IRCloggy #git 2019-07-08

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2019-07-08

← kreyren left00:02
→ tufteddeer joined00:03
→ kreyren joined00:04
← orbyt_ left00:04
→ esran joined00:09
→ orbyt_ joined00:11
→ fphilipe_ joined00:11
→ averell joined00:14
→ tomboy64 joined00:14
← acercle left00:16
← xcm left00:20
→ xcm joined00:21
← kjartan left00:30
→ oatmealraisin joined00:31
→ kjartan joined00:31
→ Sasazuka joined00:40
← fphilipe_ left00:43
← beastDiv_ left00:43
← kreyren left00:45
→ kreyren joined00:46
→ mat001 joined00:48
← Peetz0r left00:48
← thiago left00:49
→ thiago joined00:50
← MoonJaCk- left00:51
→ elsheepo joined00:51
← mat001 left00:52
← kreyren left00:55
→ kreyren joined00:55
← AtumT_ left00:57
← yyy left00:58
← kreyren left01:01
→ kreyren joined01:02
← CryptoDavid left01:04
← bashfulshell left01:10
← ephemer0l left01:11
→ ephemer0l_ joined01:11
← elichai2 left01:17
← tufteddeer left01:17
→ Peetz0r joined01:17
→ tufteddeer joined01:18
← elsheepo left01:21
→ elichai2 joined01:22
← F0rTex left01:24
→ F0rTex joined01:25
→ fstd_ joined01:49
→ seds joined01:52
← fstd left01:53
fstd_ → fstd01:53
→ ferdna joined01:54
→ cfoch joined01:54
← seds left01:58
mns_ → mns02:03
← Sasazuka left02:07
← nic-hartley left02:07
← toothe left02:10
← mowcat left02:16
→ rotaerk joined02:20
→ esrse joined02:22
→ Rubin joined02:22
→ toothe joined02:27
lagothrix → Guest3980702:27
← Guest39807 left02:27
→ lagothrix joined02:27
← Viridian left02:28
← altendky left02:28
→ envex joined02:35
← jas4711 left02:39
← Cabanossi left02:39
→ barteks2x joined02:40
→ xorander joined02:42
→ fphilipe_ joined02:43
← orcainbox left02:47
← tomboy64 left02:49
→ Cabanossi joined02:49
→ igemnace joined02:49
← valuemachine left02:50
→ valuemachine joined02:51
→ tomboy64 joined02:51
→ jas4711 joined02:55
← kjartan left03:02
→ kjartan joined03:03
← ferdna left03:08
← netj left03:14
← fphilipe_ left03:14
→ netj joined03:14
← Rubin left03:15
← rotaerk left03:15
→ rotaerk joined03:20
→ Newami joined03:21
→ Rubin joined03:22
← thiago left03:24
← kjartan left03:28
→ kjartan joined03:29
← rkta left03:34
→ rkta joined03:35
← xcm left03:38
→ xcm joined03:39
← jayjo left03:40
→ jayjo joined03:40
← elichai2 left03:47
← kjartan left03:49
→ kjartan joined03:50
→ nowhereman joined03:55
← nowhere_man left03:55
→ fjellfras joined03:57
→ martylake joined04:01
← valuemachine left04:04
← martylake_ left04:04
← cfoch left04:04
→ valuemachine joined04:04
← mrCyborg left04:05
→ mrCyborg joined04:05
← donofrio left04:06
← mud left04:08
← Anthaas left04:12
→ Anthaas joined04:12
→ durham joined04:12
← finn0 left04:13
← kjartan left04:13
→ kjartan joined04:14
→ beastDiv_ joined04:15
← fatalhalt left04:19
→ AugustusCaesar24 joined04:20
→ mat001 joined04:22
← mat001 left04:27
← publio left04:28
← kjartan left04:29
→ TheBatzMan joined04:29
→ kjartan joined04:31
→ greggerz joined04:33
← RoSoDude left04:36
← valuemachine left04:37
→ valuemachine joined04:38
← valuemachine left04:42
← durham left04:42
→ durham joined04:44
← durham left04:45
→ durham joined04:46
← durham left04:46
→ durham joined04:47
← durham left04:47
→ durham joined04:47
← durham left04:48
→ durham joined04:48
← BeerLover left04:49
← durham left04:53
→ Ori_B joined04:54
→ SwiftMatt joined04:56
← greggerz left04:56
→ matheusmoreira joined04:58
→ pks_ joined05:01
→ BeerLover joined05:02
← pks left05:02
pks_ → pks05:02
← Karyon left05:03
→ greggerz joined05:08
← BeerLover left05:11
← SwiftMatt left05:11
→ Karyon joined05:11
→ fphilipe_ joined05:13
→ valuemachine joined05:14
→ LordVoldePort joined05:17
← xcm left05:25
← kjartan left05:25
→ xcm joined05:25
→ kjartan joined05:26
→ keepLearning512 joined05:28
← keepLearning512 left05:29
→ keepLearning512 joined05:30
→ Sasazuka joined05:31
← matheusmoreira left05:34
← Karyon left05:34
→ matheusmoreira joined05:35
← AugustusCaesar24 left05:35
← agowa338 left05:42
← fphilipe_ left05:45
→ AugustusCaesar24 joined05:46
→ g00s joined05:47
→ dartmouthed joined05:50
← orbyt_ left05:57
← greggerz left06:01
← johnny56 left06:05
→ Karyon joined06:05
→ SkarmoutsosV joined06:05
← Sasazuka left06:13
← cd left06:14
← Inline left06:14
← kjartan left06:14
→ kjartan joined06:15
← xcm left06:16
→ xcm joined06:16
→ SwiftMatt joined06:19
→ Noti joined06:21
← SwiftMatt left06:23
→ salad joined06:23
→ rfold joined06:24
→ TomTom joined06:30
← SkarmoutsosV left06:31
← g00s left06:34
← nowhereman left06:44
← satifant left06:46
→ fphilipe_ joined06:47
← Newami left06:48
→ rfuentess joined06:50
→ enoq joined06:50
→ Sasazuka joined06:51
← jcbitter left06:53
→ n3wborn joined06:53
← rfold left06:54
→ jcbitter joined06:54
← fphilipe_ left06:55
→ carro7 joined06:55
→ Jackneill joined06:56
→ fphilipe_ joined06:59
→ howdoi joined07:01
→ mohsen_1 joined07:02
← carro7 left07:02
→ finn0 joined07:02
← tufteddeer left07:08
← kreyren left07:09
→ tufteddeer joined07:11
← beastDiv_ left07:11
→ kreyren joined07:11
→ ave_milia joined07:12
← fphilipe_ left07:12
← xcm left07:12
→ xcm joined07:13
→ g00s joined07:13
→ atomi joined07:15
← ave_milia left07:15
→ ave_milia joined07:15
← netj left07:16
→ netj joined07:17
← keepLearning512 left07:17
← ave_milia left07:18
→ keepLearning512 joined07:18
→ ave_milia joined07:18
→ MylesBorins joined07:19
← oatmealraisin left07:19
→ mayurvpatil joined07:22
← mayurvpatil left07:22
← keepLearning512 left07:22
→ mayurvpatil joined07:22
← mayurvpatil left07:27
