Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-07-07

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

All times shown according to UTC.

Time Nick Message
00:03 pioiytr joined #git
00:04 nick[0] ojacobson: upstream package gets imported to the (upstream) branch using git-buildpackage, which then merges upstream on top of master.  I'm trying to prevent the merge of a directory which conflicts with our build system
00:04 nick[0] ojacobson: from what I've read, using attributes and a merge=ours driver/strategy is the correct approach, and .gitignore is not
00:05 jaguarmagenta joined #git
00:05 ojacobson so this is purely generated content?
00:06 nick[0] what do you mean?
00:07 MattMaker joined #git
00:10 Lipp joined #git
00:10 nick[0] 1. merging the upstream build system creates a merge conflict 2. allowing that directory to be merged causes generation of untracked files and leads to regular PITA inconveniences, which are completely unecessary, since I'm 95% certain git attributes can be used to prevent this.
00:10 romerocesar joined #git
00:11 CheckDavid joined #git
00:14 jimi_ joined #git
00:17 anuxivm joined #git
00:17 MattMaker joined #git
00:17 piling joined #git
00:18 pur3eval joined #git
00:21 romerocesar joined #git
00:26 EvilPeng1 joined #git
00:26 mkopriva joined #git
00:26 fstd_ joined #git
00:27 shgysk8zer0_ joined #git
00:27 shgysk8zer0_ joined #git
00:27 anuxivm joined #git
00:28 nnyk joined #git
00:28 azerus joined #git
00:31 ekinmur joined #git
00:31 EvilPenguin joined #git
00:34 nick[0] maybe I need to combine this with filter-branch?
00:43 mellernoia joined #git
00:45 sctskw joined #git
00:45 xaviergmail joined #git
00:46 pur3eval joined #git
00:48 MattMaker joined #git
00:54 azbarcea joined #git
00:54 Literphor joined #git
00:57 mwhooker joined #git
00:57 MattMaker joined #git
00:59 ag4ve joined #git
00:59 rominronin joined #git
01:00 MattMaker joined #git
01:00 tesuji_ joined #git
01:03 rbern joined #git
01:06 tesuji joined #git
01:12 ochorocho__ joined #git
01:17 arescorpio joined #git
01:17 devbug joined #git
01:18 dreiss joined #git
01:18 cqi joined #git
01:20 mkopriva joined #git
01:22 cqi joined #git
01:23 SpeakerToMeat joined #git
01:23 cqi joined #git
01:24 Kaisyu joined #git
01:26 SpeakerToMeat joined #git
01:26 pks joined #git
01:28 cdg_ joined #git
01:32 xaviergmail joined #git
01:37 regfop joined #git
01:38 regfop left #git
01:40 HoierM joined #git
01:40 ekinmur joined #git
01:40 ag4ve so someone copied the files in a repo and then created a new repo (basically) - what's the easiest way to merge their commits on top of the original repo?
01:41 Hello71 git am
01:41 Hello71 er, format-patch
01:46 rpd you could also make the original repo a remote and `git pull --rebase` if you want to fix inconsistencies/conflicts that might come up
01:46 ag4ve can I make /lay down a patch /commit?
01:46 ahmedelgabri joined #git
01:46 rpd you can use git-am to create a commit from a patch created by git format-patch
01:46 rpd you can also use git-apply to apply a patch to a set of files
01:48 ilbot3 joined #git
01:48 Topic for #git is now Welcome to #git, the place for git help and self-reflection | Public logs at http://goo.gl/BuUi5o | Current stable version: 2.9.0 | First visit? Read: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | Fork by yourself, shame on you. Fork with a friend, now we're getting somewhere.
01:48 Hello71 it was my understanding that format-patch would handle that.
01:49 ag4ve Probably check out the date right before the mess was created and merge on top and then see about bringing things current
01:49 Hello71 although you might need to rewrite the root commit.
01:49 boombatower joined #git
01:50 rpd format-patch still makes a patch per commit even if you run it over a range of commits
01:53 ag4ve thanks - I've got two options (I kinda knew I could add a local remote and merge but didn't know about format-patch) I'll read up and tias
01:53 nick[0] what is the best way to prevent conflict when merging from upstream.  eg: local/dir must always take precedence over remote/dir, but remote should be prefered for everything else.  Also, if local/dir2 does not exist, what is the best way to prevent remote/dir2 from being merged?
01:54 ag4ve ours
01:54 rpd When you say "must always take precedence," you mean, if local/dir and remote/dir disagree, local/dir's patch is taken automatically?
01:55 rpd yeah just use the merge strategy "ours" like git merge -s ours
01:56 mwhooker joined #git
01:57 nick[0] rpd: but it must only use "ours" for that one dir.  Isn't git merge -s ours for everything?  I've also tried using .gitattributes, but following the following doc didn't work:
01:57 nick[0] https://git-scm.com/book/en/v2/Customizing-Git-Git-Attributes
01:58 nick[0] rpd: I am using git 2.1.4
01:59 ag4ve huh, I'd be interested in how to set a strategy per path - no use but...
02:01 rominronin joined #git
02:03 ahr3n joined #git
02:04 rgrinberg joined #git
02:04 tenc joined #git
02:04 rkazak joined #git
02:06 jaguarmagenta joined #git
02:09 harish joined #git
02:09 ag4ve ah nifty - idk you could define per path policies in attributes (which means I'm no help)
02:10 chachasmooth joined #git
02:12 nick[0] ag4ve, rpd: maybe it's one of those rarely used features that wasn't really usable until a certain release?  I'm exhausted from spending a couple of hours trying to get it to work.  Would either of you quickly be able to find if it was usable in git-2.1.4?
02:14 rpd uh, I was afk but I guess I could try
02:14 rpd but yeah, you can set .gitattributes in a directory to make it do something different
02:15 rpd anyways, if you want to be selective, you could always just use git add -p to narrow down the set of files being altered
02:15 rpd then when you merge, just use -s ours
02:16 mwhooker joined #git
02:17 nonconvergent joined #git
02:22 ejb joined #git
02:25 watabou joined #git
02:28 cqi joined #git
02:28 hexagoxel joined #git
02:32 fscala joined #git
02:32 stantonk joined #git
02:40 devbug joined #git
02:42 tjone270 joined #git
02:42 calamity_man joined #git
02:44 stantonk joined #git
02:44 a_thakur joined #git
02:44 Lipp joined #git
02:46 Goplat joined #git
02:46 stantonk joined #git
02:47 chrisshattuck joined #git
02:49 rnsanchez joined #git
02:49 liuzhen joined #git
02:49 rkazak joined #git
02:50 ahr3n joined #git
02:52 a_thakur_ joined #git
02:53 d4rklit3 joined #git
02:54 mwhooker joined #git
02:56 oich joined #git
02:57 snowkidind joined #git
02:58 tgunr joined #git
02:59 pioiytr joined #git
03:00 Gsham joined #git
03:03 fahadash joined #git
03:04 cbreak joined #git
03:05 Rodya_ joined #git
03:05 eight58 joined #git
03:07 eight58 joined #git
03:08 mkopriva joined #git
03:11 xcesariox joined #git
03:12 ochorocho__ joined #git
03:12 rnsanchez_ joined #git
03:14 texinwien_ joined #git
03:15 fracting joined #git
03:15 devbug joined #git
03:15 eight58 joined #git
03:18 _rnsanchez_ joined #git
03:19 zhiwliu_ joined #git
03:20 pur3eval joined #git
03:25 rkazak joined #git
03:31 drbean joined #git
03:35 RabbitKing joined #git
03:35 RabbitKing <RabbitKing> hi
03:35 RabbitKing * Jucato has quit (Quit: Konversation terminated!)
03:35 RabbitKing <RabbitKing> want tutorial phyton lover
03:35 RabbitKing <RabbitKing> come to ##xWindow
03:35 RabbitKing <eevee> stealth_: it is much more confusing.  if i see "say_hi" and i want to know what it is, i can just look at the top of the file to see where it's imported from
03:35 RabbitKing <RabbitKing> top notch document from me
03:35 RabbitKing <RabbitKing> there is proxy in it
03:35 RabbitKing <RabbitKing> you need to join ##xWindow to open it
03:36 RabbitKing left #git
03:36 stantonk joined #git
03:38 MZAWeb joined #git
03:39 inflames joined #git
03:43 zacts joined #git
03:44 Isla_de_Muerte joined #git
03:45 FuzzySockets joined #git
03:45 mehola joined #git
03:45 oich joined #git
03:45 Lipp joined #git
03:46 Rodya_ joined #git
03:48 Rodya_ joined #git
03:48 Dbk joined #git
03:53 pioiytr joined #git
03:56 Rodya_ joined #git
03:58 chrisshattuck joined #git
03:59 hk238 joined #git
04:00 unbalancedparen joined #git
04:01 stantonk joined #git
04:01 settermjd joined #git
04:02 rominronin joined #git
04:02 mkopriva joined #git
04:05 akushner joined #git
04:05 xlegoman joined #git
04:05 Rodya_ joined #git
04:06 pioiytr joined #git
04:06 jaguarmagenta joined #git
04:07 cwar joined #git
04:09 phanimahesh joined #git
04:09 chrisshattuck joined #git
04:11 rscata joined #git
04:12 rbern joined #git
04:18 rkazak joined #git
04:19 Rodya_ joined #git
04:23 romerocesar joined #git
04:24 sword joined #git
04:25 hahuang61 joined #git
04:26 muthu joined #git
04:27 robotroll joined #git
04:28 sangy joined #git
04:29 Rodya_ joined #git
04:29 Literphor joined #git
04:33 romerocesar joined #git
04:35 kellytk joined #git
04:35 Chill_Surf joined #git
04:36 kellytk I see that Git commit messages include the timezone.  Is it possible to base it on UTC?
04:36 shgysk8zer0_ joined #git
04:37 oich joined #git
04:38 stantonk joined #git
04:40 Rodya_ joined #git
04:41 gelei007 joined #git
04:42 ShapeShifter499 joined #git
04:43 milki kellytk: its stored as unix time. you can change the display format with options
04:43 milki hmm
04:43 milki iirc
04:43 milki >.>
04:43 thiago kellytk: why do you ask?
04:44 milki o
04:44 milki is the timezone... seperate?
04:45 thiago $ git cat-file -p HEAD | grep author
04:45 thiago author Thiago Macieira <thiago.macieira@intel.com> 1467840888 -0700
04:46 milki https://github.com/git/git/blob/master/Documentation/date-formats.txt#L11
04:47 milki unixtime + tz
04:47 milki why -.-
04:47 thiago why what?
04:47 milki i thought unixtime didnt have a timezone
04:47 thiago it doesn't
04:47 thiago that's why it's stored separately
04:49 milki hm, but it also doesnt feel like necessary information if we already have unixtime
04:49 thiago then you wouldn't know what timezone I committed on
04:50 cqi joined #git
04:50 Cabanossi joined #git
04:50 zumba_addict joined #git
04:50 ajf- joined #git
04:51 Rodya_ joined #git
04:54 milki i suppose i dont see the need to know that
04:55 quetz joined #git
04:55 kellytk Why does the timezone matter?  It seems everything being relative to UTC would be simpler
04:57 zacts joined #git
04:58 oich joined #git
05:00 pur3eval joined #git
05:00 sudarshans joined #git
05:01 ramanuja joined #git
05:01 NeoZeed joined #git
05:01 ramanuja left #git
05:02 ramanuja joined #git
05:02 ramanuja left #git
05:04 romerocesar joined #git
05:04 romerocesar joined #git
05:05 sudarshans left #git
05:05 a_thakur joined #git
05:06 eight58 joined #git
05:07 subhojit777 joined #git
05:08 freimatz joined #git
05:08 Lipp joined #git
05:11 lindenle joined #git
05:12 thiago kellytk: because some people want to know the timezone of the committer
05:12 thiago why not have the extra information?
05:12 drawkula joined #git
05:13 ochorocho__ joined #git
05:13 atrus some people might want to know the gps coordinates of the committer, that doesn't necessarily mean they should get it :) that said, knowing timezones tells you something about how to coordinate with the author meaningfully, so that's valuable
05:15 Rodya_ joined #git
05:16 pur3eval joined #git
05:16 Macaveli joined #git
05:16 Literphor joined #git
05:17 lindenle joined #git
05:17 Terens Hello
05:18 Terens having a branched merged to
05:18 Terens master then
05:18 Terens rebasing this branch may cause problems?
05:18 thiago yes
05:18 thiago unless you rebase it on master, that is
05:19 thiago that will be a no-op until you add more commits
05:21 chrisshattuck joined #git
05:21 Terens so I have merged branch A to master. What is the correct way to update my A branch with master changes?
05:21 lordjancso joined #git
05:22 thiago merge master to A
05:22 nick[0] thiago: this seems like a learning
05:22 Terens okay
05:23 nick[0] moment for me too.  do you checkout A, then merge -X theirs master, or use some other method?
