IRCloggy #git 2019-11-01

Provider of IRC logs since 2005.

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

	2019-11-01
	→ Graypup_ joined	00:00
	← Sasazuka__ left	00:02
	← brethil left	00:03
	← soju left	00:04
	→ brethil joined	00:06
	← Fernando-Basso left	00:08
	→ throstur joined	00:08
	← Codaraxis left	00:15
	← storrgie left	00:17
	→ perrier-jouet joined	00:17
	← durham left	00:20
	→ SwiftMatt joined	00:20
	→ soju joined	00:21
	← mat001 left	00:23
	→ LearnAllTheTime joined	00:23
	← soju left	00:26
	← Mattiaslndstrm left	00:29
	← JanC left	00:29
	→ JanC joined	00:30
	PacKetSlayer → PocketKiller	00:33
	→ rustyshackleford joined	00:36
	← mekhami left	00:37
	← rafasc left	00:42
	PocketKiller → PacKetSlayer	00:44
Atlenohen	if there could be a git and/or github option to disable comparison showing whitespace changes so I could clang-format all my project to a longer column width	00:44
	or some kind of an automatic internal conversion to the proper version on github while leaving my local tree in my mode intact, just like line endings, now that's an idea !!!	00:45
	So when pushing, clang-format all to clang-format-remote.cfg when writing to tree use clang-fromat-local.cfg settings, boom there you go	00:46
	← RoriconKnight left	00:46
	← narendraj9 left	00:49
b0nn	Atlenohen: unfortunately, though, some languages rely on the whitespace being of certain type (Python, Go to a lessor extent)\	00:50
	Atlenohen: however, this looks promising https://stackoverflow.com/questions/33159394/ignore-all-whitespace-changes-with-git-diff-between-commits/33159593	00:51
	→ dtux joined	00:53
	← Sleepy63 left	00:56
	← m0viefreak left	00:57
	← dtux left	00:58
	← nioncode left	00:59
	→ yn_ joined	00:59
	→ blackmesa joined	00:59
	← CalimeroTeknik left	00:59
	→ tjbp joined	01:00
	→ nioncode joined	01:00
	← SwiftMatt left	01:00
	→ blackmesa1 joined	01:01
	← kerframil left	01:01
	← blackmesa left	01:05
	← wymillerlinux left	01:07
	← leeN left	01:09
	← yawkat left	01:12
Atlenohen	I'll check it out thanks	01:15
	→ duderonomy joined	01:20
	→ Heirlung joined	01:23
	→ kerframil joined	01:24
	→ g00s joined	01:25
	← dartmouthed left	01:25
	→ durham joined	01:29
	← durham left	01:33
	← bvdw left	01:38
	→ bvdw joined	01:39
	→ nurdie joined	01:42
	→ Goplat joined	01:42
	← nurdie left	01:47
	→ yawkat joined	01:47
	→ bambanx joined	01:50
	← tsujp left	01:53
	← sentriz left	01:54
	→ ketas- joined	01:56
	← libertyprime left	01:57
	← ketas left	01:58
	← khisanth_ left	01:59
	→ sentriz joined	02:03
	← F0rTex left	02:07
	→ F0rTex joined	02:08
	→ libertyprime joined	02:10
	→ khisanth_ joined	02:12
	← throstur left	02:12
	→ jstimm joined	02:17
	→ tufteddeer joined	02:22
	← libertyprime left	02:31
	← g00s left	02:33
	→ dtux joined	02:33
	→ libertyprime joined	02:33
	← n3wborn left	02:34
	→ thebope joined	02:37
	→ JanC_ joined	02:38
	← JanC left	02:40
	JanC_ → JanC	02:40
	→ dmtucker joined	02:45
	← dtux left	02:46
	← lagothrix left	02:55
	→ lagothrix joined	03:00
	→ gioyik joined	03:08
	← tufteddeer left	03:09
	→ tufteddeer joined	03:10
	← gioyik left	03:10
	← Phruis left	03:10
	lagothrix → Guest86642	03:11
	← Guest86642 left	03:11
	→ lagothrix joined	03:11
	← Wetmelon left	03:15
	← tufteddeer left	03:16
	← R2robot left	03:20
	→ r2ro joined	03:21
	← stitched_unicorn left	03:24
	← Cabanossi left	03:25
	→ tufteddeer joined	03:28
	← MACscr left	03:30
	→ blackmesa joined	03:31
	→ Cabanossi joined	03:32
	← blackmesa1 left	03:33
	→ nurdie joined	03:34
	← nurdie left	03:37
	→ nurdie joined	03:38
	← thebope left	03:41
	← nurdie left	03:44
	→ nurdie joined	03:45
	→ Wetmelon joined	03:45
	→ toothe joined	03:46
	→ beastDiv_ joined	03:46
	← nurdie left	03:49
	← Atlenohen left	03:52
	← Decobus left	03:52
	← boombatower left	03:59
	← tufteddeer left	03:59
	→ tufteddeer joined	04:00
	→ g00s joined	04:01
	← blackmesa left	04:01
	Hoffmanbearpig → Hoffman	04:03
	← tufteddeer left	04:09
	→ mat001 joined	04:17
	→ tufteddeer joined	04:18
	← greatgatsby left	04:20
	→ greatgatsby joined	04:27
	→ chachasmooth joined	04:30
	→ ferdna joined	04:31
	← justanotheruser left	04:32
	→ justanotheruser joined	04:37
	r2ro → R2robot	04:38
	← fatalhalt left	04:40
	→ fatalhalt joined	04:41
	→ akemhp_ joined	04:45
	← mat001 left	04:46
	→ akemhp joined	04:47
	← akemhp_ left	04:50
	← R2robot left	04:50
	← orbyt_ left	04:51
	← Wetmelon left	04:51
	← akemhp left	04:52
	→ R2robot joined	04:52
	→ macaronus joined	04:56
	← realies left	05:01
	→ pks_ joined	05:02
	← pks left	05:02
	pks_ → pks	05:02
	← sgn left	05:03
	→ sgn joined	05:03
	→ xall joined	05:03
	→ altendky joined	05:04
	→ realies joined	05:09
	← cbreak left	05:17
	→ cbreak joined	05:19
	← sgn left	05:20
	→ sgn joined	05:20
	← jstimm left	05:21
	→ MACscr joined	05:33
	→ iam730 joined	05:52
	← g00s left	05:55
	← ferdna left	05:59
	→ cd joined	06:01
	→ g00s joined	06:03
	→ Wetmelon joined	06:07
	→ dijitol joined	06:08
	← thiago left	06:08
	← topdownjimmy left	06:11
	→ topdownjimmy joined	06:12
	← Wetmelon left	06:13
	→ dre_ joined	06:15
	← gpanders left	06:16
	→ Decobus joined	06:18
	← roadie left	06:24
	→ roadie joined	06:32
	← Decobus left	06:34
	← bambanx left	06:35
	← dijitol left	06:37
	→ Decobus joined	06:39
	← iam730 left	06:44
	→ Aelius_ joined	06:45
	← Aelius left	06:47
	→ iam730 joined	06:48
	← fxrs left	06:49
	← xall left	06:50
	→ fxrs joined	06:51
	→ akemhp joined	06:52
	← roadie left	06:54
	← dre_ left	06:58
	← Goplat left	06:59
	→ xall joined	06:59
	← Heirlung left	07:00
	→ Heirlung joined	07:05
	← xall left	07:06
	→ xall joined	07:08
	← duderonomy left	07:12
	← xall left	07:13
	→ veegee joined	07:18
	→ Noti joined	07:19
	→ xall joined	07:21
	← xall left	07:26
	← clemons left	07:29
	← bigfondue left	07:31
	→ gareppa joined	07:31
	→ clemons joined	07:32
	← bvdw left	07:38
	→ xall joined	07:38
	← xall left	07:38
	→ bvdw joined	07:39
	← g00s left	07:42
