Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-01-27

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

All times shown according to UTC.

Time Nick Message
00:04 BlueProtoman joined #git
00:04 mikey1 joined #git
00:05 Noldorin joined #git
00:07 manuelschneid3r joined #git
00:07 ikelso joined #git
00:08 mikeymop do i need to generate a new recovery codes everytime I want to make a personal access token to commit
00:09 dmc joined #git
00:10 mikeymop left #git
00:11 stickperson joined #git
00:13 [0xAA] joined #git
00:14 Cabanossi joined #git
00:14 mischat joined #git
00:14 phale joined #git
00:14 a_thakur joined #git
00:15 anuxivm joined #git
00:15 phale hello.. I have had a lot of files in my root directory in my git repository, and now I moved it all to another directory in the git but when I pushed it was still in the root? How do I fix this
00:17 Derperpe1 joined #git
00:19 jimi_ joined #git
00:20 idlemind joined #git
00:22 madewokherd joined #git
00:25 d1z joined #git
00:30 a_thakur joined #git
00:30 jeffspeff joined #git
00:35 roelmonnens joined #git
00:36 blueyed joined #git
00:37 phale left #git
00:41 achowe joined #git
00:43 Darren_ joined #git
00:44 Hudu joined #git
00:46 KV joined #git
00:50 malide n:b 1
00:50 bcc joined #git
00:50 bcc joined #git
00:57 mehola joined #git
00:58 krillr joined #git
01:02 ShalokShalom joined #git
01:07 qpdb joined #git
01:10 msonntag joined #git
01:10 gusgg joined #git
01:10 gusgg joined #git
01:10 gusgg joined #git
01:12 Nilesh_ joined #git
01:14 jccn joined #git
01:15 mischat joined #git
01:17 mda1_ joined #git
01:17 adino joined #git
01:25 memfrob I knew the answer the phale's question but he left
01:25 memfrob I was out..
01:25 arescorpio joined #git
01:25 memfrob *answer to
01:26 guyz joined #git
01:27 darkSeid_ memfrob, you could send him a msg?
01:27 darkSeid_ might save him hours of torture
01:28 Cabanossi joined #git
01:29 dmc https://media.giphy.com/media/sgVb9gX9DpYEo/giphy.gif
01:30 roelmonnens joined #git
01:30 pur3eval joined #git
01:31 bremner I suspect that answer to phales question is really about deployment strategies
01:32 memfrob no his commit was probably wrong. If you move files from one dir into another and run git add -A && git commit -as it _should_ just work
01:32 reverse_light joined #git
01:33 matoro joined #git
01:33 memfrob so I'll ask him to check his git commit --amend to make sure the files were moved properly.
01:33 * bremner shrugs. "when I pushed some files didn't change" is a red flag for me.
01:34 memfrob yeah his commit probably sucked.
01:34 memfrob to put it bluntly
01:34 prg3 joined #git
01:34 memfrob I never had that issue which tells me its a problem between chair and keyboard
01:35 kristofferR joined #git
01:35 pR0Ps joined #git
01:35 memfrob git should always track files when added to be tracked. if it's not tracking them, then he probably didnt run git add -A
01:35 memfrob files being moved to/from etc should always at least show up in the commit summary
01:36 memfrob now if it's showing up in the commit but not the server, then it's the other end's fault for not handling the git protocol properly.
01:37 memfrob people get too cute with server hosting i swear, like that stupid SourceTree or whatever people were talking about yesterday.
01:37 memfrob keep it simple and stupid and everything just works. remove the virtual layers and cloud and hypervisor xen stuff and wow! EVERYTHING JUST WORKS!
01:38 memfrob People have such a personal vendetta against bare metal and it makes me hate the ones who buy into the gimmick of using servers which aren't really there, but there, but also there, and here, and over there, all at once, while also being nowhere at the same time.
01:39 darkSeid_ that sounds like the plot of inception
01:39 memfrob XD
01:39 darkSeid_ :)
01:40 brent__ joined #git
01:41 redhedded1 joined #git
01:47 KV joined #git
01:48 ResidentBiscuit joined #git
01:50 BlueProtoman joined #git
01:52 d1z joined #git
01:52 almy joined #git
01:54 xlegoman joined #git
01:57 boombatower joined #git
01:58 Cabanossi joined #git
02:00 elsevero joined #git
02:02 nickabbey joined #git
02:02 danimal2 joined #git
02:06 webdev007 joined #git
02:06 tirej joined #git
02:08 cdg joined #git
02:09 aidalgol joined #git
02:10 snowkidind joined #git
02:10 matoro joined #git
02:11 heroux joined #git
02:12 nckpz joined #git
02:16 mischat joined #git
02:21 Gsham joined #git
02:23 anuxivm left #git
02:23 pks joined #git
02:24 d1z joined #git
02:25 roelmonnens joined #git
02:26 z3t0 joined #git
02:26 kristofferR joined #git
02:26 finalbeta joined #git
02:27 WakiMiko joined #git
02:31 TyrfingMjolnir joined #git
02:31 thiago joined #git
02:33 kyan joined #git
02:37 MuffinMedic joined #git
02:38 meadmoon joined #git
02:38 d5sx43 joined #git
02:41 canuk_ joined #git
02:41 robattila256 joined #git
02:41 cdg joined #git
02:46 d1z joined #git
02:46 Mavericks joined #git
02:47 JBlazeCoder21 joined #git
02:48 Dougie187 joined #git
02:48 a_thakur joined #git
02:49 KV joined #git
02:49 MuffinMedic joined #git
02:49 cfoch joined #git
02:50 Vampire0_ joined #git
02:54 d^sh joined #git
02:56 Lyqyd joined #git
02:57 aavrug joined #git
02:58 XenophonF joined #git
02:58 Cabanossi joined #git
02:59 aavrug joined #git
03:00 keepguessing joined #git
03:02 keepguessing So there is a repo I am working on which has "a master branch" and tags used for versioning.
03:02 keepguessing I want to push a change to an older tag and not on the master. Is there a way to do that?
03:03 keepguessing Any suggestions would be helpful.
03:03 cdg joined #git
03:05 sbulage joined #git
03:09 rewt keepguessing; checkout the tag, create a branch, commit to that branch
03:12 Limes joined #git
03:12 Isaac joined #git
03:13 chachasmooth joined #git
03:13 cdg_ joined #git
03:16 keepguessing rewt: ok thanks. The target repo is on the upstream. can I have a pull request that requests a branch?
03:16 keepguessing rewt: the change works also on the master so I could push to master and then crossport it to this new branch right?
03:17 rewt the pull request is the request to merge your changes in; you'd make the branch yourself
03:17 mischat joined #git
03:17 rewt then again, i'm not 100% on how github works, and !github
03:17 gitinfo Note that git != github. Feel free to ask us about Github-specific features (Forks, Pull Requests, Wikis, etc), but there are no guarantees. There is a #github channel, which might help too (again, no guarantees)
03:17 keepguessing rewt: ah Thanks. You are right.
03:17 Vortex34 joined #git
03:18 rewt maybe best to contact the maintainer and ask how they'd prefer you to do it
03:20 sbulage joined #git
03:20 roelmonnens joined #git
03:20 keepguessing rewt: thanks.
03:22 lektrik joined #git
03:22 fstd_ joined #git
03:27 mehola joined #git
03:31 gusgg Does this seem like a bug with check-ignore? I wasn't able to reproduce it unless in /. http://pastebin.com/raw/97qi81bC
03:32 GodGinrai joined #git
03:35 cdg joined #git
03:41 matoro joined #git
03:43 gusgg indeed it's not an issue in 2.11 !
03:47 mwlang joined #git
03:47 otiose joined #git
03:47 matoro joined #git
03:48 a_thakur joined #git
03:48 otiose I'm not sure this exists, but is there a way to print out something like a `git blame`, except it's kind of like a diff, in that only lines by a certain author are highlighted? And you can do this for the entire repo, or a list of files?
03:49 otiose My use case is that I'm trying to find all code written by a certain person in order to do a review
03:50 KV joined #git
03:50 howdoi joined #git
03:50 a_thakur joined #git
03:51 howdoi what is the one git command/flag that blew your mind when you realized it?
03:52 Puhe joined #git
03:52 sbulage joined #git
03:54 phroa rebase -i and request-pull for wildly different reasons
03:56 atomi joined #git
03:56 otiose oh well, just grepping the blame should be good enough I guess
03:57 otiose phroa: whoa, is that basically making a pull request through git itself?
03:57 phroa otiose: it creates an email message literally requesting someone to pull your code from your personal git repo
03:57 z3t0 joined #git
03:58 phroa I never even realized that was _a_ workflow, let alone the original one
03:58 otiose whoa
03:59 Puhe left #git
03:59 howdoi nice
04:00 phroa here's what it looks like https://marc.info/?l=linux-kernel&m=148547556118145&w=2
04:01 otiose yeah that kind of blows my mind
04:01 otiose no idea that existed
04:01 phroa like, damn, that's how serious work happens
04:02 phroa anyway for what it's worth if you use grep with a pattern like '.*The Author's Name.*' and --color=always it will highlight the entire lines in red for you ;)
04:03 sunri5e joined #git
04:04 hexagoxel joined #git
04:04 menip joined #git
04:04 Goplat joined #git
04:05 Raed joined #git
04:06 howdoi looking for the format for --date on commit in osx
04:07 howdoi something like date --date="2 days ago"
04:07 phroa see Date Formats in man git-commit
04:07 gitinfo the git-commit manpage is available at http://jk.gs/git-commit.html
04:07 duderonomy joined #git
04:08 howdoi phroa: oh well, date --date is not available on osx, I must install gnu utils with brew or something
04:09 jimi_ joined #git
04:09 phroa oh you just want to use the date command.  yeah, the purge of really useful GPL software is rather irritating
04:10 howdoi shudder
04:10 phroa I install the GNU userland stuff with --overwrite-system-binaries or whatever it is to replace the apple BSD ones
04:12 ShalokShalom joined #git
04:13 howdoi https://www.irccloud.com/pastebin/6N9dnAb5/
04:13 howdoi $ date +%N
04:13 howdoi N
04:13 howdoi :D
04:13 phroa rip
04:14 Cabanossi joined #git
04:14 howdoi BTW brew installed gdate
04:14 mehola joined #git
04:14 gusgg regarding the check-ignore bug when a git repo is rooted at "/", i've determined it's broken in 1.9.3 and fixed in 1.9.4
04:14 phroa yeah if you don't want the g prefix for GNU commands, you need to use another option when you install.  I think it's listed in the package description for most GNU things
04:15 phroa meanwhile git is at (at least) version 2.11
04:17 howdoi cool, will check that
04:17 jeffreylevesque joined #git
04:18 phroa ah here http://apple.stackexchange.com/a/69332
04:18 mischat joined #git
04:19 howdoi phroa: thanks! --default-names
04:19 phroa I think I was thinking of the --override-system-vim option for MacVim, --default-names is way less threatening of an option name
04:19 batrick joined #git
04:20 howdoi also we need to do a PATH="/usr/local/opt/coreutils/libexec/gnubin:$PATH"
04:22 d1z joined #git
04:24 atomi joined #git
04:27 Cabanossi joined #git
04:29 matoro joined #git
04:34 PaulCapestany joined #git
04:38 cagedwisdom joined #git
04:39 solrize joined #git
04:40 solrize i think i've asked this before but what's the best way to reverse clone a repo?  i.e. repo is on the local machine and i want to make a clone on the remote machine, without pulling from the remote since it doesn't have auth credentials to the local (the local has credentials to the remote)
04:40 solrize just scp -r the repo?