05:23 thiago drop the -X theirs part
05:23 thiago but yes
05:24 nick[0] thiago: why would the -X theirs be dropped in this case, if the desired effect is similar to a rebase?
05:24 ome joined #git
05:25 thiago because a) it's unnecessary and b) that's not what -X theirs does
05:25 nick[0] because it would overwrite what is unique to the local branch?
05:25 thiago no, it wouldn't
05:26 hahuang61 joined #git
05:28 nick[0] thiago: so the best approach is to manually reconcile the conflicts, file by file?  (this is a point I'm still weak on, so thank you very much for your patience)
05:28 thiago there will be no conflicts
05:29 nick[0] oh!  For some reason I run into them very often, to the point that it's starting to become a bit of phobia ;-)
05:30 Lipp joined #git
05:30 eight58 joined #git
05:30 jaguarmagenta joined #git
05:31 AndreSomers joined #git
05:32 nick[0] thiago: do think think git-2.1.4 is new enough to not cause undue headaches, or would it be a substantially worthwhile to upgrade to 2.9.0?
05:33 thiago in this case, there will be no conflicts
05:33 thiago remember we started with "so I have merged branch A to master"
05:33 thiago why would you not run the latest?
05:33 cqi joined #git
05:35 nick[0] touché wrt "so I have merged branch A to master"!  I'm a Debian stable (aka: stale ;-)) user, a backport of a newer version isn't available, and I haven't yet decided if the effort of maintaining a bpo for two years is worth the payoff
05:36 thiago I don't think there's any specific need to update
05:36 thiago it's old, but functional
05:36 xcesariox joined #git
05:37 nick[0] do you know if .gitattributes is better supported in newer versions?  I was almost pulling out my hair today trying to get them to work properly. (merge strategy ours, recursively, for one specific directory)
05:37 oich joined #git
05:37 romerocesar joined #git
05:37 thiago I don't know anything about it
05:38 nick[0] hmm, I'm starting to get the impression it's a very rarely used feature haha
05:38 Vauff joined #git
05:38 nick[0] thiago, thank you for your time. :-)
05:41 drbean joined #git
05:43 nnyk joined #git
05:47 diogenese joined #git
05:48 Alienpruts joined #git
05:49 monty joined #git
05:49 naxgid joined #git
05:50 mkopriva joined #git
05:52 The_Phoenix joined #git
05:52 jikz joined #git
05:52 cerberii joined #git
05:52 watersoul joined #git
05:53 sctskw joined #git
05:54 stantonk joined #git
05:54 anohigisavay joined #git
05:55 anohigisavay hi. how can i remove a merge commit?
05:55 anohigisavay i want to undo the merge, but i have also merged from other branches since
05:55 nilg joined #git
05:55 naxgid joined #git
05:55 thiago anohigisavay: move on
05:56 thiago anohigisavay: that's my advice
05:56 naxgid left #git
05:56 thiago I can teach you to apply the reverse of the changes that were introduced by that merge. But it's not the same as un-merge. You can't merge that branch again.
05:58 anohigisavay yea sure. i can even drop those commits because in my specific situation that branch has been abandoned
05:58 anohigisavay was just curious
05:58 MrMojito1 joined #git
05:58 anohigisavay if it's possible it'll make the history cleaner
05:58 kellytk How common is use of "git flow"?
06:01 thiago anohigisavay: do you understand the implications of a history !rewrite ?
06:01 gitinfo anohigisavay: 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
06:01 zeroed joined #git
06:03 rominronin joined #git
06:04 coolguy_ joined #git
06:06 Nitesh joined #git
06:06 phaleth joined #git
06:07 oich joined #git
06:10 Neobenedict_ joined #git
06:10 Neobenedict_ joined #git
06:11 anohigisavay thiago: yea i know that
06:12 thiago anohigisavay: ok, then do this:
06:12 thiago 1) create a new branch starting from the commit prior to the bad merge
06:12 zhiwliu joined #git
06:12 thiago 2) make sure you really are in the commit prior to the bad merge
06:13 thiago 3) replay all the activity that happened after the bad merge. If it's a normal commit, cherry-pick it; if it's a merge, do the merge again.
06:13 thiago 4) delete the master branch, rename this branch to master
06:13 thiago 5) tell everyone else to delete their clones and re-clone
06:14 anohigisavay cool that makes sense
06:14 pur3eval joined #git
06:14 freimatz joined #git
06:15 thiago oh, forgot one step: push -f the master branch to your public repository
06:15 jknetl joined #git
06:15 Rodya_ joined #git
06:17 zacts joined #git
06:17 soee joined #git
06:17 soee hi, if i have commit pushed that has one file that i woudl like to remove form it and push with next commit, how can i do this ?
06:18 thiago soee: remove from history?
06:18 soee huh ? :)
06:19 thiago soee: do you want to remove the commit from existence?
06:19 freimatz3 joined #git
06:22 soee nope, one file it was containing
06:22 thiago soee: so you just want to remove the file? git rm
06:22 soee chnages that were made in one file and pushed with commit X and than push this changes with commit Y
06:23 thiago not following
06:23 romerocesar joined #git
06:24 moei joined #git
06:24 martin_g joined #git
06:27 englishstudent joined #git
06:29 settermjd joined #git
06:31 Sceorem joined #git
06:32 b1tchcakes joined #git
06:32 MattMaker joined #git
06:34 Oebele joined #git
06:36 Alienpruts joined #git
06:41 vuoto joined #git
06:41 ashokrajar joined #git
06:43 miklcct joined #git
06:44 fscala joined #git
06:44 mkopriva joined #git
06:45 Lipp_ joined #git
06:47 aep4Ayai joined #git
06:47 nnyk joined #git
06:48 ferr joined #git
06:49 TomyWork joined #git
06:49 aep4Ayai hi
06:49 gitinfo aep4Ayai: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.
06:49 aep4Ayai is it possible to make git add -e not break functions in half?
06:49 aep4Ayai I have a file with many changes and want to have a commit that only adds the new functions
06:49 aep4Ayai without changing the old ones
06:49 kj_ joined #git
06:50 aep4Ayai https://gist.github.com/anonymous/258dcfc7b37b929a293485d904a6e1e0
06:50 aep4Ayai for example
06:51 neanderslob joined #git
06:51 aep4Ayai I want line 7 to appear in line 2
06:54 anohigisavay joined #git
07:01 vmiklos joined #git
07:02 romerocesar_ joined #git
07:04 rominronin joined #git
07:04 zeroed joined #git
07:07 irco joined #git
07:08 anohigisavay well eventually i did git rebase -ip and removed the merge commit. so far so good
07:08 mkopriva joined #git
07:08 anohigisavay i wonder how git determines which parent the child commit should attach to though?
07:09 tvw joined #git
07:11 xeviox_ joined #git
07:11 garet joined #git
07:11 thiago aep4Ayai: ignore that. There's nothing wrong.
07:12 thiago anohigisavay: each commit lists its parents
07:13 garet If I have my api and frontend separated into separate repos, and then create a master repo that imports both the api and fe and submodules, is it possible to commit changes to the individual repos form inside the submodules
07:13 garet or will I have to have duplicates on my hd? trying to brainstorm a way to reduce redundancy
07:13 xeviox_ hi guys, is it possible to check in which branches a specific branch is already merged? (so the "reverse" of git branch --merged)
07:13 garet but I want to have them as submodules in a single folder to ease of container creation using docker-compose
07:13 kellytk How common is use of "git flow"?
07:14 ochorocho__ joined #git
07:14 thiago garet: if you're using submodules, there will be no duplicates
07:14 thiago garet: each submodule is a repository
07:14 romerocesar joined #git
07:14 jast garet: possible, but slightly more tedious ('git submodule update' detaches HEAD, and to make changes you have to re-attach it)
07:14 thiago xeviox_: I think it's --contains
07:14 jast kellytk: there are a fair number of people using it, or something like it
07:14 thiago kellytk: I think the lack of answer is your answer
07:15 xeviox_ thiago: like "git branch --contains" ?
07:15 thiago xeviox_: yes
07:15 ajf- If I make a new origin out of a repo I want to keep track of fetching new commits every now and then but still do my own development on top of it, what workflow should I use ?
07:15 xeviox_ thiago: thanks, whill try it :)
07:15 thiago ajf-: any workflow
07:16 kellytk That's a good point thiago
07:16 circ-user-2lJvp joined #git
07:16 ajf- I mean for updating. I'm doing git fetch other dev; git merge --squash other/dev
07:16 ajf- and also merge alone
07:17 ajf- but then any change I made on top of it appear as conflicts again any time I attempt to do that again
07:17 kellytk The more I've learned about Git the less appeal flow has to me, because what it does is easy to do with Git manually, and my workflow is a bit different.  For example I don't use bugfix or support branches
07:17 fees joined #git
07:17 pioiytr joined #git
07:17 thiago ajf-: you can try rebasing instead of merging
07:17 GavinMagnus joined #git
07:17 ajf- thiago can I not bring the entire commit history using rebase ?
07:17 jast in the end, 'flow' is just a more formalized variant of a sound idea, and IMO the idea itself is more important than the formalism
07:18 thiago ajf-: but it's not likely to produce fewer conflicts. It might produce more.
07:18 kellytk Rather than using Git flow I think it's smarter for me to use Git directly
07:18 jrahmy joined #git
07:18 james3015 joined #git
07:18 keltvek joined #git
07:19 kellytk jast: What would you say is the key of the "sound idea"?
07:20 realz joined #git
07:20 garet thiago: so when I pull in a submodule, it will pull in that repos .git folder, allowing me to commit changes directly in that folder?
07:21 thiago garet: when you fetch a repository, it updates that repository's .git.
07:21 dreiss joined #git
07:21 garet yeah, I just dont' want to have to commit changes to my api's repo, push the changes to github, and then fetch the changes into my main project's repo
07:22 aep4Ayai thiago: I'm not sure what you mean by "there is nothing wrong". I'm asking how to make it so that I don't have to move those lines around myself.
07:22 thiago garet: if they are submodules, they are separate repositories
07:22 thiago aep4Ayai: you don't have to. There's nothing wrong.
07:22 rkazak joined #git
07:23 aep4Ayai thiago: But I don't want to commit those changes. I only want to commit lines 2-6 in my gist above.
07:23 jast kellytk: there are many ways to look at it, but to me the key insight is the advantage of separating bugfixes from features, so you can fix maintenance and active branches at the same time. flow is a rather complex way of doing that, though, it's just as easy to do fixes on the maint branch and merge to the active branch (and never the other way around).
07:23 mkopriva joined #git
07:23 thiago aep4Ayai: so you want the the line "void do_update_cache_job(void *data)" (without static) to remain?
07:24 thiago aep4Ayai: that is, "void do_update_cache_job(void *data)" and in the next line "void job_schedule_update(struct update_data *data)" ?
07:24 thiago aep4Ayai: that's not valid C
07:24 thiago aep4Ayai: are you sure you want that?
07:24 aep4Ayai thiago: I know. that's why I'm moving lines around manually.
07:24 thiago aep4Ayai: well, don't change the lines you don't want to see in the diff
07:24 kellytk jast: Thank you.  Does that mean the develop branch would have both feature and bugfix branches off of it, and hotfixes would branch off of the master branch?
07:25 aep4Ayai thiago: handling that case is what git add -e is for, isn't it?
07:25 jast kellytk: the master branch in the flow model is pure convenience for end users, so there are other ways to do that, too. the feature branches are simply the idea that you can do more experimental stuff without polluting a main development branch... minus using common sense, because in practice there is no real advantage in creating an extra branch for a trivial feature
07:25 thiago aep4Ayai: either edit the diff to look like what you want or edit the source
07:25 thiago aep4Ayai: yes. But it isn't smart enough to read your mind.
07:25 thiago aep4Ayai: all it knows is that some lines changed.
07:25 aep4Ayai thiago: It doesn't have to read my mind. It only has to be smart enough to not splice functions like it did in the gist above
07:26 thiago aep4Ayai: it doesn't know about "functions"
07:26 thiago aep4Ayai: it knows about lines. Some lines got removed, some others got added
07:26 thiago aep4Ayai: it doesn't know that it should list the removal in the middle of the additions. It never does that.
07:26 jast kellytk: well, there are different naming schemes. let's take the one used in git's own repo. master is things that will end up in the next release and have gone through a fair bit of testing already. next is things that will almost certainly end up in the next release but may need some more changes. pu is "move fast and break things". finally, maint is a previous release series that still receives patches.
07:26 aep4Ayai thiago: it already seems to have some special knowledge about functions. e.g. it gives me the names of functions the hunks are in sometimes even if the the function name is not in the diff itself
07:27 thiago aep4Ayai: and that's very often wrong
07:27 thiago aep4Ayai: it just tries to find the last line that didn't start with spaces
07:27 aep4Ayai aha, that would have fixed the splicing problem I'm talking about.