	→ Mattiaslndstrm joined	07:44
	→ aw1 joined	07:45
	← bitwinery left	07:48
	← kerframil left	07:51
	→ bigfondue joined	07:53
	← yonder left	07:54
osse	GUISE	07:57
	I learned something today that I didn't know. You can (if you're lucky) actually find out when a branch was created. reflog records it. git reflog --date	07:58
	→ f8e3 joined	07:59
f8e3	whats a god alternative to gitk?	07:59
	← clemons left	07:59
osse	tig is popular	07:59
_ikke_	+1	08:00
f8e3	i currently also try to mine data: lines per commiter, commits over timeline, i looked into gitinspector, git summary/etc and wonder, how to from python or shell extract data that i can later plot	08:00
	eg https://camo.githubusercontent.com/9eef035e55075eb1c446168de3c780696a170217/68747470733a2f2f7261772e6769746875622e636f6d2f77696b692f656a77612f676974696e73706563746f722f696d616765732f68746d6c5f6578616d706c652e6a7067 but just the data for matplotlib	08:00
	→ SteffanW joined	08:01
	→ clemons joined	08:01
	← Noti left	08:01
osse	maybe you will find this interesting: https://github.com/src-d/hercules	08:02
	→ Kartagis joined	08:02
	→ Makaveli7 joined	08:04
	→ im0nde_ joined	08:04
	→ dege joined	08:04
	← im0nde1 left	08:06
	→ TomyWork joined	08:07
f8e3	thank you	08:11
	→ blackmesa joined	08:12
	← altendky left	08:14
	→ z8z joined	08:29
	← POQDavid left	08:30
	→ throstur joined	08:31
	→ POQDavid joined	08:32
	→ greggerz joined	08:34
	→ Assault joined	08:37
Assault	hey, I added git LFS to my repo and make it track all the .png files. But after that, the .png files seem to be broken - they show in the the file explorer, but when I open them, the image viewer says "it looks like we dont support this file format"	08:38
	← crose left	08:39
Assault	heh, the content of the .png file is "version https://git-lfs.github.com/spec/v1	08:40
	oid sha256:asdsad7sdasdfasd831saddas9a9d6ghjd49987374ddd9196d076ba722c6a49e2ba67f3002e4ff	08:40
	size 1021" text	08:40
	do I have to do some additional step to get the actual .png files?	08:40
	← dmtucker left	08:42
	← greggerz left	08:42
	→ dtux joined	08:44
Assault	turns out I have to do git lfs pull	08:46
	so...do I have to do it every time a file tracked by git lfs is changed?	08:46
j416	Assault: you probably forgot to install whatever clean/smudge filters or hook it wants to add	08:48
	dunno if you need to run git-lfs pull every time	08:49
Assault	j416, what filter? I am not aware that I should add something liek that	08:49
j416	Assault: cat .git/info/attributes	08:49
_ikke_	clean/smudge filters are how LFS hooks into git	08:49
	← Joel left	08:49
j416	or are those in .git/config; I forget	08:50
_ikke_	both	08:50
Assault	attributes just contains *.png filter=lfs diff=lfs merge=lfs -text	08:50
_ikke_	the filters are defined in the config, and attributes enables them	08:50
j416	ah, right.	08:50
Assault	actually thats .gitattributes file	08:50
	→ Joel joined	08:50
Assault	i just did git lfs install on my repo	08:50
j416	Assault: I don't use lfs, tried it but it was so coupled to github and so buggy that I removed it; maybe it's better now	08:51
Assault	and then made it git lfs migrate import --include="*.png" --include-ref=refs/heads/master	08:51
j416	note that lfs is not part of git	08:51
_ikke_	I don't use it either, I have no usecase of it	08:51
	for it*	08:51
Assault	yeah i know..but Unity projects are impossible with it	08:51
	and Bitbucket has 2 GB repo size limti	08:52
	and Bitbucket has 2 GB repo size limit	08:52
_ikke_	yes, understood	08:52
	← z8z left	08:52
Assault	maybe git is just not the right tool for version control for Unity	08:52
j416	if your repo is huge, then something like lfs seems like a good idea to try	08:53
	indeed.	08:53
Assault	ok i guess I have to read some more on git lfs...	08:53
	← hqdruxn08 left	08:55
	→ Fernando-Basso joined	08:55
	← glance left	08:58
	→ hqdruxn08 joined	08:59
	← libertyprime left	09:01
	→ libertyprime joined	09:02
	→ _ADN_ joined	09:03
	← beastDiv_ left	09:03
	← hqdruxn08 left	09:03
	→ hqdruxn08 joined	09:03
	← hqdruxn08 left	09:05
	→ hqdruxn08 joined	09:05
	← LearnAllTheTime left	09:07
	← dtux left	09:09
	→ beastDiv_ joined	09:10
	→ ricks_ joined	09:10
	→ RaceCondition joined	09:11
	→ Newami joined	09:11
	← Newami left	09:12
	← hqdruxn08 left	09:12
	→ n3wborn joined	09:13
	→ akemhp_ joined	09:16
	→ Zevv joined	09:17
	→ TheFuzzball joined	09:17
	← akemhp left	09:18
Zevv	Hi #git, I have a somewhat interesting migration from svn to Git which I cannot get quite right: a 10 year old tree that only moved to trunk/branches/tags 3 years ago. I now migrated this into two separate git repos, 1 for the linear history of commit 1 to the day before the split (commit 'S'-1), and one with the branches/trunk/tags properly imported as branches from 'S' to HEAD. I merged these two repos, but I	09:19
	fail to 'connect' the histories	09:19
	→ dionysus69 joined	09:19
Zevv	I tried various permutations of git rebase --ont, but I'm not sure if that is the right way to go	09:19
	← Kartagis left	09:21
	→ vdamewood joined	09:22
	→ Bigcheese joined	09:22
	← aw1 left	09:23
	← Bigcheese left	09:24
canton7	Zevv, it's going to be easier to use a graft/replacement to join the two histories, then git filter-branch make the graft permanent	09:24
	→ Bigcheese joined	09:24
Zevv	ok I suspected that much, that makes sense	09:25
	im0nde_ → im0nde1	09:25
canton7	(rebase is going to want to rewrite your history to get rid of merges)	09:25
Zevv	right, and that is my problem - it wants to go too deep	09:25
	I'm not familiar with filter-branch yet, so I'll do some reading up. Thanks so far	09:26
canton7	I'm not sure what the state-of-the-art is with grafts vs replace at the moment -- reading around it seems that both can do what you want, but replace is the newer option	09:26
