IRCloggy #git 2021-04-08

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2021-04-08

← AnAverageHuman left00:00
amerlyq computeiro: git checkout .00:00
is you really want to lose your changes, of course00:00
→ AnAverageHuman joined00:00
amerlyq Is there any way to limit git-log and git-show message to <=N lines? I'm looking through linux-kernel commit messages and want to speed-up the process somewhat.00:01
← arcatech_ left00:03
→ DevAntoine joined00:05
← codebam left00:06
→ codebam joined00:07
computeiro amerlyq: ty00:09
← DevAntoine left00:10
← tiin57 left00:13
→ tiin57 joined00:16
← codebam left00:16
→ codebam joined00:17
→ causasui joined00:18
← Sazhen86 left00:20
→ DevAntoine joined00:21
← codebam left00:22
→ codebam joined00:24
← misuto left00:24
← kh31d4r left00:24
← DevAntoine left00:25
→ jstimm joined00:28
← stats4647 left00:28
→ stats4647 joined00:29
← codebam left00:31
→ kh31d4r joined00:32
← stats4647 left00:33
→ codebam joined00:34
← cd left00:35
← nivag left00:35
→ DevAntoine joined00:37
→ misuto joined00:41
← codebam left00:41
← DevAntoine left00:41
→ nivag joined00:42
→ nivag_ joined00:42
← tiin57 left00:42
→ Hello71 joined00:44
→ codebam joined00:44
→ tiin57 joined00:46
← nivag_ left00:49
← codebam left00:50
← zebrag left00:52
→ zebrag joined00:52
→ DevAntoine joined00:52
→ codebam joined00:54
→ xissburg joined00:56
← bin4ry81 left00:57
← DevAntoine left00:57
→ bin4ry81 joined00:57
→ stats4647 joined00:59
← krofe left01:00
← codebam left01:01
← gearnode left01:03
→ codebam joined01:03
← stats4647 left01:05
← thomas_25- left01:05
← thomas_25 left01:07
→ DevAntoine joined01:08
→ budonyc joined01:08
→ thomas_25 joined01:09
→ drbean_ joined01:10
← codebam left01:10
← ARoxdale left01:10
→ codebam joined01:11
← tiin57 left01:13
← DevAntoine left01:13
← gast0n left01:14
→ tiin57 joined01:16
← codebam left01:18
→ gast0n joined01:18
← Jackneill left01:20
→ thomas_25- joined01:20
→ codebam joined01:21
← vicfred left01:24
→ DevAntoine joined01:24
← PJBoy left01:25
← gast0n left01:27
← cads left01:27
→ cads joined01:28
← DevAntoine left01:28
→ gast0n joined01:31
← maxel left01:32
← budonyc left01:32
→ Jackneill joined01:32
← cads left01:32
→ cads2 joined01:33
← codebam left01:35
→ codebam joined01:35
← gast0n left01:40
→ DevAntoine joined01:40
→ nsc joined01:41
← shush left01:41
← codebam left01:42
← zebrag left01:42
← nsc_ left01:42
← tiin57 left01:42
→ royal_screwup21 joined01:43
→ gast0n joined01:44
← DevAntoine left01:44
→ shush joined01:44
→ maxel joined01:45
← magic_ninja left01:45
→ codebam joined01:45
→ magic_ninja joined01:46
→ tiin57 joined01:46
← Xenguy left01:47
→ Xenguy joined01:48
← royal_screwup21 left01:48
→ magic_ninja1 joined01:48
← magic_ninja left01:48
→ dimi1947 joined01:48
← dimi1947 left01:50
← codebam left01:52
→ dimi1947 joined01:52
→ codebam joined01:53
← gast0n left01:53
← dimi1947 left01:54
→ dimi1947 joined01:54
→ stats4647 joined01:55
→ DevAntoine joined01:55
← dimi1947 left01:57
→ dimi1947 joined01:57
→ gast0n joined01:57
← dimi1947 left01:58
→ dimi1947 joined01:58
← cads2 left01:59
← stats4647 left01:59
← gast0n left02:00
← DevAntoine left02:00
→ rosco joined02:02
← rosco_y left02:04
← computeiro left02:05
← magic_ninja1 left02:06
→ magic_ninja joined02:07
→ va5c0 joined02:07
← codebam left02:09
← kadoban_ left02:10
← causasui left02:11
→ DevAntoine joined02:11
→ codebam joined02:12
← tiin57 left02:12
→ gast0n joined02:12
← va5c0 left02:14
← DevAntoine left02:15
→ tiin57 joined02:16
← codebam left02:19
→ causasui joined02:20
← dimi1947 left02:20
→ dimi1947 joined02:21
→ codebam joined02:22
← gast0n left02:23
← Fokst left02:26