04:40 Gsham joined #git
04:45 Sasazuka joined #git
04:50 ayogi joined #git
04:50 KV joined #git
04:51 bocaneri joined #git
04:52 Gsham joined #git
04:53 msonntag joined #git
04:57 dreiss joined #git
04:59 raijin joined #git
05:00 dermoth joined #git
05:03 govg joined #git
05:05 Random832 solrize, or you could create a blank repo on the remote and then push to it
05:11 solrize Random832, hmm thanks, will try that
05:12 Cabanossi joined #git
05:14 thebinary joined #git
05:19 mischat joined #git
05:22 armyriad joined #git
05:22 d1z joined #git
05:25 aidalgol joined #git
05:26 bilb_ono joined #git
05:27 solrize aha had to say git push --all.  got into a confused state with just git push
05:30 solrize so i have a program to do X and i hacked it to do Y (related) instead.  is it good practice to make a separate repo for Y, or just a branch?
05:32 thebinary joined #git
05:32 ckruczek joined #git
05:32 connectkushal joined #git
05:33 z3t0 joined #git
05:36 bket joined #git
05:36 z3t0 joined #git
05:38 TomTomTosch joined #git
05:39 z3t0 joined #git
05:41 Loomaanaatii joined #git
05:42 nachoman joined #git
05:42 coco joined #git
05:44 adino joined #git
05:46 sbulage joined #git
05:47 UTAN_dev joined #git
05:47 bket joined #git
05:48 Loomaanaatii joined #git
05:49 cagedwisdom joined #git
05:50 grawity at first it should be *at least* a branch
05:50 grawity but splitting off a branch into its own repo is trivial
05:51 KV joined #git
05:52 nidr0x joined #git
05:52 tirej joined #git
05:53 wizardedit joined #git
05:53 wizardedit joined #git
05:53 rominronin joined #git
05:55 solrize grawity, thanks.  yeah it's in a branch now and flipping back and forth seems weird.  probably should factor out the common parts to be able to push stuff "upstream" separately from the forks
05:55 grawity well, you *could* have two clones of the same repo, with different branches checked out
05:56 grawity or e.g. use `git worktree`
05:56 solrize hmm what's that?  not built into git
05:56 grawity not built into your old git version, that is
05:57 solrize git version 2.1.4
05:57 solrize newer?
05:57 grawity git version 2.11.0
05:57 solrize ic
05:57 solrize ok i'll look for it
05:57 grawity `git worktree` is kinda like having two clones, but they all share the same settings (like remotes)
05:57 Cabanossi joined #git
05:57 grawity but, anyway
05:57 solrize is it ok to have more than one remote in config?  so if i push it goes to both places?
05:57 grawity well yes, that's why you have `git remote add`
05:57 solrize aha :) thanks
05:57 grawity but you'll have to push to both manually
05:58 solrize oh hmm
05:58 solrize well close enough
06:00 xissburg_ joined #git
06:01 dreiss joined #git
06:04 dec0n joined #git
06:07 SwiftMatt joined #git
06:07 bjpenn joined #git
06:08 yazdmich joined #git
06:10 UTAN_dev joined #git
06:12 msonntag joined #git
06:12 bjpenn how do i figure out all files that were remvoed from a git repo with the contents "blah"
06:12 bjpenn or deleted with file names containing "blah"
06:12 miczac joined #git
06:13 Rodya_ joined #git
06:13 rscata joined #git
06:13 osse bjpenn: 1) git log -Sblah and look for deletions
06:14 bjpenn osse: nice
06:14 osse bjpenn: 2) git log -diff-filter=D -- '**/*blah*'
06:14 osse I think
06:14 WayToDoor joined #git
06:15 bjpenn osse: ive been using git for years... how do you know all these commands?
06:15 bjpenn i mean i use just use the default ones over and over
06:15 bjpenn add / commit / push / merge
06:15 bjpenn etc
06:15 bjpenn anything that i dont use within 2 weeks, its automatically forgotten :P
06:15 Thorn joined #git
06:17 osse You learn a lot by teaching other people
06:19 mischat joined #git
06:20 bjpenn hm osse , i still can't find it with that command. I don't think it's in the commit messages. That did get me some matches though... How about searching for filenames?
06:21 osse bjpenn: that was 2)
06:23 bjpenn ah ok
06:23 dec0n joined #git
06:24 m0ltar joined #git
06:25 f0rks joined #git
06:25 ColdKeyboard joined #git
06:26 nidr0x joined #git
06:27 Cabanossi joined #git
06:33 daurnimator joined #git
06:35 overlord_tm joined #git
06:36 qt-x joined #git
06:37 otiose git log --no-merges --author="name" --oneline --since=2016-10-01 --name-only --format="" | sort -u | xargs -I {} git blame {} | grep -E ".*name.*|$" --color=always
06:37 otiose Thats how you see all the files that a person has changed since the beginning of october with the lines that they changed being highlighted
06:37 otiose In case anybody was paying attention
06:41 d1z joined #git
06:49 vkipy joined #git
06:51 ThomasLocke joined #git
06:51 ThomasLocke joined #git
06:52 ckruczek otiose: with the appropriate git alias one can replace the datetime easily to any date he wants. nice stuff, thx!
06:52 KV joined #git
06:52 zeroed joined #git
06:52 zeroed joined #git
06:53 otiose yeah, i forgot to take that one out when copy/pasting, lol
06:56 Raging_Hog joined #git
06:57 JeroenT joined #git
06:57 fhackdro- joined #git
06:59 titankiller joined #git
06:59 adino joined #git
07:00 titankiller joined #git
07:01 dreiss joined #git
07:04 Rodya_ joined #git
07:04 bjpenn joined #git
07:05 JeroenT joined #git
07:07 Mikerhinos joined #git
07:07 matsaman joined #git
07:09 matoro joined #git
07:09 CIAguy joined #git
07:10 stickperson joined #git
07:11 zeroed joined #git
07:13 Maple__ joined #git
07:13 Maple__ joined #git
07:13 Maple__ joined #git
07:13 bernardio joined #git
07:14 bjpenn joined #git
07:14 chele joined #git
07:15 jknetl joined #git
07:16 lordjancso joined #git
07:16 roelmonnens joined #git
07:17 zeroed joined #git
07:17 zeroed joined #git
07:17 XenophonF joined #git
07:20 mischat joined #git
07:20 ihradek joined #git
07:23 lindenle joined #git
07:27 nckpz joined #git
07:28 lindenle joined #git
07:30 jagob joined #git
07:34 Tobbi joined #git
07:35 Loomaanaatii joined #git
07:42 jnavila joined #git
07:42 fpqc joined #git
07:43 metachr0n joined #git
07:46 inflames joined #git
07:46 syg_ joined #git
07:46 alesan joined #git
07:47 alesan hello. I sometimes use "git checkout <SHA1>" to get the code at a given commit and have a look
07:47 alesan what if I want to get only a specific file from that commit, and leave the rest of the repository at the current state?
07:47 alesan is it even possible?
07:48 vkipy joined #git
07:48 localh0st alesan: add pat hto the file after your sha
07:48 localh0st path*
07:49 localh0st git checkout sha1 /path/to/resource.file
07:50 alesan oh that easy? :)
07:51 leftriver05 joined #git
07:52 CheckDavid joined #git
07:53 ckruczek I have some question to gitflow. We are currently trying to adapt gitflow with regards to following changes. Imagine you have product which is released for hospitals(medical equipbment) we release the product and so on master there is a release tagged with this version.
07:53 KV joined #git
07:53 ckruczek now 5 years later and several release later, this device comes back with a critical software bug and we have to fix this in the version 5 years ago because the device was not updated in the mean time because of customer definitions. how would you represent this in gitflow
07:54 ckruczek actually one must create a new branch from the 5 years back tagged version or any other ideas?
07:55 bjpenn joined #git
07:56 fpqc left #git
07:57 Singmyr joined #git
07:57 svm_invictvs joined #git
07:57 leftriver05 Evening all. Quick question... I'm used to pushing my commits from my active feature branch onto a stage branch. Is there some drawback to that, rather than switching to that stage branch and merging in from my feature branch?
07:58 pur3eval joined #git
07:59 jast leftriver05: using the 'feature:stage' syntax? that's fine, if it works (if both branches have moved on after your branching off or last merge, that kind of push will fail, so then you do need to do some kind of merge)
07:59 matsaman ckruczek: seeing that question here in there in years past online, but no ideal solution
07:59 matsaman here & there*
07:59 ngm joined #git
08:00 jast ckruczek: I'd go with your suggestion
08:00 jast whenever it turns out you have to maintain some old version, create a maintenance branch after the fact, makes sense to me
08:01 leftriver05 jast: I believe that's what it's doing, though I use sourcetree to do it, switching the remote branch from origin's version of my feature (tracked) to origin's version of stage
08:01 jknetl joined #git
08:01 matsaman ckruczek: http://stackoverflow.com/questions/16386323/following-git-flow-how-should-you-handle-a-hotfix-of-an-earlier-release
08:02 ckruczek ok jast
08:02 ckruczek thx matsaman
08:02 Tobbi joined #git
08:02 leftriver05 jast: Part of the confusion is that doing it the other way... switching to stage, merging in feature, commit and push to remote stage... that seems to create an additional commit. As such, the two branches look to be at different commits.
08:04 jast leftriver05: right, but that's the price of merge commits. sometimes they're actually preferable, if you want to keep a record of what was merged when, for posterity
08:04 matsaman I think people worry about the impossible task of making git look perfect from the back end too much =)
08:05 jast readable history is quite helpful, really
08:05 matsaman get it done, if you have to revert or change things to make it cleaner later, you can
08:05 jast you can't get it perfect in all situations, though
08:05 matsaman you don't want to become the VCS's servant =P
08:05 matsaman that'd be a proper role reversal
08:05 leftriver05 jast: Understood. And sure, better fidelity in history is obviously fine.
