IRCloggy #git 2024-08-16

Provider of IRC logs since 2005.

WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

	2024-08-16
	← Gab36 left	00:01
	← vrach left	00:05
	← intelfx left	00:06
	→ thuna` joined	00:25
	→ jacobk joined	00:29
	← Sazhen86 left	00:38
	← error27_ left	00:44
	→ error27__ joined	00:44
	← dlh404 left	00:44
	→ dlh4041 joined	00:44
	→ jaredce joined	00:52
	→ hbautista joined	00:53
	← Che-Anarch left	00:53
	→ Che-Anarch joined	00:54
	← davispuh left	00:54
	← memset left	00:55
	→ memset joined	00:57
	← jaredce left	00:57
	← tocrlisoal^ left	00:57
	← ptrck5 left	01:02
	← Che-Anarch left	01:02
	→ ptrck5 joined	01:02
	→ Che-Anarch joined	01:02
	→ troyt joined	01:08
	← mingdao left	01:10
	→ mingdao joined	01:10
	→ pete443_ joined	01:23
	← pete443 left	01:24
	→ peirik joined	01:24
	→ skered joined	01:25
	← hbautista left	01:27
	← peirik left	01:29
	→ dviola joined	01:48
	→ spit joined	01:52
	← Che-Anarch left	02:07
	→ Che-Anarch joined	02:08
	→ lucasta joined	02:10
	← gas51627 left	02:15
	← thuna` left	02:24
	→ bbailey joined	02:25
	← wobbol left	02:44
	← gast0n left	02:47
	→ peirik joined	02:47
	← peirik left	02:52
	← ChmEarl left	02:55
	← hdh left	02:55
	→ jmd joined	02:55
	← lucasta left	02:57
	→ gsi joined	03:09
	← gsi_ left	03:11
	← Lord_of_Life left	03:14
	→ JupiterBig joined	03:14
	→ Lord_of_Life joined	03:16
	← JupiterBig left	03:20
	→ cero joined	03:22
	→ kiwiirc joined	03:33
	← jmd left	03:36
	→ jmd joined	03:37
	← Betal left	03:37
	← kiwiirc left	03:48
	→ jaredce joined	03:50
	← dlh4041 left	03:53
	← jaredce left	03:55
	← Erisa1 left	03:57
	→ Erisa1 joined	03:57
	← ackyshake left	03:58
	← Erisa1 left	03:58
	→ Erisa1 joined	03:59
	← Erisa1 left	04:05
	→ Erisa1 joined	04:06
	← Minvera left	04:07
	← Erisa1 left	04:10
	← jmd left	04:10
	→ jmd joined	04:11
	→ Erisa1 joined	04:12
	← jmd left	04:14
	→ jmd joined	04:15
	← JordiGH left	04:16
	← jmd left	04:20
	→ jmd joined	04:20
	← jmd left	04:21
	→ jmd joined	04:21
	← mooz left	04:41
	← Erisa1 left	04:45
	→ Erisa1 joined	04:47
	→ stagas joined	04:49
	← ali1234 left	04:49
	→ ali1234 joined	04:50
	→ Spr0cket joined	04:52
	→ lemuria joined	04:53
	← gryffus_ left	04:56
	← kjartan_ left	04:56
	→ kjartan_ joined	05:02
	→ mooz joined	05:10
	← stagas left	05:12
	← pabs3 left	05:15
	→ pabs3 joined	05:16
	→ gryffus_ joined	05:17
	→ cixx joined	05:19
	→ kale joined	05:29
	→ maret joined	05:34
	← maret left	05:34
	→ maret joined	05:35
	← jmd left	05:39
	← jiffe3 left	05:41
	→ jiffe3 joined	05:43
	→ Guest71 joined	05:47
	→ pedahzur joined	05:48
	→ jaredce joined	05:51
	→ peirik joined	05:52
	← maret left	05:54
	← jaredce left	05:56
	← FH_thecat left	06:00
	← Erisa1 left	06:01
	→ Erisa1 joined	06:02
	→ Blasius joined	06:09
	← SArpnt left	06:15
	← pedahzur left	06:18
	→ vrach joined	06:26
	→ bookworm_ joined	06:28
	→ masoudd joined	06:29
	→ FH_thecat joined	06:31
	← peirik left	06:32
	bookworm_ → bookworm	06:32
	→ Blasius_ joined	06:37
	← demonspork left	06:41
	← Blasius left	06:41
	→ theoceaniscool joined	06:46
	→ peirik joined	06:49
	← cycoder left	06:51
	→ danse-nr3 joined	06:51
	→ jaredce joined	06:52
	→ becm joined	06:52
	→ ackyshake joined	06:54
	→ fdan joined	06:54
	→ cycoder joined	06:55
	← jaredce left	06:58
	→ gandhibobandhi_ joined	07:03
	→ Baniola joined	07:03
	← Baniola left	07:07
	→ Baniola joined	07:11
	→ Mooncairn_ joined	07:14
	← Baniola left	07:15
	→ Baniola joined	07:16
	→ stagas_ joined	07:16
	→ bob_x1 joined	07:16
bob_x1	good day. I have bare repository and trying to clone it. I see next warning. Does it mean I lost my code ?	07:17
	Warning: Remote HEAD referst to nonexistent ref, unable to checkount	07:18
	→ jaredce joined	07:18
osse	Do you have the repo locally?	07:18
	→ ferdinandyb joined	07:19
	← Baniola left	07:21
bob_x1	yes I have it somewhere. but these are backups. and backups seems not working	07:21
	and I have several repos with this error	07:21
j416	bob_x1: what's under objects/ in those repos?	07:22
bob_x1	j416: folders 02,11,1b ... etc...	07:23
	← jaredce left	07:23
j416	so there's _something_ at least, that's reassuring.	07:23
	does 'git log' work?	07:23
bob_x1	in bare repo ?	07:23
j416	yeah	07:23
	← a6IO left	07:24
bob_x1	yes it works	07:24
j416	git fsck --full	07:25
ferdinandyb	Hi, is there a setting for clone that essentially runs "git remote set-head -a origin" on cloning?	07:25
j416	(aha, --full is now default, didn't know; 'git fsck' then)	07:25
	← theoceaniscool left	07:25
j416	ferdinandyb: since clone is a read thing, I would doubt it	07:26
	ferdinandyb: curious, what do you need that for?	07:26
bob_x1	j416: one dangling commit	07:27
ferdinandyb	j416: wdym by "read thing"? it already sets a bunch of refs and does a checkout, or is there some inherent thing I'm missing here?	07:28
j416	bob_x1: dangling commits are fine	07:28
ferdinandyb	j416: the use case is that I never know if a project is main/master/trunk whatever	07:28
	→ Baniola joined	07:29
bob_x1	j416: then git fsck works fine	07:29
ferdinandyb	j416: so I try not to hard code things like origin/master in an alias an instead use "refs/remotes/origin/HEAD"	07:30
j416	ferdinandyb: ah, I was under the impression that that updated HEAD on the remote, but I suppose it's local	07:30
