Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-10-23

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

All times shown according to UTC.

Time Nick Message
00:00 jrabe joined #git
00:01 dre_ joined #git
00:01 lorenzsj joined #git
00:01 lorenzsj hi git
00:01 gitinfo lorenzsj: 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.
00:02 upperdeck joined #git
00:02 drodger joined #git
00:02 lorenzsj lol - I've been using this computer to commit to my repo. I'm the only contributor but my email and name were messed up. Is there a way to ammend the author of all my commits that aren't offical?
00:03 rafasc lorenzsj: yes. there is a way. What do you mean by commits that aren't official?
00:04 lorenzsj the email was the name of the computer
00:04 rafasc are those commits that you didn't push yet?
00:04 lorenzsj no, they've all been pushed - i just didnt realize it
00:04 lorenzsj It's not all that big of deal - it's a hassle, it's not worth it
00:05 lorenzsj if it's a hassle*
00:06 valize joined #git
00:07 rafasc it isn't, but it will rewrite all commits. And that comes with all the issues of rewriting history.
00:07 lorenzsj as in, doubling commits?
00:07 rafasc The command you're looking for is filter-branch.
00:07 nooooone joined #git
00:08 rafasc yes, more or less. The old commit would be unreachable and later discarded by the garbage colector.
00:08 nick123 joined #git
00:08 goksinen joined #git
00:08 lorenzsj probably better to just resolve the issue and not worry about it then
00:09 rafasc you wouldn't have duplicate commits in your history. But you would need to replace every single commit.
00:10 upperdeck joined #git
00:10 rafasc https://help.github.com/articles/changing-author-info/
00:11 rafasc is the history linear?
00:11 lorenzsj no
00:11 rafasc I think $git rebase --no-ff will rewrite every commit. Even those set to pick.
00:11 sunri5e joined #git
00:12 rafasc Not sure if something like $git rebase --no-ff --preserve-merges would do it.
00:13 rafasc lorenzsj: but using filter-branch is the correct way to go about it.
00:13 lorenzsj i'll give it a shot
00:13 lorenzsj thank you
00:19 asphyxia joined #git
00:20 peterbecich joined #git
00:21 goksinen joined #git
00:22 slimetrap joined #git
00:22 rafasc lorenzsj: the git rebase version command I gave you is missing the revision. But it's probably a bad idea anyways. Try the solution on the github guide instead.
00:22 TheVice joined #git
00:24 netj joined #git
00:24 TheVice left #git
00:25 upperdeck joined #git
00:26 Cabanossi joined #git
00:26 qpdb joined #git
00:27 GrailKn1ght joined #git
00:31 nooooone_ joined #git
00:33 alyce joined #git
00:34 nooooone__ joined #git
00:34 goksinen joined #git
00:35 upperdeck joined #git
00:41 durham joined #git
00:45 peterbecich joined #git
00:49 upperdeck joined #git
00:54 goksinen joined #git
00:54 Feral_Cat joined #git
00:55 cam27 joined #git
01:02 goksinen joined #git
01:04 cqi joined #git
01:06 upperdeck joined #git
01:07 dreiss joined #git
01:08 nowhereman joined #git
01:09 goksinen joined #git
01:12 finalbeta joined #git
01:12 upperdeck joined #git
01:16 cdg joined #git
01:17 pbrewczynski joined #git
01:21 Orbixx[m] joined #git
01:22 nahtnam joined #git
01:22 robogoat joined #git
01:23 fjasle joined #git
01:23 plongshot joined #git
01:23 upperdeck joined #git
01:23 shinnya joined #git
01:26 goksinen joined #git
01:27 |ifei5g00d joined #git
01:28 unine parsnip: sorry the delay, nope, I don't care
01:28 unine just want to remove that commit but preferably without getting rid of the whole shebang
01:29 rafasc is that commit on the tip of the branch?
01:29 parsnip unine: in general you can change old commits, and it will have a cascading effect that you may not notice but for all the changed hashes.
01:29 unine nope
01:29 upperdeck joined #git
01:29 rafasc unine: https://help.github.com/articles/changing-author-info/
01:30 parsnip so the shape of the commit graph doesn't need to change, but anyone with their own divergent branches or continuation of work will be very frustrated to reconcile the difference in all the hashes.
01:30 parsnip rebase -i is one way
01:30 parsnip iiuc
01:30 d^sh joined #git
01:31 unine rafasc: thank you. parsnip currently there are no forks/branches other than the branch of the commit i want to get rid of
01:32 parsnip unine: any commit that can reach that commit by parents of parents of ... etc, their hash will change, so they will be a "new" commit (the tree object that the commit points to will still be the same)
01:32 madewokherd joined #git
01:34 rafasc parsnip: rebase will probably change the commiter, but not the author.
01:34 parsnip unine: you can use filter-branch (rafasc linked to an example) or `git rebase -i`.
01:34 parsnip oh
01:34 unine i'll try that github script
01:34 rafasc so he will probably need to rebase, chose edit. Do commit --amend --author <new author>; git rebase --continue.
01:34 rafasc unine: that will change EVERY commit.
01:35 upperdeck joined #git
01:35 unine well, that's fine isn't it?
01:35 parsnip ah, internet says you can use rebase -i with commit --amend --author='...'
01:35 rafasc depends.
01:36 rafasc parsnip: already said that above :)
01:36 parsnip ah, was typing
01:37 parsnip and the obligatory !backup
01:37 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
01:37 kfoonamalik joined #git
01:37 sunri5e joined #git
01:37 rafasc unine: if the every commit is already yours. It won't matter. But if you're on a shared repo, you'll steal commits from everyone.
01:37 DataComputist joined #git
01:37 neil_cn joined #git
01:38 parsnip filter-branch is like napalm
01:38 parsnip if i may be so insensitive :(
01:38 robogoat joined #git
01:39 goksinen joined #git
01:39 peterbecich joined #git
01:39 Topic for #git is now Welcome to #git, the place for git help and poltergits | Public logs at https://gitirc.eu/log | First visit? Read: https://gitirc.eu | Current stable version: 2.14.2 | Getting "cannot send to channel"? /msg gitinfo .voice | 8 days until the skeleton war
01:42 kfoonama1ik joined #git
01:54 upperdeck joined #git
01:54 stan_man_can left #git
01:54 unine rafasc: stealing commits is good too :) haha, just kidding, i just ran the script and it worked, but didn't change all emails, only the ones configured in the script
01:57 ilbot3 joined #git
01:57 Topic for #git is now Welcome to #git, the place for git help and poltergits | Public logs at https://gitirc.eu/log | First visit? Read: https://gitirc.eu | Current stable version: 2.14.2 | Getting "cannot send to channel"? /msg gitinfo .voice | 8 days until the skeleton war
01:57 unine parsnip: well honestly I generally use sourcetree so let's hope it handles that in a safe way for me :)
01:58 converge joined #git
01:58 goksinen joined #git
01:58 unine judging by the overall quality of it i'm not very hopeful though :P
01:58 maheshpec joined #git
01:59 neel joined #git
01:59 parsnip unine: is it here: `~\AppData\Local\Atlassian\SourceTree\passwd`?
02:01 neel joined #git
02:02 neel joined #git
02:02 jrm pijul (https://pijul.org/) sounds like an interesting git alternative
02:03 unine parsnip: hes, but not in plain text
02:03 unine yes*
02:03 piremi joined #git
02:04 unine git has a 'git gui' application though, i'll give it a try
02:04 rafasc real man use $watch -cn.1 'git log --graph --online --all --color'
02:04 rafasc :^)
02:06 cdg joined #git
02:07 unine rafasc: lol.... actually git-gui although it looks ugly seems to do the job, i'll try it for a while
02:07 rafasc unine: also gitk
02:08 lagothri1 joined #git
02:09 dreiss joined #git
02:10 unine rafasc: yep, git-gui launches gitk to view the history
02:10 unine not the prettiest thing in the world but apparently is usable
02:11 rafasc yeah
02:11 rafasc I often use gitk to browse for dangling commits.
