IRCloggy #git 2019-01-10

Provider of IRC logs since 2005.

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

	2019-01-10
	← sanscoeur left	00:00
	← Toadisattva left	00:02
	→ foul_owl joined	00:04
	← finalbeta left	00:04
	← mat001 left	00:04
supernovah	jast: ty	00:05
	→ mat001 joined	00:06
	← CodeSlingerPaul left	00:07
	← mat001 left	00:08
	→ mat001 joined	00:08
	← n000g left	00:10
	← dfee left	00:11
	→ aw1 joined	00:13
	← jhulten left	00:13
	→ dfee joined	00:14
	← dslegends left	00:15
	→ jhulten joined	00:16
	← m0viefreak left	00:16
	← dfee left	00:17
	→ TheRuckus joined	00:20
	→ Kaisyu joined	00:21
	→ andyhuzhill joined	00:22
	← jhulten left	00:23
	→ n000g joined	00:23
	→ causasui joined	00:24
causasui	I do `git merge --no-commit master`. There's a merge conflict. I fix the conflict, add the file that had the conflict, and do `git commit`. Now I have to enter a commit message. Why? I don't want a merge commit.	00:24
	Do I have to fix the merge conflict in advance before merging?	00:25
GreenJello	causasui, you have to rebase if you want to avoid the merge conflict	00:25
	merge commit*	00:25
	causasui, you need a commit that resolves the conflict, unless you rewrite history so that your local changes are modified such that there isn't a conflict, which a rebase lets you do	00:26
causasui	GreenJello: I've already pushed -.- I think this means I have to add another commit to resolve the conflict in advance. I don't mind figuring that out, but it will look weird in the history I think	00:27
	← sylario left	00:27
GreenJello	causasui, I don't think you can do that	00:28
causasui	I can cherry-pick the commits out of master lol	00:28
GreenJello	causasui, unless you remove your local change to the file entirely by having it match what you're merging in	00:28
	→ jhulten joined	00:29
GreenJello	and even then, you still need a merge commit I think	00:29
	← Boobie left	00:30
GreenJello	causasui, either merge and accept that there will be a commit, or rebase your changes to avoid it	00:30
causasui	GreenJello: don't see how rebase is an option without force pushing	00:30
	maybe I misunderstand what you mean	00:30
GreenJello	causasui, not at this point, but in the future	00:30
causasui	Ironically, there's no way I could anticipate this, because I'm making changes to the CI configuration so I have to push to test them	00:31
GreenJello	you can rebase the commits that haven't been pushed	00:31
causasui	but that's neither here nor there	00:31
GreenJello	it's always safe to do that, and doesn't require a force push	00:31
	but now that you pushed, you'd be rewriting the remote history, which is best to avoid	00:32
	causasui, the simple workflow is: try to push, it gets rejected, rebase on the remote, a push will work unless someone ninja commits while you're doing this	00:33
	in which case... your back at step 1	00:33
	you're*	00:33
causasui	yeah I get that but thanks	00:34
	tbh this is all an artifact of the dumb CI system that stores its project configuration in git itself	00:34
	but w/e, I'll just bite this bullet, thanks	00:34
	← jhulten left	00:37
j416	rullie: I don't quite understand. Do you want to stay on A and stage the changes of B?	00:38
	rullie: git checkout A && git read-tree B # ?	00:38
	→ thiago joined	00:39
	← Phylock left	00:39
	← CheckDavid left	00:40
	← lxk left	00:43
	→ CheckDavid joined	00:45
	→ jhulten joined	00:46
	← jhulten left	00:52
	← donofrio left	00:52
	← SGrover left	00:53
	→ jhulten joined	00:58
	← dege left	00:58
	← TheRuckus left	01:00
	← flying_penguin left	01:00
	→ TheRuckus joined	01:03
	← jhulten left	01:03
	→ justan0theruser joined	01:07
	→ jhulten joined	01:07
	→ dfee joined	01:07
	← justanotheruser left	01:07
	← mat001 left	01:09
	→ mat001 joined	01:10
	← mat001 left	01:14
	→ ferdna joined	01:15
	→ mat001 joined	01:16
	← dpyro left	01:16
	← mat001 left	01:20
	← wildlander left	01:24
	← \void left	01:27
	→ Bobdude joined	01:32
	← alyptik left	01:39
	→ apragmatist_ joined	01:40
	← orbyt_ left	01:43
	← Zexaron left	01:44
	xcm → Guest21039	01:47
	← Guest21039 left	01:47
	→ xcm joined	01:47
	→ greatgatsby joined	01:48
rullie	j416: eh, precisely! Thanks! :D	01:52
	rullie googles what the hell's read tree	01:52
	← causasui left	01:53
	→ rkazak joined	01:54
	← veegee left	01:54
	→ redrambles joined	01:56
	← durham left	01:56
	→ libertyprime joined	01:56
	→ Goplat joined	02:00
rullie	j416: git read-tree -mu is exactly what i needed	02:00
	is read-tree well known and i'm just under a rock?	02:01
bremner	it's plumbing	02:01
	so it depends on your definition of well known	02:02
rullie	guess i don't plumb enough	02:03
	← jhulten left	02:04
	← jnewt left	02:05
	← andyhuzhill left	02:06
	← libertyprime left	02:07
	→ andyhuzhill joined	02:07
	← jakogut left	02:07
	→ orbyt_ joined	02:08
	← andyhuzhill left	02:11
	→ andyhuzhill joined	02:11
	← greatgatsby left	02:12
	← rkazak left	02:12
	← a-chymera left	02:15
	→ afidegnum joined	02:16
	→ mat001 joined	02:16
afidegnum	hello, i clone a repo, how do i reset all available tags and commits to 1.0.0 ?	02:16
	→ KeyJoo joined	02:19
	→ finalbeta joined	02:19
	→ dreiss joined	02:20
	→ rkazak joined	02:22
	← KeyJoo left	02:24
	→ KeyJoo joined	02:25
afidegnum	exit	02:25
	← afidegnum left	02:26
	← royal_screwup21 left	02:27
	← KeyJoo left	02:28
	→ alyptik joined	02:34
	← strugee left	02:35
	← Sasazuka left	02:36
	→ strugee joined	02:42
	← mowcat left	02:44
	← al-damiri left	02:45
	← duderonomy left	02:46
	← seni left	02:46
	← manuels left	02:46
	← manuelschneid3r left	02:46
	→ KeyJoo joined	02:49
	← khisanth_ left	02:50
	← strugee left	02:50
	← sweatsuit left	02:54
	← ketas left	02:55
	→ strugee joined	02:56
	→ sweatsuit joined	02:57
	→ Hathadar joined	02:59
	→ ketas joined	02:59
	← CheckDavid left	03:00
	← dreiss left	03:01
	→ duderonomy joined	03:02
	→ dpyro joined	03:03
	← waveclaw left	03:03
	→ dreiss joined	03:03
	← Essadon left	03:04
	← R2robot left	03:07
	→ shaarad joined	03:07
	→ waveclaw joined	03:08
	→ khisanth_ joined	03:08
	← YuGiOhJCJ left	03:10
	← dfee left	03:10
	→ moei joined	03:12
	← mooasaurus left	03:14
	→ mooasaurus joined	03:14
	← unreal left	03:15
	← boombatower left	03:15
	→ stitched_unicorn joined	03:18
	→ nowhere_man joined	03:19
	← Mattiaslndstrm left	03:20
	→ kerframil joined	03:21
	← shaarad left	03:22
	← Dirak left	03:25
	→ dfee joined	03:26
	← stitched_unicorn left	03:27
	← freeman42x left	03:28
	→ shaarad joined	03:32
	→ thebope joined	03:35
	→ veegee joined	03:35
	← diogenese left	03:45
	→ diogenese joined	03:46
	← rkazak left	03:48
	→ jubal_ joined	03:49
	→ jhulten joined	03:52
	← jubal left	03:52
	justan0theruser → justanotheruser	03:54
	→ rkazak joined	03:54
	→ Dirak joined	03:55
	→ sanscoeur joined	03:56
	← shaarad left	03:58
	lagothrix → Guest87941	03:58
	← Guest87941 left	03:58
	→ lagothrix joined	03:58
	→ Wetmelon joined	03:59
	← topdownjimmy left	04:06
	→ topdownjimmy joined	04:06
	← xlued left	04:08
	← sanscoeur left	04:08
	→ Mattiaslndstrm joined	04:09
	→ xlued joined	04:09
	→ sanscoeur joined	04:10
	← rkazak left	04:18
