IRCloggy #git 2022-11-13

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.

2022-11-13

NorrinRadd the feature branch merges should not have been lost00:00
lmat nedbat: Yes, it's a very pared down git flow. Basically, develop branch, master branch, feature branches. That's about it :-D00:00
← thebombzen left00:00
NorrinRadd as long as you rebase on top of and do not edit any commits, those commits will still be there00:00
in a big enough team, i think git-flow can be useful. for teams that don't release often and don't have good CI and potentially still use manual testing, it can be useful00:02
lmat they most likely need hotfix branches also00:02
→ hbautista joined00:03
NorrinRadd but back to the original issue, github is saying development is not updated because somehow master updated from a source that is not development. a hotfix is a good enough reason. but either someone commited to it directly, which is most likely against the team room. So they screwed up the repo.00:04
i said rebase development but that's not necessarily good because it'll disturb the development branch that people have locally. you can also merge master into development and it should solve the problem00:05
meh just review the merge commit thoroughly to make sure it isn't reverting any development work00:05
→ thebombzen joined00:06
NorrinRadd in the future any hotfixes should merge both into master and the development branch at roughly the same time00:06
BtbN the merge commit alone is enough to make it says it's not up to date00:07
← xandris left00:10
← TonyStone31 left00:11
→ xandris joined00:12
→ hoochmonger joined00:17
→ _natrys_ joined00:18
lmat NorrinRadd: "somehow master updated from a source that is not development." I am unable to verify that. Like I said, the only commit on master that is NOT on development is the last merge from development to master.00:18
BtbN: Thank you.00:19
NorrinRadd yeah master should have been fast-forwarded instead of a merge commit00:19
lmat https://github.com/STARTedUP-Foundation/IEDC-Webserver/pull/19700:20
NorrinRadd: Oh...hmm00:20
NorrinRadd: Does Microsoft Github do FF?00:20
NorrinRadd yes00:20
← natrys_ left00:21
NorrinRadd BtbN but if its merged back to dev, it should get rid of the github complaint i think00:21
BtbN With a Spaghetti-History like that, who knows how Github interprets things00:22
I've just learned to avoid merge commits, since they always cause confusion and complications00:22
NorrinRadd ^ that part00:22
lmat NorrinRadd: I agree: merging that merge commit back to dev should get rid of it. But I don't want to do that.00:23
BtbN: No merges at all?00:23
BtbN no, they confusing00:23
lmat BtbN: Do you work on a team (of more than one ^_^)?00:23
→ TonyStone31 joined00:23
NorrinRadd it's possible to work without merges00:23
lmat Certainly,00:23
NorrinRadd on a large team even00:23
BtbN Not using merge commits does not mean not working in branches...00:24
NorrinRadd git-flow generally only moves in one directly00:24
from devs up to master00:24
not the reverse00:24
s/directly/direction00:24
lmat The way I generally work is basically the same, except that we use merge commits (that could be FFs) to help delineate feature branches: https://imgur.com/a/k3Ems9R00:25
NorrinRadd keeps your git history nice and linear00:25
BtbN doing bisects with merge commits is more painful as well00:25
lmat BtbN: certainly!00:25
BtbN I'd rather have single, isolated commits, each doing one thing, in one long line00:25
lmat With the pattern I posted above, the merge commits are helpful because they point back to the Microsoft Github pull request so that I can see the code review comments and discussion.00:26
NorrinRadd lmat all that can be part of commit messages though00:26
the link to the PRs, etc00:27
i guess what i mean to say is, you can configure github so that when a PR is "merged", it'll so a fast forward with a commit message that contains the first message of the PR.00:28
lmat NorrinRadd: Yeah...I don't think I could get my team to be that tidy xD00:28
specing remember the days when git was supposed to store project history00:28