→ fphilipe_ joined07:28
← g00s left07:28
→ Mattiaslndstrm joined07:29
→ g00s joined07:29
← tufteddeer left07:29
→ SkarmoutsosV joined07:29
→ tufteddeer joined07:30
← Goplat left07:31
→ agowa338 joined07:32
→ mimajof joined07:32
→ Anticom joined07:33
→ chele joined07:38
← xcm left07:44
← queip left07:47
→ xcm joined07:47
→ shored joined07:48
→ jabular joined07:48
← Mattiaslndstrm left07:50
→ agowa339 joined07:51
→ Mattiaslndstrm joined07:51
← agowa338 left07:55
← Mattiaslndstrm left07:55
← fjellfras left07:55
→ queip joined07:57
`whoami` hello, when doing a rebase i selected the wrong option (fixup instead of reword), is there a way to go backwards ? :x07:59
← g00s left08:00
jast `whoami`: git rebase --abort if the rebase is still in progress, otherwise undo it and start over08:01
`whoami` mhm, ok :/08:01
jast http://undo-failed-rebase.gitrecipes.de/08:02
`whoami` (rebase is over)08:02
ah nice, tyvm will read08:02
→ Mattiaslndstrm joined08:04
rfuentess jast: nice way to redo the branch :)08:09
I normally do the trick with git reflog, (or by pushing the changes before the rebase and just deleting my local branch)08:10
→ T_UNIX joined08:12
← martylake left08:13
← xcm left08:14
→ xcm joined08:14
→ martylake joined08:15
→ joe4 joined08:16
→ sudosays joined08:19
→ yyy joined08:22
`whoami` well, it seems `git reset --hard @{1}` fixed the issue08:26
that's awesome.. :)08:27
→ keepLearning512 joined08:27
→ floppydh joined08:33
← geekodour08 left08:35
→ Ploppz joined08:36
Ploppz I have a branch with several commits and I want to reduce it to a single commit and apply it on top of master. What is the right command?08:36
← fieldmapper left08:37
canton7 Ploppz, you can merge wish --squash?08:37
→ Nothing4You joined08:38
sudosays yeah squash the commits08:39
Ploppz ok thanks08:41
← yonder left08:42
← gde33 left08:44
← Sasazuka left08:45
→ rosco_y joined08:46
→ gde33 joined08:46
← barteks2x left08:49
→ durham joined08:50
← durham left08:54
→ wu7c7i joined08:57
← n3wborn left08:58
← Mattiaslndstrm left08:59
→ joe1 joined08:59
Ploppz many repos seem to not like merge but rather that you always rebase instead, right? So as to avoid a commit dedicated to merging. How would I do what I asked without merging?09:00
← joe4 left09:02
Ploppz I tried to checkout my branch, then `git rebase -i master` but that didn't work09:02
What I did was leave the first line with "pick" and all the rest "squash". But it did not retain most of my changes it seems. I added files that suddenly were not there anymore09:03
jast typically you'd use a non-interactive rebase (i.e. without '-i')09:03
Ploppz oh wait I see now that it tries to apply each commit individually! Ok let me try that09:03
jast squashing is an option, of course, if you have a lot of "garbage" commits that fell out of just committing randomly while you were working :)09:04
Ploppz I do want to both squash and rebase09:04
I did `git rebase master` but again the files I added are not there anymore09:04
→ johnny56 joined09:05
→ Lucas_Gray joined09:05
cdunklau Ploppz: did you merge other stuff onto your branch?09:05
← luckman212 left09:05
cdunklau or just make commits on it09:05
jast did rebase ask you about fixing conflicts? because if so, you're not actually done yet :)09:05
→ akdev joined09:07
Ploppz Yes it did ask me to fix conflicts, but I had expected that it be the conflict between my branch and master? But files from my branch were not present09:08
→ luckman212 joined09:08
Ploppz cdunklau: Earlier I did a merge from branch to master, but I reverted it with `git reset --hard @~1` but I don't think this should matte?r09:08
cdunklau Ploppz: rebase applies the changes of each commit in turn, not all at once09:09
so if you get conflicts along the way, it might be in a commit that came *before* you added those files09:09
Ploppz Hm.. maybe I shoud squash first and then rebase then09:09
cdunklau Ploppz: if you didn't merge master onto your branch, then yes that's probably the nicest option09:10
Ploppz is it right to use rebase -i for squashing a series of commits? or is there a better way?09:11
→ stdedos joined09:11
← zamba left09:12
→ medfly joined09:12
medfly Hi. I am trying to git-minus-gitlab/github/whatever. Someone replied to me saying I should make a change and with it "reviewed-by". Am I supposed to add reviewed by to my own commit message?09:13
Like what does that even mean?09:13
→ planetcall|work joined09:13
cdunklau Ploppz: i like using -i for squashing, gives more control. just gotta be careful not to remove lines09:14
← valuemachine left09:14
→ zamba joined09:15
→ valuemachine joined09:15
Ploppz thanks cdunklau !09:16
after rebase -i all commits I thought I had squashed still show up in `git log`09:16
cdunklau Ploppz: what's git status say09:17
Ploppz ah wait that's because they are still on remote. Sorry09:17
← catsup left09:18
→ joe2 joined09:18
→ catsup joined09:18
← catsup left09:19
→ catsup joined09:20
→ martylake_ joined09:20
← hussam left09:20
← joe1 left09:21
← martylake left09:24
← sudosays left09:25
→ sudosays joined09:25
← sudosays left09:26
← kjartan left09:27
→ kjartan joined09:28
← sunri5e_ left09:30
→ sunri5e joined09:31
← AugustusCaesar24 left09:33
davve hey, i messed up and need to revert commits between somehash and anotherhash09:38
i googled how to do a ranged revert but i am getting error: empty commit set passed09:38