	← nowhere_man left	04:20
	← Crell left	04:22
	→ Sasazuka joined	04:23
	← alyptik left	04:24
	→ Pistos joined	04:25
	← neunon left	04:26
	← Sasazuka left	04:27
Pistos	Can anyone tell me how, given two repositories, to cherry-pick/rebase/whatever B's history into a subdirectory of A? Details at https://stackoverflow.com/questions/54121752	04:27
	→ neunon joined	04:28
	→ ziyourenxiang joined	04:29
	→ carb0n joined	04:32
	→ Cabanossi joined	04:32
	← Cabanoss- left	04:32
	→ crimastergogo joined	04:35
	← Anthaas left	04:39
	→ Anthaas joined	04:40
	→ jessica_b joined	04:42
	← MACscr_ left	04:48
	→ midnightmagic joined	04:49
	← ngui left	04:50
	→ ngui joined	04:51
	← jhulten left	04:51
	→ rkazak joined	04:51
	← dfee left	04:51
	← cdown left	04:52
	← dpyro left	04:52
	← Mattiaslndstrm left	04:53
	→ dfee joined	04:54
	→ alyptik joined	04:54
	← finalbeta left	04:55
	← aw1 left	04:55
	← andyhuzhill left	04:56
	→ andyhuzhill joined	04:57
	→ keepLearning512 joined	04:57
	← duderonomy left	04:58
	→ jhulten joined	04:59
	→ pks_ joined	05:02
	← sanscoeur left	05:02
	← bdonnahue left	05:03
	← pks left	05:04
	pks_ → pks	05:04
	← ngui left	05:05
	→ ngui joined	05:05
	← andyhuzhill left	05:06
	→ andyhuzhill joined	05:06
	← jhulten left	05:07
	← snowgoggles left	05:07
	→ duderonomy joined	05:08
	← sharkboy left	05:08
	→ sharkboy joined	05:08
	→ aw1 joined	05:08
	← aw1 left	05:09
	← duderonomy left	05:12
	← jubal_ left	05:17
	← Tiyanki left	05:19
	← ferdna left	05:20
	→ jubal joined	05:20
	→ Tiyanki joined	05:21
	← thebope left	05:21
	← Tiyanki left	05:22
	→ thebope joined	05:22
	→ Tiyanki joined	05:22
	→ cdown joined	05:22
	← Tiyanki left	05:23
	← comptroller left	05:23
	→ Tiyanki joined	05:23
	→ Sasazuka joined	05:24
	← thebope left	05:26
	← Sasazuka left	05:28
	← rkazak left	05:30
	← ngui left	05:30
	→ ngui joined	05:31
	→ pwnz0r joined	05:31
	← mikecmpbll left	05:31
	← andyhuzhill left	05:31
	→ andyhuzhill joined	05:32
	→ rnmhdn joined	05:33
rnmhdn	how can I turn a directory into a github repo?	05:33
orbyt_	git init	05:34
	git push	05:34
	done	05:34
	← RustyShackleford left	05:35
	← ngui left	05:35
	→ ngui joined	05:35
	← andyhuzhill left	05:36
rnmhdn	can you elaborate on the options?	05:36
	→ andyhuzhill joined	05:36
rullie	rnmhdn: when you make a new repo on github there's a short paragraph	05:36
	→ nowhere_man joined	05:38
atrus	Pistos: would it matter if the commits were the same shas, and in repo A, you'd see the actual commit adding file.txt? because if that's okay, you could add a commit to B that moves everything in B into a sub-directory B (mkdir B; git mv * B/; git commit -A -m 'move B contents into B subdir'), then add the B repo to A as a remote (git remote add B ../b/; git fetch B), and just merge that new commit in (git merge B/master), and you'd be done.	05:39
	← dfee left	05:40
	→ ToBeFree joined	05:41
	← ToBeFree left	05:41
	← xcm left	05:42
	← redrambles left	05:42
rnmhdn	rullie: I can't find that short paragraph	05:44
	→ xcm joined	05:45
	← carb0n left	05:45
	→ carb0n joined	05:46
rullie	rnmhdn: After you click on Create repository, the entire next page is dedicated to helping you set it up	05:47
	can't miss it	05:47
	← crimastergogo left	05:50
craftyguy	I frequently find that I am working on a branch, make many edits/commits, some of which are related to the overall spirit of the branch but should be submitted as a separate patch series (err, aka 'PR'). are there any good articles, suggestions, etc on workflows to account for this, avoid this, or generally deal with this?	05:51
	← psprint left	05:52
	→ cdown_ joined	05:52
	← keepLearning512 left	05:52
rnmhdn	I did it using https://stackoverflow.com/questions/3311774/how-to-convert-existing-non-empty-directory-into-a-git-working-directory-and-pus#3311824 and https://stackoverflow.com/questions/37937984/git-refusing-to-merge-unrelated-histories-on-rebase#37938036	05:53
craftyguy	Creating a new branch for each, cherry-picking relevant commits, and maintaining many of these special branches is tough. especially if the original branch which inspired them ultimately depends on those patches	05:53
	← orbyt_ left	05:54
	← cdown left	05:54
	cdown_ → cdown	05:54
	← carb0n left	05:59
	→ carb0n joined	06:02
	← Dirak left	06:03
	← cdown left	06:06
	← andyhuzhill left	06:06
	→ andyhuzhill joined	06:07
rullie	craftyguy: branch the branches?	06:07
	← Fusl left	06:07
craftyguy	rullie: ok. but managing a patch series on branches of branches when a branch depends on it (and managing the branch + other branches of branches, etc) is tedious.	06:09
	→ Fusl joined	06:09
craftyguy	Maybe that's the way of git. I'm just curious of others have found a better way than what I currently do	06:09
	→ crimastergogo joined	06:09
	→ Dirak joined	06:09
	→ redrambles joined	06:10
	← andyhuzhill left	06:11
	→ andyhuzhill joined	06:11
	→ cdown joined	06:12
	→ pionen joined	06:12
	← pionen left	06:13
	→ esrse joined	06:14
	← redrambles left	06:15
	← Dirak left	06:15
	← cdown left	06:16
	→ Dirak joined	06:18
	← midnightmagic left	06:19
	→ fairuz joined	06:23
	← fairuz left	06:24
	← carb0n left	06:26
	→ carb0n_ joined	06:26
	← ahmed89 left	06:27
	→ fairuz joined	06:29
	→ lamont joined	06:30
	← ngui left	06:30
	← netj left	06:30
	→ ngui joined	06:31
	→ netj joined	06:31
	→ midnightmagic joined	06:33
	→ ahmed89 joined	06:34
	← fairuz left	06:34
	← ngui left	06:35
	→ ngui joined	06:35
	← aleph- left	06:37
	→ sanscoeur joined	06:37
	← nowhere_man left	06:38
	→ thebope joined	06:38
	← kjartan left	06:39
	← sanscoeur left	06:40
	→ kjartan joined	06:41
	→ n3wborn joined	06:44
	← Dirak left	06:46
	→ Dirak joined	06:47
	→ dfee joined	06:52
	→ stitched_unicorn joined	06:52
	→ FermiDirak joined	06:53
	→ andrzejv joined	06:56
	← FermiDirak left	06:58
	← Inline left	06:59
	→ keepLearning512 joined	07:00
	→ oxymoron93 joined	07:02
	→ Leopere joined	07:05
	→ redrambles joined	07:06
	← pwnz0r left	07:09
	→ FermiDirak joined	07:10
	← redrambles left	07:11
	← thiago left	07:12
	← FermiDirak left	07:15
	→ AWindowsKrill joined	07:18
	→ duderonomy joined	07:20
	→ Envil joined	07:23
	← Envil left	07:24
	→ Xiti` joined	07:25
	← Xiti` left	07:26
	← Xiti left	07:26
	← Bobdude left	07:29
	→ R2robot joined	07:32
	→ AlHafoudh joined	07:33
	→ Xiti joined	07:33
	→ jas4711 joined	07:36
j416	rullie: if you actually want to merge and not just read data into the index, perhaps git-merge with --no-commit is easier to remember.	07:36
	(you'll have to abort the merge if you don't need the commit, of course)	07:36
	← dreiss left	07:37
	→ eigenvalue joined	07:38
	← z\|bandito left	07:41
	→ sQVe joined	07:42
	← cbrt64 left	07:43
	← crimastergogo left	07:44
	→ crimastergogo joined	07:44
rnmhdn	how can I ignore some files?	07:47
	put a .ginignore next to .git	07:47
	or use .git/info/exclude	07:47
	→ cbrt64 joined	07:47
	→ lxk joined	07:47
	← Goplat left	07:48
	← mat001 left	07:49
	← kerframil left	07:50
	← ahmed89 left	07:50
oxymoron93	.gitignore is usually tracked by repo so everyone can have the same ignore rules rnmhdn while .git/info/exclude has rules which you don't share	07:51
