Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-08-01

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

All times shown according to UTC.

Time Nick Message
00:02 mellernoia joined #git
00:02 MattMaker joined #git
00:06 fuchstronaut joined #git
00:07 MattMake_ joined #git
00:08 Lyqyd joined #git
00:11 tesuji joined #git
00:12 MattMaker joined #git
00:13 shgysk8zer0 joined #git
00:16 zumba_ad_ joined #git
00:16 spudowiar joined #git
00:17 MattMaker joined #git
00:20 cdg joined #git
00:20 zumba_ad_ joined #git
00:21 afuentes joined #git
00:22 dsdeiz joined #git
00:22 MattMaker joined #git
00:26 zumba_ad_ joined #git
00:26 afuentes joined #git
00:27 BSaboia joined #git
00:28 RedneckDonut joined #git
00:29 tristanp joined #git
00:30 Samual joined #git
00:31 MattMake_ joined #git
00:32 Chunk2 joined #git
00:33 goTAN joined #git
00:34 zumba_ad_ joined #git
00:37 rubie joined #git
00:41 rbern joined #git
00:41 RustyShackleford joined #git
00:42 ktosiek joined #git
00:45 BenderRodriguez joined #git
00:46 MattMaker joined #git
00:46 mingrammer joined #git
00:46 BrianBlaze420 joined #git
00:47 Rodya_ joined #git
00:49 zumba_ad_ joined #git
00:49 hahuang61 joined #git
00:59 ckubrak joined #git
00:59 dreiss joined #git
01:00 aidalgol How do I push a branch other than the one I have checked out?
01:00 aidalgol If I have fixup02 checkedout, will `git push origin master` do it?
01:01 preaction that will push your current branch, fixup02, to the remote "master". is that what you want?
01:01 aidalgol Nope.
01:01 Rodya_ joined #git
01:01 preaction otherwise, the complete command is: git push <remote> <local_ref>:<remote_ref>
01:02 aidalgol I want to push master to master without having to bother with stashing my changes and checking out master.
01:02 preaction so, to push the local fixup02 to the remote fixup02, it would be: git push origin fixup02:fixup02
01:02 aidalgol And what's the difference between origin/fixup02 and remotes/origin/fixup02 ?
01:02 bremner aidalgol: git push remote localbranch:remotebranch. err. what preaction said
01:03 cqi joined #git
01:03 preaction aidalgol: i don't know where you got origin/fixup02, but remotes/origin/heads/fixup02 is the local copy of what you think the remote branch looks like
01:04 preaction there's lots of ways to specify refs/commits, omitting certain redundancies (like how i added "heads", but you didn't have that)
01:04 preaction !refspec
01:04 gitinfo [!refspecs] Refspecs are used by fetch/push to *spec*ify which *ref*s to transmit where. They have the form "source:destination". They can be prefixed with a "+" to force the update, possibly displacing existing history. More info: http://jk.gs/git-fetch.html or http://i.qkme.me/3tke7r.jpg
01:05 aidalgol I've seen some people merge from origin/BRANCH_NAME instead of just BRANCH_NAME, and I say `gitk master origin/master` to explicitly show the head of both local and remote.
01:06 preaction right, because they want to ensure they're merging from the remote's BRANCH_NAME, not whatever their local BRANCH_NAME happens to be
01:06 preaction which of course only works if they've fetched the remote recently
01:07 ulkesh joined #git
01:08 ulkesh joined #git
01:08 danslo joined #git
01:09 aidalgol yeah, it's always been right after a fetch.
01:10 aidalgol so why does `git branch -a` list "origin/blah" prefixed with "remotes/" ?
01:10 plasticle joined #git
01:11 catag87 joined #git
01:12 sn131678 joined #git
01:15 afusion joined #git
01:17 preaction aidalgol: because you have, locally, a copy of the ref that the remote has. you keep this from the last time you fetched, so your local repo knows when things change (this is how "your branch is X commits [ahead|behind] remote" works)
01:18 ekinmur joined #git
01:19 djapo joined #git
01:19 djapo how can i know which remote is my current branch attached to
01:21 anuxivm left #git
01:22 preaction djapo: git config branch.<name>.remote
01:22 ahmedelgabri joined #git
01:23 ImJune joined #git
01:26 ejb joined #git
01:29 Atrumx joined #git
01:29 dsdeiz joined #git
01:29 amen12 joined #git
01:31 cyphase joined #git
01:32 dopesong joined #git
01:33 crayon joined #git
01:35 crayon joined #git
01:39 mycyber joined #git
01:40 harish_ joined #git
01:41 Lunatrius` joined #git
01:42 pks joined #git
01:46 dreiss joined #git
01:49 ilbot3 joined #git
01:49 Topic for #git is now Welcome to #git, the place for git help and nursery rhymes | Public logs at http://goo.gl/BuUi5o | Current stable version: 2.9.2 | First visit? Read: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | Ring around the repo, pockets full of plumbing, objects, objects, we all commit.
01:50 hahuang61 joined #git
01:51 shaggycat__ joined #git
01:54 danimal2 joined #git
01:58 d^sh joined #git
02:02 israelzuniga joined #git
02:03 liuzhen joined #git
02:06 fuchstronaut joined #git
02:09 liuzhen joined #git
02:11 catag87 joined #git
02:17 jaguarma_ joined #git
02:19 cqi joined #git
02:21 a_thakur joined #git
02:22 goTAN joined #git
02:24 syndikate joined #git
02:24 ahmedelgabri joined #git
02:26 gtristan joined #git
02:26 dreiss joined #git
02:27 interne7_ joined #git
02:28 jaguarmagenta joined #git
02:29 _sjs joined #git
02:32 Derperperd joined #git
02:34 Lunatrius` joined #git
02:34 mingrammer joined #git
02:36 Goplat joined #git
02:39 a_thakur_ joined #git
02:39 a_thakur joined #git
02:40 Gsham joined #git
02:41 a_thakur joined #git
02:41 ljc joined #git
02:46 boombatower joined #git
02:47 danimal2 joined #git
02:50 a_thakur joined #git
02:51 chachasmooth joined #git
02:52 _leb joined #git
02:55 lagothrix joined #git
02:58 hexagoxel joined #git
02:59 danimal2 joined #git
03:01 ome joined #git
03:06 JanC joined #git
03:09 danslo joined #git
03:14 david007 joined #git
03:16 aavrug joined #git
03:22 rgrinberg joined #git
03:25 Bennux joined #git
03:27 trobotham joined #git
03:28 SwiftMatt joined #git
03:29 mingrammer joined #git
03:29 MZAWeb_ joined #git
03:31 nnyk joined #git
03:32 mingrammer joined #git
03:32 kadoban joined #git
03:33 dopesong joined #git
03:36 rubie joined #git
03:36 fstd_ joined #git
03:43 texinwien_ joined #git
03:44 fahadash joined #git
03:50 catag87 joined #git
03:51 hahuang61 joined #git
03:57 wwalker which manpage talks about the different ways to show a commit  (sha, HEAD, thing~, thin^^, etc.)?
03:57 ojacobson man gitrevisions
03:57 gitinfo the gitrevisions manpage is available at http://jk.gs/gitrevisions.html
03:58 wwalker ojacobson: thanks!
04:00 moneylotion joined #git
04:00 thecomedian joined #git
04:02 rscata joined #git
04:03 snowkidind joined #git
04:03 kulelu88 joined #git
04:04 titanium17 joined #git
04:04 jaguarmagenta joined #git
04:07 fuchstronaut joined #git
04:11 goTAN joined #git
04:16 arescorpio joined #git
04:16 zumba_ad_ joined #git
04:21 Literphor joined #git
04:22 mingramm_ joined #git
04:25 ahmedelgabri joined #git
04:28 rubie joined #git
04:28 Limes_ joined #git
04:29 danimal2 joined #git
04:31 rubie_ joined #git
04:33 jaguarmagenta joined #git
04:35 titanium17 joined #git
04:38 cebor joined #git
04:38 zumba_ad_ joined #git
04:42 Cabanossi joined #git
04:45 p1k joined #git
04:45 jaguarma_ joined #git
04:47 macefr joined #git
04:47 jaguarmagenta joined #git
04:47 MattMaker joined #git
04:49 RustyShackleford joined #git
04:50 ffabi joined #git
04:50 adrien_ joined #git
04:50 a_thakur joined #git
04:51 catag87 joined #git
04:51 johnmilton joined #git
04:52 rgrinberg joined #git
04:54 adrien__ joined #git
04:56 jottr joined #git
05:00 govg joined #git
05:01 adrien__ joined #git
05:02 SwiftMatt joined #git
05:03 diogenese joined #git
05:03 ImJune joined #git
05:04 lordjancso joined #git
05:05 jaguarmagenta joined #git
05:06 jetpack_joe joined #git
05:07 afusion_ joined #git
05:08 ffabi joined #git
05:10 subhojit777 joined #git
05:12 moneylotion left #git
05:13 jaguarma_ joined #git
05:13 pinewoman joined #git
05:13 a_thakur joined #git
05:19 howdoi joined #git
05:19 jaguarmagenta joined #git
05:19 ffabi joined #git
05:22 pinewoman joined #git
05:27 jaguarmagenta joined #git
05:27 stuh84 joined #git
05:29 BrianBlaze420 joined #git
05:30 ljc joined #git
05:32 diogenese joined #git
05:33 Kulrak joined #git
05:33 Kulrak joined #git
05:33 byteflame joined #git
05:34 ffabi joined #git
05:34 jaguarmagenta joined #git
05:35 dopesong joined #git
05:36 ljc joined #git
05:36 zeroed joined #git
05:37 Macaveli joined #git
05:38 afusion joined #git
05:41 moneylotion joined #git
05:42 emPi joined #git
05:44 MattMaker joined #git
05:46 ahmedelgabri joined #git
05:50 ffabi joined #git
05:51 nilg joined #git
05:51 hahuang61 joined #git
05:56 freimatz joined #git
05:58 ahmedelgabri joined #git
05:59 aavrug joined #git
06:00 goTAN joined #git
06:00 aavrug joined #git
06:00 Raging_Hog joined #git
06:00 satya4ever joined #git
06:01 aavrug joined #git
06:03 ytl joined #git
06:04 dopesong joined #git
06:05 mariuscc joined #git
06:06 skraito joined #git
06:06 skraito WELCOME TO ##FURUDORA AND ##WINDOW , OFFICIAL CHANNEL FOR MICROSOFT AND LINUX CHANNEL ... .
