IRCloggy #git 2021-10-22

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.

2021-10-22

→ The_Jag joined00:04
perrierjouet thanks for the helpo00:05
→ Xenguy joined00:06
→ juliopcrj joined00:09
→ rgrinberg joined00:10
← jazzy left00:12
→ jetchisel joined00:13
← juliopcrj left00:15
→ Lord_of_Life_ joined00:15
← Lord_of_Life left00:15
Lord_of_Life_ → Lord_of_Life00:16
← Sven_vB left00:18
→ gordonfish- joined00:19
← gordonfish left00:22
← Guest4 left00:27
← Ilyu_ left00:33
→ Ilyu joined00:33
→ jazzy joined00:35
← ArtVandelayer left00:47
→ ttree joined00:50
quazimodo elibrokeit: just saw what you said re. git-flow00:54
agree00:54
← Gustavo6046 left00:58
→ Gustavo6046_ joined00:58
quazimodo the ceremony is a bit nonsensical too00:59
→ pyevery__ joined00:59
→ pyeve____ joined01:00
← pyevery__ left01:00
Gustavo6046_ → Gustavo604601:01
← pyeve____ left01:05
→ pyevery__ joined01:07
← Gustavo6046 left01:11
← pyevery__ left01:12
← jetchisel left01:12
→ Gustavo6046 joined01:13
← ZacSharp left01:13
← thebombzen left01:16
→ ferdna joined01:25
→ mud joined01:28
→ The_Jag_ joined01:33
→ mgedmin joined01:33
← The_Jag left01:36
← bambanxx left01:39
→ DoofusCanadensis joined01:44
← awmv left01:47
← snedd left01:47
← durham left01:50
→ stats4647 joined01:53
← stats4647 left01:58
nedbat quazimodo: what ceremony do you mean?02:04
→ jetchisel joined02:04
→ mp3 joined02:05
elibrokeit nedbat: it's a reply to this comment of mine from quite a while back:02:07
[10:45:34 PM] <elibrokeit> quazimodo: the other problem with git-flow is that it assumes "sprinkle more branches on top" is the solution to all problems. It adds a lot of ritual ceremony that you generally don't need, and when you do need it, you don't need git-flow as a prerequisite (e.g. "branch a maintenance lineage off of the latest major release" is trivially done after the fact)02:07
git-flow has very particular branching "recommendations", when all you probably need is a) master, b) maint-$MAJOR.$MINOR02:08
nedbat elibrokeit: i was wondering in particular what steps were branded as ritual ceremony?02:09
→ thebombzen joined02:13
→ durham joined02:17
elibrokeit well, you have a develop branch which is the git-flow version of traditional master branches. This is then branched into "release-MAJOR.MINOR" and then carefully merged --no-ff into "master", because merge commits are apparently next to godliness... that branch is atrociously named, it should be named "release". Then to seal the deal, you merge --no-ff back into develop too! Then delete release-MAJOR.MINOR02:20
Want to make a maintenance fix? Branch master this time, to hotfix-MAJOR.MINOR.PATCH, then merge that back into master and then back into develop again. You can't branch off of release-MAJOR.MINOR because you deleted it lol02:20
← fling left02:24
← bn_work left02:26
← jcbjoe2021 left02:27
← Darkfoe_ left02:27
← merpnderp left02:28
← sa left02:28
← ec left02:28
← ^3^ left02:28
← LordLion left02:28
→ fling joined02:28
→ LordLion joined02:29
→ jcbjoe2021 joined02:29
← durham left02:29
→ Darkfoe_ joined02:29
→ merpnderp joined02:29
→ sa joined02:29
→ ec joined02:29
→ ^3^ joined02:29
← igemnace left02:34
→ bn_work joined02:42
← FinnElija left02:43
→ FinnElija joined02:45
→ dsrt^ joined02:51
← BSaboia left02:55
← alzgh left02:57
← Gustavo6046 left02:58
→ Gustavo6046 joined02:58
← skapata left03:10
quazimodo you make it sound worse than it is03:14
in practice the biggest problems are not the thinking, the style or naming or even the ceremony but the effect it has on development professional practice & the business as a whole03:15
← ChmEarl left03:15
quazimodo devs tend to do longer, more complex features that are harder to 'get into prod ready state'03:15
they often merge into `develop` stuff that is questionable03:15
and frequently it gets hard to cut a release because `develop` is blocked by a feature that's not quite ready, or even more insidiously, is ready but marketing says don't release it yet03:16
sensible cicd + excellent automated testing + trunk based dev + feature flags + elephant carpaccio is the way to go, i think03:17
→ Gustavo6046_ joined03:41
← Gustavo6046 left03:42
Gustavo6046_ → Gustavo604603:44
→ yeirr joined03:52
→ coot joined03:56
← zebrag left04:06
← reset left04:11
← pulse left04:17
→ saroy joined04:19