07:27 aep4Ayai if it did that for the diff itself
07:28 jast kellytk: in that layout, if you need to fix a bug, you do it in maint and then merge maint to master. new features go to master, next or pu, depending on how robust they seem to be. because, in the git model, features can "graduate" from pu to next to master, git uses feature branches for anything that isn't added straight to master, simply because that way the maintainer can graduate individual changes without, say, merging all of next into master.
07:28 thiago aep4Ayai: you can try to use the minimal, patience or histogram diff algorithms. Unlikely to produce any difference, though.
07:29 jast kellytk: and perhaps that makes the purpose of feature branches clearer: they are for when you don't know yet when/where you will want to merge them, nor in what order. :)
07:30 kellytk Thanks jast.  I've read http://nvie.com/posts/a-successful-git-branching-model/ and it doesn't describe the difference between how the bugfix and hotfix branches work in Git flow, do you know?
07:30 kellytk I know that hotfixes come off of the main branch, so I'm guessing bugfixes come off of the develop branch the same as feature branches do
07:30 gwokamollaye joined #git
07:30 jast kellytk: if you look at flow, it seems to promise that master only ever points at a tagged release, and never anything between two releases. if you don't care about that, and personally I don't, you can simply put hotfixes on master
07:30 thiago bugfixes come in the branch that the bug exists in
07:31 jast flow doesn't distinguish between bugfix and hotfix branches btw
07:31 kellytk `git flow init` seems to disagree
07:31 jast oh, you mean the tool, not the underlying model
07:31 thiago for me, a hotfix is something that branches off the last release and triggers an immediate release
07:31 jast I don't really know anything about the too
07:32 jast l
07:32 kellytk Correct, pardon me
07:32 thiago instead of taking in all the bugfixes that have happened since that release
07:32 nilg joined #git
07:32 xcesariox joined #git
07:32 jast wouldn't dream of using it
07:32 kellytk thiago: Likewise
07:32 kellytk jast: That's what I'm thinking
07:33 kellytk I have an idea of how I'd like Git workflow to go, so I think what I'll do is document that, and create branches/complexity as-needed
07:33 jast sounds good to me
07:33 djb-irc joined #git
07:33 kellytk A benefit to using Git flow the tool is it let me set an alternative name for the master branch before it's created, something I can't figure out on my own
07:33 jast I don't really like "complexity first" approaches much :}
07:33 aswen joined #git
07:33 jast oh, that's simple
07:34 kellytk I prefer to use "release" as opposed to master
07:34 jast before your first commit, 'git branch -m newname'
07:34 xeviox_ thiago: exactly what I was looking for, thanks :D
07:34 kellytk I'll try that now!
07:34 jast or after your first commit, doesn't matter :) as long as you do it before pushing
07:35 kellytk error: refname refs/heads/master not found, fatal: Branch rename failed
07:37 kellytk That's after `git init`, using `git branch -m release`
07:37 DevAntoine joined #git
07:38 mischat joined #git
07:40 jrahmy_ joined #git
07:40 rominronin joined #git
07:41 jast kellytk: weird, I could have sworn I've used that in the past. oh well. you can still rename the branch after your first commit, or use this slightly more archaic alternative: git symbolic-ref HEAD refs/heads/newname
07:42 jast (don't do that if you've already committed things, by the way :))
07:42 netj joined #git
07:43 GavinMagnus joined #git
07:43 ToBeCloud joined #git
07:43 djb-irc joined #git
07:43 romerocesar joined #git
07:44 kellytk jast: That worked!
07:45 kellytk Would you mind explaining to me what's going on behind the scenes with a newly initialized Git repo, to update the default primary branch name from master?
07:46 jast it's fairly simple, actually
07:46 kellytk I see that in a new repo, .git/refs/heads is empty BTW
07:47 jast HEAD is a "symbolic reference" to the currently checked out commit branch. that means that, unlike branch references, it can point to a commit *or* another branch
07:47 styx_ joined #git
07:47 jast in a fresh repo it points to 'refs/heads/master', i.e. the master branch, even though it doesn't exist yet
07:47 jast when you make a commit and HEAD points to something that doesn't exist, it is created automatically
07:48 kellytk As defined in .git/HEAD as "ref: refs/heads/master"
07:48 jast exactly
07:49 lindenle joined #git
07:49 eagle-owl joined #git
07:49 multi_io joined #git
07:50 Alienpruts joined #git
07:51 kellytk `git symbolic-ref HEAD refs/heads/release` updated .git/HEAD but there's nothing in refs/heads/.  I add a file, commit, and "-rw-r--r--  1 kellytk  staff   41 Jul  7 00:50 release" is created.  I understand now thanks jast
07:52 kellytk It's an elegant solution to update the default HEAD path and then rely on the commit -> branch automatic creation mechanism
07:54 Balliad joined #git
07:56 hahuang61 joined #git
07:56 stantonk joined #git
07:59 grift joined #git
08:00 eagle-owl left #git
08:00 oich joined #git
08:01 sentriz joined #git
08:01 kurkale6ka joined #git
08:02 circ-user-A26a8 joined #git
08:03 Gonzo89 joined #git
08:03 romerocesar joined #git
08:05 Lipp_ joined #git
08:05 basti_ left #git
08:06 nnyk joined #git
08:07 Kulrak joined #git
08:07 Kulrak joined #git
08:07 marcogmonteiro joined #git
08:07 Andrew_K joined #git
08:08 bongjovi joined #git
08:08 styx_ joined #git
08:08 Lipp_ joined #git
08:12 MattMaker joined #git
08:17 a_thakur joined #git
08:18 _noblegas joined #git
08:19 pur3eval joined #git
08:19 leehambley joined #git
08:20 tannart joined #git
08:22 MattMaker joined #git
08:23 ochorocho__ joined #git
08:23 Lipp_ joined #git
08:27 davisonio joined #git
08:30 cdown joined #git
08:31 MattMaker joined #git
08:31 cdown joined #git
08:32 wrouesnel joined #git
08:32 shinnya joined #git
08:33 subhojit777 joined #git
08:33 pur3eval_ joined #git
08:34 romerocesar joined #git
08:35 stantonk joined #git
08:37 rkazak joined #git
08:37 Beams joined #git
08:38 Literphor joined #git
08:39 Softeisbieger joined #git
08:39 texinwien_ joined #git
08:45 devbug joined #git
08:45 Lipp_ joined #git
08:47 zincrokx joined #git
08:48 clmsy joined #git
08:49 tannart joined #git
08:49 tannart left #git
08:49 ahr3n joined #git
08:53 frog_ joined #git
08:53 dionysus69 joined #git
08:54 frog_ Hi. I accidently "git commit --amend" code to a commit that was already pushed. How do revert that without losing my changes?
08:54 dionysus69 what happened here?? I was working on a branch and then switched to master and I expected all the changes and unstaged files to be left on that branch but all the changes were on master too
08:54 tobiasvl dionysus69: !float
08:54 gitinfo dionysus69: If you have made a change in your working directory and have NOT YET COMMITTED, you may "float" that change over to another (`git checkout oldbranch`) or new (`git checkout -b newbranch`) branch and commit it there.  If the files you changed differ between branches, the checkout will fail.  In that case, `git stash` then checkout, and `git stash apply` and go through normal conflict resolution.
08:54 tobiasvl dionysus69: commit (or stash) if you don't want to carry the changes across
08:55 dionysus69 now I did git reset --hard
08:55 dionysus69 is it irreversible ?
08:55 tobiasvl not likely, since you didn't commit and git didn't know about the changes
08:56 tobiasvl why would you hard reset
08:56 dionysus69 ok :\ I guess
08:56 netj joined #git
08:56 dionysus69 dunno thats the only way I know how to go back to last commit and remove changes
08:56 tobiasvl yeah, but that wasn't what you wanted to do here, was it
08:57 dionysus69 it kind of was on master only though
08:57 mnghqf joined #git
08:57 dionysus69 now it removed changes on other branch also xD
08:57 dionysus69 damn it yesterday lost
08:57 tobiasvl uncommited changes don't belong to any branch
08:57 tobiasvl they belong to the working directory
08:57 dionysus69 :s
08:57 chachasmooth joined #git
08:57 dionysus69 ok thanks I guess next time I will know :)
08:58 tobiasvl hehe. yeah
08:58 tobiasvl in general, --force (or -f) and --hard should only be used if you're sure of what you're doing
09:02 pioiytr joined #git
09:03 liuzhen joined #git
09:03 oich joined #git
09:06 notebox joined #git
09:06 frog_ I found this to undo an accidential "commit --amend" .. git reset --soft HEAD@{1} .. so what does actually HEAD@{1} .. where can I learn more about the HEAD syntax? Looks like I accidentally did "commit --amend" twice, so I need to find the right HEAD to soft reset to.
09:06 troulouliou_div2 joined #git
09:07 i7c git reflog
09:07 FunkyAss joined #git
09:08 Peng joined #git
09:08 selckin !reflog
09:08 gitinfo The git reflog (`git log -g`) records the SHAs of your HEADs for 2+ weeks. `git checkout -b myrestore OLDSHA` and `git reset --hard OLDSHA` will relink to that state via a new and current branch respectively, see http://sethrobertson.github.com/GitFixUm/ for full details.  WARNING: reset --hard will trash any uncommitted changes!  Visualize with: gitk --all --date-order `git log -g --pretty=%H`
09:09 Drzacek joined #git
09:10 Drzacek left #git
09:11 netj joined #git
09:12 phanimahesh joined #git
09:13 frog_ i7c: thx! that helped!
09:14 vuoto joined #git
09:16 jantje joined #git
09:16 mischat joined #git
09:16 Lingo_ joined #git
09:17 GavinMagnus joined #git
09:19 bruce_lee joined #git
09:19 Lingo_ left #git
09:19 lugzer joined #git
09:19 lugzer hello
09:21 lugzer when doing git log, is it possible to retrieve a full path to a commit? i mean a weblink, in case you're using some tool like Stash or similar
09:22 jrabe joined #git
09:23 settermjd joined #git
09:23 tobiasvl lugzer: well git doesn't know the url, but you can add it with git log --format=string:<string>
09:23 lugzer tobiasvl: oh that's a nice tip
09:23 tobiasvl something like --format=string:"http://stash.yourcompany.com/commit=%h" or whatever suits you
09:24 lugzer yes that's perfect - thank you!
09:25 tobiasvl (it can be set in git config too so you don't have to pass it as a parameter to git log all the time)
09:25 Jellyg00se joined #git
09:27 ferr1 joined #git
09:28 texinwien_ joined #git
09:29 lugzer wow, that's just awesome :) it's just that i need to submit a list of commits done in a particular month, and it is required to provide a clickable link
09:30 ramsub07 joined #git
09:35 Andrew_K joined #git
09:36 kedare joined #git
09:36 lugzer i'm trying to figure out if there is a way to emulate something like: git log --since="2016.[thismonth].1" -- do you think would that be doable with --grep ?
09:36 lugzer my goal is to get commits from current month
09:37 tobiasvl why do you need to emulate git log --since? can't you just use that?
09:37 lugzer not emulate
09:38 lugzer i am just lazy, and would like to have a variable for "current month" ;)
09:38 lugzer so that i don't have to do --since="2016.07.01" every time (replacing month number)
09:39 jimi_ joined #git
09:39 lugzer ok i can see now that --grep will only look into log message... that won't cut it
09:39 tobiasvl --since="2016.`date +%m`.01" ? :P
09:39 PCatinean joined #git
09:39 selckin -S
09:40 lugzer oh man, let me try. if that works, i'm buying you a beer
09:42 lugzer yes! :)
09:45 King_Hual joined #git
09:48 cyan__ joined #git
09:48 voxadam joined #git
09:49 Literphor joined #git
09:49 lugzer now, one more thing - i know i can do "git config user.name" to show my git username. how can i use that in --author ? i would simply like to get only my commits without needing to specify the name, using the username from config
09:50 lugzer found it
09:50 lugzer --author="$(git config user.name)" :)
09:51 pur3eval joined #git
09:53 netcarver joined #git
09:53 cqi joined #git
09:56 hahuang61 joined #git
09:57 notebox joined #git
09:57 afuentes joined #git
09:59 notebox joined #git
10:00 zeemz joined #git
10:01 pioiytr joined #git
10:02 pioiytr joined #git
10:03 ahr3n joined #git
10:03 jrabe joined #git
10:04 nettoweb joined #git
10:04 nilg joined #git
10:07 rahtgaz joined #git
10:09 ahmedelgabri joined #git
10:10 muthu joined #git
10:12 texinwien_ joined #git
10:13 dopesong joined #git
10:15 romerocesar joined #git
10:15 hk238 joined #git
10:16 dopesong joined #git
10:16 tarkus_ joined #git
10:18 jaguarmagenta joined #git
10:20 tinajohnson___ joined #git
10:20 ome joined #git
10:21 Darcidride joined #git
10:23 fuchstronaut joined #git
10:24 ochorocho__ joined #git
10:28 pioiytr joined #git
10:28 rnsanchez joined #git
10:30 mehola joined #git
10:31 rnsanchez joined #git
10:32 Chiko joined #git
10:36 eijk_ joined #git
10:36 stantonk joined #git
10:41 dsdeiz joined #git
10:41 dsdeiz joined #git
10:42 fuchstronaut joined #git
10:46 settermjd joined #git
10:47 a3Dman joined #git
10:48 mariuscc joined #git
10:48 fracting joined #git
10:51 fuchstronaut joined #git
10:51 sangy joined #git
10:53 johnmilton joined #git
10:54 lugzer joined #git
10:55 johnmilton joined #git
10:57 jrabe left #git
10:57 jrabe joined #git
10:57 fscala joined #git
10:58 freimatz3 joined #git
10:59 doebi joined #git
11:02 dhruva joined #git
11:06 marcogmonteiro joined #git
11:11 cyan__ joined #git
11:14 notebox joined #git
11:15 ojdo joined #git
11:17 Limes_ joined #git
11:19 Arban joined #git
11:21 ExoUNX joined #git
11:22 HoierM joined #git
11:24 Ordentlig joined #git
11:26 pioiytr joined #git
11:26 jrabe joined #git
11:26 Optiprism joined #git
11:26 Optiprism Hey I'm having some trouble with git svn, I can access the svn via browser using the url https://<url>/folder
11:26 Optiprism But when I try to "git svn clone <usrname>@https://<url>/folder" I get "Bad URL passed to RA layer: Illegal repository URL"
11:27 Optiprism What am I doing wrong?