06:06 skraito IF YOU'R GOOD WE WILL HIRE YOU FOR $300 BUCK PER WEEK FROM SUPPORT FROM IRC ... .
06:07 skraito left #git
06:08 gtristan joined #git
06:08 fuchstronaut joined #git
06:11 goTAN joined #git
06:13 Guest23219 joined #git
06:14 lucasem joined #git
06:15 Oebele joined #git
06:16 tortib_ joined #git
06:19 jceb joined #git
06:20 makinen " Merge branch 'develop' into feature/syl"
06:20 makinen how can this happen
06:21 GavinMagnus joined #git
06:21 makinen I checked out a wrong branch first
06:21 lucasem makinen:  what are you asking?
06:22 makinen but when I checked out feature/syl it opened up an editor and tried to merge the wrong branch into
06:22 dreiss joined #git
06:22 makinen it
06:22 star_prone joined #git
06:22 Khisanth joined #git
06:23 lucasem tell us: what you did, exactly; what happened; what you expected to happen
06:23 afusion joined #git
06:23 lucasem makinen: ^
06:24 rik316 joined #git
06:26 makinen $ git checkout?feature/syl $ git pull
06:26 ambro718 joined #git
06:27 Arkreuz joined #git
06:27 makinen Switched to branch 'feature/syl'
06:27 makinen Your branch and 'develop' have diverged,
06:27 makinen and have 10 and 3 different commits each, respectively.
06:28 makinen this is what git said after git checkout feature/syl
06:29 solrize joined #git
06:29 zeroed joined #git
06:29 ojacobson You asked git to merge the upstream of the current branch (feature/syl) into it.
06:30 solrize http://git.savannah.gnu.org/cgit/lightning.git  <-- anyone know what web front end this is?  is it gitweb?
06:30 ojacobson That's what git pull does, in tandem with fetching it from the upstream repository.
06:30 ojacobson solrize: it's cgit. Right in the URL. :)
06:30 solrize oh.  hehe :)
06:30 rubie joined #git
06:30 solrize i didn't know about it
06:30 ojacobson https://git.zx2c4.com/cgit/ eg
06:30 lucasem git pull can be dangerous
06:30 solrize lol, i typed cgit into the search bar and got     Canadian Girls in Training, or CGIT, is a church-based program for girls and young women aged 11–17 throughout Canada.
06:31 solrize https://git.zx2c4.com/cgit/   is that the main repo for cgit?
06:32 bbb joined #git
06:32 solrize yow it's written in C, i guess i should have expected that ;)
06:33 makinen ojacobson: has someone else made modifications to the upstream?
06:33 ojacobson possibly
06:34 solrize anyway thanks, i think i'll try to use this
06:34 makinen but why git said something aboyut the branch develop?
06:34 ojacobson 'develop' is the upstream branch of 'feature/syl'
06:34 spudowiar joined #git
06:35 ojacobson presumably because you set it up that way, possibly without meaning to
06:35 makinen feature/syl was branched from develop originally
06:37 makinen could I have just ignored the message about diversion and continue working
06:37 makinen on feature/syl
06:39 makinen git basically was just nice and informed me about the diverged upstream branch
06:40 nikitasius joined #git
06:41 ojacobson yes
06:41 ojacobson git provides mechanism, not policy; what you do with the information it gives you is, by design, your call :)
06:42 Sceorem joined #git
06:43 makinen okay now I'm starting to get how git works
06:44 hendry joined #git
06:45 hendry what's the best command to search for a file that was deleted a while back?
06:46 selckin 'git log -S' will search content, or maybe like git log --stat | grep filename, or if you know the full path just log that ?
06:48 piling joined #git
06:49 danslo joined #git
06:49 JeroenT joined #git
06:51 hendry selckin: thanks! now i just need to figure out how to recover it http://s.natalian.org/2016-08​-01/1470034297_2548x1380.png
06:52 * hendry wonders if I list logs where I am removed an entire file
06:52 selckin git checkout sha1 -- file
06:52 selckin git revert that commit
06:52 selckin the sha:filename is for git show
06:53 catag87 joined #git
06:54 arnas joined #git
06:54 arnas hello
06:54 selckin !hi
06:54 gitinfo [!welcome] Welcome to #git, a place full of helpful gits. If you have a question,  feel free to just go ahead and ask—somebody should answer shortly.  For more info on this channel, see http://jk.gs/git/  Take backups (type !backup to learn how) before taking advice.
06:54 hendry selckin: i was doing git checkout sha1 -- file.. do i need the commit before or something ?
06:54 [Brain] joined #git
06:55 selckin hendry: uhm yeah, try sha^
06:55 arnas okay, so something got bugged with my repo, because  it says that I have one modified file in a directory that doesn't even exist. I tried 'git checkout .', 'git clean -f', 'git reset --hard' and 'git stash -u' but none of those helped. Any ideas?
06:55 arnas I need to switch between branches and I can't because of this file
06:56 selckin its probably a file that is not tracked in the current branch and exists, and is tracked in the other branch
06:56 selckin and it doesn't want to overwrite it,s ince its not tracked
06:56 arnas selckin, it says it's "modified"
06:56 selckin show the actual output, of git status and when you try to change branches
06:56 arnas selckin, and if I try to "git checkout <branch name>" it says: error: Your local changes to the following files would be overwritten by checkout:"
06:57 selckin sounds like what i said
06:57 selckin arnas: git status says modified?
06:57 arnas selckin, http://hastebin.com/qecuyuyeji.vhdl
06:58 ImJune joined #git
06:58 selckin arnas: those changes should go away with 'git checkout -- .'
06:58 arnas selckin, tried like 100 times, didn't work :(
06:58 selckin arnas: what does diff show? could be crlf bullshit
06:59 arnas selckin, how do I check? just git diff?
06:59 selckin yeah
06:59 arnas it's just the whole file in red text
06:59 the_ktosiek joined #git
07:00 selckin and git diff -w ?
07:00 tvl joined #git
07:00 arnas selckin, that's git diff: http://hastebin.com/cofohizubi.xml
07:00 arnas git diff -w is the same
07:01 selckin then checkout should work, don't know
07:01 arnas but it doesn't
07:01 selckin 'git reset HEAD .; git checkout -- .'
07:01 star_prone joined #git
07:01 arnas selckin, didn't help
07:02 selckin no idea then, sorry
07:02 arnas :(
07:02 arnas wtf :/
07:02 jast checkout and reset don't help with CRLF issues
07:02 jast !crlf
07:02 gitinfo to fix problems with line endings on different platforms, check out http://line-endings.gitrecipes.de/
07:02 selckin but the diff doesn't try to readd the lines
07:02 selckin so its not htat?
07:02 cyan__ joined #git
07:03 selckin arnas: delete the file and checkout?
07:03 Eryn_1983_FL joined #git
07:03 jast oh, right, I didn't read it as "only red", kind of assumed a wall of red followed by a wall of green :)
07:03 selckin jast: yeah thats why i asked -w, and the diff he pasted didn't look like it added, unless i don't remember how the header line works
07:04 arnas selckin, seems like the only option
07:04 arnas thanks
07:04 jast is the affected file below your current dir (CWD)?
07:05 arnas jast what do you mean?
07:05 selckin ../filename
07:05 hendry selckin: thank you btw
07:05 hendry left #git
07:05 jast well, suppose your current dir in the shell is ~/myproject/bar. is the file in there, too, or is it in the parent dir or something like that?
07:06 jknetl joined #git
07:07 Macaveli joined #git
07:07 arnas it's in myproject/bar/something/that_file
07:07 arnas but I already deleted it and it solved my problem :)
07:07 titanium17 joined #git
07:07 Zamerick joined #git
07:08 testild joined #git
07:09 jknetl_ joined #git
07:10 ikmaak joined #git
07:11 Balliad joined #git
07:12 p4trix joined #git
07:15 bocaneri joined #git
07:17 crayon joined #git
07:18 crayon joined #git
07:19 JeroenT joined #git
07:20 PCatinean joined #git
07:20 Sonarpulse joined #git
07:21 hahuang61 joined #git
07:21 Sonarpulse joined #git
07:21 Sonarpulse is there anyway to get the tree hash for a directory without any object database?
07:22 Silenced_v2 joined #git
07:22 Silenced_v2 joined #git
07:22 Sonarpulse so just recursively traverse and keep hashes in memory?
07:22 selckin !xy
07:22 gitinfo Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...
07:23 Sonarpulse hah
07:23 Sonarpulse so Cargo, the package manager for Rust, currently hash source tarballs and calls it a day
07:24 Sonarpulse but tarballs aren't really content addressable (many different tarballs store essentially the same stuff)
07:24 Sonarpulse so want to implement git tree object as an alternative checksum strategy
07:24 shaggycat_ joined #git
07:24 lucasem Sonarpulse: just use a regular checksum
07:24 lucasem md5 my-dir
07:24 djb-irc joined #git
07:24 moriko joined #git
07:25 Sonarpulse md5 works on directories?
07:25 nikitasius Sonarpulse: nope `md5sum: beer: Is a directory`
07:26 Sonarpulse ok that's what I thought
07:26 nikitasius under debian
07:26 Sonarpulse https://github.com/NixOS/nix/issues/1006
07:26 Sonarpulse I also opened that
07:27 Sonarpulse basically I want to move the world in a more content addressable direction, and git tree objects have a enough inertia to make a for a good defactor standard :)
07:27 ImJune joined #git
07:27 lucasem well, you can use md5 on tar at least!
07:28 nikitasius lucasem: better sha1, sha1 works faster on large files
07:28 Sonarpulse lucasem: that is basically what is done today
07:28 lucasem nikitasius:  huh, TIL
07:29 nikitasius lucasem: `TIL` - ?)
07:29 nikitasius it mean...?
07:29 lucasem today I learned
07:29 nikitasius ahh ok!
07:30 selckin suspect you can use libgit2 to hack something together, but i don't really see the benefit being worth it over the extra complexity
07:31 Seveas also, git is slowly (glacially) trying to move away from sha1
07:31 MrWoohoo joined #git
07:32 Sonarpulse Seveas: anything to read about that?
07:32 Seveas the odd mailinglist thread demanding that git moves
07:32 Seveas and brian m. carlson's commits around struct object_id
07:32 selckin the hashing algo used isn't really relevant
07:33 Seveas it is. If your hashing algorithm is broken, content can be substituted without anybody noticing
07:33 selckin his goal isn't security, its reducing disk space
07:33 selckin else he needs to gpg sign or whatever
07:33 Seveas the tricky bit here is the definition of 'broken'. sha1 is showing cracks, but to break it in the context of git, that's still quite a way away
07:33 Sonarpulse my idea was to have seperate fields for hash algo and "strategy" e.g. flat tar hash, tree object etc
07:34 arnas joined #git
07:34 Sonarpulse that is reasonably forward compatable in a vacuum
07:34 Sonarpulse but may or may not jive with what git does
07:35 nikitasius lucasem: or at least (sha1) gives +- same perfomance. real 0m7.986s for 4.0Gb file sha1sum and real 0m6.509s for md5sum on xeon1270v3 . But hash is longer (+8).
