-filter?
06:20 zq huh, never mind
06:21 dsantiago joined #git
06:21 sea-gull_ joined #git
06:23 leb joined #git
06:24 overlord_tm joined #git
06:24 hendry little confused how to pull a git repo into a repo i just initted ($HOME thing) https://gist.github.com/kaihendry​/640bb90e72a1153ff38e96d3f98b670f
06:25 rscata joined #git
06:26 jfr joined #git
06:28 jeffreylevesque joined #git
06:30 rendar joined #git
06:34 gmn8719 left #git
06:36 blackwind_123 joined #git
06:37 venmx joined #git
06:39 brillenfux joined #git
06:39 ayjay_t joined #git
06:45 a_thakur joined #git
06:45 xall joined #git
06:47 ThomasLocke joined #git
06:47 OnBrokenWings joined #git
06:49 howdoi joined #git
06:51 akushner joined #git
06:53 mischat joined #git
06:54 Anja joined #git
06:54 LucaTM joined #git
07:00 venmx joined #git
07:01 Raging_Hog joined #git
07:01 ngperfex joined #git
07:04 mycyber joined #git
07:05 p4trix joined #git
07:06 qt-x joined #git
07:08 vdv joined #git
07:08 krillr joined #git
07:10 Guest2_ joined #git
07:12 spykins joined #git
07:12 JJjack joined #git
07:12 _ikke_ hendry: https://gist.github.com/kaihendry/640bb90e72​a1153ff38e96d3f98b670f#file-gistfile1-txt-L4
07:12 JeroenT joined #git
07:12 _ikke_ You first need to create a remote
07:12 _ikke_ git remote add <name> <url>
07:13 _ikke_ set-url is for updating an already existing remote
07:13 zeroed joined #git
07:13 zeroed joined #git
07:14 nowhere_man joined #git
07:15 Repox joined #git
07:15 qt-x joined #git
07:16 gniourf joined #git
07:20 PCatinean joined #git
07:20 lb1c joined #git
07:22 freimatz joined #git
07:24 JJjack joined #git
07:25 qws-user-1229 joined #git
07:27 byte512 joined #git
07:27 mrpackethead joined #git
07:27 dreiss joined #git
07:28 ToBeCloud joined #git
07:29 Mowgli joined #git
07:30 robattila256 joined #git
07:31 elect joined #git
07:34 morenoh149 joined #git
07:37 chele joined #git
07:38 Sceorem joined #git
07:40 seishun joined #git
07:42 jnavila joined #git
07:43 rorro joined #git
07:44 vuoto joined #git
07:47 MrMojito joined #git
07:47 MrMojito joined #git
07:47 davimore__ joined #git
07:50 a_thakur joined #git
07:54 mischat joined #git
07:54 adac joined #git
07:55 meauses joined #git
07:59 SwiftMatt joined #git
08:04 hendry _ikke_: thank you btw
08:04 danslo joined #git
08:07 venmx joined #git
08:08 moei joined #git
08:08 jknetl joined #git
08:08 netj joined #git
08:11 acetakwas joined #git
08:13 ahmedelgabri joined #git
08:14 troulouliou_div2 joined #git
08:15 Ciblia joined #git
08:15 TomyWork joined #git
08:16 finalbeta joined #git
08:17 zq left #git
08:17 spykins joined #git
08:18 Rodya_ joined #git
08:20 overlord_tm joined #git
08:21 andrew710 joined #git
08:23 PtxDK_ joined #git
08:24 Alienpruts joined #git
08:27 a_thakur joined #git
08:31 apotry joined #git
08:31 rendar joined #git
08:33 GT4066 joined #git
08:33 afuentes joined #git
08:34 encod3 joined #git
08:34 jaguarmagenta joined #git
08:37 byte512 joined #git
08:37 fees joined #git
08:38 JJjack joined #git
08:42 pks joined #git
08:42 spykins_ joined #git
08:45 jaguarmagenta joined #git
08:47 geoid joined #git
08:49 JeroenT joined #git
08:51 JeroenT joined #git
08:51 brk_ joined #git
08:53 _ikke_ np
08:53 djb-irc joined #git
08:55 LukasT joined #git
08:57 LukasT Hello, we have got weird problem in our git repo: after commiting .gitattributes and normalizing line endings https://help.github.com/articl​es/dealing-with-line-endings/ when I switch to some branch, some files are immediately shown as changed
08:57 LukasT only whitespace changes are visible
08:57 LukasT I mean git diff -w returns empty output , so the changes are only white-space
08:58 LukasT git stash does not get rid of that change
08:58 LukasT I have to commit it
08:58 LukasT Any ideas what's wrong?
08:59 kedare joined #git
09:00 osse LukasT: what does 'file example/file' say ?
09:01 osse wher example/file is one of the changed ones
09:06 g_zen joined #git
09:08 ExeciN joined #git
09:08 venmx joined #git
09:08 marcogmonteiro joined #git
09:10 kurkale6ka joined #git
09:10 rorro joined #git
09:12 LukasT osse: I followed this advice and it helped http://stackoverflow.com/a/18792689/1068199
09:13 LukasT osse:  C++ source, ASCII text, with CRLF line terminators
09:13 lindenle joined #git
09:17 jaguarmagenta joined #git
09:17 mozzarella joined #git
09:17 manuels joined #git
09:18 Snugglebash joined #git
09:18 LukasT left #git
09:19 lektrik joined #git
09:19 mozzarel1 joined #git
09:20 LukasT joined #git
09:22 gahbu_ joined #git
09:22 Nilesh_ joined #git
09:22 wrksx joined #git
09:23 morenoh149 joined #git
09:23 berndj joined #git
09:27 zeroed joined #git
09:31 fbwnd joined #git
09:33 BadSnappa joined #git
09:34 Snugglebash joined #git
09:35 ferr joined #git
09:36 Endarked joined #git
09:39 blahdodo joined #git
09:39 JeroenT joined #git
09:41 nettoweb joined #git
09:42 JeroenT_ joined #git
09:44 xocket joined #git
09:48 davimore_ joined #git
09:53 ivh_ joined #git
09:55 xaa joined #git
09:56 oskarkv joined #git
09:57 chll_ joined #git
09:57 sunri5e joined #git
09:58 snthsnth joined #git
09:59 Serpent7776 joined #git
09:59 darkbit joined #git
10:00 p4trix joined #git
10:02 mindriot101 joined #git
10:03 ivh_ Hi. I added post-receive hook to a non-bare repo and according to the docs, the working directory for the script should be the repo's root dir. But it is .git/ instead for some reason. Any insights? (Running 2.10 on both ends, but repo has been created with older version, in case this matters)
10:03 ocbtec joined #git
10:05 ViEwHoSANyC42Kt6 left #git
10:06 GavinMagnus joined #git
10:07 done joined #git
10:08 tribly_ joined #git
10:09 venmx joined #git
10:09 mablae joined #git
10:11 mindriot101 joined #git
10:12 osse ivh_: it might be a bug in the docs, or the hook. Usually post-receive isn't used in non-bare repos
10:12 osse but with that said, cd "$(git rev-parse --show-toplevel)" should do the trick
10:13 mischat joined #git
10:13 subhojit777 joined #git
10:13 romerocesar joined #git
10:14 ivh_ thanks! with receive.denyCurrentBranch updateInstead it seems to wokr nicely to push into non-bare repos. I use this to deploy my little site.