	filter-branch can do many things, but one of the things it can do is to rewrite all of history after a graft/replacement to keep its structure, but "bake" that graft/replacement into the actual commit history	09:26
j416	git-replace indeed; it took a long while for me to try it but it actually works nicely (and it's less of a hassle than the old way)	09:27
	→ leeN joined	09:27
Zevv	sounds like what I'm looking for. I do not care for the old history very much, apart from that I want to keep for historical and archealogical reasons	09:27
canton7	looking at the examples, grafts do seem less painful for this case	09:28
	echo "$first-commit-of-new-history $last-commit-of-old-history" >> .git/info/grafts && git filter-branch $last-commit-of-old-history..HEAD (from https://stackoverflow.com/a/3811028/1086121)	09:30
Zevv	So my initial setup make sense, ie, making 2 separate imports, a flat one of the 'old' part and a branched one of the 'new' part, getting both into the same repo, and then replace/graft to fuse them together??	09:30
canton7	you might want some extra filter-branch commands, like `--tag-name-filter cat`	09:30
	Zevv, yep, that's exactly what I'd have done	09:30
Zevv	ok thanks!	09:30
canton7	(when you were asking your initial question I was thinking "Oh god, I'm going to have to explain how to do two separate imports from different regions of history, and I can't remember the magic git-svn incantations", then "phew" when you said you'd already figured it out ><)	09:33
j416	:D	09:33
	I second that; seems like a sane way to go about it. I'd have done the same, too.	09:33
Zevv	canton7: I've done my git over the years, but some of the deeper trenches I have not seen the bottom of yet	09:34
	← leeN left	09:36
	→ gpanders joined	09:36
	→ hqdruxn08 joined	09:37
	← hqdruxn08 left	09:38
	← vdamewood left	09:39
	→ hqdruxn08 joined	09:39
	← RaceCondition left	09:41
	← hqdruxn08 left	09:41
	→ aw1 joined	09:41
	→ akemhp joined	09:46
	→ akem__ joined	09:48
	← akemhp_ left	09:48
	← akemhp left	09:50
	→ akemhp_ joined	09:52
	← TheFuzzball left	09:53
	← akem__ left	09:54
	← plasticle left	09:56
	→ plasticle joined	09:59
	→ TheFuzzball joined	10:03
	→ dartmouthed joined	10:03
	→ akem__ joined	10:03
	← akemhp_ left	10:06
	→ vdamewood joined	10:08
	← beastDiv_ left	10:09
	→ dtux joined	10:10
	← dtux left	10:14
	→ Noti joined	10:14
	→ feriman joined	10:14
	← SteffanW left	10:15
	← BlackMaria left	10:18
	→ BlackMaria joined	10:18
	← akem__ left	10:21
	→ akemhp joined	10:22
	LVK267 → LasseVK	10:25
Zevv	and of course I messed up now - after the graft, do I need to do the filter-branch for all individual branches?	10:25
	← gxt left	10:26
	→ nabil_ joined	10:26
canton7	Zevv, you can make filter-branch do everything at once	10:27
	← nabil left	10:29
	← greatgatsby left	10:30
	← gareppa left	10:33
	→ hqdruxn08 joined	10:37
	← hqdruxn08 left	10:38
	→ hqdruxn08 joined	10:38
	← hqdruxn08 left	10:39
	← ricks_ left	10:42
	→ akemhp_ joined	10:46
	← tufteddeer left	10:46
	→ tufteddeer joined	10:47
	← akemhp left	10:49
	→ aw1_ joined	10:51
	→ akemhp joined	10:53
	← aw1 left	10:53
	yn_ → yn	10:53
	← thomasross left	10:54
	← akemhp_ left	10:55
	→ Barrt joined	10:56
	→ storrgie joined	11:00
	→ Luyin joined	11:06
Luyin	hi, I want to revert changes from a branch merged into master, and later I'll want to revert the revert again. So I have a temporary measurement that needs to be in master for a while (the revert), and later I'll want that temporary solution to be gone from master again. I've read https://mirrors.edge.kernel.org/pub/software/scm/git/docs/howto/revert-a-faulty-merge.html and a lot on stackoverflow, man git-revert	11:08
gitinfo	the git-revert manpage is available at https://gitirc.eu/git-revert.html	11:08
Luyin	and more, and it seems like it's a really complicated thing. What would you recommend?	11:08
	→ dtux joined	11:11
	→ emsjessec joined	11:12
	← aw1_ left	11:13
	← dtux left	11:15
	→ ricks_ joined	11:18
	← libertyprime left	11:20
	→ omnireq_ joined	11:21
	→ libertyprime joined	11:21
	← omnireq left	11:23
	← tufteddeer left	11:23
	→ tufteddeer joined	11:23
Luyin	could I revert all the single commits from the branch but not the merge commit? would that make sense?	11:23
	not nice, maybe, but easier?	11:23
	← macaronus left	11:30
	→ roflin joined	11:32
	← Barrt left	11:37
	← caskd-dev left	11:52
	← nowhere_man left	11:52
	→ caskd joined	11:52
	→ m0viefreak joined	11:54
	PipeItToDevSda → PipeItToDevNull	11:55
	→ jguddas-tr joined	11:57
	← TheFuzzball left	11:58
	→ TheFuzzball joined	12:04
	← TheFuzzball left	12:05
snappy	thoughts on empty commits in history to use as signals/marks to delineate chunks of commits?	12:05
_ikke_	what about tags?	12:06
	→ Barrt joined	12:07
snappy	this is more like saying "start refactoring" "end refactoring" "start adding tests" "end adding tests" or something like that.	12:07
	i know these should probably be distinct branches, but the branch juggling is a bit annoying	12:07
_ikke_	sounds like branches	12:07
	which get merged in	12:07
snappy	yeah fair enough, agreed actually	12:08
	nested branches with merge commits are probably nicer for that.	12:08
	→ usm4n joined	12:08
	← usm4n left	12:08
snappy	don't think i've ever seen it used before.	12:09
	so probably not a good idea	12:09
	← Barrt left	12:10
	→ Wetmelon joined	12:11
	← zetneteork left	12:25
	← libertyprime left	12:27
	→ fk joined	12:27
	← cd left	12:30
	← BonusPlay left	12:31
	← jguddas-tr left	12:31
	→ TheFuzzball joined	12:31
	→ thomasross joined	12:31
	← Wetmelon left	12:35
	← fk left	12:37
	→ fk joined	12:38
	→ BonusPlay joined	12:39
	→ stutter joined	12:44
	→ tmager joined	12:44
	→ cd joined	12:44
	← swalladge left	12:52
Bombe	I have a repo that has mulitple remotes, some fast ones, some slow ones. Is there a way to not fetch the slow ones with “git remote update”?	12:54
Habbie	you can make groups	12:54
	!man git-remote	12:54