rnmhdn	I have my django db_password in a file	07:52
	which one should I use?	07:52
	there repo is public	07:52
oxymoron93	exclude	07:52
rnmhdn	so dir is next to .git and pass.txt is in dir	07:53
	→ dreiss joined	07:53
	← stitched_unicorn left	07:53
rnmhdn	I add a single line 'dir/pass.txt' to the end of exclude?	07:53
oxymoron93	tho you have another option, if you wan't to make all repos on your machine to ignore that file	07:53
	← WolfOrwell left	07:54
rnmhdn	that file is not in any other repo	07:54
	it's only in this repo	07:54
oxymoron93	ok then use ./git/info/exclude and play with it, I guess dir/pass.txt should do	07:55
	→ chele joined	07:55
	← ziyourenxiang left	07:58
oxymoron93	by the way, in case you misunderstood, when you ignore some file, git will ignore it, it won't ask you to start tracking such file, etc, thus you will never share the file, git would only have information about what it is ignoring, and not about the content of ignored files, so you could have used any kind of gitignore file, but for the sake if .gi	08:02
	tignore changes in future, it is better to use `exclude` file	08:02
	→ jubal_ joined	08:06
	← AWindowsKrill left	08:08
	→ AWindowsKrill joined	08:08
	← jubal left	08:09
	→ mloskot joined	08:09
	← sauvin left	08:12
	← satifant left	08:12
wadadli	exclude?	08:14
	TIL	08:14
	→ sauvin joined	08:17
	→ dege joined	08:19
	← shentino left	08:20
	→ T_UNIX joined	08:20
	→ jmd joined	08:28
	← hahuang65 left	08:29
jmd	I seem to recall there is an option to rebase which updates all the timestamps of unpushed commits. Can someone remind me what that option is?	08:29
	→ satifant joined	08:30
	→ mona_x joined	08:30
	← ngui left	08:30
	→ ngui joined	08:31
oxymoron93	--ignore-date jmd	08:31
	check man git rebase for future reference	08:31
gitinfo	the git-rebase manpage is available at https://gitirc.eu/git-rebase.html	08:31
jmd	oxymoron93: I am aware of the man page.	08:32
	Thanks for the pointer anyway.	08:32
	→ jason69 joined	08:35
	← Wetmelon left	08:35
	→ hahuang65 joined	08:35
	→ hofmann3900 joined	08:38
	→ Makaveli7 joined	08:38
	← Dirak left	08:40
	← andyhuzhill left	08:41
	→ andyhuzhill joined	08:42
	← crimastergogo left	08:44
	→ thm joined	08:45
	← ngui left	08:45
	→ crimastergogo joined	08:45
	→ ngui joined	08:45
	← andyhuzhill left	08:46
	← regiontog left	08:46
	→ andyhuzhill joined	08:46
	→ regiontog joined	08:48
	→ Mo joined	08:51
	→ segnior joined	08:51
	← Makaveli7 left	08:52
	← wyre left	08:53
Mo	Hi, I have a clone of a P4 repo, and continuously do git p4 sync. Now the email adresses of the remote p4 users have changed, but my local clone still has the old adresses. Where does git keep those adresses?	08:53
	I can't re-sync the repo from scratch, because the rev ids are not guaranteed to be stable, and I need those for an existing bug tracker..	08:54
grawity	each commit stores its author information in full (and it's part of the data that comprises the commit id)	08:54
	→ floppydh joined	08:55
grawity	it's not an external lookup like in p4 as you're describing	08:55
Mo	grawity: Does that mean, new commits coming from the git p4 sync those are still passing the old adresses from P4?	08:55
	Then I would need to check that on P4 site. But there, p4 users have all the new adresses.	08:56
grawity	if `git p4 sync` gets old information for new commits from somewhere, then it's probably cached somewhere in .git/p4* would be my guess	08:56
	← plasticle left	08:56
grawity	or possibly in .git/config under [git-p4]	08:57
Mo	Initially was created by git p4 clone --bare //TEST@all	08:57
	grawity: No p4* there, config has only 3 lines in [core]	08:58
	→ wyre joined	08:58
	→ dcpc007 joined	08:59
Mo	I have no idea why git log still shows the old adresses.	08:59
grawity	do you mean it shows old addresses for new commits that are synced after the change, or for old ones?	09:00
	→ nowhere_man joined	09:00
Mo	grawity: No, new commits after new syncining.	09:04
	And as I don't know if users meta data are only transferred at initial clone or on every sync, I don't know if this is cached.	09:05
grawity	do native p4 tools on the same machine show the updated information, at least?	09:05
	→ mikecmpbll joined	09:05
	→ fairuz joined	09:06
Mo	Yes, they do. p4 users shows the new mail adresses	09:06
	← eigenvalue left	09:08
Mo	p4 changes (like git log) doesn't show mail adresses, only internal user names.	09:08
	A new clone has the old adresses as well, so actually I expect the error on P4 site, need to check that first..	09:10
	→ stitched_unicorn joined	09:11
	→ Noti joined	09:13
	→ wxg joined	09:14
	← dreiss left	09:15
	← andyhuzhill left	09:16
Mo	grawity: Is there any $HOME directory cache for these things? A new clone should actually know nothing about the P4 users and fetch new ones.	09:17
	→ andyhuzhill joined	09:17
	← andyhuzhill left	09:21
	→ andyhuzhill joined	09:21
	→ plasticle joined	09:24
	← andyhuzhill left	09:31
	→ oojacoboo joined	09:31
	→ andyhuzhill joined	09:32
	← fairuz left	09:33
	← andyhuzhill left	09:35
	→ Envil joined	09:36
	→ redrambles joined	09:39
	← crimastergogo left	09:40
	→ crimastergogo joined	09:40
	← hussam left	09:40
	← jmd left	09:41
	← netj left	09:43
	→ Sasazuka joined	09:44
	← redrambles left	09:45
	→ blackntan joined	09:47
	← Sasazuka left	09:49
	→ ziyourenxiang joined	09:51
	← crimastergogo left	09:55
	← ngui left	09:55
	← stitched_unicorn left	09:55
	← GreenJello left	09:55
	→ ngui joined	09:56
	→ crimastergogo joined	09:56
	→ mobile_c_ joined	09:56
	← mobile_c left	09:58
	← wxg left	10:00
	→ planetcall\|work joined	10:00
	← Mo left	10:02
	→ Mo_ joined	10:02
	→ Mo joined	10:02
	→ a-chymera joined	10:02
	→ Zexaron joined	10:05
	→ enn_ joined	10:06
	→ jstein_ joined	10:11
	jstein_ → jstein	10:11
	← Kaisyu left	10:11
	→ elichai2 joined	10:13
	← oojacoboo left	10:13
	→ acidjnk joined	10:14
acidjnk	hello	10:14
	I have a modified file that got "stuck" somehow. It is not staged, and "add" does not stage it. What can I do?	10:14
	← ngui left	10:15
	→ stitched_unicorn joined	10:15
	→ ngui joined	10:15
	← crimastergogo left	10:16
	→ afuentes joined	10:16
	→ crimastergogo joined	10:16
	← catsup left	10:17
	→ catsup joined	10:17
Mo	grawity: sry, was disconnected. Is there any caching in git when creating a new clone?	10:18
	← catsup left	10:18
	→ oojacoboo joined	10:19
	→ catsup joined	10:19
_ikke_	acidjnk: what does git status -s return?	10:19
	→ Sasazuka joined	10:20
acidjnk	It shows it as "M"	10:21
	← crimastergogo left	10:21
acidjnk	Another strange thing: I pushed this change already yesterday, to the same branch I'm at now.	10:22
	→ crimastergogo joined	10:22
_ikke_	and git ls-files --eol <filename>?	10:22
	→ stub joined	10:22
acidjnk	i/lf w/crlf attr/	10:23
	I think I fixed it just by fiddling around.	10:24
	← stitched_unicorn left	10:24
acidjnk	first, I reverted the change, so nothing to stage. Then again git checkout (same branch I'm already at). Then it showed as changed again, but as reverse of what it was before. Then reverted again. Now nothing to commit, which is good.	10:25