10:16 osse You might end up hosting the .git dir too. But I suppose that's fixable with .htaccess or equivalent
10:16 Snugglebash joined #git
10:16 osse The traditional approach is a bare repo and some inspirating from here: !deploy
10:16 gitinfo Git is not a deployment tool, but you can build one around it (in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/deploy.html
10:16 osse But if it works...
10:17 xall joined #git
10:17 ivh_ osse: sure. the hook is to build the static html from the repo and then sync it locally to the hosting-directory.
10:18 osse ah
10:20 ivh_ btw, git rev-parse --show-toplevel does not work when inside .git/
10:25 cdown joined #git
10:27 nowhereman joined #git
10:28 einnjo joined #git
10:29 dviola joined #git
10:30 troulouliou_div2 joined #git
10:30 venmx joined #git
10:31 meauses joined #git
10:40 mindriot_ joined #git
10:42 encod3 joined #git
10:47 ayjay_t joined #git
10:48 clmsy joined #git
10:48 rnsanchez joined #git
10:51 courrier joined #git
10:53 k_sze[work] joined #git
10:53 k_sze[work] Is there a fast way to git diff two different files from two different commits?
10:54 Seveas k_sze[work]: git diff commit1-sha1:path/to/file1 commit2-sha1:path/to/file2
10:55 k_sze[work] Thanks
10:56 vuoto_ joined #git
10:58 vuoto joined #git
10:59 vuoto_ joined #git
11:00 a_thakur joined #git
11:01 vuoto joined #git
11:08 marvi joined #git
11:10 rorro joined #git
11:11 marvi joined #git
11:12 morenoh149 joined #git
11:13 Darcidride joined #git
11:15 re1 joined #git
11:15 cdown joined #git
11:16 vuoto joined #git
11:16 hhee joined #git
11:17 vuoto joined #git
11:17 jaguarmagenta joined #git
11:23 Milos joined #git
11:25 Oatmeal joined #git
11:27 GavinMagnus left #git
11:31 denisMone joined #git
11:32 a3Dman joined #git
11:32 BackEndCoder joined #git
11:33 sdelic joined #git
11:33 elastix joined #git
11:37 JeroenT joined #git
11:37 polyrob_ joined #git
11:43 re1 joined #git
11:44 converge joined #git
11:46 dodobrain joined #git
11:48 anonymuse joined #git
11:49 Snugglebash joined #git
11:49 GavinMagnus joined #git
11:49 hariel joined #git
11:50 paladinn joined #git
11:55 GavinMagnus joined #git
11:55 brillenfux joined #git
11:57 cerkauskas joined #git
11:57 cerkauskas joined #git
11:58 cerkauskas hello. is there a way to force pull and ignore conflicts (pull latest version available) ?
12:00 shinnya joined #git
12:00 tobiasvl cerkauskas: do you want to merge, favoring the remote's version? or do you want to scrap your local history and replace it with the remote's history?
12:01 tobiasvl I assume the latter based on your wording, but you never know
12:01 cerkauskas tobiasvl: I completely don't care about local changes.
12:02 tobiasvl cerkauskas: easiest way then is to fetch and then just "git reset --hard REMOTE_NAME/BRANCH_NAME"
12:04 lamont joined #git
12:07 bauruine joined #git
12:08 dsdeiz joined #git
12:10 hahuang65 joined #git
12:11 cerkauskas tobiasvl: thank you :)
12:12 King_Hual joined #git
12:12 Achylles joined #git
12:14 Shoplifting joined #git
12:15 isysd joined #git
12:15 GavinMagnus joined #git
12:18 vuoto joined #git
12:24 al-damiri joined #git
12:25 johnmilton joined #git
12:28 xall joined #git
12:29 brk_ joined #git
12:31 mycyber joined #git
12:35 UniFreak joined #git
12:36 UniFreak I want to revert my repo (already pushed) to a previous commit id. but when I do `git rever <sha>` it give me `didn't supply a -m option` error
12:36 UniFreak what's that? how to fix?
12:38 szgyg joined #git
12:38 zeroed joined #git
12:38 zeroed joined #git
12:41 tobiasvl UniFreak: you need to specify which parent of the merge to revert to. likely you want to do "git revert -m 1 <sha>" for first parent of the mainline
12:42 cdown joined #git
12:43 UniFreak tobiasv1: under what condition it will need the -m option?
12:44 noecc joined #git
12:45 tobiasvl when reverting merges
12:47 datamaan joined #git
12:48 osse since he's saying "*to* a previous commit" I'm thinking git-revert is not the tool for the job here
12:48 osse !revert
12:48 gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?
12:49 bathtub_shark joined #git
12:49 datamaan hi. I've got a folder with some version of a software from github but I have no idea which commit my version of the code corresponds to (I have just the code. no git repo). is there a way to find the corresponding commit for my code-version?
12:49 f3r70rr35f joined #git
12:51 snufft joined #git
12:51 steelsw0rd joined #git
12:53 zeroed joined #git
12:53 JJjack joined #git
12:54 adac joined #git
12:54 WayToDoor joined #git
12:55 porsche944 joined #git
13:01 morenoh149 joined #git
13:03 bathtub_shark joined #git
13:03 synthroid joined #git
13:06 Learath2 joined #git
13:08 cdown joined #git
13:08 kesslerfrost joined #git
13:10 texinwien joined #git
13:10 sdothum joined #git
13:12 dminuoso joined #git
13:15 barduck joined #git
13:15 Seveas datamaan: commit it as a new root commit and diff it against all the existing commits :)
13:18 jaguarmagenta joined #git
13:18 datamaan Seveas: I just started writing a script that iterates all commits, does a checkout and diffs (ignoring the .git folder). but committing as new root commit is a nice improvement
13:19 Raed joined #git
13:21 boombatower joined #git
13:26 snufft I've got an email from a friend saying "See latest commit (tag/v7.01)" and he's since gone to bed. Can anyone tell me how I would see tag/v7.01? I've never had to look for anything like that before :S
13:27 bremner git tag -l
13:27 bremner probably after "git fetch --tags"
13:27 serialoverflow joined #git
13:28 mrkake joined #git
13:29 jost joined #git
13:29 snufft bremner, thanks for that :) git tag -l gives me the list of versions and I can see v7.01 in there, but what do I do with it now? do I (can I) diff v7.01 against the previous version?