08:06 alesan once I add a file to the index, how can I see the "diff" that will be put in the commit if I proceed to do so?
08:06 jast alesan: git diff --staged
08:06 alesan ok thanks for the dumb question
08:06 leftriver05 jast: So summary version is really that I can rest easy that doing it the way I have is fine, understanding that if I go to push onto a branch that's ahead (without having merged into my feature first), it might reject it. Right?
08:07 jast leftriver05: yes. unless sourcetree is doing something abominable in the background, which I have no way of knowing :}
08:07 jast but if the authors are even just slightly sane, that shouldn't be the case
08:09 leftriver05 jast: It has been fine so far. I think it's doing the push <remote> <feature>:<stage> type syntax that you mentioned. Thanks so much for the help... it was a "I'd better check" type question and I feel better about it.
08:09 tirej_ joined #git
08:11 dsdeiz joined #git
08:11 jast leftriver05: yeah, never hurts to get a second or third opinion :)
08:12 Hudu joined #git
08:12 leftriver05 jast: For sure. :) Thanks again. Night all!
08:12 Cabanossi joined #git
08:13 xissburg joined #git
08:13 vkipy left #git
08:14 cagedwisdom joined #git
08:14 jmpp joined #git
08:15 skynxnex joined #git
08:15 ThomasLocke_ joined #git
08:18 dersand joined #git
08:19 nckpz joined #git
08:20 mischat joined #git
08:20 Hudu joined #git
08:21 nckpz joined #git
08:22 eduardas_m joined #git
08:22 sea-gull joined #git
08:24 govg joined #git
08:26 sdx23 joined #git
08:27 Tobbi joined #git
08:29 sdx23 Hi. Is it evil to do a commit --ammend after merging a branch?
08:29 Snugglebash joined #git
08:29 osse sdx23: depends on why you do it
08:30 sdx23 osse: because the changes in the branch we're not complete, but I just realized afterwards.
08:30 matsaman depends on whether you're doing it for evil? =P
08:30 cdown joined #git
08:30 sdx23 *were
08:30 osse sdx23: introducing unrelated changes into a merge commit is evil. doing stuff related to the merge it self is not.
08:30 grawity sdx23: in that case I'd suggest finishing the changes *in* that branch, and merging again
08:30 sdx23 matsaman: I'm always trying to be as evil as possible, but this is another story :p
08:30 osse grawity++
08:30 navidr joined #git
08:30 sdx23 grawity: and reverting the merge?
08:31 grawity nah
08:31 sdx23 ok, but in case the merge already happened?
08:31 sdx23 (but was not pushed)
08:31 grawity then it'll just merge the new commits and that's it
08:31 grawity well, if you want to keep things squeaky clean
08:31 grawity then you could throw away the unpushed merge with some `reset --hard`
08:31 osse sdx23: git reset --hard @~; g co otherbranch; do stuff; git commit; git checkout master; git merge otherbranch
08:32 sdx23 ah, so after the merge I'd checkout the branch, commit some, merge again.
08:32 grawity but it's not required; you can merge from the same branch multiple times, and each time it'll just add new commits
08:32 sdx23 ok, osse sounds cleaner in terms of history of merges later.
08:32 osse if the merge was not difficult then it's no problem redoing it
08:32 osse unless you lie about it not being pushed :p
08:33 sdx23 :)
08:33 sdx23 Thanks for all the comments, osse, grawity!
08:34 djb-irc joined #git
08:35 ac3takwas joined #git
08:38 mns joined #git
08:38 elect joined #git
08:38 ugvfxc joined #git
08:42 sdx23 Another thing: Is there a trick to not forget a stash pop? Can I show it in status?
08:43 bannakaffalatta joined #git
08:43 matsaman prompt?
08:44 matsaman alias?
08:44 PioneerAxon joined #git
08:44 matsaman while true; do notify-send "don't forget"; done
08:45 osse sdx23: git config --global alias.s '!git status; git rev-parse --quiet --verify refs/stash >/dev/null 2>&1 && echo YOU HAVE A STASH && :'
08:45 matsaman PS1="(stash)\n$PS1"?
08:46 Balliad joined #git
08:46 sdx23 osse: looks dirty, but maybe the best solution. Or I indeed do something to my prompt.
08:47 osse it's as clean as it's gonna get
08:47 osse unless you write your own replacement for git status
08:47 sdx23 matsaman: I havn't up to now, I remember that some solution I tried was quite slow.
08:47 osse but i condone the prompt idea. if you use the prompt stuff that git is bundled with then it's very easy to do
08:48 matsaman some prompts are incredibly overloaded
08:48 matsaman what ist hat one all the kiddies like
08:48 matsaman oh-my-zsh?
08:48 osse BEGONE, SATAN
08:48 matsaman heh
08:48 _ikke_ I have omz, but a very simple prompt
08:49 matsaman and bash's command_prompt thing is unfortunately a pain to use with, for example, awful things like virtualenv
08:49 * osse looks for holy water, garlic cloves and silver bullets
08:50 matsaman I doubled the complexity of my prompt recently after years
08:50 stickperson joined #git
08:50 matsaman from indicating whether I am root or not, to indicating whether I'm root or not and whether the last command failed or not
08:50 matsaman the rest is down to ginkgo
08:51 osse There are many prompts, but this one is mine: https://asciinema.org/a/dwz0w1kq9yznlysuluaosu2tr
08:52 ThomasLocke_ joined #git
08:53 ferr1 joined #git
08:53 KV joined #git
08:53 Junior Hello! I would like a second opinion about: new branch for each issue that i have then merge to dev or just use a single branch "dev" in which i will do all modifications? (Team of 3 people at this point)
08:53 matsaman heh
08:54 djb-irc joined #git
08:54 matsaman this is my prompt, this is my gun
08:54 matsaman osse: I would have to have an \n or two in there so it didn't vary in length
08:54 PioneerAxon joined #git
08:55 osse matsaman: what do you mean?
08:55 matsaman ~/d/myproject
08:55 osse oh, the distance from the left edge until where you start typing?
08:55 matsaman if you are elsewhere the location of the cursor changes, no?
08:55 matsaman yeah
08:55 osse doesn't bother me. it's what i'm used to
08:56 matsaman I like a fixed title/bar for path, otherwise I _usually_ use pwd... though in some installs I have the path followed by a newline
08:56 osse it's no accident it looks veeeery similar to the default prompt on Ubuntu :P (except for the git stuff)
08:56 Tobbi joined #git
08:56 matsaman it has an economy
08:57 matsaman the dots are kewl, though
08:57 osse lil' traffic light
08:57 matsaman yar
08:59 d1z joined #git
09:00 b100s joined #git
09:00 bongjovi joined #git
09:03 osse matsaman: what it doesn't show is that I added some cute arrows that show the "3 behind 5 ahead" stuff
09:03 osse but I've found it's kinda annoying so I think I'll remove it
09:06 aard_ joined #git
09:08 matsaman =)
09:08 cyphase joined #git
09:08 matsaman that does happen
09:09 matsaman experimentation ftw, though
09:10 Takumo joined #git
09:10 Takumo joined #git
09:10 mikecmpbll joined #git
09:11 jmpp joined #git
09:11 jmpp joined #git
09:11 osse ALSO: GUISE
09:11 marcogmonteiro joined #git
09:12 osse I made a thing that makes my stupid workflow easier: Sometimes I copy file paths directly from diff hunk headers
09:12 osse So I made a thingy that automatically translates a/foo/bar or b/foo/bar into :/foo/bar when you paste
09:12 osse It is my greatest achievement so far
09:12 Cabanossi joined #git
09:13 Dumblez0r joined #git
09:13 babilen joined #git
09:14 rvgate joined #git
09:15 bl0w3d_0ut joined #git
09:15 a_l_e joined #git
09:16 jknetl joined #git
09:16 matsaman osse: with a:?
09:17 sunri5e joined #git
09:17 jast osse: but what if you have to work with a subdir named 'a' or 'b', huh? HUH?
09:17 osse matsaman: yes.  magick git status to provide a path relative to the root of the repo
09:17 matsaman ah
09:17 osse jast: in that case the diff header will be a/a/foo/bar :P
09:18 jast I mean when typing it manually
09:18 nowhere_man joined #git
09:18 courrier joined #git
09:18 matsaman what'd you use to interface with your buffer?
09:18 osse jast: it only affects paste
09:18 jast that's evil
09:18 mvgfwd joined #git
09:19 jast so if you paste a path from 'git status' output instead of 'git diff' output...
09:19 osse it won't start with a/ or b/ so nothing will happen
09:20 jast new file:   a/c
09:20 osse matsaman: http://sprunge.us/XEiZ
09:20 jast NOW WHAT
09:20 osse now i'm screwed
09:20 jast hagh
09:20 jast -g
09:21 jast I win
09:21 quackgyver joined #git
09:21 osse but the benefit is larger than the cost
09:21 jast doesn't matter as long as I win
09:21 osse but you don't win
09:21 jast I do, and there's my crown
09:21 jast ChanServ has spoken
09:22 Singmyr joined #git
09:22 denisMone joined #git
09:22 Tobbi joined #git
09:23 jast nifty thing you made there, though
09:24 osse heheehehehh
09:24 osse Not sure how I got in the habit of copying paths from git diff in the first place, but yes I like it a lot
09:24 jast devoice wars!
09:25 NeverDie_ joined #git
09:25 osse I wasn't voiced in the first place, so I win
09:25 masuberu joined #git
09:25 osse You cannot take from me what I don't have!
09:25 osse unless I'm in debt...
09:26 jast I'm an op in disguise, so I win
09:26 * jast implements voice debt in gitinfo
09:26 * osse declares bankruptcy
09:27 osse It could be a lot smarter, for example: If I'm not in a git repo then using :/ probably doesn't make sense. In that case just remove. But then it could check if I'm in a subdir of the path I'm pasting and blablabla
09:28 KV joined #git
09:28 osse but the perfect is the enemy of the good and what-have-you
09:28 jast before you know it you've implemented a git user interface in zsh
09:29 osse yup. "Press and hold the Back button on the mouse while tapping the keyboards Pause button N times to stage the N most recently edited files"
09:29 afuentes joined #git
09:30 osse OMG! Press the Back and Forward buttons on the mouse to reblame on the current file on the previous commit!!