BtbN Github is smart enough to keep a refernece to the PR if you rebase-merge via its UI anyway00:28
lmat specing: curated project history ;-)00:29
BtbN "Project history" like... "Fixes", "More Fixes", "Ooops", "seriufgbsrgf", "", ...00:29
Each commit with a random amount of changes ranging from -1+1 to -3000+500000000:30
← _natrys_ left00:30
lmat BtbN: No kidding!00:30
BtbN: "Added css property" "that didn't work, using inline styles" "forgot comma". I mean, for real. I look at coworkers' 50-commit feature branches and almost all of them are fixing typos from previous commits. That is not helpful.00:31
BtbN There is nothing wrong with making commits pretty. That's what feature staging branches are for00:31
nedbat lmat: why do you type "Microsoft GitHub" each time instead of just "github"?00:31
BtbN Each commit should be precise, concise and work on its own.00:32
lmat nedbat: Not sure...been doing that for many years. Same reason I type Microsoft Outlook and Microsoft Windows and GNU/Linux and Atlassian Jira and Chevrolet Camaro :shrug:00:35
NorrinRadd BtbN++00:36
lmat BtbN: Oh if only...I will rejoice when I see the day in one of my projects!00:36
NorrinRadd you make it a rule & work doesn't get "meged" until they fix their commit messages00:36
giving them a format to fill out helps00:37
BtbN You want me to merge your code? You at least follow basic git hygiene.00:37
nedbat lmat: you can save some typing00:38
→ brezanac joined00:45
→ mandeep joined00:47
← brezanac left00:49
→ ThorMojito joined01:09
← TonyStone31 left01:16
← forgotmynick left01:17
← thuna` left01:17
→ thuna` joined01:18
← eligos left01:25
← gvg left01:30
← rgrinberg left01:36
← wootehfoot left01:38
→ xx joined01:38
← ghost43 left01:39
← FinnElija left01:39
← Aminda left01:39
→ ghost43 joined01:39
→ Aminda joined01:40
→ thiago joined01:40
→ FinnElija joined01:43
→ lucasta joined01:43
→ rgrinberg joined01:44
→ va5c0 joined01:45
→ TonyStone31 joined01:45
← ghost43 left01:49
→ ghost43 joined01:50
← va5c0 left01:50
← crabbedhaloablut left01:51
→ crabbedhaloablut joined01:51
→ va5c0 joined01:52
← ThorMojito left02:00
← va5c0 left02:00
→ va5c0 joined02:01
← va5c0 left02:03
← DaNeenjah left02:03
→ va5c0 joined02:06
← TonyStone31 left02:06
← va5c0 left02:09
→ va5c0 joined02:09
← clime left02:12
→ YuGiOhJCJ joined02:13
← va5c0 left02:13
→ elastic_1 joined02:17
← elastic_dog left02:17
elastic_1 → elastic_dog02:17
← ___nick___ left02:19
← Aminda left02:20
← YuGiOhJCJ left02:20
→ Aminda joined02:20
→ YuGiOhJCJ joined02:20
→ TonyStone31 joined02:20
→ ___nick___ joined02:21
← ___nick___ left02:21
→ ___nick___ joined02:23
← atrigent_ left02:23
→ atrigent joined02:24
→ small joined02:26
← YuGiOhJCJ left02:30
→ YuGiOhJCJ joined02:31
← TonyStone31 left02:34
← thuna` left02:36
← loulou_ left02:37
→ hiroot joined02:38
← tonymke left02:39
← thiago left02:40
→ wallabra joined02:45
← funhouse__ left02:47
→ atrigent_ joined02:59
← atrigent left03:00
→ wallabra_ joined03:04
← wallabra left03:05
wallabra_ → wallabra03:05
← atrigent_ left03:10
← audio left03:11
← vaaleyard left03:12
→ audio joined03:12
← moldorcoder7 left03:12
→ thiago joined03:12
→ TonyStone31 joined03:13
← TonyStone31 left03:19
→ ikwyl6 joined03:21
→ gvg_ joined03:22
← gvg_ left03:22
→ vdamewood joined03:25
← ferdna left03:26
→ gvg joined03:29
← gvg left03:31
← hoochmonger left03:32
← lucasta left03:32
→ gvg joined03:37
→ SirRippovMaple5 joined03:38
→ lucasta joined03:38
→ TonyStone31 joined03:39
→ vaaleyard joined03:41
← audio left03:44
← terrorjack left03:48
→ terrorjack joined03:49
→ ThorMojito joined03:52
← ThorMojito left03:53
← hiroot left04:17
← thebombzen left04:19
smol-hors → pony04:19
→ vyryls joined04:27
→ gsi_ joined04:29
→ ThorMojito joined04:31
← ThorMojito left04:32
← gsi left04:32
← vdamewood left04:32
← gast0n left04:36
← vyryls left04:40
→ Milos joined04:41
← zebrag left04:44
→ hiroot joined04:44
← zulutango left04:51
→ vdamewood joined04:51
← lucasta left04:54
← small left04:56
→ audio joined04:59
→ small joined04:59
← hiroot left05:05
← wallabra left05:07
→ valeyard joined05:08
← vaaleyard left05:12