	Still no idea what happened :-)	10:25
	← Sasazuka left	10:25
acidjnk	thanks	10:25
stub	Is there standard URL format for referencing a particular branch in a repository? I've seen proto://host/path/to/repo:branch , but I don't know if there is consensus or even if it is valid (suspect : is a valid character in the repo path)	10:26
_ikke_	no, there is ot	10:26
	not*	10:26
stub	ta. I'll make one up if I can't get this field split in two.	10:27
	← keepLearning512 left	10:30
	→ clime joined	10:33
clime	hello, is it possible to lock git repo against writes?	10:33
canton7	clime, so no other git repos can make changes to the file(s) that you lock?	10:33
	or so nobody can make changes in that specific git repo?	10:34
clime	i.e. prevent creating new commits, creating/deleting tags, ...	10:34
_ikke_	clime: filesystem permissions	10:34
clime	canton7: i want to prevent anyone to change the state while my script is running	10:34
	← armyriad left	10:35
clime	i need kind of a database transaction but git probably doesn't have it	10:35
	→ wxg joined	10:36
_ikke_	clime: git does have atomic ref updates	10:36
clime	_ikke_: i am invoking git client several times from a script and i would the state to stay the same between the calls	10:36
stub	I think the best you can do is force your scripts' updates to overwrite any other changes.	10:37
_ikke_	clime: right, that's not possible	10:37
clime	_ikke_: ok	10:37
	→ keepLearning512 joined	10:37
_ikke_	clime: You could clone the repo, then you get that	10:37
clime	ye i have it cloned	10:37
	even if cloned it is still accessible for writes by other processes if they have permissions	10:38
	i would kind of like to have "consistent reads"	10:39
	→ qswz joined	10:39
clime	but it is not like i would expect somebody writing to the repo while my script is running	10:40
	(not like web server running multiple processes upon database)	10:40
	← ngui left	10:40
qswz	What can I do? I want to strangle my coworkers because they always merge master into their fucking feature branches, instead of rebase or pull -r	10:40
	→ ngui joined	10:41
clime	i know there is .git/HEAD.lock but i haven't found so far something similar for tags	10:41
	← crimastergogo left	10:42
qswz	I asked why, once, explaining that rebase is less work, cleaner, less duplication	10:42
	→ crimastergogo joined	10:42
qswz	they replied that in their previous experiences, they always agreed on merging instead of rebasing	10:43
	like when you learn to do something wrong, it makes sense to continue, right?	10:43
	→ mra90 joined	10:43
	← ngui left	10:45
	→ ngui joined	10:45
qswz	I leanred to walk backwards, so now I keep doing it	10:46
	what's the problem, it works	10:46
	← keepLearning512 left	10:47
	← crimastergogo left	10:48
	→ yyy joined	10:48
	← kjartan left	10:49
	← n3wborn left	10:49
	← wxg left	10:49
	→ armyriad joined	10:49
	← ngui left	10:50
	→ ngui joined	10:50
	→ kjartan joined	10:52
	→ cdown joined	10:53
	→ manuelschneid3r joined	10:53
	→ finalbeta joined	10:54
	← ngui left	10:55
	→ ngui joined	10:55
	→ rafasc joined	10:56
	← armyriad left	10:58
	← cdown left	11:00
	← rafasc left	11:01
	← DrillTherapy left	11:02
	→ igemnace joined	11:05
	→ yn joined	11:07
	→ tgunr joined	11:09
	→ jungsubk joined	11:11
	← gxt left	11:28
	← jungsubk left	11:31
	← ngui left	11:35
	→ ngui joined	11:36
	← dermoth left	11:39
	← acidjnk left	11:39
	→ dermoth joined	11:40
	← ngui left	11:40
	→ ngui joined	11:40
	→ redrambles joined	11:41
	← cbrt64 left	11:44
	→ dockyyy joined	11:45
	← redrambles left	11:46
	→ cbrt64 joined	11:47
	→ aiena joined	11:49
	→ netj joined	11:50
aiena	with git shallow clones can I pull to older commits and have git repo update the code to that state?	11:50
	or do shallow clones only allow you to work with the current and future commits	11:50
	→ jungsubk joined	11:51
haasn	Is there a good, simple tool for managing a directory of git repoositories? Basic operations should be "check for updates", "update all", "freeze / thaw specific repo"	11:56
	← dockyyy left	11:56
haasn	it should require no state or configuration other than simply being inside the correct working dir	11:56
	(for freezing repos, you could do that inside a .dotfile of the parent dir)	11:56
	it should correctly handle conflicts due to unclean working directory, needing merge commits, etc.	11:57
	and either prompt you to merge interactively or refuse	11:57
	if the git repository is in a detached HEAD state, it would tell you that updates to `master` are available but warn you about the repository being detached	11:58
	→ enn__ joined	11:58
haasn	I could write a tool like this but I'd rather avoid reimplementing work, and I know it already exists in some domain-specific contexts (e.g. portage essentially does this for git-based overlays)	11:58
	seems like http://myrepos.branchable.com/ comes up	11:59
	← enn_ left	11:59
	enn__ → enn_	11:59
haasn	but this already violates my requirements slightly in that it requires configuration (I need to `mr register`)	12:00
bremner	how is your magic dotfile not configuration?	12:00
haasn	bremner: it's state, not configuration; and only needed to store the state of which repository is frozen	12:00
	it wouldn't need to exist otherwise	12:01
	i.e. no "setup" required	12:01
	but I'm happy to drop this requirement if `mr` fulfils all of my other requirements, I suppose; since initial setup only happens once anyway	12:01
	and you can just for d in *; mr register $d	12:01
	← craigo left	12:02
	← yyy left	12:02
	← Noti left	12:03
	→ yyy joined	12:03
	← ngui left	12:05
	→ ngui joined	12:06
haasn	seems to do what I want, good enough	12:06
	← enn_ left	12:06
	→ gxt joined	12:08
	→ mowcat joined	12:09
	← ngui left	12:10
	→ ngui joined	12:10
vlt	Hello. If someone installs Git on a Windows machine, do they have automatically SSH access (and can generate an ssh pubkey, for example) or are there additional steps necessary?	12:12
	← tchan left	12:14
	→ CodeSlingerPaul joined	12:14
	→ redrambles joined	12:15
	← dermoth left	12:17
	→ acidjnk joined	12:18
acidjnk	hello	12:18
	No, my problem from earlier is resurfacing occasionally. It's a "crazy" file. Sometimes shows as modified, rarely as staged, often showing as diff a single line which was committed yesterday already. When I revert it, that seems to solve the problem, but then the "removal of that line" suddenly is the diff, and any commit would destroy the line, which I don't want.	12:20
	Last time it came back was when I did checkout -b myNewBranch	12:20
	← redrambles left	12:21
	→ D630 joined	12:22
	← jungsubk left	12:23
	← KeyJoo left	12:24
	← ngui left	12:25
	→ ngui joined	12:25
	← R2robot left	12:26
clime	_ikke_: what do you think about using inotify to detect changes?	12:37
	_ikke_: i cannot lock ok but i can at least watch for modify changes and bail out if there is any	12:38
	→ lucasb joined	12:39
	→ redrambles joined	12:40
	← ngui left	12:40
	→ keepLearning512 joined	12:40
	→ ngui joined	12:40
	→ SerialSmile joined	12:40
	→ Makaveli7 joined	12:40
	← yyy left	12:41