02:11 unine any ideas git-gui handles github/bitbucket login?
02:11 rafasc with something like this $gitk --all --argscmd="git fsck --unreachable --dangling --no-reflogs | cut -d$' ' -f3"
02:14 YuGiOhJCJ joined #git
02:14 goksinen joined #git
02:15 nooooone joined #git
02:15 unine oh wow... git-gui works a treat
02:16 upperdeck joined #git
02:16 unine what's that sign-off thing though? does that only add that note to the commit message or does it do something else?
02:21 nooooone joined #git
02:21 dre_ joined #git
02:21 chachasmooth joined #git
02:25 Cabanossi joined #git
02:26 goksinen joined #git
02:28 Mattiaslndstrm joined #git
02:29 tarkus joined #git
02:31 upperdeck joined #git
02:32 nydel07 joined #git
02:33 einnjo joined #git
02:34 str joined #git
02:38 upperdeck joined #git
02:38 diogenese joined #git
02:39 fahadash joined #git
02:44 Oatmeal joined #git
02:45 alyce joined #git
02:45 Milos joined #git
02:46 nooooone_ joined #git
02:46 goksinen joined #git
02:46 white_knight joined #git
02:52 justanotheruser joined #git
02:52 fstd_ joined #git
02:53 justanotheruser joined #git
02:57 upperdeck joined #git
03:01 gtristan joined #git
03:01 goksinen joined #git
03:02 jstimm joined #git
03:04 dreiss joined #git
03:05 pks joined #git
03:06 govg joined #git
03:10 upperdeck joined #git
03:11 piremi joined #git
03:13 goksinen joined #git
03:16 nooooone joined #git
03:18 upperdeck joined #git
03:22 Bobdude joined #git
03:23 Murr joined #git
03:24 svm_invictvs joined #git
03:25 Bobguy joined #git
03:28 shoogz joined #git
03:30 goksinen joined #git
03:31 Andrew_K joined #git
03:32 subnet257 joined #git
03:35 asphyxia joined #git
03:38 diogenese joined #git
03:39 Cabanoss- joined #git
03:39 steeze joined #git
03:43 upperdeck joined #git
03:44 peterbecich joined #git
03:44 Random832 joined #git
03:44 goksinen joined #git
03:47 diqidoq joined #git
03:47 cdg joined #git
03:52 Raed|Laptop joined #git
03:55 |ifei5g00d joined #git
03:56 goksinen joined #git
03:57 alyce joined #git
04:00 cqi joined #git
04:00 Rodya_ joined #git
04:00 goweol joined #git
04:00 piremi joined #git
04:01 Dragon092 joined #git
04:03 cam27 joined #git
04:03 bumblebeer joined #git
04:03 upperdeck joined #git
04:05 asphyxia joined #git
04:05 dreiss joined #git
04:07 goksinen joined #git
04:10 surfist_ joined #git
04:19 cathedral joined #git
04:19 Bobdude joined #git
04:21 durham joined #git
04:21 igemnace joined #git
04:21 goksinen joined #git
04:24 OutreEE joined #git
04:25 mud Just that. It has meaning in certain projects (most notably linux kernel)
04:26 lexruee joined #git
04:26 ti2 joined #git
04:27 shoogz joined #git
04:29 Nizumzen joined #git
04:30 nooooone joined #git
04:30 shoogz joined #git
04:32 schleppel joined #git
04:35 safe joined #git
04:35 luckman212 joined #git
04:43 upperdeck joined #git
04:48 knop_ joined #git
05:02 ankush joined #git
05:02 bumblebeer joined #git
05:04 upperdeck joined #git
05:07 imack joined #git
05:08 piremi joined #git
05:10 tarkus joined #git
05:12 _kk22kk_ joined #git
05:17 upperdeck joined #git
05:18 xender left #git
05:21 ankush joined #git
05:23 Murr joined #git
05:31 StephenS joined #git
05:31 StephenS joined #git
05:33 Dragon092 joined #git
05:33 _kk22kk_ joined #git
05:35 tarkus joined #git
05:36 subnet257 joined #git
05:36 slimetrap joined #git
05:41 cur8or joined #git
05:41 fatihgenc joined #git
05:45 Garen joined #git
05:46 ochorocho joined #git
05:46 subnet257 joined #git
05:52 cqi joined #git
05:53 Raging_Hog joined #git
05:56 Saichovsky joined #git
06:00 diogenese joined #git
06:01 freimatz joined #git
06:03 drodger joined #git
06:03 upperdeck joined #git
06:08 JeroenT joined #git
06:08 dec0n joined #git
06:08 JeroenT_ joined #git
06:09 nisseni joined #git
06:09 rscata joined #git
06:10 JeroenT_ joined #git
06:11 luckman212 joined #git
06:12 Bobguy joined #git
06:13 surfist joined #git
06:13 Puffball joined #git
06:19 Rodya_ joined #git
06:19 chele joined #git
06:24 sunri5e joined #git
06:25 luckman212 joined #git
06:27 abba joined #git
06:28 fatihgenc joined #git
06:35 ogny joined #git
06:35 ogny joined #git
06:40 Cabanossi joined #git
06:41 Raging_Hog joined #git
06:41 dreiss joined #git
06:41 opalepatrick joined #git
06:43 nooooone_ joined #git
06:44 zefferno joined #git
06:45 nooooone_ joined #git
06:45 alyce joined #git
06:51 str joined #git
06:52 tkore joined #git
06:54 wandering_vagran joined #git
06:58 nd joined #git
06:59 justkitsu joined #git
07:00 heftig joined #git
07:01 Rodya_ joined #git
07:03 nd- joined #git
07:06 JanC_ joined #git
07:07 nd joined #git
07:08 LocaMocha joined #git
07:09 _ikke_ .version
07:09 Topic for #git is now Welcome to #git, the place for git help and poltergits | Public logs at https://gitirc.eu/log | First visit? Read: https://gitirc.eu | Current stable version: 2.14.3 | Getting "cannot send to channel"? /msg gitinfo .voice | 8 days until the skeleton war
07:09 neel joined #git
07:09 Rodya_ joined #git
07:10 bumblebeer joined #git
07:11 jaggz joined #git
07:12 rominronin joined #git
07:13 nd joined #git
07:13 ki0_ joined #git
07:15 bumblebeer joined #git
07:19 asphyxia joined #git
07:20 luckman212 joined #git
07:21 ankush left #git
07:23 Murr joined #git
07:24 Guest1587 joined #git
07:24 Abbott left #git
07:24 |ifei5g00d joined #git
07:25 cur8or joined #git
07:26 jameser joined #git
07:28 ertes-w joined #git
07:28 tarkus joined #git
07:30 jceb joined #git
07:31 Disgeae joined #git
07:31 raynold joined #git
07:32 dege joined #git
07:35 upperdeck joined #git
07:40 Cabanossi joined #git
07:41 dege_ joined #git
07:41 oxymoron93 joined #git
07:44 |ifei5g00d joined #git
07:45 rtjure joined #git
07:45 neel joined #git
07:46 paul424 joined #git
07:47 eduardas_m joined #git
07:47 neel joined #git
07:48 cdg joined #git
07:48 jcarpenter2 joined #git
07:48 Sonderblade joined #git
07:48 neel joined #git
07:48 jcarpenter2 what's the skeleton war?