command: git revert -n f8a139c4eea662c7f3864375279d2bf7fca059eb^..8f4447f73999c2cfdc62cb8e3527b39d8ecc0eb509:38
medfly imo it's easiest to rebase -i, but to each their own09:39
davve oh you can revert from there?09:39
I have pushed it though so I can't just rebase09:39
medfly ah I see09:39
davve ah I needed to have the older commit first09:40
→ fjellfras joined09:42
← fjellfras left09:42
→ m1sosoup joined09:43
→ fjellfras joined09:44
← xcm left09:46
→ xcm joined09:47
← unreal left09:47
→ fjellfras_ joined09:49
← buggytom left09:49
← fjellfras left09:49
← fjellfras_ left09:49
→ fjellfras_ joined09:50
→ arunkumar413 joined09:50
→ unreal joined09:54
← fjellfras_ left09:54
→ dre_ joined09:55
phinxy Did I do something wrong when there was <<<< foo and >>>>> foo, stuff in the source code after stashing, pulling and applying stash?10:01
medfly well, merge conflicts happen10:01
phinxy Is the syntax <<<< from git or from the source code?10:02
medfly from git, it's having trouble merging your changes because nearby lines have changed10:03
← serialoverflow left10:03
medfly (Although it won't stop you from committing <<<< >>> === stuff)10:03
this is a "most likely" answer10:05
→ figuinhas joined10:05
phinxy Come to think of how complex things could get when the source code of git is being worked on. Like, there must be a lots of cases where the source code is interpreted as "git commands".10:05
← lankanmon left10:05
phinxy eh, perhaps not. Double quotes and escape codes if probably suffucient10:06
→ BeerLover joined10:07
← kjartan left10:07
← yyy left10:08
→ kjartan joined10:09
→ serialoverflow joined10:09
→ fieldmapper joined10:12
→ bsanford joined10:14
← bsanford left10:14
← chele left10:16
← tufteddeer left10:16
→ tufteddeer joined10:17
→ chele joined10:18
← armyriad left10:24
→ chele_ joined10:26
← chele left10:26
← tufteddeer left10:26
→ fjellfras joined10:28
← dre_ left10:31
← mooasaurus left10:33
→ rchavik joined10:33
← Anthaas left10:37
→ Anthaas joined10:39
← xcm left10:40
→ eigenvalue joined10:41
→ xcm joined10:41
← fjellfras left10:41
← vsbogd left10:42
→ dege joined10:45
← moldybits left10:47
← valuemachine left10:49
← keepLearning512 left10:49
→ valuemachine joined10:49
→ lacrymology joined10:53
← inquisitiv3 left10:55
→ armyriad joined10:55
← ChrisLane left10:56
→ ChrisLane joined10:56
→ hussam joined10:57
xcm → Guest4256110:58
→ xcm joined10:58
← Guest42561 left10:59
← BeerLover left11:00
→ Mattiaslndstrm joined11:02
→ BeerLover joined11:02
← BeerLover left11:04
← valuemachine left11:06
→ valuemachine joined11:07
→ BeerLover joined11:10
← BeerLover left11:11
← johnny56 left11:15
→ BeerLover joined11:16
← kjartan left11:16
→ kjartan joined11:17
→ lankanmon joined11:24
→ emsjessec joined11:25
← rchavik left11:25
← arunkumar413 left11:30
← kjartan left11:30
← finn0 left11:30
→ kjartan joined11:31
← xcm left11:31
→ xcm joined11:32
← gde33 left11:37
← Lucas_Gray left11:37
← BeerLover left11:39
← SkarmoutsosV left11:39
→ Lucas_Gray joined11:40
← lacrymology left11:42
→ TheWild joined11:43
TheWild hello11:43
← medfly left11:43
TheWild I cloned (--mirror) repository A to X. There's also repository B (which is supposedly a fork?). Thing is that B contains pending pull request, commit of which is 5511ce0 and I want to set my master to it.11:45
I can't simply reset to it because my repository has no knowledge of 5511ce011:45
← fphilipe_ left11:46
→ BeerLover joined11:47
rfuentess TheWild: you need to put repository B as a remote repository11:47
with that you can run `git fetch` and you should be able to see the remote branch at B with your wished commit11:47
← Lucas_Gray left11:48
→ tcorneli joined11:48
→ fretta joined11:51
→ gde33 joined11:51
→ arunkumar413 joined11:51
→ Lucas_Gray joined11:52
→ fphilipe_ joined11:52
→ finn0 joined11:53
tcorneli Hi, i have something weird going on with git. I'm using one repository for multiple worktrees. When I pull from origin branch_a to the local branch_a in the corresponding worktree, the other worktrees are changed too, but git worktree list shows no changes in the current commit. When I reset the other worktrees and do a pull on one of them, they ar11:54
e updated correctly, but git worktree list still show no change. I am working from a bare repository if that's important.11:54
← BeerLover left12:00
→ oatmealraisin joined12:00
→ deathwishdave joined12:01
→ yo-wan joined12:02
TheWild rfuentess: oh, man. Figured that out.12:03
git fetch B refs/pull/1/head:master12:03
thanks anyway12:03
shame that it only works for branches (heads) and not commits known by hash :/12:04
Tried by hash: git fetch B 5511ce0:master12:05
but I got: error: Server does not allow request for unadvertised object 5511ce012:05
→ BeerLover joined12:05
→ johnny56 joined12:05
→ rfold joined12:09
← tcorneli left12:10
← fretta left12:12
← BeerLover left12:14
rfuentess TheWild: once that your local repository knows about B you can use cherry-pick12:14
TheWild meh, cherry-pick would alter the hash12:14
rfuentess mmm then a git merge ?12:14
→ Hi-Angel joined12:15
→ thiago joined12:16
Hi-Angel I wonder, can I rewind an interactive git-rebase a few commits back? For example, I'm doing an complex rebase over many commits, and I figured that a few commits back I made a mistake. Is it possible not to start all over again, but, kind of, rewind it back in time to that commit?12:18
rfuentess Hi-Angel: I think you could do something with reflog and reset --hard... BUT, I would not do it .12:21