	→ malmalmal joined	12:42
	→ GreenJello joined	12:42
	← redrambles left	12:45
	← esrse left	12:45
	← ngui left	12:45
	→ ngui joined	12:46
	→ comptroller joined	12:46
	→ redrambles joined	12:46
	← GreenJello left	12:47
	→ wxg joined	12:47
clime	that actually seems to be an ok approach here	12:48
	← D630 left	12:49
clime	e.g. invoking inotifywait -e modify -r git will give me a modify watch for all files in "git" repo. I should be able to use this to assure that there was no change during my script processing.	12:49
	i don't have anything better than this anyway	12:49
	this is also quite fast	12:49
	← ngui left	12:50
	→ ngui joined	12:50
	→ donofrio joined	12:53
	← stub left	12:53
	← ngui left	12:55
	→ ngui joined	12:56
	→ irrgit joined	12:58
	→ rafasc joined	12:58
	← rafasc left	12:58
	← malmalmal left	12:58
	← ngui left	13:00
	→ ngui joined	13:00
	← wxg left	13:02
	← aiena left	13:03
	← duderonomy left	13:04
	→ rafasc joined	13:04
	← ngui left	13:05
	→ ngui joined	13:06
	← nowhere_man left	13:08
	← ngui left	13:10
	→ ngui joined	13:10
	→ malmalmal joined	13:17
	im0nde → donlaton	13:21
	donlaton → im0nde	13:23
	→ wxg joined	13:24
	← redrambles left	13:24
	← ngui left	13:25
	→ kapil____ joined	13:25
	→ ngui joined	13:26
	→ howdoi joined	13:29
	← ngui left	13:30
	→ ngui joined	13:30
	→ yyy joined	13:30
	→ al-damiri joined	13:32
manuelschneid3r	is git able to version file system trees without data?	13:32
	← rafasc left	13:33
	→ xerxas joined	13:33
	→ troulouliou_div2 joined	13:33
manuelschneid3r	I want to track the dir structure of my personal data to monitor changes	13:33
_ikke_	manuelschneid3r: afaik, no	13:34
	it requires blob hashes	13:34
	← Makaveli7 left	13:34
	→ rafasc joined	13:35
	← ngui left	13:35
	→ ngui joined	13:36
	→ gnanesh joined	13:36
grawity	track the output of `find ~ -ls` if that's what you want	13:38
	← ngui left	13:40
	→ ngui joined	13:40
	→ R2robot joined	13:40
	→ Makaveli7 joined	13:40
	→ Diggsey joined	13:41
	← gnanesh left	13:42
	→ PlasmaHH joined	13:42
	→ PEG joined	13:43
	→ tchan joined	13:47
	← troulouliou_div2 left	13:47
	← pks left	13:48
	→ redrambles joined	13:51
	→ flashmozzg joined	13:54
flashmozzg	Hello. Is there some way in Git, or perhaps in services such as GitLab, to make history rewrite smooth? I.e. to somehow notify all concerned parties and reduce the friction?	13:55
	← ngui left	13:55
	← redrambles left	13:56
	→ ngui joined	13:56
dawik	git is extendable to probably do what you want.. but innately I don't think so	13:56
	→ ableto joined	13:57
flashmozzg	To be specific, my situation is: I have a repo with a very limited number of developers, that needs to have it's history rewritten. It's not the problem for us, however, there are more people using it as a read-only. I wonder if there some built-in way to communicate to them the steps they need to do to reduce problems.	13:57
dawik	just tell them to pull with rebase	13:58
PlasmaHH	Hi, I am trying to get an overview of a repo with gitk and came across the case that for a commit gitk shows multiple branches (lower left pane, git diff-tree I presume) and I am a bit puzzled as to what that really means to have so many branches there, can someone enlighten me on what exactly those entries should tell me?	13:58
_ikke_	flashmozzg: no, there is nothing built-in in git to communicate these kinds of things	13:58
flashmozzg	yeah, that's the problem: I don't know/can't efficiently communicate with the all potential users.	13:58
	← Makaveli7 left	13:59
PlasmaHH	flashmozzg: I don't quite understand what you really want. have a way to email everyone who has (anonymously?) cloned that repo?	14:00
flashmozzg	I was hoping there was some way to reduce the friction, since simple pull could break already working things for them. it'd be perfect if there was a way to temporarily transfer them to the "old" branch or to tell the git not to try and merge diverging commits, but suggest pulling with the --rebase instead.	14:00
	→ Essadon joined	14:01
flashmozzg	I.e. to make git aware that there was an intentional history rewrite.	14:02
	→ redrambles joined	14:03
_ikke_	flashmozzg: agree that would be nice, but there is currently nothing for that	14:03
benharri	aiui you'd want to email/notify the other developers, and maybe put a notice in the README.md if you've got it hosted somewhere	14:04
	→ troulouliou_div2 joined	14:04
	← ngui left	14:05
	→ dermoth joined	14:05
	→ ngui joined	14:05
flashmozzg	benharri: yeah, the devs are not the problem. They are aware. The problem is undetermined amount of users that only pull the latest version of the repo and do nothing else. Some of them might have some automation scripts, so I'd really want to avoid any unexpected breakages.	14:07
	→ dslegends joined	14:07
benharri	i'm not aware of any way to do that apart from a note on the README or something in your release notes	14:08
	← ngui left	14:10
	→ ngui joined	14:11
	→ kerframil joined	14:11
	← troulouliou_div2 left	14:11
	→ [twisti] joined	14:12
flashmozzg	ok, thanks anyway	14:13
[twisti]	how can i tell git to not try to merge but instead overwrite with my stash ? scenario: i stashed some work to file A, then pulled and got a (totally changed due to autoformat) version of file A. i confirmed with the dev who checked it in that it was an accident and that there is nothing new in there. so i would like to OVERWRITE the recently pulled file A with the version i put in my stash. but git attempts	14:14
	to merge, which is a clusterfuck due to the auto formatting. i know i could reset to the HEAD version, but how do i reset to the stash version isntead ?	14:14
	← veegee left	14:15
flashmozzg	you can use stash{N} or something like that, IRC	14:15
	← ngui left	14:15
flashmozzg	instead of HEAD	14:15
[twisti]	oh, i can treat stash like just another remote ?	14:15
	→ ngui joined	14:15
[twisti]	so then something like git checkout stash -- filename ?	14:16
parsnip	maybe revert the autoformat first?	14:17
	or revert, pop, squash?	14:17
flashmozzg	stash@{N}	14:18
dawik	flashmozzg: what about stop rewriting history and stick to merges?	14:18
e	:\|	14:18
flashmozzg	dawik: it had to be done. The alternatives were worse.	14:19
	→ goodafternoon joined	14:19
[twisti]	thanks	14:21
	→ Matze202 joined	14:22
	← shicks2525 left	14:23
	← rafasc left	14:24
	← ski7777 left	14:24
	→ rafasc joined	14:24
	→ lacrymology joined	14:28
	← ngui left	14:35
	→ ngui joined	14:36
Pistos	atrus: That's an interesting proposal. However, I worry that there may be overwrites or conflicts. e.g. both A and B have a package.json that is mostly unrelated.	14:36
	← andrzejv left	14:37
	← rafasc left	14:39
	→ smrtz joined	14:39
	← flashmozzg left	14:40
	← ngui left	14:40
	→ ngui joined	14:40
	→ GinoMan joined	14:41
	← qswz left	14:41
	→ publio joined	14:45
	→ ohcibi joined	14:45
	← malmalmal left	14:48
	→ czart joined	14:49
atrus	Pistos: well, if you've moved everything out of the way before merging them, there shouldn't be any conflicts.	14:50
	← czart left	14:50
	← goodafternoon left	14:51
	← tomfarr left	14:53