07:49 _ikke_ Ask Eugene
07:49 neel joined #git
07:50 osse http://knowyourmeme.com/memes/events/skeleton-war
07:51 neel joined #git
07:51 neel joined #git
07:52 Dragon092_ joined #git
07:52 gfurlan joined #git
07:52 jcarpenter2 oh, so it's just a bunch of people posting pictures of doots
07:52 f2k joined #git
07:52 oxymoron93 I like how peaks have that exponential fall over years...
07:53 jcarpenter2 no programming competition
07:53 xaa joined #git
07:53 goksinen joined #git
07:56 mikecmpbll joined #git
07:56 masuberu joined #git
08:00 DaGoaty joined #git
08:01 blackmesa1 joined #git
08:03 Floflobel joined #git
08:06 lks joined #git
08:06 pbandark joined #git
08:07 ki0 joined #git
08:08 lss8 joined #git
08:08 gfurlan joined #git
08:09 pxfgod joined #git
08:09 pxfgod how to write a merge-driver for git?
08:10 Cabanossi joined #git
08:10 osse pxfgod: man gitattributes, see "Defining a custom merge driver"
08:10 gitinfo pxfgod: the gitattributes manpage is available at https://gitirc.eu/gitattributes.html
08:10 luckman212 joined #git
08:11 Dragon092_ joined #git
08:11 ki0 joined #git
08:13 DaGoaty joined #git
08:13 jsears joined #git
08:14 upperdeck joined #git
08:15 seschwar joined #git
08:16 asphyxia joined #git
08:17 pxfgod osse, gitinfo, thx a lot
08:17 gitinfo pxfgod: you're welcome, but please note that I'm a bot. I'm not programmed to care.
08:17 riemann joined #git
08:19 koell joined #git
08:19 pxfgod left #git
08:20 canton7 I love how often that response gets triggered
08:20 rtjure joined #git
08:22 King_Hual joined #git
08:22 overlord_tm joined #git
08:22 |ifei5good joined #git
08:22 chinandy joined #git
08:23 Gurkenglas joined #git
08:24 Oatmeal joined #git
08:24 cdown joined #git
08:28 jediraster joined #git
08:30 chinandy_ joined #git
08:30 rtjure joined #git
08:31 ki0 joined #git
08:32 ki0 joined #git
08:35 joakimk joined #git
08:36 joakimk I'm having problems pushing to github: "git subtree push --prefix dist origin gh-pages" gives me "error: failed to push some refs to https://github.com/joakimgk/angular-datepicker.git Updated were rejected because a pushed branch tip is begind its remote counterpart"
08:37 joakimk What do I do to handle this? The push is meant to "publish a demo page" to my GitHub Pages (I was asked to do this by the repo owner, so he can have a look at my contribution)
08:37 upperdeck joined #git
08:39 joakimk so I have a fork of his repo, which I've cloned to my laptop and made local changes. When I push, I'm pushing to *my* remote (github). Do I need to make sure that my remote origin (github) is up-to-date with *his* remote (github) before being able to push etc??
08:40 rtjure joined #git
08:41 goksinen joined #git
08:47 koell joined #git
08:48 aielima joined #git
08:53 Soni joined #git
08:54 ox3ch0 joined #git
08:54 rtjure joined #git
08:55 MarioBranco joined #git
08:55 |ifei5g00d joined #git
08:56 _ADN_ joined #git
08:56 goksinen joined #git
08:56 murii joined #git
08:58 overlord_tm joined #git
09:03 |ifei5good joined #git
09:05 jnavila joined #git
09:07 rtjure joined #git
09:08 Floflobel joined #git
09:10 mariobranco__ joined #git
09:10 asphyxia joined #git
09:11 floppydh joined #git
09:11 cstratak joined #git
09:13 venkat joined #git
09:14 _ADN_ joined #git
09:15 goksinen joined #git
09:15 Tobbi joined #git
09:15 cdown joined #git
09:17 rtjure joined #git
09:17 asphyxia joined #git
09:17 cdown joined #git
09:20 jakesyl_mobile joined #git
09:22 demego joined #git
09:23 Murr joined #git
09:28 cdown_ joined #git
09:29 rtjure joined #git
09:31 cdown__ joined #git
09:31 SwiftMatt joined #git
09:32 robW_HK joined #git
09:36 gfurlan_ joined #git
09:38 blackmesa1 joined #git
09:39 rvgate joined #git
09:40 MarioBranco joined #git
09:41 rtjure joined #git
09:48 venkat joined #git
09:52 luckman212 joined #git
09:52 Raging_Hog joined #git
09:52 rtjure joined #git
09:54 cyber37 joined #git
09:57 Puffball joined #git
09:58 JeroenT joined #git
10:03 JeroenT_ joined #git
10:03 rtjure joined #git
10:04 Puffball joined #git
10:09 re1 joined #git
10:12 neel joined #git
10:16 ki0 joined #git
10:16 rtjure joined #git
10:19 Milos joined #git
10:23 goksinen joined #git
10:24 Cabanossi joined #git
10:25 rtjure joined #git
10:26 Cavallari joined #git
10:26 goksinen_ joined #git
10:27 przemoc joined #git
10:28 netj joined #git
10:30 JeroenT joined #git
10:31 goksinen joined #git
10:35 rtjure joined #git
10:36 koell joined #git
10:36 leeN joined #git
10:37 sdnv joined #git
10:38 goksinen joined #git
10:39 leeN joined #git
10:41 mattheweigand joined #git
10:46 blackmesa1 joined #git
10:46 Fernando-Basso joined #git
10:48 re1_ joined #git
10:49 feodaron joined #git
10:50 linuxconformer joined #git
10:50 linuxconformer anyone here?
10:50 oxymoron93 yes
10:51 oxymoron93 but don't count on me :D
10:51 tmm88 joined #git
10:51 tmm88 joined #git
10:52 linuxconformer i've got a very strange website setup (inherited) where every change is saved as a new branch, now the issue is the repo is getting very large which is starting to cause some long load times, i'm wondering what's the best way of pruning the repo to reduce its size, without causing issues in the application
10:52 grimdox joined #git
10:52 _ikke_ Did you already try the typical house keeping (Running git gc, etc)?
10:53 bin joined #git
10:53 linuxconformer _ikke_: ah no i haven't, didn't realize there were such tools
10:54 cdg joined #git
10:55 rtjure joined #git
10:56 ochorocho joined #git
11:00 grimdox_ joined #git
11:00 alyce joined #git
11:00 nedbat linuxconformer: when you say load times, you mean the time to load the site in the browser? or something else?
11:04 rtjure joined #git
11:05 GavinMagnus joined #git
11:06 GavinMagnus left #git
11:06 Spydar007 joined #git
11:07 tvw joined #git
11:08 boxrick Good afternoon folks. I am ssh'd to a jumpbox using 'ssh -A'. When I use a git clone from a repo using ssh from my own machine it is fine, when I do it from the jumpbox I get 'Permission denied (publickey).'
11:08 boxrick If I place my public key on the jumpbox in the standard place, it weirdly works. Any explanation?
11:09 boxrick so public key living in /USER/.ssh/id_rsa.pub
11:09 upperdeck joined #git
11:10 finalbeta joined #git
11:12 Tobbi joined #git
11:12 _ikke_ boxrick: can you check what ssh -vv <host> returs?
11:14 nooooone__ joined #git
11:14 boxrick In this particular case I am trying to clone from Github, but I think it still provides output
11:14 boxrick One moment
11:15 osse does agent forwarding mean that keys are magically supposed to work?