→ Fokst joined02:27
→ DevAntoine joined02:27
→ jimklimov joined02:27
→ Muimi joined02:28
← codebam left02:29
← jimklimov1 left02:30
← mindCrime_ left02:31
← DevAntoine left02:31
→ codebam joined02:32
→ gast0n joined02:32
← codebam left02:36
→ jaziz1 joined02:37
← jaziz left02:40
→ codebam joined02:42
→ Whistler joined02:42
← tiin57 left02:42
→ DevAntoine joined02:43
← gast0n left02:43
← lawt left02:44
→ tiin57 joined02:46
→ lawt joined02:47
← DevAntoine left02:47
← codebam left02:48
→ snowkidimd joined02:50
→ gast0n joined02:51
← shush left02:52
→ codebam joined02:52
→ shush joined02:52
→ arcatech joined02:53
→ InfiniteZero_ joined02:55
→ FinnElija joined02:56
finn_elija → Guest8275002:56
FinnElija → finn_elija02:56
← Aquazi left02:57
← shush left02:58
← InfiniteZero left02:58
→ DevAntoine joined02:58
← Guest82750 left02:59
→ prg3 joined03:00
← AlpineLlama left03:00
← gast0n left03:00
← gschanuel left03:00
← codebam left03:01
→ codebam joined03:01
→ gschanuel joined03:01
→ AlpineLlama joined03:02
← DevAntoine left03:03
← rosco left03:04
← arcatech left03:04
← codebam left03:06
→ gast0n joined03:08
→ powerhouse_ joined03:08
← Loomaanaatii left03:08
→ codebam joined03:09
← causasui left03:09
→ Loomaanaatii joined03:12
← tiin57 left03:12
← gast0n left03:12
→ DevAntoine joined03:14
→ tiin57 joined03:16
→ arcatech joined03:18
← DevAntoine left03:19
← batrick left03:19
→ shush joined03:23
← arcatech left03:23
← thomas_25 left03:23
→ thomas_25 joined03:24
→ Goplat joined03:28
→ DevAntoine joined03:30
← DevAntoine left03:35
← tiin57 left03:42
→ royal_screwup21 joined03:43
← CryptoClub left03:45
→ tiin57 joined03:45
→ DevAntoine joined03:46
← royal_screwup21 left03:48
→ CryptoClub joined03:48
→ Loomaanaatii2 joined03:48
← hussam left03:49
← Loomaanaatii left03:49
Loomaanaatii2 → Loomaanaatii03:49
← snowkidimd left03:52
← DevAntoine left03:53
→ justanotheruser joined03:55
→ tmz_ joined04:00
← tmz left04:00
→ cads2 joined04:07
← nrh^ left04:09
→ nrh^ joined04:10
← tiin57 left04:12
→ tiin57 joined04:16
← nrh^ left04:17
→ nrh^ joined04:17
→ DevAntoine joined04:20
← drbean_ left04:21
← powerhouse_ left04:22
→ snowkidimd joined04:22
← DevAntoine left04:24
→ batrick joined04:27
← shush left04:27
← rangergord left04:30
→ rangergord joined04:33
→ segnior joined04:34
→ ferdna joined04:38
→ shush joined04:40
→ lagrange joined04:41
← tiin57 left04:43
← rangergord left04:45
← segnior left04:45
Shentino_ → Shentino04:45
→ tiin57 joined04:46
→ segnior joined04:46
→ rangergord joined04:47
← thinkpad left04:50
→ thinkpad joined04:50
→ DevAntoine joined04:51
← ferdna left04:52
← Whistler left04:53
← hqdruxn08_ left04:54
← DevAntoine left04:56
← lagrange left04:58
← HumanG33k left05:01
→ HumanG33k joined05:01
← kyshoc left05:08
→ kyshoc joined05:09
← kyshoc left05:09
→ kyshoc joined05:09
← dimi1947 left05:10
→ hqdruxn08 joined05:11
← durham left05:12
← tiin57 left05:12
← Heirlung left05:14
→ tiin57 joined05:15
→ Guest20 joined05:17
→ Heirlung joined05:17
← austin987 left05:19
← madewokherd left05:25
← chemsmith left05:28
← thiago left05:29
→ shabius joined05:30
→ chemsmith joined05:30
→ ferdna joined05:31
← jimklimov left05:32
→ austin987 joined05:32
→ jimklimov joined05:34
← Mmike left05:36
← ferdna left05:37
→ vorona joined05:40
← AnAverageHuman left05:42
← tiin57 left05:42
→ FH_thecat joined05:43
→ royal_screwup21 joined05:44
← Goplat left05:44
→ crose joined05:44
→ tiin57 joined05:46
← royal_screwup21 left05:48
← edlou left05:58
← FH_thecat left06:03
← tiin57 left06:12
← sudoforge left06:14
→ tiin57 joined06:16
← k3yp left06:17
Tabmow If I have this sort of checkout - git checkout --progress --force -B feature/branch refs/remotes/origin/feature/branch06:21