	ferdinandyb: I see, makes some sense	07:30
	my bad.	07:30
	bob_x1: then by the looks of it, you didn't lose your project	07:31
	← Baniola left	07:31
j416	bob_x1: I assume you were able to successfully clone it. Try to check out a branch.	07:31
bob_x1	can I checkout a branch from the bare repo ?	07:32
j416	from your clone of it	07:32
	ferdinandyb: fwiw, when I run that, it changes nothing in my local clone	07:35
	→ Baniola joined	07:35
j416	ferdinandyb: i.e. .git/refs/remotes/origin/HEAD already existed and running 'git remote set-head -a origin' did not change its contents	07:36
	it did touch it though (it's mtime changed)	07:36
	its*	07:36
bob_x1	j416: no luck. clone of bare repo produces only .git folder and empty list of branches	07:36
j416	bob_x1: git branch -r	07:36
bob_x1	but something changed. now it does not show error 'Unable to checkount'	07:37
ferdinandyb	j416: hmm, I'm pretty sure I ran into this at some point, but it may have been in an older version? let me check	07:37
j416	this is 2.46.0 fwiw	07:37
bob_x1	j416: ah yeah... now works. and code restored	07:37
	thank you very much.	07:37
j416	ferdinandyb: repo cloned from github, not sure if that matters or not	07:37
	bob_x1: o/	07:38
	← danse-nr3 left	07:38
bob_x1	j416: what has happened? can you explain ?	07:39
j416	bob_x1: I don't know how you ended up like that. sorry.	07:40
bob_x1	ok	07:40
j416	I don't think I've seen that error before	07:40
bob_x1	thank you once again. very good channel	07:40
	← bob_x1 left	07:40
j416	but, remote's HEAD is just a reference	07:40
	and he left	07:40
	→ danse-nr3 joined	07:40
	← skapata left	07:45
	← danse-nr3 left	07:45
	← Baniola left	07:45
	→ john_johnk joined	07:46
	← kjartan_ left	07:48
ferdinandyb	j416: I just checked on a github repo and indeed it is set	07:51
	← john_johnk left	07:52
ferdinandyb	and it's also set on an sr.ht clone, so I guess what I ran into earlier was a one-time or something	07:52
	→ kjartan joined	07:53
j416	I wonder how it might have happened	07:53
	I bet it wasn't set in bob's case above	07:54
ferdinandyb	j416: actually I do have an idea	07:57
	git blame to the rescue 😄	07:57
j416	do tell if you find it	07:57
ferdinandyb	https://github.com/ferdinandyb/dotfiles/commit/a3eb7d948b70e8cc52760e698f213163c21da6df	07:57
j416	ooh	07:58
	obviously	07:58
	not set on 'git init' even if a remote is added, then	07:58
	but set if you clone the repo; makes sense	07:58
ferdinandyb	yeah, so in this case, I'm looking for a git add setting 😄	07:59
	I mean git remote add setting	07:59
j416	I like your relatively neat history for your dotfiles.	08:00
	helps in times of these :)	08:00
ferdinandyb	there's already a "git remote add -m master origin" that manually adds HEAD, I think "git remote add -a origin" would make sense	08:00
j416	huh	08:01
	why master	08:01
	that's odd	08:01
	→ jaredce joined	08:01
ferdinandyb	j416: thanks, I got into the habit in general, and try to do it for my dotfiles especially when I feel that it's something not quite obvious	08:02
j416	sane unless remote's HEAD can be found to be something else, I suppose	08:02
	← Gamah left	08:02
ferdinandyb	j416: master is just the examples in man git-remote	08:02
gitinfo	j416: the git-remote manpage is available at https://gitirc.eu/git-remote.html	08:02
j416	ferdinandyb: yeah, I do that too. best ever.	08:02
ferdinandyb	LOL	08:03
	nice bot	08:03
	🙂	08:03
osse	gitinfo: cookie	08:03
j416	ferdinandyb: oh, my bad. I was reading the man page and didn't understand that by <master> they meant whatever you put	08:03
osse	aww	08:03
j416	!botsnack	08:03
gitinfo	Om nom nom	08:03
osse	ah	08:03
j416	!cookie	08:03
	→ warmana joined	08:03
	→ Gamah joined	08:03
ikke	!ping	08:05
gitinfo	I only respond to !reply_time	08:05
ferdinandyb	j416: sooo, what's the preferred way of requesting features	08:05
	?	08:05
	← daru left	08:05
j416	ferdinandyb: mailing list	08:05
osse	If I understand you correctly you want to change the config of a remote repo upon clone?	08:09
j416	they want a way to set refs/remotes/origin/HEAD automatically to the remote HEAD not only on clone, but also when adding a remote, I think	08:10
osse	Ah	08:10
	I thought that already happened, heh	08:10
ferdinandyb	osse: j416 exactly, yes, I just want to automatically have git remote add set refs/remotes/<remote>/HEAD	08:11
j416	or, maybe you want fetch to do it	08:11
	remote add afaik does not use the network	08:11
ferdinandyb	j416: that maybe makes more sense actually, yes	08:12
	considering I made the alias fetchall to do just that	08:13
j416	:)	08:13
ferdinandyb	ok, so the actual feature I want is for "git fetch" to have a setting, flag whatever to set the remote/HEAD	08:14
j416	it'd need to be updated on every fetch then I suppose	08:15
	I have a feeling this should already be the way it works; maybe it's not. ¯\_(ツ)_/¯	08:15
	also afaik (I may be mistaken), there's no way from the remote HEAD to know for sure which branch it corresponds to, because it points at a hash, not a ref, so Git has to guess which ref it might be	08:16
	← Guest71 left	08:16
	→ Baniola joined	08:17
ferdinandyb	j416: I'm not sure I follow, cat .git/refs/remotes/origin/HEAD -> ref: refs/remotes/origin/maste	08:17
	← Baniola left	08:18