09:30 osse That would actually be soemwhat cool
09:31 mischat joined #git
09:31 apotry joined #git
09:32 clmsy joined #git
09:32 troulouliou_div2 joined #git
09:32 sunny256 joined #git
09:33 simesy joined #git
09:34 roelmonn_ joined #git
09:34 t0by joined #git
09:34 t0by joined #git
09:35 matsaman s/reblame/rebase remote/
09:38 ThomasLocke_ joined #git
09:38 ThomasLocke_ joined #git
09:40 matsaman ninite
09:46 alcohol joined #git
09:46 Tobbi joined #git
09:46 Hello71 joined #git
09:50 sunny256 joined #git
09:50 lindenle joined #git
09:51 chll_ joined #git
09:51 msonntag joined #git
09:52 skunkz joined #git
09:52 mischat_ joined #git
09:53 vktec joined #git
09:54 ismaris joined #git
09:54 TomyWork joined #git
09:55 roelmonnens joined #git
09:56 diogenese joined #git
09:58 nowhere_man joined #git
09:59 skunkz joined #git
10:00 Tobbi joined #git
10:01 m4sk1n joined #git
10:02 mischat joined #git
10:02 Nitesh How do I undo git reset HEAD~9 ?
10:02 tirej joined #git
10:03 dminuoso !reflog
10:03 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`
10:03 dminuoso I wonder.
10:03 rnsanchez joined #git
10:03 dminuoso Perhaps?
10:03 jast if it's the last thing you did, git reset HEAD@{1}
10:03 dminuoso Nitesh: Or you can git reset back to wherever you want your pointer to point
10:03 jast (which is a reference to a reflog entry)
10:05 Nitesh reflog gives me this. eebec16 HEAD@{0}: reset: moving to HEAD~9  , so git reset 'HEAD@{0}' shd work?
10:06 skunkz joined #git
10:06 Nitesh I want all the 9 commits back as 9 commits in the git log entry.
10:07 Dhanya[m] left #git
10:07 enjunear[m] left #git
10:07 kineticrootkit[m left #git
10:07 M-shine left #git
10:07 bjpenn joined #git
10:08 dikiaap left #git
10:08 mehola joined #git
10:09 mehola joined #git
10:09 skunkz joined #git
10:10 mehola joined #git
10:11 mehola joined #git
10:11 achadwick joined #git
10:11 mehola joined #git
10:12 cdown joined #git
10:12 mehola joined #git
10:13 Tobbi joined #git
10:13 mehola joined #git
10:14 mehola joined #git
10:15 mehola joined #git
10:15 flaviodesousa joined #git
10:15 mehola joined #git
10:17 rvgate_ joined #git
10:18 alex1a joined #git
10:21 merlac joined #git
10:22 SilverBe joined #git
10:22 notebox joined #git
10:22 durham joined #git
10:23 manuelschneid3r joined #git
10:23 theoceaniscool joined #git
10:24 Serpent7776 joined #git
10:24 theoceaniscool joined #git
10:26 mar77i joined #git
10:31 d1z joined #git
10:34 minikN joined #git
10:36 KV joined #git
10:37 tirej joined #git
10:39 bruce_lee joined #git
10:39 bruce_lee joined #git
10:39 osse Nitesh: then it's HEAD@{1}
10:39 osse HEAD@{0} is the same as HEAD
10:39 Nitesh osse: yes that worked
10:39 Nitesh :-)
10:42 DrSlony Hello, is there a reliable way of getting teh number of commits since the last tag without having to parse "git describe --tags --always --long"?
10:43 osse git rev-list --count tag..HEAD
10:43 DrSlony osse without knowing the tag name, just "the latest tag"
10:44 osse DrSlony: hmm, maybe git rev-list --count HEAD --not tags
10:45 dminuoso left #git
10:46 JeroenT joined #git
10:46 DrSlony osse "tags" would still need to be expanded to known tags
10:47 osse oops
10:47 osse I mean
10:47 osse DrSlony: hmm, maybe git rev-list --count HEAD --not --tags
10:47 katherine joined #git
10:48 DrSlony oh, that works great, thank you
10:48 JeroenT_ joined #git
10:48 marcogmonteiro joined #git
10:51 sarri joined #git
10:51 sarri joined #git
10:52 chrishan joined #git
10:52 DrSlony osse why would "git describe --tags --always --long" show a different commit count to "git rev-list --count HEAD --not --tags" ?
10:53 mischat_ joined #git
10:53 DrSlony i found two branches that do that, describe shows 4.2-1227-g3eeeb151 while rev-list shows 3
10:53 jast describe may be skipping merged in commits, try rev-list with an additional --first-parent
10:54 jast though that might miss tags depending on how your repo is structure...
10:54 jast d
10:54 mischat joined #git
10:55 DrSlony same with --first-parent
10:55 dikiaap joined #git
10:56 osse you run describe with --tags right
10:56 osse I can't think of a reason
10:56 mar77i joined #git
10:56 DrSlony I'm not much worried about this as both branches are old, possibly imported into git from mercurial, just curious
10:57 osse I cannot reproduce
10:58 adamru joined #git
11:00 kurkale6ka joined #git
11:00 Darcidride joined #git
11:02 brent__ joined #git
11:03 Tobbi joined #git
11:04 aw1 joined #git
11:04 chipotle joined #git
11:05 DrSlony seems "git rev-list --count HEAD --not --tags" returns the number of commits since the branch was created
11:05 ShalokShalom_ joined #git
11:07 osse :S
11:07 osse what does "since the branch was created" mean
11:07 t0by joined #git
11:07 t0by joined #git
11:08 DrSlony git checkout foo; git rev-list --count HEAD --not --tags; git describe --tags --always --long
11:08 DrSlony first returns "2", second returns "4.2-1167-g216af81b"
11:09 DrSlony foo was branched off from master and two commits were made
11:10 osse is this a public repo ?
11:10 DrSlony yes, git@github.com:Beep6581/RawTherapee.git, checkout denoise_uniform
11:12 Hudu joined #git
11:13 osse I get it
11:13 osse my rev-list assumes that the current commit is newer than any tag
11:14 osse but only two commits on this branch are not reachable from any tag
11:14 osse the rest are reachable from 5.0-gtk2 for example
11:15 DrSlony i believe only those two commits were made on that branch, the rest are from master
11:15 chrishan joined #git
11:15 osse yep
11:15 osse i believe so to
11:15 osse i can see it
11:16 DrSlony the more i know about git the less i feel i know, so i confine myself within the scope of belief :]
11:16 osse it makes sense why this happens
11:16 DrSlony i'm writing a faith-based cmake script
11:18 tvw joined #git
11:19 JeroenT joined #git
11:20 osse i'd parse git describe
11:20 osse i can't think of a way to get it done without lots of shell magic that is much harder than what you want to avoid
11:22 Sheogorath[m] joined #git
11:24 pbrewczynski joined #git
11:24 osse DrSlony: halleluja
11:24 osse git rev-list --count HEAD --not $(git tag --merged HEAD)
11:25 osse but then you need $() which means a shell. or you need to run git tag separately from cmake
11:26 osse and "--merged" is somewhat recent
11:29 pR0Ps joined #git
11:31 courrier joined #git
11:31 dan2k3k4 joined #git
11:31 afuentes joined #git
11:31 thunderbirdtr joined #git
11:32 DrSlony osse thank you
11:33 DrSlony Only reason I'm trying to get this info is to make our windows packagers happy who want a numeric version, but a numeric version by itself is complete nonsense and I think I will instead try to detect whether a tag was checked out. Tag implies a release. If not a release, then set numeric version to 9.9.9.9 and add a "-dev" suffix somewhere and be done with it. A commit hash is always included somewhere.
11:35 jast so you don't actually need the exact number of commits you've got on top of the tag?
11:35 YuGiOhJCJ joined #git
11:39 wotm8 joined #git
11:39 [0xAA] joined #git
11:40 jknetl joined #git
11:40 mar77i_ joined #git
11:44 DrSlony jast since getting them is difficult (parsing, shell, etc) and since even if i have them they still mean nothing without knowing the branch (feature branches, more than one stable branch, etc), then i think its better to ditch the numeric version idea altogether.
11:45 osse but if it IS a release you still need to parse
11:45 jast for a release you'd have a tag on that commit
11:46 osse how does that avoid scraping for version numbers?
11:47 jast it doesn't avoid it but reduces it so a simple 'git describe HEAD' or 'git name-rev --tags --name-only HEAD'
11:47 jast s/so/to/
11:47 DrSlony the contents of "git describe --tags --always --long" are always included in all builds, for our dev needs. As far as users are concerned, when releases are made a tag is all the info you need ("oho this is version 6.0, i need to upgrade")
11:47 jeffreylevesque joined #git
11:48 osse so what do you give the windows guys for their numeric version?
11:48 DrSlony It seems that Inno Setup, which we use, requires a numeric version of the form 0.0.0.0 http://www.jrsoftware.org/ishelp/index.php?topic=setup_versioninfoversion
11:49 DrSlony osse releases will get numeric versions from tags, such as 6.0
11:49 DrSlony nightly builds will use 9.9.9.9 for the installer
11:49 jast that's because the version field in windows binaries has the same format
11:49 jast I'd suggest 0.0.0.0 actually
11:49 DrSlony ok, 0.0.0.0 it is
11:49 seni joined #git
11:49 jast because there will be some dummy assuming that this is a newer stable version :)
11:49 DrSlony i could make up some number but without knowing the branch that number would mean nothing, it would be misleading
11:50 ShapeShifter499 joined #git
11:50 Tobbi joined #git
11:50 Anticom joined #git
11:51 osse so you've already written this parsing code that you want to avoid writing? :S
11:51 DrSlony the stable gtk3 branch is some 200 commits ahead of the stable master branch, but the changes only relate to the GTK+ code. The engine of the program is the same in both. So one branch would show version 5.0.42, the other would show 5.0.242, but they'd be the same thing aside from the different ui
11:52 DrSlony the parsing code was written in mercurial times or before, now it blew up with a tagged released which contained alphabetical characters (5.0-gtk3, 5.0-rc1, that type of thing)
11:52 ertesx joined #git
11:53 osse we have tags like v1.2.3 and v1.2.4-beta1
11:54 osse for a windoze thing I ended up using 1.2.4.547  for the numeric version, where 547 is the number of commits since
11:54 DrSlony you don't provide feature branch builds?