How do I see what has changed since last commit?06:22
git diff --name-only --exit-code HEAD^ HEAD -- . gives me: fatal: bad revision 'HEAD^'06:23
→ k3yp joined06:26
→ DevAntoine joined06:26
→ sord937 joined06:27
→ sudoforge joined06:27
← Betal left06:27
← DevAntoine left06:31
→ john_johnk joined06:34
← Ranhir left06:34
→ LetoThe2nd joined06:37
→ xrisovalantis joined06:38
← sozuba left06:39
→ sozuba joined06:41
→ guerra joined06:42
← tiin57 left06:42
j416 Tabmow: what is it that you are actually trying to do?06:43
→ tiin57 joined06:46
← sozuba left06:46
→ KnoP joined06:47
→ sozuba joined06:47
→ Ranhir joined06:48
Tabmow j416: in my github action I need to check if there are any changes to a certain directory.06:48
I found the issue, the checkout action defaults to fetch-depth 1 so there was no other history :) - Changing it to 2 fixed my issue.06:49
The error message was throwing me...06:49
→ lunatic_wei joined06:52
j416 you don't need checkout for that; diff should be enough06:53
git diff --exit-code origin/feature/branch^..origin/feature/branch -- .06:54
← raymond left06:54
Tabmow j416: true, I need checkout for other things anyway so was just easier to use that until I ran into this slight issue.06:56
j416 check.06:57
→ rcoote joined06:57
→ DevAntoine joined06:58
→ rfuentess joined07:02
← DevAntoine left07:02
→ dpl joined07:02
← unluckyshrubbery left07:03
→ thansen9 joined07:07
← thansen left07:07
thansen9 → thansen07:07
← tiin57 left07:12
→ DevAntoine joined07:13
→ tiin57 joined07:16
→ bolovanos joined07:16
→ Haohmaru joined07:25
→ otisolsen70 joined07:25
→ fandre1986 joined07:26
← Guest20 left07:26
← sudoforge left07:28
← xissburg left07:33
→ Ignacio joined07:34
→ Guest20 joined07:34
→ sudoforge joined07:36
← howdoi left07:36
← shush left07:39
→ jimmiehansson joined07:40
← tiin57 left07:42
← mw_ left07:42
→ mw_ joined07:43
← DevAntoine left07:44
→ tiin57 joined07:45
← Mattiasl_ left07:46
← rednplueplaid96 left07:48
→ rednplueplaid96 joined07:49
→ powerhouse joined07:49
← fandre1986 left07:49
→ vorona1 joined07:50
← vorona left07:51
→ DevAntoine joined07:51
→ wrobinso1 joined07:52
→ accesspoint joined07:53
→ Aquazi joined07:53
← wrobinson left07:54
← rednplueplaid96 left07:55
→ rednplueplaid96 joined07:56
← CryptoClub left07:58
← InfiniteZero_ left07:59
→ CryptoClub joined08:00
← [twisti] left08:01
← duderonomy left08:10
← ghost43 left08:10
→ ghost43 joined08:11
→ duderonomy joined08:12
← tiin57 left08:12
← accesspoint left08:14
→ stats4647 joined08:14
→ tiin57 joined08:15
← Ignacio left08:16
→ Ignacio joined08:16
← stats4647 left08:18
→ PCatinean joined08:18
PCatinean Hi all08:18
osse Hi one08:19
→ cd joined08:22
PCatinean XD08:23
PCatinean giggles08:23
PCatinean I was wondering if I could get an idea on how to solve an issue. My purpose is to check version numbers in a python project that has modules which contains __manifest__.py which contain dictionaries and have a "version" inside. I want to do a pipeline which compares the versions and if the version of the change is <= version being merged into then fail.08:25
→ lbracher joined08:25
PCatinean There are several ways to approach this, checking the diff directly or just getting the data from current branch and the brach merged into08:25
Not sure if there's some git functionality that would make this easier08:25
← Muimi left08:27
osse sounds like git grep to me08:28
→ krofe joined08:31
PCatinean git diff and grep pipe?08:32
← mat001 left08:33
→ amahl joined08:35
← sudoforge left08:37
osse I personally would get all the versions from master, the versions from the other branch, and then compare one by one08:38