gitinfo	The git man pages are available online at https://gitirc.eu/git.html. Or were you looking for the "man git-foo" syntax (without the !)?	12:54
	the git-remote manpage is available at https://gitirc.eu/git-remote.html	12:54
Bombe	Hmm, I just found remote.<name>.skipFetchAll.	12:55
Habbie	or that	12:55
	i think	12:55
Bombe	How do I configure groups? Nothing about that in git-remote’s man page…	12:56
Habbie	groups are mentioned under 'update'	12:56
	i do not know more than what it says there	12:56
Bombe	Ah, it’s a single config property.	12:56
	Okay, that should help, thanks. :)	12:56
	→ Kartagis joined	12:58
	← tmager left	12:59
	→ treehug88 joined	13:00
	→ lucasb joined	13:01
	→ nika_ joined	13:05
	→ cvcxc joined	13:06
	← magic_ninja_work left	13:08
	← Makaveli7 left	13:08
	← cd left	13:10
	→ is_null joined	13:12
	→ dtux joined	13:13
	→ cd joined	13:13
	← fk left	13:15
	→ TomyLobo2 joined	13:17
	← TomyLobo2 left	13:17
	← thomasross left	13:17
	← dtux left	13:18
	→ fk joined	13:21
	← staafl left	13:23
	← caskd left	13:24
	→ tmager joined	13:28
	→ staafl joined	13:28
	→ jguddas-tr joined	13:29
	→ royal_screwup21 joined	13:32
	← jguddas-tr left	13:35
	→ caskd joined	13:35
	→ nowhere_man joined	13:36
	← igemnace left	13:36
	← bvdw left	13:38
	→ jguddas-tr joined	13:38
	→ bvdw joined	13:39
	→ malmalmal joined	13:41
	← blackmesa left	13:42
	← tufteddeer left	13:43
	→ tufteddeer joined	13:44
	← royal_screwup21 left	13:45
	→ CodeSlingerPaul joined	13:45
	→ nurdie joined	13:47
Assault	hmm..is submodules configuration not part of the repository?	13:48
	→ pedroteosousa joined	13:48
Assault	I mean, I cloned my repo, but the config file contains nothing about its subdmoules	13:48
	→ thebope joined	13:49
BtbN	they're submodules, they have their own .git	13:49
	→ greggerz joined	13:49
Assault	yes...but isnt my main repo supposed to know which submodules are part of it?	13:49
	→ thblt joined	13:49
osse	Assault: yes	13:50
	there should be a .gitmodules file	13:50
	→ nurdie_ joined	13:51
Assault	there is a .gitmodules in my original repo (on my computer), but on where I re-cloned it	13:51
	← nurdie left	13:52
osse	Assault: you should add and commit it	13:52
Assault	ooh, my other repo was on master, whereas the submodules are on another branch	13:53
	→ roadie joined	13:54
thblt	I asked on #github, but just in case: anybody knows of a way (on GitHub) to import/recreate pending PRs from an upstream repo to a fork, besides doing it by hand?	13:55
	→ igemnace joined	13:56
Habbie	thblt, you don't want to move the entire repo?	13:57
	← Luyin left	13:58
thblt	Habbie: Owner is unresponsive on GitHub, and I haven't yet decided to assume maintainership.	13:58
	← nowhere_man left	13:58
Habbie	thblt, ah	13:58
	thblt, best answer i have is 'github has a great API'	13:58
thblt	Habbie: thanks, that the answer I wanted to avoid, but if that's the only way so be it :)	13:59
fury	i think you can manually merge them, but it won't be connected to any PRs	14:00
Habbie	if i understand the question correctly, the point is not to immediately merge them all	14:00
thblt	fury: what Habbie says :) I just want to have them handy while I start to experiment with the existing code.	14:01
fury	ah	14:01
	→ akemhp_ joined	14:01
Habbie	thblt, have them handy locally?	14:01
thblt	on my clone on github, with the conversation, or anyway somewhere that isn't just upstream. I was hoping for a magic import script.'	14:02
	← tufteddeer left	14:02
	← akemhp left	14:04
	→ tufteddeer joined	14:04
	→ lembron joined	14:05
	im0nde1 → im0nde	14:05
fury	thblt: does this sound like it might work? older, but maybe the APIs it uses are still there... https://github.com/IQAndreas/github-issues-import	14:06
	← jguddas-tr left	14:07
lembron	uhm, git on windows - a while ago i did "a git split" (aka branch,rename,modify,merge 1 file to 2 to "split with history") - thats all fine, but now gitk stops working from time to time and throws me "filename to long" errors - for the last weeks hard-reload helped, but now im getting a full errormessage and things "break" - where do dig from here? https://pastebin.com/wifMwiy3 'couldn't execute	14:07
	"git": file name too long'	14:07
thblt	fury: thanks, I'll look into it, it looks like a great start!	14:08
	→ spuc joined	14:09
	→ royal_screwup21 joined	14:09
	← thebope left	14:12
	→ thebope joined	14:12
	→ dtux joined	14:13
	→ nowhere_man joined	14:14
	→ thiago joined	14:14
	→ javaJake joined	14:18
	← dtux left	14:18
	← roadie left	14:19
	← royal_screwup21 left	14:21
	← pedroteosousa left	14:24
	→ sq-one joined	14:24
	→ thomasross joined	14:25
	← vdamewood left	14:28
	→ royal_screwup21 joined	14:29
	← Stoot left	14:29
	→ Stoot joined	14:30
	← omnireq_ left	14:37
	→ welovfree joined	14:38
	← welovfree left	14:38
	← Zevv left	14:41
	← TomyWork left	14:45
	→ TomyWork joined	14:46
	← thiago left	14:46
	← nowhere_man left	14:48
	→ jguddas-tr joined	14:49
lembron	oh great, git-fow-windows itself also is broken in the current release? - all issues closed, but nether stable nor the nightlyRC fixes anything ;(	14:49
	← Inline left	14:49
	← TomyWork left	14:50
	← royal_screwup21 left	14:50
	→ TomyWork joined	14:51
	← nika_ left	14:52
	→ jlosito joined	14:54
	← jguddas-tr left	14:54
	← malmalmal left	14:55
lembron	oh wtf... did i brick my system now or what is this =(	14:56
	← cdown left	14:56
	→ nika_ joined	14:56
	→ royal_screwup21 joined	14:56
	→ cdown joined	14:57
lembron	"bash: /dev/fd/63: No such file or directory" -- while cleanly showing a link "/dev/fd -> /proc/self/fd/" and all of `<(cmd)` is now broken.	14:57
	→ boombatower joined	15:00
thblt	lembron: your day sounds fun	15:01
	← Assault left	15:02
	← royal_screwup21 left	15:02
lembron	literally just turned super-shit-hole ;( -- from "gitk nudges for 2 seconds every few days" to "gitk is broken" to "bash is broken" - cant do anything now	15:02
	→ Assault joined	15:05