11:27 Literphor joined #git
11:28 Optiprism Err did I get voice on the channel?
11:28 Peng_ Yes, you're voiced.
11:30 moritz Optiprism: I don't think username@https://... is the correct format
11:30 freimatz3 joined #git
11:30 cdown joined #git
11:31 Optiprism Oh I see, thank you, appearently the username isn't required and git asks for it seperately
11:33 allcentury joined #git
11:35 azbarcea joined #git
11:42 sdothum joined #git
11:43 digidog joined #git
11:44 b1tchcakes joined #git
11:44 vuoto_ joined #git
11:45 fracting joined #git
11:46 _nwkj86 joined #git
11:48 vuoto_ joined #git
11:49 notebox joined #git
11:50 espindola joined #git
11:51 Sceorem joined #git
11:56 jottr joined #git
11:56 leehambley joined #git
11:57 hahuang61 joined #git
11:57 _ikke_ And if you want to provide it, use: http://username@...
11:58 eijk_ joined #git
11:59 synthroid joined #git
12:00 [diablo] joined #git
12:02 tomog999 joined #git
12:03 martin_g joined #git
12:08 Guest83886 joined #git
12:09 pioiytr joined #git
12:09 ag4ve joined #git
12:10 ag4ve does a bare repo keep any type of reflog?
12:11 Hello71 !tias?
12:11 gitinfo Try it and see™. You learn much more by experimentation than by asking without having even tried. If in doubt, make backups before you experiment (see !backup). http://gitolite.com/tias.html may help with git-specific TIAS.
12:11 osse ag4ve: not by default
12:11 ag4ve we ran into an issue yesterday (solved because my coworker kept the commit message) but basically he committed stuff he didn't mean to
12:12 King_Hual joined #git
12:14 ProbabilityMoon joined #git
12:14 ag4ve k, so it can - the remote is ghe - is this something I can enable remotely or is this something in the bare config they'd need to support?
12:14 Hello71 !sensitive?
12:14 gitinfo [!filter_sensitive] You can use filter-branch to remove sensitive data from a repository's history. https://help.github.com/articles/remove-sensitive-data/
12:14 fuchstronaut is there a way to split up the last commit (wich was kind of a draft commit) that was already pushed?
12:15 _ikke_ ag4ve: yes: git config core.logAllRefUpdates true
12:15 osse fuchstronaut: whether it's pushed or not doesn't change how you do it. but it maybe changes whether you want to
12:15 ag4ve ty
12:16 osse _ikke_: is that the default? :O
12:16 fmcgeough joined #git
12:16 _ikke_ For non-bare repos it is
12:16 fuchstronaut osse: It would be nice, because how it is now it is one big messy commit, and I would love to split it up into handsome chunks that can be checked and used step by step
12:17 osse fuchstronaut: git reset HEAD~; git add ...; git commit; git add ...; git commit; ...;
12:17 osse fuchstronaut: then git push -f
12:17 osse !rewrite
12:17 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
12:18 fuchstronaut push -f will give trouble to anyone who already pulled that branch, right?
12:18 _ikke_ correct
12:18 fuchstronaut I see
12:18 fuchstronaut thanks!
12:19 jaguarmagenta joined #git
12:20 Cata|Work joined #git
12:24 leeN joined #git
12:24 ochorocho__ joined #git
12:26 ag4ve hmmm if I delete a branch and rename another to match the old branch's name, would this mess people up or would everyone fetch the new history?
12:26 Meteorhead joined #git
12:26 jim_carry joined #git
12:27 ag4ve (I am thinking of doing this in the next few days and talking about rewriting history made me think I might need more verbal communication about this)
12:28 cdown_ joined #git
12:28 jim_carry I have fethed pull-request and now I want to do a minor change and do git commit --amend. Will pull-request page will be marked as merged by doing it?
12:28 ag4ve I'm still thinking it should be fine since it's just moving head, but...
12:28 _ikke_ ag4ve: They will probably merge in that branch into their current branch
12:29 fracting joined #git
12:29 fahadash joined #git
12:29 ag4ve Mmmm ok, communication than :)
12:31 Literphor joined #git
12:32 mehola joined #git
12:33 nilg joined #git
12:34 alpha joined #git
12:35 liquidjamm joined #git
12:37 wrouesnel2 joined #git
12:38 stantonk joined #git
12:38 sirius joined #git
12:38 Heisenberg1276 joined #git
12:39 relipse joined #git
12:40 a_thakur joined #git
12:41 zumba_addict joined #git
12:42 tabrez joined #git
12:43 rgrinberg joined #git
12:45 tvw joined #git
12:45 daniel_rose joined #git
12:46 BSaboia joined #git
12:47 phanimahesh joined #git
12:49 vuoto joined #git
12:54 sirius joined #git
12:55 MutantMahesh joined #git
12:55 peepsalot joined #git
12:55 JustinHitla joined #git
12:56 JustinHitla so that page https://wireless.wiki.kernel.org/en/users/drivers/ath9k_htc says: "This driver requires firmware. The firmware can be obtained from firmware tree: http://git.kernel.org/?p=linux/kernel/git/firmware/linux-firmware.git", how to obtain it ?
12:57 griffindy joined #git
12:58 a_thakur_ joined #git
12:58 delboy1978uk joined #git
12:59 delboy1978uk how can i checkout a specific file from a specific commit?
12:59 _ikke_ git checkout <commit> fiel
12:59 _ikke_ file
12:59 delboy1978uk ah cool, thanks _ikke_
13:00 a3Dman joined #git
13:00 JustinHitla it's like I never asked the question
13:01 delboy1978uk ask again?
13:01 JustinHitla no
13:01 JustinHitla I will keep my dignity
13:01 JustinHitla not begging
13:03 cd-rum joined #git
13:03 pioiytr joined #git
13:04 jimi_ joined #git
13:06 xpk joined #git
13:07 carif_ joined #git
13:07 diegoviola joined #git
13:08 diegoviola hi
13:08 delboy1978uk joined #git
13:09 fuchstronaut joined #git
13:11 delboy1978uk left #git
13:11 allcentury joined #git
13:12 ejb joined #git
13:15 marscher joined #git
13:18 CEnnis91 joined #git
13:18 ekinmur joined #git
13:18 wodim JustinHitla: click on "tree", then find the firmware file you need, then click on "plain"
13:20 lmat joined #git
13:20 i7c diegoviola, hi.
13:24 ndim joined #git
13:25 rnsanchez_ joined #git
13:26 Rodya_ joined #git
13:28 LiohAu joined #git
13:28 snowkidind joined #git
13:28 LiohAu hi guys
13:28 i7c Hi guy
13:28 LiohAu is there a way when merging 2 branches, to solve conflicts by telling git to use 1 folder coming from one branch ?
13:28 JustinHitla gi huys
13:29 LiohAu i'm merging my develop branch which has updates in folder A into my feature branch which has update in folder B.
13:30 LiohAu When I merge, I have a lot of conflicts on files in A
13:30 LiohAu but since I don't care about my A folder state in my feature branch, I would like to force taking changes from develop for that A folder.
13:30 Rodya__ joined #git
13:30 LiohAu is there a way to do that?
13:31 i7c checkout --theirs?
13:32 oich joined #git
13:32 narendraj9 joined #git
13:33 LiohAu i7c: can I indicate a folder with that command?
13:34 i7c uuuhm, I'd say you can do git checkout --theirs -- path/to/directory
13:34 mehola joined #git
13:34 LiohAu well it is not solving the conflicts
13:35 LiohAu I did a "git merge --squash develop" just before
13:35 LiohAu should I pass an option to the merge ?
13:37 i7c Well
13:37 i7c you run git checkout --theirs -- some/path
13:37 i7c And then you add stuff to the index (git add) AKA staging the changes.
13:37 i7c And then you commit.
13:37 Gsham joined #git
13:38 i7c Git tells you this btw: Automatic merge failed; fix conflicts and then commit the result.
13:39 i7c So you probably forgot about the git add some/path
13:39 madewokherd joined #git
13:44 chitopunk joined #git
13:44 rainbyte joined #git
13:47 tinajohnson___ joined #git
13:49 miklcct joined #git
13:50 garet joined #git
13:50 sangy joined #git
13:50 settermjd joined #git
13:52 lugzer joined #git
13:53 nonconvergent joined #git
13:54 Ordentlig joined #git
13:55 NeverDie joined #git
13:55 Ordentli1 joined #git
13:55 Ordentli2 joined #git
13:56 stantonk joined #git
13:56 NeverDie joined #git
13:57 romerocesar joined #git
13:57 LiohAu i7c: when I run git status after the merge, the files are marked as "both modified"
13:58 LiohAu when I use git checkout --theirs -- Folder/*
13:58 LiohAu and run git status again, nothing change
13:58 hahuang61 joined #git
13:58 courrier joined #git
14:01 GodGinrai joined #git
14:02 AaronMT joined #git
14:02 i7c LiohAu, don't use the *
14:02 i7c Your shell will do funny thigns.
14:03 LiohAu ok, but that does not solve the problem :(
14:03 i7c mhm
14:03 azbarcea joined #git
14:03 LiohAu these files are still in the "unmerged paths" section marked as "both modified"
14:04 i7c Well it will update them in the working directory. You have to git add Folder
14:04 i7c You need to update the index to tell git that you resolved stuff.
14:05 D-Boy joined #git
14:05 cuebix-wk joined #git
14:06 pioiytr joined #git
14:06 Dougie187 joined #git
14:07 dionysus69 joined #git
14:10 mischat_ joined #git
14:10 ResidentBiscuit joined #git
14:10 lmat joined #git
14:11 sdothum joined #git
14:11 nonconvergent joined #git
14:11 mischat_ joined #git
14:11 notebox joined #git
14:12 CheckDavid joined #git
14:12 pioiytr joined #git
14:13 cdown joined #git
14:13 LiohAu i7c: no the merge and checkout did shit I just have unreadable project files now
14:13 lmat joined #git
14:14 lmat joined #git
14:14 djwnoh joined #git
14:15 daniel_r_ joined #git
14:16 hck89 joined #git
14:16 harish joined #git
14:16 LiohAu I may be really dumb to waste 2 hours to merge 2 different versions containing 4 folders.
14:16 nilg joined #git
14:17 ash_workz joined #git
14:17 jungsubk joined #git
14:18 gechr joined #git
14:18 delboy1978uk joined #git
14:18 delboy1978uk how do i delete the last commit from a remote?
14:18 gechr joined #git
14:18 delboy1978uk i reset --hard HEAD^ in my working folder
14:19 gechr joined #git
14:19 oich joined #git
14:19 EvilDMP left #git
14:20 jaguarmagenta joined #git
14:24 Guest80 joined #git
14:24 delboy1978uk got it just had to push -f
14:24 delboy1978uk left #git
14:24 happy-dude joined #git
14:25 ochorocho__ joined #git
14:27 romerocesar joined #git
14:28 Asiajey joined #git
14:28 Asiajey joined #git
14:30 CalimeroTeknik_ joined #git
14:30 AaronMT_ joined #git
14:31 Neo joined #git
14:31 stantonk joined #git
14:31 phanimahesh joined #git
14:31 MZAWeb joined #git
14:32 Darcidride_ joined #git
14:33 jeffreylevesque joined #git
14:33 synthroi_ joined #git
14:33 Literphor joined #git
14:35 mehola joined #git
14:36 JustinHitla left #git
14:36 freimatz3 joined #git
14:38 stantonk joined #git
14:38 nonconve1gent joined #git
14:39 fuchstronaut joined #git
14:39 phanimahesh joined #git
14:41 nonconve1gent left #git
14:43 Neo joined #git
14:43 mdw joined #git
14:43 treaki joined #git
14:44 synthroid joined #git
14:44 ncd|leeN joined #git
14:45 ProbabilityMoon joined #git
14:47 ekinmur joined #git
14:47 pioiytr joined #git
14:49 d10n-work joined #git
14:50 zacts joined #git
14:51 pioiytr joined #git
14:52 Neo joined #git
14:55 subhojit777 joined #git
14:56 lindenle joined #git
14:57 Neo_ joined #git
14:58 romerocesar joined #git
15:01 PCatinean Hey guys, if I have a repository with 4 folders and a fork with just two of them, How do I merge in just changes from the common folders and not add the other ones to the fork as well?