→ nate4 joined05:16
→ wallabra joined05:18
← nate4 left05:20
→ Lord_of_Life_ joined05:28
← Lord_of_Life left05:29
Lord_of_Life_ → Lord_of_Life05:30
← small left05:34
← rgrinberg left05:41
← YuGiOhJCJ left05:46
→ YuGiOhJCJ joined05:47
→ madewokherd` joined05:52
← chexum left05:53
→ chexum joined05:53
← madewokherd left05:54
← jundran left05:55
→ ThorMojito joined05:55
← ThorMojito left05:55
→ jmd joined05:58
← reset left06:00
← audio left06:03
← valeyard left06:03
→ xlirate__ joined06:06
← xlirate_ left06:09
→ small joined06:12
→ ThorMojito joined06:19
← ThorMojito left06:19
← Murr left06:25
→ Murr joined06:26
← hbautista left06:28
→ igemnace joined06:37
→ zenobius joined06:37
→ ThorMojito joined06:40
← YuGiOhJCJ left06:40
← ThorMojito left06:40
→ YuGiOhJCJ joined06:41
← Aminda left06:41
← YuGiOhJCJ left06:42
→ Aminda joined06:42
→ YuGiOhJCJ joined06:43
→ vyryls joined06:51
→ nate4 joined06:52
→ thebombzen joined06:53
← YuGiOhJCJ left06:54
→ YuGiOhJCJ joined06:55
← nate4 left06:56
← vyryls left06:58
→ ThorMojito joined06:59
← ThorMojito left06:59
pony → smol-hors07:00
← jmd left07:04
→ jmd joined07:05
← jmd left07:05
→ jmd joined07:05
→ IceMichael joined07:11
IceMichael hi guys07:11
cross-posting from docker, but I think, it's really just a git question: I would like to retrieve the revision from git, which I can do. But which files do I need for that? Currently I copy .git to my container but... that seems huge07:12
Celeo Revision as in current commit SHA? If so: ./.git/refs/heads/<branch_name>07:17
IceMichael Celeo: yeah! hm, let me try07:18
← small left07:20
→ ThorMojito joined07:20
IceMichael well, but it will still yield "not a git repository"07:20
← Aminda left07:20
IceMichael I mean, I could just directly access the files07:20
← ThorMojito left07:20
→ Aminda joined07:21
IceMichael so I would basically "git ref-parse HEAD" to work07:21
Celeo If you have the .git folder, yeah, that command would work. The _latest_ SHA is in that file, if that's all you need.07:24
IceMichael yeah, my issue is: I don't want to COPY my entire .git folder into my docker container just to get this command working07:24
.git is clearly large :)07:24
so I was thinking: maybe copy a subset...07:25
but now I'm thinking, maybe it's cleaner to just run the git command before and ship the result07:25
← elastic_dog left07:26
→ elastic_1 joined07:26
elastic_1 → elastic_dog07:26
gnoo you could try a shallow clone07:32
git clone --depth 107:33
IceMichael gnoo: yeah, but it's rather not 3rd party07:36
depth 0 would be nice lol07:37
← igemnace left07:38
→ zeenk joined07:38
→ ThorMojito joined07:41
← ThorMojito left07:42
Celeo Could also give `git clone --no-checkout` a run; depends on where the size you don't want is located - the git history, or the repo content07:43
IceMichael ah sorry, misunderstood depth07:44
Celeo: ah, I could combine it07:44
← thiago left07:48
→ dnh joined07:48
IceMichael ok, I need history but no checkout, so that helps, thank you guys :) 478M -> 22MB07:48
Celeo Great!07:53
IceMichael yes, thank you very much :)07:58
← rustyshackleford left07:59
IceMichael well, that shifts more from docker stuff into a outside-docker bash file but cannot be helped07:59
(with disadvantage that this does not work well on Windows, not sure how to solve this... maybe use Python?!)07:59
← derjanni left08:01
→ derjanni joined08:02
→ han-solo joined08:04
→ feriman joined08:14
← potash left08:18
→ igemnace joined08:20
→ fling_ joined08:24
← fling left08:25
← Techcable left08:25
→ potash joined08:30
fling_ → fling08:31
→ dksnd joined08:34
← YuGiOhJCJ left08:40
← jazzy left08:44
← dksnd left08:45
← jmd left08:48
→ jmd joined08:48
→ atrigent joined08:53
→ sedzcat joined08:54
← fling left08:59
→ fling joined08:59
← jfsimon1981 left09:05
→ jfsimon1981 joined09:06
→ nate4 joined09:10
→ sa0 joined09:12
← nate4 left09:15
→ YuGiOhJCJ joined09:16
→ bpaster joined09:17
bpaster is there a way to put a badge on README.md that shows the amout of git clones in github?09:17
or unique visitors09:17
→ EvilDMP joined09:19
← bodiccea left09:20
bookworm bpaster: #github is over there -->09:20