thblt	fwiw googling "git: file name too long" returns a few intersting things	15:06
	← Kartagis left	15:10
lembron	whelp, leveling _everything_ and on the 5th reinstall things started working again	15:12
	→ omnireq joined	15:13
	→ dtux joined	15:15
	← dtux left	15:19
	← qqx left	15:20
	→ qqx joined	15:20
	← thomasross left	15:23
	→ algiss joined	15:23
	← nika_ left	15:28
sangy	okay, I bet I'm being silly but my brain is being stupid today	15:32
	doesn't this transport/remote-helper work │fatal: unable to find remote helper for 'git+https' ?	15:32
	(I know I can use this with vanilla https)	15:32
	← dionysus69 left	15:33
	→ nika_ joined	15:33
	→ roadie joined	15:36
	→ gareppa joined	15:37
	← TheFuzzball left	15:40
	→ dtux joined	15:46
	→ thiago joined	15:50
	← diogenese left	15:51
	→ akem__ joined	15:51
	→ diogenese joined	15:51
	← greggerz left	15:51
	→ greggerz joined	15:52
	← akemhp_ left	15:53
	→ TheFuzzball joined	15:54
	→ pedroteosousa joined	15:56
	← akem__ left	15:57
	→ blackmesa joined	15:57
	→ HumanG33k joined	16:00
	← Noti left	16:05
	← fflam left	16:07
	→ jguddas-tr joined	16:07
	→ fflam joined	16:09
	← jguddas-tr left	16:15
	→ mat001 joined	16:18
jrnieder	sangy: yeah, I think you're looking for https://	16:18
	sangy: there's git+ssh:// and ssh+git:// but those are considered to have been a mistake	16:18
thblt	r	16:21
	← pedroteosousa left	16:21
thblt	(sorry)	16:21
sangy	jrnieder: thanks! yeah I realized this is an upper-layer speak. I thought it was just blindly passed to git	16:23
	← TomyWork left	16:30
	→ fannagoganna joined	16:36
	← _ADN_ left	16:38
	→ orbyt_ joined	16:38
fannagoganna	Hi, I am trying to change history, by changing one email address of commit to another. I follow instructions in https://www.git-tower.com/learn/git/faq/change-author-name-email and section "Editing the author of past commits." Specifically, suggested command that starts with "git filter-branch."	16:39
	I run that command, but when I do a "git log" I can still see old email addresses. Old email addresses haven't changed.	16:39
_ikke_	fannagoganna: What is the actual command that you ran	16:40
fannagoganna	https://www.irccloud.com/pastebin/UwgUh4K7/example+git+filter+branch.txt	16:41
	Old email is one I want to change to new email, in this case "tanim.islam@gmail.com"	16:41
	→ nicexe joined	16:43
	← Exec1N left	16:45
_ikke_	fannagoganna: there is a new tool called filter-repo	16:45
	https://github.com/newren/git-filter-repo	16:45
	→ tgummerer joined	16:45
_ikke_	It should make these kinds of operations easier	16:45
	→ bitwinery joined	16:45
_ikke_	for your usecase, there is a --mailmap option	16:45
fannagoganna	_ikke_: thanks I will try it out	16:46
	← igemnace left	16:46
	→ gxt joined	16:48
	→ dionysus69 joined	16:57
	→ prepend joined	17:02
	← Xiti left	17:02
	→ Xiti joined	17:03
	→ igemnace joined	17:06
	← sq-one left	17:07
	→ g00s joined	17:11
	← fk left	17:12
	→ fk joined	17:14
	← captain42 left	17:19
	→ fromBeyond joined	17:19
	← TheFuzzball left	17:19
	← cbreak left	17:20
	→ cbreak joined	17:21
fromBeyond	Hi. Suppose I have 3 Projects, each in its own repo. We have that R2 depends on R1, and R3 depends on both R1 and R2. How should I structure this? When all I had was R2 and R1, I simply added R1 as a submodule of R2. But i am unsure how to do it now. I hope my wording was understandable	17:22
	→ Goplat joined	17:22
	← paraxial left	17:22
	→ paraxial joined	17:23
	← orbyt_ left	17:25
f8e3	git summary --line does it count per lines remaining or lines commited ?	17:26
	← paraxial left	17:28
f8e3	the doc is silent: Summarize with lines other than commits. , so probably total	17:31
	← qqx left	17:32
	← nika_ left	17:32
	→ qqx joined	17:32
	→ paraxial joined	17:35
	← g00s left	17:36
	→ nika joined	17:41
	→ cvcxc_ joined	17:41
	→ orbyt_ joined	17:42
j416	fromBeyond: git is not a dependency manager; use a proper dependency manager or put them in one repo	17:43
	← JanC left	17:43
	← cvcxc left	17:44
	→ JanC joined	17:45
	← m0viefreak left	17:48
	→ tang^ joined	17:49
	← tang^ left	17:49
	← nika left	17:53
	→ dijitol joined	17:55
	← roadie left	17:57
	→ roadie joined	17:58
	→ Sasazuka joined	18:06
	← greggerz left	18:12
	→ greggerz joined	18:12
	← prepend left	18:15
	→ wootehfoot joined	18:16
	→ fission6 joined	18:18
	→ Phylock joined	18:20
	← nurdie_ left	18:22
	→ nurdie joined	18:22
	← algiss left	18:23
	→ Blackpajamas joined	18:25
	← nicexe left	18:26
	← Phylock left	18:27
	← nurdie left	18:27
	← gareppa left	18:27
	→ nicexe joined	18:28
	← blackmesa left	18:29
	← tufteddeer left	18:31
	→ tufteddeer joined	18:31
	→ Keytap_ joined	18:33
	→ royal_screwup21 joined	18:33
	← Blackpajamas left	18:35
Keytap_	Afternoon all. Just making sure, but there's no flag I can give to 'git update-index --assume-unchanged [...]' to make it only last until the next time the specified files change, is there?	18:35
	← BlackMaria left	18:35
_ikke_	No, that's kind of contradictory :)	18:35
cbreak	Keytap_: it doesn't change	18:35
	← tufteddeer left	18:36
Keytap_	WOuldn't have thought. I'm probably going about what I"m trying to do the wrong way anyways.	18:36
	→ Blackpajamas joined	18:36
	← fission6 left	18:37
_ikke_	there is --skip-worktree, which is better suited for this purpose, but still, there is no option to have that flag disappear when the file changes	18:37
cbreak	the point of assume-unchanged is to prevent git from wasting time checking things for changes that don't change	18:38
	a performance optimization	18:38
	unless you have gigantic repositories or a very slow file system, you probably don't need it	18:38
_ikke_	yes, indeed, that's the contradictory part	18:38
cbreak	Keytap_: have you considered getting a better file system? :)	18:39
_ikke_	What is it what you are trying to achieve?	18:39
	→ Phylock joined	18:40
	→ beastDiv_ joined	18:42