15:02 zivester joined #git
15:03 Seveas you don't. Let me throw an !xy at you :)
15:03 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...
15:03 stantonk joined #git
15:03 clemf joined #git
15:03 _abc_ joined #git
15:04 _abc_ Hello. I am not using git regularly, need to get a git project from github, can't. I can't install git for technical reasons, only have an old version which refuses to work with the site. How does one get a zip or other format of a snapshot (clone) from a site like https://github.com/mogenson/USBTempLogger ? I have no way to use git on this machine
15:04 dsdeiz joined #git
15:04 dsdeiz joined #git
15:05 _abc_ Anyone please?
15:06 ResidentBiscuit There's a Download zip button on the page
15:06 Seveas _abc_: old versions of git should work with github just fine, but you can also download a zip from github
15:06 sangy joined #git
15:08 a_thakur joined #git
15:09 CalimeroTeknik joined #git
15:09 rkazak joined #git
15:10 LeMike joined #git
15:11 rgrinberg joined #git
15:12 texinwien_ joined #git
15:13 SwiftMatt joined #git
15:13 garet joined #git
15:14 Neo_ joined #git
15:15 tokage joined #git
15:15 lugzer joined #git
15:15 Xandaros Hmm, maybe someone here has a tip for me to improve my workflow: I have a fork of a project where I renamed a file and translated it from javascript to Lua. Is there any chance to get git to track changes still?
15:17 mhinz_ left #git
15:17 Noldorin joined #git
15:18 _abc_ So, anyone, how does one pull or clone git projects from github without using git? The download/pull button on the right does nothing in my browser.
15:18 _abc_ On this URL and not only https://github.com/mogenson/USBTempLogger
15:18 Ordentlig joined #git
15:18 jfran joined #git
15:18 Ordentli1 joined #git
15:18 Ordentli2 joined #git
15:19 Neo_ joined #git
15:22 GodGinrai _abc_: turn on your js.
15:22 irco joined #git
15:22 GodGinrai _abc_: the "clone or download" button brings up a dialog which has a "download zip" link
15:25 quetz joined #git
15:25 i7c LiohAu, "unreadable project files"?
15:25 mmmveggies joined #git
15:25 pioiytr joined #git
15:25 i7c If you did what I told you, git can't possibly be the reason for that, except if your other branch had unreadable project files.
15:26 LiohAu i7c: i followed what you did, and basically my .csproj files were containing >>>>HED ... and so on
15:26 stantonk joined #git
15:26 GodGinrai Xandaros: Use `git mv`
15:26 i7c You did git checkout --theirs on the folder and then checked the files?
15:27 PCatinean Seveas, I don't think I understand
15:28 romerocesar joined #git
15:28 PCatinean Basically I have a repository with 5 modules, I work at all of them and push updates/fixes etc continuously. I have sold 3 of the modules to a customer and future updates, which means he should have a separate repo with just 3 modules and their updates
15:28 naxgid joined #git
15:29 eijk_ joined #git
15:29 _abc_ GodGinrai: I turned on js, my browser & this system are ancient. Switched to another machine, got the location of the packages since there the button worked, and got the packages.
15:30 Xandaros GodGinrai: Will it be able to determine the context, though? Since the whole file changed, more or less
15:30 Neo_ joined #git
15:30 _abc_ left #git
15:30 daniel_rose joined #git
15:30 Gonzo89 joined #git
15:30 perlpilot PCatinean: or ... you just need a "build" program that packages those 3 modules just for them and don't give them access to a repo  :)
15:31 PCatinean perlpilot, well I have to give them access to a repo so they can clone and even submit issues, maybe even propose a few fixes if they are technical
15:31 tyreld joined #git
15:31 chrisshattuck joined #git
15:31 GodGinrai Xandaros: Well, you're asking afer the fact.  You should have used `git mv` to rename the file, then made your changes.
15:31 cdown joined #git
15:32 GodGinrai Xandaros: best you can do now is to copy the changed file out of the repo, checkout the old version of the file, `git mv` it to the new filename, copy your file back in, commit
15:32 GodGinrai or maybe do a commit after the mv as well, if you really want to be careful
15:33 Xandaros Alright, I have about 20 such repos and I've only pulled in upstream changes in one of them. Let's see how well it works
15:35 PCatinean so any idea?
15:35 Neo_ joined #git
15:38 robotroll joined #git
15:38 pioiytr joined #git
15:38 perlpilot PCatinean: submodules?
15:39 perlpilot PCatinean: or, if you don't want the hassles of that, perhaps you need to get intimate with subtree merges
15:40 thiago joined #git
15:40 TomyLobo joined #git
15:40 Xandaros GodGinrai: The git mv commands are in my history, so I definitely used that to move them. I probably didn't make an individual commit for them, though
15:40 Xandaros s/them/move
15:41 Xandaros the move, even
15:41 allcentury joined #git
15:41 madewokherd joined #git
15:41 Alienpruts joined #git
15:42 dmto joined #git
15:43 yaewa joined #git
15:47 Xandaros GodGinrai: Nah, that unfortunately doesn't help. All it does is give me the original JS file, but it can't tell me what actually changed. I'm not really surprised, though - this is a hard problem
15:48 unbalancedparen joined #git
15:49 Rodya_ joined #git
15:49 kaldoran joined #git
15:49 Neobenedict joined #git
15:49 Neobenedict joined #git
15:50 GodGinrai Xandaros: http://thisbythem.com/blog/preserving-history-when-renaming-files-in-git/ <-- maybe you just need to query the log in the right way?
15:52 Xandaros That only talks about git log, though. I'm merging in upstream changes - which are in a differently named file, in a different (albeit similar) language
15:52 FuzzySockets joined #git
15:54 GodGinrai Xandaros: why not `git mv` it back to its original name, commit, merge, `git mv` again?
15:54 moei joined #git
15:55 Xandaros ... Yeah, I guess I could do that. To be honest, just looking at the diff and merging it manually is probably easier since the changes are small. I'll keep it in mind if a big merge comes up, though
15:55 lindenle joined #git
15:56 oich joined #git
15:56 xpk_ joined #git
15:58 ekinmur joined #git
15:58 lindenle joined #git
15:58 romerocesar joined #git
15:58 lindenle joined #git
15:59 hahuang61 joined #git
15:59 EvilPenguin joined #git
16:02 stantonk joined #git
16:02 cdg joined #git
16:03 editshare-ericf joined #git
16:04 fuchstronaut joined #git
16:04 HoierM joined #git
16:04 Gsham joined #git
16:04 ochorocho__ joined #git
16:07 ojacobson Xandaros: fundamentally git has no idea about renames, at any point, so git-log is the best you get
16:07 ojacobson (git is purely snapshot-oriented)
16:08 StuartMI joined #git
16:08 Xandaros I am aware. I'm not trying to track changes manually, I'm merging upstream changes. I was just asking if there was an easier way to do it. “No” is a valid answer to that question :P
16:08 al-damiri joined #git
16:12 tematibr joined #git
16:14 aard_ joined #git
16:16 pur3eval joined #git
16:17 catag87 joined #git
16:17 stantonk joined #git
16:18 LionsMane joined #git
16:18 b1tchcakes joined #git
16:18 romerocesar joined #git
16:19 catag87 Hi all. I have a quick question. If I have a branch, release/2.x, and I made a new branch off of that branch, feature/X, added 3 commits and did a PR and had it merged back into release/2.x, and I want to cherry-pick the merge I did into a different branch, release/2.2.1 (just an example), what is the best, cleanest way to do that?
16:20 fuchstronaut joined #git
16:21 jaguarmagenta joined #git
16:21 spudowiar joined #git
16:21 Smirnoff joined #git
16:22 Neo_ joined #git
16:22 BSaboia joined #git
16:23 blackwind_123 joined #git
16:23 dsdeiz joined #git
16:23 dsdeiz joined #git
16:26 Guest80 joined #git
16:27 Neo_ joined #git
16:27 Seveas catag87: the best way would have been to base that feature branch on a common on ancestor of release/2.x and that other branch
16:27 kj_ joined #git
16:28 durham joined #git
16:28 Seveas failing that, just cherry-pick the 3 commits (and not the merge) into your branch.
16:28 catag87 Seveas thanks!
16:29 mdw joined #git
16:30 toogley joined #git
16:30 mada joined #git
16:31 durham_ joined #git
16:31 stantonk joined #git
16:31 notebox joined #git
16:32 mada hi, is there any option to git-show that'll allow me to either exclude an entire directory, or at least select which files' changes to see on some commit?
16:32 jumpman joined #git
16:33 dsdeiz joined #git
16:34 Electrometro joined #git
16:35 Literphor joined #git
16:36 totte joined #git
16:37 mehola joined #git
16:39 devbug joined #git
16:39 dsdeiz joined #git
16:39 dsdeiz joined #git
16:40 The_Phoenix joined #git
16:42 dopesong joined #git
16:42 daniel_rose joined #git
16:46 SteffanW joined #git
16:47 osse mada: git show accepts all the same arguments as git-diff
16:48 romerocesar joined #git
16:49 t0by joined #git
16:51 piling hi guys
16:52 piling can we re-open closed PRs ?
16:52 piling on github specifically
16:52 jokajak joined #git
16:54 mada osse: thank you
16:54 lugzer joined #git
16:55 JanC_ joined #git
16:56 toogley joined #git
16:57 Filystyn joined #git
16:57 Filystyn guys how do i us esomethign like this: git://site/some
16:57 Filystyn ?
16:58 Filystyn anyone ?
16:59 stantonk joined #git
16:59 hahuang61 joined #git
16:59 jast Filystyn: what do you mean by "use"? do you want to clone a git repository and someone gave you a URL like that?
17:00 Brain joined #git
17:01 cdown joined #git
17:02 MacWinner joined #git
17:04 Lvjasha joined #git
17:04 mada Got it, by doing git show <hash> -- . '!:path/file'; thanks osse.
17:04 fracting joined #git
17:05 BSaboia joined #git
17:05 Filystyn clone worked after second tiem gg
17:05 Filystyn thx anyway cu
17:07 dopesong joined #git
17:07 Gsham joined #git
17:09 nnyk joined #git
17:10 shinnya joined #git
17:11 gajus How to separately commit just adding a file to the index?
17:11 gajus and then separately adding content
17:13 nate_c joined #git
17:13 Seveas git add -N
17:16 A5101 joined #git
17:16 Filystyn left #git
17:16 tinajohnson___ joined #git
17:17 deadnull_ joined #git
17:18 jstimm joined #git
17:21 d4rklit3 joined #git
17:22 garet joined #git
17:27 stantonk joined #git
17:27 acetakwas joined #git
17:28 Didac joined #git
17:28 acetakwas joined #git
17:28 sanct joined #git
17:28 circ-user-A26a8 joined #git
17:31 Lvjasha joined #git
17:33 ekinmur joined #git
17:33 thiago joined #git
17:33 tuxlovesyou joined #git
17:33 slayerjain joined #git
17:34 cr joined #git
17:34 SwiftMatt joined #git
17:35 daniel_rose joined #git
17:36 cr left #git
17:36 ascarter joined #git
17:36 mada left #git
17:38 mehola joined #git
17:38 sanct Hi, I'm trying to merge two branches, and a merge conflict came up (http://imgur.com/IEVcFbo).  How is this a merge conflict if there is nothing in the HEAD section?
17:38 mdw_ joined #git
17:39 spudowiar sanct: I think one branch added stuff and one branch removed stuff there
17:41 mdw joined #git
17:41 sanct spudowiar, sorry, meant rebase.
17:42 dreiss joined #git
17:42 daniel_rose joined #git
17:43 sanct would it still be the same cause if it was a rebase instead?
17:43 GavinMagnus left #git
17:44 stantonk_ joined #git
17:44 foist joined #git
17:44 dsdeiz joined #git
17:47 spudowiar idk
17:47 spudowiar probably
17:48 SimonNa joined #git
17:49 TooLmaN joined #git
17:49 allcentury joined #git
17:49 lugzer joined #git
17:51 akushner joined #git
17:52 dopesong joined #git
17:53 thiago sanct: only you can tell. You have to try to see if it would conflict.