13:30 tobiasvl you can do basically anything with it that you can do with other refs
13:30 snufft that's what I mean though, I've never had to deal with other refs before. I literally write my code, pull, commit, push. that's all I've ever needed to do with git, haha
13:30 brk_ joined #git
13:31 mizu_no_oto joined #git
13:31 snufft the repo master (the guy who emailed me) is the one who takes care of all of the other stuff
13:31 bremner snufft: yes. try "git diff v7.01"
13:32 snufft awesome, thanks bremner :D
13:33 drbean joined #git
13:34 jeffreylevesque joined #git
13:37 bathtub_shark joined #git
13:39 johnmilton joined #git
13:42 JeroenT joined #git
13:44 fmcgeough joined #git
13:45 Snugglebash joined #git
13:46 porsche944 joined #git
13:48 mindriot101 joined #git
13:51 jimi_ joined #git
13:53 JamesBaxter joined #git
13:54 empedokles78 joined #git
13:55 jaguarmagenta joined #git
13:56 bathtub_shark joined #git
13:56 przemoc joined #git
13:56 empedokles78 When I git push I receive a Warning: "'push.default' is not set. the implicit value was changed from matching to simple in git 2.0. It gives me the options matching or simple. What do I need to do?
13:57 tobiasvl follow the instructions?
13:57 Starky joined #git
13:57 tobiasvl as far as I remember (it's been a while since 2.0) it's pretty extensive
13:58 mindriot101 joined #git
13:58 xock_stream joined #git
13:58 empedokles78 tabiasvl, it gives me both options. it's a new git install.
13:59 empedokles78 e.g. git config --global push.default matching or --global push.default simple
13:59 xall joined #git
13:59 tobiasvl yeah but it should also explain them, right?
14:00 djb-irc joined #git
14:00 tobiasvl only you can choose which one you want, but "simple" is the new default so I'd go for that if you're not extremely used to "matching"
14:01 empedokles78 if it's matching, local branches willbe sent to the remote branches with the same name.
14:01 Gsham joined #git
14:02 empedokles78 in simple the actual branch will be sent to the remote-branch, which git-pull uses to actualize the actual branch.
14:02 empedokles78 that's what it says.
14:02 empedokles78 I don't understand that though :)
14:03 mizu_no_oto joined #git
14:04 osse what don't you understand about it
14:05 timvisher joined #git
14:06 denisMone joined #git
14:06 empedokles78 osse, the difference.
14:07 osse empedokles78: your repo and the remote repo probably have a bunch of branches in common.  With "matching", all those in common will be pushed when type 'git push' and nothing else. With "simple", only the one you have checked out will be pushed
14:07 davimore__ joined #git
14:08 empedokles78 what do you mean with checked out?
14:08 osse The one you're working on now
14:09 osse the one whose name is printed when you do git status and it says "On branch foo"
14:09 Gsham joined #git
14:10 sdothum joined #git
14:10 empedokles78 osse, why does it make sense to push files that are already on gitlab if they are the same with matching?
14:10 osse that question does not make sense. you don't push files
14:10 osse you push a series of commits
14:10 JeroenT joined #git
14:11 osse the BRANCH names are the same, not the FILE names
14:14 empedokles78 osse, but aren't my files on gitlab?
14:15 romerocesar joined #git
14:15 kesslerfrost joined #git
14:15 osse yes
14:15 tobiasvl (although no, because it's likely a bare repo, but that doesn't really matter)
14:16 osse tobiasvl: they're probably in /tmp somewhere :p you can click on them!
14:16 tobiasvl hehe
14:17 empedokles78 osse, I'm the single developer. It is more or less just used to be able to develop on my laptop and on my local machine.
14:17 empedokles78 so simple?
14:18 osse yes
14:18 osse it's the new default, it's probably what people expect.
14:18 osse go with that
14:18 irco joined #git
14:20 empedokles78 git add --all does check out?
14:20 ExoUNX joined #git
14:20 oaao joined #git
14:21 tobiasvl no
14:21 bathtub_shark joined #git
14:22 venmx joined #git
14:23 cdg joined #git
14:24 JeroenT joined #git
14:25 hhee joined #git
14:25 djh joined #git
14:26 ShekharReddy joined #git
14:26 osse checkout does check out
14:26 Darren_ joined #git
14:26 JeroenT joined #git
14:26 junw left #git
14:26 JeroenT joined #git
14:27 claudiu04 joined #git
14:30 empedokles78 osse, tobiasvl, I never used a checkout command. is it needed?
14:30 govg joined #git
14:30 osse that's a very broad question
14:30 osse if you only work on one branch it's almost never needed. but it can be useful.
14:32 empedokles78 osse, it hasn't been used in the tutorial. I guess it only uses one branch. git status says: On branch master. your branch is on the same level als 'irigin/master'. nothing to commit.
14:32 osse if you have uncommitted changes it can say that use can use checkout to get rid of the changes
14:33 Gsham joined #git
14:34 lektrik joined #git
14:34 peeps[lappy] joined #git
14:34 empedokles78 I don't understand.
14:34 mda1 joined #git
14:34 osse change a file. run git status again. you will see.
14:36 mindriot_ joined #git
14:36 empedokles78 I see it says that too, because I haven't made a commit?
14:36 j`ey joined #git
14:36 empedokles78 what do you mean by checkout?
14:37 j`ey is there a simpler way to do this: git stash ; git checkout X ; git stash pop ; git diff HEAD~1
14:37 tobiasvl empedokles78: man git checkout
14:37 gitinfo empedokles78: the git-checkout manpage is available at http://jk.gs/git-checkout.html
14:37 jaguarmagenta joined #git
14:38 Gloomy joined #git
14:39 nicksloan joined #git
14:40 osse j`ey: depending on the changes between the current branch and X then stash might not be necessary
14:40 empedokles78 thanks. that's hard to grasp, I guess I have to do that codeacademy course.