11:15 osse do you use the same key to access the jumpbox as you do to access github ?
11:16 _ikke_ Do you use the same user?
11:16 boxrick Yes and it also works for getting onto a further box beyond the jumpbox
11:16 boxrick This case I am just playing with root, but I changed the ssh config to use git@
11:16 because joined #git
11:17 rtjure joined #git
11:17 boxrick Well 'ForwardAgent yes
11:17 boxrick User git'
11:17 |ifei5g00d joined #git
11:18 gurrkiin if I have commits (chronologically)  A-B-C-D-E and I want to combine B into D (A-C-D-E), is this a cherry-pick?
11:18 oxymoron93 it is squash, kinda
11:19 _ikke_ interactive rebase reorder + squash
11:19 gurrkiin i thought squash just squashes into the next commit ?
11:19 gurrkiin ah i didnt know rebase had reorder
11:19 _ikke_ That's why you need to reorder
11:19 oxymoron93 well yeah, with reorder you will squash it into D
11:19 gurrkiin so i reorder to A C B D E
11:19 _ikke_ You can do it in one step
11:20 _ikke_ mark D as squash
11:20 tobiasvl !interactive_rebase
11:20 gitinfo Interactive rebase sounds similar to rebase but has completely different abilities. It can do this to commits: change the order, squash some of them together, remove some, add random existing commits from other branches, split them, and more... it's very powerful. Documentation is in the section "Interactive Rebase" in 'man git-rebase'.
11:21 gurrkiin oh i literally just move the lines in the "rebase-file" around to reorder?
11:21 tobiasvl yep
11:21 gurrkiin I was looking for a reorder keyword
11:21 gurrkiin cool
11:21 oxymoron93 :D
11:21 boxrick In this case these are the differences of the ssh output one with *only* the private key present for the user which works, the other is with no keys at all present.
11:21 boxrick https://gist.github.com/boxrick/aea1c40c82a7396c373f9d61e89ab14b
11:21 boxrick then using ssh -A to connect to this box.
11:21 tobiasvl gurrkiin: just move them and put squash in front of the one you want to squash
11:22 Murr joined #git
11:23 gurrkiin ez
11:23 gurrkiin now for the real magic - force pushing to my remote
11:24 Cabanossi joined #git
11:25 Isla_de_Muerte joined #git
11:27 upperdeck joined #git
11:28 rtjure joined #git
11:28 linuxconformer nedbat: yeah the time to load the site, im not too familiar with the backend code, but everytime a commit is made to master, (i believe) the whole repo is loaded, and as it's 1.5GB in size, it takes a few minutes before the website is usable
11:29 linuxconformer _ikke_: git gc managed to reduce the repo size by about 8%, thanks for the tip
11:30 cdown joined #git
11:31 boxrick IdentitiesOnly <-- was in the config which explains it. <_ikke_> thanks for the -vv suggestion
11:33 _ikke_ boxrick: np
11:34 upperdeck joined #git
11:35 whyme joined #git
11:36 carl--- joined #git
11:37 fmcgeough joined #git
11:37 carl--- can i not if there is a gitconfig with another username use my own username when doing git pull
11:37 carl--- so that i can force git pull --user.name somehow
11:38 rtjure joined #git
11:41 Acerific joined #git
11:41 Acerific joined #git
11:43 osse carl---: git -c user.name=blah pull
11:44 malt3 joined #git
11:45 floppydh joined #git
11:46 carl--- osse, aah . thanks
11:46 whyme joined #git
11:47 BernhardPosselt joined #git
11:47 BernhardPosselt hi, we want to add a commit after a certain tag (hotfix)
11:47 BernhardPosselt is that possible?
11:47 BernhardPosselt same branch
11:48 rtjure joined #git
11:48 cbreak BernhardPosselt: sure
11:48 BernhardPosselt git checkout hash, git commit, git tag?
11:48 cbreak BernhardPosselt: but obviously you can't insert commits between other commits
11:49 cbreak just after
11:49 cbreak BernhardPosselt: make a branch for the hotfix
11:49 cbreak git checkout -b myhotfix somecommit
11:49 cbreak then commit it there
11:49 cbreak somecommit is the tag after which you want to add your changes
11:50 piremi joined #git
11:58 rtjure joined #git
12:00 Raging_Hog joined #git
12:03 SkyPatrol joined #git
12:06 BernhardPosselt thanks!
12:10 |ifei5g00d joined #git
12:10 rtjure joined #git
12:13 upperdeck joined #git
12:13 pbrewczynski joined #git
12:15 venkat joined #git
12:21 rtjure joined #git
12:22 ShekharReddy joined #git
12:25 asphyxia joined #git
12:27 str joined #git
12:30 clickjack joined #git
12:31 Mattiaslndstrm joined #git
12:35 whyme joined #git
12:36 ssarah joined #git
12:37 koell joined #git
12:39 phluks joined #git
12:39 NightMonkey joined #git
12:40 synthroid joined #git
12:41 nikivi joined #git
12:44 alyce joined #git
12:49 ahrs joined #git
12:49 venkat_ joined #git
12:50 raffo joined #git
12:50 chakra joined #git
12:55 fission6 joined #git
12:55 ogny joined #git
12:59 asphyxia joined #git
13:01 sbeex joined #git
13:01 jsears_ joined #git
13:02 sbeex hello ! Is it possible to push to a new branch ? Suppose I am on develop (local) and I want to push to release/1.1.1 (this branch does not exists currently) ? Thank you
13:03 nowhereman joined #git
13:05 osse sbeex: yes
13:05 osse git push origin develop:refs/heads/release/1.1.1
13:05 oxymoron93 just add the name of the destination to git push
13:05 oxymoron93 oh boi
13:06 sbeex osse: that easy? haha I was looking for something more cmplicated >< thank you ! :D
13:06 osse sbeex: if you want it more complicated you can solve a rubik's cude with your feet at the same time
13:07 Jack678 joined #git
13:07 cdown_ joined #git
13:07 sbeex osse: haha developers sarcasm has no limits :)
13:09 venkat joined #git
13:13 CussBot joined #git
13:14 tkore\ joined #git
13:15 czart_ How can I check what refs (commits) are present at remote repository? Locally, I can use 'git reflog'. I need this to delete commits that I do not need anymore permanently.
13:16 cdg joined #git
13:17 tlaxkit joined #git
13:17 relipse__ joined #git
13:18 chinandy joined #git
13:18 oxymoron93 with git fetch you can update your remote branches to resemble situation on remote repo
13:18 jceb joined #git
13:19 relipse___ joined #git
13:19 whyme joined #git
13:20 czart_ oxymoron93: Thanks, will check out this.
13:24 gugah joined #git
13:26 selckin joined #git
13:27 nedbat czart_: why delete commits?
13:28 osse to deny involvement
13:29 Es0teric joined #git
13:29 czart_ nedbat: for example Redmine (project management) refers to commits that are not in master anymore (e.g. due to amendment, rebase, etc.). I wish to delete them, so that Redmine do not refers to unused commits.
13:29 jnavila joined #git
13:30 czart_ Redmine (project management software) does not refer*
13:30 cdg_ joined #git
13:31 oxymoron93 it sounds like more of a Redmine problematic, not something that should git worry about
13:32 nedbat czart_: what does that mean, it refers to commits?
13:32 uebera|| joined #git
13:32 uebera|| joined #git
13:33 MeltedLux joined #git
13:33 jast czart_: the Git protocol prevents you from accessing unreferenced objects, so the only way to do that is any specific thing your server/hosting offers/allows
13:33 finalbeta joined #git
13:33 ogny joined #git
13:34 czart_ nedbat: it seems like it shows (in a web interface) even those commits that are not in master anymore (like a historical data, and so-called "dangling" commits). If I find out why it shows a full list, I will let you know :)
13:34 nedbat czart_: i agree with oxymoron93 that this seems like a Redmine issue, not something to solve by deleting history in git
13:36 czart_ nedbat: yeah, it might be that Redmine records some data, and then does not show the actual state of repo...