11:55 Sri[m] joined #git
11:55 kikijiki[m] joined #git
11:55 am2on joined #git
11:55 elwan7[m] joined #git
11:55 saintaquinas[m] joined #git
11:55 KevinMGranger[M] joined #git
11:55 anand[m] joined #git
11:55 M-meznak joined #git
11:55 alaeri[m] joined #git
11:55 unclechu joined #git
11:55 srkrishna[m] joined #git
11:55 Guest36054 joined #git
11:56 Remramm joined #git
11:56 SpEcHiDe joined #git
11:56 aviraldg joined #git
11:56 musicmatze[m] joined #git
11:56 spinningarrow[m] joined #git
11:56 monomon[m] joined #git
11:56 UncleCJ joined #git
11:56 unilovelight[m] joined #git
11:56 daopu[m] joined #git
11:56 patrickr[m] joined #git
11:56 Salander27 joined #git
11:56 hallogallo[m] joined #git
11:56 freimatz joined #git
11:57 osse not in the "huge QA department" sense
11:57 pR0Ps joined #git
11:58 mniip_ joined #git
11:58 JeroenT joined #git
11:58 [0xAA] joined #git
11:58 osse if I did I'd maybe but the branch name in the About dialog or something like hat
11:58 DrSlony what about the exe version?
11:59 gajus joined #git
12:00 osse hmm
12:01 Guest36054 left #git
12:01 DrSlony i'm also considering using the commit date
12:01 osse I set the AssemblyVersion, AssemblyFileVersion and AssemblyInformationalVersion attributes
12:01 osse The two first are the quasi thingy I mentioned above. The last one is the describe output as-is
12:02 Anticom Hi all. I just can't remember this. How do i load all submodules for a repo **after** it has been cloned already? I forgot the --recursive flag during clone :/
12:02 osse look decent enough when you right-click on the thing
12:02 osse Anticom: git submodule init && git submodule update
12:03 Anticom Cheers :]
12:04 neunon joined #git
12:05 nyo2 joined #git
12:09 bjpenn joined #git
12:09 bl0w3d_0ut joined #git
12:10 askb joined #git
12:10 robertsmall joined #git
12:12 SilverBe joined #git
12:12 robertsmall Is there any reason that anyone can think of that when I do `git describe` on my workstation it works as expected showing the tag, even with a clean checkout but when I'm running it on another (CircleCI container) it apparently doesn't know about the tag. I've checked and the tags are listed the same on both repos. I've pulled, I've pushed, etc.
12:13 stroybat joined #git
12:13 robertsmall I did have to force my tag because it hadn't originally been annotated.
12:13 grawity the tags are *listed*, but do they both point to the same object?
12:14 grawity by default a fetch won't update tags
12:15 robertsmall I have done `git pull --tags --force` and that didn't change anything
12:15 rvgate_ joined #git
12:15 cheater left #git
12:16 robertsmall I've also done a fresh checkout on my workstation and it seems fine there.
12:16 hobodave joined #git
12:16 grawity a fresh checkout or a fresh cone?
12:16 grawity fresh clone*
12:16 mehola joined #git
12:17 robertsmall Clone, sorry, still a little Perforce in my brain
12:18 DrSlony osse you wouldn't happen to know how to store the output of "git tag --merged HEAD" in a variable in CMake, would you? CMake's execute_process doesn't use an intermediate shell.
12:19 DrSlony Otherwise I could call a bash script to return the number of commits since the tag, though I don't know whether all systems where this program gets built support bash scripts.
12:19 DrSlony I suppose they do.
12:19 robertsmall As far as I understand it the CircleCI build is a fresh clone as well
12:20 grawity what does the server have for your tag?
12:20 grawity compare `git ls-remote .` on a working copy with `git ls-remote theserver`
12:20 noecc joined #git
12:20 osse DrSlony: execute_process(... OUTPUT_VARIABLE FOO ...)
12:24 robertsmall Seems to be the same on my local machine
12:24 robertsmall Had to release the CircleCI build so others could do some builds, it is repeatable though.
12:24 [0xAA] joined #git
12:25 seni joined #git
12:27 robertsmall OK, built a new one. On my local and the server I get `refs/tags/5.12.0^{}` and `refs/tags/5.12.0` but on the other box it has `refs/tags/5.12.0` but no `refs/tags/5.12.0^{}`
12:27 Cabanossi joined #git
12:27 KV joined #git
12:28 DrSlony osse I tried that but it fails, it tries to use the first returned tag as an argument http://paste2.org/DA15LNwe
12:28 DrSlony execute_process(COMMAND ${GIT_CMD} tag --merged HEAD OUTPUT_VARIABLE GIT_TAG WORKING_DIRECTORY "${PROJECT_SOURCE_DIR}")
12:28 kettlewell joined #git
12:29 DrSlony ah no, it tries to use all the tags as one argument
12:29 dec0n joined #git
12:29 maks__ joined #git
12:29 osse DrSlony: yes you need to split it and whatnot
12:30 osse DrSlony: string(REPLACE "\n" ";" GIT_TAG_LIST "${GIT_TAG}")
12:30 robertsmall grawity, not sure if you saw above
12:30 osse when use GIT_TAG_LIST unquoted in the next execute_process
12:30 osse *then
12:31 kettlewell joined #git
12:31 raijin joined #git
12:31 leeN joined #git
12:32 alex1a joined #git
12:32 DrSlony osse woah, thanks a million!
12:43 MrWoohoo joined #git
12:48 Cavallari joined #git
12:51 eduardas_m joined #git
12:52 kristoff_ joined #git
12:54 kristofferR joined #git
12:55 DrSlony osse if you were to make a numeric version of tag "5.1-rc1", how would you do that? stripping -rc1 and replacing it with the number of commits since the last tag would result in e.g. 5.1.1234
12:55 osse that's what I do because I was too lazy to come up with something better
12:55 Tobbi joined #git
12:56 osse I don't think I'm the correct guy to talk to about proper handling of this stuff
12:57 DrSlony as I see it, with git's flexibility, assuming one builds feature branches or tags release candidates, any numeric version is nonsense and should not be used.
12:59 osse As long as it's correct when you build a tag and it's better than completely useless when you build anything else I am happy
12:59 DrSlony same
12:59 ojdo joined #git
12:59 osse and imho 5.1.1234 is better than 0.0.0.0 but that's just me
12:59 osse even if it looks like it's after 5.1 when it's not
13:01 reynierpm left #git
13:02 davimore joined #git
13:03 Raging_Hog joined #git
13:05 synthroid joined #git
13:05 fragMental joined #git
13:05 osse jast: Oops. vim :/blabla/rofl.txt
13:10 reynierpm joined #git
13:10 reynierpm hello, I don't know if this topic is related to this channel or I should go to #linux or any other specific channel, if so then let me know
13:11 Raging_Hog joined #git
13:11 reynierpm I have two different Github users, rperez (used at office with a private repository) and reynierpm (used anywhere with my public repositories) I want to be able to push to those repositories over SSH but I know only how to setup one of them but not both
13:11 reynierpm can any give me some ideas for this problem?
13:12 jast osse: WIP, right :)
13:12 _ikke_ reynierpm: How do you normally setup one of them?
13:13 jast reynierpm: yeah, that's slightly tricky but doable
13:13 reynierpm in other words: rperez => should push to private repo only and with proper username/email, reynierpm => should push to public repo only with reynierpm as username and my email
13:13 reynierpm _ikke_, jast using SSH keys
13:13 reynierpm but for one it works like a charm
13:13 jast since github doesn't have your username in the URL, you have to pick a default and change git's config in the other repos
13:13 reynierpm the problem is when become two different ssh keys and two different repos
13:13 jast (the alternative is to make up a second hostname and use that for your second user)
13:13 reynierpm jast, the problem is not the username but the authentication
13:14 reynierpm jast, have two different keys could be a mess
13:14 jast I just mean that's what makes it impossible to have ssh choose the correct authentication based only on the URL
13:14 _ikke_ reynierpm: comes down the same thing
13:15 jast git config core.sshcommand 'ssh -i mysecondkey'
13:15 moritz reynierpm: http://stackoverflow.com/questions/22768517/how-to-manage-one-only-key-per-each-git-repository
13:15 moritz you could use github2 (or something) as hostname in one of the the repos
13:15 reynierpm moritz, _ikke_ jast ok allow me to review this info and try, brb thx for the moment guys
13:16 moritz and define github2 as an alias for github.com with IdentityFile ~/.ssh/otherkey in your .ssh/config
13:17 kristofferR joined #git
13:18 King_Hual joined #git
13:18 King_Hual joined #git
13:18 rivarun joined #git
13:18 sdothum joined #git
13:21 seni joined #git
13:21 courrier joined #git
13:22 thierryp joined #git
13:22 maks joined #git
13:23 CussBot joined #git
13:24 rvgate_ joined #git
13:24 ecuanaso joined #git
13:25 maks how to make some file untracked or undo git add operation?
13:25 raspado joined #git
13:25 moritz git rm --cached thefile
13:26 maks ty
13:26 ronny joined #git
13:27 JeroenT joined #git
13:28 KV joined #git
13:29 Hudu joined #git
13:31 denisMone joined #git
13:32 tsdev joined #git
13:37 inovas joined #git
13:39 synthroi_ joined #git
13:39 theoceaniscool joined #git
13:40 rwb joined #git
13:41 blackwind_123 joined #git
13:41 thierryp left #git
13:42 jsolano joined #git
13:42 Gsham joined #git
13:44 cousteau joined #git
13:44 cousteau how do I .gitignore the content of the build/ directory, but not the build/ directory itself?  build/* ?
13:44 _ikke_ You can't
13:44 _ikke_ git does not track jus directories
13:44 _ikke_ just
13:44 cousteau ah
13:45 dave0x6d joined #git
13:45 _ikke_ (you can create a dummy .gitkeep file or something and track that)
13:45 cousteau so git only tracks regular files, not directories
13:45 stackofcats joined #git
13:47 _ikke_ right
13:48 cousteau I see, thanks!
13:49 ThiefMaster so, add the directory to .gitignore and git add -f a .keep file in there
13:49 ThiefMaster or even better, make your build tools create the directory if it doesn't exist
13:51 * cousteau makes the build tools create the directory if it doesn't exist
13:52 DrSlony left #git
13:52 cousteau ThiefMaster, thanks for the advice :)
13:52 otiose joined #git
13:53 bl0w3d_0ut joined #git
13:55 bl0w3d_0ut joined #git
13:55 jeffreylevesque joined #git
13:55 govg joined #git
13:57 hasc joined #git
13:59 seni joined #git
14:01 tcorneli joined #git
14:02 brent__ joined #git
14:03 jimi_ joined #git
14:03 sbulage joined #git
14:09 ToxicFrog ThiefMaster: better than .keep: put a .gitignore file inside that directory that ignores everything inside it
14:10 ThiefMaster i prefer a top-level .gitignore but i guess that's am atter of preference
14:13 reynierpm guys take a look to this setup I've https://paste.fedoraproject.org/537664/ why when I git push to docker-pyrocms it goes as another different user? you can see here https://github.com/reypm/docker-pyrocms/commit/139c22ed6ca31968918de8c41fd51a0a02eacee2 as it was push using the user I have at office :-\
14:13 SilverBe joined #git
14:14 Guest99287 joined #git
14:14 masuberu joined #git
14:14 osse jast: previously I thought that whoever decided that git rev-parse --show-cdup should print nothing instead of '.' was a moron. But now I realize that person is a genius
14:15 reynierpm jast, moritz _ikke_ any advise?