j416	ferdinandyb: locally yes, but the remote HEAD. See output of git ls-remote	08:18
	→ derpydoo joined	08:19
ferdinandyb	https://0x0.st/X44m.txt, you mean if multiple branches were pointed to a726909879d we could not know which branch?	08:20
j416	that's my understanding	08:20
ferdinandyb	most forges only allow for a single "default" branch, so I'd guess there's at least a convention of not doing that	08:20
	← kale left	08:22
j416	I created a github repo just now with three branches 'master', 'x', 'y' pointing at the same commit, and set the default branch in github's UI to 'x'. A fresh clone checks out branch 'x'. I changed the default branch (in the UI) to be 'y', and a fresh clone again correctly checks out branch 'y'. I don't know what sorcery is happening here.	08:23
	→ warmana1 joined	08:23
	→ gas51627 joined	08:23
j416	ls-remote says https://x0.at/SKpz.txt	08:24
ferdinandyb	then there's probably somethign else going on, yeah	08:24
	← warmana left	08:25
	warmana1 → warmana	08:25
	→ warmana1 joined	08:26
j416	ah, it actually stores it, it just doesn't show it in the output of ls-remote	08:27
	odd, I wonder where I picked up that it tries to guess it.	08:27
	git ls-remote --symref	08:27
	→ danse-nr3 joined	08:28
j416	sorry for the noise and thanks, I learnt something.	08:28
ferdinandyb	j416: me too 🙂	08:29
j416	\^^/	08:30
	→ KnoP joined	08:30
	← warmana left	08:30
	warmana1 → warmana	08:30
ferdinandyb	ok, so fetch does _not_ set remote/HEAD	08:30
	for sure	08:30
	← rustyshackleford left	08:31
j416	--symref is new-ish (only 8 years old) git::99c08d4eb28f	08:33
gitinfo	Git web link: https://github.com/git/git/commit/99c08d4eb28f	08:33
ferdinandyb	"newish"	08:35
	TIL that detached head is when your head is _not_ a symref	08:36
	so, next up writing to the ML	08:38
	← cixx left	08:38
j416	indeed; detached HEAD means HEAD points directly to a commit	08:39
	→ Baniola joined	08:39
	← fdan left	08:40
	← Baniola left	08:42
	← jmjl left	08:44
	→ jmjl joined	08:44
	← gryffus_ left	08:44
	→ redbool joined	08:45
	→ gryffus_ joined	08:48
	→ Baniola joined	08:55
	← ferdna left	08:59
	→ kale joined	09:01
	← danse-nr3 left	09:03
	→ danse-nr3 joined	09:04
	← jacobk left	09:04
	→ jacobk joined	09:05
	← Baniola left	09:08
	← Cork left	09:08
	→ Baniola joined	09:11
	→ ferdna joined	09:12
	← Baniola left	09:13
	← danse-nr3 left	09:19
	→ danse-nr3 joined	09:21
	→ Baniola joined	09:26
	→ hiroot joined	09:30
	← hiroot left	09:31
	→ tlatelolco3 joined	09:32
	→ john_johnk joined	09:42
	← Baniola left	09:44
	→ Baniola joined	09:45
	→ gnoo joined	09:54
	← Che-Anarch left	09:55
	pikapika_lunar → militantorc	09:55
	→ Che-Anarch joined	09:55
	← danse-nr3 left	10:07
	← ferdna left	10:11
	→ danse-nr3 joined	10:11
	← xandris left	10:12
	→ xandris joined	10:14
	→ epicout joined	10:14
	← Baniola left	10:15
	→ Baniola joined	10:17
	← Baniola left	10:24
	← stagas_ left	10:26
	→ Baniola joined	10:26
	← masoudd left	10:26
Walex	ferdinandyb: "if multiple branches were pointed to a726909879d we could not know which branch?": 'git' has no branches, as such, they are just a way of saying... Usually that way of talking works, in some cases it does not reflect well what is happening.	10:31
	→ haliaeetus joined	10:31
	← derpydoo left	10:32
	← spit left	10:36
	→ Saksham_Baba joined	10:45
	← Saksham_Baba left	10:45
	→ Saksham_Baba joined	10:45
	← john_johnk left	10:47
	→ spit joined	10:48
	→ Betal joined	10:55
	← epicout left	10:56
	→ Maestro joined	10:59
	→ john_johnk joined	11:05
	← danse-nr3 left	11:05
	← Baniola left	11:05
	← peirik left	11:10
	← gryffus_ left	11:13
	← dviola left	11:14
osse	ferdinandyb: y u no write to list yet >:((((	11:16
krumelmonster	I messed up in git mergetool (vimdiff) and I want to go back to before I called git mergetool. I did not write any changes but still the file moved from "both modified" to "modified" so I cannot just git mergetool over	11:16
	→ gryffus_ joined	11:16
osse	krumelmonster: git checkout --merge path/to/file	11:17
	→ Baniola joined	11:17
ferdinandyb	krumelmonster: git reset ORIG_HEAD	11:17
osse	That'll "restore" the file to the state it had immediately after failing the merge	11:17
ferdinandyb	osse: not yet, but will do so	11:18
krumelmonster	thank you osse that did it	11:18
	← john_johnk left	11:21
	→ dviola joined	11:24
	→ lacrymology joined	11:24
	→ JordiGH joined	11:24
	← lacrymology left	11:25
	→ lacrymology joined	11:25
ferdinandyb	osse: which one is the correct mailing list for this? [email@hidden.address]	11:27
osse	ferdinandyb: yes	11:27
	!list	11:27
gitinfo	[!mailing_list] The mailing list can be reached via [email@hidden.address] You don't need to subscribe to the list, you will always be put in cc on reply. Read archives at http://public-inbox.org/git	11:27
ferdinandyb	ack	11:27
osse	AFAIK that is the only list	11:27
	← ackyshake left	11:27
osse	Remember to send plaintext or else the grumpy old guys don't like it	11:28
	ferdinandyb uses aerc and is also grumpy about unnecessary html 😄	11:30
	→ Gab87 joined	11:30
Gab87	hey guys , i have a question	11:30
	say i have created a hotfix branch , and i have merged it already into my master branch	11:31
	but i want to undo that action	11:31
	← mrpops2ko left	11:31