Also, is not more easy to fix the problem after the rebase ? with another `git rebase -i` ?12:21
→ thefatma joined12:23
→ achen_ joined12:23
Hi-Angel Hmm, I don't think so. I gonna have complex conflicts in future commits, and I won't know if I solved them correctly if I can't build project. But I can't build it because of the mistake. I can though just do some changes that looks correct, and then do the second rebase… But I'm not sure I'd really like it, I'm not even sure that these changes gonna play out, i.e. that I not gonna have to call --abort on some commit.12:26
→ tufteddeer joined12:26
→ Sir_Cheer joined12:29
rfuentess Hi-Angel: ok, that sounds odd, are you rebasing your master branch ?12:29
Hi-Angel rfuentess, no, not master… Why does it sound odd?12:30
→ CodeSlingerPaul joined12:30
rfuentess Hi-Angel: normally when you are doing the `rebase -i` its over your own branch. So, you can run multiple times if necessary12:31
← tufteddeer left12:31
Hi-Angel Oh12:31
I didn't know it. Let me try12:31
→ tufteddeer joined12:32
← oatmealraisin left12:32
rfuentess of course, each time you are working in the current state of the branch. But if a specific commit requires extra work, then it's OK12:32
Hi-Angel rfuentess, nope, it says "fatal: It seems that there is already a rebase-merge directory, and I wonder if you are in the middle of another rebase."12:32
cdunklau Hi-Angel: what mistake did you make12:32
Sir_Cheer ".esaber rehtona fo elddim eht ni era uoy fi rednow I dna ,yrotcerid egrem-esaber a ydaerla si ereht taht smees tI :lataf" syas ti ,epon ,ssetneufr12:32
rfuentess Hi-Angel: You first need to finish your current rebase in progress (or abort it )12:33
Hi-Angel cdunklau, in short, I just didn't constify some part of code, and at future commits this causes build failure.12:34
rfuentess, ah, in this case as I said I'd rather just do --abort, and start all over12:34
cdunklau Hi-Angel: hmm, editing code in a rebase sounds like a bad idea :)12:35
Hi-Angel Okay, anyway, nvm. It's not that important, I can just do abort. I was just wondering if it's possible to rewind rebase back.12:35
cdunklau, hmm, why? This is a well established practice.12:35
Like, how are you supposed to address a review otherwise? :)12:35
cdunklau make a new commit with the fix?12:35
if you're trying to make it so every commit that hits master is compileable, squashing is probably the right thing12:37
(i'm probably assuming a whole bunch about your situation i have no business doing :D )12:37
err, to be clear, i mean squashing the branch before merging to master12:38
Hi-Angel cdunklau, if your new commit fixes previous commit, and none of them have been merged upstream, then reviewers gonna require them squashed. Now, if the same situation happens with a commit, say, HEAD~5, then you gonna need to edit the HEAD~5 commit :)12:39
Sir_Cheer (: timmoc 5~DAEH eht tide ot deen annog uoy neht ,5~DAEH ,yas ,timmoc a htiw sneppah noitautis emas eht fi ,woN .dehsauqs meht eriuqer annog sreweiver neht ,maertspu degrem neeb evah meht fo enon dna ,timmoc suoiverp sexif timmoc wen ruoy fi ,ualknudc12:39
→ exitcode1 joined12:39
cdunklau Sir_Cheer: that's pretty annoying12:39
Hi-Angel At least Sir_Cheer put a smile at the beginning :)12:40
cdunklau Hi-Angel: what advantage does editing a commit give you though?12:40
Hi-Angel cdunklau, less work for reviewers. Also, assuming you didn't do it, and they notice your mistake in HEAD~5 commit, they gonna require you to fix it anyway.12:41
cdunklau, have you contributed to open source projects?12:42
cdunklau hmm i'm still not getting why that'd be something worthwhile, especially if the commits are gonna get squashed anyway12:42
Hi-Angel: yes12:42
Hi-Angel cdunklau, squashing commits is a bad practice though.12:43
cdunklau Hi-Angel: it *can* be, sure12:43
→ inkbottle joined12:43
→ lacrymology joined12:43
rfuentess Hi-Angel: Depends :P I do a lot of --fixup, --squash and then rebase -i --autosquash12:43
cdunklau say i make a feature, and submit it as a PR. the branch has several commits from me wanting to save ongoing progress12:44
rfuentess but maybe Hi-Angel need that HEAD~5 separated because it's a specific logical change that doesn't mix with the other four commits12:44
Hi-Angel rfuentess, yeah, sure12:44
cdunklau then the maintainer does a review and wants me to fix some things. i do that, make a new commit12:44
rfuentess in that case, ammend that commit is required12:44
cdunklau then everything's okay in the whole changeset, but to keep the whole feature frrom being spread out across several commits in `master`'s history, squashing is the right approach12:45
rfuentess cdunklau: yeah, the complexity for squashing that new commit will depends if the following four commits also work on the same file :) (I normally do like this)12:46
← netj left12:46
Hi-Angel cdunklau, but here's the thing: you do want to "spread out" the feature in history. These commits are separate logical changes, they need to be kept separately so one could git-bisect, or just quickly figure out what was happening in that specific commit.12:46
→ bashfulshell joined12:47
Hi-Angel If you squash them, the history gonna have one large commit with multiple logical changes, and if someone find regression that leads to that large commit, it would be harder to figure out why was that happening12:47
→ netj joined12:47