14:40 j`ey osse: fair enough. I cant really explain what I want to do
14:41 j`ey basically I committed something, and I got some comments, so I'm re-doing the patch
14:41 j`ey I want to show the new patch, as if I never commited anything in the first place
14:42 osse sounds about right
14:46 WayToDoor joined #git
14:47 derk0pf joined #git
14:47 ResidentBiscuit joined #git
14:48 guampa joined #git
14:48 kesslerfrost joined #git
14:50 morenoh149 joined #git
14:51 EricPoe joined #git
14:51 AaronMT joined #git
14:55 cdg_ joined #git
14:55 rorro joined #git
14:58 xocket joined #git
15:03 converge joined #git
15:03 Dweezahr joined #git
15:06 ech0s7 joined #git
15:06 kpease joined #git
15:07 aax joined #git
15:07 adac joined #git
15:11 Es0teric joined #git
15:12 JeroenT_ joined #git
15:12 JanC_ joined #git
15:13 JeroenT joined #git
15:16 Phylock joined #git
15:16 Dougie187 joined #git
15:17 lindenle joined #git
15:17 d0nn1e joined #git
15:20 nettoweb joined #git
15:20 Learath2 joined #git
15:21 davidfetter_ge joined #git
15:22 netj joined #git
15:27 venmx joined #git
15:31 jeffreylevesque joined #git
15:33 Darren_ joined #git
15:33 dar123 joined #git
15:38 charles` fragMental: yeah I have it as a bare repo, but it wasn't created with --group=shared and I want to convert it; lots of answers I have found seem pretty complicated and a lot of disagreement about how to do it
15:38 IrishPrime joined #git
15:38 Endarked joined #git
15:38 xall joined #git
15:39 aarobc joined #git
15:40 manuels joined #git
15:40 joki- joined #git
15:41 bathtub_shark joined #git
15:43 LionsMane joined #git
15:43 morenoh149 joined #git
15:43 byte512 joined #git
15:46 davimore_ joined #git
15:48 mindriot101 joined #git
15:49 nikivi joined #git
15:53 synthroid joined #git
15:54 Shentino joined #git
15:54 synthroid joined #git
15:55 synthroid joined #git
15:56 synthroid joined #git
15:56 Mowgli joined #git
15:57 synthroid joined #git
15:57 mizu_no_oto joined #git
15:58 git_guest joined #git
15:58 xock_stream joined #git
16:01 git_guest I have a computer without network connectivity, and I need to sync a git repo to that machine. Is there an easy way to update the offline machine without having to copy the entire repo to media each time?
16:01 darkbit joined #git
16:04 des_consolado joined #git
16:04 davidfetter_ge joined #git
16:04 bathtub_shark joined #git
16:04 canton7 git_guest, seen git bundles?
16:05 des_consolado is there such thing as some kind of git package manager? it would be cool if there was something like python virtualenv but for git stuff, and a gitty kind of package manager that could maintain and keep up to date your repo's if you want, and also if you do stuff like build them there's the virtual env where it installs the binaries and all this and that...
16:05 des_consolado what do you think and has anything similar been done?
16:06 des_consolado I've got a new macbook and I'm concerned about how I'm going to keep my system clean if/when something isn't on brew and I have to install manually some code from some git
16:06 canton7 I don't understand why git has anything to do with installing stuff
16:06 King_Hual joined #git
16:06 git_guest canton7: I wasn't aware of them. Reading the help, I could bundle giving the revision on the unconnected machine to get a file that has everything from that rev up to latest?
16:06 des_consolado often projects come with build scripts which will build stuff and output it to system directories
16:07 mischat_ joined #git
16:07 King_Hual joined #git
16:07 King_Hual joined #git
16:07 eivarv joined #git
16:08 Seveas package management and version control are two different things. Don't try to do the one with the other.
16:08 canton7 git_guest, another option is to put a bare repo on a USB stick, add it as a remote to git repos on the offline machine and another, and push to / fetch from it
16:08 hfp joined #git
16:08 des_consolado yeah that was a silly thought actually, but the idea of a generic virtual environment I think would be cool
16:08 canton7 des_consolado, why git specifically? why not hg, or SVN, or anything? Why does this have anything to do with git?
16:08 des_consolado I don't see why virtualenv is just a python thing
16:09 empedokles78 left #git
16:09 des_consolado well okay point taken
16:09 git_guest canton: unfortunately I can only use write-once media, and I don't need to push to it from the unconnected machine, I'm just getting the latest for some third party projects
16:09 canton7 git_guest, ah, air-gapped machines? fun
16:10 git_guest des_consolado: bsd has jails and linux has chroots
16:10 canton7 git_guest, and yeah, your understanding of bundles is right. Tbh it's probably just easier to create a bundle with everything in it (unless size is an issue) - if you get the 'from' revision wrong, you're screwed
16:10 oncall-pokemon joined #git
16:11 ojdo joined #git
16:11 qqx Depends which way you get the 'from' revision wrong.
16:11 git_guest canton7: I currently just use a script to update a bunch of repos and 7zip the whole thing, and the size is getting prohibitive. would the bundles be much smaller that the whole repo?
16:12 canton7 git_guest, a bundle of everything might be slightly smaller than the whole repo (.git dir). If you specify a 'from' revision, then the size will get smaller
16:13 git_guest if not, it seems like it'd be simple enough to have the script record the version last bundled and use that, and I could manually edit the revisions if somehow the recorded one was ever wrong
16:13 qqx I just created both a bundle and a bare clone of a repo, the bundle was about 1/3 the size.
16:14 zeroed joined #git
16:14 zeroed joined #git
16:14 GavinMagnus joined #git
16:14 qqx Actually, I made a mistake there using a local clone.
16:14 clemf joined #git
16:14 qqx Cloning using a file:// URL made the bare clone much smaller.
16:15 qqx Although still a bit larger than the bundle.
16:15 j`ey left #git
16:16 crose joined #git
16:16 git_guest is there an easy command-line to get the current latest revision(s)?
16:16 canton7 they should be broadly comparable, as both use the same pack format, iirc
16:16 canton7 git_guest, `git rev-parse HEAD` for the current branch
16:19 git_guest thanks
16:19 git_guest I wish channels for other projects were so helpful =-)
16:20 Muzer joined #git
16:21 Thorn joined #git
16:23 lordjancso joined #git
16:23 tyreld joined #git
16:25 blackwind_123 joined #git
16:26 rockstar___ joined #git
16:27 Dandelock joined #git
16:28 Vgr joined #git
16:28 Vgr joined #git
16:30 moriko joined #git
16:30 Snugglebash joined #git
16:31 sagerdearia joined #git
16:32 nikivi joined #git
16:32 lordjancso_ joined #git
16:33 chardan joined #git
16:34 mikecmpbll joined #git
16:35 chardan Hi, everyone... I have a branch that's getting tons of bad merges, and basically I want to make the head of my branch identical to the head of master in another repo. I want to not have any of "my" changes. I've tried rebase and merge and keep winding up with conflicts. What's the right incantation?
16:35 eivarv joined #git
16:35 a_thakur joined #git
16:35 bathtub_shark joined #git
16:35 kulelu88 joined #git
16:36 zeroed joined #git
16:36 zeroed joined #git
16:36 subhojit777 joined #git
16:37 thomasreggi joined #git
16:37 a_thakur joined #git
16:38 ResidentBiscuit reset --hard master
16:40 thebope joined #git
16:41 chardan ResidentBiscuit: I'll read the man page for reset! Thanks!