Keytap_	cbreak,_ikke_: I'm building a docker container and I'm replacing one of the files with a proxy script that calls the command inside the container (as a convenience tool for the developer), but the proxy script completely rewrites the content of the original file, so I'm currently doing --assume-unchanged so the proxy script doesn't accidentally get added to source control	18:42
	→ BlackMaria joined	18:42
cbreak	just give the script a different name	18:42
	and gitignore it	18:43
	→ tufteddeer joined	18:43
	→ blackmesa joined	18:55
	← Blackpajamas left	18:56
	→ nurdie joined	18:56
	→ rath joined	18:58
	→ l4sh joined	18:59
	← dionysus69 left	19:01
	→ dionysus69 joined	19:01
	← iam730 left	19:04
	→ iam730 joined	19:06
	→ duderonomy joined	19:09
	← duderonomy left	19:10
	← lucasb left	19:10
	→ Blackpajamas joined	19:13
	← Phylock left	19:14
	← fannagoganna left	19:15
	← thiago left	19:18
	← sentriz left	19:22
	→ soju joined	19:23
	→ sentriz joined	19:23
	→ pedroteosousa joined	19:24
cyberpear	Keytap_: you could probably do some magic with git replace	19:25
	→ leeN joined	19:25
	← Blackpajamas left	19:26
cyberpear	Several years ago, filter Branch did not properly rewrite replaced blobs	19:26
	← nurdie left	19:26
	→ Blackpajamas joined	19:27
	→ gioyik joined	19:27
	← gioyik left	19:27
	→ nurdie joined	19:27
	→ gioyik joined	19:29
	→ crose joined	19:30
	← gioyik left	19:30
	← ricks_ left	19:31
	← steven__ left	19:31
	← nurdie left	19:32
	← royal_screwup21 left	19:33
cyberpear	you could lie to git and tell it that your modified file is the file in the most recent commit. Be aware that if someone commits changes to the file, it will look like they are committing a smaller change than they actually are committing as your previously hidden modifications would be included in the new commit	19:34
	(In theory)	19:34
	Keytap_: ^	19:35
	( attempting to answer "how could you do it", not "should you do it")	19:36
	← bvdw left	19:38
	→ bvdw joined	19:39
	→ fstd_ joined	19:41
	← emsjessec left	19:42
	← fstd left	19:45
	fstd_ → fstd	19:45
	→ nabil joined	19:48
	← nabil left	19:49
	→ nabil joined	19:49
	← nabil_ left	19:50
	← tufteddeer left	19:54
	→ jfariasf joined	19:54
	→ tufteddeer joined	19:54
	→ swalladge joined	19:56
	← tufteddeer left	19:58
	← feriman left	19:59
	→ nabil_ joined	19:59
	← throstur left	20:01
	← nabil left	20:02
	→ nabil__ joined	20:03
	← nabil_ left	20:05
	← soju left	20:08
	→ mszyszko joined	20:10
	← mszyszko left	20:10
	→ mszyszko joined	20:11
	→ tufteddeer joined	20:12
	→ thiago joined	20:12
	← Blackpajamas left	20:13
	→ throstur joined	20:14
	→ fattredd_ joined	20:15
	→ Blackpajamas joined	20:15
	→ mszyszko_ joined	20:15
	← mszyszko left	20:16
	→ Inline joined	20:17
	← mszyszko_ left	20:19
	← fattredd_ left	20:20
	← Decobus left	20:22
	← Blackpajamas left	20:24
	→ soju joined	20:24
	← is_null left	20:25
	→ Exec1N joined	20:28
	→ n001g joined	20:30
	← ExeciN left	20:30
	→ mcerb joined	20:30
mcerb	I appear to be in a detached head state	20:30
	← n000g left	20:30
	n001g → n000g	20:30
Sasazuka	if I do cherry-pick with "-n" is there a way to undo that?	20:31
	→ igghibu joined	20:31
	→ greatgatsby joined	20:31
	← throstur left	20:33
mcerb	I've never been more afraid of losing my HEAD than I am in this moment	20:34
Habbie	mcerb, if you are worried about losing your head, 'git branch pleasesavemyhead' will give it a name	20:35
_ikke_	!detached	20:35
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	20:35
Habbie	!reattach	20:35
gitinfo	Letters refer to !detached. (a) and (b): 'git checkout branchname' to continue working on another branch, or 'git checkout -b branchname' to start a new one here; (c) git am --continue; (d) git rebase --continue	20:35
	→ throstur joined	20:36
	→ Wetmelon joined	20:36
mcerb	ah good, there it is	20:36
	← beastDiv_ left	20:36
	← tufteddeer left	20:36
mcerb	so now I'm back to my real problem, I commited something locally to the wrong branch	20:36
	→ tufteddeer joined	20:36
mcerb	and I want my last two commits to be in a different branch	20:36
	← dionysus69 left	20:37
	→ dionysus70 joined	20:37
	→ jgrim joined	20:38
	dionysus70 → dionysus69	20:39
vishal	mcerb: do you care about the history (i.e. the sha IDs of the commits)?	20:39
mcerb	I suppose not particularly	20:39
vishal	IOW, have those been pushed anywhere in a way someone could have used them for basing more work off?	20:39
	you cherry-pick them to whatever branch you want, but that will !rewrite those commits	20:40
gitinfo	Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum	20:40
	→ soju_ joined	20:40
mcerb	ok so I suppose I do care about the history	20:41
	but only inasmuch as I don't want it to mess up anyone else	20:41
	I think I can just make my new branch and then delete the old branch	20:42
vishal	if the branch on which they are is significant, and you want them gone from this branch, then you will have to revert them	20:42
mcerb	they're not significant	20:43
	← dionysus69 left	20:43
vishal	you can still cherry-pick them to the new branch, they will look like new commits with new sha IDs	20:43
	← soju left	20:43
vishal	then you can just reset --hard the old branch if they're just on top, or rebase -i and drop them if they are further back in the history	20:43
	reset --hard is considered as a history rewrite and you will have to force push the branch	20:44
	→ nabil joined	20:44
mcerb	I think that last part is innaccurate	20:46
vishal	hm?	20:46
mcerb	I just did git reset --hard and then git push and nothing errored on me	20:46
vishal	did you have it pushed /before/ the reset?	20:46
	← nabil__ left	20:46
mcerb	no	20:46
vishal	well then it is not a problem :)	20:47
mcerb	ok good	20:47
	I have thus far avoided having to force push anything	20:47
vishal	its only a rewrite if you end up deleting the commits you pushed	20:47
mcerb	I suppose that makes me not a very experienced Jedi	20:47