14:16 moritz reynierpm: which setting would you think changes that?
14:17 reynierpm moritz, I expect git push to docker-pyrocms would go as reynierpm and not as mmi-rperez
14:17 perlpilot joined #git
14:17 timvisher joined #git
14:18 hasc joined #git
14:19 aw1 joined #git
14:19 t0by joined #git
14:19 reynierpm moritz, I though if I push to a certain repo it will take the local setup instead of the global but apparently I am wrong how do I avoid be pushing with mmi-rperez when I don't want to?
14:20 moritz reynierpm: the settings user.name and user.email only affect new commits that you create
14:21 reynierpm moritz, it's weird I had create the commit after changes them but .... it went with the wrong user
14:21 reynierpm :-\
14:21 Knirch osse: just curious, how so? and also, when is --show-cdup prefered over --show-toplevel?
14:22 osse Knirch: heh, in my specific usecase it was just easier to get the desired result that way
14:22 osse Knirch: one gives you a relative path, one gives you an absolute. Different usecases I guess
14:23 mischat_ joined #git
14:24 Raging_Hog joined #git
14:24 Knirch osse: mm, just having a hard time trying to figure out one where relative path would be better over absolute (from a git repository perspective)
14:24 osse Knirch: if you use it in a makefile, for example
14:25 osse generated files, maybe
14:25 tirej joined #git
14:25 _geoff joined #git
14:25 Vampire0 Knirch, or if you use Git for Windows in a Cygwin environment and want to give that path to a Cygwin tool that would not be able to handle an absolute windows path, but it could handle a relative path
14:26 Knirch osse: right, thanks :)
14:27 mohabaks joined #git
14:27 mda1 joined #git
14:28 osse Knirch: you cannot do   cd "$(git rev-parse --show-cdup)"   for example, since it might print nothing
14:28 osse but in mycase I want to prepend that string to another path, so prepending nothing makes sense.
14:28 osse Especially since it also print's e.g. '../' and not '..'
14:29 KV joined #git
14:29 Vampire0 osse you could do `cd "./$(git rev-parse --show-cdup)"` though :-)
14:29 osse that's true
14:32 osse But for *purely* aesthetic reasons I wanted to have nothing if I could
14:33 JeroenT_ joined #git
14:37 GodGinrai joined #git
14:37 ghormoon hi, how bad idea is to clone repository with windows (mingw) git, use it with linux one and push with windows again?
14:38 DrSlony joined #git
14:38 ghormoon i.e. I have stupid vpn that works only with windows and cuts down routes. so I can't reasonably tunnel the linux git to the repo :)
14:38 snowkidind joined #git
14:39 DrSlony I'm on branch foo, I made some commits and accidentally pushed them. Now I squashed them into one and don't want to force-push, I want to put it on a new branch, bar, How do I do that?
14:39 Ryanar joined #git
14:39 perlpilot ghormoon: aside from the tediousness, it shouldn't be a bad idea at all.
14:39 Vampire0 DrSlony, `git branch bar` or if you want to checkout the branch at the same time `git checkout -b bar`
14:39 ghormoon perlpilot: thanks, I was just checking if windows one won't use something too differently :)
14:40 ghormoon if you hve a better idea how to deal with the vpn, I'm in. it took me 3 days to figure how to hack it to be able to RDP to the box at least. but  doing something more complex like routing won't be feasible with hacks like this (i.e. rebuilding routes every 50ms :D)
14:40 bongjovi joined #git
14:40 perlpilot DrSlony: Vampire0's first way let's you specify where you want the branch to branch from by saying "git branch bar foo"  (where "bar" and "foo" can be any commitish things)
14:41 DrSlony thank you
14:42 shinnya joined #git
14:43 Vampire0 perlpilot, DrSlony both ways do, just add " foo" in the end
14:43 gitinfo DrSlony: This channel tracks karma based on who has gotten lots of thanks for being helpful. If you want to help someone reach karmic nirvana, please mention their name when thanking them with "thank you", "thankyou", "thanks", "thx" or "cheers". Try ".karma <nick>" or ".topkarma" to show karma status of a person.
14:43 AaronMT joined #git
14:45 ThumbBone cheers
14:45 ThumbBone .topkarma
14:45 gitinfo top karmic beings of past 30 days ('all' arg to see totals): o‍sse: 4,  _‍ikke_: 2,  j‍ast: 2,  t‍obiasvl: 1,  k‍adoban: 1
14:45 marcogmonteiro joined #git
14:46 ThumbBone .topkarma all
14:46 gitinfo top karmic beings of all time: o‍sse: 263,  _‍ikke_: 227,  c‍break: 143,  s‍eveas: 115,  c‍anton7: 110
14:46 Es0teric joined #git
14:46 otiose joined #git
14:47 ski7777 hi
14:47 Raging_Hog joined #git
14:48 ski7777 I forked a repository and I want to merge all upstream changes into my repository but I don't want the currnt coomits, I want to merge a tag. Is this possible?
14:48 skunkz joined #git
14:48 Hudu joined #git
14:50 Vampire0 ski7777, you can merge any commit-ish you like
14:51 SoRoot joined #git
14:52 ski7777 Vampire0: could you give me a command to merge a tag into the current branch?
14:52 skunkz joined #git
14:52 Vampire0 ski7777, `git merge tag`
14:52 mischat joined #git
14:53 nowhere_man joined #git
14:53 Hudu joined #git
14:54 d10n-work joined #git
14:54 nickabbey joined #git
14:54 noecc joined #git
14:56 ash_workz joined #git
14:57 Cabanossi joined #git
14:58 aard__ joined #git
15:01 kpease joined #git
15:01 zzz_ joined #git
15:01 bannakaffalatta joined #git
15:02 m0viefreak joined #git
15:05 WayToDoor joined #git
15:07 cdg joined #git
15:08 cyan___ joined #git
15:09 cyan___ joined #git
15:09 Es0teric joined #git
15:10 cyan___ joined #git
15:11 cyan___ joined #git
15:13 AlexP11223|2 joined #git
15:14 SoRoot1 joined #git
15:14 mozzarel1 joined #git
15:14 mischat_ joined #git
15:14 JanC_ joined #git
15:14 cyan___ joined #git
15:15 cyan___ joined #git
15:17 nickabbey joined #git
15:17 oven joined #git
15:17 mischat_ joined #git
15:18 MikeChernev joined #git
15:18 apeiros joined #git
15:20 synthroid joined #git
15:20 mischat joined #git
15:21 texinwien_ joined #git
15:22 ecuanaso joined #git
15:25 mischat__ joined #git
15:25 tirej joined #git
15:25 diget joined #git
15:26 Cabanossi joined #git
15:26 mischat_ joined #git
15:27 le_melomane joined #git
15:27 fahadash joined #git
15:28 kadoban joined #git
15:30 mohabaks joined #git
15:30 masuberu joined #git
15:31 KV joined #git
15:33 mischat joined #git
15:34 crose joined #git
15:35 SoRoot1 joined #git
15:36 mischat__ joined #git
15:38 CheckDavid joined #git
15:38 dikiaap joined #git
15:38 Dougie187 joined #git
15:39 cdg_ joined #git
15:41 mischat joined #git
15:42 griffindy joined #git
15:44 le_melomane joined #git
15:45 bongjovi joined #git
15:46 zarubin joined #git
15:49 mischat_ joined #git
15:51 ceprius joined #git
15:52 skunkz joined #git
15:56 ceprius left #git
15:57 diogenese joined #git
15:59 Ryanar joined #git
15:59 bongjovi joined #git
16:00 BackEndCoder joined #git
16:02 kristofferR joined #git
16:04 jnewt joined #git
16:04 buzzLightBeer joined #git
16:06 tang^ joined #git
16:06 dreiss joined #git
16:09 zarubin joined #git
16:12 GarethAdams joined #git
16:13 jagob joined #git
16:15 Jackneill joined #git
16:15 e14 joined #git
16:17 seni joined #git
16:17 Noldorin joined #git
16:18 jknetl joined #git
16:18 mehola joined #git
16:19 Nilesh_ joined #git
16:21 brent__ joined #git
16:21 GodGinrai joined #git
16:21 mohabaks joined #git
16:22 GodGinrai joined #git
16:25 troulouliou_div2 joined #git
16:25 troulouliou_div2 joined #git
16:26 Snugglebash joined #git
16:26 Jujo1 joined #git
16:30 regedit joined #git
16:30 SoRoot1 joined #git
16:31 KV joined #git
16:31 adino joined #git
16:32 madewokherd joined #git
16:33 stickperson joined #git
16:34 Jujo1 left #git
16:34 pulec damn I am super confused
16:34 pulec I was editing one file on some branch
16:34 pulec then did git checkout and all those changes seem to be lost in that file
16:34 Gsham joined #git
16:34 perlpilot pulec: did you "git checkout filename"?
16:35 pulec no git checkout branch
16:35 Dougie187 Is that the branch you were on?
16:36 tyreld joined #git
16:36 pulec hm I got the whole history, let me summarize
16:36 jozwior joined #git
16:37 cust joined #git
16:37 Jordan joined #git
16:38 pulec I was on branch develop, had 5 modified, files, I did git diff on all of them to see and confirm the changes, then I did git checkout db, but db branch is not localy avaliable, so it didn't do anything
16:39 pulec but after I did git status again, and was still on branch develop, one of the files disappeared in modified
16:39 pulec maybe that git checkout db reverted the changes?
16:39 pulec or maybe atom editor is somehow broken and makes mess between the branches
16:40 jknetl joined #git
16:41 pulec hm in on all of those branches that file stayed old
16:41 chipotle joined #git
16:41 Snugglebash joined #git
16:42 jccn joined #git
16:45 dave0x6d joined #git
16:45 qqx joined #git
16:45 pulec hm whatever I will just write it again, not sure what to learn form this though
16:45 Electrometro joined #git
16:46 nanga joined #git
16:46 nanga joined #git
16:47 ResidentBiscuit joined #git
16:51 Darren_ joined #git
16:52 multi_io joined #git
16:53 pulec OH LOL
16:53 pulec it makes sense
16:53 otiose joined #git
16:53 pulec i had folder db
16:53 pulec and I thought I was switching to db branch
16:53 pulec so as I did git checkout db
16:53 pulec which returned nothing, so success
16:53 pulec and that reverted everything
16:53 pulec damn! :D
16:53 pulec can I at least enable some warning?