PCatinean I need to compare just changed ones though08:40
And exclude other files which are not python modules like ".gitlab-ci.yaml" / README.MD etc08:40
← lunatic_wei left08:41
→ ARoxdale joined08:42
→ jimklimov1 joined08:42
← opal left08:42
osse of course08:42
← tiin57 left08:42
PCatinean and with current branch I can do a regular python import which makes it ezpz but for the other I would have to...checkout?08:43
osse git gives you git diff and git diff --name-only and git grep, all of which can be filtered by filenames.08:43
But you have to glue them together yourself. git does not give you git version-cmp08:43
PCatinean yeah that's the route I have so far, git diff --name-only08:43
oh there's an actual git grep nice, not piping grep on git diff08:44
← jimklimov left08:44
→ tiin57 joined08:46
← reen left08:48
→ reen joined08:48
→ PJBoy joined08:52
PCatinean this is way better, thank you for the great tip osse08:53
→ unluckyshrubbery joined08:53
→ royal_screwup21 joined08:55
→ accesspoint joined08:55
→ mozgalom joined08:56
→ gearnode joined08:57
→ calcul0n_ joined08:58
← kadobanana left09:00
← royal_screwup21 left09:00
← wallacer left09:01
← Shentino left09:03
→ wallacer joined09:04
→ Shentino joined09:04
← tiin57 left09:12
→ tiin57 joined09:16
← otisolsen70 left09:16
→ fstd_ joined09:18
→ jast joined09:18
← fstd left09:18
fstd_ → fstd09:18
→ Deknos joined09:21
Deknos hey, can someone with git 1.8 and git 2.20 work on the same repository without getting conflicts? or are there some backwards incompatibilities in the last years since 1.8? (i am aware, 1.8 is OLD)09:22
asking for a friend (no, really :D)09:22
→ Mattiaslndstrm joined09:22
← hnOsmium0001 left09:25
→ igopi joined09:27
→ solarjoe4 joined09:27
gitinfo set mode: +v09:27
→ Rashad joined09:28
igopi is the any project that actually rewriting git into python?09:29
*there09:29
← jetchisel left09:30
← Rashad left09:33
→ Rashad joined09:34
← lawt left09:36
osse Deknos: no problems09:38
Deknos: some defaults have changed and stuff. but the underlying data structures and all that is exactly the same09:38
← nutron left09:41
← cd left09:42
← tiin57 left09:42
→ royal_screwup21 joined09:43
← jaziz1 left09:44
→ tiin57 joined09:45
→ tm007 joined09:46
Deknos thank you! is this also valid for later versions than 2.20? like up to 2.30?09:47
← gigasu_shida left09:49
osse yep09:50
→ jetchisel joined09:52
→ TheMeltdown joined10:03
TheMeltdown → Guest9318910:03
← Guest84395 left10:04
← heftig left10:06
← elibrokeit left10:06
← rcoote left10:08
→ heftig joined10:11
→ elibrokeit joined10:11
→ opal joined10:11
← gearnode left10:12
← tiin57 left10:12
← Rashad left10:14
→ Rashad joined10:15
← opal left10:15
→ tiin57 joined10:15
→ opal joined10:16
← igopi left10:18
→ gearnode joined10:19
→ zebrag joined10:21
← royal_screwup21 left10:22
→ royal_screwup21 joined10:23
CustosLimen hi, are there alternatives to this: https://pre-commit.com/ ?10:24
← sozuba left10:24
→ sozuba joined10:27
osse you can always put any logic you want into .git/hooks/pre-commit10:27
→ igopi joined10:31
CustosLimen yes but everyone has to do it by themselves then10:32
pre-commit manages this for you10:32
osse Not completely automatically. Each person still has to run "pre-commit install", right?10:35
Anyway, I don't know of an alternative to this with this feature set10:35
CustosLimen yes you still have to do that10:40
← codebam left10:40
← tiin57 left10:42
→ computeiro joined10:43
← igopi left10:44
→ tiin57 joined10:45
← sozuba left10:46
→ sozuba joined10:48
nedbat CustosLimen: why did you want alternatives?10:49