17:54 acetakwas joined #git
17:54 sanct thiago, the image above is the result of the rebase
17:54 shaggycat joined #git
17:56 thiago btw, please don't use images for text
17:57 Sasazuka joined #git
17:57 pioiytr joined #git
17:58 nettoweb joined #git
17:58 pioiytr joined #git
17:58 sanct sorry, first time here
17:59 SwiftMatt joined #git
17:59 romerocesar joined #git
18:00 n_blownapart joined #git
18:01 stantonk joined #git
18:03 n_blownapart hello please help me undestand grey folders. I've read about submodules but cannot revive these directories. the problem seems to be happening without any changes in my (ignorant) use of github.
18:03 rbern joined #git
18:04 durham joined #git
18:05 Geo joined #git
18:05 thiago git has no concept of colours
18:05 thiago you're asking a github question. Please see #github or their online support and FAQ.
18:05 Geo When using git log, is there a way to only show commits up to the previous tag?
18:05 thiago personally, I've often wondered what they are, but never needed to find out
18:05 thiago Geo: git log that_tag
18:05 Geo I'm looking at man gitrevisions but can't seem to get the right syntax
18:05 gitinfo the gitrevisions manpage is available at http://jk.gs/gitrevisions.html
18:05 n_blownapart thiago well isn't that swell
18:06 thiago Geo: where "up to" is "from the beginning of time up to"
18:06 daniel_rose joined #git
18:07 Geo thiago: yeah, heh, thats backwards of what I'm trying to do
18:07 Geo I'm trying to gen a Changelog for this version
18:07 thiago Geo: so you want "from now backwards to" that tag?
18:07 thiago Geo: git log that_tag..
18:07 thiago including the ..
18:07 Geo from the current commit, back to the previous tag
18:07 Geo perfect!
18:08 Geo thank you
18:08 thiago you may want to look at the git shortlog command
18:08 Geo yep, have that already
18:08 Geo and --decorate! :)
18:08 thiago shortlog has no such option
18:09 netj joined #git
18:09 Darcidride joined #git
18:09 Geo oh
18:09 Geo sorry, i was thinking oneline
18:09 Geo shortlog is helpful too, thanks
18:09 Geo didnt know that existsed
18:10 dionysus69 joined #git
18:10 Alienpruts joined #git
18:11 satifant joined #git
18:11 Geo wow, that really is helpful!
18:12 Geo You just saved me a lot of time, thank you
18:12 Geo Who knew people were still helpful on IRC?!?! ;)
18:12 preaction ... everyone?
18:12 ekinmur joined #git
18:13 refried_ joined #git
18:13 MTecknology joined #git
18:13 MTecknology I miss you you so much git... :(
18:13 MTecknology I feel like hg is the land of randomly deleted files
18:13 stantonk joined #git
18:14 monokrome joined #git
18:17 ToxicFrog MTecknology: could be worse, you could be using SVN
18:18 rgrinberg joined #git
18:19 crose joined #git
18:20 Seveas MTecknology: try git-cinnabar to get the git experience with hg repos :)
18:21 stantonk joined #git
18:21 fuchstronaut joined #git
18:21 elastix joined #git
18:21 Geo Does anyone know where I can find the mailmap.file ?
18:21 jaguarmagenta joined #git
18:22 MTecknology Seveas: 1) Howdy! 2) Awesome! :D
18:22 DolphinDream how to find out what common files two commits touch ?
18:25 perlpilot DolphinDream: I can think of a complicated way to do it.  Nothing concise though.
18:27 Seveas DolphinDream: comm -12 <(git show --name-status commit1 | sed -e 's/^[AD]/M/') <(git show --name-status commit2 | sed -e 's/^[AD]/M/')
18:28 Seveas Geo: that question doesn't parse in my head. mailmap.file is a configuration variable, you find it in your git config :)
18:28 perlpilot DolphinDream: mine is: git show --pretty="" --name-only commit1 commit2  | sort | uniq -c | awk '$1 > 1 { print $2 }'
18:29 stantonk joined #git
18:29 Seveas perlpilot: ooh, name-only is better than my seddery
18:29 Seveas DolphinDream: comm -12 <(git show --name-only commit1) <(git show --name-only commit2)
18:29 nedbat if I have "ci" aliased to commit, then "git ci --help" just shows me `git ci' is aliased to `commit'.  Is there a way to have "git ci --something" show me the man page for commit?
18:30 eycsigfy joined #git
18:30 romerocesar joined #git
18:30 perlpilot Seveas: mine is also slightly better if you need to find the common files between N > 2 commits  ;)
18:31 ramsub07 joined #git
18:31 Seveas nedbat: nope.
18:31 TooLmaN joined #git
18:31 nedbat :(
18:32 DolphinDream nice
18:32 ramsub07 hello, while squashing one commit into the other, which one should be squash and which one should be pick ? Suppose there are 2 commits A and B. A precedes B. I want B to get merged into A and be one single commit. What should i use against A ?
18:32 Geo Seveas - great, thank you
18:32 DolphinDream thx guys .. the one with the pretties formula wins
18:33 DolphinDream i like Seveas'
18:33 Seveas nedbat: builtin/help.c:479
18:34 DolphinDream though i get a comm: file 1 is not in sorted order error
18:34 DolphinDream and for file
18:34 DolphinDream file2
18:34 Seveas ramsub07: pick A, squash B
18:34 perlpilot DolphinDream: yep, I was just about to mention that:)
18:34 aep-shoutlet joined #git
18:34 perlpilot DolphinDream: comm assumes the files are sorted, so you'll have to add | sort to each of the args to comm
18:35 gajus Seveas Thank you
18:35 perlpilot comm -12 <(git show --name-only commit1 | sort) <(git show --name-only commit2 | sort)
18:35 Seveas perlpilot: neh, gotta suppress the commit message instead
18:35 perlpilot (in case I wasn't clear enough)
18:35 Seveas git show --name-only --pretty=''
18:35 perlpilot Seveas: that too I guess.   Does "git show" always output the files in sorted order?  I'm still not sure.
18:35 DolphinDream was clear
18:35 Seveas it does
18:36 Seveas comm -12 <(git show --name-only --pretty='' commit1) <(git show --name-only --pretty='' commit2)
18:36 perlpilot Seveas++
18:36 Rodya_ joined #git
18:36 pso joined #git
18:36 DolphinDream i get a lot of empty lines though :) not a biggy but not the cleanest way i guess
18:36 voxadam joined #git
18:37 DolphinDream ah.. ok.. with pretty you dont' get empty lines..
18:37 DolphinDream Seveas, you still need to add |sort
18:38 Seveas then comm and git disagree about sorting order :)
18:39 perlpilot surely setting LC_COLLATE would make them agree?
18:39 mehola joined #git
18:40 Seveas setting LC_COLLATE=C works
18:40 Seveas hurricane:~/code/python-hpilo (master=)$ comm -12 <(git show --name-only --pretty='' HEAD) <(git show --name-only --pretty='' HEAD^)
18:40 Seveas hpilo.py
18:40 Seveas and that is correct :)
18:41 perlpilot The more we talk about this, the more I like my version  ;-)
18:41 AndroidLoverInSF joined #git
18:42 perlpilot Though, mainly I think I like it because it's more generalizable:  git show --pretty="" --name-only commit1 commit2 commit3 | sort | uniq -c | awk '$1 == 3 { print $2 }'  # for instance
18:42 Seveas perlpilot: I don't. Your use of sort | uniq | awk is abysmal. Can all be done in a single awk script :P
18:43 perlpilot Seveas: sure!  I contemplated doing it that way, but then I already *knew* how to do it this way  :)
18:43 perlpilot Seveas: and I think you mean, "... in a single perl script"   ;)
18:44 lugzer joined #git
18:44 Seveas no, I meant awk :)
18:46 stantonk joined #git
18:46 Heisenberg1276 joined #git
18:46 eycsigfy joined #git
18:47 pur3eval joined #git
18:47 perlpilot Here's a Perl 6 version that I didn't think too much about:  git show --pretty="" --name-only commit1 commit2 | perl6 -e 'lines.Bag.grep(*.value == 2).map(*.key)>>.say'
18:48 perlpilot that grep/map business could probably be simplified.
18:51 sangy joined #git
18:51 javajoe joined #git
18:54 settermjd joined #git
18:55 GodGinrai joined #git
18:56 stantonk joined #git
18:56 dreiss joined #git
18:56 circ-user-A26a8 joined #git
18:59 pur3eval joined #git
18:59 garet joined #git
19:00 romerocesar joined #git
19:00 hahuang61 joined #git
19:04 gajus How to commit only file rename without commit the contents changes
19:04 gajus Seveas I have tried git add -N both files
19:04 Geo I'm looking through the pretty= options; is there a way to display files modified in a commit?
19:04 gajus but it does not recognize the file rename
19:05 perlpilot Geo: git show --pretty='' --name-only COMMITISH
19:06 perlpilot Geo: if you want the other stuff too, just omit the "--pretty=''" pary
19:06 perlpilot er, part
19:06 TooLmaN joined #git
19:06 Geo cool - and bonus points, how about from inside the format:<options> ?
19:06 bunniez joined #git
19:06 Geo I don't see a % option there
19:07 perlpilot Why does it need to be done with a %-specifier?
19:07 leehambley joined #git
19:08 Geo That's how it's done inside the format option
19:08 stantonk joined #git
19:08 Geo --pretty=format:bla bla %aN etc etc
19:09 Geo but I certainly don't know if there's a better/different way
19:09 perlpilot I'm asking why you need it to be in the format in the first place
19:09 WayToDoor joined #git
19:10 shgysk8zer0_ joined #git
19:10 perlpilot Geo: either way, you may need to do 2 separate calls to use the format and get a list of files changed.
19:10 bunniez left #git
19:11 Geo perlpilot: because I'm creating a custom msg in format
19:12 kaldoran joined #git
19:12 Geo git log --pretty=format:"This is my custom msg, by %a, then %b, etc"
19:13 Geo so i don't think 2 calls will work there, unfortunately
19:14 tematibr joined #git
19:14 perlpilot no, not without you manually replicating some of the logic.
19:14 WayToDoor joined #git
19:14 nnyk joined #git
19:14 bjoe2k4 joined #git
19:14 sanketdg joined #git
19:17 voxadam joined #git
19:17 Ordentlig joined #git
19:18 stantonk joined #git
19:18 Ordentli1 joined #git
19:18 lugzer joined #git
19:18 Ordentli2 joined #git
19:19 vuoto joined #git
19:19 WayToDoor joined #git
19:19 Literphor joined #git
19:19 perlpilot Geo: Here's something *extremely* ugly:  git log --pretty=format:"Custom %cd: %H" | perl -pe 's/([a-z0-9]+)$/join " ", split "\n", qx(git show --pretty='' --name-only $1)/e;'
19:19 MZAWeb_ joined #git
19:20 perlpilot Geo:  Maybe someone else has something better.
19:20 mischat joined #git
19:21 MattMaker joined #git
19:21 Jason__ joined #git
19:22 Gsham joined #git
19:23 Geo heh, thanks!
19:28 Ordentlig joined #git
19:28 Ordentli1 joined #git
19:28 Ordentli2 joined #git
19:29 durham_ joined #git
19:29 cdown joined #git
19:29 Literphor joined #git
19:33 SimonNa joined #git
19:36 f3r70rr35f joined #git
19:36 nettoweb joined #git
19:36 catag87 joined #git
19:37 dsdeiz joined #git
19:37 catag87 joined #git
19:37 Kobaz what'
19:37 Lvjasha joined #git
19:37 Kobaz what's the best way to 'bring a file up to date'
19:37 Kobaz say i'm behind origin/master by like 200 someodd commits
19:38 Kobaz and i want, 10 of those commits, to merge in the latest and greatest of a specific file
19:38 Kobaz i have all the git revs of the commits i want, and i try and cherry pick and i have nothing but conflicts
19:38 moritz cherry-pick those commits
19:38 Kobaz and i fix the conflicts, and then cherry pick the next one, and i have the *same* conflicts
19:38 moritz or if you really want just one file from the state of that branch, git checkout origin/master path/to/file
19:38 moritz and then commit
19:39 espindola joined #git
19:39 Kobaz so i wind up giving up, and just scping the file over everything
19:39 Kobaz from an updated repo
19:39 moritz that's what the git checkout thingy does
19:39 moritz except you don't need separate copy of the repo that way
19:40 Kobaz yeah
19:40 mehola joined #git
19:40 Kobaz so
19:40 Kobaz how would i kind of 'fast forward' just that file
19:40 Kobaz so i have a git checkout of that file then...