bpaster kk09:20
→ srinidhi joined09:21
→ ThorMojito joined09:22
← ThorMojito left09:22
→ bodiccea joined09:25
← YuGiOhJCJ left09:25
→ YuGiOhJCJ joined09:25
bpaster bookworm, are you a git pro?09:32
bookworm I can manage... wouldn't call myself a pro09:33
→ small joined09:33
bookworm if you ask an actual question, someone out of ~850 people probably knows the answer09:33
bpaster bookworm, i had a question but it solved itself lol. my git remote show origin didnt show my branches in 'Local refs configured for 'git push'... but after i restarted my computer it shows now for some reason o_o09:34
bookworm sounds like a GUI ?09:35
bpaster im using cli09:35
git version 2.34.109:35
it updated the09:35
'git pull side' correctly, but for some reason didnt show on push09:36
← t0mm13b left09:36
bookworm ah, I never issue git remote show09:36
bpaster you just have upstream in config?09:36
bookworm if I want to associate a branch, I push with -u... from then on I don't need to care09:37
and generally I know what's mapped to what, so I don't need to query that information09:37
(it's normally a one to one mapping anyhow)09:37
bpaster well i was working on another branch and i did the --upstream-set-to so remote show origin updated for pull, i could git pull without extras, but git push didnt work, but now it seems it works after i restarted...09:38
bookworm restarting your OS shouldn't impact git in the slightes t09:39
bpaster I agree 100% thats why i was confused09:39
bookworm considering that it dies between invocations, every new git command starts from scratch09:39
(it uses the fs as a cache, but those survive a reboot)09:39
→ atrigent_ joined09:39
bpaster bookworm, if you make pseudo branch now and remote it to another pseudo branch09:39
does it update both push and pull for you?09:40
← atrigent left09:40
bookworm s/remote/rename/ ?09:40
bpaster in 'git show remote origin'09:40
im a git newb so i have some wrong commands in histry... but i did 'git checkout -b tracker2' then 'git branch --set-upstream-to origin/fake2' / 'git push --set-upstream origin tracker2:fake2'09:42
and so on... and 'git remote show origin'09:42
it updated the git pull side09:42
but not push for whatever reason09:42
osse tried to help me but i think he had the same problem09:43
so i just resorted to changing confing to 'push.default upstream'09:43
bookworm well, _osse is what I'd call a git pro, so if they reproduced it, it is in fact that way (might be buggy)09:43
bpaster it still didnt show in 'git remote show origin' in push section09:43
oh okay09:44
bookworm, are you able to reproduce it?09:44
can i make a bug report somewhere or09:44
bookworm you can ask the git mailing list09:44
bpaster i have no idea why it shows now though, like you said after restarting my pc09:45
Local refs configured for 'git push': main pushes to main ... was only shown, but now it shows the other two branches as well09:45
← crabbedhaloablut left09:46
→ crabbedhaloablut joined09:46
→ rsx joined09:48
→ dionysus69 joined09:49
bookworm you'll need to get a reproducer (minimal), preferably something that can be run from an small script to get to the same state09:50
that is, if you want to poke the mailing list09:51
← sa0 left09:53
→ sa0 joined09:53
→ coot joined09:53
← Betal left09:56
← coot left09:59
→ t0mm13b joined09:59
→ coot joined10:00
← bodiccea left10:02
→ bodiccea joined10:02
← coot left10:05
← bodiccea left10:08
→ Techcable joined10:09
→ bodiccea joined10:10
→ ThorMojito joined10:12
← ThorMojito left10:12
→ vitali64 joined10:18
← vitali64 left10:19
→ jackneilll joined10:22
→ vitali64 joined10:24
← vitali64 left10:24
→ vitali64 joined10:24
vitali64 → f_10:25
← Murr left10:25
← bodiccea left10:26
→ Murr joined10:26
→ bodiccea joined10:26
← bodiccea left10:28
→ bodiccea joined10:33
← bodiccea left10:41
→ sedzcat1 joined10:43
← sedzcat left10:44
sedzcat1 → sedzcat10:44
→ bodiccea joined10:45
← teddyc left10:47
← skapata left10:47
→ coot joined10:47
→ peirik_ joined10:52
→ teddyc joined11:01
→ liefer39 joined11:08
→ memoryleak joined11:16
← peirik_ left11:18
→ peirik_ joined11:20