→ forgotmynick joined04:27
← jazzy left04:35
← bloody left04:37
← tirnanog left04:37
← Thanatermesis left04:43
← mp4 left04:43
→ ArtVandelayer joined04:46
← ArtVandelayer left04:48
→ ArtVandelayer joined04:48
← FH_thecat left04:49
→ BSaboia joined04:50
→ madewokherd` joined04:52
← BSaboia left04:52
→ mp4 joined04:53
→ BSaboia joined04:54
← madewokherd left04:55
→ betelgeuse joined04:56
← vishal left04:57
← Murr left04:59
→ Murr joined04:59
← yeirr left04:59
→ vishal joined05:00
→ daoudr joined05:03
← crabbedhaloablut left05:06
→ crabbedhaloablut joined05:06
← DoofusCanadensis left05:10
← Gustavo6046 left05:13
← junktext_ left05:13
→ Null_A joined05:13
→ Gustavo6046 joined05:14
← emf_ left05:14
→ jazzy joined05:15
← arcatech left05:15
← ferdna left05:18
← BSaboia left05:25
← sudoforge left05:26
→ d1b_ joined05:30
d1b_ → d1b05:30
← hbautista left05:31
← saroy left05:31
→ A_Dragon joined05:31
→ Gustavo6046_ joined05:33
← Gustavo6046 left05:33
Gustavo6046_ → Gustavo604605:36
→ mei joined05:37
← zmt00 left05:37
Jong Question, if I see a bug that I can fix with one line, but it's in someone else's code, is it a problem if I commit the one line change to master? I ask because if everyone made a commit per one-line change, there'd be far too many commits05:41
← gxt_ left05:41
→ gxt_ joined05:42
← thiago left05:43
→ gatty joined05:46
← rgrinberg left05:51
← gxt_ left05:52
→ gxt_ joined05:53
→ yeirr joined05:57
→ igemnace joined06:03
← gxt_ left06:06
→ rgrinberg joined06:07
→ gxt_ joined06:09
→ steam joined06:10
→ bkircher joined06:16
← ttree left06:17
← Xenguy left06:26
→ aniruddha joined06:27
→ ckeshav joined06:30
j416 osse: nice! ty @ patch06:30
aniruddha hello everyone I have a question that I want to get last 3 commits https://github.com/flatcar-linux/locksmith/pull/14/commits to another branch. I am not able to figure out that how can i get the last 3 commits if i merge it's also getting the 4th one from last06:35
can anyone help me :)06:35
→ palasso joined06:37
ikke aniruddha: man git cherry-pick06:37
gitinfo aniruddha: the git-cherry-pick manpage is available at https://gitirc.eu/git-cherry-pick.html06:37
aniruddha thanks :D06:38
ikke Should those commits still be in the original branch?06:40
ckeshav → ckeshav_06:41
ckeshav_ → ckeshav__06:42
ckeshav__ → ckeshav06:43
aniruddha yes they are in the integrate-fleetlock branch and i want those in the tmp-integrate-fleetlock branch to play with some of the logic06:44
← ckeshav left06:46
→ ckeshav joined06:46
← Null_A left06:49
ikke you can first cherry-pick them on the correct branch, then reset the first branch to the 4th commit06:49
→ theoceaniscool joined06:51
→ The_Blode_ joined06:52
→ rfuentess joined06:53
→ elkalamar_ joined06:54
→ otisolsen70 joined06:55
aniruddha yes its done thanks alot :D06:56
← elkalamar__ left06:57
ikke n;p06:57
np*06:57
← gast0n left06:58
→ kenanmarasli joined06:59
← ckeshav left06:59
← elkalamar_ left06:59
→ vlado_ joined07:00
← rgrinberg left07:03
← unluckyshrubbery left07:04
→ mallkrampus joined07:09
→ mat001 joined07:16
→ mannequin joined07:16
→ arcatech joined07:18
→ alkino joined07:19
← arcatech left07:23
← jazzy left07:27
→ shored joined07:27
← vlado_ left07:28
→ vladoski joined07:31
← dsrt^ left07:31
← bkircher left07:33
osse j416: my motives were questioned :(07:34
oh I misread a bit.07:38
← forgotmynick left07:47
← mat001 left07:52
← alkino left07:55
→ alkino joined07:55
→ RiFo joined08:11
→ bkircher joined08:13
← gatty left08:14
osse Jong: no. Sometimes bugfixes are small, but they certainly deserve to stand out as separate commits08:19
→ pcarphin joined08:28
→ danbak joined08:31
→ jimklimov joined08:31
→ lgc joined08:33
← jimklimov left08:36
← coot left08:42
→ BenjiProd joined08:44
← Ilyu left08:48
→ Ilyu joined08:48
→ spare joined08:58
← spare left09:02
→ spare joined09:03
← Swahili left09:04
→ Swahili joined09:04
→ coot joined09:04
→ alzgh joined09:05
← alkino left09:05
→ alkino joined09:06
→ unluckyshrubbery joined09:08
← daoudr left09:09
→ elkalamar joined09:10
A_Dragon → Awoobis09:12
← yeirr left09:13
Timvde Jong: The idea is that a commit should be one semantic "change". How you define how much one "change" can entail is up to the users, but I prefer a more granular approach. I try to make sure that every commit passes the build and could be deployed on its own.09:14