CustosLimen don't want to use it if there is something better, not that there is something wrong with it, but it is all I know about10:49
← krofe left11:00
nedbat CustosLimen: i see11:00
← Rashad left11:10
← igemnace left11:12
← tiin57 left11:12
← wallacer left11:14
→ wallacer joined11:15
→ tiin57 joined11:16
← vorona1 left11:20
→ planetcall|work joined11:21
→ segurb joined11:23
← sozuba left11:23
← royal_screwup21 left11:23
→ royal_screwup21 joined11:23
segurb hi everyone, suddenly all my git repos, even the one with 5 files in it, is terribly slow on git pull, it takes like 5 minutes to check and have no changes at all, using bitbucket as remote, and it doesnt seem to have issues11:23
any way I can trace/debug this issue?11:23
← royal_screwup21 left11:28
osse segurb: GIT_TRACE=1 git pull11:29
sounds like suddenly your internet is slow11:29
← justanotheruser left11:32
← Guest20 left11:34
segurb osse: 200/20 internet aint slow at all11:34
osse That's not what I meant11:35
segurb osse: https://0x0.st/-c7X.jpg11:35
← hololeap left11:36
segurb its just stuck on that step11:36
→ hololeap joined11:36
osse try 'ssh -T [email@hidden.address]11:36
← meropyri left11:36
→ shush joined11:37
segurb with git trace or?11:37
because it just hangs11:37
osse not with git trace. just plain ssh11:37
ok, we're getting somewhere11:37
segurb just hangs11:37
osse time to be verbose AF11:37
ssh -vvvT [email@hidden.address]11:38
segurb https://0x0.st/-c7P.jpg11:38
→ meropyri joined11:38
← hethw left11:39
→ hethw joined11:39
osse ok then11:39
ssh -4T [email@hidden.address]11:39
segurb You can use git to connect to Bitbucket. Shell access is disabled11:40
before that it says logged in as <myusername>11:40
→ sweatsuit_ joined11:40
osse there we go11:40
Seems like something somewhere between your computer and bitbucket's server doesn't like IPv611:40
segurb nothing changed, except I havent used this PC for 10 days11:41
osse try this: GIT_SSH_COMMAND='ssh -4' git pull11:41
segurb worked like a charm before that11:41
← shush left11:41
segurb osse: that worked really fasty11:41
← sweatsuit left11:41
sweatsuit_ → sweatsuit11:41
segurb what is ssh -4?11:42
osse force IPv411:42
← tiin57 left11:42
→ drag0nius joined11:43
drag0nius how can i set username and password for a single repository only?11:43
i found git config credential.origin.username11:43
what would be equivalent for password?11:44
segurb osse: I assume I can set 4 as global for git untill I find whats up with 6?11:45
osse you can do it in .ssh/config for example11:46
→ tiin57 joined11:46
← FFY00_ left11:46
→ FFY00_ joined11:46
osse or git config --global core.sshCommand 'ssh -4'11:46
segurb its a global issue on my machine11:46
→ Muimi joined11:46
segurb osse:thank you very much11:47
atleast a temp fix so I can work11:47
and Il play with the 6 on weekend11:47
thanks !11:47
← led_belly left11:48
← zebrag left11:52
→ rcoote joined11:52
→ zebrag joined11:52
→ rsx joined11:55
→ led_belly joined11:56
→ royal_screwup21 joined11:57
← drag0nius left11:59
← royal_screwup21 left12:02
← nrh^ left12:04
→ rapiz joined12:05
← tm007 left12:05
← tiin57 left12:12
→ dsrt^ joined12:15
→ stats4647 joined12:15
← rcoote left12:15
→ tiin57 joined12:16
→ royal_screwup21 joined12:17
← stats4647 left12:19
→ cdown joined12:19
← rustyshackleford left12:20
→ viko joined12:23
→ jim87 joined12:24
jim87 Hello! Is there a way to clone a repository with all the tags, but without downloading all the changes? My use-case is to be able to get a list of tags of a repository, I don't need the sources themselves12:25
seanBE_ → seanBE12:27
→ kenanmarasli joined12:29
→ Anticom joined12:30
osse git clone --filter=... I think12:32