cdunklau makes sense12:48
→ BeerLover joined12:48
cdunklau the situation i try to avoid by squashing is where the various commits from one feature branch appear interleaved with others12:48
(after merging)12:48
← BeerLover left12:49
Hi-Angel Btw, I recently read a nice blog post from libinput maintainer on commits organization, maybe it's a nice moment I can share it http://who-t.blogspot.com/2009/12/on-commit-messages.html12:50
cdunklau maybe i'm extra-sensitive to this because i just spent a week trying to figure out wtf happened with a borked-up merge12:50
Hi-Angel: oldie-but-goodie right there12:52
Hi-Angel (=12:53
rfuentess cdunklau: that issue (the search of bad commits in disorder) is more or less mitigated with git bisect (AFAIK)12:53
cdunklau rfuentess: if you have good tests :)12:53
→ exitcode_ joined12:54
cdunklau i'm mostly sour about the debacle last week because it's my fault12:54
protip: don't let PRs get stale! :D12:55
→ tcorneli joined12:55
← exitcode1 left12:56
Hi-Angel cdunklau, I'm a little unclear on what happened, but it sounds like you needed in your branch commits from another feature branch. And the only situation I can think of where you may have ended up with commits "interleaved" is if you tried to "git-merge" changes from another branch to yours. Well, in this case you ought to use "git rebase" instead of "git merge".13:01
By doing "git rebase" you always make sure that your new commits are still on the top.13:01
cdunklau Hi-Angel: the dev branched off master, did some things, submitted PR. then they branched off their PR branch, and did more13:01
then submitted another PR13:02
i reviewed the first PR, they changed stuff, i came back months later and merged it13:02
then merge conflicts showed up in the second PR13:02
Hi-Angel Oh, I see13:02
cdunklau yeah :D13:02
← magic_ninja_work left13:03
cdunklau looks like their fault, isn't though13:03
had i not left this alone for months, it would have been more obvious to the dev what they did13:04
but since i did, i had to clean up the mess, since they didn't remember what they did13:04
and they're not all that experienced13:05
← esrse left13:05
Hi-Angel That's IMO a typical situation, because sometimes you need to continue working on a feature without waiting for the older branch to get merged. They shouldn't have sent the 2nd PR until the 1st one got merged though… Anyway, what I do in such situations: just use "git cherry-pick" to pick up the new commits from the 2nd branch.13:05
cdunklau there were other issues that prevented such a clean fix13:07
i wound up just going commit-by-commit, annotating all the changes, until i was able to get the whole picture13:07
then manually fixed the conflicts in the second PR13:07
lots of ammo for my "things not to do" list, for sure :)13:08
→ BeerLover joined13:11
← joe2 left13:12
← tcorneli left13:12
→ joe2 joined13:13
← tufteddeer left13:14
← BeerLover left13:14
→ tufteddeer joined13:15
← LordVoldePort left13:19
→ mat001 joined13:19
← tufteddeer left13:20
Hi-Angel Btw, in fact right now I'm working on a code that branched off a branch that has a few weeks old merge request that nobody bothers to review.13:20
← inkbottle left13:20
← polyrob left13:21
← finn0 left13:24
→ SwiftMatt joined13:24
→ keepLearning512 joined13:26
→ BeerLover joined13:27
→ gxt joined13:29
← BeerLover left13:29
→ tufteddeer joined13:32
→ altendky joined13:32
← queip left13:33
← Mattiaslndstrm left13:33
← fphilipe_ left13:34
cdunklau yaaay13:35
→ fphilipe_ joined13:36
→ queip joined13:40
← ave_milia left13:41
rfuentess rebase over rebase over rebase : )13:42
that is where one learns to love --autosquash :)13:42
→ czart joined13:43
→ magic_ninja_work joined13:45
→ Eryn_1983_FL joined13:46
← fphilipe_ left13:46
→ Fabricio20 joined13:47
Hi-Angel rfuentess, even ignoring for a second the fact that I'd be required to break that squashed big commit back into logical changes, I don't see why squashing the commits would change much for me.13:47
rfuentess Hi-Angel: ah sorry, not for your case13:48
Hi-Angel Ah, okay13:48
← magic_ninja_work left13:48
rfuentess but having the squash! and fixup! on the PR's make more easy to force the last changes into the other PRs13:49
→ achen__ joined13:50
← mohsen_1 left13:51
← lacrymology left13:51
← achen_ left13:52
← thefatma left13:52
→ thefatma joined13:52
← xcm left13:54
Hi-Angel Hmm… I don't know, maybe I'm just not familiar with such workflow, but for me it's much easier to do this: given branch1 with PR, and a work-in-progress branch2 that is branched off branch1. If branch1 gets changed, I'd rather do (while on branch2): "git format-patch -10 --stdout > 1.patch", then "git reset --hard branch1", then "git am -3 1.patch". (and there's a less destructive way with cherry-pick…)13:55
→ xcm joined13:55
← thiago left13:55
→ fphilipe_ joined13:56
Hi-Angel (the "-10" here assumes that branch2 has 10 new commits on top of branch1)13:56
→ lacrymology joined13:56
rfuentess Hi-Angel: doesn't sound bad. I should try it :)13:57
→ thiago joined13:57
→ oxymoron93 joined13:57
← joe2 left14:00
← xcm left14:02
← sweatsuit left14:02
→ xcm joined14:03
→ joe2 joined14:05
→ sweatsuit joined14:05
→ elichai2 joined14:05
← floppydh left14:08
→ AbleBacon joined14:08
← xcm left14:10
→ xcm joined14:10
→ Es0teric joined14:11
← Lucas_Gray left14:12
← enoq left14:14
→ Lucas_Gray joined14:14
← fphilipe_ left14:17
→ ghoti joined14:17