Gab87	like it didnt exist before	11:31
	how can i achieve this	11:31
	i have read there are several ways , git revert , git reset but i am not sure which one is the proper one to choose for my case	11:32
Foxboron	Gab87: Are others reading from master?	11:32
Gab87	what do you mean ?	11:32
Foxboron	Are other developers using "git pull" on the master branch?	11:33
Gab87	ah i'm not there yet , i am just trying to create a repository that represents this diagram , like a task to learn about gitflow	11:33
	https://cdn.hashnode.com/res/hashnode/image/upload/v1668069961266/fI6dAXt_8.png	11:33
Foxboron	okay, then the proper way to do this is with "git revert" and get a revert commit	11:34
Gab87	so you see , in this diagram , i merged hotfix too early , i should have created feature for after 1.0 first	11:34
Foxboron	"git reset" should only be done on local branches that others are not reading from	11:34
Gab87	according to the timeline	11:34
	→ alexherbo2 joined	11:34
Gab87	if you don't mind , can you explain a little more	11:35
	about git reset	11:35
	why do we use git revert and not git reset	11:35
	← Saksham_Baba left	11:35
Foxboron	"git reset" alters history, while "git revert" creates a commit that removes the changes with a new commit	11:35
	You don't alter history of branches that other people are also working towards	11:36
	→ Saksham_Baba joined	11:36
Gab87	by creating a new commit , wouldn't that change the structure of the repository ?	11:36
	and prevent me from achieving the goal , thus is having a repo similar to the diagram	11:36
osse	ferdinandyb: you mean you "use aerc btw" ? :P	11:38
	→ AmR joined	11:39
krumelmonster	I think I finally did the merge in vimdiff as mergetool. I mostly copied over "REMOTE". I'm still in the four pane vimdiff mode (and it's very hard to read for me still). I would now like to check a final diff of REMOTE and the merged file to check for mistakes. Either still in vim or out in the shell	11:39
	← lacrymology left	11:39
Gab87	but if i use git reset to delete my hotfix branch , how does that alter history of branches that other people are working also on . is it because the merge commit will be deleted in the master branch , which the is the branch that other people are working on	11:39
Chipzz	krumelmonster: you can commit locally and compare your local and remote commit btw	11:40
ferdinandyb	osse: lol 😄 we should make that a catchphrase	11:40
Chipzz	if your lcoal commit is not what you want you cna still amend it	11:40
	→ alfredb joined	11:41
krumelmonster	hmm	11:42
ferdinandyb	is there a way to reference the current branches remote? like @{u}, but not the actual upstream branch, rather basically I'd like to access refs/remotes/[currentremote]/HEAD	11:42
Foxboron	ferdinandyb: @{push}?	11:42
krumelmonster	status says modified: mergedfile. can't I git diff mergedfile against the commit I'm still in the process of merging in?	11:43
Gab87	Foxboron can you look at my question	11:43
Chipzz	krumelmonster: also: "git diff file origin/master -- file" I think on top of my head	11:43
	← Saksham_Baba left	11:43
	→ Saksham_Baba joined	11:43
	← Baniola left	11:44
Foxboron	Gab87: You have several. I'm not sure how to answer them in an educational fashion as you don't really want to use "git reset" to fix these mistakes.	11:44
Chipzz	Gab87: git reset doesn't delete any branches. And even if it would, that would only affect your local repo. If you pushed to a repo before and other people cloned that, they would not be affected. But you would be "behind"	11:45
ferdinandyb	Foxboron: "In a non-triangular workflow, @{push} is the same as @{upstream}, and there is no need for it."	11:45
Gab87	okay , so as i am understanding , git revert does not undo the history , it brings it forward , say i want to delete commit A , it will create a new commit where Commit A was deleted	11:45
	but then , that changes the structure of my repository , doesn't it ?	11:46
	i am trying to achieve this :	11:46
	https://cdn.hashnode.com/res/hashnode/image/upload/v1668069961266/fI6dAXt_8.png	11:46
Foxboron	ferdinandyb: hrmmmm	11:46
ferdinandyb	Foxboron: consider you're on a branch that has upstream as "gitlab/mybranch", the project also has another remote called origin. I want to "git log gitlab/HEAD^.."	11:47
Gab87	Chipzz i dont want to delete the branch , i just want to undo the merge commit that was merged into master branch , without changing the structure of my repository , if you get what i mean	11:47
	so by using git revert , wouldn't it create extra commits on the main branch ?	11:48
	maybe , if you check out this image , you'll understand what i mean	11:48
	https://cdn.hashnode.com/res/hashnode/image/upload/v1668069961266/fI6dAXt_8.png	11:48
	→ Baniola joined	11:48
krumelmonster	Chipzz: `git diff --staged <hash of the cherry picked commit> -- file` did the trick, wonder if there'd been a way that didn't have me research the commit hash first... (I'm cherry picking a series)	11:49
Foxboron	ferdinandyb: whats the purpose?	11:49
ferdinandyb	Foxboron: well that particular example I gave lists the commits on your branch not on remote/HEAD, if there are some forks or not automatically updated mirror lying around it might make sense to not check this against origin/HEAD but currentremote/HEAD	11:51