→ atrigent joined11:20
← atrigent_ left11:21
→ clime joined11:21
→ terminalpusher joined11:22
← zer0bitz left11:29
→ cdown joined11:33
→ mw_ joined11:36
← memoryleak left11:36
← mw_ left11:37
→ bgs joined11:38
→ memoryleak joined11:39
← f_ left11:40
← teddyc left11:40
→ vitali64 joined11:41
← EvilDMP left11:43
→ lucasta joined11:43
→ EvilDMP joined11:43
→ zer0bitz joined11:45
→ wootehfoot joined11:45
← peirik_ left11:50
← vdamewood left11:50
→ vdamewood joined11:51
→ teddyc joined11:53
← cdown left11:54
→ momomo joined11:58
→ humanface joined11:59
humanface hello11:59
if I set a github repo to public, then everybody can search and find that repo easily, and possibly add anything to that?11:59
so if I go to sleep and when I wake up, there could be modifications or the whole repo could have been deleted by a random user?12:00
or how does it work?12:00
← clime left12:00
← vitali64 left12:00
→ vitali64 joined12:01
→ roadie joined12:02
← vitali64 left12:03
← stemid left12:03
→ vitali64 joined12:04
→ john_johnk joined12:04
← jmd left12:05
→ stemid joined12:05
osse bpaster: I think it's just git remote show being weird, not that the configuration is weird12:12
nedbat humanface: a github public repo is not public-writable.12:13
vitali64 → f_12:14
→ ThorMojito joined12:16
← feriman left12:16
→ duxsco joined12:16
← ThorMojito left12:16
→ peirik_ joined12:17
← bpaster left12:20
← wootehfoot left12:25
→ jmd joined12:27
← YuGiOhJCJ left12:29
← jmd left12:34
← peirik_ left12:50
→ gareppa joined13:07
← duxsco left13:09
→ duxsco joined13:11
→ ThorMojito joined13:11
← ThorMojito left13:12
→ nate4 joined13:12
← f_ left13:16
← nate4 left13:17
← gareppa left13:18
→ gareppa joined13:19
← gareppa left13:20
→ gareppa joined13:20
← yappar left13:21
← coot left13:22
← OnkelBlade left13:22
→ jmd joined13:24
← roadie left13:24
→ OnkelBlade joined13:24
→ peirik_ joined13:25
→ szkl joined13:28
← OnkelBlade left13:29
→ yappar joined13:29
← wallabra left13:30
← peirik_ left13:30
→ OnkelBlade joined13:31
→ ThorMojito joined13:32
← ThorMojito left13:33
← yappar left13:34
gsi_ "public repo" in github means it's just publicly visible, readable13:36
but only the owner (you) or the people that _you_ allow to can write there13:37
when in doubt, than check _that_ provider's documentation (github happens to use git, but "isn't git")13:38
→ yappar joined13:43
← lucasta left13:46
← yappar left13:48
→ ThorMojito joined13:52
→ yappar joined13:53
← ThorMojito left13:53
← yappar left13:54
→ yappar joined13:54
← chexum left14:01
→ chexum joined14:02
→ mw_ joined14:05
→ kkawg joined14:07
← jmd left14:07
humanface gsi_ thanks14:07
→ jmd joined14:07
← jmd left14:08
humanface gsi_ so in general, if you want to share a test project with a recruiter (a lead dev), for example, then you usually set the repo to public, and give him the url, right?14:08
← rsx left14:08
→ peirik_ joined14:08
humanface and not the other way around (private repo, inviting the recruiter lead dev as collaborator)14:09
→ duxsco75 joined14:10
duxsco → Guest359014:11
→ jmd joined14:11
duxsco75 → duxsco14:11
→ ferdna joined14:11
← chexum left14:11
← Guest3590 left14:12
→ ThorMojito joined14:17
← ThorMojito left14:18
nedbat humanface: right. The recruiter can read but not write if you make it public. He can write to it if you invite them as a collaborator, and also that introduces a hurdle for them (they need a GitHub account, they need to accept, etc.)14:19
humanface nedbat ok thanks <314:19
← jmd left14:23
→ jmd joined14:24
← elastic_dog left14:30
→ lucasta joined14:33
→ ThorMojito joined14:34
← ThorMojito left14:34
← small left14:34
→ elastic_dog joined14:35
→ chexum joined14:36
→ ThorMojito joined14:39
← peirik_ left14:41
← dionysus69 left14:41
→ vitali64 joined14:42
vitali64 → f_14:42
← ___nick___ left14:42
← ThorMojito left14:43
→ rurtty joined14:43
→ ___nick___ joined14:43
← john_johnk left14:47
← duxsco left14:50
← mpldr left14:53
→ mpldr joined14:53
→ peirik_ joined14:53
← lucasta left14:54
← vdamewood left14:55
→ nosferandu joined14:55
→ vdamewood joined14:56
← peirik_ left14:58
→ Davidian joined15:01
← mw_ left15:07