→ AtumT joined14:21
→ fphilipe_ joined14:22
→ magic_ninja_work joined14:25
← Sir_Cheer left14:26
→ fphilipe joined14:27
→ oatmealraisin joined14:27
→ boombatower joined14:28
← fphilipe_ left14:28
→ greggerz joined14:29
→ pyfgcr joined14:34
← agowa339 left14:36
← oxymoron93 left14:37
→ davidfetter joined14:38
davidfetter hi14:38
haymawk waves hi14:39
davidfetter is there a way to parallelize fetching remotes?14:39
← lacrymology left14:39
davidfetter has an origin (always fetched first) that's a little pokey14:39
haymawk davidfetter: no I don't think you can do that14:41
https://stackoverflow.com/questions/43019999/git-pull-multiple-remotes-in-parallel14:41
davidfetter ok14:42
→ rsrx joined14:43
foo Here: https://gitolite.com/deploy.html - it says under "4.1 checkout -- the "fight club" of git deployment": you always checkout to the same deploy directory you always checkout the same branch ... I want to deploy to a different directory and/or a different branch... it seems to have worked out. Anyone else done this before and seen issues?14:47
→ Phylock joined14:48
joe2 → revoltingPeasant14:50
→ tarkus joined14:52
jast foo: it can, in theory, cause issues because the index will be updated to match each time, so git's of what files should be deleted when switching/updating a second work tree could be off14:54
*git's knowledge14:54
→ polyrob joined14:54
← xcm left14:54
jast you can work around this by also passing different index files for each work tree (GIT_INDEX_FILE I think)14:54
→ xcm joined14:55
← mimajof left14:56
→ stutter joined14:56
← bashfulshell left14:57
foo jast: thank you. In theory, as in, it can happen but rarely does? I wonder if there's a guide on how to do this in a sane way.14:57
jast well, personally I think it's enough of a problem to avoid doing this14:58
→ achen_ joined15:00
← polyrob left15:00
← mat001 left15:01
← thefatma left15:01
← achen__ left15:01
→ thefatma joined15:02
→ polyrob joined15:04
← Noti left15:04
→ theoceaniscool joined15:06
← vdl left15:08
→ SkarmoutsosV joined15:09
→ CyberX joined15:09
→ vdl joined15:10
← Anticom left15:11
foo jast: thank you for sharing. Is there *any* work around? eg. right now, I just cp -Rv 'd the whole repo to a backup, then I set GIT_WORK_TREE, git checkout -f anotherbranch ... then I moved the old git repo back... heh. Probably overkill?15:12
← tufteddeer left15:13
→ tufteddeer joined15:13
→ navidr joined15:15
← thiago left15:16
→ thiago_ joined15:16
thiago_ → thiago15:16
→ arunkumar413_ joined15:23
← deathwishdave left15:23
← tufteddeer left15:24
← raffo left15:24
→ tufteddeer joined15:24
← arunkumar413 left15:24
← keepLearning512 left15:28
→ raffo joined15:28
← wu7c7i left15:32
← SwiftMatt left15:33
← exitcode_ left15:34
← chele_ left15:34
→ yyy joined15:35
→ orbyt_ joined15:35
→ cjohnson joined15:35
→ tang^ joined15:36
← m1sosoup left15:38
sybariten so, a newbie question. I don't know if this is git or gitlab related. We have two environments for something that i would describe as config files. test and production. The files are almost identical between the environments, save for some paths or strings in them, like test.database.com and prod.database.com15:40
← tufteddeer left15:41
sybariten ideally i'd like to be able to work on everything in the test environment, do my push and pulls there, and when i see that it works, just simply check out the same things in the prod environment.... do versioning syustems have some sort of functionality that can dynamically change certain values like that?15:41
← Cabanossi left15:46
→ Cabanossi joined15:46
davve sybariten: a workflow using git branches is the norm i would say15:48
sybariten davve: aha, could you explain a little further how this would be solved or what i need to search for?15:48
← miklcct left15:49
davve usually you would let master be production, and have development in a another branch, maybe even named development or testing15:49
sybariten okay15:49
davve where you would then merge to master when you are in a happy state... maybe even use tags and consider them releases15:49
← rfuentess left15:49
→ miklcct joined15:50
← AbleBacon left15:50
← orbyt_ left15:54
sybariten ok, and.. with a simple example. If these are two files here https://pastebin.com/GRFBt65V and in the test case i have added "delta" as you can see. Could i check that in somehow, and check it out at the production place, with the delta added but the database path still being correct (prod.database.com) ?15:55
→ AbleBacon joined15:55
← TheWild left16:01
→ TheWild joined16:02
→ mat001 joined16:02
← mat001 left16:02
→ mat001 joined16:03
→ BeerLover joined16:06
osse !config16:06
gitinfo [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas16:06
← BeerLover left16:10
→ mowcat joined16:10
← thiago left16:12
→ gareppa joined16:13
← SkarmoutsosV left16:13
→ lacrymology joined16:14
← gareppa left16:14
← Lucas_Gray left16:18
→ durham joined16:20
→ nerosnm joined16:21
← tarkus left16:24
← TheWild left16:24
→ cdown_ joined16:28
← cdown left16:30
cdown_ → cdown16:30
← arunkumar413_ left16:31
← rsrx left16:34
← durham left16:34
← Es0teric left16:36
← revoltingPeasant left16:42
→ elsheepo joined16:44
← hellauer left16:45
← johnny56 left16:45
→ MACscr joined16:47
← yyy left16:48
→ TheWild joined16:51
→ bluesmonk joined16:52
← n000g left16:54
→ subopt joined16:54
← pyfgcr left16:55
→ gracecr joined16:58
→ iam730 joined16:58