Foxboron	ferdinandyb: so is the push remote origin or currentremote? and is upstream suppose to be origin or currentremote?	11:53
	because to me it sounds like you are describing the triangular setup as the manpage is explaining	11:54
	(vim-fugitive actually displays this properly in :Git status if you set it up properly. Which is quite neat)	11:55
Chipzz	Foxboron: you should note that the name origin is merely a convention (and a pretty vague one at that too), you can name it whatever you want	11:55
	(and in the case of a triangular setup you should probably rename it to be less confusing)	11:56
Foxboron	Chipzz: Sure, but is that relevant to this problem?	11:56
ferdinandyb	the @{upstream} of the current branch is "currentremote/mybranch". I want to run "git log currentremote/HEAD^..", but in an alias. This means I need to do some scripting to get the current upstream, find the name of the remote from there and construct the string"currentremote/HEAD". In practice, I currently just have "origin" hardcoded.	11:58
	Foxboron: so this is not the difference between pull and push URL-s if that is what you are asking	11:59
	@{push} and @{origin} are the same	11:59
	→ JupiterBig joined	12:00
Foxboron	right, got it	12:01
	← JupiterBig left	12:02
Chipzz	krumelmonster: right I assumed you saved the file to disk	12:05
	→ dlh404 joined	12:07
	← Gab87 left	12:09
	← Saksham_Baba left	12:11
Foxboron	ferdinandyb: Okay, i got busy with geoguessr, but I also can't seem to figure out how to generally get the push remote name?	12:15
	← jaredce left	12:21
	→ davispuh joined	12:23
	← Baniola left	12:24
	← Maestro left	12:25
pagios	hello all, i am looking for a book and/or resource to understand how to design my web application. mainly would like to know when to use classes, functions, helper functions etc, something that can organize and help me create modular app. Any recommendations? i am using js	12:26
gsish	got better chances asking that in a place where web developers are?	12:29
	→ stagas joined	12:29
	← yctn left	12:29
	→ Baniola joined	12:33
	← AmR left	12:33
	→ AmR joined	12:34
	→ JupiterBig joined	12:34
	→ jaredce joined	12:34
	→ ws2k3 joined	12:36
	→ user_oreloznog joined	12:37
	← Baniola left	12:38
ferdinandyb	Foxboron: yeah, I think you have script that with something currently	12:38
	error27__ → error27	12:38
Foxboron	ferdinandyb: I suspect something like git config --local branch.$(git branch --show-current).remote	12:38
	to get the value, else you need to parse the output of show-refs or something	12:38
	→ mrpops2ko joined	12:39
	← jaredce left	12:39
ferdinandyb	osse: https://public-inbox.org/git/D3HBD7C1FR14.74FL1Q1S9UCB@ferdinandy.com/T/#u	12:40
Foxboron	oh, this is a problem I've had!	12:41
	→ Cork joined	12:41
	← error27 left	12:42
	→ error27 joined	12:42
osse	ferdinandyb: sniff it's beautiful :´)	12:47
	→ PlasmaHH joined	12:47
	← skered left	12:48
	→ skered joined	12:49
	→ rsx joined	12:49
PlasmaHH	Hi, I am doing something like code quality in a project, where I create commits way faster than we can review and especially merge stuff... so my idea was to bundle things into small enough pieces so that people feel able to rewview and merge them chunk by chunk. now in terms of my branch of things, I was wondering how to set it up or if there is a common thing done in similar situations:	12:51
	I could either have my branch, then when its "full" create a new one on top of that, and when its merged rebase onto mainline. or I could have one huge branch with all my changes and from that cherry pick into a "presentation branch" and when that is merged rebase from mainline...	12:51
	→ gandhibobandhi__ joined	12:51
	→ jaredce joined	12:52
	→ osc4rpt joined	12:54
Foxboron	PlasmaHH: the latter approach is often refered to as "stacked branches", fwiw	12:54
	← jmjl left	12:54
	→ Baniola joined	12:54
Foxboron	git recently got better support for this with rebase.updateRefs so it's a reasonable approach	12:55
	← gandhibobandhi_ left	12:55
Foxboron	or rather, what you are describing isn't exaclty stacked branches. But it's the approach you probably want	12:56
	exactly*	12:56
	← gryffus_ left	12:57
	→ gryffus_ joined	12:57
Chipzz	Foxboron: you should be aware that different web frontends (including github) have varying levels of support for stacked branches when recommending them	12:58
PlasmaHH	Foxboron: ah intresting, update-refs seems to be a really useful thing there	12:58
Chipzz	if he happens to be on one that doesn't support stacked branches very well it may not be a good suggestion	12:59
	← Baniola left	12:59
PlasmaHH	Chipzz: we use gitlab however I don't care much about how it is displayed there, all git operations are to be done locally usually	12:59
Foxboron	Chipzz: This isn't relevant for the question	13:00
ikke	PlasmaHH: I sometimes use git machete	13:00
Foxboron	forge is orthogonal to the question "how do i solve this in git"	13:00
	the forge is*	13:00
Chipzz	Foxboron: I think it is	13:01
ikke	gitlab-cli recently also included a feature to help with a stacked workflow	13:01
Chipzz	his diffs still need to be reviewed	13:01