if you only need tags I would rather investigate some sort of API12:32
→ cdown_ joined12:33
→ bremner joined12:34
→ calcul0n__ joined12:34
jim87 osse: by API you mean vendor-specific git server APIs?12:34
bremner if you really just want the list of tags, you can use git-ls-remote12:34
jim87 bremmer: I was looking into it indeed12:35
← cdown left12:35
cdown_ → cdown12:35
← yawkat left12:35
← calcul0n_ left12:36
← snowkidimd left12:39
→ royal_screwup215 joined12:39
← tiin57 left12:42
→ rosco_y joined12:43
→ snowkidimd joined12:44
→ tiin57 joined12:45
← sgn left12:46
→ sgn joined12:47
→ rafasc joined12:47
→ CodeSlingerPaul joined12:48
→ xgfvc joined12:49
← snowkidimd left12:49
← sgn left12:49
→ AnAverageHuman joined12:50
→ sgn joined12:50
← royal_screwup21 left12:50
← royal_screwup215 left12:50
→ kalin_ joined12:51
→ royal_screwup21 joined12:51
← kenanmarasli left12:51
← zebrag left12:52
→ xlei joined12:52
→ zebrag joined12:52
← kalin___ left12:53
← cads2 left12:55
← royal_screwup21 left12:55
→ Betal joined12:56
→ royal_screwup21 joined12:58
osse jim87: yes12:59
but ls-remote might be even better. I keep forgetting it exists12:59
→ finalbeta joined13:07
→ Duikb00t joined13:07
Duikb00t Hello13:07
→ miczac joined13:08
Duikb00t My main project has a .gitignore file. In the project I can go to storage/app/public/.gitignore13:08
By default: storage is set to not-tracked in the main .gitignore file.13:08
But I want to track 1 folder in that public folder ( name: static ) So I added !static into the gitignore inside the public directory. But nothing seems to get tracked. Where am I going wrong?13:09
Thanks!13:09
osse the directory storage is ignored completely, so git won't even recurse into it to check for further .gitignore files13:10
lI think you need to do this in the outer gitignore:13:11
storage/*13:11
!storage/app13:11
Duikb00t Currently: /public/storage = in .gitignore ( the main one) (outer)13:11
Sorry13:11
Currently: /storage/*.key = in .gitignore ( the main one) (outer)13:12
→ bsanford joined13:12
← tiin57 left13:12
← daurnimator left13:12
← bsanford left13:13
→ tiin57 joined13:15
← Xenguy left13:16
osse !gitignore_whitedir13:16
gitinfo A global/directory .gitignore blacklist with a (potentially) subdirectory whitelist is not easy to specify in git. However, something like `printf '%s\n' '/*' '!/a/' '/a/*' '!/a/b/' '/a/b/*' '!/a/b/c/' > .gitignore` (ignore everything but a/b/c directory) or `printf '%s\n' '*' '!*/' '!*.txt' > .gitignore` (ignore everything but *.txt files) may do what you want.13:16
→ yawkat joined13:17
← xgfvc left13:21
→ royal_screwup217 joined13:25
→ xgfvc joined13:25
Duikb00t Not easy.13:25
I tought: !storage/app/public/static would fix it.13:25
osse you need to unignore app, then unignore public, then unignore static13:29
if the files are truly static you can simply: git add -f storage/app/public/static13:29
rafasc that unignores the 'static' file. But since storage is ignored, git never looks what is inside.13:29
osse but you need to use -f to add new files inside that dir13:29
→ mat001 joined13:30
Duikb00t I fixe dit13:30
yeah13:30
lmat I am rebasing and I see conflicts. I have been working to resolve them, but went too far. How can I get the original conflicts back?13:31
Duikb00t Graet13:32
Thanks osse and rafasc13:32
rafasc storage/*; !/storage/app; /storage/app/*; !/storage/app/public; /storage/app/public/*; !/storage/app/public/static;13:32
you would need something like the above, but probably a good candidate for add -f as osse mentioned.13:32
→ muhaha joined13:32
← rapiz left13:33
rafasc lmat: probably: git checkout -m <path-to-file>;13:34