→ daoudr joined09:15
← spare left09:19
→ spare joined09:19
dviola I'm trying to rebase two commits, when doing `git rebase -i HEAD~2' I get: fatal: invalid upstream 'HEAD~2'09:19
any ideas what I'm doing wrong?09:19
this is a new repo I just created09:19
osse dviola: use --root09:24
if you only have two commits then only HEAD and HEAD~1 exist09:24
→ txtsd joined09:24
dviola osse: oh, that explains it, thanks :)09:25
← tejr left09:28
← ghost43 left09:28
→ tejr joined09:29
→ ghost43 joined09:29
→ rsx joined09:42
→ wender joined09:43
→ vlado_ joined09:45
← computeiro left09:47
← vladoski left09:47
← af left09:50
→ Ryan__ joined10:02
← Ryan_ left10:05
← AbleBacon left10:09
→ af joined10:23
→ caveman joined10:23
← igemnace left10:33
→ ckeshav joined10:38
→ stats4647 joined10:52
← stats4647 left10:56
← Gustavo6046 left11:00
→ velix joined11:01
velix Man... I hate this: git for Windows randomly connects via curl.exe to IP 104.21.24.9 ... Any way to turn this off? I think that's against law in many countries.11:02
Is it sending out any data?11:02
I bet it's telemetry...11:02
I my country, you have to CONSENT to do so!11:02
bremner I guess there's a git-for-windows github repo where you can complain?11:03
→ Gustavo6046 joined11:03
→ Rhvs joined11:10
velix bremner: Okay, thanks.11:11
I should start a firewall on linux, too. Maybe it does the same here.11:12
I bet this was Micro$oft's idea.11:12
They like sending telemetry.11:12
← spare left11:14
velix I think the problem is bigger than expected. Seems like it affects binary builds only! It's not in the source code: https://github.com/git-for-windows/git/search?q=104.21.24.911:15
← velix left11:19
→ pulse joined11:20
→ cdown_ joined11:21
cdown_ → cdown11:24
→ BSaboia joined11:26
← pah left11:28
→ pah_ joined11:28
← caveman left11:32
→ frobnic joined11:35
← quazimodo left11:37
→ cdown_ joined11:38
← cdown left11:41
cdown_ → cdown11:41
← ckeshav left11:49
→ natrys joined11:52
→ skapata joined11:58
← skapata left11:58
→ skapata joined11:58
← wender left11:58
→ caveman joined11:59
→ Sazhen86 joined12:00
← BSaboia left12:00
→ BSaboia joined12:02
→ computeiro joined12:02
→ nyah joined12:03
→ rewrit3 joined12:04
← vlado_ left12:15
Seveas "telemetry" yes. Probably just checking for updates...12:15
→ vladoski joined12:15
← coot left12:16
bookworm (why would you put an IP in the source... chances are it would be there as a hostname)12:18
Seveas yeah12:18
→ Betal joined12:19
→ jiehuan joined12:19
→ speckz joined12:23
→ jwillikers joined12:24
→ Lunatrius joined12:25
bremner fwiw, they left12:28
← alzgh left12:29
→ mei6 joined12:30
← mei left12:31
mei6 → mei12:31
← jwillikers left12:33
Dragnslcr In case they come back, that's the IP address for gitforwindows.org12:34
→ burakcank joined12:34
mgedmin checking for updates maybe?12:35
← rahl left12:37
→ rahl joined12:39
→ alzgh joined12:43
← caveman left12:44
→ causa_sui joined12:51
← ano left12:52
→ forgotmynick joined12:53
→ ano joined12:53
→ Rashad joined12:57
← causa_sui left12:57
→ coot joined13:01
← ano left13:01
→ ano joined13:02
→ bloody joined13:04
← speckz left13:13
→ meator joined13:16
→ sudoforge joined13:17
→ spare joined13:17
← spare left13:25
← ano left13:27
→ spare joined13:27
→ ano joined13:28
→ Thanatermesis joined13:29
→ causa_sui joined13:30
← spare left13:30
→ spare joined13:31
→ juliopcrj joined13:32
→ gh34 joined13:38
← jiehuan left13:38
→ harpia joined13:40
← RiFo left13:41
→ utis joined13:43
→ RiFo joined13:44
← otisolsen70 left13:46
← pulse left13:48
→ pyevery__ joined13:55
→ BUSY joined13:55
→ pulse joined13:56
→ pepperoni joined14:01
← pyevery__ left14:03
→ pyevery__ joined14:04
← ArtVandelayer left14:06
→ cyberpear joined14:06
Portugol9_ → Portugol914:08
→ yeirr joined14:08
← pyevery__ left14:09
→ vlado_ joined14:13
← vladoski left14:15
← spare left14:16
← meator left14:16
→ spare joined14:18
← Ilyu left14:21
→ Ilyu joined14:22
← spare left14:27
→ spare joined14:27
→ zebrag joined14:29
← causa_sui left14:34
→ pflanze joined14:36
← Sazhen86 left14:36
→ pyevery__ joined14:37
→ Xaldafax joined14:44
→ pretty_dumm_guy joined14:53
→ Argorok joined14:54
→ thiago joined14:55