07:35 eijk joined #git
07:35 jaguarmagenta joined #git
07:36 lucasem nikitasius: yeah, I mean sha1 is definitely less vulnerable than md5. It's neat they're that close, I kinda wonder how it scales asymptotically
07:36 selckin sha1 is 3x slower then md5
07:37 selckin 10billion/s to 3billion/s on modern gpu
07:37 nikitasius selckin: nope: watch my examples
07:37 nikitasius 0m7.986s Vs. 0m6.509s for 4Gb not 3 times
07:37 selckin nikitasius: 2nd time the file was cached in memory and all other things you didn't include in your crappy benchmark
07:38 nikitasius selckin: file was cached on md5sum. and i run after md5sum on file cached in memory, sure. AND sha1sum too.
07:38 nikitasius both results for files cached in memory.
07:38 selckin nikitasius: check hashcat benchmarks
07:39 nikitasius selckin: when you (debian) cached file in memory (32Gb), and you run md5sum or sha1sum few times, you can see real perfomance.
07:39 nikitasius no hdd bottleneck
07:40 nikitasius selckin: and it was on CPU, not on GPU.. there is no GPU on my server, especially modern. But its unrelated to #git, and im sorry for offtop.
07:41 lordjancso joined #git
07:45 pwnz0r joined #git
07:46 t0by joined #git
07:47 BlessJah nikitasius: actually you should hash /dev/zero to avoid bottlenecks
07:50 sagerdearia joined #git
07:52 rominronin joined #git
07:53 catag87 joined #git
07:57 texinwien_ joined #git
08:02 circ-user-AKHej joined #git
08:04 kedare joined #git
08:07 marcogmonteiro joined #git
08:09 fuchstronaut joined #git
08:13 tomog999 joined #git
08:13 a_thakur joined #git
08:13 ahmedelgabri joined #git
08:16 Literphor left #git
08:18 arnas joined #git
08:19 star_prone joined #git
08:20 tvw joined #git
08:21 shaggycat_ joined #git
08:21 zumba_ad_ joined #git
08:23 mischat joined #git
08:26 JeroenT joined #git
08:26 goTAN joined #git
08:28 dopesong joined #git
08:30 clmsy joined #git
08:36 DevAntoine joined #git
08:36 DevAntoine hi
08:37 cdown joined #git
08:39 maxorator joined #git
08:40 _ikke_ Hey
08:41 jknetl_ joined #git
08:43 Tachyon_ joined #git
08:43 govg joined #git
08:44 lpapp joined #git
08:45 lpapp hello, I am having a repository which is about 3.5 GB on the server!
08:45 lpapp I guess this is because some of my predecessors used to check out binary blobs, etc, which I cleant up.
08:45 lpapp but it is still harming the clones, etc?
08:45 lpapp what would you suggest me to do? To recreate the repo after the binary blob cleanup to have it smaller?
08:46 lb lpapp my condolences. do you know the BFG? https://rtyley.github.io/bfg-repo-cleaner/
08:47 maroloccio joined #git
08:47 cdown_ joined #git
08:47 lpapp lb: cheers, seems these condolences have gone a long way ... ;-)
08:47 lb and yes, your history will clutter all clones with the binary stuff
08:47 selckin lpapp: well if you want a smaller repo, removing them by rewriting history is the only way, could maybe keep a copy of the old one around for archival purposes tho
08:48 Zarthus "The BFG treats you like a reformed alcoholic: you've made some mistakes in the past, but now you've cleaned up your act.", well that's a good way to start off my morning :)
08:48 lb so if you want a smaller repo and smaller history you can clean/rewrite it with the BFG or git filter-branch tools
08:48 maroloccio hi. git config user.mail a@a.com in a repo. git config --list shows my global value first, my local value after. is that expected behaviour? of course local overrides global but.. why list it?
08:48 lb but this will "harm" all existing clones
08:48 maroloccio related but different question: can i not --list only effective values?
08:48 lb since it will rewrite the history of your project !rewrite
08:48 gitinfo Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum
08:48 lpapp BFG looks like a good thing, but it will not help with many many small blobs.
08:48 lpapp basically our Yocto tree was checked in with all the "non-vanilla" artefacts.
08:49 lpapp so there are so many small blobs that git could not even handle renaming, etc.
08:49 lpapp just saying that it has got a lot of small stuff.
08:49 lb lpapp if you dont care about history ( or maybe just archive it somewhere) you can also just restart the repo and just add the current checkout (manually)
08:49 lpapp lb: no, I need git log --follow and co
08:50 davisonio joined #git
08:50 lpapp but yes, it should be possible to restart the repo after the blob was removed.
08:50 lpapp and keep the history afterwards
08:50 lpapp it was done 2-3 years ago
08:50 Softeisbieger joined #git
08:50 lpapp BFG and GFB are my friends for that?
08:51 selckin !filter-branch
08:51 selckin !filter
08:51 gitinfo [!filters] Filters can be very useful for automatically making changes to files at checkout(smudge) and checkin/add(clean). See 'man gitattributes' for more information
08:51 selckin pff
08:51 lb selckin not exactly the right trigger ^^
08:52 selckin lpapp: you can remove them using git-filter-branch, for example i have this little script http://pastie.org/10925754
08:52 lb lpapp if you can start fresh, i'd suggest to do so. if you want to keep all history from the past, you'll have to use filter-branch to filter out stuff that you want to remove
08:52 lb but lpapp !backup
08:52 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.c​om/GitBestPractices/#backups
08:52 chll_ joined #git
08:52 lb filter-branch and friends can mess up your repo so be sure to have a backup
08:52 lpapp lb: yes, I think, in my case, I can warn my colleagues to do a brand new clone.
08:52 lpapp the reduced history would get them there soon enough, anyway
08:52 scoopr left #git
08:54 davisonio joined #git
08:56 maroloccio joined #git
08:57 lpapp I guess GFB will also break git bisect
08:57 chrysa joined #git
08:57 lpapp but if not, that is g reat.
08:57 lpapp great*
08:59 david007 joined #git
09:01 lpapp left #git
09:01 eijk joined #git
09:03 dopesong joined #git
09:03 dhollinger joined #git
09:05 luxflow joined #git
09:06 arnas joined #git
09:10 JeroenT joined #git
09:10 notebox joined #git
09:12 chrysa joined #git
09:13 cdown joined #git
09:14 Macaveli joined #git
09:15 p4trix joined #git
09:16 Arban joined #git
09:18 ifra201 joined #git
09:21 hahuang61 joined #git
09:23 afuentes joined #git
09:23 courrier joined #git
09:24 Anticom joined #git
09:30 polyzen joined #git
09:33 syndikate joined #git
09:35 wwwi joined #git
09:36 ahmedelgabri joined #git
09:36 clmsy joined #git
09:40 phanimahesh joined #git
09:42 fuchstronaut joined #git
09:43 mahendra_ joined #git
09:46 noncom joined #git
09:49 noncom when i want to switch to a branch, i do `git branch` to find the correct name for the branch and then i copy-paste it with mouse into the `git checkout ...` parameter. my question is: how can i avoid using the mouse? the names are pretty complex, including numbers and several words, i don't feel like typing them in... is there any other way to list the branches and then pick one of them for
09:49 noncom a checkout
09:49 noncom ?
09:50 ajf- joined #git
09:50 selckin git gui -> branch -> checkout
09:50 selckin git checkout $(git branch | grep foo)
09:51 selckin oh right not using the mouse, forget the fist one ;)
09:51 luxflow left #git
09:51 noncom hmm, git gui  "is not a git command".. but the second one seems a great idea!
09:51 themill tab completion can help
09:52 selckin noncom: its standard, debian doesn't install it by default tho
09:52 noncom i'm on mac...
09:53 goTAN joined #git
09:53 noncom themill: that's what i would think too, but it does not work for some reason... so i concluded there's no tab completion for this case.. i don't know..
09:53 selckin there is
09:53 selckin maybe not installed/enabled
09:54 esc joined #git
09:55 noncom hmm, yeah, it does not work in the terminal...
09:55 catag87 joined #git
09:55 noncom i'll try installing these tools, thanks :)
09:56 selckin source /usr/local/git/contrib/comp​letion/git-completion.bash #somewhere
09:57 selckin source  $(brew --prefix)/etc/bash_completion
09:57 selckin not a mac wizard
09:57 sdothum joined #git
10:12 quadruple joined #git
10:13 maroloccio joined #git
10:13 Guest1100101 joined #git
10:15 gugas joined #git
10:18 eijk_ joined #git
10:23 govg joined #git
10:29 tvw joined #git
10:30 quadruple hi, I have downloaded a repo.zip and extracted it somewhere. now I realized that a clone would have been better. Is it possible to copy .git into the newly created folder so it behaves like a clone?
10:32 rudi_s quadruple: Sure. Simplest solution is to just git clone .. foo and then move foo/.git into the directory.
10:32 rudi_s (Where .. is the clone URL.)
10:33 quadruple rudi_s, ty. that was also my assumption
10:35 cdg joined #git
10:39 mostlybadfly joined #git
10:40 astrofog joined #git
10:40 nuc joined #git
10:40 rudi_s np
10:42 JeroenT joined #git
10:46 ShekharReddy joined #git
10:46 texinwien joined #git
10:47 shaggycat_ joined #git
10:55 nnyk joined #git
10:56 lmatteis left #git
10:56 catag87 joined #git
10:58 cqi joined #git
11:01 sylario joined #git
11:01 JeroenT_ joined #git
11:02 adc-soft-dev-03 joined #git
11:04 Silenced_v2 joined #git
11:10 Lomex joined #git
11:11 Silenced_v2 joined #git
11:12 dopesong joined #git
11:13 Guest1100101 Hi, I have a github repository A, which was github forked to B; A has tags; how do I transfer A's tags to B ?