→ shush joined13:37
→ daurnimator joined13:38
← tiin57 left13:42
← shush left13:42
planetcall|work Anyone familiar with `gitversion` ? I am curious how does it maintain state?13:44
rafasc what is gitversion?13:44
this: https://github.com/GitTools/GitVersion?13:45
→ tiin57 joined13:46
→ hussam joined13:46
planetcall|work rafasc, yes13:49
rafasc I think it is the first time I stumble over this. And just reading the intro, there are things that I already dislike about it. :D13:51
E.g. don't tell me how I should name my own remotes: https://gitversion.net/docs/learn/git-setup13:52
← zebrag left13:52
→ zebrag joined13:52
→ budonyc joined13:56
_flow_ → flow13:56
lmat rafasc: Ahh, thanks!13:57
planetcall|work I think it elegantly solves the issue of versioning the binaries13:57
→ vimIsJustBest joined13:59
rafasc planetcall|work: from the looks of it, seems like they keep state on their own config files?13:59
planetcall|work rafasc, not really. It actually is deriving the state from commits and tags somehow14:00
it does have a config file in yml format which guides us which strategy to use along with other defaults.14:01
rafasc then it probably does it on the fly, like git describe would do it.14:01
planetcall|work it seems so. I don't have a complete grasp on it. I guess I will have to dig deeper.14:01
rafasc I looked briefly at: git log --no-walk --tags -p; and so next-version:14:01
But seems that was in a test file.14:02
← royal_screwup21 left14:02
→ royal_screwup21 joined14:03
← thansen left14:04
→ madewokherd joined14:05
← hethw left14:05
→ vorona1 joined14:05
← axisys left14:06
rafasc "the tag is used and build metadata (excluding commit count) is added"14:06
added where?14:06
→ thansen joined14:07
← tiin57 left14:12
← CryptoClub left14:15
→ tiin57 joined14:15
→ CryptoClub joined14:17
→ rcoote joined14:18
→ rapiz joined14:21
← royal_screwup217 left14:22
← royal_screwup21 left14:22
→ royal_screwup21 joined14:23
← solarjoe4 left14:24
← rapiz left14:25
→ DevAntoi_ joined14:26
← elichai2 left14:27
← royal_screwup21 left14:27
← DevAntoine left14:29
planetcall|work to the version14:33
→ thiago joined14:37
→ snowkidi_ joined14:38
→ arcatech joined14:39
→ rustyshackleford joined14:40
→ royal_screwup21 joined14:41
← thansen left14:41
→ thansen joined14:41
← tiin57 left14:42
→ tiin57 joined14:45
→ stats4647 joined14:45
← zebrag left14:52
→ zebrag joined14:52
← royal_screwup21 left14:53
→ royal_screwup21 joined14:53
← royal_screwup21 left14:58
→ RaIn joined14:58
→ henri joined14:59
→ mowcat joined15:00
← RaIn left15:00
→ noobcoder joined15:01
← noobcoder left15:01
→ noob-RAIN joined15:02
← snowkidi_ left15:03
→ snowkidimd joined15:03
← noob-RAIN left15:03
→ airgap joined15:06
→ hnOsmium0001 joined15:08
→ maroloccio joined15:11
← Serus left15:11
← tiin57 left15:12
← rcoote left15:12
← austin987 left15:15
→ tiin57 joined15:16
← arcatech left15:16
→ Serus joined15:17
← xrisovalantis left15:21
← miczac left15:22
← mingdao left15:22
→ miczac joined15:23
→ kadobanana joined15:25
→ austin987 joined15:29
cYmen when I create a repo with init --bare --shared=all I get sharedrepository=2 if I git config core.sharedRepository all I get sharedRepository=all ...is this a bug or what is going on here?15:30
→ royal_screwup21 joined15:31
← haymawk left15:32
→ haymawk joined15:34
← royal_screwup21 left15:35
→ mindCrime_ joined15:36
→ shush joined15:38
→ causasui joined15:38
→ sgn1 joined15:39
← miczac left15:39
← sgn left15:40
rafasc I get sharedrepository=2, but it should be fine.15:41
← tiin57 left15:42
→ howdoi joined15:43
rafasc https://github.com/git/git/blob/a0dda6023ed82b927fa205c474654699a5b07a82/t/t1301-shared-repo.sh#L48-L5315:43