Argorok Is it possible to create a branch from master (let's call it featureX), then compare a featureXY branch against master and cherry pick code easily? The featureXY branch include featureX and featureY, I want to separate them accordingly but there are too many changes to do it manually14:57
← shokohsc8 left14:57
← Murr left14:58
→ Murr joined14:58
← pyevery__ left15:00
→ arcatech joined15:00
→ tsdh joined15:06
← mud left15:08
→ shokohsc8 joined15:09
→ mud joined15:12
→ stats4647 joined15:12
bookworm Argorok: simplest in that case is to rebase twice, so that you have both X and Y separate15:13
as in, interactive rebase down to master, delete anything that isn't the thing you want (either X or Y parts) then merge the thing as needed back to master15:14
Argorok Never used/did a rebase. Lemme read about it15:14
bookworm as for "comparing", you mean diff? git diff / git log can both take a range of commits. start..stop15:14
Argorok Only delete? Because I wil also need to readd some code that was deleted15:15
Yeah, git diff between master and featureXY15:15
bookworm >I wil also need to readd some code that was deleted , what do you mean?15:17
→ marw joined15:18
← stats4647 left15:19
Argorok FeatureXY started with featureX and then featureY started after featureX finished. But, for featureY, some code of featureX was deleted. So, when separating them into two distinct branches, I need to get all the featureX code back15:19
← arcatech left15:20
bookworm ah, but then you already have a clean separation yes? X1->X2->X3->Y1->Y2->Y3 that's how your commit graph looks then?15:20
if so, simply checkout a new branch with X3, that's X.15:21
if Y messes with X / needs stuff from it, you either need to keep it on top of it or remove the stuff that depends on X, your choice15:21
Argorok Mooooore or less... There are some X4 and X5 among the Ys15:23
bookworm then you need to rebase15:23
so that you can sort them to the rest of the x changes15:23
← hackinghorn left15:24
→ ChmEarl joined15:24
Argorok Hmmmmm... Thanks. I will read about and watch some videos to not fuck up everything. More than 115 files, don't want to do the separation manually x.x15:25
→ zmt00 joined15:29
→ causa_sui joined15:30
→ alfredb joined15:32
alfredb → abu15:33
→ pyeveryt_ joined15:33
→ rgrinberg joined15:33
→ Gustavo6046_ joined15:34
← Gustavo6046 left15:34
← natrys left15:34
← harpia left15:35
← causa_sui left15:35
← causasui left15:36
Gustavo6046_ → Gustavo604615:37
← Ryan__ left15:37
→ Ryan_ joined15:38
→ Null_A joined15:38
← pyeveryt_ left15:38
← vdamewood left15:38
← rfuentess left15:40
→ vdamewood joined15:41
→ m0viefreak joined15:42
← RiFo left15:44
→ durham joined15:49
← madewokherd` left15:50
→ rafspiny joined15:53
→ arcatech joined15:54
← Gustavo6046 left15:54
→ Gustavo6046 joined15:55
← arcatech left15:57
← rafspiny left16:00
→ arcatech joined16:00
← mp4 left16:05
← kenanmarasli left16:06
← arcatech left16:06
→ arcatech joined16:07
← arcatech left16:08
← vlado_ left16:11
→ odoood joined16:14
← odoood left16:14
→ Krematorium joined16:16
Krematorium hello guys, question: if i rename my local and origin branch but i already have an PR (Gitlab, Github, Bitbucket) opened with, the existing PR would recognize the branch name just fine or i would have to pen a new PR or so?16:17
ikke Krematorium: No, you would still need to push to the old branch name, or open a new MR/PR16:18
You cannot push a branch rename16:18
Krematorium oh16:18
ok, thanks16:18
→ hackinghorn joined16:20
Dragnslcr I've used git-svn to migrate a Subversion repository to git, and everything seems to have worked fine. I'm now trying to setup mirroring commits to the git repository back to Subversion, but I don't have the original directory that I used for the migration, and I haven't been able to get it to work.16:21
← mei left16:21
← hackinghorn left16:21
Dragnslcr I've tried doing git svn clone, git remote add, and git pull, but I end up with two copies of every branch. Trying to do git svn rebase gives me "Unable to determine upstream SVN information".16:22
Does anyone have any pointers to get this to work?16:22
← Gustavo6046 left16:25
→ Gustavo6046_ joined16:25
→ madewokherd joined16:27
Gustavo6046_ → Gustavo604616:28
→ brianward joined16:28
← rsx left16:28
→ causa_sui joined16:29
→ harpia joined16:32
← alzgh left16:35
→ alzgh joined16:36
← m0viefreak left16:39
→ zer0bitz joined16:41