11:14 JeroenT joined #git
11:14 eijk_ joined #git
11:18 heftig Guest1100101: git clone --mirror A
11:18 interne7y joined #git
11:19 heftig Guest1100101: git push --mirror B
11:19 heftig Guest1100101: this will overwrite B's state with A's state, including all branches and tags
11:19 nettoweb joined #git
11:20 heftig Guest1100101: otherwise, if you already have a clone of A, you could git push --tags B to just push all tags to B
11:21 star_prone joined #git
11:22 fahadash joined #git
11:22 danslo joined #git
11:22 Guest1100101 heftig, I have full access to both A and B
11:22 star_prone joined #git
11:22 Linuxriley joined #git
11:24 _ikke_ Have you already cloned A?
11:25 Linuxriley left #git
11:27 Guest1100101 note: A is older and has tags , B forked from A and contains new development
11:29 _ikke_ You can just push the tags
11:29 Guest1100101 could I just add A as remote into B, and the  git push --tags A ???
11:30 xaa joined #git
11:34 _ikke_ git fetch A refs/tags/*:refs/tags/* && git push --tags B
11:34 goTAN joined #git
11:38 johnmilton joined #git
11:39 Peuc joined #git
11:51 cyphase joined #git
11:52 jaguarmagenta joined #git
11:55 Kartagis left #git
11:56 jokajak joined #git
11:56 fmcgeough joined #git
11:57 catag87 joined #git
11:58 SonikBoom joined #git
11:59 Guest1100101 _ikke_, this worked, thanks!
11:59 synthroid joined #git
12:00 _ikke_ yw
12:02 davisonio joined #git
12:03 goTAN joined #git
12:03 Lomex joined #git
12:04 Sceorem joined #git
12:04 davis left #git
12:09 diegoviola joined #git
12:10 jaguarma_ joined #git
12:10 Gamecubic joined #git
12:11 eijk_ joined #git
12:12 star_prone joined #git
12:14 rwb joined #git
12:15 JeroenT joined #git
12:16 Noldorin joined #git
12:17 phanimahesh joined #git
12:19 interne7y joined #git
12:22 rnsanchez joined #git
12:23 dsdeiz joined #git
12:23 dsdeiz joined #git
12:25 perlpilot joined #git
12:25 star_prone joined #git
12:26 zarubin_ joined #git
12:26 eijk joined #git
12:28 millerti joined #git
12:30 lss8 joined #git
12:30 TooLmaN joined #git
12:31 sn131678 joined #git
12:32 jimi_ joined #git
12:33 cdown joined #git
12:34 SourTrout joined #git
12:35 sanketdg joined #git
12:37 cdg joined #git
12:39 evie_hammond joined #git
12:41 syndikate joined #git
12:42 danslo joined #git
12:43 Shentino joined #git
12:44 ekinmur joined #git
12:46 jimi_ joined #git
12:53 zivester joined #git
12:55 josuebrunel joined #git
12:55 str joined #git
12:56 daniel_rose joined #git
12:57 jeffreylevesque joined #git
12:59 axisys joined #git
12:59 rgrinberg joined #git
13:00 axisys left #git
13:00 jetpack_joe joined #git
13:01 byteflame joined #git
13:02 Codfection joined #git
13:03 djb-irc joined #git
13:04 rahtgaz joined #git
13:08 pwnz0r joined #git
13:14 millerti joined #git
13:14 ExoUNX joined #git
13:15 axl_ joined #git
13:15 darkbit joined #git
13:16 star_prone joined #git
13:17 Sworden joined #git
13:17 Sworden2 joined #git
13:22 garethdaine joined #git
13:22 govg joined #git
13:23 hahuang61 joined #git
13:23 kinlo joined #git
13:24 axl_ joined #git
13:24 kinlo how does git commit deletes?  I was under the impression that git just references a tree in a commit, so if a client sees a new tree with a file missing, it knows the file has been deleted and will delete the file itself when doing a git pull?
13:25 byteflame joined #git
13:25 kinlo I've got a corrupt client that has pushed "something" to my master, causing a git pull to make files as "not known by git" instead of deleting them, is this normal behaviour?  is my git corrupt?
13:26 cqi joined #git
13:26 _ikke_ kinlo: Were these files even tracked? (might be that they were ignored, but later not anymore)
13:26 millerti joined #git
13:28 kinlo _ikke_: what happened is the following, a new git clone of an existing repo has been done, the fetch part succeeded, the checkout part didn't.  it was in a script and my fault, I forgot to check exit codes so the script continued.  it changed a file, did a git add + git push, and the commit output clearly saw a huge amount of files as "deleted".  Doing a git pull on a checkout of that repo suddenly marks those files not as deleted, as git would delete them, b
13:28 _ikke_ kinlo: You are right that when someone else 'untracks' files from the repository, they will be removed in other working trees
13:29 _ikke_ kinlo: Your message was truncated
13:29 JeroenT joined #git
13:29 kinlo eh, does freenode cut off lines?
13:30 _ikke_ Might be
13:30 _ikke_ Might also be a limit in the IRC protocol
13:30 kinlo shouldn't happen, got len.pl installed to protect me from that :/
13:30 vassagus joined #git
13:31 kinlo in any case, another git clone of the same repo, which had the files under version control, behaves strangely: I do a git pull and suddenly those files are not deleted but "unknown"
13:31 kinlo which makes me doubt how I understand git, as I understand a new tree is committed, not the delete, so this should be impossible
13:32 griffindy joined #git
13:33 Samual joined #git
13:33 ekinmur joined #git
13:33 kinlo _ikke_: you did get the part until "saw a huge amount of files as "deleted"." right ? :)
13:34 rnsanchez joined #git
13:34 djb-irc joined #git
13:34 _ikke_ yes
13:36 _ikke_ What part confuses you?
13:37 kinlo well, if you push a commit that has fewer files in it, how is it possible a git pull does not delete those files but instead just marks them not part of the repo?
13:37 kbs joined #git
13:40 CheckDavid joined #git
13:40 rnsanchez joined #git
13:42 ajf- joined #git
13:43 quenode joined #git
13:44 veilig2000 joined #git
13:44 jaguarmagenta joined #git
13:45 tabrez joined #git
13:46 nnyk joined #git
13:46 Anticom Hi all. I was reading into subtrees and found an old blog post talking about .gittrees file. Has this file been removed in later versions of git?
13:46 Anticom (refering to this blog post: http://ruleant.blogspot.de/2013/06/git-s​ubtree-module-with-gittrees-config.html )
13:47 f3r70rr35f joined #git
13:49 _ikke_ kinlo: That's not default behaviour
13:50 _ikke_ kinlo: What OS are you on?
13:50 mrkat joined #git
13:50 kinlo ubuntu 14.04, git from package
13:51 mrkat joined #git
13:52 kinlo I should add that the git pull might also been done during a low disk space condition - I'm trying to figure out what my collegues have screwed up so it might just be something they did wrong
13:54 garethdaine joined #git
13:56 spudowiar joined #git
13:57 vassagus joined #git
13:58 vassagus joined #git
13:59 howdoi joined #git
14:01 brandonw joined #git
14:01 GodGinrai joined #git
14:02 Macaveli joined #git
14:02 garethdaine joined #git
14:03 goTAN joined #git
14:07 grill joined #git
14:07 kjnilsson joined #git
14:07 janx joined #git
14:07 nikitasius joined #git
14:08 le_melomane joined #git
14:09 faheem joined #git
14:09 howdoi joined #git
14:09 parsnip joined #git
14:09 phaleth joined #git
14:10 cmosguy joined #git
14:10 troyt joined #git
14:11 tank9988 joined #git
14:11 jason237 joined #git
14:12 Ineentho[m] joined #git
14:12 M-meznak joined #git
14:12 flupke joined #git
14:13 GameGamer43 joined #git
14:13 rfv joined #git
14:13 DarkMukke joined #git
14:14 johtso joined #git
14:14 ToeSnacks joined #git
14:14 zeller joined #git
14:14 Anticom Anyone?
14:14 nezZario joined #git
14:15 XDS2010 joined #git
14:15 Awesomecase joined #git
14:16 zacts joined #git
14:16 addshore joined #git
14:17 martin_g joined #git
14:17 henriquev joined #git
14:17 tfnico joined #git
14:17 epic joined #git
14:17 sylario joined #git
14:17 beckyconning_ joined #git
14:17 noahsussman joined #git
14:17 theresajayne joined #git
14:17 samsonizzle joined #git
14:17 bedouin joined #git
14:17 yena joined #git
14:17 ljharb joined #git
14:17 imanc joined #git
14:17 {AS} joined #git
14:18 MLM joined #git
14:18 apipkin_ joined #git
14:18 lebster joined #git
14:18 Finnoloid joined #git
14:18 trig-ger joined #git
14:18 JackMc joined #git
14:18 jleclanche joined #git
14:18 elmcrest joined #git
14:18 Lewix joined #git
14:18 shennyg joined #git
14:18 zero7 joined #git
14:18 DuncanT joined #git
14:18 boxrick1 joined #git
14:18 imdsm_ joined #git
14:18 cspotcode joined #git
14:18 mehdi_ joined #git
14:18 etcinit joined #git
14:18 vcoinminer joined #git
14:18 sfan786 joined #git
14:18 timwis joined #git
14:18 wizonesolutions joined #git
14:18 teadrop_ joined #git
14:18 harleypig joined #git
14:18 elijah joined #git
14:18 CEnnis91 joined #git
14:18 Lowell joined #git
14:18 stealthii joined #git
14:19 ulkesh joined #git
14:19 braderhart joined #git
14:19 hecliunyx joined #git
14:19 huhlig joined #git
14:19 pizzaops joined #git
14:19 kwmiebach joined #git
14:19 angular_mike joined #git
14:19 nickstenn joined #git
14:19 ggherdov joined #git
14:19 barq joined #git
14:19 fahadash joined #git
14:19 angelsl joined #git
14:19 [diecast] joined #git
14:19 CheckDavid joined #git
14:19 interne7y joined #git
14:20 scalability-junk joined #git
14:20 cloudcell__ joined #git
14:20 kutenai joined #git
14:20 mmlb joined #git
14:21 regfop joined #git
14:21 rom1504 Anticom: no, you are alone on the internet, sorry
14:22 cmosguy joined #git
14:22 cstrahan joined #git
14:22 Anticom rom1504: sudo rm -rf / then...
14:23 Anticom :)
14:23 janx joined #git
14:23 smakarov_ joined #git
14:25 timvisher joined #git
14:26 timvisher is there a way to get a diff that excludes certain named files?
14:26 ggherdov joined #git
14:26 timvisher for instance, in a tree of varying depth there are files 'foo', i'd like to get a diff but with no mention of 'foo' anywhere.