16:43 Vampire0 joined #git
16:45 crayon joined #git
16:45 panzon joined #git
16:46 ResidentBiscuit reset is one of the more complicated commands thats used a lot
16:47 chardan ResidentBiscuit: Yeah, it looks like all the options have different implications. I don't want to /lose/ my history, but do want to essentially "reset" the head.
16:48 chardan ResidentBiscuit: I think it's probably what I want, in some variation-- thank you!
16:48 NeverDie joined #git
16:49 ResidentBiscuit Oh so you tried to just rebase onto master?
16:50 ayjay_t joined #git
16:50 chardan ResidentBiscuit: Yeah, I've tried rebase and merge, neither seems to do quite what I want. I wind up with a broken and complicated set of things that I won't know how to merge-- but /do/ know that the other repo is what I "would" want. I have a small number of my own changes that I want to replay over that newer repo.
16:51 ResidentBiscuit There's no magic for getting around conflicts if you want to keep both your changes and take whats changed on mater
16:51 ResidentBiscuit master*
16:51 hfp joined #git
16:51 tang^ joined #git
16:51 chardan ResidentBiscuit: I can manage conflicts with /my/ code, but I don't know what to do with all the other ones.
16:52 ResidentBiscuit You might have to bring in someone else then. Frequent rebasing helps this situation
16:53 paladinn joined #git
16:53 paladinn joined #git
16:53 LarArT joined #git
16:54 ayogi joined #git
16:55 einnjo joined #git
16:56 sim642 joined #git
16:56 LionsMane Huh, ok, I am trying to cherry pick SVN commits from a svn-remote branch into a git branch.  The line that says "git-svn-id" has a GUID looking number after the svn path.  Does anyone know what that is?  I thought it was the check SHA1 hash.
16:57 davimore joined #git
16:58 kulelu88 left #git
16:58 lindenle joined #git
16:58 canton7 LionsMane, it's the UUID of the SVN repo, and the revision of that commit. It's used by git-svn to match git commits up to SVN commits
16:59 kesslerfrost joined #git
16:59 LionsMane canton7: Ah ok.  Phew!  I thought I had figured out how to do the git-svn tool and then it model broke down somewhere.
16:59 xocket joined #git
17:00 LionsMane I also had a git and svn upgrade in the middle of it.
17:01 thomasreggi joined #git
17:02 meauses joined #git
17:03 dooglius joined #git
17:03 Hobbyboy joined #git
17:03 cdown joined #git
17:05 hhee joined #git
17:05 safe joined #git
17:06 overlord_tm joined #git
17:08 WayToDoor joined #git
17:08 bathtub_shark joined #git
17:10 hanthings joined #git
17:11 Masber joined #git
17:11 jstein_ joined #git
17:12 Cavallari joined #git
17:13 Siegfried joined #git
17:13 nicksloan joined #git
17:16 eivarv joined #git
17:17 mmlb joined #git
17:17 crayon joined #git
17:18 mmlb joined #git
17:20 mrpackethead joined #git
17:24 kourier joined #git
17:24 bathtub_shark joined #git
17:25 yanome joined #git
17:27 vuoto joined #git
17:28 ochorocho__ joined #git
17:29 hfp joined #git
17:29 lexxis joined #git
17:31 xock_stream joined #git
17:32 algae joined #git
17:32 InfoTest joined #git
17:32 mizu_no_oto joined #git
17:32 Rodya_ joined #git
17:33 kourier Not sure which doc to study in-depth first. I think I can do this with a rebase fixup back to the first commit in the new repo, but was hoping you could tell me I'm on the right path. I want to open one subdirectory of a parent repo to the community for collaboration as a new, pristine repository with no trace of my prior commits. I want to be able to share changes between the parent repository.
17:33 git_guest canton7: what syntax would I use when creating a bundle to mean 'everything after 1234...'? just give that rev numer?
17:33 eivarv joined #git
17:34 vuoto joined #git
17:34 KekSi joined #git
17:34 vuoto joined #git
17:35 gahbu_ joined #git
17:35 thomasreggi joined #git
17:36 bathtub_shark joined #git
17:36 kourier To start, I checked out a branch in the parent repo and did `git rm` on the irrelevant bits, converted .gitignore to a whitelist selecting the modulus, and commit. Then I did `git init` new repo, `git add remote origin`, and `git pull origin subdironly:master`.
17:37 tmz joined #git
17:38 kourier Then I did in the new child repo `git rebase -i --root -s recursive -X theirs ` and did `fixup` to squash all commits to the root.
17:40 hfp joined #git
17:40 prg3_ left #git
17:40 kourier It is at this point that it is not clear how to proceed with sharing changes between the now-rewritten log and the parent.
17:41 prg3 joined #git
17:41 hfp_work joined #git
17:42 qqx kourier: Have you looked at git-subtree ?
17:42 kadoban Yeah, that sounds like exactly what 'git subtree' is for
17:43 inflames joined #git
17:43 dreiss joined #git
17:43 kourier Oh ok, I hadn't yet picked apart the distinction between the different subtree or sub-tree commands for git, but I'll definitely go have a look at `git subtree`. Thanks y'all.
17:43 Cavallari joined #git
17:45 git_guest whenever I try git bundle --create file.bundle rev, I get an "error: could not open 'file.bundle'"
17:47 seishun joined #git
17:48 git_guest as far as I can tell, there is no reason the file couldn't be created. I'm on windows and I have full access to the folder
17:49 xPucTu4 joined #git
17:49 kourier qqx: kadoban: Score. This is precisely what I need to share objects separately from commits. Thankya.
17:49 qqx git_guest: That should be "create" rather than "--create"
17:49 svm_invictvs joined #git
17:50 kadoban kourier: 'welcome, good luck. It's a bit funky at first, don't give up ... it gets easier to use/understand.
17:50 Mobutils joined #git
17:51 git_guest qqx: thanks
17:52 dsdeiz joined #git
17:52 CEnnis91 joined #git
17:53 hfp_work joined #git
17:53 mischat joined #git
17:53 neilthereildeil joined #git
17:54 nikivi joined #git
17:54 ELFrederich joined #git
17:55 neilthereildeil hi. whats the right codereview workflow if i am maintaining a tree? Should i pull all remote changes into a local branch that tracks the remote repo and only merge to master when the remote repo has my comments fixed?
17:55 git_guest so I can apparently use rev..master to create an update bundle. Is there a syntax I can use if I want to include the entire repo (for cases where there isn't anything to update)?
17:55 ELFrederich What are the options for having user/group kind of security with Git?  I know GitLab and Gerrit do this, but is there anything more lightweight?