Foxboron	Chipzz: Then you should inform them instead of trying to interject the person providing the support	13:02
PlasmaHH	Chipzz: we use gitlab merge requests for reviews, I don't see how that could be a problem	13:02
Chipzz	PlasmaHH: there's several things to consider wrt the forge, ie how does it handle changes in lower parts of your branch stack	13:04
PlasmaHH	ikke: it seems that git machete only works on local branches?	13:04
Chipzz	how does reviewing work when you're reviewing different parts	13:04
PlasmaHH	Chipzz: there would only ever be one merge request from the stack be active	13:05
Chipzz	PlasmaHH: it really depends on your workflow (and that of your colleagues), but people may want to review different parts of your stack independently and at different times\	13:07
PlasmaHH	Chipzz: we only do reviews of merge requests	13:07
Chipzz	PlasmaHH: that's bad then	13:08
PlasmaHH	Chipzz: why?	13:08
Chipzz	lets say you're working on a feature, and your stack is 1) add new API to existing classes 2) use the new API to actually build the new feature	13:08
	you would want to review the new API separately to see if it makes sense	13:09
	the reason to have stacked diffs is to split up a big change into smaller changes that are easier to digest, and that may affect different teams in your company	13:10
PlasmaHH	Chipzz: which is why I dont issue a merge request for the branch that is using the api before I added the new api (besides that these kinds of changes wont happen for my current use case anyways)	13:10
	it would not make sense to merge in the usage of the new api when the code implementing it is not yet there...	13:11
Chipzz	so you're not really using stacked diffs then?	13:11
PlasmaHH	Chipzz: I am not using anything, I have asked for what is the best way to do in my situation	13:11
	and while certainly sometimes one branch branches from another, this isnt a particular named or defined workflow here	13:12
ikke	PlasmaHH: it syncs them with the upstream branches as well	13:13
PlasmaHH	ikke: just skimmed the docs etc. and tried "git machete discover" which only uses locals but couldnt find anything to make it see those on origin too	13:14
	→ danse-nr3 joined	13:14
Chipzz	PlasmaHH: I would say it all really depends on what the desired workflow is, for both you and your colleagues	13:16
PlasmaHH	Chipzz: well thats why I asked above from my point which works better. for the colleagues it is as usual, reviewing the merge requests, and I control which pieces of my code I "reveal" to them via those.	13:17
	anything not in merge requests is considered nonexistent for official code reviews	13:18
	→ rainystorm joined	13:18
Chipzz	maybe committing the API part of your feature separately works for you, but it's probably not a good idea. Using stacked branches allows your colleagues to request last minute changes to your API for example	13:18
	it greatly improves the chance both parts of your feautre are coherent, and you won't need commits later to fix design mistakes	13:19
PlasmaHH	Chipzz: sure, that doesnt seem to be a problem. we do that, and then have to adapt the part using that api too.	13:19
kjetilho	PlasmaHH: since you call it merge requests, have you looked at merge trains in GitLab?	13:20
	it is a thorny problem to be able to continue working while waiting for MR's to be accepted and merged :/	13:20
	← kale left	13:21
Chipzz	PlasmaHH: it all really depends on what kind of software you're working on I guess. it may work in your case, as you say, and it may not work in other cases	13:23
	or maybe you just haven't encountered cases yet where it won't work out very well	13:23
PlasmaHH	kjetilho: for one I am quite limited in what I can use and configure there, and also I think I remember those only working witihn the same repo?	13:23
	→ Minvera joined	13:24
Chipzz	PlasmaHH: in the end it's sth you'll have to evaluate for yourself, only you (and your colleagues) can be the judge of that	13:25
kjetilho	PlasmaHH: hrm - but if it is in different repos, you don't need them to reduce rebase work?	13:25
Chipzz	PlasmaHH: one concrete example is this: you have a WIP branch with part 1 and 2, someone is reviewing part 1 and comments on certain parts of the diff, you make changes and rewrite that part. The commit the reviewer was reviewing is no longer relevant, where do the comments he made go?	13:27
PlasmaHH	kjetilho: its not to reduce rebase work but to make smaller "packages" of changes for review and merges to happen... its was basically impossible the last time I did something similar, where I had ~ 1700 changes all over the codebase and I had to find ways to break it into smaller changes to have people review and merge them	13:28
Chipzz	is gitlab smart enough to recognize your rebased part 1 is the same part as the previous part 1 and retain the comments?	13:28
kjetilho	Chipzz: yes, if you commit the change to the MR branch	13:28
Chipzz	anyway those are the kind of things you need to keep in mind when evaluating if stacked diffs are right for you or not	13:29
PlasmaHH	Chipzz: isnt that example more of a question whether to use rebases or not? we almost always rebase before a merge	13:31
Chipzz	different kind of rebase	13:31
	I'm guessing you're rebasing on master before merging?	13:31