b0nn	--force-with-lease	20:47
	← roadie left	20:47
	→ beastDiv_ joined	20:48
	← rath left	20:48
	→ Ploppz joined	20:50
	→ cahoots joined	20:52
cahoots	hi, i want to have a hook run on cherry-picks, is there any way to do this?	20:53
Ploppz	I have a feature branch with many commits ahead of (and 1 commit behind) master. I want to rebase it on top of master. I find it overly difficult to do this the usual way which seems to be per commit. These commits span back many months ago and I can't remember exactly what change I did back then. Is there any alternative to just sort of accept	20:53
	all changes done IN TOTAL by the feature branch, and only resolves when there is some conflict between CURRENT master and CURRENT feature branch?	20:53
	Hope you understand my query	20:53
	← soju_ left	20:54
	→ rafasc joined	20:54
Ploppz	When I started rebasing the first commit, it seemed like I was working with very old code, and for example it had files that are now long gone ... so I wonder if there is a way where I don't have to deal with these intermediate files that are not even relevant anymore	20:54
	← rafasc left	20:54
	← dtux left	20:56
	→ g00s joined	20:59
	← thiago left	21:01
cdunklau	Ploppz: git isn't magic, it's just as dumb as any program	21:02
	→ nurdie joined	21:03
	→ gareppa joined	21:03
cdunklau	Ploppz: first step is really understanding the differences in the history (git log --oneline --decorate yourbranch)	21:04
	→ ferdna joined	21:04
cdunklau	Ploppz: if it's simple enough, you could squash all the commits not on master to make it simpler to look at	21:04
	← tmager left	21:05
Ploppz	ahh I think it would be beneficial to keep the history :(	21:05
cdunklau	"accept all changes done IN TOTAL by the feature branch" <- this can't be done in general, with any real reasoning	21:05
Ploppz	ok I do understand it given I want to rewrite the entire history of the feature branch..	21:06
	thanks	21:06
cdunklau	in some circumstances it could	21:06
	Ploppz: but !dots might be helpful	21:06
gitinfo	Ploppz: A..B = stuff that happened between A and B (if A and B are related; otherwise refer to "man gitrevisions"), A...B = (a) in history viewers: stuff that isn't in both A and B yet; (b) in "git diff": stuff that happened in B since the two diverged; (c) in "git checkout": the merge base of A and B. "master.." is the same as "master..HEAD" and "..master" is the same as "HEAD..master", and so forth.	21:06
cdunklau	say, git log master..feature	21:06
	Ploppz: if you don't have any merges in your feature branch, it'll be pretty easy to figure out what happened	21:07
	← nurdie left	21:08
cdunklau	git log master...feature might be enlighting too, especially in comparison with master..feature (two dots)	21:08
	→ rafasc joined	21:10
	→ blackmesa1 joined	21:10
	→ roadie joined	21:12
	← blackmesa left	21:12
	← g00s left	21:16
	→ GunArm joined	21:16
GunArm	I'm struggling to deal with a bunch of problems caused by differeing autocrlf settings on a series of machines working on the same repo. am I correct that I should be able to use .gitattributes to make per-repo settings that override whatever their local git config says about that?	21:18
	← gareppa left	21:19
GunArm	just wanted to bounce this off someone before I do it and make a ruckus about everyone merging it	21:19
	→ thiago joined	21:19
	→ g00s joined	21:19
rafasc	GunArm: more or less. Once the the project is in good place, and a gitattributes is in effect, it should be harder to introduce eol errors.	21:20
	→ sazop joined	21:20
	← igghibu left	21:20
rafasc	but it's not a panacea.	21:21
	← crose left	21:21
GunArm	sure. i realize someone could still force it	21:22
rafasc	'* text=auto' will fix most of the problems.	21:23
	but you'll need to git add --renormalize . ; once after you change it.	21:23
	← g00s left	21:26
	→ g00s joined	21:28
	→ thomasross joined	21:31
	→ soju joined	21:33
	← beastDiv_ left	21:35
	← Ori_B left	21:35
	← soju left	21:38
	→ Ori_B joined	21:40
	→ blackmesa joined	21:41
	← blackmesa1 left	21:41
Ploppz	thanks	21:44
	Now, during rebase: `git rebase master` (in the feature branch). In merge conflicts, what does HEAD refer to?	21:44
	there is one part called "HEAD" and then one part called with the name of the commit currently being rebased	21:45
	not sure which part belongs to master and which to the feature branch	21:45
	← g00s left	21:46
Ploppz	seems like the HEAD part is the master part	21:46
	Honestly why can I not just for every commit, select the past form the feature branch (that is not HEAD / belonging to master)?	21:48
	→ g00s joined	21:48
	← rafasc left	21:48
_ikke_	HEAD is the commit that already exists, the other part is the commit that is being applied	21:50
Ploppz	ok thanks	21:53
	I'm just going to go ahead and systematically remove the HEAD part anyway	21:53
	→ soju joined	21:54
	→ beastDiv_ joined	21:55
	← beastDiv_ left	21:55
	← g00s left	21:56
	→ blackmesa1 joined	21:57
	→ dtux joined	21:57
	← cahoots left	21:58
	← soju left	21:58
	← blackmesa left	21:59
	← cbreak left	22:01
	→ cbreak joined	22:02
	← dtux left	22:03
	← thblt left	22:05
	→ crose joined	22:05
	→ mattfly joined	22:06
mattfly	How do i delete my last local comit?	22:07
_ikke_	git reset --hard HEAD~ # cannot be undonde	22:08
	undone	22:08
mattfly	and how do I go 2 back?	22:08
_ikke_	git reset --hard HEAD~2	22:08
mattfly	okay...	22:08
_ikke_	man gitrevisions	22:08
gitinfo	the gitrevisions manpage is available at https://gitirc.eu/gitrevisions.html	22:08
	→ yonder joined	22:13
	← pedroteosousa left	22:13
	← tufteddeer left	22:16
	→ tufteddeer joined	22:16
	← l4sh left	22:20
	← tufteddeer left	22:21
	← greggerz left	22:21
	→ tufteddeer joined	22:22
	→ greggerz joined	22:25
	→ Shentino joined	22:27
	← greggerz left	22:27
	← Shentino left	22:28
	→ m0viefreak joined	22:28
	→ Shentino joined	22:28
	→ LearnAllTheTime joined	22:28