woss[m] Hi guys, I'd appreciate any constructive comments on the article i'm writing. it's a first part of the series `Rehosting on IPFS`.16:44
I mainly seek the comments on the stuff i say about git, is it in the correct form, and is it correct in terms `i don't want to confuse people and tell lies`. Thank you in advance16:44
https://kelp.notion.site/WIP-Part-1-Rehosting-git-repositories-on-IPFS-db5ab8e567064eba8fd10237e6b1fecc16:44
→ orbyt joined16:44
→ emf joined16:49
→ ArtVandelayer joined16:53
→ ___nick___ joined16:57
← ___nick___ left16:57
→ ___nick___ joined16:58
← Murr left16:58
← The_Blode_ left16:59
→ Murr joined16:59
→ junktext_ joined17:00
← BenjiProd left17:02
→ natrys joined17:02
→ darkdrgn2k joined17:03
darkdrgn2k is it possible to submodle into a folder of a repo, or does it always have to be root17:04
ikke no, it can be in any subdirectory17:04
darkdrgn2k how would you define that?17:04
thiago it's always the root of the submodule's repo17:05
darkdrgn2k yeh thats what i was asking :/17:05
you cant go deper into a folder in the repo17:06
i guess that makes sense tough since git refrence everythig from that :17:06
← natrys left17:07
← forgotmynick left17:09
thiago note that a repository does not have a root17:09
the submodule points to a commit and the commit has a root tree17:09
so you can just make a commit that has something as as the root17:10
→ kenanmarasli joined17:12
→ ash_worksi joined17:13
← causa_sui left17:15
→ causa_sui joined17:16
← theoceaniscool left17:21
← durham left17:21
← causa_sui left17:21
→ durham joined17:22
→ causa_sui joined17:22
← vdamewood left17:25
← causa_sui left17:27
← SuperLag left17:28
→ pyeveryt_ joined17:34
← pyeveryt_ left17:39
← coot left17:40
relipse I configured my .gitignore file, then I did git add public_html only to realize it took a good 15 minutes because of all the image files, how can I add uploads/ to my .gitignore and unstage all the matching uploads/* files17:42
ikke add uploads/ to your .gitignore file using your favorite editor17:44
then git rm --cached uploads17:44
relipse so just to be clear. That will not delete anything.17:46
→ masber joined17:46
→ SuperLag joined17:46
→ wojciech joined17:47
ikke correct17:47
make sure you add --cached17:47
← wojciech left17:48
→ wojciech joined17:48
← wojciech left17:48
→ wojciech joined17:49
← wojciech left17:49
→ dolor_avis joined17:49
→ kaartic joined17:50
← dolor_avis left17:53
→ dolor_avis joined17:53
← dolor_avis left17:54
→ dolor_avis joined17:54
← dolor_avis left17:55
→ causa_sui joined17:58
← causa_sui left18:02
← kaartic left18:04
← ano left18:06
← Null_A left18:07
→ rsrx joined18:07
← rsrx left18:07
→ rsrx joined18:07
→ pyeveryt_ joined18:09
→ ano joined18:09
→ Null_A joined18:10
→ gast0n joined18:17
→ vdamewood joined18:17
→ meator joined18:18
→ vinleod joined18:19
← Null_A left18:20
← rgrinberg left18:20
← spare left18:21
→ spare joined18:21
← vdamewood left18:22
vinleod → vdamewood18:22
→ Null_A joined18:22
→ reset joined18:23
→ The_Blode joined18:24
→ humky joined18:30
→ john_johnk joined18:31
← meator left18:33
← rsrx left18:35
→ bloodElk joined18:40
→ stats4647 joined18:45
→ kd5678__ joined18:45
→ goldfish joined18:48
← spare left18:48
→ spare joined18:48
← spare left18:48
→ spare joined18:49
→ theoceaniscool joined18:49
← spare left18:50
→ spare joined18:50
← kd5678__ left18:51
→ kd5678__ joined18:51
→ jonosterman joined19:03
relipse is it possible to see the sum total of all the sizes of the files that are staged (added with git add)19:04
about to be committed19:04
so i can see if my repository will get too big from this commit19:04
also I tried doing git rm --cached public_html/_uploads/* and i'm getting fatal: pathspec 'public_html/_uploads/1032568974.png' did not match any files so it is failing?19:12
and there are still hundreds of them in the staging area19:13
← Null_A left19:15
→ Null_A joined19:15
← theoceaniscool left19:19
← yeirr left19:22
← john_johnk left19:23
← BSaboia left19:25
← af left19:29
→ af joined19:30
← vdamewood left19:31
← Krematorium left19:31
→ ZacSharp joined19:34
→ rond_ joined19:34
relipse how do I clone a responsitory that is stored over ssh19:37
is the idea of fork a git thing or is that only github and bitbucket?19:37
← masber left19:38
j416 not a git thing19:38