16:54 mischat joined #git
16:55 TomyLobo joined #git
16:56 hahuang65 joined #git
16:58 jgornick Hey folks, is there a way that I can specify a pattern of branches that I want to fetch/pull?
16:59 jgornick Basically, I'd like to only fetch and pull branches that match master and */project/* (i.e. feature/project/cool-feature).
17:00 otiose joined #git
17:00 ResidentBiscuit Probably write a script for it
17:01 a_thakur joined #git
17:01 sgrover joined #git
17:02 jason237 joined #git
17:02 jgornick ResidentBiscuit: Is that something on the client or server side? Also if it's on the client side and I have a script to intercept what branches to fetch/pull, it's probably a fair assessment that it'll still download all of the branches?
17:03 sgrover Simple merge conflict help?  I'm on Branch A, and did git merge BranchB.  Conflict.  I want the changes in BranchB.  So BranchB is considered --theirs... right?
17:03 thiago jgornick: * can only be the last token. You can fetch refs/heads/project/*, but not refs/heads/*/project/*
17:03 nyo2 joined #git
17:04 ikelso joined #git
17:05 jgornick thiago: You could list multiple refspecs, right?
17:05 sdothum joined #git
17:05 thiago jgornick: yes
17:05 jgornick Perfect, that'll work.
17:05 thiago jgornick: use ls-remote to get the remote branches, then filter and fetch them
17:05 jgornick We have a limited number of topic branches, so I don't necessarily need 2 wildcards.
17:07 a_thakur joined #git
17:07 bilb_ono joined #git
17:07 stomplee joined #git
17:07 mda1_ joined #git
17:08 tvw joined #git
17:08 jgornick So, in my .gitconfig, where would I specify the refs for my remote?
17:09 _ikke_ jgornick: Instead of editing .gitconfig, you can beter use the existing tools to do it
17:09 jgornick Go on _ikke_ :)
17:09 svm_invictvs joined #git
17:10 _ikke_ Hold on
17:11 tonyj joined #git
17:11 tcorneli joined #git
17:11 thierryp joined #git
17:12 _ikke_ git config fetch.<remote>.fetch "refs/heads/master:refs/heads/master"
17:12 _ikke_ git config --add fetch.<remote>.fetch "refs/heads/feature/project/*:refs/heads/feature/project/*"
17:12 _ikke_ and go on
17:16 Domhack_ joined #git
17:17 ispn joined #git
17:20 jgornick _ikke_: Thank you so much!
17:20 Ryanar joined #git
17:23 e14 joined #git
17:24 lindenle joined #git
17:24 mischat joined #git
17:24 israelzuniga joined #git
17:30 raspado joined #git
17:31 yohnnyjoe joined #git
17:32 KV joined #git
17:34 Murii joined #git
17:35 snowkidi_ joined #git
17:35 adino joined #git
17:36 e14 joined #git
17:37 mda1 joined #git
17:39 Tobbi joined #git
17:41 jgornick Another question, is there a way to move files/folders in a stand-alone project repository to another repository that contains multiple projects folders, all while keeping commit history?
17:41 Snugglebash joined #git
17:43 Cabanossi joined #git
17:43 aard_ joined #git
17:44 LBV_User joined #git
17:44 nettoweb joined #git
17:44 byte512 joined #git
17:46 perlpilot jgornick: is there way?  yes.  you probably want subtree merge
17:46 jgornick perlpilot: I'll take a look at that. Thank you!
17:46 jgornick Another option I was thiking about would be to create a patch and then apply with --directory applied?
17:47 otiose joined #git
17:53 nickabbey joined #git
17:55 lindenle joined #git
17:57 LeBlaaanc joined #git
17:57 e14 joined #git
17:58 seni joined #git
18:00 jgornick _ikke_: The fetch.origin.fetch option doesn't seem to be working...
18:00 synthroid joined #git
18:00 zeroed joined #git
18:01 ecuanaso joined #git
18:01 ikelso joined #git
18:01 cagedwisdom joined #git
18:03 chipotle joined #git
18:04 matoro joined #git
18:05 bilb_ono joined #git
18:05 jgornick It doesn't seem to be fetching any of the branches that follow that pattern, even though I can see the ref when doing ls-remote.
18:05 KV joined #git
18:06 systemovich joined #git
18:06 grawity it's remote.<remote>.fetch
18:07 e14 joined #git
18:10 _ikke_ yeah, my bad
18:12 Ryanar joined #git
18:14 denisMone joined #git
18:14 a_thakur joined #git
18:16 a_thakur joined #git
18:17 mischat joined #git
18:19 nettoweb joined #git
18:19 durham joined #git
18:20 dreiss joined #git
18:21 Limes joined #git
18:25 jeffreylevesque joined #git
18:26 ikelso joined #git
18:26 spacelord_ joined #git
18:27 ShapeShifter499 joined #git
18:29 nickabbey joined #git
18:29 thiago joined #git
18:31 navidr joined #git
18:33 nickabbe_ joined #git
18:37 ikelso joined #git
18:38 jgornick grawity and _ikke_  Thanks!
18:39 noecc joined #git
18:39 ambro718 joined #git
18:40 prg3 joined #git
18:41 ambro718 Hi, is there something that would allow having different "classes" of ignored files - one for files that are generated and that I would frequently like to delete all at once (to ensure a clean build), and one for files that I do not want to have deleted (like local configuration files of IDEs)?
18:42 otiose joined #git
18:42 bannakaf_ joined #git
18:42 ambro718 usually I gitignore the former class and don't gitignore the latter class, that way I can do "git clean -Xdf" to clean up the build outputs. But this leaves the IDE config files always show up as "Untracked files" in git status, which is annoying.
18:45 Wjnm1 joined #git
18:45 sbeller_ joined #git
18:47 matoro joined #git
18:53 MineCoins joined #git
18:54 GodEater joined #git
18:54 GodEater joined #git
18:55 Murii joined #git
18:55 systemovich joined #git
18:56 Anja joined #git
18:57 ankit01ojha joined #git
18:58 brent__ joined #git
19:00 YuGiOhJCJ joined #git
19:00 theoceaniscool joined #git
19:00 KevinDF joined #git
19:01 KevinDF joined #git
19:01 Vampire0 ambro718, add a "clean" task to your build script that cleans up the build output files and add all files to .gitignore
19:03 btree joined #git
19:03 inflames joined #git
19:05 jeffreylevesque joined #git
19:05 duderonomy joined #git
19:06 al-damiri joined #git
19:06 ikelso joined #git
19:08 timvisher how do i list a given set of deleted files recursively based on a sha?
19:08 grawity based on what sha?
19:08 timvisher something like `git show sha^:foo/{bar,bat,bin}` where bar and bat are files and bin is a directory
19:08 grawity it's probably `git ls-tree -r <treeish>`
19:09 grawity where <treeish> is a tree sha, commit sha, or generally anything that somehow resolves to a directory (tree object)
19:10 timvisher grawity: yep that was it
19:10 dikiaap joined #git
19:10 timvisher hadn't made the connection that a treeish could be a revision spec
19:10 rnsanchez joined #git
19:11 grawity a commit can always be resolved to its root tree
19:11 ojacobson Don't the docs refer to that operation as "unpeeling"?
19:11 LeBlaaanc joined #git
19:11 grawity un- sounds like the opposite of that operatioin
19:13 causasui joined #git
19:13 causasui joined #git
19:13 jimi_ joined #git
19:13 Sasazuka joined #git
19:15 Rodya_ joined #git
19:15 fscala joined #git
19:16 irco joined #git
19:17 jeffreylevesque_ joined #git
19:17 Random832 grawity, unpeel and peel are synonyms in english
19:17 grawity goddamnit
19:17 Random832 like inflammable and flammable
19:18 jaafar joined #git
19:18 Random832 bone and debone
19:18 Derperpe1 joined #git
19:19 ceprius joined #git
19:20 ojacobson I guess "wrap" would be the antonym
19:21 nickabbey joined #git
19:21 ceprius left #git
19:21 GT4066 joined #git
19:23 nimbleark joined #git
19:24 ikelso joined #git
19:25 stickperson joined #git
19:26 ikelso_ joined #git
19:26 sea-gull joined #git
19:26 Ryanar joined #git
19:26 hwrdprkns joined #git
19:27 MineCoins joined #git
19:28 Derperpe1 joined #git
19:30 bilb_ono joined #git
19:32 jccn joined #git
19:33 energizer joined #git
19:33 energizer In git diff --color-words is it possible to hide lines that have no differences?
19:35 mikecmpbll joined #git
19:36 otiose joined #git
19:38 sns joined #git
19:39 Snugglebash joined #git
19:40 qqx energizer: You can use -U0 to tell it to not include any context lines.
19:41 energizer qqx: id dont think they're context lines with color words, it just shows all lines, highlighting differences within each line
19:43 qqx energizer: Have you tried it? The only unchanged lines that should be shown by diff are ones shown for context.
19:43 energizer qqx: git diff -U0 -w --color-words='[^[:space:],]|([[:alnum:]]|UTF_8_GUARD)+' --no-index sorted_k7_space sorted_both_space
19:44 energizer returns some lines with color diffs and some lines with no color
19:44 Balliad joined #git
19:45 e14 joined #git
19:45 KV joined #git
19:46 qqx In my testing I do see some lines with no color, although only ones there --color-words is causing some differences to be hidden completely.
19:46 morenoh149 joined #git
19:46 ambro718 Vampire0: the reason I'm using git clean is that git can ensure everything is cleaned, irrespective of the cleaning capabilities of the build system in use
19:46 morenoh149 is there a way to show every git commit message that contains some pattern?
19:47 morenoh149 not the pattern being in the changes of a commit
19:47 apotry joined #git
19:49 energizer morenoh149: git log --all --grep='Build 0051'
19:49 publio joined #git
19:50 morenoh149 energizer: thanks
19:51 morenoh149 and how about the entire commit body for some pattern?