13:37 czart_ jast: OK, thanks for a tip.
13:37 czart_ oxymoron93: yeah, I will investigate that too.
13:37 jast I'm not sure how Redmine's Git integration works, I've never used it... but I suspect they fetch the data from the repository
13:37 whyme joined #git
13:37 jast so they have a local copy
13:38 miklcct_ joined #git
13:39 Cabanossi joined #git
13:40 gareppa joined #git
13:41 czart_ jast: ah yeah, you may be totally right. I will try to find the local repo Redmine fetched and clean reflog for that repo...
13:42 sbeex osse: git push origin develop:refs/heads/release/1.1.1 doesnt works it says "develop" is not know I guess my CI tools didn't do a pure "git clone"
13:42 sbeex can i simple write :dest without from ?
13:42 osse no
13:42 osse you can use HEAD:refs...
13:43 osse empty from part means delete
13:43 sbeex osse: ah yes not a good idea then :)
13:43 blackmesa1 joined #git
13:43 pR0Ps joined #git
13:44 EvilPigeon joined #git
13:44 guideline joined #git
13:45 diqidoq joined #git
13:45 ochorocho1 joined #git
13:45 sbeex osse -> error: src refspec develop does not match any.
13:45 osse was that the error after using HEAD ?
13:46 sbeex osse: yes I did: git push origin develop:HEAD/refs/release/1.1.1
13:46 YuGiOhJCJ joined #git
13:46 oxymoron93 not that
13:46 jast HEAD:refs/heads/release/1.1.1
13:46 jast HEAD = current local branch (or commit if HEAD is detached)
13:47 osse sbeex: btw, what does  git symbolic-ref HEAD  print ?
13:47 sbeex jast: worked like a charm
13:47 sbeex osse: I give a look 2 sec
13:48 Plouescat joined #git
13:49 sbeex osse: fatal: ref HEAD is not a symbolic ref
13:50 sbeex [solved] I will read some docs about HEAD/refs it seems something I need to learn more :p thank you guys
13:51 al-damiri joined #git
13:52 chachasmooth joined #git
13:52 manuelschneid3r joined #git
13:56 osse sbeex: i thought you said you were on develop
14:04 czart_ jast: I found a config file of a repo fetched by a Redmine plugin: https://pastebin.com/gjpx8syC, it's a bare repo fetched with: "fetch = +refs/*:refs/*". How can I "clean" such a bare repo?
14:06 _ikke_ possibly git prune --expire=noew
14:06 _ikke_ possibly git prune --expire=now
14:07 czart_ _ikke_: let me check...
14:09 dviola joined #git
14:09 jast czart_: bare repos typically don't have reflogs, so typically you'd 'git repack -ad' followed by 'git prune --expire=now'
14:10 Cabanossi joined #git
14:10 _ikke_ jast: Does prune only work with loose objects?
14:10 jast I guess the order doesn't matter that much, actually
14:10 jast _ikke_: yes
14:10 _ikke_ ah ok
14:10 _ikke_ Well, then the order does match, right?
14:10 jast otherwise it'd have to rewrite all packfiles that have unreferenced objects
14:10 ResidentBiscuit joined #git
14:10 jast if I read the docs correctly, repack -ad destroys unreferenced objects in packs
14:11 jast by repacking all referenced objects, then deleting the old packfiles
14:11 _ikke_ ah ok
14:11 jast -Ad, on the other hand, turns all unreferenced objects loose
14:11 jast so if you want to be safe about deleting unreferenced things, that's what you use :)
14:12 Raging_Hog joined #git
14:12 Gustavo6046 joined #git
14:12 Cabanossi joined #git
14:13 blackmesa1 joined #git
14:13 _ikke_ normally it's the other way around, lower case safe, uppercase less safe
14:14 jast -d (delete) is what makes it unsafe in the first place :)
14:14 Cabanossi joined #git
14:14 jast without -d, -A and -a do the same thing. I suspect -A simply was added after -a.
14:15 AaronMT joined #git
14:16 Cabanossi joined #git
14:17 peacememories joined #git
14:18 a3Dman joined #git
14:18 gfurlan_ joined #git
14:18 cbonobo joined #git
14:18 MarioBranco joined #git
14:20 jceb joined #git
14:25 mattcen joined #git
14:26 finalbeta joined #git
14:28 MineCoins joined #git
14:29 cbonobo joined #git
14:30 raulp joined #git
14:33 gfurlan_ joined #git
14:33 ozcanesen joined #git
14:36 Puffball joined #git
14:37 royiv_ joined #git
14:38 jt__ joined #git
14:38 MeltedLux joined #git
14:40 manuelschneid3r joined #git
14:41 Gurkenglas joined #git
14:42 bongjovi joined #git
14:42 raf1 joined #git
14:43 bsanford joined #git
14:43 re1 joined #git
14:44 stutter joined #git
14:50 kculpis joined #git
14:51 oxymoron93 joined #git
14:51 bartman joined #git
14:52 thiago joined #git
14:52 tmsmith joined #git
14:55 synthroid joined #git
14:58 wuseman joined #git
14:59 rasto joined #git
15:01 hahuang65 joined #git
15:01 Es0teric joined #git
15:04 guideline joined #git
15:06 newpy joined #git
15:06 venkat joined #git
15:07 gfurlan joined #git
15:09 flashdel joined #git
15:10 rafalcpp joined #git
15:11 lipvig joined #git
15:11 venkat joined #git
15:13 flashdel hi folks, i got a question: I have a git/gitlab installation which is on a server inhouse and i want to create a second installation which will be on a server which will be located at an external hoster. I got two repositories which should be in sync between those two places, is that possible? What would be the best way to do it?
15:14 jt__ joined #git
15:15 z1haze joined #git
15:15 lipvig so if I have separate branch for deployment, how do I avoid cherry picking and duplicated commits?
15:15 rafalcpp joined #git
15:16 jt__ joined #git
15:18 durham joined #git
15:18 cverna joined #git
15:19 jnavila joined #git
15:19 evie_hammond joined #git
15:19 stan_man_can joined #git
15:19 stan_man_can left #git
15:20 z1haze hey guys - When looking at files during a rebase/merge conflicts.. can anyway give me some tips to help speed up the process and understand what is being changed?
15:20 jsears__ joined #git
15:21 phluks joined #git
15:21 lipvig left #git
15:22 _wuseman joined #git
15:23 Murr joined #git
15:23 MeltedLux joined #git
15:25 rafalcpp joined #git
15:25 kpease_ joined #git
15:27 GavinMagnus joined #git
15:28 neel joined #git
15:29 shinnya joined #git
15:29 nick123_ joined #git
15:30 Es0teric joined #git
15:32 matoro_ joined #git
15:32 zivester joined #git
15:32 gtristan joined #git
15:32 cagomez joined #git
15:35 regedit joined #git
15:35 MineCoins joined #git
15:35 koell joined #git
15:36 King_Hual joined #git
15:37 madewokherd joined #git
15:38 YuGiOhJCJ joined #git
15:38 jottr joined #git
15:38 kpease joined #git
15:41 orbyt_ joined #git
15:42 dave0x6d joined #git
15:43 Seveas flashdel: have a post-receive hook on one sync to the other, and make the other read-only
15:43 mattheweigand joined #git
15:44 nil__ joined #git
15:45 rkta joined #git
15:45 tmsmith joined #git
15:45 nil__ Hi there, for some reason my .gitignore with contents "build" doesn't prevent the "build/*" contents from getting added in git add *. Is there a different way to write the ignore rule?