mackerman fork workflow was invented by git hosting. There is no fork operation native to git.19:38
j416 a fork is just another repo as far as git is concerned19:38
mackerman Regarding rm --cached, you asked to "unstage" files. This will leave them in the working directory.19:38
j416 relipse: the * in public_html/_uploads/* is expanded by your shell, not git. If you want to remove that entire directory with contents, you can git rm -r --cached public_html/_uploads19:40
relipse *update* yep -r worked fine thanks19:40
j416 might be able to quote it like 'public_html/_uploads/*' too.19:40
← Rashad left19:41
relipse I have a server that I log in as relipse and then do sudo su to get to /home/foobar/myrepo/ , how do I clone that from my home computer19:42
git clone [email@hidden.address] ??19:43
mackerman A user that can log in with ssh, and has file permissions there.19:43
relipse so i'll need root19:44
mackerman No, not root19:44
You can grant user relipse read and write to ~foobar/myrepo19:45
relipse how do I do that mackerman19:45
mackerman Try asking a support channel of your operating system.19:45
File permissions aren't really a git thing.19:46
Would have the same problem if you wanted sftp access to upload and download files.19:47
→ Krematorium joined19:48
→ jimklimov joined19:48
← stats4647 left19:49
→ euouae joined19:52
euouae Hello19:52
I'm confused about git workflows. Here's an example:19:52
Suppose there's a function A() I want to modify. I make some changes, but it turns out that A() uses B() and B() is broken too. However, I don't want to make a gigantic PR with /too many/ changes, because it's hard to review.19:53
so I make another branch, I fix B(), then I wish to continue working on A(), and I'd like to push two PRs, fix-B and fix-A. In github, this means I need the fix-A commented with "depends on #fix-B", for the sake of having some continuity -- helps reviewers.19:54
My question is, after I have fixed B, do I merge A with it to continue working? Is that what I'm supposed to do?19:54
mackerman euouae: There are multple ways to do anything19:55
A rebase of fix-A to the main branch might make sense after fix-B is merged.19:55
euouae That's true mackerman, I guess, I'm struggling because the project has a "clean commits" soft-policy which I guess means the merges are ugly. But rebases are good?19:56
Yeah unfortunately that means I have to wait for fix-B to be merged; I want to continue working on fix-A19:56
← Krematorium left19:58
mackerman If you can prove fix-A fixes the problem at hand, you may not have to wait on fix-B. Depends on your workflow, reviewers, and tests.19:58
→ rgrinberg joined19:58
ZacSharp If you are sure fix-B will be merged you can merge it for yourself locally and then rebase fix-A onto that19:58
← furrymcgee left19:59
ZacSharp as long as you and the upstream maintainer do the same thing the commits will end up being the same19:59
euouae ZacSharp ah! nice. Why should I merge it locally?19:59
Hm... ok. to repeat the actions. Got it19:59
ZacSharp only problem might be authorship19:59
if the merge requires a signed commit you have no chance to do it or you discovered a severe security problem20:00
relipse do you need write permissions to clone a repo?20:00
you'd think you only need read20:00
euouae ZacSharp, mackerman thanks for your help20:01
← ___nick___ left20:04
→ theoceaniscool joined20:05
→ evocatus joined20:05
euouae although I sign my commits, this is an open-source project with no security requirements20:06
ZacSharp the problem is if the merge is a commit signed by someone else20:06
lantech19446 / autojoinem.py add\n20:06
euouae I think the merges are done automatically by github that signs them right?20:07
ZacSharp but I just noticed you can just cherry-pick your commits onto theirs and are fine20:07
probably20:07
I think github singns its commits20:07
euouae github kind of gets in the way sometimes :( it's annoying20:07
ZacSharp but that just prevents creating the same commit20:08
euouae Another question I have is, how to maintain the order of branches?20:08
ZacSharp you can still merge locally, work on that merge and the cherry-pick/repase/whatever the files onto the actual merge20:08
← kd5678__ left20:08
euouae For example if I have a main and a devel, how to ensure devel is /always/ ahead of main?20:08