→ thiago joined17:04
← greggerz left17:06
→ tarkus joined17:07
← nerosnm left17:11
← elsheepo left17:11
→ johnny56 joined17:11
→ revoltingPeasant joined17:13
← tarkus left17:14
← xcm left17:15
→ fission6 joined17:16
← Jackneill left17:16
→ xcm joined17:16
→ Es0teric joined17:18
← planetcall|work left17:21
→ gloomy joined17:22
→ cfoch joined17:23
→ n000g joined17:26
→ unlimiter joined17:26
← figuinhas left17:27
→ mohsen_1 joined17:28
← unlimiter left17:29
← czart left17:31
→ CryptoDavid joined17:33
→ orcainbox joined17:34
← fphilipe left17:35
→ Laxtour joined17:36
← Laxtour left17:36
→ Laxtour joined17:36
→ keepLearning512 joined17:37
← jwest left17:42
← fission6 left17:43
→ appll joined17:43
← lacrymology left17:44
← keepLearning512 left17:46
← Hi-Angel left17:47
→ agowa338 joined17:50
← TheWild left17:51
→ Inline joined17:54
← Laxtour left17:55
→ |Fitz| joined17:55
→ Laxtour joined17:56
← Laxtour left17:56
→ fphilipe joined17:57
→ GeraldW joined17:57
→ rsrx joined17:57
← fphilipe left18:01
→ fission6 joined18:05
→ Dirak joined18:06
Dirak is it possible to have two git branches open at the same time on a single computer?18:06
My usecase is I have a long running process that requires me to be checked into one branch, and I want to work on another18:07
cdunklau Dirak: clone the repo to somewhere else on the machine18:08
Dirak ah good idea. ty!18:08
← agowa338 left18:10
→ SwiftMatt joined18:10
bin_bash just dont forget that you did that ;)18:10
← Cabanossi left18:14
→ jwest joined18:14
→ SkarmoutsosV joined18:15
← jwest left18:15
→ wu7c7i joined18:17
→ Cabanoss- joined18:19
→ durham joined18:19
→ m0viefreak joined18:20
← Karyon left18:21
← Cabanoss- left18:23
← T_UNIX left18:23
→ Karyon joined18:26
→ czart joined18:28
→ Cabanossi joined18:28
→ unlimiter joined18:28
← sauvin left18:28
← armyriad left18:30
← unlimiter left18:31
← theoceaniscool left18:40
→ czart_ joined18:41
← xcm left18:43
→ xcm joined18:44
← czart left18:44
← SwiftMatt left18:53
→ acercle joined18:54
osse Dirak: git worktree is nice for this18:58
→ finn0 joined19:05
← Karyon left19:06
→ bashfulshell joined19:07
→ n3wborn joined19:20
← Dirak left19:21
← revoltingPeasant left19:21
→ elsheepo joined19:24
→ mat001_ joined19:24
← mat001_ left19:25
← czart_ left19:26
← gloomy left19:26
→ mat001_ joined19:26
→ alonkli joined19:27
← mat001 left19:27
← igemnace left19:30
← giaco left19:31
→ tufteddeer joined19:34
← sunri5e left19:34
→ sunri5e joined19:35
→ mat001 joined19:37
← emsjessec left19:40
← mat001_ left19:40
→ fphilipe joined19:42
← mingdao left19:43
← fission6 left19:43
→ mingdao joined19:44
→ fission6 joined19:48
→ relipse joined19:55
→ lf94 joined19:57
lf94 How do paths work in a git hook?19:57
Apparently one of the programs specified is running...but it really isnt or something.19:58
cdunklau lf94: you probably did something wrong. maybe!20:00
Ori_B lf94: not sure what your question is, specifically -- but you should be able to debug with a bunch of 'echo' and 'set -x'20:00
cdunklau (tell us more, then we can help)20:00
lf94 http://leefallat.ca/wiki/20:00
That script in the bottom :)20:00
If I invoke it manually, within the repository...it works20:01
When I git push, the echos I put into a custom version print....20:01
(post-receive hook)20:01
I have also replaced ~ with absolute paths20:02
Same story20:02
→ dino82 joined20:02
osse You're being too vague. What precisely is happening when you push?20:03
← oatmealraisin left20:03
→ Dirak joined20:03
relipse Hi, I'm getting a message about my repository: This repository is currently 2.0 GB in size. If it exceeds 0 bytes it will be put into read-only mode.What can I do to free up space?20:04
← rsrx left20:05
lf94 osse: https://gist.github.com/lf94/56ac4596fea74f3e3b177579d1f8531c20:05
I'm not sure how I can be any more specific20:05
→ Sasazuka joined20:05
lf94 It clearly echos "ran asciidoctor"...20:06
→ giaco joined20:06
cdunklau lf94: server-side repos are almost always "bare", they don't have a working tree20:07
osse lf94: yes, and ?20:07
lf94 Ok. I think the problem is my git commands are not running as I expect them to...20:07
cdunklau: I know20:07
osse the scrupt you linked to doesn't have "echo ran asciidoctor" in it20:07
cdunklau so if the script expects a working tree, it probably won't work on the server without extra hacking20:07
lf94 cdunklau: that's why post-receive comes into play...20:08
osse: I know - I just added it after the asciidoctor command in post-receive hook20:08
relipse is there a way to delete all my git history except the last commit?20:08
that will free up space of my repository somehow20:09
cdunklau lf94: "On the remote server you need this setting so git will accept pushes on a non-bare repository "20:09
lf94 I think it _is_ working, and git checkout HEAD ; git reset --hard HEAD -- ; are not working as I expect...20:09
cdunklau that sounds not-usual :)20:09
→ gloomy joined20:09
lf94 It is unusual - I don't say it's normal20:09
I guess I'll be explicit :p20:09
How can I checkout a completely fresh version of head within the post-receive hook, then?20:10
(if what I wrote earlier doesnt work...)20:10
cdunklau relipse: back up the repo, delete it from the server, clone from the backup locally, remove the .git dir from the clone (not the backup!), git init, add all the files, commit, add the remote, force push20:11
osse lf94: !deploy20:11
cdunklau err, not force push.20:11