→ madewokherd joined15:07
← madewokherd` left15:11
→ peirik_ joined15:12
→ lucasta joined15:12
→ ThorMojito joined15:13
← ThorMojito left15:14
→ fweht joined15:15
← peirik_ left15:16
← srinidhi left15:17
→ thuna` joined15:18
→ zmt01 joined15:20
← zmt00 left15:21
← lucasta left15:21
← yappar left15:23
← ferdna left15:27
→ alfredb joined15:33
→ zmt00 joined15:33
→ duxsco joined15:33
← zmt01 left15:34
→ rostero joined15:34
← sedzcat left15:41
← Xenguy left15:43
→ sedzcat joined15:45
→ adamlaska joined15:47
→ yappar joined15:55
→ lucasta joined15:55
← yappar left15:56
← adamlaska left15:57
→ moldorcoder7 joined15:59
→ ThorMojito joined16:17
← ThorMojito left16:17
→ Mordag joined16:21
→ Xenguy joined16:22
← kkawg left16:25
← Murr left16:25
→ Murr joined16:25
→ ThorMojito joined16:40
← ThorMojito left16:40
→ phil170 joined16:40
→ roadie joined16:40
← sedzcat left16:42
→ thiago joined16:43
← duxsco left16:44
→ duxsco joined16:47
→ geri joined16:50
← warmana left16:53
→ peirik_ joined17:00
→ nate4 joined17:13
← mcon left17:14
→ ThorMojito joined17:17
← ThorMojito left17:18
← roadie left17:18
← nate4 left17:18
humanface nedbat and do I need to send anything else to the recruiter near the github repo link?17:27
if the repo is public17:27
← Xenguy left17:32
← peirik_ left17:34
→ roadie joined17:37
← duxsco left17:39
gsi_ did you check the github provider's docs? this channel being git ...17:41
don't wait for others to tell you, check yourself17:41
use the URL (from a different account, from a friend's machine) and see whether it's accessible17:42
humanface gsi_ sorry for being stressful a bit, I need to send something tomorrow...17:44
gsi_ alternatively try accessing other people's public repos with just their URL17:45
it's the same situation, this time "you are the recruiter"17:46
or ask github specific questions in the github specific channel, they are in a much better position to help you17:46
humanface ok it seems, HTTPS url is working17:47
I have no idea why the SSH version is not17:47
← dsrt^ left17:51
← TonyStone31 left17:51
humanface gsi_17:55
→ Xenguy joined17:56
← lucasta left17:57
humanface so is it normal that a HTTPS url can be used to clone a public github repo, and the SSH url can't be used without some sort of authentication?17:57
imMute humanface: SSH *requires* authentication for git repos. I suppose it's technically possible to have anon access, but it would be tricky and I've never seen it done17:58
humanface: basically, SSH was built with client authentication from the get go, while HTTPS was not.17:59
humanface hmm I see17:59
thanks for the clarification17:59
→ rgrinberg joined18:03
→ TonyStone31 joined18:04
← rostero left18:04
→ coot joined18:05
← thuna` left18:12
← rurtty left18:12
← TonyStone31 left18:12
→ rurtty joined18:16
← rgrinberg left18:16
→ skapata joined18:18
← cweiss076532124 left18:18
→ cweiss076532124 joined18:21
→ TonyStone31 joined18:24
← Murr left18:25
→ Murr joined18:25
→ thuna` joined18:27
→ lucasta joined18:28
← rurtty left18:30
← atrigent left18:32
→ atrigent joined18:32
→ LuKaRo joined18:35
osse imMute: I think the "git" is like an anonymous ssh. never really understood how it worked18:39
oh, it's just a raw tcp thing18:40
← zeenk left18:41
mjt0k heh. what *is* "raw tcp"? :)18:41
osse I dunno man!18:42
networking confuses me!18:42
← thiago left18:42
mjt0k tcp is just a bi-directional stream, that's it. stream of bytes. how it's interpreted is up to the application. smtp, http, pop3, etc etc - that's raw tcp too, in your definition :)18:43
→ thiago joined18:43
ikke Yes, not everything will go over http or ssh :P18:43
mjt0k and git protocol ("language") is quite simple too. To me it has been sort of mistery how git lays out on top of http18:44
← LuKaRo left18:44
ikke though people try to shove more and more over just http(s)18:44
osse mjt0k: what I mean is that when you use git:// there is no procol "between" tcp itself and the git data, like there is with ssh or http(s)18:44
protocol18:44
ikke the protocol is git18:45
mjt0k well yes, this is git protocol18:45
simple and effective18:45