ikke on github you do not have a lot of control over that20:09
→ kd5678__ joined20:09
euouae Got it. Thanks20:09
ikke Only thing you can do is prevent pushes to main and always require pull requests20:10
relipse i can't believe they are getting rid of master20:10
that's so silly20:10
ikke It's not worth the effort to argue over it20:10
← theoceaniscool left20:11
j416 do you know if git.git is also moving to make main the default name?20:11
ikke I have not seen any discussions about it20:12
→ AbleBacon joined20:14
← kd5678__ left20:16
euouae This project I work on uses master instead of main20:18
I think it's too much work for the devs to change it right now20:19
ikke That was and still is the default for git20:19
euouae I have another project that I work on myself where I made two mistakes20:20
1) I didn'g sign the commits and 2) I forgot to include a license until after20:21
how can I reapply everything, and reorder the license to be commit #1? With rebase, right?20:21
(reapply everything so that it's signed)20:21
mackerman You don't need for a license to be the first commit20:21
euouae I've been told that's the case mackerman , not legal advice but general advice on a gpl channel20:22
ikke but in general, rebase can be used for these kinds of things, yes20:23
mackerman Do you have users you are having arguments with about the license? If not it doesn't matter.20:23
euouae I agree that it doesn't matter, but I want to do it regardless :P20:24
nedbat euouae: if someone says the license has to be first, they will also say that rebasing to make it first doesn't count.20:24
euouae: the point is, if someone grabbed your code now, there's no license. You can't actually go back in time to fix that.20:25
Puyo → puyoxyz20:25
euouae This is local code only nedbat20:25
I have not yet made my move to upload it20:25
nedbat euouae: then it doesn't matter.20:25
euouae: the first moment anyone can see the code, it has a license on it.20:25
→ Crispy joined20:26
mackerman As soon as you create a thing, it may have copyright. Depending on local laws.20:26
→ Krematorium joined20:27
pah_ → pa20:27
← pa left20:27
→ pa joined20:27
puyoxyz → Puyo20:30
→ m0viefreak joined20:31
→ rafspiny joined20:31
→ causa_sui joined20:32
← ArtVandelayer left20:35
← aniruddha left20:36
← causa_sui left20:37
→ ghost43_ joined20:39
→ remyabel joined20:40
← Null_A left20:40
← ghost43 left20:41
← rgrinberg left20:41
← junktext_ left20:41
← Krematorium left20:42
→ junktext_ joined20:42
gordonfish- → gordonfish20:44
→ lantech19446 joined20:46
← madewokherd left20:47
← lantech19446 left20:48
→ lantech19446 joined20:48
→ mat001 joined20:50
→ ArtVandelayer joined20:51
← rond_ left20:55
← zer0bitz left20:58
← Murr left20:58
→ Murr joined20:58
→ vdamewood joined20:59
← euouae left20:59
← Gustavo6046 left21:04
← evocatus left21:07
→ Gustavo6046 joined21:07
← johncs left21:09
→ vinleod joined21:10
→ johncs joined21:10
← vdamewood left21:11
← Ilyu left21:12
← mat001 left21:16
→ rgrinberg joined21:22
← zebrag left21:22
→ mat001 joined21:24
→ forgotmynick joined21:24
← orbyt left21:25
← kenanmarasli left21:26
→ madewokherd joined21:27
→ ZacSharp63 joined21:30
← Sario left21:31
→ ericnoan joined21:31
→ SpacePlod joined21:31
→ avar joined21:31
→ starfarer joined21:31
→ xaxas joined21:31
→ NoImNotNineVolt joined21:31
→ mattf joined21:31
→ mfiano joined21:31
→ makara joined21:31
→ Ugrastil joined21:31
→ sahilister joined21:31
→ blmt joined21:31
→ eschwartz joined21:32
→ alissa joined21:32
→ Tabmow joined21:32
→ gjnoonan joined21:32
→ hendry joined21:32
→ Piraty joined21:32
→ wilz joined21:32
→ Sario joined21:32
→ eamanu joined21:32
→ hwrd joined21:32
→ NiKaN joined21:32
→ shush joined21:32
→ paruh joined21:32
→ Byteflux joined21:33
→ det joined21:33
→ dgw joined21:33
→ themill joined21:33
→ jrm joined21:33
→ jinsun joined21:33
→ Anarchic joined21:33
→ sunyibo joined21:33
→ jiffe joined21:33
→ mrkajetanp joined21:33
→ lemonsnicks joined21:33
→ Cyp joined21:33
→ Ram-Z joined21:33
→ Timvde joined21:33
→ amosbird joined21:33
→ webstrand joined21:33
→ dslegends joined21:33
← ZacSharp left21:33
← computeiro left21:34
ZacSharp63 → ZacSharp21:36
→ RichiH joined21:36
→ hanneso joined21:36
→ tureba joined21:36
→ cryptonector joined21:36
← jimklimov left21:36
→ arand joined21:36
← ash_worksi left21:42