Pistos	atrus: Okay, seems to be a good solution, though I was forced to use --allow-unrelated-histories . Thank you for your suggestion.	14:53
atrus	Pistos: glad it worked. i forgot about that flag :)	14:54
	→ Siecje joined	14:55
Pistos	atrus: I'll mention that I read about git-subtree in other SOF answers, and that was VERY close to what I wanted, but `git log -p ...` did not track across the "border" between the histories/repos. In contrast, your solution does, which is nice.	14:55
Siecje	I have a branch that is in a detached HEAD. I created a branch and made a commit. Then I created another branch and made a commit. Why is it in a detached HEAD state? This means the garbage collector will remove this branch right?	14:56
jast	hard to tell why it happened with no extra information	14:57
	!repro	14:57
gitinfo	[!transcript] Please paste (using https://gist.github.com/ or similar) a transcript ( https://git.io/viMGr ) of your terminal session so we can see exactly what you see	14:57
	→ boombatower joined	14:57
	→ JordanThedev joined	14:58
Siecje	jast: I created these branches a few days ago.	14:58
	Each head has a branch though.	14:58
	→ maryo joined	14:59
Siecje	These are only local branches.	15:00
	← kjartan left	15:00
	→ xypnox joined	15:00
oxymoron93	!detached	15:00
gitinfo	A detached HEAD (aka "no branch") occurs when your HEAD does not point at a branch. New commits will NOT be added to any branch, and can easily be !lost. This can happen if you a) check out a tag, remote tracking branch, or SHA; or b) if you are in a submodule; or you are in the middle of a c) am or d) rebase that is stuck/conflicted. See !reattach	15:00
	→ tombusby joined	15:00
	→ kjartan joined	15:02
	← mra90 left	15:03
	← ngui left	15:05
	→ ngui joined	15:06
	← mzf left	15:06
	→ bdonnahue joined	15:09
	← xypnox left	15:10
	→ crimastergogo joined	15:10
	← sQVe left	15:10
	← Hello71 left	15:12
Pistos	atrus: I posted your solution, and gave you credit. Thanks again. https://stackoverflow.com/a/54131590/28558	15:12
	← crimastergogo left	15:13
	→ crimastergogo joined	15:13
	→ Hello71 joined	15:15
	← ngui left	15:15
	→ ngui joined	15:15
	← oxymoron93 left	15:16
	→ cdown joined	15:17
	← d^sh_ left	15:18
	→ greggerz joined	15:18
	← lucasb left	15:19
	← ngui left	15:20
	→ ngui joined	15:21
	← carb0n_ left	15:21
	→ czart joined	15:22
	← crimastergogo left	15:23
	→ pks joined	15:23
	→ crimastergogo joined	15:24
	← ngui left	15:25
	→ ngui joined	15:25
	← jason69 left	15:29
	← clime left	15:30
	→ shicks2525 joined	15:32
	→ d^sh joined	15:33
	← cdown left	15:33
	→ TJ- joined	15:36
Siecje	Well I got to the commit by git checkout branch_name. Maybe I'm not in detached HEAD state? If I do git status I see HEAD detached at ...	15:37
	→ st-gourichon-f joined	15:38
	→ Tibian joined	15:38
	← floppydh left	15:40
	← ngui left	15:40
	→ ngui joined	15:41
	← yyy left	15:43
	→ Makaveli7 joined	15:43
	← ngui left	15:45
	Tibian → ad0ze	15:45
	→ ngui joined	15:45
	→ AbleBacon joined	15:47
	→ d^sh_ joined	15:47
	→ yyy joined	15:48
	← ngui left	15:50
	→ nshp joined	15:50
	→ ngui joined	15:51
	→ spongie joined	15:51
	→ durham joined	15:51
	← d^sh left	15:51
	← durham left	15:51
spongie	I'm trying to sync a fork, added the repos as upstream, fetched upstream, but im not getting the branches.	15:52
	branch -r shows a lot of remote branches im not getting locally. i dont understand exactly how this is supposed to work, any hints?	15:53
canton7	!local-branch-from-remote	15:53
	→ Wetmelon joined	15:53
canton7	!local_branch_from_remote	15:54
gitinfo	The following commands are all equivalent, assuming <branch> doesn't yet exist: 'git checkout -b <branch> <remote>/<branch>', 'git checkout -t <remote>/<branch>', 'git checkout <branch>'. The latter invokes some magic.	15:54
canton7	https://git-scm.com/book/en/v2/Git-Branching-Remote-Branches	15:54
	spongie, ^	15:54
	→ thiago joined	15:57
	→ Dougie187 joined	15:58
	← crimastergogo left	15:59
	→ crimastergogo joined	15:59
	→ cdown joined	16:00
	← ngui left	16:00
	→ ngui joined	16:00
spongie	canton7: thanks!	16:01
	← spongie left	16:01
	→ st-gourichon-fid joined	16:01
	→ fission6 joined	16:02
	← st-gourichon-f left	16:02
	← ad0ze left	16:03
	← nicoulaj left	16:03
	→ nicoulaj joined	16:08
	→ GinoMan2440 joined	16:11
	→ psprint joined	16:12
psprint	Huh, just reading apply/am/format-patch docs for the first time, great tools	16:12
	→ ozcanesen joined	16:14
	← GinoMan left	16:14
fission6	whats the best way to approach this, two months ago there was a commit which deleted about 8 files. I want to restore the files in a specific new folder, I have the commit, how can i pop these files into the new folder, they are all in one folder so i need to basically grap the folder and restore into a new folder	16:17
canton7	fission6, 'git show <commit>^:path/to/file1 > path/to/new/file1', will do it file-by-file	16:18
	I'm not aware of a way to do all the files in one go which restores into a different folder	16:18
	'git checkout <commit>^ -- path/to/folder' will restore the whole folder into the same place	16:18
	← rnmhdn left	16:19
	← acidjnk left	16:19
fission6	let me just do that last one and i can maually clean up once "out of git"	16:19
	why do i need commit ^	16:20
	why the ^	16:20
canton7	it refers to the first parent of the specified commit. Since the commit removed the files, the commit just before will have the files in	16:21
	→ GreenJello joined	16:21
	→ Inline joined	16:22
	← logithack left	16:22
	← nixdork left	16:22
fission6	i see	16:22
im0nde	Hi, can anybody tell me how to do a post-receive git-hook in gitea that notifies wathcers about a new push?	16:22
fission6	let me try thanks canton7	16:22
canton7	im0nde, one of the built-in examples sends an email iirc?	16:22
	→ ColdKeyboard joined	16:22
im0nde	canton7: Build in examples?	16:23
	→ AtumT joined	16:23
canton7	im0nde, look in .git/hooks, there are hooks ending in '.sample'	16:23
	→ timothywcrane joined	16:23
	→ dreiss joined	16:24
canton7	im0nde, ah, not quite. From man githooks: "The default post-receive hook is empty, but there is a sample script post-receive-email provided in the contrib/hooks directory in Git distribution, which implements sending commit emails.	16:24
gitinfo	im0nde: the githooks manpage is available at https://gitirc.eu/githooks.html	16:24
canton7	im0nde, here: https://github.com/git/git/blob/master/contrib/hooks/post-receive-email	16:24
	ah, looks like this replaces it: https://github.com/git/git/tree/master/contrib/hooks/multimail	16:25
	→ mat001 joined	16:25
	← d^sh_ left	16:25
	← GreenJello left	16:25
	→ cswl joined	16:28
	→ Nexilva joined	16:28
	← chele left	16:29
Nexilva	Hello, it is possible when you do 'git status' that it doesn't tell you about untracked files? I have setup git in my ~ home directory, and I've added a few config .dotfiles, and I when I do git status, it showed me all the other files in my home directory which are not tracked.	16:29
	Thank you!	16:29
_ikke_	Nexilva: add * to .gitignore	16:29
	← mat001 left	16:29
	← crimastergogo left	16:30