19:52 tobiasvl have a good weekend https://twitter.com/isoiphone/status/824771226585296896
19:52 energizer i just googled for that idk
19:52 energizer git diff -U0  --word-diff-regex='[^[:space:],A-Z]' --no-index sorted_k7_space sorted_both_space
19:52 energizer fatal: unable to generate word diff
19:52 energizer qqx why would that be
19:53 canuk joined #git
19:54 davimore joined #git
19:54 mischat joined #git
19:57 MineCoins joined #git
19:58 crose joined #git
19:58 Cabanossi joined #git
20:01 pks joined #git
20:01 suttonwilliamd joined #git
20:09 dsdeiz joined #git
20:09 dsdeiz joined #git
20:11 jnavila joined #git
20:12 systemovich joined #git
20:13 satifant joined #git
20:15 garethdaine joined #git
20:15 Balliad joined #git
20:16 systemovich joined #git
20:21 ThomasLocke joined #git
20:21 ThomasLocke joined #git
20:23 Akimb joined #git
20:23 MillerBoss joined #git
20:24 noecc left #git
20:30 otiose joined #git
20:31 sargas joined #git
20:33 Prisoner-2460_1 joined #git
20:33 Prisoner-2460_1 git
20:33 ceprius joined #git
20:34 ceprius left #git
20:36 _ikke_ usage: git [--version] [--help] [-C <path>] [-c name=value] [--exec-path[=<path>]] [--html-path] [--man-path] [--info-path] [-p | --paginate | --no-pager] [--no-replace-objects] [--bare] [--git-dir=<path>] [--work-tree=<path>] [--namespace=<name>] <command> [<args>]
20:37 Prisoner-2460_1 just what i was looking for
20:40 darkSeid_ hi, i forked a repo and modified some stuff, did git status and its showing a bunch of staged files which i didn’t add to the staging area, any idea whats happened?
20:40 jccn joined #git
20:44 tobiasvl darkSeid_: what kind of files?
20:45 tobiasvl hard to theorize about what happened from that little info. !repro
20:45 gitinfo [!transcript] Please paste (using https://gist.github.com/ or similar) a transcript ( https://git.io/viMGr ) of your terminal session so we can see exactly what you see
20:45 cdown joined #git
20:45 bl0w3d_0ut joined #git
20:46 darkSeid_ ok ill get that done, ty tobiasvl
20:46 KV joined #git
20:49 oskarkv joined #git
20:51 darkSeid_ https://gist.github.com/BlindPacemaker/c2c3f8bac2b6b510c8f5ca754e0fc1e5
20:54 Jackneill_ joined #git
20:55 cdg joined #git
20:55 Singmyr joined #git
20:56 ikelso joined #git
20:56 sbeller joined #git
20:58 Atm0spher1c joined #git
20:59 Prisoner-2460_1 made four commits to develop instead of my feature branch. nothing’s pushed. the history of both branches is the same except these four new commits
21:00 Tobbi joined #git
21:01 morenoh149 joined #git
21:02 tobiasvl Prisoner-2460_1: all right, and then?
21:02 tobiasvl oh you want to move them
21:02 Prisoner-2460_1 indeed, into that feature branch
21:02 systemovich joined #git
21:02 tobiasvl how does the feature branch look?
21:02 MineCoins joined #git
21:03 tobiasvl easiest is probably just to checkout feature and cherry pick the commits over, then it doesn't matter how feature looks
21:03 Prisoner-2460_1 the feature branch doens’t even have anything commited on it
21:03 MineCoins joined #git
21:03 Prisoner-2460_1 it’s the same as it was when i created it off develop. so it’d be a clean transfer from devlep to feature.. if something like that is possilbe
21:04 SteffanW joined #git
21:04 MineCoins joined #git
21:05 MineCoins joined #git
21:05 sonOfRa joined #git
21:06 MineCoins joined #git
21:06 tobiasvl Prisoner-2460_1: oh, sure, that's even easier, if you're sure you don't have anything you need on develop: git branch -f develop
21:06 ikelso joined #git
21:07 Prisoner-2460_1 while on the feature branch?
21:07 tobiasvl ugh sorry
21:07 tobiasvl don't run that
21:07 davimore_ joined #git
21:07 tobiasvl that would ruin develop instead :P
21:08 tobiasvl while on develop: git branch -f feature_branch; git reset --hard HEAD^4
21:08 jstimm joined #git
21:08 leak1t joined #git
21:09 tobiasvl but that will rewrite both develop and the feature branch, so make sure you want to do what you said you wanted to do
21:09 doener joined #git
21:09 Sasazuka__ joined #git
21:09 tobiasvl (in particular that you want to throw away the current feature and overwrite it with develop, and that you want to erase the last 4 commits from develop)
21:10 tobiasvl Prisoner-2460_1: aaand HEAD~4, sorry
21:10 Prisoner-2460_1 corrrect. the feautre was created and forgotten about, so there’s nothing on there
21:12 ecuanaso joined #git
21:12 tobiasvl then you can do what I said. "git branch -f feature" will create a branch called "feature" pointing to the commit develop now points to, and -f means it will ignore that there already is a branch there. then "git reset --hard HEAD~4" will reset your current branch back to the fourth grandparent of HEAD.
21:13 FreakySneaky joined #git
21:14 * Prisoner-2460_1 >ignore that there is already a branch there
21:14 tobiasvl there are other ways to do it too, "git checkout feature; git merge develop; git checkout develop; git reset --hard HEAD~4" is probably less scary, since it would be a fast forward merge? but the first one is shorter
21:14 Prisoner-2460_1 does this mean -f is optional if i already have a branch though?
21:15 tobiasvl no, without -f it will refuse to overwrite the existing branch
21:15 tobiasvl you can also look at "git branch --move" which renames branches
21:15 leftriver01 joined #git
21:17 ferr1 joined #git
21:18 griffindy joined #git
21:22 Prisoner-2460_1 seems to have worked. this is great
21:23 Prisoner-2460_1 i wonder, since nothing was pushed, if i could’ve cherry picked. and then prevent the server repo from having that ugly duplicate commit?
21:23 jccn joined #git
21:24 otiose joined #git
21:25 hahuang61 joined #git
21:29 e14 joined #git
21:29 manuelschneid3r joined #git
21:29 otiose joined #git
21:29 WifiBrain joined #git
21:31 TomTomTo1 joined #git
21:32 nickabbey joined #git
21:35 mp4 joined #git
21:36 tobiasvl Prisoner-2460_1: yes, as long as you haven't pushed you can do whatever you want
21:37 e14 joined #git
21:37 tobiasvl (and even if you have pushed, you can maybe do whatever you want if the server allows you to push rewritten history, although that might mess up for your collaborators)
21:37 oskarkv_ joined #git
21:39 yohnnyjoe joined #git
21:42 jimi_ joined #git
21:44 overlord_tm joined #git
21:45 Bob8989|2 joined #git
21:46 apeiros left #git
21:47 KV joined #git
21:49 todd_dsm joined #git
21:50 nezZario i'm trying to split a git repo into smaller self contained repos ...
21:51 nezZario is there any way to retain commit history on only selected parts?  (rewriting history/etc is fine -- they'll be new distinct repos)
21:51 todd_dsm hey all, new on deck. I just did a git push and I can see my changes in the web ui. however, when I click on raw it shows the previous version of the file. Is there some documentation that describes this behavior?
21:51 nezZario todd_dsm: maybe you pushed to a separate branch than you are viewing ?
21:52 todd_dsm it's still pretty new; there's only a master at this point.
21:53 todd_dsm Uhhhh, ok - it just displayed the correct content in the page now. Is there some lag that I'm not aware of?
21:54 perlpilot nezZario: filter-branch is your friend
21:55 tobiasvl todd_dsm: probably some cache – what web UI are you talking about?
21:55 todd_dsm tobiasvl: just the github web ui.
21:55 tobiasvl oh, github. no idea then
21:55 todd_dsm ah, okay
21:56 tobiasvl but caching stuff like that is prettty common
21:56 nezZario perlpilot: thank you -- I thought that was probably the right way, but just wanted to confirm I was ballpark correct in the assumption
21:56 tobiasvl todd_dsm: ask in #github maybe
21:56 dsdeiz joined #git
21:56 todd_dsm fair enough, just enough time to for more coffee :-)
21:56 todd_dsm will do
21:57 pbrewczynski joined #git
21:59 BackEndCoder joined #git
21:59 Prisoner-2460_1 left #git
22:03 e14_ joined #git
22:05 systemovich joined #git
22:05 jimi_ joined #git
22:07 canuk joined #git
22:10 Vinnie_win joined #git
22:11 jimi_ joined #git
22:13 Cabanossi joined #git
22:13 LeBlaaanc joined #git
22:14 adino joined #git
22:17 brent__ joined #git
22:18 cdown joined #git
22:24 Derperpe1 joined #git
22:26 ccolorado left #git
22:28 Ryanar joined #git
22:32 pumafied joined #git
22:34 cagedwisdom joined #git
22:34 davimore joined #git
22:36 energizer1 joined #git
22:38 anuxivm joined #git
22:42 rwb joined #git
22:42 Cabanossi joined #git
22:45 timthowtdi joined #git
22:45 Ragnor joined #git
22:45 MarioBranco joined #git
22:48 KV joined #git
22:51 MarioBranco_2 joined #git
22:51 MarioBranco joined #git
22:51 jccn joined #git
22:54 Boyenn_ joined #git
23:03 treehug88 joined #git
23:03 tsdev joined #git
23:03 YuGiOhJCJ joined #git
23:03 druonysus_ joined #git
23:07 robotroll joined #git
23:07 fahadash joined #git
23:07 Naan joined #git
23:10 Jackneill_ joined #git
23:12 shinnya joined #git
23:15 jccn joined #git
23:17 ecuanaso joined #git
23:20 hasc_ joined #git
23:20 btree left #git
23:26 CheckDavid joined #git
23:30 otiose joined #git
23:31 mp_ joined #git
23:32 Dougie187 left #git
23:32 bannakaffalatta joined #git
23:34 bl0w3d_0ut joined #git
23:35 bl0w3d_0ut joined #git
23:36 sargas joined #git
23:37 mp4 joined #git
23:39 t0by joined #git
23:41 techntoke__ joined #git
23:41 ikelso joined #git
23:42 matsaman joined #git
23:42 menip joined #git
23:43 braderhart joined #git
23:44 dsdeiz joined #git
23:44 dsdeiz joined #git
23:45 adino joined #git
23:46 le_melomane joined #git
23:46 mp_ joined #git
23:48 LaptopBrain joined #git
23:49 KV joined #git
23:51 WifiBrain joined #git
23:51 frostythesnowman joined #git
23:52 madewokherd joined #git
23:54 almy joined #git
23:55 brent___ joined #git
23:55 jccn joined #git
23:56 slash77 joined #git
23:57 jstimm joined #git
23:57 druonysus_ joined #git
23:57 Cabanossi joined #git
23:58 adino joined #git

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