PlasmaHH	we rebase on the branch that the merge targets	13:32
Chipzz	right, but that's not what I was referring to	13:32
	what I was referring to is rewriting parts of your stack	13:32
	or maybe inserting extra commits out of order	13:33
kjetilho	ideally these parts should be independent. if not, well, more rebasing.	13:34
Chipzz	lets say you have Base <- API <- feature	13:34
	colleague reviews API and requests changes	13:34
	now you have Base <- API' <- feature'	13:35
	or maybe you have Base <- API1 <- API2 <- feature'	13:36
PlasmaHH	and your point is that comments attached to API MR will not be there anymore because gitlab doesnt know how to put them in API' ?	13:37
kjetilho	that would be false. GitLab will keep track of that.	13:38
Chipzz	like kjetilho pointed out they will be there if you do it right, but yes	13:38
	that's historically been a problem with some of the forges, not sure what the current status is and if things improved	13:38
PlasmaHH	seems to be a nonproblem then for me	13:40
Chipzz	a search in the group of former engineers of a company I worked in before has had multiple discussions about how to work with stacked diffs over the years	13:40
PlasmaHH	I would be happy if people would at least try to get their heads around submodules here... people dont like git much mostly here	13:41
Chipzz	PlasmaHH: what programming language are you working in?	13:42
PlasmaHH	Chipzz: C++	13:42
	(mostly)	13:42
Chipzz	PlasmaHH: cmake + vcpkg. vcpkg enables you to make overlay ports. no more submodules needed	13:43
	or chances are the need for them will be reduced, at least	13:43
PlasmaHH	Chipzz: vcpkg has been rejected because it makes the build process slow and more importantly it makes it much more involved to do the API+feature simultaneous change thing	13:44
	→ Baniola joined	13:44
Chipzz	personally I dislike using strategies like submodules for embedded code copies	13:44
	PlasmaHH: if it makes it slow then your engineers did a bad job. If they would have done their job corretcly they would have set up a cache which would actually have increased build sppeds	13:45
PlasmaHH	Chipzz: I dont know the specifics they tried. but the main point was really making the development too much involved when changing libraries	13:48
Chipzz	WDYM changing libraries?	13:49
PlasmaHH	Chipzz: you see, our code base consists of several product repositories and several library repositories for shared parts	13:49
	← alfredb left	13:51
Chipzz	"several library repositories" sounds like a good match for a vcpkg registry	13:53
PlasmaHH	Chipzz: I wasnt involved in that experiment but it seems that it was deemed too convoluted to properly develop a feature in a library together with the code that is using it because you had to publish it etc. instead of just pushing it to the right branch in the submodule	13:55
Chipzz	hmmm I get what you're saying	13:57
	→ Left_Turn joined	13:57
Chipzz	but it sounds like a mess either way	13:58
	because in the end you would still have to push from your submodule	13:58
	doesn't sound like fun either	13:59
PlasmaHH	its not the greatest thing and at all kinds of places and we gradually try to improve it but getting enough developers to change is hard ;)	13:59
	well its at least just one operation instead of whatever is necessary to create the package and then possibly have to wait for it to be ready or I dont know	14:00
Chipzz	I'm really not sure if using a submodule is that much better than using a separate checkout for the library involved	14:00
	→ lucasta joined	14:00
Chipzz	You don't have to wait for a package to be ready at all	14:00
	You can use overlay ports in vcpkg	14:00
PlasmaHH	the submodule has the advantage that with one git command everything is there	14:00
Chipzz	which are ports thaht are local to the git repo you're working in	14:00
PlasmaHH	you meann like pip install -e and virtual environments? well either way I dont have the influence there to revise that decision at the moment	14:03
	also I dont see what the advantage could be, you have seperate repos either way, its just how they are glued together, you still have to check into one before being able to use it in the other	14:04
	← alexherbo2 left	14:14
	← JupiterBig left	14:18
	→ jensen joined	14:18
	→ JupiterBig joined	14:21
	← jfsimon left	14:26
	→ JupiterB2g joined	14:26
	→ jfsimon joined	14:26
	→ Turn_Left joined	14:26
	← Left_Turn left	14:30
	← spit left	14:31
	← dsimic left	14:34
	→ john_johnk joined	14:35
	→ dsimic joined	14:36
	← Baniola left	14:39
	→ misanthropos joined	14:46
	→ gblfxt joined	14:47
	→ rustyshackleford joined	14:47
	← JupiterBig left	14:48
	← JupiterB2g left	14:48
	→ jmjl joined	14:48
	← AmR left	14:49
	← Mooncairn_ left	14:50
	← tlatelolco3 left	14:50
	← kurahaupo left	14:56
	→ tavares joined	14:56
	→ epicout joined	15:00
	← PlasmaHH left	15:02
	→ JupiterB1g joined	15:03
	→ JupiterB2g joined	15:04
	← haliaeetus left	15:04
	← john_johnk left	15:06
	→ thuna` joined	15:08
	→ jmd joined	15:11
	← danse-nr3 left	15:12
	← error27 left	15:14
	→ error27 joined	15:14
	← jmd left	15:14
	→ jmd joined	15:14
	← stagas left	15:15
	→ AmR joined	15:16
	→ fonini joined	15:22
	← JupiterB2g left	15:26
	← JupiterB1g left	15:26
	← KnoP left	15:36