gitinfo lf94: Git is not a deployment tool, but you can build one around it (in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/deploy.html20:11
cdunklau relipse: recreate the repo on the server, then push your adjusted repo20:12
lf94 osse: yeah I know. Alternatively I could just use rsync20:12
cdunklau relipse: and stop adding big files to your repo20:12
:)20:12
lf94 osse: but damn, I was not far off :D20:13
Thanks for the link20:13
← ghoti left20:14
← cfoch left20:15
← wu7c7i left20:18
→ def_jam joined20:18
← eb0t left20:19
← eblip left20:19
→ eb0t_ joined20:20
← im0nde| left20:20
→ im0nde_ joined20:20
→ im0nde- joined20:20
← im0nde left20:21
← eigenvalue left20:23
lf94 ah ha20:24
All working!20:24
→ rsrx joined20:24
lf94 Thanks again osse20:24
← gloomy left20:24
← thefatma left20:25
← achen_ left20:25
← subopt left20:28
← lf94 left20:28
deltab relipse: using bitbucket?20:30
→ lacrymology joined20:31
→ vdv joined20:31
← vdv left20:31
← elsheepo left20:32
→ Lucas_Gray joined20:32
→ leeN joined20:33
← gracecr left20:34
← rfold left20:36
→ Jackneill joined20:44
← fphilipe left20:45
← mat001 left20:46
→ mat001 joined20:49
← Dirak left20:51
← Sasazuka left20:51
→ Dirak joined20:56
→ martylake joined20:56
← netj left20:56
← mat001 left20:57
→ netj joined20:57
→ mat001 joined20:58
← mat001 left20:58
← mort left20:59
→ mat001 joined20:59
← martylake_ left20:59
← budonyc left20:59
← finn0 left21:00
← mat001 left21:01
→ mat001 joined21:02
← martylake left21:04
→ nerosnm joined21:07
← Inline left21:09
← nerosnm left21:12
→ fphilipe joined21:12
← salad left21:14
→ paws_ joined21:14
→ nerosnm joined21:19
← wagle left21:23
← queip left21:24
→ wagle joined21:24
← Jackneill left21:24
← TomyLobo left21:26
← yo-wan left21:27
→ beastDiv_ joined21:28
→ queip joined21:29
← xcm left21:31
→ xcm joined21:31
← AbleBacon left21:31
← relipse left21:33
← rsrx left21:33
→ relipse joined21:34
← Phylock left21:38
elichai2 Hi, how do I apply a commit on a subtree?21:38
I want to apply this commit: https://github.com/sipa/bitcoin/commit/3fe71bbd95f68ac03de9757397706570d3ee6a6b but it applies in on `/` instead of on `/src/secp256k1`21:39
(either with cherry-pick or rebasing)21:39
← valuemachine left21:43
← xcm left21:43
→ valuemachine joined21:43
→ xcm joined21:43
→ afowler joined21:45
← fphilipe left21:46
→ fphilipe joined21:47
← afowler left21:51
→ jstein_ joined21:56
jstein_ → jstein21:56
← fphilipe left21:57
← mohsen_1 left21:58
→ seds joined21:58
← relipse left21:59
← fission6 left22:00
→ theoceaniscool joined22:00
deltab elichai2: I don't know, but maybe by editing the patch22:02
← dartmouthed left22:03
← Es0teric left22:08
← Shentino left22:09
→ mat001_ joined22:14
← seds left22:16
← mat001 left22:17
→ seds joined22:18
← dege left22:18
← seds left22:18
← nerosnm left22:20
← bluesmonk left22:21
→ jubal joined22:29
→ fphilipe joined22:30
← jstein left22:31
← theoceaniscool left22:35
← tango_ left22:35
→ oatmealraisin joined22:35
← Lucas_Gray left22:37
→ ghoti joined22:37
← lacrymology left22:37
← kjartan left22:37
→ Lucas_Gray joined22:37
→ lacrymology joined22:38
→ kjartan joined22:38
← hussam left22:39
→ Shentino joined22:44
→ SwiftMatt joined22:45
← tufteddeer left22:45
→ tufteddeer joined22:46
← tang^ left22:48
→ jubal_ joined22:49
def_jam → eblip22:49
← tufteddeer left22:51
← jubal left22:51
→ tango_ joined22:52
→ jubal joined22:55
→ mat001 joined22:56
← Shentino left22:57
← mat001_ left22:57
← jubal_ left22:58
→ Shentino joined22:58
→ cd joined22:58
→ mat001_ joined22:58
← mat001_ left22:59
→ mat001_ joined22:59
→ yyy joined22:59
→ RoriconKnight joined22:59
← mat001 left23:00
→ tufteddeer joined23:02
← Shentino left23:02
← leeN left23:03
causaui1 in gitignore "**/thing" is the same as just "thing" right? also "thing/*" == "thing/**" or no? I'm trying to figure out what use "**" has except sandwiched somewhere inside a path23:03
← fphilipe left23:04
→ ferdroid_ joined23:05
→ eigenvalue joined23:05
← SkarmoutsosV left23:06
← Ferdroid left23:07
ferdroid_ → Ferdroid23:07
← Ferdroid left23:11
← m0viefreak left23:13
deltab causaui1: note that "It is not possible to re-include a file if a parent directory of that file is excluded.", so I guess you might exclude with foo/** so that you can include foo/bar23:15
← karoshii left23:15
← canton7 left23:15
← demize left23:15
→ karoshii joined23:15
→ demize joined23:15
→ seds joined23:16
deltab and as I understand it thing/* will match thing/foo but not thing/foo/bar23:17
→ nic-hartley joined23:17
← thiago left23:18
→ canton7 joined23:19
causaui1 hm I'm testing this, I made `foo/bar/baz` and it's ignored by `foo/` and `foo/*` and `foo/**` all the same23:19
← seds left23:19
causaui1 okay prepending "**bar/" ignores both "bar/" and "foo/bar/" etc23:22
→ elsheepo joined23:23
→ AugustusCaesar24 joined23:24
causaui1 er, I feel dumb23:26
← mowcat left23:27
← misuto left23:28
← elsheepo left23:30
→ misuto joined23:32
← perrier-jouet left23:32
→ orbyt_ joined23:32
← queip left23:33
← acercle left23:33
← n3wborn left23:33
← kjartan left23:34
→ kjartan joined23:35
→ oatmealr1 joined23:37
← oatmealraisin left23:37
→ oatmealraisin joined23:37
→ queip joined23:40
← magic_ninja_work left23:41
← CyberX left23:42
← oatmealr1 left23:42
← nic-hartley left23:45
→ subopt joined23:45
← oatmealraisin left23:48
← mat001_ left23:53
← SwiftMatt left23:53

Logs Search ←Prev date Next date→ Channels Documentation