15:49 nixjdm joined #git
15:49 Seveas nil__: are the contents of build/ already tracked?
15:50 nil__ Seveas, I think so, that's probably the issue
15:51 Seveas ayup
15:51 Seveas .gitignore doesn't work for already-tracked things.
15:52 Seveas you may want to remove those files from the repository. git rm --cached build will untrack them but not remove the files from disk
15:53 nil__ Seveas, I did git checkout -- build, was that the wrong move?
15:53 Seveas no, but it doesn't help either :-)
15:54 tombusby joined #git
15:55 upperdeck joined #git
15:56 Murii|osx joined #git
15:57 MeltedLux joined #git
15:57 matsaman joined #git
15:57 ki0 joined #git
15:58 tohuw joined #git
15:58 tohuw joined #git
15:59 SwiftMatt joined #git
16:00 graff joined #git
16:00 Cabanossi joined #git
16:01 graff I'd like to make a very small patch to the musl C library using git's interface. I have cloned their repository and made my change. I really want to do this right and not make a fool of myself, would you guys walk me through the steps I need to take?
16:01 thiago joined #git
16:01 upperdeck joined #git
16:02 matsaman graff: this? https://git.musl-libc.org/cgit/musl
16:02 graff well I used:   https://www.musl-libc.org/download.html
16:03 matsaman the git clone one?
16:03 graff wait sorry. bad paste buffer
16:03 graff git clone git://git.musl-libc.org/musl
16:03 graff I used that ^
16:03 dminuoso graff: If you want to not make a fool of yourself, simply ask in #musl
16:03 graff from here  https://www.musl-libc.org/download.html
16:04 dminuoso graff: They can tell you best how contributions work for their project.
16:04 dminuoso This channel is not equipped to answer what conventions and rules some random project may have.
16:04 graff the change is just to one file, and is a grammatical fix. I really hope to use this to learn the right way to make a real professional change. So any advice even on the right way to send an email woudl also help. In the past I have had a lot of trouble using mailing list. perhaps from problems with gmail but I am not sure
16:04 matsaman graff: best ask #musl how they do it; their website is not useful
16:05 matsaman ah dminuoso beated me
16:05 svm_invictvs joined #git
16:05 codecutter joined #git
16:05 codecutter how do i undo a remote push
16:05 matsaman well you won't really learn that, as they don't seem very real professionally organized
16:05 codecutter ?
16:05 matsaman probably they're very oldschool and take patches via mailing lists
16:05 graff matsaman: really? hmm. interesting
16:05 matsaman but just ask them
16:06 dminuoso codecutter: What did you do?
16:06 paskl joined #git
16:06 matsaman modern projects using modern VCS systems have a straightforward mechanism to allow end users to submit patches
16:06 jast codecutter: by figuring out which commit was on the branch before and force pushing it back. for instance, suppose your branch 'foo' was at commit abc123 before on remote 'origin': git push --force origin abc123:refs/heads/foo
16:06 matsaman github has pull requests, etc.
16:06 dminuoso jast: NO...
16:06 dminuoso !force
16:06 gitinfo [!force_push] If you need to overwrite the history of a remote git repository (very bad idea, see !rewrite), you can do so with `git push --force-with-lease`. Note the remote server may reject this. See man git-config and search for receive.denyNonFastForwards. Best practice is for upstream servers to denyNonFastForwards.
16:06 dminuoso !rewrite
16:06 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
16:06 jast well yes
16:07 graff matsaman: i really need to get the help here, or somewhere else. for odd reasons not worth mentioning
16:07 * matsaman rolls eyes
16:07 jast matsaman: and if you want to contribute to Git, you send a patch to its mailing list :)
16:07 dminuoso graff: Then go somewhere else.
16:07 drodger joined #git
16:07 graff matsaman: it is eye rolling worthy. i deserve it
16:07 dminuoso graff: If you refuse to use the one place that can help you, then we cant help you either.
16:07 matsaman graff: "How can I send bug reports, patches, or support requests?" — https://www.musl-libc.org/faq.html
16:08 graff matsaman: so no suggestions on the right command to run to generate the patch? I can just use classis diff/patch if you think that would be better
16:08 dminuoso codecutter: Just create a new commit to revert the change, and push it.
16:08 matsaman so yeah, they use mailing list antiquity
16:08 graff matsaman: ah ok I will read that
16:08 matsaman graff: https://en.wikipedia.org/wiki/Patch_(Unix)#Usage_examples
16:08 jast graff: git format-patch, possibly even git send-email
16:08 roro256 joined #git
16:08 matsaman efficiency doesn't seem to be their jam, despite musl's rep
16:08 graff thanks jast and matsaman. let me work with this for a bit
16:09 graff matsaman: I think that it's just that fleker wants to maintain autonomous control for various reasons.
16:09 jast matsaman: efficiency isn't purely an advantage sometimes
16:09 graff perhaps due to bad experiences with other libc vendors in the far past
16:09 jesst joined #git
16:09 graff but yeah. I d see what you mean
16:10 dminuoso codecutter: Either you pushed sensitive information, in which case the information is basically compromised (push a new commit to revert, and change the credentials), or you didn't, in which case a simple additional commit is easier.
16:10 matsaman jast: when, for example, you want to be less efficient
16:10 graff may I ask another question. I seem to have problems using gmail to reply to mailing lists. what is the actual correct way to reply
16:10 dminuoso codecutter: Changing public history is often not allowed, or will make it quite obvious that someone did something silly.
16:10 dminuoso ;-)
16:10 dminuoso And forcing everybody to do work.
16:10 jast matsaman: to give just one example, code reviews make including patches less efficient
16:10 graff I don't exactly get the difference between "reply to all" "reply to sender" and "CC .."
16:11 moritz again it depends on the mailing list
16:11 matsaman jast: you're suggesting one might make contributing a patch difficult so that fewer patches are provided?
16:11 graff I mean I get it to work -- I think. But I am often left feeling that I made an error and that it was corrected by the admins
16:11 jast matsaman: what's difficult about sending an e-mail?
16:11 graff moritz: hmm yeah
16:11 matsaman what's difficult about setting up a patch submission system
16:12 matsaman they already setup a git frontend
16:12 moritz on some mailing lists, it's normal to "reply to all"
16:12 graff matsaman: would you please relax
16:12 moritz on others, you should only send to the list
16:12 matsaman graff: text doesn't relax, it's just little lines
16:12 graff I am going to place you on ignore, you're confusing me
16:12 jast depends on what you mean by patch submission system
16:13 jast something like github's pull requests? requires me to sign up for github. sending an e-mail doesn't. it's arguable which is preferrable.
16:13 graff this is a git channel, If i wanted opinions about using github I would have gone there
16:13 tldry joined #git
16:13 graff thanks jast. I will try that command. sorry to have been a bother
16:14 jast graff: good question. normally "reply to all" simply sends a response to all addresses from the previous mail's headers (all to, cc, ...)