14:27 JeroenT joined #git
14:28 texinwien joined #git
14:28 regfop just throwing this question out -- if I've merged a develop branch into master and my master is now a few commits ahead of develop, in some cases is it recommended to merge master back into develop?
14:28 tank9988 left #git
14:29 smakarov__ joined #git
14:29 regfop we use develop for our release patch
14:29 ShekharReddy joined #git
14:30 Cooler joined #git
14:30 nnyk joined #git
14:30 Cooler hey can someone help me understand the basics?
14:30 cdown joined #git
14:31 rom1504 Cooler: Anticom can help you
14:31 Cooler especially the undo changes commands
14:33 vktec Cooler: Pro Git https://git-scm.com/book/en/v2
14:34 BrianBlaze420 joined #git
14:34 gtristan joined #git
14:36 rubie joined #git
14:36 garethdaine joined #git
14:36 le_melomane joined #git
14:36 catag87 joined #git
14:36 Anticom Cooler: https://www.atlassian.com/git/tutorials/ and http://rypress.com/tutorials/git/index
14:37 eijk_ joined #git
14:38 timvisher specifying every other file in the repo doesn't seem to work either? `find . -name foo -prune -o -print0 | xargs -0 git diff --stat origin/master` still lists `foo`
14:39 discopatrick joined #git
14:40 invex joined #git
14:40 batrick joined #git
14:42 catag87 joined #git
14:42 jkstrick joined #git
14:42 timvisher with globstar I can get some of the files to be excluded but not all: `git diff --stat origin/master -- **/!(foo)` ignores `foo` at the root of the repo but not `foo` at the leaves, which is really strange
14:43 timvisher also I wouldn't expect that to be any different than the `find | xargs` command
14:44 selckin then your find is incorrect
14:44 jaguarmagenta joined #git
14:44 star_prone joined #git
14:45 timvisher selckin: i can pipe both of them to a pager and i see what i expect
14:45 catag87 joined #git
14:47 kj_ joined #git
14:48 kj_ left #git
14:48 anohigisavay joined #git
14:48 anohigisavay hi
14:48 anohigisavay i have some trouble in understanding how interactive rebase works
14:49 jknetl joined #git
14:49 anohigisavay afaik git commits contain not diff but fresh new file contents
14:50 anohigisavay how does dropping commits have any effects then?
14:50 moritz anohigisavay: rebase doesn't use the actual commit, but the diff from the commit and its parent
14:51 moritz anohigisavay: there are two competing mental models of commits. When talking about rebase, using the "commits represent code changes + meta data" is easier
14:51 anohigisavay moritz: i see
14:51 moritz whereas when talking about branching, the staging area etc, the "commits contain a complete snapshot of the repository contents" is easier
14:52 zivester joined #git
14:52 anohigisavay moritz: thanks :)
14:53 danslo joined #git
14:53 weckl joined #git
14:54 anohigisavay oh another question. how do patches actually work? if i drop a commit the subsequent commits can still apply  without conflicts?
14:54 dhollinger joined #git
14:54 anohigisavay say the have changes on the same file?
14:54 moritz depends on the actual patches
14:54 grawity sometimes
14:54 Darcidride joined #git
14:54 moritz if they touch the same part of the file, usually not
14:54 grawity depends on how much the subsequent commits actually depended on the dropped commit
14:55 moritz if they touch different parts of the file, they tend to work without conflict
14:55 qpdb joined #git
14:55 grawity if diff 1 is +foo, diff 2 is -foo +bar, and diff 3 is -bar +baz, obviously removing #2 would break it
14:56 anohigisavay woa magic
14:57 rominron_ joined #git
14:57 anohigisavay i'm doing some experiments with rebase. i have two branches and they merge without conflict
14:57 grawity darcs people would know more about things like patch theory and stuff, but this is the general basics
14:57 cyphase joined #git
14:57 grawity I mean, take a look at the textual diffs, and consider whether you could follow the instructions in there
14:58 anohigisavay i did `rebase -ip` and removed the merge commit, and some successive commit fails with conflicts
14:59 star_prone joined #git
14:59 crayon joined #git
15:00 anohigisavay i mean if a merge happens without conflict i suppose it's safe to remove it and put all commits in a single line?
15:00 d^sh joined #git
15:00 Xandaros `git diff` highlights whitespace at the end of file and warns when there is no newline at the end of a file. Is it possible to extend this so it highlights “todo” and “debug” (case insensitive), as well?
15:01 anohigisavay but if i do `git rebase -i` it works as intended.
15:01 Xandaros The number of times I included some debug modules and then committed that later is quite embarrassing :(
15:01 grawity Xandaros: no, write a commit hook that rejects this
15:02 Xandaros Is it possible to easily circumvent such a hook should the need arise?
15:03 grawity probably
15:03 grawity --no-verify, it seems
15:03 fuchstronaut joined #git
15:03 Xandaros Yup, just found it myself whilst reading up on them. I shall do this, then, thank you :)
15:05 benjwadams joined #git
15:06 kutsuya joined #git
15:06 benjwadams there's something like `git add -p` for rebasing commits, right?
15:06 benjwadams .. i think it might be `git reset -p`?
15:06 moritz both exist
15:07 benjwadams I need to "interactively undo changes"
15:07 Xandaros grawity: Looks like the sample pre-commit rejects commits with whitespace errors, too. This is rather useful! :D
15:07 benjwadams existing in a commit
15:08 Cata|Work joined #git
15:10 benjwadams I could do `git reset --soft <filename>` followed by `git add -p <filename>`
15:11 subhojit777 joined #git
15:12 axisys joined #git
15:13 sangy joined #git
15:14 charlesr joined #git
15:15 Raging_Hog joined #git
15:15 axisys joined #git
15:16 cdg joined #git
15:16 pioiytr joined #git
15:16 pioiytr joined #git
15:18 shinnya joined #git
15:19 dopesong joined #git
15:20 Gitzilla joined #git
15:20 dmto joined #git
15:20 snowkidind joined #git
15:20 dopesong joined #git
15:22 ImJune joined #git
15:23 hahuang61 joined #git
15:24 ejb joined #git
15:25 neunon joined #git
15:25 eivarv joined #git
15:26 nydel07 joined #git
15:26 elementalest joined #git
15:26 elementalest joined #git
15:26 JCaesar joined #git
15:27 fullstack joined #git
15:27 Derperperd joined #git
15:27 anohigisavay i did further experiments. seems removing a merge commit is equivalent to undoing one of it's parent commit
15:28 anohigisavay this is weird
15:28 weltall joined #git
15:28 ahmedelgabri joined #git
15:29 thiago joined #git
15:29 GT4066 joined #git
15:30 smakarov_ joined #git
15:33 tyreld joined #git
15:33 elementalest joined #git
15:33 elementalest joined #git
15:34 zacts joined #git
15:34 dlaboratory joined #git
15:35 synthroid joined #git
15:35 catag87 joined #git
15:36 smakarov__ joined #git
15:38 madewokherd joined #git
15:43 veilig2000 joined #git
15:45 EduardoMartins joined #git
15:45 markelite joined #git
15:47 jaafar joined #git
15:48 dreiss joined #git
15:50 nydel07 joined #git
15:50 rik316 joined #git
15:50 rolha joined #git
15:50 jab416171 joined #git
15:51 Guest1100101 joined #git
15:51 Electrometro joined #git
15:51 SourTrou_ joined #git
15:52 kutsuya joined #git
15:53 M-meznak joined #git
15:56 Electrom_ joined #git
15:56 nydel07 joined #git
15:56 rik316 joined #git
15:58 rominronin joined #git
15:59 pioiytr joined #git
16:06 zarubin joined #git
16:06 davisonio joined #git
16:10 aep-shoutlet joined #git
16:11 veilig2000 joined #git
16:12 daniel_rose joined #git
16:12 dreiss joined #git
16:13 oich joined #git
16:18 MacWinner if update master on server-A,  and on server-B I'm currently in branch foo, but I want to switch directly to the latest master commit, what would be the way to do it?  on server-B, I did a git pull, then git checkout master, then I needed to do a git pull again to fast-forward
16:18 MacWinner i want to go straight to the fast forwarded point
16:19 anuxivm joined #git
16:19 thiago MacWinner: git fetch server-A && git push . server-A/master:master
16:20 thiago also: git fetch server-A && git rebase server-A/master master
16:21 crayon joined #git
16:21 GodGinrai whoa whoa whoa
16:21 GodGinrai why would he need to push?
16:21 ambro718 joined #git
16:22 GodGinrai MacWinner: where are you running your commands from?
16:22 MacWinner running them from Server-B
16:23 TomyLobo joined #git
16:23 josuebrunel joined #git
16:23 MacWinner my server-B master is behind the master that was pushed on server-A.   i want to switch to master on Server-B without having hte intermediate step of being on the old master branch on server-b
16:24 ajf- joined #git
16:24 MacWinner git checkout origin/master?
16:25 stoopkid joined #git
16:25 thiago GodGinrai: pushing to .
16:26 GodGinrai MacWinner: `git checkout master && git pull`  Technically you'd have the intermediate step, but it would be executed in one command, so would that satisfy your wish?
16:26 ajf- joined #git
16:26 ansimation joined #git
16:27 MacWinner GodGinrai, not really.. because my current branch is actually the same as master.. so I basically just want to switch labels without modifying files
16:27 Lomex joined #git
16:27 thiago then what I said
16:27 Dong__ joined #git
16:28 GodGinrai MacWinner: if your current branch is the same as master, then why does server A even matter?
16:28 _Patrick_ joined #git
16:28 MacWinner sorry, i'm totoally not explaining this clearly.
16:28 thiago server A's master is the same as what server B's foo
16:28 ajf- joined #git
16:29 MacWinner thiago, that's right
16:29 tablespork joined #git
16:29 thiago he wants to make B's master be the same as those two and not touch any files
16:29 durham joined #git
16:29 danslo joined #git
16:29 thiago so, try what I suggested
16:29 Dong__ Guys, got a single PACK file in ofs Delta, now I want to make a ref Delta pack out of this, how do I do it after unpack-objects step?
16:29 MacWinner thiago, which server do I run those on?
16:30 Tad3j joined #git
16:31 ImJune joined #git
16:31 ajf- joined #git
16:31 polyzen `git pull --rebase` on B might be an option
16:32 MacWinner thiago, on server-B, I did "git pull && git rebase origin/master master  && git checkout master"
16:32 polyzen no need to checkout the branch you're on
16:32 perlpilot joined #git
16:32 emPi joined #git
16:32 polyzen ah
16:32 polyzen foo branch
16:32 Sonarpulse left #git
16:33 tablespo_ joined #git
16:34 ytl joined #git
16:37 thiago MacWinner: the second argument to git rebase indicates the branch to check out
16:37 Tachyon_ joined #git
16:37 thiago MacWinner: so you were already in master before that checkout
16:38 tablespork joined #git
16:38 crayon joined #git
16:38 MacWinner thiago, yeah, I just realized that, thanks!   so if I'm on branch foo, and I do a 'git rebase origin/master',  how does it know to rebase the master branch?