	← jgrim left	22:34
	→ soju joined	22:35
	← Shentino left	22:38
	→ Shentino_ joined	22:38
	← soju left	22:40
	→ royal_screwup21 joined	22:41
	→ localhorse joined	22:41
	→ jguddas-tr joined	22:44
	→ rafasc joined	22:44
	← CodeSlingerPaul left	22:45
	← dijitol left	22:48
	← cvcxc_ left	22:48
	← jguddas-tr left	22:49
	← omnireq left	22:53
	← ferdna left	22:55
	→ dtux joined	22:58
	→ vicfred joined	23:01
	→ soju_ joined	23:03
	← dtux left	23:03
	← shabius left	23:11
	← Rovanion left	23:12
	→ dtux joined	23:13
	→ Rovanion joined	23:16
	← tufteddeer left	23:17
	→ tufteddeer joined	23:27
	← thebope left	23:29
	→ Brainium joined	23:31
	← tufteddeer left	23:32
	← soju_ left	23:32
	← clemons left	23:32
	← tubaman left	23:33
	→ clemons joined	23:33
snappy	so i have a series of commits, i want to ensure they all build; so i do git rebase -exec ./check.sh master - it drops me into the rebase conflict; i was hoping instead it'd just print the commit and short comment, and continue	23:33
	my other thought was to use git bisect for this - but bisect doesn't seem to limit search to master, but it's also for finding a single commit instead of multiple commits.	23:34
	→ thebope joined	23:34
rafasc	snappy: git rebase flattens history, you would need at least --rebase-merges	23:34
snappy	oh good point	23:35
	but even rebase-merges isn't a guarantee if i recall	23:35
rafasc	but neither auto-resolves conflicts.	23:35
	← gxt left	23:35
rafasc	you're better off checking out each commit, and running it.	23:35
snappy	right, i just want to detect which commits have a "conflict" (i.e. exec fails)	23:36
rafasc	exec failing isn't a conflict.	23:36
snappy	er maybe i used the wrong terminology.	23:36
rafasc	snappy: what do you want? a) ensure all commits pass check.sh? b) find the commit that broke check.sh?	23:37
snappy	it'll drop you into rebase mode if the execution fails.	23:37
	rafasc: (a)	23:37
	and if a commit fails, i want (b) without dropping into rebase exec failure mode, just print the commit.	23:38
	so i can see all failing commits.	23:38
	← cd left	23:38
	← thebope left	23:39
	← mattfly left	23:39
Ploppz	More question about rebasing. Actually.. there's no way it should be as hard as it is right now, when I do `git rebase upstream/master`. I'm trying to rebase the branch of this PR ( https://github.com/TLmaK0/rustneat/pull/39 ) onto the master branch of that repo. As you can see it is ONE commit behind master. And that commit is this one	23:40
	https://github.com/TLmaK0/rustneat/commit/0a9f13683a39c0c6b0c875f0ee6159b485042dca .	23:40
rafasc	snappy: then parse git rev-list A..B; check out each commit, test it, print if fails	23:40
snappy	ok	23:40
Ploppz	when I do `git rebase upstream/master` on my feature branch, I have to go through every commit and I get conflicts in every single one of them, all of which I have no idea how to solve	23:41
	→ Newami joined	23:41
snappy	Ploppz: are they repeated conflicts?	23:41
	← dege left	23:41
	← Newami left	23:42
rafasc	snappy: git rev-list master~10..master \| while read -r rev; do git checkout "$rev"; if ! ./check.sh; then printf '%s is bad' "$rev" ; done	23:42
	bash oneliner	23:42
snappy	cheers	23:42
rafasc	snappy: with errors included, like forgetting to close the if clause.	23:42
	and you'll probably want to use checkout -f, so it doesn't fail.	23:43
Ploppz	@snappy some were repeated I think, but also new ones were introduced	23:43
snappy	Ploppz: for repeated ones, you can use git rerere -- but what might be easier if it's been so long, is to branch from where you are, reset to upstream/master and incrementally redoing your code.	23:44
	→ workingjubilee joined	23:45
Ploppz	what do you mean with reset to upstream/master?	23:45
snappy	basically start a new branch from upstream/master; git branch fixeverything upstream/master	23:46
	and then just stage your code changes on top of that.	23:46
	→ tufteddeer joined	23:47
snappy	that's how i would do it anyway, particularly if you've been working out of sync from upstream/master for way too long.	23:47
Ploppz	That would erase the history then (I mean not include it in master)?	23:48
	← royal_screwup21 left	23:48
Ploppz	But look, it's only one commit behind master, one which only changes a line in README. There must be some easy way to rebase it?	23:48
	behind upstream/master*	23:48
	if only that one line was not changes - if that latest commit on upstream master was not there, then it would be trivial!	23:49
snappy	is that what git diff upstream/master says?	23:49
	er rather git fetch upstream && git diff upstream/master	23:49
	hold up, let me pull this in.	23:50
	← workingjubilee left	23:51
	→ workingjubilee joined	23:51
	→ omnireq joined	23:51
snappy	git diff upstream/master says otherwise.	23:52
Ploppz	git diff upstream/master has thousands of lines because my own branch (let's call it origin/master since I call my fork remote origin) is many commits ahead. But only one commit behind.	23:52
	← tufteddeer left	23:53
Ploppz	if you go here https://github.com/Ploppz/rustneat it says "This branch is 74 commits ahead, 1 commit behind TLmaK0:master. "	23:53
	→ tufteddeer joined	23:53
snappy	oh yeah i see, origin last commit was 1 year ago.	23:53
Ploppz	and if it were not for that one latest commit to TLmaK0:master, rebase would be trivial	23:53
snappy	hold up.	23:53
	you can probably interactive rebase on your own and cherry-pick in the commit.	23:54
	← sazop left	23:54
	→ mattfly joined	23:55
	← fromBeyond left	23:55
Ploppz	sorry what do you mean?	23:56
	→ duderonomy joined	23:57
Ploppz	I could give up and merge upstream/master -> origin/master instead...	23:58
	perhaps I will	23:58
rafasc	Ploppz: First things first. Your branch has some merges, what do you want to do with them?	23:58
	and commits like "15b68b5 ."	23:59

Logs Search ←Prev date Next date→

Channels Documentation