19:40 Kobaz and now it's showing up as a local change
19:40 Kobaz it's like 10 commits rolled into one patch basically
19:40 romerocesar joined #git
19:41 Kobaz if i commit that, then i'll get conflicts when i git pull sometime in the future to get everything up to date
19:41 _ikke_ Kobaz: Unlike some other SCM's, git's history is not file based
19:41 _ikke_ But a snapshot of the whole tree
19:41 Kobaz yeah i know it's not file based
19:41 Kobaz but my changes tend to be file based
19:41 Kobaz so one git commit tends to be one file
19:41 Kobaz different modules and whatnot
19:42 Kobaz so each cherry pick is generally one set of changes to one file
19:42 Kobaz i usually dont have an issue doing one or two cherry picks
19:42 Ordentlig joined #git
19:42 Kobaz but if i have to do 10, i wind up with a conflict-loop
19:42 Ordentli1 joined #git
19:42 Kobaz i keep getting the same conflicts over and over and it's a waste of time
19:42 _ikke_ man git rerere
19:42 gitinfo the git-rerere manpage is available at http://jk.gs/git-rerere.html
19:42 Kobaz i wind up doing git reset
19:42 Kobaz and then scp the file over manually
19:43 Kobaz and just keep a mental note to git checkout that file before i do the next pull
19:44 ojacobson Should you be using something that -is- file-oriented, here? It sounds like Git may be the wrong tool for the (unstated, sadly) needs you have
19:45 Kobaz needs are tracking basic app development
19:45 Kobaz and supporting distributed repos (local commits)
19:46 Kobaz i really dont see why i can't just cherry-pick a series of revs
19:46 Kobaz i dont understand why there would be conflicts at all
19:47 Kobaz especially in this case i have now, where the local branch has not diverged
19:47 Kobaz it's just behind...
19:47 ojacobson if you find yourself cherry-picking regularly, you're probably following a workflow git's a poor fit for
19:47 ojacobson git inherently assumes that your workflow's primitive step is "take a snapshot, make a change to the whole project, take another snapshot"
19:47 ojacobson the versionining primitive is snapshots of the whole project
19:47 pioiytr_ joined #git
19:47 fracting joined #git
19:48 Literphor joined #git
19:48 _ikke_ Kobaz: applying commit C without commit B can cause a commit while applying both would not
19:48 _ikke_ cause a conflict*
19:49 Kobaz but that's the thing
19:49 Kobaz i'm applying them in reverse historical order
19:49 Kobaz if i generated patch files in the same order, i would not have conflicts
19:49 Kobaz but yet, when i do it with cherry pick, i do
19:49 _ikke_ Kobaz: That's odd, because that's basically what cherry-pick does
19:49 _ikke_ generate a patch, apply it, and commit it
19:50 Kobaz obviously oldest first, and not skipping any, would be the way to avoid conflicts
19:50 ojacobson applying a patch sequence backwards seems like a recipe for tons of conflicts
19:50 ojacobson unless the patches are 100% independent of one another, with no overlapping chunks at all
19:50 Kobaz overlapping chunks?
19:50 daniel_rose joined #git
19:50 ojacobson I have no reason to believe that'd be true, since it rarely has been in my experience :)
19:51 ojacobson A patch is a sequence of chunks. Each chunk is a filename, a series of context lines, a series of removed lines, a series of added lines, and a series of trailing context lines.
19:51 _ikke_ Kobaz: One patch changing the same lines as another patch
19:51 ojacobson Two chunks overlap if the regions they apply to contain any of the same lines
19:52 Kobaz sure, plenty of changes overlap
19:52 Kobaz but applied in the right order, should apply fine
19:52 _ikke_ Kobaz: I think you saying reverse chronological order is confusing us
19:52 Kobaz lines 10-20 might be changed in patch 1, and lines 5-8 might be changed in patch 2
19:52 _ikke_ we would assume you applied the newest first
19:52 Kobaz why should it conflict though
19:52 vuoto joined #git
19:52 Kobaz why would i apply the newest first
19:52 Kobaz you need to apply the old patches before the new patches
19:52 _ikke_ Kobaz: That's what reverse chronological means
19:53 Kobaz [07 07 15:50] <Kobaz> obviously oldest first, and not skipping any, would be the way to avoid conflicts
19:53 ojacobson yes
19:53 _ikke_ yes, I was just trying to point out why it caused confusion earlier
19:53 Kobaz okay
19:54 _ikke_ Have you analyzed what it's conflicting on?
19:54 Kobaz a little bit
19:54 Kobaz a small bit of code in the beginning of the file
19:55 Kobaz er, i meant  , lines 5-18 changed in patch 2... rather
19:55 Kobaz but i dont see any pattern really to the conflicts
19:56 Kobaz it'
19:56 Kobaz s just all over the place
19:56 _ikke_ Not by chance any line-ending differences that might cause it?
19:56 Kobaz and one single cherry pick tends to work just fine
19:56 Kobaz nope, same system types everywhere, same line endings
19:57 Kobaz multiple cherry picks always give me problems
19:57 Kobaz when applied in the right order
19:57 Kobaz i'll have to build some test cases at some point and upload somewhere
19:57 Literphor joined #git
19:58 Kobaz i thought maybe there was some known issues about cherry picks i wasn't aware of
19:58 _ikke_ No, not that I'm aware of
19:58 Kobaz this sounds like, not a quick fix
19:58 Ordentli2 joined #git
19:58 _ikke_ Perhaps, perhaps not, but it's hard to tell without seeing what the conflicts are
19:59 Geo can anyone run this and confirm that the first entry is buggered, for some reason? git log --no-merges --pretty=format:"%w(75)%H %n Committed %ai by %an %n %s %n %b" <reference>..
19:59 _ikke_ A cherry-pick is basically this: git diff <commit>~1 <commit> | git apply - && git commit -C <commit>
19:59 Geo The first two entries get smooshed together, but all other entries look fine
20:01 romerocesar joined #git
20:01 _ikke_ Smooshed together how?
20:01 jstimm joined #git
20:02 Geo I get the body line of the 1st commit w/o a blank line between it and the next commit hash- but *every* other body line has a space
20:02 Geo s/space/empty line after it
20:03 Geo But yours looks ok? All entries look uniform?
20:03 Geo I'm wondering if its something odd with that particular entry
20:03 Kobaz here's the weird thing
20:03 Kobaz _ikke_: i fix the conflicts
20:03 Kobaz and then i do git add, git diff --cached
20:03 Kobaz and there's no changes
20:04 _ikke_ Geo: Yeah, they look the same
20:04 _ikke_ Geo: Trying it on git.git
20:04 bjoe2k4_ joined #git
20:04 _ikke_ Kobaz: Do you happen to have a .gitattributes file
20:04 _ikke_ ?
20:04 Kobaz where... in the repo? or ~
20:05 Geo ok, thanks- i just confirmed it is that particular commit msg, for some reason
20:05 Geo added another commit, and it looks fine
20:05 Geo *shrug*
20:05 Geo sorry Changelog fans
20:05 _ikke_ Kobaz: repo
20:05 Kobaz nope
20:05 Kobaz http://pastebin.com/mhG1wWHv
20:05 Kobaz so there's the conflict
20:05 Kobaz that i get over and over and over
20:05 Kobaz on every cherry pick
20:05 Kobaz i fix the conflict, and there's nothing to commit
20:06 javajoe joined #git
20:06 _ikke_ Kobaz: What might give a bit more clarity is setting git config merge.conflictStyle diff3
20:07 Kobaz l
20:07 Kobaz k
20:07 pioiytr_ joined #git
20:08 stantonk joined #git
20:08 Kobaz http://pastebin.com/NwLce8yX
20:08 acetakwas joined #git
20:08 _ikke_ It looks like a conflict has been comitted. Is that possible?
20:08 _ikke_ Do you see the double ++
20:09 _ikke_ ?
20:09 Kobaz that's my git diff
20:09 Kobaz yeah i see the double +
20:09 _ikke_ Ah, git diff of a conflict
20:09 _ikke_ that's very confusing :P
20:09 Kobaz hehe
20:09 Kobaz okay lemme cat the file
20:10 mdw joined #git
20:10 Kobaz http://pastebin.com/sT5mZ5tn
20:10 KnightRider00 joined #git
20:11 Kobaz so how would i force that conflict to kind of go away
20:11 Kobaz and not try and merge in for every cherry pick
20:11 moritz Kobaz: so, why do you want to cherry-pick, if checking out the file from the other branch is that much easier?
20:12 nnyk joined #git
20:12 _ikke_ Looks like a whitespace conflict or somethign
20:12 lugzer joined #git
20:13 _ikke_ What stands out to me is that there are no merge-base changes
20:13 _ikke_ Between ||||||| and =======
20:14 todd_dsm_wk joined #git
20:14 DolphinDream Seveas, list common files part two :) how do you list the common files touched by two branches/refs (not just two commits)
20:14 thoraxe joined #git
20:15 garet joined #git
20:16 Geo left #git
20:16 _ikke_ DolphinDream: What is the relation between these two refs/
20:16 jayjo joined #git
20:16 _ikke_ To know what a specific branch touches, you need to know where it starts, and the only way is to know the ref the branch was created on
20:17 IWishIKnew joined #git
20:17 jayjo I have a repo that I want to duplicate to a completely new directory and start at a sepcific commit? Is this possible? If I checkout a new branch it changes my source code (which is the intention of switching branches), but I need to keep this piece of software as it is
20:17 kaldoran joined #git
20:17 Kobaz moritz: i want to keep the history
20:17 _ikke_ jayjo: man git worktree
20:17 gitinfo jayjo: the git-worktree manpage is available at http://jk.gs/git-worktree.html
20:18 KevinMGranger left #git
20:18 _ikke_ jayjo: perhaps this is what you are looking for
20:18 Kobaz moritz: i want to do basically a 'git pull on the file'
20:18 osse Kobaz: stop wanting that. git can't help you :(
20:19 Kobaz but i should be able to cherry-pick my way to that end.. shouldn't i?
20:19 Kobaz if my particular cherry-picks are all on one file
20:19 stantonk joined #git
20:19 azerus joined #git
20:19 _ikke_ Kobaz: yes
20:20 Kobaz so how do i cherry pick the next rev on my list
20:20 _ikke_ osse: Would it help if he says, hey, I get these strane conflicts when cherry-picking a bunch of commits where I don't expecy any
20:20 Kobaz without getting the same exact conflict, over and over
20:20 MattMaker joined #git
20:20 nwkj86 joined #git
20:21 osse Kobaz: enable rerere
20:21 osse Kobaz: run git rerere; then try again
20:22 _ikke_ osse: That's symptom patching
20:22 Kobaz i rerere'd
20:22 fuchstronaut joined #git
20:22 Kobaz same conflict afterwards, on a git cherry pick
20:22 _ikke_ I would not expect a sequence of commits to cause the same conflicts over and over
20:22 osse Kobaz: you have to fix them once after rerere, then never again
20:22 Kobaz k
20:23 osse (hopefully)
20:23 nidr0x joined #git
20:25 Kobaz well allrightey
20:25 Kobaz that fixed it
20:26 Topic for #git is now Welcome to #git, the place for git help and hard knocks | Public logs at http://goo.gl/BuUi5o | Current stable version: 2.9.0 | First visit? Read: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | A branch, a tag, and a reflog walk into a bar. The bartender says, "What is this, some sort of rebase?"
20:26 javajoe joined #git
20:26 nidr0x joined #git
20:27 wodim Kobaz: are you the genuine kobaz?
20:27 Kobaz yes
20:28 Kobaz [07 07 16:28] -NickServ- Information on kobaz (account kobaz):   [07 07 16:28] -NickServ- Registered : Mar 21 02:10:44 2001 (15y 16w 0d ago)
20:28 wodim gud
20:28 cwar joined #git
20:28 Arban joined #git
20:29 Kobaz what be up
20:29 _ikke_ Do you have non-genuine Kobaz's?
20:29 Kobaz sometimes there's imposters
20:30 mmmveggies joined #git
20:30 oich i have files under version control that are not mine and that I can't/shouldn't change. One file is changed by building (a timestamp is changed and nothing else). Is there a way to configure git to ignore change to a file that is under version control?
20:30 DolphinDream _ikke_ .. two diverged branches.. one touches some files in a series of commits.. the other touches some files in a series of commits.. i want to know what is the common set of files modified by either branch relative to their common ancestor.
20:31 osse Kobaz: should I have heard about you? :O
20:31 osse are you meat loaf?
20:31 Kobaz only on weekends
20:31 Kobaz osse: i dunno
20:31 Kobaz osse: i wrote this: https://sourceforge.net/projects/kobazscripts/
20:31 Kobaz it was quite popular back in the heyday
20:32 DolphinDream ABC .. A and C are branches off of the common ancestor B .. A touches N files.. C touches M files.. both A and C must have a set of common files that both modified.
20:33 moritz you can do a git diff --name-only B A and git diff --name-only B C
20:33 GavinMagnus joined #git
20:34 moritz and then use some command line magic to do the intersection
20:34 DolphinDream if i don't know B would i have to derived that via merge-base or something ?
20:34 osse Kobaz: will any of these scripts tell me what meat loaf won't do for love?