16:38 ExoUNX joined #git
16:38 ExoUNX joined #git
16:39 osse it doesn't
16:41 BrianBlaze420 joined #git
16:42 davisonio joined #git
16:42 titanium17 joined #git
16:44 davisonio joined #git
16:44 phanimahesh joined #git
16:44 anuxivm left #git
16:44 EvilPeng1 joined #git
16:45 airdisa joined #git
16:46 dwmw2_gone joined #git
16:46 axl_ joined #git
16:51 xaa joined #git
16:51 saidelike joined #git
16:51 daynaskully joined #git
16:53 tablespork joined #git
16:55 EvilPeng1 joined #git
16:56 GodEater joined #git
16:56 GodEater joined #git
16:56 star_prone joined #git
16:57 M-shine joined #git
16:58 OAC joined #git
16:58 dreiss joined #git
16:59 rgrinberg joined #git
16:59 Bert1567 joined #git
16:59 boushi joined #git
17:00 xall joined #git
17:00 rgrinberg joined #git
17:02 elementalest joined #git
17:02 elementalest joined #git
17:07 str joined #git
17:09 _W_ joined #git
17:10 jason237 joined #git
17:13 ImJune joined #git
17:14 vuoto joined #git
17:15 SteffanW joined #git
17:15 RustyShackleford joined #git
17:16 davisonio joined #git
17:16 _Patrick_ joined #git
17:17 star_prone joined #git
17:19 jeffreylevesque joined #git
17:20 thiago joined #git
17:20 crayon joined #git
17:21 pwnz0r joined #git
17:23 Ineentho[m] joined #git
17:23 M-mistake joined #git
17:23 M-meznak joined #git
17:24 xenu joined #git
17:24 xenu hello, when i'm using "git pull --rebase=preserve" it always prints "Successfully rebased and updated refs/heads/master." even when there are no changes
17:24 xenu is there any way to change it?
17:24 hahuang61 joined #git
17:25 xenu normal "git pull --rebase" would print "Your branch is up-to-date..." when there are no changes
17:25 rolha joined #git
17:26 xall joined #git
17:27 nitric joined #git
17:27 eivarv joined #git
17:28 GavinMagnus left #git
17:28 fuchstronaut joined #git
17:28 hosttor joined #git
17:29 Darcidride joined #git
17:30 maroloccio joined #git
17:32 daniel_rose joined #git
17:32 zlenpy joined #git
17:33 invex joined #git
17:34 lpapp joined #git
17:34 Darren_ joined #git
17:34 zlenpy git for windows installer isnt offering me option to specify installation location - is this possible?
17:34 lpapp hi, I would like to get all the commits in all branches from a user the last three days? Is it easily possible?
17:35 Darcidride_ joined #git
17:35 fees joined #git
17:35 dopesong joined #git
17:35 osse lpapp: git log --since='3 days ago' --author=a_user --all
17:35 lpapp basically my colleague pushed a change on Friday to some branch, but I cannot find it :D
17:36 lpapp osse: cheers
17:37 lpapp osse: it is strange that it goes back to the 17th of July
17:37 wwwi joined #git
17:37 lpapp logs are between 17th and 28th of July
17:37 lpapp ah, maybe I need to replace 3 days ago with the actual date format
17:37 lpapp whatever that is
17:38 leeN joined #git
17:38 osse lpapp: the string '3 days ago' actually works in my case.
17:38 lpapp ah, ok, not for me, for some reason
17:39 osse have you fetched? maybe you don't have it locally
17:39 lpapp not sure what I am doing wrong.
17:39 lpapp I did git pull --rebase
17:39 osse try '100 days ago' just to make sure it works at all.
17:39 lpapp but even if I do not have the commits, I expect the commits from the 28th of July or none
17:39 lpapp rather than 17-28
17:39 osse maybe it's a recent feature
17:39 lpapp git --version
17:39 lpapp git version 2.9.0
17:40 athri joined #git
17:40 osse hmm
17:40 lpapp I will upgrade it now to 2.9.2
17:40 osse je suis trés clueelss
17:40 osse no it's definitely older than that
17:40 nitric lpapp: have you git fetched?
17:40 lpapp only git pull --rebase in the current branch
17:40 nitric that includes a fetch, so you're good there...just double checking
17:41 osse do a plain 'git fetch'
17:41 osse oh
17:41 osse doesn't pull only fetch exactly what's needed?
17:41 lpapp and even if I was not doing it, like I wrote above, I still would not expect commits between 17-28
17:41 osse maybe it uses committer date instead of author date or something
17:41 osse try --pretty=fuller to see the commit date also
17:42 lpapp isn't git log showing the commit date?
17:42 osse no, author date
17:42 lpapp I did not know that there is an author date
17:42 davisonio joined #git
17:42 lpapp which is different from the commit date
17:42 osse I think it's the other way around; you didn't know there was a commit date :P
17:42 lpapp I knew about "commit date" and "push date"
17:43 rwb joined #git
17:43 osse anyway, the author date is supposed to be constant. it never changes after the commit has been created. but the committer date can be updated by rebase and if you send in a patch by email etc.
17:43 lpapp hmm, yeah, it seems these are commit dates
17:43 osse for the same reason there is a difference between author and committer
17:44 lpapp how about git commit --amend?
17:44 osse if I make patch, send it you and you apply it then i'm the authort and you're the committer
17:44 lpapp although that is probably a rebase.
17:44 osse i'm not sure.
17:45 lpapp osse: right, I believe these are recent cherry-picks
17:45 djb-irc joined #git
17:45 osse i would expect --amend to also change the author date since it actually modifies the changes the commit introduces but I don't know
17:45 lpapp I guess cherry-pick modifies the committer date, but not the author?
17:45 osse yes I think so
17:45 osse do they have different names too?
17:46 lpapp names of?
17:46 osse author/committer
17:46 lpapp hmm, that is an interesting question
17:47 osse a recent commit in git itself: http://sprunge.us/CCXh
17:47 lpapp btw, git fetch and git rebase --pull do not quite update all the branches
17:48 lpapp not the git log output anyway; so for instance, if I check out a different branch than the current one, it does not have the latest stuff in the output of gitlog.
17:48 diogenese joined #git
17:48 lpapp so I guess prior to searching, I would also need some kind of global pull
17:48 osse git fetch does by default, but it can be configured otherwise. do you ahve more than one remote?
17:48 lpapp only one remote as far as I am aware.
17:48 osse and git log by default only lists from HEAD and backwards, which is why i suggested log --all
17:49 lpapp right.
17:49 lpapp still, git log does not show the latest.
17:49 ABM_Dan joined #git
17:49 lpapp which confuses me
17:49 lpapp osse: in case of a cherry-pick, the author remains, but the committer is the person doing the cherry-pick.
17:50 lpapp so the commit thing really seems to be what I thought of as "pusher".
17:50 ABM_Dan Hi, I have a common situation that I have to deal with pretty often, I tried writing a command to quickly deal with it
17:50 osse lpapp: what does 'git config remote.origin.fetch' show?
17:50 ABM_Dan git fetch origin master && git merge FETCH_HEAD && git checkout --ours ./composer.lock && git add ./composer.lock && git commit -m "composer.lock merge" && git push origin HEAD
17:50 lpapp osse: +refs/heads/*:refs/remotes/origin/*
17:50 ABM_Dan but when I run git merge FETCH_HEAD it finds the conflict, and it stops the execution
17:50 osse lpapp: that indeed fetches all branches from origin, so that part should be ok
17:50 lpapp osse: I see
17:50 ABM_Dan I obviously know about the conflict, and I would resolve it with the rest of the commands
17:51 ABM_Dan but the error stops the execution, is there a way to prevent it?
17:51 osse ABM_Dan: use ; instead of &&
17:51 ABM_Dan Hm, makes sense.
17:52 m0viefreak joined #git
17:53 Guest83 joined #git
17:54 Darren__ joined #git
17:54 osse ABM_Dan: dare I suggest removing that file from the repo?
17:54 osse is it actually just a lock file?
17:54 veilig2000 joined #git
17:54 ABM_Dan We're using it as a way to trigger a specific build.
17:54 Mannn joined #git
17:55 ABM_Dan Like, we work on separate bundles, and when it's time to build, composer update the central bundle, and merge the updated composer.lock to the repo.
17:55 ABM_Dan And the artifact is built using that specific version.
17:57 ABM_Dan Like, we don't want to update on staging or production, it might have versions that have untested flaws, we update locally, check stuff, if we're happy with everything, we push that lock, and are sure that the staging/production is using tested versions.
17:58 aard joined #git
17:58 led_belly joined #git
17:59 ojacobson ABM_Dan: if you use && to chain commands, then yes, the chain will stop at the first failing command.
17:59 ojacobson Merge conflicts cause git-merge to exit with a failing status.
17:59 dranve joined #git
17:59 ojacobson You are, at some level, getting what you asked for, though perhaps not what you meant to ask for
18:00 ABM_Dan ojacobson: is there, potentially, a way to tell git merge to "expect" a conflict?
18:00 ojacobson No.
18:00 ABM_Dan (the issue is fixed, ; does the job)
18:01 GodGinrai ABM_Dan: what would telling it to expect a conflict do?
18:01 ABM_Dan When a conflict is met, it's marked as such, but no error status is returned.
18:02 cyphase joined #git
18:02 nydel07 joined #git
18:03 lpapp left #git
18:03 davisonio joined #git
18:04 israelzuniga joined #git
18:05 fuchstronaut joined #git
18:05 ShekharReddy joined #git
18:05 altendky joined #git
18:06 disi joined #git
18:09 editshare-ericf joined #git
18:13 eivarv joined #git
18:14 grayjoc joined #git
18:16 robotroll joined #git
18:17 yoan joined #git
18:19 Mikerhinos joined #git
18:21 interne7y joined #git
18:24 Sasazuka__ joined #git
18:29 ORiON-- joined #git
18:30 fahadash joined #git
18:31 rominronin joined #git
18:31 danslo joined #git
18:32 batrick joined #git
18:32 Repox joined #git
18:32 dreiss joined #git
18:36 ningu joined #git
18:36 ningu how do I get a remote branch rename to apply to my local repo?