	→ SArpnt joined	15:36
	→ gast0n joined	15:39
	← Ox7C5 left	15:42
	→ JupiterB1g joined	15:43
	→ JupiterB2g joined	15:43
	← JordiGH left	15:46
	← jfsimon left	15:46
	→ Mooncairn joined	15:47
	→ ChmEarl joined	15:48
	→ IceMicha- joined	15:53
	← IceMicha- left	15:56
	→ IceMicha- joined	16:00
	← Che-Anarch left	16:00
	→ Che-Anarch joined	16:00
	← rsx left	16:01
	← JupiterB2g left	16:01
	← JupiterB1g left	16:01
	← rainystorm left	16:01
	← de-facto left	16:02
	→ de-facto joined	16:06
	→ john_johnk joined	16:07
	← osc4rpt left	16:13
	← Traneptora left	16:14
	← jaredce left	16:15
	→ darth joined	16:16
	← jmd left	16:20
	← warmana left	16:20
	← thiago left	16:21
	← darth left	16:22
	→ pgnd joined	16:22
	→ jmd joined	16:22
	→ DarthChillash joined	16:22
	→ Traneptora joined	16:26
	← jmd left	16:26
	→ jmd joined	16:26
	jensen → aNr2	16:36
	→ Left_Turn joined	16:36
	← jmd left	16:36
	→ jmd joined	16:37
	← jmd left	16:38
	← Turn_Left left	16:38
	→ hdh joined	16:42
	→ jmd joined	16:45
	← john_johnk left	16:45
	→ skapata joined	16:47
	→ jaredce joined	16:52
	→ Turn_Left joined	16:52
	← gandhibobandhi__ left	16:54
	→ ackyshake joined	16:56
	← Left_Turn left	16:56
	← jaredce left	16:57
	← thuna` left	17:01
	← mooz left	17:07
	→ mooz joined	17:13
	→ thelounge6897480 joined	17:18
	→ ferdna joined	17:25
	← hdh left	17:27
	→ hdh joined	17:30
	← gryffus_ left	17:35
	← crabbedhaloablut left	17:40
	← Che-Anarch left	17:42
	→ crabbedhaloablut joined	17:42
	← hamburgler left	17:42
	→ Che-Anarch joined	17:42
	→ gryffus_ joined	17:43
	→ hamburgler joined	17:46
	→ jaredce joined	17:53
	← becm left	17:53
	← SArpnt left	17:58
	← jaredce left	17:58
	→ PantelicGR joined	18:01
	→ jaredce joined	18:11
	← gryffus_ left	18:16
	← jaredce left	18:16
	→ wootehfoot joined	18:35
	→ AmR_ joined	18:38
	→ duxsco joined	18:39
	→ gryffus_ joined	18:40
	← AmR_ left	18:40
	→ AmR_ joined	18:41
	→ alexherbo2 joined	18:41
	← wootehfoot left	18:41
	→ wootehfoot joined	18:42
	← AmR left	18:42
	→ thuna` joined	18:43
	← ash_worksi left	18:43
	→ rainystorm joined	18:46
	→ thiago joined	18:57
	← Che-Anarch left	19:00
	→ Che-Anarch joined	19:00
	→ Left_Turn joined	19:04
	← Turn_Left left	19:07
	→ Turn_Left joined	19:13
	← Left_Turn left	19:17
	← Turn_Left left	19:18
	→ Left_Turn joined	19:18
	← duxsco left	19:21
	← Jck_true left	19:24
	→ Jck_true joined	19:25
	← alexherbo2 left	19:31
	← pgnd left	19:37
quesker	how can I pop the stash and just overwrite this file, not merge?	19:38
j416	quesker: git checkout yourstash -- yourfile	19:39
quesker	ok thanks	19:39
j416	...then delete the stash.	19:40
quesker	oh I want to keep it. weird after pop it was still there	19:41
	← ferdna left	19:45
	← toric left	19:49
	← wootehfoot left	20:00
	← jacobk left	20:00
	→ jacobk joined	20:01
	→ duxsco joined	20:13
	← duxsco left	20:15
	← thiago left	20:19
	← memset left	20:23
	→ memset joined	20:23
	← lucasta left	20:26
	→ haliaeetus joined	20:26
	← SuperLag left	20:27
	→ SuperLag joined	20:34
	cbreak_ → cbreak	20:34
	← gryffus_ left	20:34
	← AmR_ left	20:35
	→ gryffus_ joined	20:38
	→ greybaloon joined	20:39
	→ duxsco joined	20:45
	← theobjectivedad left	20:50
	→ qgnox joined	20:51
	← duxsco left	20:52
	→ theobjectivedad joined	20:52
	→ duxsco joined	20:53
	← rainystorm left	20:54
	← jacobk left	20:55
	→ jacobk_ joined	20:55
	← greybaloon left	20:58
	→ m0viefreak joined	21:00
	← gast0n left	21:04
	← duxsco left	21:06
	← jmd left	21:08
	→ gast0n joined	21:08
	→ greybaloon joined	21:10
	→ masoudd joined	21:16
	→ YuGiOhJCJ joined	21:17
	← memset left	21:32
	→ memset joined	21:32
	→ wootehfoot joined	21:40
	← fonini left	21:46
	← memset left	21:48
	→ memset joined	21:49
	← Che-Anarch left	21:50
	→ Che-Anarch joined	21:50
	← greybaloon left	21:59
	← Che-Anarch left	22:05
	← wootehfoot left	22:05
	→ DrVortex joined	22:05
	→ randm joined	22:06
	→ Che-Anarch joined	22:06
	→ Gab2 joined	22:07
Gab2	hello	22:07
	need some urgent help	22:07
	i have created a local repo that looks like this :	22:07
	https://cdn.hashnode.com/res/hashnode/image/upload/v1668069961266/fI6dAXt_8.png	22:07
	→ howdoi joined	22:08
Gab2	then i wanted to push the local repo to github	22:08
	i first added an origin	22:08
	then i accidently did git push origin master	22:08
	which pushed only the master branch	22:08
	how do i undo this	22:08
	i have to push other branches as well ( everything ) . how to do this	22:09
	can anyone assist?	22:09
	→ shailangsa joined	22:10
	← DrVortex left	22:14
weaksauce	Gab2 just push the other branches or use --all to push every branch	22:21
Gab2	yeah i got it already ! thankss	22:22
	so now in my repo	22:22
	there's alot of button saying	22:22
	compare and pull request	22:22
	whats that for	22:22
	← Gab2 left	22:29
	← masoudd left	22:31
	→ lucasta joined	22:31
	→ paddymahoney joined	22:33
	← user_oreloznog left	22:33
	← Left_Turn left	22:39
	← PantelicGR left	22:54
	← epicout left	22:58
	→ demonspork joined	23:02
	← Vonter left	23:11
	→ thiago joined	23:31
	← SJrX left	23:32
	→ SJrX joined	23:34
	→ jaredce joined	23:41
	→ JupiterBig joined	23:43
	← jaredce left	23:45
	← qgnox left	23:46
	← JupiterBig left	23:48
	→ Gary_Sinise joined	23:51
	← davispuh left	23:54

Logs Search ←Prev date Next date→

Channels Documentation