16:14 jast I suspect gmail doesn't have specific buttons for "proper" responses to mailing lists
16:14 matsaman jast: sending an email does require you to sign up for email
16:14 jast though in git's mailing list, the convention is to reply to all
16:14 matsaman github makes you get an account, but that needn't be the case
16:14 jast matsaman: uh, no
16:14 graff jast: it seems to get very confusing. I have seen felker be critical of gmail. perhaps this is part of why
16:14 matsaman jast: yup
16:15 jast getting a response requires a mail account. sending an e-mail doesn't
16:15 jokajak joined #git
16:15 matsaman jast: that sounds like something an end user would find convenient
16:15 upperdeck joined #git
16:15 jast matsaman: at any rate I still don't see what makes e-mail more complicated than going to some arbitrary website, register yet another account, and pasting your patch there
16:16 jast sending an e-mail with git is actually easier since no copy&paste is involved
16:16 jast okay, I guess pull requests make that possible, too, if you happen to use the same hoster
16:16 matsaman git push & a pull request vs open email client, type nonsense, copy & paste like cave man, mail, nevermind security
16:16 graff Part of the reason I am able to make advanced changes to mathematics libraries is that I never assume anything is "easy"
16:16 jast but if you didn't even want to upload your branch anywhere, github is actually quite a bit more work than sending an e-mail
16:16 matsaman nevermind there are instructions just for using the mailing list, in this case
16:17 SwiftMatt joined #git
16:17 jast matsaman: git includes a tool to send patches via mail, without any copying and pasting
16:17 graff conversely, I assume that it was difficult for everyone and from that much is revealed
16:17 graff thank you all. be back later
16:17 matsaman jast: how would you send an email to the musl mailing list?
16:17 sbeex joined #git
16:17 roro256 I have this warning Untracked files:   (use "git add <file>..." to include in what will be committed)  "\n" how to remove this file ?
16:18 GavinMagnus left #git
16:18 jast matsaman: git send-email --to=musl@example abc123
16:18 matsaman I don't think that's their email address
16:18 jast you don't say
16:18 matsaman yap
16:19 durham joined #git
16:19 matsaman if your argument is that git send-email is simple if you walk through a lot of extra steps to make it more complicated, then it isn't simple
16:19 * matsaman stating them obviouses
16:19 jast roro256: if there's a strange file like that, probably you should just delete it from disk
16:19 demego joined #git
16:19 ToBeCloud joined #git
16:19 synthroid joined #git
16:20 jast matsaman: if your argument is that github pull requests are simple just because I have to register, fork, add their remote, push, and fill out a bunch of web forms, then it isn't simple
16:20 raynold joined #git
16:20 jast unless you've already done the one-time setup work. hey wait, the same is true for send-email!
16:20 sammyo joined #git
16:20 matsaman github was an example you latched onto for some reason
16:21 matsaman but yes, it's more simple to setup a github account and git clone from any of thousands of projects and do a pull request
16:21 Dalek_ joined #git
16:21 jast you were talking about "a patch submission system" but never provided any specific example, so I just came up with one
16:21 matsaman no I did provide one, github
16:21 matsaman you just got a real stranglehold on it
16:21 jast well, feel free to provide another one
16:21 matsaman okay
16:22 sammyo joined #git
16:22 orbyt_ joined #git
16:22 roro256 okay I just had to do clean -f
16:23 jast honestly I find it strange that if I ask for an example, you give one, and I give you my view on potential disadvantages of that system you specifically mentioned, somehow that means I've "latched onto" it
16:23 sammyo joined #git
16:23 sammyo joined #git
16:23 matsaman neat
16:24 sammyo joined #git
16:25 ankush joined #git
16:26 grimdox joined #git
16:27 ankush left #git
16:27 mattheweigand joined #git
16:28 venkat joined #git
16:28 Khisanth joined #git
16:30 in1t3r joined #git
16:33 venkat joined #git
16:33 upperdeck joined #git
16:33 Tobbi joined #git
16:36 Dreadnaught joined #git
16:36 Dreadnaught joined #git
16:37 rafasc joined #git
16:38 venkat joined #git
16:42 venkat joined #git
16:46 EvilPigeon left #git
16:51 fission6 joined #git
16:52 grimdox joined #git
16:53 mattheweigand joined #git
16:58 peterbecich joined #git
17:00 bumblebeer joined #git
17:00 Dalek_ joined #git
17:02 nixjdm joined #git
17:03 ankush joined #git
17:04 hahuang65 joined #git
17:05 dsdeiz joined #git
17:06 codecutter joined #git
17:07 codecutter if someone forks your public project and begins to make changes to it, is it normal for you to do a pull/push on that forked project with your original?
17:07 matsaman normal enough
17:07 matsaman probably more normal to only do pulls, though, if you find things you like in the fork
17:08 codecutter whats the protocal in this kind of situation?
17:08 matsaman it'd have to be some special use case fork for you to want to bother helping them out
17:08 matsaman the protocol for pull requests is the same everywhere, AFAIK
17:08 rafasc normal protocol is trying to figure out why the dev isn't doing PR to your repo.
17:08 matsaman request pull, done
17:09 rafasc but you're free to snatch the code without notice. Provided the license allows it.
17:09 energizer joined #git
17:10 jnavila joined #git
17:10 matsaman well, provided it is possible =P
17:10 tmm88 joined #git
17:10 tmm88 joined #git
17:10 codecutter request pull? what does that mean?
17:10 yohnnyjoe joined #git
17:11 _ikke_ man git request-pull
17:11 gitinfo the git-request-pull manpage is available at https://gitirc.eu/git-request-pull.html
17:11 codecutter actually they did open a issue
17:11 codecutter i responded by pull they repo and pushing it onto mine
17:11 codecutter thats normal, right?
17:11 matsaman pulls and pushes are normal
17:11 matsaman what's the problem?
17:11 kpease_ joined #git
17:11 feodaron joined #git
17:12 codecutter no problem just as long its the normal protocol
17:12 codecutter :)
17:12 rafasc If he has a solutions to the problem, tell him to to make a Pull request.
17:12 rafasc then you can merge that request in.
17:12 rafasc this is the "normal protocol"
17:12 _ikke_ Note that pull request is just an interface to do what you can also do manually
17:13 _ikke_ So a pull request is in no way a requirement
17:13 _ikke_ If anything, a pull request is mostly a notification
17:13 _ikke_ (with possibility to comment / review)
17:14 codecutter rafasc: i'll keep that in mind next time.
17:14 codecutter thanks
17:14 rafasc yes, true. When git is concerned. But when talking about its use in github, is a bit weird.
17:15 rafasc Usually upstream just "accepts" the PR. They don't go out into your repo and merge it themselves.
17:15 rafasc Which I think was the question.
17:15 _ikke_ rafasc: I say it's no weirder then them forking your repo and making changes
17:16 czart__ joined #git
17:16 rafasc _ikke_: But often those changes are made with a pull request in mind.
17:17 _ikke_ Sure, but that's mostly because they want to notify upstream of the chagnes
17:17 _ikke_ changes
17:17 _ikke_ If they just create a branch, upstream is usually not aware of it
17:18 _ikke_ The pull request is basically a notification saying "hey, I got this branch with a nice improvement, please merge it"
17:18 mattheweigand joined #git
17:18 _ikke_ It's not a goal on it's own
17:19 _ikke_ The only issue might be that the change was not finished yet
17:20 rafasc Also github PR also kinda of work as code review.
17:21 PioneerAxon joined #git
17:21 manuelschneid3r joined #git
17:22 cyan__ joined #git
17:23 Murr joined #git
17:23 _ikke_ yup. aready mentioned that
17:23 rafasc If it's a critical bug fix, and the author didn't issue the PR. Sure, go ahead. But usually if the author didn't issue the PR is due to some reason. That being incomplete changes/ wrong solution etc.
17:23 rafasc _ikke_: I type slowly :D
17:23 _ikke_ but if upstream already merges the branch, apparently they accepted the change
17:24 _ikke_ rafasc: right, but they *choose* to merge it. I hope you reviewed the change before you did that and decided it's mergable
17:24 rafasc yes right...