18:36 preaction rename the branch on your local repo
18:37 ningu hmm
18:37 ningu looks like prune may do it
18:37 ningu preaction: yeah, I can do that, the issue was removing the old branch
18:37 preaction prune will remove the old branch, which isn't what you asked
18:37 ningu true
18:38 ningu either would have worked, I guess
18:39 attente is there a way to force git-am a patch with conflicts marked with <<<<<<< >>>>>>>?
18:39 attente i've tried -3 without luck
18:40 mkoskar joined #git
18:43 catag87 joined #git
18:43 attente ah, never mind. just manually patching with the rebase patch
18:43 kulelu88 joined #git
18:46 Guest83 joined #git
18:47 ningu joined #git
18:48 ningu not sure if this is a directly git-related question but here goes. suppose someone wants to propose a change to a single file in a repo, and they have commit rights to the repo but aren't ready to push it yet. would the standard procedure for proposing such a change be a pull request? and if so would that require creating a new branch?
18:48 davisonio joined #git
18:50 ejb joined #git
18:50 ojacobson That's up to the project, ultimately, but if the project is using pull requests for changes, using a pull request for a small change would make sense
18:51 ningu what would the alternative be?
18:51 ojacobson Discussing the change via mailing lists/irc/slack/whatever, mailing a patch, making the change and asking "hey what do you think" after the fact… many options
18:51 Limix joined #git
18:51 ningu ok
18:51 ojacobson and yes, each pull request generally starts from a new branch based on the target of the pull request (a pull request for origin/master would start with a new branch based on origin/master)
18:51 ningu I know you can mail patches around
18:51 grayjoc joined #git
18:51 ningu just making sure I wasn't missing anything
18:52 Limix Hi all, is there a way to revert a commit, but also get those changes back in the working dir?
18:52 Limix I comitted files that I wanted to pull into a branch along with files that I intended to stage and commit
18:53 soerenb !fixup
18:53 gitinfo So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!
18:53 jimi_ joined #git
18:53 josuebrunel joined #git
18:54 dopesong joined #git
18:55 maroloccio hi. can i checkout branch2 in my working directory but then position my head and branch on branch1 without affecting my working directory?
18:55 ojacobson yes
18:56 ojacobson there are a couple of ways to accomplish that
18:56 maroloccio ojacobson: cool. how :)
18:56 ojacobson just to recap: you want the tree from branch2^{commit} staged and in the work tree, and then branch2 set to be equal to branch1?
18:56 nitric Limix: you could also just create a new branch, reset your current branch to some previous commit, and then checkout the new branch and continue working in the new branch
18:57 maroloccio ojacobson: maybe something simpler than that sorry
18:57 jknetl joined #git
18:58 maroloccio ojacobson: i would like go be "on branch1" (head set to the tip of branch1) but to have at my disposal the content of branch2 in my working directory. so that i can git add --patch parts of that working directory to branch1 and make commits with the incremental adds
18:58 johnmilton joined #git
18:58 Limix nitric, thank you that sounds perfect
18:59 ojacobson ah
18:59 ojacobson I do this all the time
18:59 cdown joined #git
18:59 ojacobson git checkout branch1; git checkout --patch branch2 -- paths/of/interest
18:59 ojacobson commit, repeat the checkout step
18:59 ojacobson continue until satisfied
19:00 jimi_ joined #git
19:00 Ragnor joined #git
19:01 maroloccio ojacobson: i did not know i coudl checkout --patch. but still.. doesnt do it right quite yet :)
19:01 ojacobson The other way you can do it, _starting from a clean checkout of branch1_, is `git rm -rf --cached :/ && git checkout branch2 :/ && git reset HEAD :/`
19:01 ojacobson I'm sure there's a shorter way to spell that, but that's one I use
19:02 ojacobson remove and unstage everything, then stage and check out all of branch 2, then restore the staging area to what it would be if you'd just checked out branch 1
19:02 GavinMagnus joined #git
19:02 maroloccio ojacobson: my problem is that i need the whole of branch2 "inside" branch1 or the commit will fail because CI fails
19:02 ojacobson leaving everything from branch 2 as uncommitted changes
19:02 maroloccio ojacobson: i can't, in other words, checkout only the paths of interest
19:03 yoan_ joined #git
19:05 cyan__ joined #git
19:06 stickyboy joined #git
19:08 GavinMagnus left #git
19:08 hahuang65 joined #git
19:09 stickyboy I get conflicts in `git rebase` that shouldn't happen because the file with the conflicts doesn't exist on the branch I'm rebasing on/to.
19:09 hahuang61 joined #git
19:09 robotroll joined #git
19:10 stickyboy I'm confused. :(
19:10 cyphase joined #git
19:10 maroloccio git rebase --interactive
19:10 maroloccio see what's happening in there
19:11 maroloccio also git log file-that-should-not-be-there to see if it ever appears in the other history..
19:12 yoan__ joined #git
19:12 stickyboy maroloccio: I did `git show targetbranch:path/to/file` and it's "fatal, exists on disk, but not on targetbranch"
19:13 stickyboy Err, ok, let me try the log to be sure.
19:13 paul424 joined #git
19:14 nydel07 joined #git
19:14 stickyboy git log targetbranch:path/to/file      "does not exist in targetbranch"
19:14 stickyboy Hmm.
19:15 stickyboy And the "both modified" bit is weird. The other branch would never modify this file (it's upstream FLOSS project and my file doesn't exist there).
19:16 gtodd joined #git
19:16 johnmilton joined #git
19:18 gtodd I have made a number of small changes in a branch and broken something ... how can "manually bisect"  to find when it last worked (test didn't catch things)?
19:19 gtodd git reset fa81b2a (try the feature ... broken ... go back further ... etc)  iterating back till I find when it last worked
19:20 hahuang61 joined #git
19:20 gtodd then diff against the HEAD or tip of the branch ?
19:20 maroloccio stickyboy: agree, it's weird :(
19:21 gtodd then go back to HEAD and manually fix the files
19:22 gtodd (or undo changes on those files only by checking out a copy of the file from the working revision)
19:23 stickyboy maroloccio: Thinking something weird with graph. Checking a few commits in `git rebase -i` that aren't there but should be. Hmm.
19:24 stickyboy Maybe I need to git prune, recreate my branch from remote, git gc, etc...?
19:25 ki0 joined #git
19:25 star_pro_ joined #git
19:25 ahmedelgabri joined #git
19:27 nitric gtodd: sounds like you want a straight git bisect, but if you really insist on doing it manually, do `git checkout` instead of `git reset`
19:28 nitric git checkout HEAD^ && ./run-your-tests, and repeat until tests fail, and then `git checkout your-branch` and do your fixes
19:29 maroloccio stickyboy: git gc is the last thing i would do in that case. you dont want to look at your reflog? e.g. git log -g
19:30 Ice_Strik2 joined #git
19:31 Ice_Strik2 left #git
19:31 carif joined #git
19:31 Ice_Strike joined #git
19:31 Ice_Strike How to undu the file changes that has not been staged?
19:31 Ice_Strike modified:   app/Http/Kernel.php
19:31 maroloccio git checkout file?
19:32 Ice_Strike What that undu what it was?
19:32 maroloccio i have trouble understanding you sorry
19:32 Ice_Strike maroloccio Need to undu all the files has been changed before staged
19:32 maroloccio before staged?
19:33 Ice_Strike maroloccio I have not executed: git add .
19:33 Ice_Strike Let me explain
19:33 rgrinberg joined #git
19:33 Ice_Strike Yesterday I was working on the project, made little changes but I did not do: git add .
19:34 Ice_Strike At work I was working on the same project
19:34 stickyboy maroloccio: Eh, not worried I'll lose commits or something, just confused why this is conflicting as upstream doesn't care about stickyboy's local modifications to files outside their tree.
19:34 Ice_Strike Now at home, I need to undu the changes from yesterday
19:35 maroloccio Ice_Strike: are home and work 2 different repositories?
19:35 fuchstronaut joined #git
19:36 Ice_Strike maroloccio Yea but using github
19:36 Ice_Strike But I did not git push to github yesterday
19:36 Ice_Strike I did some work at work and pushed to github
19:37 Ice_Strike Now at home, I don't want to do git pull from github because of changes I made yesterday
19:37 maroloccio which you want to keep?
19:37 sangy joined #git
19:38 circ-user-HPmjC joined #git
19:38 Ice_Strike maroloccio http://pastebin.com/C6Cvgzpb
19:38 Ice_Strike I want to move all this modified from yesterday
19:38 Ice_Strike and use latest git pull from yesterday.
19:38 gtodd nitric: OK I'm going to trey the bisect approach
19:39 maroloccio Ice_Strike: sorry afk for a while
19:39 gtodd nitric: does reset cause be to lose changes?
19:39 davisonio joined #git
19:40 nilg joined #git
19:41 stickyboy maroloccio: Maybe it's rebasing some merge of mine on top of the upstream one and gets confused somehow.
19:41 gtodd nitric: I wanted to do git reset HEAD~5
19:41 stickyboy hmm
19:42 gtodd nitric: or git reset --hard HEAD~5  ... and move forward commit by commit til things fail
19:43 nitric gtodd: as long as you've committed everything at some point, `git reset` will not lose changes, but be careful with `git reset --hard`
19:43 nitric gtodd: and again, do "git checkout" instead of "git reset": reset will only change your index, NOT your working filfes
19:44 stickyboy maroloccio: Eh, no. If I look at the commit that is mysteriously conflicting, it's a simple commit that shouldn't conflict. Argh.
19:45 nitric gtodd: and while you *can* modify your working files with `git reset --hard`, you don't want to be doing that, you want (again): "git checkout"
19:45 gtodd oh
19:46 satifant joined #git
19:46 gtodd not sure why I would want to change only an index ok thanks will avoid til I grok it
19:46 gtodd nitric: cheers
19:48 paul424 joined #git
19:48 stickyboy maroloccio: Now I'm manually cherry picking the commits to see where the problem lies,
19:51 stickyboy I hate merges. I wish people would rebase more.
19:51 ki0_ joined #git
19:55 nitric gtodd: yw, good in-depth article on the reset command: https://git-scm.com/blog/2011/07/11/reset.html
19:55 gtodd thanks
19:56 Zamerick joined #git
19:56 jeffreylevesque joined #git
19:56 eijk_ joined #git
19:56 rnsanchez_ joined #git
19:57 stickyboy maroloccio: Ok, I manually cherry-picked the commits up to the troublesome one and ignored any merge commits.