17:56 qqx git_guest: You can just use "master"
17:57 kadoban ELFrederich: You've looked at gitolite? It provides quite configurable security options
17:58 ELFrederich kadoban: I haven't, I will now... anythign else I should look at?  We're using Git as an underlying data store for an application, not using it directly, but still want to have a concept of users/groups
17:59 kadoban Not that comes to mind
17:59 ToBeFree joined #git
17:59 ELFrederich kadoban: GitLab is great, but it offers a ton of stuff we don't really need.  Ideally we just need users/groups and a web interface isn't necesary but would be nice to have
18:00 grawity sounds like gitolite
18:00 Es0teric joined #git
18:00 qqx ELFrederich: It's also possible to use OS users and groups, although you have to be careful with how permisisons get applied to newly created files.
18:02 kadoban I have had quite a bad time with OS users/groups in the past with git. It was ... annoying. Though I was using it for something quite different from what it seems you are, so maybe it'd work better for you.
18:02 crose joined #git
18:03 rorro joined #git
18:07 madewokherd joined #git
18:09 unreal joined #git
18:09 bathtub_shark joined #git
18:09 vdv joined #git
18:11 aarobc joined #git
18:14 git_guest after unbundling, do I need to do something to update the working copy to the latest?
18:15 fmcgeough joined #git
18:15 romerocesar joined #git
18:15 robotroll joined #git
18:16 cek left #git
18:16 xer0x_ left #git
18:16 sea-gull_ joined #git
18:18 tvw joined #git
18:23 a3Dman joined #git
18:24 zeroed joined #git
18:26 waterCreature joined #git
18:26 davidfetter_ge joined #git
18:26 waterCreature hello, if I do commit, which branch will it be pushed to?
18:27 nikivi joined #git
18:27 Akimb joined #git
18:27 tobiasvl waterCreature: if you commit, it will be commited to the currently checked-out branch. if you then later push, it will be pushed to the branch you push to, or if you don't push to a specific branch, it depends on default settings
18:28 nettoweb joined #git
18:28 NeverDie joined #git
18:29 tobiasvl but in most scenarios, to the branch on the remote called "origin" which is called the same as the local branch
18:30 adac joined #git
18:31 xocket joined #git
18:31 rockstar_ joined #git
18:31 Vinnie_win_x joined #git
18:32 dsantiago joined #git
18:32 nikivi joined #git
18:33 pitastrudl joined #git
18:33 navidr joined #git
18:34 waterCreature joined #git
18:35 pitastrudl joined #git
18:39 Bass10 joined #git
18:40 averell joined #git
18:42 dreiss joined #git
18:42 jast joined #git
18:43 Domhack joined #git
18:44 seishun joined #git
18:44 JeremyD joined #git
18:45 railssmith joined #git
18:46 cyan__ joined #git
18:48 pitastrudl joined #git
18:49 nikivi joined #git
18:56 ShapeShifter499 joined #git
18:57 qqx git_guest: Normally you don't specifically unbundle. Instead you'd use `git fetch` or `git pull` to get data out of the bundle.
18:57 kadoban I wasn't even aware that "unbundle" was a thing
18:58 git_guest for the purpose of integrating a bundle, what is the difference between fetch and pull?
18:58 qqx kadoban: It's intended as a plumbing for use by fetch. But is somewhat prominently documented in the bundle man page.
18:59 qqx git_guest: Same as the difference for working with other repositories.
18:59 Jackneilll joined #git
18:59 qqx fetch will just bring the data into your current repo, pull will do that as well as merge or rebase to integrate the fetched commits into your current branch.
18:59 LBV_User joined #git
19:01 git_guest thank you for the explanation, I was having a bit of trouble mentally parsing the help
19:01 qqx Basically once you have the bundle file on the destination system you just treat it as if it's a bare git repo.
19:02 NeverDie_ joined #git
19:02 mindriot101 joined #git
19:03 Siegfried joined #git
19:04 encod3 joined #git
19:04 kourier joined #git
19:05 crose joined #git
19:05 git_guest The only git features I've used are clone and pull (from the original that was cloned), so I'm not at all familiar with inter-repo or multi-repo stuff stuff
19:07 qqx git_guest: `git fetch /path/to/bundle` should create a ref named FETCH_HEAD.
19:07 thebope joined #git
19:08 git_guest I really appreciate all the help. I'm transforming a several-hour-long repo sync into a few minute process thanks to the assistance
19:08 qqx Then you can do `git log ..FETCH_HEAD` to show what's new. And `git merge FETCH_HEAD` or `git rebase FETCH_HEAD` to actualy pull it in.
19:09 nikivi joined #git
19:09 gargola joined #git
19:09 git_guest in my script, I save off the last-transferred revision number and create a bundle from that number to master. Sometimes, that will result in an empty set and I get an error message, git fails to delete the lockfile, and there is a prompt asking whether it should try again. Is there a flag I can use to suppress the prompt?
19:11 nettoweb joined #git
19:13 qqx git_guest: I don't think there's a way to prevent git bundle from issuing an error if asked to create an empty bundle.
19:14 git_guest qqx: I don't mind the error message, just the prompt about trying to remove the lockfile again (because it prevents it from being fully automated)
19:14 qqx You could check if `git rev-list -1` outputs anything with the ref specs you're going to pass to bundle.
19:14 qqx And only proceed if that generates output.
19:16 nicksloan joined #git
19:16 Endarked joined #git
19:16 netj_ joined #git
19:17 adac joined #git
19:18 git_guest qqx: thanks, that works perfectly
19:19 zeroed joined #git
19:21 jimi_ joined #git
19:22 snthsnth joined #git
19:23 fmcgeough joined #git
19:29 bathtub_shark joined #git
19:30 jorj joined #git
19:30 SwiftMatt joined #git
19:32 xock_stream joined #git
19:32 rossome joined #git
19:35 danslo joined #git
19:35 einnjo joined #git
19:37 manuels joined #git
19:37 neilthereildeil left #git
19:38 git_guest git bundle unbundle seemed to work correctly, but then I had to manually update-ref the local head. When I try git pull or git fetch on the bundle, they complain "fatal: Couldn't find remote ref HEAD"
19:39 johnny56 joined #git
19:40 todd_dsm joined #git
19:41 jancoow joined #git
19:41 encod3 joined #git
19:41 kriskropd joined #git
19:43 jancoow Hi. For a report i want a nice git log diagram. However, I want the name of EVERY branch name when they are merged in the left column
19:43 Spec-Chum joined #git
19:43 jancoow how can i do that
19:43 chardan joined #git
19:45 Flowinho joined #git
19:47 sea-gull_ joined #git
19:47 neilthereildeil joined #git
19:47 neilthereildeil if i have tags in my repo, and i pull from another repo, what happens to my tags?
19:47 jancoow they will stay. tags are bound to commit hashes
19:48 neilthereildeil what if someone pulls from me?