rewt just like the http protocol is on "tcp itself"18:45
ikke but yes, it's not wrapped18:45
mjt0k and git-over-http is kind of like translating the language into foreign language and back18:45
osse hmm, you seem to know more than me. what does the payload look like?18:46
mjt0k heh. how a git object look like? :)18:47
ikke dumb http is just http requests18:47
smart http I'm not familiar with the details18:47
osse no i mean for the git protocol18:47
ikke It's a binary protocol18:47
mjt0k https://git-scm.com/docs/protocol-v218:48
← terminalpusher left18:48
→ LuKaRo joined18:49
→ hamburgler joined18:51
osse hmm18:52
if I'm reading this right then apart from the initial request there is no difference between what the TCP payload is (when using git://) and what the e.g. HTTP body is18:54
→ alkhwarizmi joined18:55
mjt0k what *is" "tcp payload" ?18:55
ikke the byte of streams18:55
→ Ajax1465 joined18:55
→ chimbosonic1 joined18:55
ikke stream of bytes* :PP18:55
mjt0k tcp is a stream of bytes. it needs to be structured - either in http: or git:18:56
but the same objects are transferred in either case by git, yes18:56
← TonyStone31 left18:57
← Ajax146 left18:57
Ajax1465 → Ajax14618:57
mjt0k it's just the surrounding in "git-over-http" is much bigger usually :)18:57
btw, where 'hyper-text' in there is really interesting question ;))18:58
it has become an "etp", or "everything-transfer-protocol" :)18:59
← OnkelBlade left18:59
osse I think we're just talking past eachother here...19:00
← lucasta left19:01
→ OnkelBlade joined19:01
osse by tcp payload I mean the stuff that is not the tcp header19:01
mjt0k you don't see headers at the application level, it is all done by the kernel for you19:02
be it git or http or whatever19:02
osse of course19:03
mjt0k so everything you see is the tcp payload. and the git structures and commands are there *too*19:03
alkhwarizmi Anyone else fed up of personal access tokens and SSH keys? I wrote a Git credential helper that securely authenticates to GitHub, GitLab, BitBucket and other forges using OAuth. Help wanted to test https://github.com/hickford/git-credential-oauth19:03
ikke ssh keys work perfectly fine for me19:04
osse mjt0k: we're definitely talking past eachother here.19:05
i'm not at the application level, or at the command line19:06
my initial question about git:// is basically this: I open Wireshark, and capture that I clone over git:// and clone over http:// What are the difference in the captures?19:07
alkhwarizmi git is a higher level protocol ? the layering is: git over http over TCP over IP19:08
→ donofrio joined19:09
→ TonyStone31 joined19:10
← alkhwarizmi left19:12
→ lgc joined19:13
← lgc left19:13
mjt0k and they run away :)19:16
← igemnace left19:20
← jmd left19:28
→ jmd joined19:28
→ srinidhi joined19:31
→ peirik_ joined19:31
← EvilDMP left19:34
→ jazzy joined19:35
→ squirrel joined19:35
← han-solo left19:35
← gnoo left19:39
→ zer0bitz_ joined19:39
← zer0bitz left19:40
→ lucasta joined19:44
→ stef204 joined19:51
→ gnoo joined19:54
← coot left19:54
osse actually, I could just do that experiment myself I guess :p19:55
but I am organizing my cables19:55
→ zeenk joined20:00
← peirik_ left20:04
← thebombzen left20:04
← zeenk left20:05
← lucasta left20:06
→ theoceaniscool joined20:07
← perrierjouet left20:09
← Davidian left20:10
→ goldfish joined20:11
→ jetchisel joined20:12
← thiago left20:15
← alfredb left20:16
→ peirik_ joined20:16
→ DoubleJ joined20:16
← agowa339 left20:17
→ vqueiroz joined20:18
→ agowa339 joined20:18
← peirik_ left20:21
→ EvilDMP joined20:25
→ thiago joined20:26
← f_ left20:27
← BlessJah left20:27
← zenobius left20:31
→ BlessJah joined20:31
← srinidhi left20:32
→ lucasta joined20:35
← randm left20:40
→ randm joined20:41
← gareppa left20:53
→ feriman joined20:57
→ obiwahn joined20:58
obiwahn Hey - Why would I use `git worktree` when I can clone using `--reference`?20:58
ikke obiwahn: worktrees share configuration20:59
reference clones do not20:59
obiwahn ikke: Do you like use the feature?21:00
ikke I do use worktrees, yes21:01
at least in one case21:01
→ wallabra joined21:01