→ euouae joined21:42
euouae I have21:42
I have origin and upstream and it's a bit misconfigured21:42
I originally had origin to be my gh clone, but I want to only push to my gh clone and do PRs, instead of fetching from gh clone21:42
so I'm trying to have branch foo track upstream/foo instead of origin/foo. How can I do this? I used --unset-upstream, but I don't understand how to use --set-upstream-to21:43
← mat001 left21:43
→ jimklimov joined21:45
← gh34 left21:46
→ jazzy joined21:47
← jimklimov left21:47
→ cahoots joined21:48
jazzy → jazz-o-lantern21:48
ZacSharp you just checkout the branch and run `git branch --set-upstream-to=remote/branch`21:49
do you have write permissions for the repo? Otherwise it might be better to set up a triangular setup where you fetch from upstream but push to origin21:50
→ mat001 joined21:51
← spare left21:52
→ spare joined21:52
← junktext_ left21:52
← spare left21:55
→ zebrag joined21:57
cahoots hi, i'm trying to make a tool for github CI. for various legacy reasons, it can only have the name of the base branch and sha of the topic branch that the pr is trying to merge in. given this, can i reliably determine the sha of the base branch with which topic has been merged for the CI job? i mean, github always just merges topic with some base branch, so i can just look at the head commit and use the22:00
parent sha that's not the topic sha i've been given, right?22:00
← eschwartz left22:00
→ eschwartz joined22:00
← rewrit3 left22:01
← euouae left22:01
ZacSharp In fact, if you can rely on Github always merging the pr into the base branch and never the base branch into the pr you can just take the first parent22:02
← harpia left22:04
← mat001 left22:07
→ causa_sui joined22:10
cahoots can i rely on that?22:14
← causa_sui left22:14
→ LuKaRo joined22:22
→ jazzy joined22:23
← jazz-o-lantern left22:25
← Swahili left22:29
→ ash_worksi joined22:29
→ Swahili joined22:30
← rafspiny left22:31
vinleod → vdamewood22:34
← rgrinberg left22:37
→ DoofusCanadensis joined22:37
← jazzy left22:39
← nyah left22:40
→ BisexualWolf joined22:42
→ mat001 joined22:42
BisexualWolf Hello, all. I have a personal project and I'm trying to work on documenting it. I am hosting the documentation on gh-pages and was wanting to use the jekyll theme "documentation" from tomjoht/documentation-theme-jekyll on github. My question is how best to make use of the repo.22:42
My repo has a docs directory in the master branch that will be used for the documentation. My first thought was to just clone the repo into that directory and then make the required changes and i need and then push my changes back to my repo. But I would like to maintain upstream changes/improvements to the theme in my theme, so I thought to use a git submodlule after cloning into the docs directory.22:42
Would using git submodules be the proper way to make use of theme but still be able to receive upstream changes? Is there a better way to achive what I'm after?22:43
← mat001 left22:46
← daoudr left22:46
→ jazzy joined22:48
← palasso left22:51
BisexualWolf Or having just finished reading about git subtree, which one sounds more true to what I'm trying to accomplish?22:53
→ hbautista joined22:57
→ ChemicalRascal joined22:58
→ magic_ninja_work joined23:00
← pizdets left23:02
→ pizdets joined23:03
← det left23:07
→ Xenguy joined23:08
→ det joined23:12
← brianward left23:14
← ash_worksi left23:15
← darkdrgn2k left23:19
→ howudodat joined23:24
howudodat I have a weird problem synchronizing repos.23:24
I have a local repo on my computer, and I created a central repo on my web server. On another computer I did git clone to retrieve from the web server23:25
Computer A: added some files and made some changes, did a commit and push to local repo and push to central repo23:25
Computer B: pulled those files down, and now they show as unstaged / untracked on Computer B23:26
← goldfish left23:27
← vdamewood left23:31
← Gustavo6046 left23:35
→ rafspiny joined23:36
→ Gustavo6046 joined23:36
→ ninjin joined23:39
→ rgrinberg joined23:40
← ZacSharp left23:41
← magic_ninja_work left23:41
← DoofusCanadensis left23:42
→ stats4647 joined23:46
→ ZacSharp joined23:47
← ZacSharp left23:48
← stats4647 left23:50
→ dsrt^ joined23:55

Logs Search ←Prev date Next date→ Channels Documentation