Nexilva	https://stackoverflow.com/questions/11542687/git-how-to-ignore-all-present-untracked-files is this an applicable workaround?	16:31
	To use git status -uno intead?	16:31
_ikke_	\sure	16:31
Nexilva	Ok, now I know two. Thank you. I did the ignore bit.	16:32
	← thebope left	16:33
	← ngui left	16:35
	→ ngui joined	16:36
	← nuc left	16:37
	→ nuc joined	16:39
	← ziyourenxiang left	16:39
	← ngui left	16:40
	→ ngui joined	16:40
	→ CheckDavid joined	16:40
	→ mat001 joined	16:41
	← ngui left	16:45
	→ ngui joined	16:46
	→ orbyt_ joined	16:46
	→ GinoMan joined	16:46
	← howdoi left	16:49
	→ \void joined	16:49
	→ sanscoeur joined	16:49
	← GinoMan2440 left	16:50
	← ngui left	16:50
	→ ngui joined	16:50
	→ nixdork joined	16:54
	→ GreenJello joined	16:55
	← ngui left	16:55
	← PlasmaHH left	16:55
	← SerialSmile left	16:55
	→ ngui joined	16:56
	→ logithack joined	16:56
	→ crimastergogo joined	16:58
	← GreenJello left	16:59
	← ngui left	17:00
	→ ngui joined	17:00
	→ flashmozzg joined	17:00
flashmozzg	is there a flag/option to add a remote if it's not already added?	17:01
	So it doesn't complain on successive applications?	17:01
	← dreiss left	17:01
	← Hathadar left	17:01
	CheckDavid → CryptoDavid	17:01
	→ shentino joined	17:03
	← mat001 left	17:05
osse	flashmozzg: Sounds harmless. You can probably just silence it.	17:05
	Alternatively git config --get remotes.foo \|\| git remote add foo... Etc	17:06
	--get remote.foo.url	17:07
flashmozzg	osse: ok, I just don't want some real error to slip through	17:08
	→ khisanth__ joined	17:09
	→ d^sh joined	17:09
	← shentino left	17:10
	→ kaeptmblaubaer joined	17:10
	→ nowhere_man joined	17:10
	← ngui left	17:10
	→ shentino joined	17:10
	→ ngui joined	17:11
	→ jhulten joined	17:12
	→ thebope joined	17:13
	← khisanth_ left	17:13
	→ dreiss joined	17:16
	→ mat001 joined	17:17
	← orbisvicis left	17:17
	← keepLearning512 left	17:18
	← ngui left	17:20
	→ ngui joined	17:20
	→ sanscoeu_ joined	17:23
	← mona_x left	17:24
	← dreiss left	17:24
	→ leeN joined	17:25
	← sanscoeu_ left	17:25
	← kapil____ left	17:25
	→ keepLearning512 joined	17:25
	← ngui left	17:25
	→ ngui joined	17:26
	→ jakogut joined	17:26
	← sanscoeur left	17:27
	→ Revan- joined	17:27
	→ GreenJello joined	17:27
	→ bdonnahue1 joined	17:31
	← bdonnahue left	17:31
	← flashmozzg left	17:33
	← GreenJello left	17:33
	→ dviola joined	17:34
	→ durham joined	17:35
	← durham left	17:36
	→ sanscoeur joined	17:38
	← crimastergogo left	17:38
	→ crimastergogo joined	17:38
	← ngui left	17:40
	→ ngui joined	17:40
	→ pwnz0r joined	17:41
	← pwnz0r left	17:43
	← crimastergogo left	17:43
	→ crimastergogo joined	17:43
	← orbyt_ left	17:45
	← sanscoeur left	17:45
	← mikecmpbll left	17:46
	← kerframil left	17:48
	→ kerframil joined	17:49
rullie	what would cause git merge-base --fork-point A B to return nothing when git log clearly shows A and B fokred at C?	17:50
	→ gruetzkopf joined	17:51
	← zulutango left	17:51
	← smrtz left	17:52
	→ zulutango joined	17:52
	← ngui left	17:55
	→ ngui joined	17:56
	← mat001 left	17:57
	→ mat001 joined	17:57
	→ mikecmpbll joined	17:58
	← crimastergogo left	18:01
	→ mat001_ joined	18:01
	← mat001 left	18:02
	khisanth__ → Khisanth	18:02
	← kaeptmblaubaer left	18:02
	← bdonnahue1 left	18:05
	← ngui left	18:05
	→ ngui joined	18:05
	← keepLearning512 left	18:06
	→ wootehfoot joined	18:06
	← fission6 left	18:06
	→ bdonnahue joined	18:06
	← lacrymology left	18:07
	← jhulten left	18:11
	→ mitchellsalad__ joined	18:12
	→ jhulten joined	18:12
	→ orbyt_ joined	18:15
	→ kroon joined	18:16
	→ jstimm joined	18:19
	→ bdonnahue1 joined	18:22
	← bdonnahue left	18:22
	← gkatsev left	18:23
	← paraxial left	18:23
kroon	Does Paul Mackerras hang around in here ? Is there a special ml for sending gitk patches ?	18:23
	← manuelschneid3r left	18:24
	← ngui left	18:25
	→ ngui joined	18:26
	← blackntan left	18:27
	← cdown left	18:30
	← ngui left	18:30
	→ ngui joined	18:30
	→ cdown joined	18:30
	→ kezimo joined	18:31
	← yyy left	18:34
	→ Tibian_ joined	18:36
	← orbyt_ left	18:37
	→ causasui joined	18:38
	→ d^sh_ joined	18:38
causasui	I can set an alias to '!$@' to execute a command at the root of the repo. but it looks like that's in a subshell that doesn't inherit aliases and env vars and so on. is there any way around that?	18:38
	← kezimo left	18:38
causasui	do I have to do something like '!bash --login -c "$@"'	18:39
	(I did try that, it didn't behave differently)	18:39
	← T_UNIX left	18:40
	Tibian_ → Tibian	18:41
	← d^sh left	18:41
	Tibian → ad0ze	18:41
	→ Mattiaslndstrm joined	18:42
nedbat	causasui: can you show the alias?	18:42
	← mooasaurus left	18:42
	→ mooasaurus joined	18:42
causasui	nedbat: currently it's `git config --global alias.exec '!$@'` but I've tried a couple versions	18:42
nedbat	causasui: I don't understand what !$@ means? Whose syntax is that?	18:43
causasui	nedbat: mine, I'm just messing around. I saw that '!' executes what follows at the root dir of the repo you're in. I use that in a few other aliases (which work fine), but I had the idea to turn it into something more generic	18:45
RichiH	59	18:50
	sorry	18:50
	→ wildlander joined	18:50
nedbat	causasui: it's not your syntax, you are expecting git to do things based on ! and $@. The git docs say that ! means the rest is a shell command, rather than a git command, and they will execute at the root of the repo. What did you expect $@ would do?	18:53
	→ orbyt_ joined	18:53
	→ virmaha joined	18:54
	← qqx left	18:55
	→ qqx joined	18:55
	→ mat001 joined	18:55
	← planetcall\|work left	18:55
	→ sanscoeur joined	18:55
virmaha	Hello. I've a git repo and say i always work on branch proj_64. i then created another branch from it, called proj_32. Whenever, I make a new commit to proj_64, I want same to be present on proj_32 branch	18:56
	how can i do that? right now, i am manually cherry-picking the commit	18:56
causasui	nedbat: Exactly what you just said...	18:58
nedbat	causasui: I didn't say what $@ would do?	18:58
	virmaha: how do you want the branches to differ?	18:58
	← mat001_ left	18:59
causasui	nedbat: I expect that $@ would interpolate whatever arguments came after the alias. Which is what happens. But it looks like it's executed in a subshell (so my shell aliases don't work etc	18:59
	→ jrnieder joined	18:59
	← d^sh_ left	18:59
nedbat	causasui: right, it's definitely executed in a new shell	19:00
	← Dougie187 left	19:00
nedbat	causasui: sorry, i don't know how to make it more seamless	19:00
causasui	np, thanks anyhow	19:00
virmaha	nedbat: only in their name	19:02
nedbat	virmaha: hmm, why? What's the point of have two branches that are always identical?	19:02
	← orbyt_ left	19:03
virmaha	nedbat: internal infrastructure restrictions	19:04
	← timothywcrane left	19:04
	→ Dirak joined	19:04
	← Nexilva left	19:05
	→ Narrat joined	19:05
	→ GreenJello joined	19:06
	→ manuelschneid3r joined	19:07
nedbat	virmaha: no need to cherry-pick, you can just make them refer to the same commit.	19:10
virmaha	nedbat: how?	19:10
nedbat	virmaha: git checkout proj_32; git reset --hard proj_64	19:10
	→ orbyt_ joined	19:10