19:48 bathtub_shark joined #git
19:49 jancoow did you pushed them to the remote git server?
19:49 chrisawemoore joined #git
19:51 kaldoran joined #git
19:52 converge joined #git
19:52 converge joined #git
19:52 msonntag joined #git
19:55 xall joined #git
19:56 manuels joined #git
19:57 johnny56 joined #git
19:57 vuoto joined #git
19:57 NeverDie_ joined #git
19:58 noecc left #git
19:59 m0viefreak joined #git
19:59 nettoweb joined #git
20:01 griffindy joined #git
20:01 manuels joined #git
20:06 dendazen joined #git
20:06 nidr0x joined #git
20:06 johnny56_ joined #git
20:07 skunkz joined #git
20:07 porton joined #git
20:07 porton I wrote some software but then we have decided to remove some features from my software. I want to save the old version however. Should I save it as a tag or as a branch?
20:08 grawity well, it'd be saved by Git no matter what
20:08 cdown joined #git
20:08 grawity but what you're describing sounds a bit like release versions, which are usually tagged
20:09 porton grawity: what is the advantage of a tag over a branch?
20:10 grawity tags are meant to be immutable, and describe one *specific* version
20:10 grawity e.g. v1.0.4
20:10 porton grawity: I know this
20:10 grawity and that's the advantage
20:10 grawity if you get a specific tag, it'll have the exact same features every time
20:10 mindriot101 joined #git
20:11 grawity branches might be specific to e.g. v1.0.x or v1.x but they can still change
20:11 grawity and who knows, maybe next month you'll decide to remove some of those features from v1.0 as well
20:14 panzon joined #git
20:15 neilthereildeil jancoow: no i didnt push them. but what if someone pulls from me?
20:17 moriko joined #git
20:19 rodyhaddad joined #git
20:19 zeroed joined #git
20:19 qqx git_guest: The bundle will only contain refs that you specifically added. You can either include HEAD when creating the bundle, or specify a different ref (e.g. master) when fetching or pulling.
20:20 bathtub_shark joined #git
20:20 qqx If you include HEAD in the bundle you don't need to specify the starting point multiple times. You can do `git bundle create /path/to/create last-update..master HEAD`.
20:22 aidalgol joined #git
20:22 qqx Although if you're doing that it might be better to specify the revisions as: \^last-update master HEAD
20:22 p4trix joined #git
20:22 snufft joined #git
20:22 MrWoohoo joined #git
20:23 rodyhaddad Hey fellas, does anybody know of a command to deletes all remote branches that don't exist locally?
20:23 rodyhaddad Most searches bring up the opposite, deleting local branches that aren't in remote, but I'm looking to delete remote branches that aren't in local
20:23 kriskropd joined #git
20:23 qqx rodyhaddad: `git fetch --prune <remote>`
20:24 qqx Can be shortened to -p
20:24 stux|RC-only joined #git
20:24 rodyhaddad qqx: hm, that deletes remote branches that are on my machine
20:24 ochorocho__ joined #git
20:24 rodyhaddad qqx: I'm looking to delete the branches remotely, usually done by pushing an empty reference to the remote branch
20:25 [twisti] left #git
20:25 qqx Ahh right, `git push --prune <remote>`
20:25 qqx That one can't be shortened.
20:26 rodyhaddad qqx: Ohh, awesome!
20:26 canton7 well, probably '--pr'? iirc, you can abbreviate any of the longer flags
20:26 mizu_no_oto joined #git
20:27 johnny56 joined #git
20:28 qqx canton7: Actually it would need to be --pru since there's --progress as well.
20:29 qqx But I generally try to avoid depending on those, since they can always be broken when new options are added.
20:30 canton7 ah, didn't spot --progress
20:30 rodyhaddad I'm okay with not abbreviating it :D, but thanks for the the last tip :)
20:30 padde joined #git
20:30 ixjlyons joined #git
20:31 qqx I only spotted it by trying, (along with -n)
20:32 rorro joined #git
20:32 xocket joined #git
20:34 mischat joined #git
20:36 synthroid joined #git
20:36 Fissure joined #git
20:39 salamanderrake joined #git
20:40 matsaman joined #git
20:41 goTAN joined #git
20:45 panzon joined #git
20:45 Jackneill joined #git
20:46 anth0ny joined #git
20:46 P4nther joined #git
20:48 processkiller joined #git
20:51 Jackneill_ joined #git
20:53 cagedwisdom joined #git
20:56 wrouesnel1 joined #git
20:57 anuxivm joined #git
21:00 nikivi joined #git
21:01 pothepanda joined #git
21:03 nwkj86 joined #git
21:04 Guest2 joined #git
21:05 bathtub_shark joined #git
21:06 BadSnappa joined #git
21:07 Fissure joined #git
21:09 satifant joined #git
21:09 Vinnie_win joined #git
21:11 Raed joined #git
21:14 nwkj86 joined #git
21:14 Atm0spher1c joined #git
21:15 zeroed joined #git
21:16 dreiss joined #git
21:19 texinwien joined #git
21:20 nikivi joined #git
21:21 cdown joined #git
21:26 kourier joined #git
21:32 xock_stream joined #git
21:35 shinnya joined #git
21:36 danslo joined #git
21:37 dsantiago joined #git
21:41 chardan joined #git
21:41 mischat joined #git
21:41 neilthereildeil left #git
21:44 adac joined #git
21:45 dsantiago joined #git
21:47 Fissure joined #git
21:47 matsaman joined #git
21:47 padde joined #git
21:49 Gsham joined #git
21:49 nikivi joined #git
21:49 nwkj86 joined #git
21:49 mischat joined #git
21:52 stux|RC-only joined #git
21:53 Fissure joined #git
21:54 Samual joined #git
21:54 Lyqyd joined #git
21:56 rodyhaddad joined #git
21:58 jimi_ joined #git
21:59 mindriot101 joined #git
22:01 Fissure joined #git
22:03 nikivi joined #git
22:06 mizu_no_oto joined #git
22:07 eivarv joined #git
22:08 dreiss joined #git
22:09 acetakwas joined #git
22:11 Fissure joined #git
22:11 stick-shift joined #git
22:11 fatalhalt joined #git
22:13 goTAN joined #git
22:13 Jackneilll joined #git
22:14 bathtub_shark joined #git
22:14 NeverDie joined #git
22:16 Snugglebash joined #git
22:18 harwood joined #git
22:18 nikivi joined #git
22:19 shgysk8zer0 joined #git
22:24 nettoweb1 joined #git
22:24 d5sx43 joined #git
22:25 pothepanda joined #git
22:26 kballou joined #git
22:27 panzon joined #git
22:27 BlaXpirit joined #git
22:32 xocket joined #git
22:34 dendazen joined #git
22:34 krillr joined #git
22:35 rwb joined #git
22:37 stux|RC-only joined #git
22:37 whlai joined #git
22:38 ELFrederich Curious how something like gitolite works.  It seems to just use hooks and stuff.  When you "git clone git@some-host:some-project" how does "Git" know to use "Gitolite"?