obiwahn thank you! I'll have a look at it. Until now I was a bit undecided if its worth the time / even thought it is probably not mega complex:)21:02
→ ferdna joined21:03
osse with --reference you still have to fetch in all the different repos you have21:04
← ___nick___ left21:04
→ oriba joined21:05
← wallabra left21:07
→ wallabra joined21:08
← oriba left21:10
← wallabra left21:11
← EvilDMP left21:14
→ nate4 joined21:15
→ EvilDMP joined21:16
← epony left21:16
→ wallabra joined21:18
← nate4 left21:20
→ madewokherd` joined21:23
← derpydoo left21:23
← madewokherd left21:26
→ tonymke joined21:30
← gnoo left21:33
← dirtydan left21:34
→ dirtydan joined21:34
← roadie left21:36
← geri left21:39
→ thebombzen joined21:41
← jmd left21:42
fat_owl → hoho21:43
→ Guest83_ joined21:45
← theoceaniscool left21:45
← Guest83_ left21:46
← bgs left21:50
← vdamewood left21:53
← stef204 left21:54
← Aminda left21:58
→ Aminda joined21:58
← feriman left21:59
→ kexec joined22:01
← drogas left22:03
dansan_ → dansan22:03
← tejr left22:04
→ epony joined22:04
→ coot joined22:04
→ gasbag joined22:04
← m0viefreak left22:06
→ peirik_ joined22:17
← gasbag left22:18
→ vyryls joined22:19
← lucasta left22:19
← Murr left22:25
→ Murr joined22:25
← coot left22:33
← vyryls left22:33
← phil170 left22:35
→ rgrinberg joined22:38
← R2robot left22:41
← elastic_dog left22:41
→ elastic_dog joined22:41
→ austin987 joined22:42
→ R2robot joined22:48
← austin987 left22:48
→ lucasta joined22:49
← peirik_ left22:51
← fweht left22:52
→ tejr joined22:54
← lucasta left22:58
→ lechner joined23:01
→ XV8 joined23:04
→ lagash joined23:05
→ peirik_ joined23:05
→ madewokherd joined23:06
→ acresearch1 joined23:08
acresearch1 people, i have a question, on github website, is it possible to delete an old commit? i own the project, i the only one in it and i want to delete a commit perminantly, is it possible?23:09
← madewokherd` left23:10
EdFletcher acresearch1: you can use interactive rebase, drop that commit, then force push23:10
← peirik_ left23:11
← sa0 left23:12
acresearch1 EdFletcher: what do you mean?23:14
is there a walk through i can read? i looked online but i haven't found an answer23:14
nedbat acresearch1: you get rid of the commit locally, then force-push to your main branch on github23:15
EdFletcher acresearch1: https://git-scm.com/book/en/v2/Git-Tools-Rewriting-History23:15
← humanface left23:15
→ agowa joined23:15
EdFletcher ah they do have anchors they're just hidden argh https://git-scm.com/book/en/v2/Git-Tools-Rewriting-History#_deleting_a_commit23:16
← charly left23:16
← agowa339 left23:19
→ peirik_ joined23:19
← chexum left23:21
acresearch1 EdFletcher: i do not have it locally, only on the website, does that mean i have to clone it locally first?23:21
EdFletcher acresearch1: yes, github is just a git hosting service (with some extra sugar on top). we are talking about git here.23:21
acresearch1 EdFletcher: ok 1 moment let me clone it locally23:21
→ gast0n joined23:22
→ chexum joined23:22
acresearch1 ok cloned,23:22
what should i do then? i have the commit hash that I want to remove23:23
EdFletcher acresearch1: you read the docs i sent https://git-scm.com/book/en/v2/Git-Tools-Rewriting-History#_deleting_a_commit23:23
acresearch1 ok 1 moment23:24
EdFletcher: so i just type drop hash ?23:25
that is it?23:25
EdFletcher: under the "delete a commit" section they do not include which command to use23:28
i think the safest thing to do is just delete the whole project23:29
thanks guys for your help23:29
→ alkhwarizmi joined23:30
→ rustyshackleford joined23:31
→ gnoo joined23:33
← alkhwarizmi left23:34
EdFletcher holy moly frijoles batman! i step away for a few minutes and you just go right for the nuclear option!?!23:35
← EvilDMP left23:38
gsi_ but it resolves the OP's question, does it not? that commit will be gone, mission accomplished23:38
doesn't take as long as reading some docs either23:38
→ ThorMojito joined23:39
← ThorMojito left23:39
← acresearch1 left23:40
← TonyStone31 left23:48
← dnh left23:49
← llh left23:52
← peirik_ left23:52
→ dnh joined23:57

Logs Search ←Prev date Next date→ Channels Documentation