virmaha	nedbat: btw proj_32 has 3 submodules which further have 5 submodules each. Still, it should be fine, right	19:11
	→ yyy joined	19:11
virmaha	proj_64*	19:11
nedbat	virmaha: i have never used submodules	19:11
virmaha	okay	19:11
	→ blackntan joined	19:12
wadadli	is there a quick cmd to find the banch/commit id that is the parent of my current branch?	19:13
	← jstimm left	19:13
nedbat	wadadli: what do you mean by parent?	19:17
	→ snowgoggles joined	19:20
	→ throwceecee joined	19:20
	→ WolfOrwell joined	19:21
	← Mattiaslndstrm left	19:23
	→ Mattiaslndstrm joined	19:25
	→ pwnz0r joined	19:26
	← profmac left	19:27
	← orbyt_ left	19:27
	← pwnz0r left	19:28
	← elichai2 left	19:30
	← czart left	19:31
	→ MACscr joined	19:31
	← hofmann3900 left	19:31
	→ aleph- joined	19:31
	→ czart joined	19:33
	→ carb0n_ joined	19:37
	→ maryo_ joined	19:38
	→ profmac joined	19:40
	→ lacrymology joined	19:40
	→ PocketKiller joined	19:40
	← maryo left	19:42
	→ Sasazuka joined	19:43
	← tomaw left	19:43
	→ tomaw joined	19:44
	← ngui left	19:45
	← throwceecee left	19:46
	→ d^sh joined	19:48
	→ ngui joined	19:50
	← leeN left	19:50
j416	webstrand: git log -1 --format=fuller	19:51
	will show all parent hashes of a commit, iirc	19:51
	mm nope, I am lying, sorry	19:52
	webstrand: git log -1 --format=raw	19:53
	← carb0n_ left	19:55
	→ fission6 joined	19:55
	← mat001 left	19:57
	← greggerz left	19:57
	← jas4711 left	19:58
	→ mat001 joined	20:01
	→ Phylock joined	20:03
	→ GinoMan2440 joined	20:04
	← gxt left	20:06
	← mat001 left	20:06
	→ n3wborn joined	20:06
	← Envil left	20:07
	← GinoMan left	20:07
	→ jnewt joined	20:07
	← czart left	20:09
	→ duderonomy joined	20:11
	→ carb0n_ joined	20:12
	← al-damiri left	20:12
	← im0nde left	20:13
	← ableto left	20:13
	← carb0n_ left	20:19
	← snowgoggles left	20:19
	mrCyborg\|zzz → mrCyborg	20:19
	← LionsMane left	20:20
webstrand	j416: I think you meant wadadli	20:26
j416	ah, indeed. apologies, webstrand.	20:27
	→ im0nde joined	20:27
	→ snowgoggles joined	20:29
lacrymology	can someone help me re-apply a stash that was already dropped?	20:33
	ah,found it	20:34
	← ngui left	20:35
	→ ngui joined	20:35
	→ gxt joined	20:36
	→ scarcry_ joined	20:38
	→ mat001 joined	20:38
	← ngui left	20:40
	→ ngui joined	20:40
j416	how'd you find it?	20:43
	scarcry_ → scarcy	20:43
	← scarcy left	20:43
	→ blackntan-airboo joined	20:44
	← Matze202 left	20:44
	→ bn_work joined	20:44
tango_	probably rolled under the sofa	20:45
j416	nod	20:45
	← ngui left	20:45
	→ scarcry_ joined	20:45
	→ Fernando-Basso joined	20:46
	→ troulouliou_div2 joined	20:49
	→ ngui joined	20:50
	→ spacesuitdiver joined	20:51
	← fission6 left	20:51
	scarcry → Guest14587	20:52
	scarcry_ → scarcry	20:52
	← troulouliou_div2 left	20:55
	← cbrt64 left	20:55
	→ cbrt64 joined	20:56
	← bdonnahue1 left	20:57
	→ bdonnahue1 joined	20:57
	→ goiko joined	20:58
	→ fission6 joined	20:58
	← fission6 left	20:58
	← CryptoDavid left	21:00
goiko	hi, i am using git with some submodules, commit some changes within the submodule and want to push it to the remote.	21:01
	← Mattiaslndstrm left	21:05
	← ad0ze left	21:14
	→ fission6 joined	21:14
	← PocketKiller left	21:15
	→ unbearable joined	21:21
	← WolfOrwell left	21:23
	→ freeman42x joined	21:23
	→ al-damiri joined	21:24
causasui	goiko: yes?	21:25
	→ m0viefreak joined	21:30
	← ngui left	21:30
	← Diggsey left	21:31
	← snowgoggles left	21:32
	← iam730 left	21:35
	→ ngui joined	21:35
	→ gini joined	21:36
goiko	causasui, sry, HEAD was detached, now I get it	21:37
	← cbrt64 left	21:39
	→ cbrt64 joined	21:39
	← igemnace left	21:40
	← gini left	21:40
	→ bongozig joined	21:41
	→ greatgatsby joined	21:41
	→ shakalaka joined	21:46
	→ blacknta_ joined	21:49
	→ snowgoggles joined	21:50
	← blackntan-airboo left	21:50
	← blacknta_ left	21:50
goiko	causasui, i forget that i had to checkout the master in the submodule and also pull before commit and push	21:54
	→ orbyt_ joined	21:56
	← TJ- left	22:02
	← Siecje left	22:03
	→ sourcream joined	22:03
	→ iam730 joined	22:05
	← shabius left	22:06
	→ shabius joined	22:07
	← cswl left	22:08
	← jsatk_ left	22:08
	→ LVK267 joined	22:09
	→ stux\|RC joined	22:09
	→ rick8024 joined	22:09
	→ worstadmin joined	22:10
	→ cswl joined	22:10
	→ fowlduck joined	22:10
	→ bl0m1 joined	22:10
	← ngui left	22:10
	→ jsatk__ joined	22:10
	→ fury joined	22:10
	→ ngui joined	22:11
	→ rsaanon joined	22:11
	← KnoP left	22:11
	→ SleePy joined	22:11
	→ blackntan-airboo joined	22:11
	← rsaanon left	22:12
	→ rsaanon joined	22:13
	← ngui left	22:15
	← rick8024 left	22:15
	→ ngui joined	22:15
	← gruetzkopf left	22:17
j416	goiko: typically you would work in the original repo of the submodule and then fetch changes into the repo that has the submodule when needed	22:17
	← mat001 left	22:19
	→ KnoP joined	22:20
goiko	j416, sure that is right but i need it in my (main) project, so I can change it directly in the submodules	22:22
	← fission6 left	22:22
goiko	just to skip that step	22:23
	→ fission6 joined	22:23
j416	sometimes that is an indicator that it shouldn't be a submodule (though I don't know your project, of course)	22:25
	→ mat001 joined	22:25
	← AtumT left	22:25
	→ Mattiaslndstrm joined	22:30
goiko	good point	22:30
	→ Toadisattva joined	22:37
	← ozcanesen left	22:37
	← orbyt_ left	22:42
	→ TJ- joined	22:43
	→ roygbiv joined	22:44
	← jhulten left	22:46
	→ jhulten joined	22:47
	→ Nexilva joined	22:52
	← cswl left	22:57
	← Olipro left	22:57
	→ escherial joined	22:57
	→ Olipro_ joined	22:57
	Olipro_ → Olipro	22:57
	← sourcream left	22:58
	← blackntan-airboo left	22:58
	← jakogut left	23:02
	← brandonkal left	23:03
	→ acercle joined	23:03
	← fission6 left	23:06
	← wootehfoot left	23:08
	← n3wborn left	23:08
	← AbleBacon left	23:11
	← publio left	23:12
	→ orbyt_ joined	23:12
	← unbearable left	23:13
	← escherial left	23:14
	← Makaveli7 left	23:14
	→ courrier joined	23:14
	→ bubo joined	23:14
	→ escherial joined	23:15
	← jelmer left	23:15
	← jstein left	23:16
	→ jelmer joined	23:16
	← goiko left	23:17
	← bdonnahue1 left	23:17
	← sanscoeur left	23:18
	→ Kaisyu joined	23:20
	← oojacoboo left	23:20
	← Fernando-Basso left	23:21
	← acercle left	23:23
	→ acercle joined	23:24
	← snowgoggles left	23:25
	← roygbiv left	23:27
	← kerframil left	23:27
	← yyy left	23:27
	← mat001 left	23:30
	← TimWolla left	23:33
	→ goiko joined	23:33
	← jhulten left	23:33
	← spacesuitdiver left	23:33
	→ RoriconKnight joined	23:34
	→ mat001 joined	23:35
	→ jhulten joined	23:35
	← jhulten left	23:38
	← maryo_ left	23:39
	→ TimWolla joined	23:39
	→ jhulten joined	23:43
	→ jakogut joined	23:43
	← escherial left	23:45
	← courrier left	23:45
	→ courrier joined	23:45
	→ escherial joined	23:45
	← jelmer left	23:45
	→ jelmer joined	23:46
	← ahf left	23:47
	← orbyt_ left	23:50
	← CodeSlingerPaul left	23:51
	← greatgatsby left	23:51
	→ cd joined	23:52
	← rsaanon left	23:52
	→ orbyt_ joined	23:55
	→ spacesuitdiver joined	23:58
	→ rsaanon joined	23:58
	→ snowgoggles joined	23:59
	← spacesuitdiver left	23:59
	→ ahf joined	23:59

Logs Search ←Prev date Next date→

Channels Documentation