22:39 tobiasvl ELFrederich: what do you mean? gitolite runs a server on some-host which serves repos there
22:39 whlai can anyone explain how to calculate what the http.postBuffer should be set to?
22:39 ELFrederich tobiasvl: I'm pretty sure it doesn't run a server
22:39 matsaman that is: it sets up a git repo
22:40 whlai I have a huge push that I'm trying to complete but don't know how to calculate what the buffer should be
22:40 whlai should it match the size of what I'm trying to push?
22:40 matsaman I imagine the default is fine, unless you actually encounter an error
22:41 matsaman (this sentiment can be applied to most situations)
22:41 whlai i.e if I'm pushing 5G to my remote repo, does the buffer need to be 5G?
22:41 matsaman version controlling binaries, are we? =)
22:41 whlai maybe :D
22:42 matsaman I would still wait for a problem to actually manifest
22:42 whlai problem has manifested
22:42 whlai error: RPC failed; result=56, HTTP code = 0
22:43 whlai Also, I'm pushing 84G, not 5. :D
22:44 matsaman whlai: okay, you can set postBuffer to the max bytes you plan to send
22:44 matsaman from the client-side
22:44 shgysk8zer0 joined #git
22:44 whlai right. so it should match.
22:45 rahtgaz joined #git
22:45 whlai The buffer required = total dat being pushed
22:45 matsaman >=
22:45 whlai right
22:45 matsaman that's my understanding
22:45 whlai Thank you sir!
22:45 mischat joined #git
22:45 whlai wait ....
22:45 matsaman strange git doesn't tell you this
22:45 matsaman it's usually so fsckin' noisy about errors, with stupid suggestions =P
22:45 whlai Where does the buffer get stored? any idea?
22:46 matsaman in git's ordinary config
22:46 matsaman so within .git for your repo
22:46 matsaman git config -l
22:46 matsaman etc.
22:46 whlai No, I mean the actual buffered data
22:46 matsaman oh
22:46 matsaman heh
22:46 whlai Where does it sit before transport?
22:46 Charliechin joined #git
22:46 whlai Ram?
22:46 whlai pagefile?
22:46 sea-gull_ joined #git
22:47 julius_ joined #git
22:47 julius_ hi
22:47 whlai Swap?
22:47 ELFrederich Curious how something like gitolite works.  It seems to just use hooks and stuff.  When you "git clone git@some-host:some-project" how does "Git" know to use "Gitolite"?
22:48 cdown joined #git
22:48 matsaman whlai: what git version is this?
22:48 matsaman hi julius_
22:48 whlai version 1.7.9.5
22:48 Rodya_ joined #git
22:49 matsaman ELFrederich: https://git-scm.com/book/en/v​1/Git-on-the-Server-Gitolite
22:49 matsaman whlai: oh okay, I've read that just using a more recent git version might avoid this issue altogether
22:49 matsaman https://blogs.msdn.microsoft.com/congyiw/20​16/06/02/git-stop-changing-http-postbuffer/
22:49 whlai ty sir!
22:49 matsaman whlai: as for the buffer I'm not personal sure
22:49 matsaman I would image it just start with ram
22:50 matsaman like any ordinary app <shrug>
22:50 elsevero joined #git
22:50 matsaman I'm not personally*
22:50 whlai I feel like swap would be more appropriate if it's available
22:50 matsaman man, really dropping bits of words today
22:50 quesker left #git
22:50 matsaman mmm, well, the purpose of swap, really, is to exist if ram is exhausted
22:50 matsaman although in these days of huge amounts of RAM, it's more and more nothing but a hibernation space
22:51 whlai well, with the amount I'm pushing, I don't think I have enough space with swap and ram combined
22:51 whlai think I'm going to had to git add in chunks
22:51 m0viefreak you really dont want pushbuffer that high...
22:52 whlai yeah
22:52 m0viefreak upgrading to a git version thats not 4 years old might be a good start
22:52 jeffreylevesque joined #git
22:52 whlai yup. on it
22:52 dsdeiz joined #git
22:53 matsaman whlai: honestly I'm assuming that will be managed on its own
22:54 whlai gonna give it a shot
22:54 matsaman where the buffer is / how it works is a separate (likely non-)issue from the bug you're probably getting because of using a really old version
22:55 nikivi joined #git
22:58 whlai update seems to be working
22:58 whlai It counted objects - didn't get that far before
22:59 preaction i'm trying to use "git am" in a script, but if I set GIT_DIR and GIT_WORK_TREE (if i don't actually "cd $GIT_WORK_TREE"), git am can't find my patches. it seems to cd itself and then look for the arguments. is this expected?
23:01 bathtub_shark joined #git
23:02 Fissure joined #git
23:06 nettoweb joined #git
23:07 goTAN joined #git
23:08 Snugglebash joined #git
23:10 ochorocho__ joined #git
23:10 joki- joined #git
23:11 snthsnth joined #git
23:11 nikivi joined #git
23:18 d5sx43 joined #git
23:19 crazycoder joined #git
23:19 crazycoder hello
23:19 crazycoder how can i only show the git merge .... here:           git log --pretty=format:"%h; %s" v2.0.1..v2.0.2
23:19 crazycoder in this way i see all the commits between 2.0.1 and 2.0.2
23:20 crazycoder but i always do     git merge --no-ff dev    (into master)
23:20 hhee joined #git
23:20 crazycoder so i would like to show that merge only
23:23 acetakwas joined #git
23:26 julius_ joined #git
23:31 raijin joined #git
23:33 xock_stream joined #git
23:39 Xethron joined #git
23:46 overlord_tm joined #git
23:47 d5sx43 joined #git
23:47 overlord_tm joined #git
23:48 multi_io joined #git
23:48 bathtub_shark joined #git
23:49 multi_io why does "git fetch remote rbranch:lbranch" NOT also set lbranch up to track remote/rbranch ?
23:49 multi_io because it seems it doesn't, according to git branch -avvv
23:50 jimi_ joined #git
23:51 multi_io it fetches remote/rbranch and creates lbranch and points it to the same commit as remote/rbranch, but the former isn't set up to track the latter
23:54 inflames joined #git
23:54 syg_ joined #git
23:54 multi_io or can I set this up afterwards?
23:55 osse multi_io: git branch -u lbranch rbranch
23:55 justanotheruser joined #git
23:57 EricPoe joined #git
23:59 Darren_ joined #git

| Channels | #git index | Today | | Search | Google Search | Plain-Text | summary