19:57 vuoto joined #git
19:57 stickyboy Now it applies fine.
19:57 stickyboy What does that mean?
19:58 peroht joined #git
20:00 aep-shoutlet joined #git
20:00 phanimahesh joined #git
20:01 star_pro_ joined #git
20:02 Dougie187 joined #git
20:03 nnyk joined #git
20:05 phanimahesh joined #git
20:05 davisonio joined #git
20:07 cyphase joined #git
20:07 kulelu88 before githubs strategy of forking and submitting PRs, how did people used to commit new code to projects? using branches?
20:08 robotroll joined #git
20:08 ojacobson therein lies a massive cultural anthropolgy study, honestly
20:09 ojacobson we've been saying "code review leads to better code, and here's the overwhelming mountain of evidence" for decades, and groups all over have tried lots of ways
20:09 ojacobson but github's the first large-scale code review system that people willingly use
20:09 thefifthsetpin joined #git
20:09 stickyboy "Yo, Linus, please pull from https://blah.com/blah.git k thx bye"
20:09 ojacobson the only even remotely widely-used predecessor was the fact that open-source projects often had closed maintainership, which meant that getting your code in meant communicating with the maintainers
20:10 ojacobson mailing them patches, or posting your code somewhere they can download, or otherwise satisfying whatever requirements the project leadership had, usually motivated more by their own convenience and operational control than by any specific desire for code review
20:11 ojacobson closed-source stuff often had - still has, but github has changed the world there, too - wildly different incentives, which often meant in practice that how you got your changes in was "commit them, then maybe review them after the fact"
20:11 nitric kulelu88: see https://ariejan.net/2009/10/26/how-t​o-create-and-apply-a-patch-with-git/, and read about the git commands "patch", "apply", and "am"
20:12 nitric ....and then be thankful github came along  :-)
20:13 pcre_ joined #git
20:13 nwkj86 joined #git
20:13 Lunatrius joined #git
20:14 kulelu88 I'm not that thankful for a closed-source solution that talks about open-source but hides its own code (but that is another discussion all together).
20:14 nitric use gitlab then
20:16 stickyboy And then you have two problems. :D
20:16 stickyboy writing code AND hosting the server where your repositories live.
20:16 jayc joined #git
20:16 Limes_ joined #git
20:16 eijk_ joined #git
20:17 nitric well, you have to suffer your principles, otherwise you don't gain enough vitriol to argue about them
20:17 eijk_ joined #git
20:19 stickyboy "suffer your principles", I like that. :)
20:19 stickyboy You can tell I'm a sysadmin.
20:19 ojacobson the real game changers for pull requests are 1. anyone can open one, you don't need to be a maintainer to _submit code for consideration_, and 2. the UI for reviewing and accepting proposals is polished and highly structured, in ways that are relatively easy to explain to newcomers (relative to, say, emailed patch serieses)
20:19 ojacobson 1.'s tricky for host-your-own, just ask wordpress folks how quickly pingbacks fell to spam
20:20 catag87 joined #git
20:20 nnyk joined #git
20:20 durham_ joined #git
20:20 catag87 joined #git
20:22 d4rklit3 joined #git
20:22 zeroed joined #git
20:23 ash_workz joined #git
20:25 mohabaks joined #git
20:31 rgrinberg joined #git
20:32 stickyboy joined #git
20:32 danslo joined #git
20:32 rominronin joined #git
20:34 dreiss joined #git
20:36 fuchstronaut joined #git
20:37 maroloccio joined #git
20:37 bambams !backup !rewrite
20:37 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.c​om/GitBestPractices/#backups
20:37 gitinfo Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum
20:39 maroloccio ojacobson: hey, what do you think of just this to achieve what we discussed before? git symbolic-ref HEAD refs/heads/master any undesirable effects?
20:40 ojacobson that's morally equivalent to 'git reset --soft master'
20:42 maroloccio ojacobson: i think not
20:42 maroloccio ojacobson: unless i don't get the sense of _morally_ equivalent? :)
20:42 crayon joined #git
20:42 osse maroloccio: he means it's the same thing
20:43 maroloccio osse: then, it's not
20:43 osse how so?
20:43 maroloccio if i do the reset soft from branch1, i am still on branch1 am i not?
20:44 osse maroloccio: you mean if you run 'git reset --soft branch1' ?
20:45 maroloccio git checkout branch1; git reset --soft master. assuming branch1 stems from master, this is only going to slide my head pointer from the tip down to master, without touching the index or working dir, right?
20:45 osse yes
20:45 osse regardless of whether branch1 stems from master or not, btw
20:45 maroloccio actually, leaving "behind" that commit i was on before which i would now like to ask.. how do i get it back optimally? git rev-list? git log -g, find it and go back to it?
20:45 jaguarmagenta joined #git
20:46 osse back to where? branch1 ?
20:46 maroloccio A -> B (master) -> C (branch1)
20:46 osse you can do  git reset --soft branch1
20:47 maroloccio it won't slide back my head to where it was before
20:47 osse no, sorry. git reset --soft branch@{1}
20:47 maroloccio yes, get it from the reflog at the position it was before the change. agree
20:48 osse ah, that's one difference between symbolic-ref and reset --soft. symbolic-ref would not update branch1, only HEAD itself.
20:48 maroloccio yes
20:48 maroloccio that's one
20:49 maroloccio what i really wanted to do was to be inside branchA, checkout the content of branchB but keeping my head pointing at branchA. then git add --patch committing changes little by little.
20:49 maroloccio osse: do you know a better way to do the above? i only discovered git symbolic-ref by trial and (lots-of) error
20:50 osse maroloccio:  git checkout branchB -- .
20:50 maroloccio it's almost like i wanted to checkout branchB whilst staying on branchA
20:50 maroloccio you mean checkout --patch, osse ?
20:50 osse no
20:50 maroloccio then i am confused
20:50 osse you said you wanted to checkout the contents of branchB
20:51 osse that's what git checkout branchB -- . does
20:52 osse but with that said, git checkout --patch branchB -- . does sound more reasonable.
20:58 aidalgol joined #git
21:00 SourTrout joined #git
21:01 Spec-Chum joined #git
21:01 WaReZ joined #git
21:03 ki0 joined #git
21:04 SourTrou_ joined #git
21:06 dsdeiz joined #git
21:07 moriko joined #git
21:08 snowkidind joined #git
21:09 maroloccio joined #git
21:13 ahr3n joined #git
21:13 maroloccio osse: again, a difference. git checkout --patch branchX -- . is actually different from git checkout branchX -- . (feels nuts to me that it is, but it seems to be, specifically: even if i accept all changes.. i am not really getting all the files from branchX by using --patch)
21:13 eijk_ joined #git
21:15 tanuki joined #git
21:15 osse maroloccio: i didn't say they where the same
21:15 osse welp, then should be if you accept all
21:15 osse are you in a subdir?
21:17 maroloccio osse: you did not say that, true
21:19 ahr3n joined #git
21:22 hahuang61 joined #git
21:26 ahr3n joined #git
21:29 maroloccio sorry, but what does this mean? warning: ignoring broken ref refs/remotes/origin/HEAD <-- ?
21:30 Derperperd joined #git
21:30 l4v2 joined #git
21:32 zacts joined #git
21:32 zumba_addict joined #git
21:33 zumba_addict i'm trying to clone a repo but I also included the directory that I only want to clone. It;s failing with "project you were looking for could not found"
21:34 osse zumba_addict: it cannot be done
21:34 zumba_addict so command is like git clone git@git.domain.com:app/leg.git/some/path
21:34 osse clone the whole thing
21:34 zumba_addict oh
21:34 ahr3n joined #git
21:34 zumba_addict yup, that's what i'll do then i'll just add another command for copying
21:34 zumba_addict thanks
21:35 tyreld joined #git
21:36 l4v2 After reading this https://help.github.com/articles/why-are-my-​contributions-not-showing-up-on-my-profile/ is it safe for me to conclude that the reason why my contributions graph on GItHub.com is not changing is because I am only working on my own projects, and not projects I’ve cloned?
21:37 l4v2 Or is it because I have to config my email?
21:41 ahr3n joined #git
21:43 shaggycat_ joined #git
21:43 robogoat joined #git
21:44 rewt l4v2, that sounds more like a !github question
21:44 gitinfo l4v2: Note that git != github. Feel free to ask us about Github-specific features (Forks, Pull Requests, Wikis, etc), but there are no guarantees. There is a #github channel, which might help too (again, no guarantees)
21:44 ColdKeyboard joined #git
21:45 l4v2 oh ok, thanks
21:45 dsdeiz joined #git
21:45 dsdeiz joined #git
21:49 jargan joined #git
21:49 kadoban joined #git
21:50 ahr3n joined #git
21:53 kbs joined #git
21:58 maroloccio thanks osse for your help earlier
22:00 Mannn joined #git
22:00 kyan joined #git
22:04 zacts joined #git
22:06 tortib_ joined #git
22:09 zacts joined #git
22:14 tortib_ joined #git
22:23 interne7y joined #git
22:25 kyan joined #git
22:25 vassagus joined #git
22:26 danmackay joined #git
22:26 tortib_ joined #git
22:26 SourTrout joined #git
22:28 bluezone joined #git
22:33 rominronin joined #git
22:34 tortib_ joined #git
22:34 rgrinberg joined #git
22:36 spudowiar joined #git
22:37 fuchstronaut joined #git
22:38 byteflame joined #git
22:43 red_ joined #git
22:44 axl__ joined #git
22:46 jimi_ joined #git
22:47 axl_ joined #git
22:49 Darren_ joined #git
22:53 pwnz0r joined #git
23:03 rbern joined #git
23:05 SourTrout joined #git
23:06 gargola joined #git
23:07 maroloccio joined #git
23:11 ejb joined #git
23:12 ahr3n joined #git
23:13 Guest75 joined #git
23:14 romerocesar_ joined #git
23:15 jaguarmagenta joined #git
23:17 ljc joined #git
23:19 BSaboia joined #git
23:23 hahuang61 joined #git
23:27 d0nn1e joined #git
23:29 Guest97561 joined #git
23:29 Celelibi joined #git
23:29 madewokherd joined #git
23:31 ksolv joined #git
23:31 vassagus joined #git
23:31 cdown joined #git
23:31 dsdeiz joined #git
23:31 dsdeiz joined #git
23:38 ash_workz joined #git
23:39 djapo joined #git
23:46 mar77i joined #git
23:47 ArchNoob joined #git
23:52 ksolv joined #git
23:57 ToBeCloud joined #git

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