17:25 cyan__ joined #git
17:25 rafasc I usually have imcomplete/messy work published, because I don't expect people will clone from upstream and not from me.
17:26 rafasc Often I have ugly commit messages that I later rebase/squash together. If I found that upstream went in my repo and merged that mess, I would be a bit puzzled.
17:26 sdnv left #git
17:27 rafasc But yeah, as you said, it's their choice.
17:27 Es0teric joined #git
17:33 blackmesa1 joined #git
17:34 jottr joined #git
17:36 jceb joined #git
17:36 ertes joined #git
17:37 notemerson joined #git
17:38 YuGiOhJCJ joined #git
17:41 leeN joined #git
17:41 GavinMagnus joined #git
17:43 mikecmpbll joined #git
17:44 groundnuty joined #git
17:45 sat-aton joined #git
17:45 sat-aton joined #git
17:46 scorpiosec joined #git
17:47 JanC_ joined #git
17:48 energizer joined #git
17:54 einnjo joined #git
17:55 ghost43 joined #git
17:57 jnavila joined #git
17:57 tmsmith joined #git
17:57 orbyt_ joined #git
17:59 GavinMagnus left #git
18:00 blackmesa1 joined #git
18:02 _ADN_ joined #git
18:03 nixjdm joined #git
18:04 salamanderrake joined #git
18:04 gopar joined #git
18:05 cbonobo joined #git
18:05 boombatower joined #git
18:07 Akimb joined #git
18:09 synthroid joined #git
18:10 cdown joined #git
18:10 wootehfoot joined #git
18:11 Sasazuka joined #git
18:12 fission6 joined #git
18:13 nisseni joined #git
18:13 Es0teric joined #git
18:14 fission6 joined #git
18:14 upperdeck joined #git
18:14 xiongmao86 joined #git
18:15 robotroll joined #git
18:16 xiongmao86 Is there an easy way to remove a submodule added by "git submodule add repo"?
18:17 xiongmao86 I search the stackoverflow and no straight answer.
18:17 peterbecich joined #git
18:18 Raed|Laptop joined #git
18:19 cagedwisdom joined #git
18:23 Mattiaslndstrm joined #git
18:25 tvw joined #git
18:26 overlord_tm joined #git
18:26 GPenguin joined #git
18:27 JimDungo joined #git
18:28 Raed|Laptop joined #git
18:28 imack joined #git
18:28 FatalNIX given the file is executable by everyone, and the file is named post-receive in a remote git bare repo, is there any reason why post-receive may not be running when we push to the repo?
18:29 Jack678 joined #git
18:32 Gurkenglas joined #git
18:33 kpease joined #git
18:33 tunekey joined #git
18:33 upperdeck joined #git
18:34 SwiftMatt joined #git
18:34 pbrewczynski joined #git
18:38 Nizumzen joined #git
18:39 nick123 joined #git
18:42 lormiz joined #git
18:43 cdown joined #git
18:44 upperdeck joined #git
18:44 spexidev joined #git
18:45 basiclaser joined #git
18:45 ToBeCloud joined #git
18:45 m0viefreak joined #git
18:45 Jackneilll joined #git
18:46 sybariten joined #git
18:46 sybariten so.... a couple of ignore questions
18:46 Gustavo6046 joined #git
18:47 sybariten in general, what is the approac when you are only interested in versioning files in the "root" level of the project dir? Nothing under any subdirs should be taken
18:47 kculpis joined #git
18:47 ScalaWilliam joined #git
18:49 sybariten and question number two: here is my current ignore file. I got help with this from some stackexchange post. What does the the last row with the !*/ actually do?
18:49 Fairy joined #git
18:50 cagomez joined #git
18:50 sybariten uh, sorry .. the paste     https://pastebin.com/qccGYgWQ
18:50 upperdeck joined #git
18:50 rafasc sybariten: first question: /*/**
18:51 kyan joined #git
18:51 sybariten rafasc: aha, so double asterisks has some special function?
18:52 rafasc yes, they transverse directories.
18:53 rafasc sybariten: $git help gitignore | less -p 'Two consecutive asterisks'
18:54 sybariten ok cool. Now, i did an experiment here... i have a root dir full of "*.def" files here, and as you can see from my gitignore that is what i want to version. But i also makde a temporary dir "test" and copied some of the def files into there. First of all i have not added the  /*/** yet. Git does _not_ seem to pick up the def files in those dirs though. Maybe because theres only !*.def defined.....  however it
18:54 andai joined #git
18:54 sybariten seemed to pick up on the dir itself... shouldnt git skip dirs that essentially have no files?
18:54 sybariten i understand this question may be a bit all over the place
18:55 rafasc the second question: ! negates previous rules. So you can ingore everything with *, then selectively uningore a file with !importantfile.txt
18:55 xiongmao86 Is git rm -f sufficient to remove a submodule?
18:56 sybariten rafasc: yeah but that unignore wont mean that ./somedir/importantfile.txt will be also versioned ?
18:56 |ifei5g00d joined #git
18:57 rafasc depends on the unignore rule
18:57 sybariten ok
18:57 anjen joined #git
18:58 rafasc if you want to only unignore the root one, I think you can do !/important.txt
18:58 cagomez joined #git
18:58 sybariten ok.. .thanks
18:58 nick123 joined #git
18:58 rafasc Im not 100% sure, but test it :D
18:59 rafasc xiongmao86: I think this is right: https://stackoverflow.com/a/16162000/2408130
18:59 sybariten rafasc: yeah, will do. It's always a bit tricky with these "inverse" mind-things. I may have had the wrong idea about how the rules worked, from the beginning.
19:00 rafasc git help gitignore is always there for you ;)
19:01 Akimb joined #git
19:01 xiongmao86 rafasc: Ah, I have read the pages, and there are actually tree ways to remove a submodule which I mistake as one way.
19:01 rafasc sybariten: I lied.
19:02 xiongmao86 r
19:02 xiongmao86 rafasc: thank you .
19:02 rafasc '/' doesn't anchor against the toplevel. It anchors it against the location of the .gitignore itself.
19:03 fission6 joined #git
19:03 nixjdm joined #git
19:04 max3 joined #git
19:04 alyptik joined #git
19:05 rafasc xiongmao86: that was because git evolved. When looking on S.O. pay attention to the version they are talking. Depending on the version you have, you probably only need $git deinit and $git rm
19:05 hahuang65 joined #git
19:06 _ikke_ git submodule deinit*
19:06 upperdeck joined #git
19:06 m0viefreak joined #git
19:06 max3 can someone explain to me something that i probably missed from git 101: on gitlab i have two branches (dev1, dev2). i merged dev2 into dev1 locally and pushed up. if i browse the repos all the files are the same but when i try to open a merge request (in either direction) each says the other is missing the same lines (those from the merge of dev2 into dev1). what the hell is going on
19:06 xiongmao86 rafasc: yeah, I have tried that out.
19:06 rafasc right _ikke_, thanks.
19:06 Phlogistique joined #git
19:06 bumblebeer joined #git
19:07 xiongmao86 _ikke_: thanks, rafasc: thank you again.
19:07 anjen joined #git
19:07 rafasc xiongmao86: you welcome.
19:07 rafasc 're
19:07 Acerific joined #git
19:07 Acerific joined #git
19:09 rafasc max3: if you merged and pushed what are you trying to do now?
19:10 bergice_ joined #git
19:10 max3 rafasc, i was just double checking
19:10 synthroid joined #git
19:11 SwiftMatt joined #git
19:12 gopar joined #git
19:13 bumblebeer joined #git
19:13 upperdeck joined #git
19:13 max3 rafasc, ideas?

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