20:34 moritz DolphinDream: yes
20:34 Kobaz quite possibly
20:35 moritz DolphinDream: ah, comm -12 seems to be the command line I was looking for
20:35 DolphinDream moritz, right on. i was using this for comparing files touched by two commits
20:35 MattMaker joined #git
20:35 moritz comm -12 $(git diff --name-only B A|sort) $(git diff --name-only B C|sort)
20:35 moritz not sure if the sort is neccessary
20:35 DolphinDream it is
20:36 troulouliou_div2 joined #git
20:36 moritz the output of git diff --name-only looks pre-sorted to me
20:36 DolphinDream i'll have to add B = git merge-base A C first
20:36 Robody joined #git
20:36 _ikke_ Using merge-base will probably not help, because that would contain commits that are not considered part of C but B for example
20:36 DolphinDream i tried it without sort and gives error at my end
20:37 moritz ok
20:38 DolphinDream _ikke_,  so i'd have to get the commits immediately after B on the A branch (A') and C branch (C') ? and then use the comm -12 $(git diff --name-only A' A | sort) $(git diff --name-only C' C | sort) ?
20:39 _ikke_ Easiest if you know B
20:39 _ikke_ git diff --name-only B...A
20:39 DolphinDream does gi diff --name-only B A also include the files touched by B itself ?
20:39 stantonk joined #git
20:39 spudowiar joined #git
20:40 _ikke_ Not with the 3 dots
20:40 DolphinDream ah. neat
20:40 _ikke_ (it uses the merge base between B and A
20:40 _ikke_ )
20:40 pioiytr_ joined #git
20:41 DolphinDream so to exclude B from both branch diffs: comm -12 $(git diff --name-only B...A | sort) $(git diff --name-only B...C | sort)
20:41 _ikke_ correct
20:41 DolphinDream but you say that finding out B via merge-base A C will not work ?
20:41 _ikke_ Typically, the merge-base between A and B is different from B and C
20:42 pioiytr_ joined #git
20:42 _ikke_ So one of the two will contains commits belonging to B, not to the branch
20:42 DolphinDream no. i mean the common ancestor of A and C .. B = git merge-base A C
20:42 _ikke_ B is usually not an ancester of either
20:43 pioiytr_ joined #git
20:43 _ikke_ If what you want is the difference between the common ancester, then that will work of course
20:43 pur3eval joined #git
20:46 MattMaker joined #git
20:47 pur3eval_ joined #git
20:47 javajoe joined #git
20:47 DolphinDream _ikke_, http://pastebin.com/ApL0Svz1
20:49 _ikke_ Ok, they both start from the same commit B
20:49 durham joined #git
20:49 _ikke_ then git diff --name-only C...A and git diff --name-only A...C should work
20:50 ramsub08 joined #git
20:51 stantonk_ joined #git
20:52 FuzzySockets joined #git
20:53 Literphor joined #git
20:54 jaggz joined #git
20:54 oich if I update a submodule to use it's new head then checkout a different branch that includes the submodule git status indicates that the submodlulde has changed. How can I get the branch to use the commit that the submodule had when it was added to the branch?
20:55 oich sync and update --remote do not seem to do it
20:55 yqt joined #git
20:55 stantonk joined #git
20:56 zeemz joined #git
20:57 jaggz how do you contribute to a github project?
20:57 jaggz they say we submit pull requests
20:58 GyrosGeier jaggz, same as with any other project: you publish your changes and ask the other person to merge them
20:58 GyrosGeier github has just formalized that in a tracker
20:58 GyrosGeier so you can say "my changes are published as this branch"
20:59 Neobenedict_ joined #git
20:59 Neobenedict_ joined #git
21:00 Gsham joined #git
21:00 jaggz how do I see all the branches to see their branch naming conventions?
21:00 FunkyAss i find it amusing that git has a manpage documenting different workflows
21:00 jaggz FunkyAss, what's that manpage?
21:00 jaggz gitworkflows.. cool
21:00 FunkyAss man gitworkflows #on my system
21:00 gitinfo the gitworkflows manpage is available at http://jk.gs/gitworkflows.html
21:00 jaggz yuck
21:01 MattMaker joined #git
21:01 hahuang61 joined #git
21:01 romerocesar joined #git
21:02 Literphor joined #git
21:03 jaggz GyrosGeier, I've worked on only a few public projects over the years, btw
21:03 kulelu88 joined #git
21:03 kulelu88 joined #git
21:03 nnyk joined #git
21:03 jaggz like with blender I submitted my own patches directly to someone who worked on that area of the software and they put the changes in for me
21:04 jaggz but nodemcu has a CONTRIBUTING.md I'm reading now
21:04 madewokherd joined #git
21:05 rgrinberg joined #git
21:05 EvilDMP joined #git
21:06 zacts joined #git
21:06 ochorocho__ joined #git
21:07 acetakwas joined #git
21:08 cdown_ joined #git
21:08 lugzer joined #git
21:08 elastix joined #git
21:10 IWishIKnew_ joined #git
21:10 Lvjasha joined #git
21:10 stantonk joined #git
21:11 romerocesar joined #git
21:15 ahr3n joined #git
21:16 Literphor joined #git
21:16 MattMaker joined #git
21:18 phantomtiger joined #git
21:18 jantje hi! what do I need to do to ensure the post-receive hooks fully execute? I already do  trap '' PIPE in the post-receive script
21:19 ramsub07 joined #git
21:19 mischat joined #git
21:19 MattMaker joined #git
21:19 qsx cat > /dev/null
21:21 paws_ joined #git
21:21 romerocesar joined #git
21:24 kpease joined #git
21:25 robotroll joined #git
21:25 IWishIKnew joined #git
21:25 jantje qsx: I don't think the hook will do much then? :-)
21:26 stantonk joined #git
21:26 qsx jantje: well it'll continue to run.
21:26 qsx and not die of sigpipe
21:26 qsx !hook_p
21:26 gitinfo [!hook_pitfalls] Guidelines for writing hooks: 1. Consume all input (cat >/dev/null if you don't want it). 2. If you use any 'cd', also 'unset GIT_DIR'. 3. Don't git-pull in a hook (or any other script).
21:26 papna left #git
21:26 GyrosGeier jaggz, branch naming conventions don't matter
21:27 GyrosGeier the branch name gets lost in the merge
21:27 jantje basically I do  trap '' PIPE; IN=$(cat /dev/stdin) ; <do lots of other stuff>
21:27 pur3eval joined #git
21:28 pur3eval hello
21:29 Sasazuka_ joined #git
21:29 srcerer joined #git
21:31 davisonio joined #git
21:32 stantonk joined #git
21:33 vuoto joined #git
21:34 bruce_lee joined #git
21:34 pioiytr_ joined #git
21:35 CheckDavid joined #git
21:36 m0viefreak joined #git
21:37 garet joined #git
21:39 chitopunk joined #git
21:40 stantonk joined #git
21:40 chitopunk joined #git
21:41 mehola joined #git
21:43 stantonk_ joined #git
21:44 mmmveggies left #git
21:44 tokage joined #git
21:46 jimi_ joined #git
21:47 pioiytr_ joined #git
21:49 jantje qsx: is there any chance the post-receive never got executed? I have had one user who somehow managed to ctrl-c during a commit and I'm sure the post-receive hook didn't run.
21:49 jantje (git 2.5.2)
21:50 fracting joined #git
21:51 MZAWeb_ left #git
21:52 romerocesar joined #git
21:53 chitopunk joined #git
21:54 chitopunk joined #git
21:54 MattMaker joined #git
21:55 javajoe joined #git
21:56 Dougie187 left #git
21:56 ShekharReddy joined #git
21:58 stantonk joined #git
22:00 cd-rum joined #git
22:01 joki joined #git
22:02 lugzer joined #git
22:02 LiohAu joined #git
22:03 jaggz GyrosGeier: thanks :)
22:04 stantonk_ joined #git
22:06 cd-rum joined #git
22:07 ochorocho__ joined #git
22:08 Literphor joined #git
22:10 circ-user-A26a8 joined #git
22:12 romerocesar joined #git
22:12 ekinmur joined #git
22:14 garet joined #git
22:15 zacts joined #git
22:16 Neobenedict joined #git
22:16 Neobenedict joined #git
22:17 stantonk joined #git
22:19 kbs joined #git
22:20 stantonk joined #git
22:20 mischat joined #git
22:20 maroloccio hi. i am in detached HEAD and want to push _that_ to origin foo, where foo isnt a branch on origin so this fails: git push HEAD:foo. what to do?
22:23 jaguarmagenta joined #git
22:23 fuchstronaut joined #git
22:23 EvilPenguin joined #git
22:24 maroloccio is it possible to push in detached HEAD mode?!
22:24 ojacobson target refs/heads/foo
22:24 ojacobson eg. `git push origin HEAD:refs/heads/foo`
22:24 BlaXpirit maroloccio, just make a branch out of that, to make life easier
22:25 ojacobson also this
22:25 maroloccio BlaXpirit: no i cnat
22:25 HoierM joined #git
22:25 maroloccio because i am in git rebase interactive mode and these commands are run using exec to-do
22:25 maroloccio makes sense?
22:25 BlaXpirit no, but it's probably just me
22:25 maroloccio no?
22:26 pioiytr_ joined #git
22:26 maroloccio imagine you have to do ci on the previous 30 commits. but ci only works on origin
22:26 BlaXpirit i'm saying i don't know enough for it to make sense
22:26 maroloccio you have to push to origin for it to do the ci for you
22:26 maroloccio ci has to pass for all 30 commits, cant pass for 29 of them
22:27 maroloccio i am checking out the last 30, pushing to origin, then reading the reports
22:27 maroloccio could i do that better?
22:27 ramsub08 joined #git
22:30 maroloccio ojacobson: that tip worked perfectly. thank you. ci is churning for me. appreciate it
22:31 ojacobson This would be easier if you could run your CI tasks locally
22:31 ojacobson then you could use git rebase --exec 'your build step'
22:31 stantonk joined #git
22:31 stantonk joined #git
22:32 maroloccio i can run my ci tasks locally
22:32 maroloccio but each one takes like 40 seconds
22:33 maroloccio of course i am using git rebase --exec as a principle.. but i have to push for it to happen remotely.. or i'll fry my laptop
22:35 sangy joined #git
22:36 editshare-ericf joined #git
22:36 maks joined #git
22:37 maks ?
22:37 maks HEY ?
22:37 davisonio joined #git
22:37 maks hi ?
22:37 Eugene maks - !welcome
22:37 gitinfo maks: 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.
22:37 maks hi ?
22:38 maks !welcome
22:38 gitinfo 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.
22:39 davisonio joined #git
22:39 stantonk joined #git
22:40 refried_ joined #git
22:40 BSaboia joined #git
22:40 acetakwas joined #git
22:41 Sasazuka joined #git
22:41 ramsub07 joined #git
22:42 pioiytr_ joined #git
22:42 romerocesar joined #git
22:48 Literphor joined #git
22:49 Faylite joined #git
22:50 oalu joined #git
22:52 chrisshattuck joined #git
22:52 espindola joined #git
22:54 Neobenedict_ joined #git
22:54 Neobenedict_ joined #git
22:58 lugzer joined #git
22:59 GavinMagnus left #git
23:00 ramsub08 joined #git
23:02 hahuang61 joined #git
23:02 romerocesar joined #git
23:04 NwS joined #git
23:04 a3Dman joined #git
23:05 neogineer joined #git
23:05 pioiytr_ joined #git
23:06 azbarcea joined #git
23:07 neogineercom joined #git
23:09 Gsham joined #git
23:10 ramsub07 joined #git
23:11 neogineeeeeeer joined #git
23:13 justanot1eruser joined #git
23:14 IWishIKnew_ joined #git
23:15 titanium17 joined #git
23:15 sweatsuit_ joined #git
23:16 Dreamer3 joined #git
23:16 zacts joined #git
23:16 frogonwheels_ joined #git
23:16 medeirosthiiago joined #git
23:20 ashokrajar joined #git
23:21 medeirosthiiago joined #git
23:21 Faylite Wow, today I lost a bunch of work on my .vimrc file, no clue what happened, it's like the commit never happened. Is there a cache directory or something I could search?
23:23 mischat joined #git
23:27 yanome joined #git
23:28 mablae__ joined #git
23:29 kugel joined #git
23:29 kugel joined #git
23:29 arescorpio joined #git
23:31 d4rklit3 joined #git
23:31 Chiko joined #git
23:33 tctara joined #git
23:34 makinen joined #git
23:34 boz_v1 joined #git
23:34 devbug joined #git
23:35 SporkWitch joined #git
23:35 Mathiasdm joined #git
23:36 inflames joined #git
23:39 daniel_rose joined #git
23:40 davisonio joined #git
23:40 steven_a_s joined #git
23:41 mehola joined #git
23:42 benwbooth joined #git
23:43 romerocesar joined #git
23:45 garet left #git
23:51 lugzer joined #git
23:53 netj joined #git
23:55 ashokrajar joined #git
23:56 zumba_addict joined #git
23:56 nettoweb joined #git
23:59 CEnnis91 joined #git

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