and there's a test for it, confirming 2=all.15:43
← shush left15:44
→ shush joined15:44
rafasc But seems this isn't documented anywhere, might be worth looking the mailing list if someone already asked about it and if not, maybe ask.15:45
→ tiin57 joined15:45
→ arcatech joined15:46
→ LuxuryMode joined15:46
→ miczac joined15:47
→ sudoforge joined15:48
→ linuxguru joined15:48
← stats4647 left15:49
← sebas057 left15:51
← sagagnon40 left15:51
← zebrag left15:52
← mantisfan left15:52
← mr_francois3 left15:52
← pcarphin left15:52
→ zebrag joined15:52
→ mantisfan joined15:52
→ mr_francois3 joined15:54
→ sebas057 joined15:54
→ pcarphin joined15:54
→ sagagnon40 joined15:54
→ olx69 joined15:57
← long4mud left15:59
→ Newami joined15:59
← mindCrime_ left15:59
← ghoti left16:03
→ axisys joined16:06
→ durham joined16:06
→ nutron joined16:07
← miczac left16:09
← Newami left16:11
→ ubuntu joined16:12
ubuntu → Guest5634616:12
← tiin57 left16:12
← sebas057 left16:12
← mr_francois3 left16:12
← pcarphin left16:12
← sagagnon40 left16:12
→ Newami joined16:13
Guest56346 Hi, is it possible for git log -- somepath/ to list all logs under some/path or any symlink it may have which is behind somepath/ ?16:13
← muhaha left16:15
→ nyxyn joined16:15
maroloccio On a flaky network I am having trouble cloning a bit repo. Is there a "resumable" clone option?16:16
→ tiin57 joined16:16
rafasc Guest56346: you want the pattern to match somepath/* where * are symbolic links?16:16
Guest56346 somepath/ may contains symlinks to other directories before somepath/16:17
rafasc maroloccio: clone with --depth=1; then progressively fetch it with git fetch --deepen=N;16:17
followed by a final git fetch --unshallow16:18
maroloccio trying.. thanks rafasc16:18
← durham left16:19
→ durham joined16:19
→ mellernoia joined16:21
← henri left16:21
→ ghoti joined16:21
maroloccio rafasc: working beatifully. thank you *very much*! :)16:23
← Haohmaru left16:23
→ Guest20 joined16:25
→ rcoote joined16:26
← vimIsJustBest left16:28
→ Psyh joined16:29
← nyxyn left16:30
← Psyh left16:31
← maroloccio left16:31
← Newami left16:31
→ maroloccio joined16:35
→ modonnell joined16:37
→ xissburg joined16:38
← KnoP left16:41
← tiin57 left16:42
← modonnell left16:43
← Duikb00t left16:44
→ emanuele6 joined16:45
→ tiin57 joined16:46
← planetcall|work left16:47
← maroloccio left16:47
← Anticom left16:50
→ Duikb00t joined16:51
→ long4mud joined16:51
← rfuentess left16:52
→ mokulus joined16:52
→ stats4647 joined16:53
← Duikb00t left16:55
← mantisfan left16:56
→ mantisfan joined16:56
← wrobinso1 left17:01
→ wrobinson joined17:01
→ Hatmad joined17:01
← PCatinean left17:02
← gordonfish left17:03
← Hatmad left17:04
← irrgit left17:06
← Muimi left17:06
← snowkidimd left17:06
→ fulsiram joined17:07
astronavt is there a way to tell git-read-tree to just overwrite any existing files?17:08
i'm trying to synchronize a subdirectory of a remote repo (that i do not control) with some files in my project's repo17:08
but if the files already exist, git-read-tree complains and quits. the docs just say that it will refuse to overwrite anything...17:09
→ julius_ joined17:09
julius_ gi17:09
→ irrgit joined17:09
julius_ hi17:09
astronavt if there's no way to do it, is there a workaround?17:09
→ snowkidimd joined17:09
astronavt i don't want to have to make 2 commits if i can avoid it17:09
julius_ i typed: git rm -r --cached syncserver/.* and i got a list of my complete git repository, somehow the * was not used inside the syncserver/ directory but on the root of my git17:10
how do i undo that action?17:10
← ghoti left17:10
astronavt julius_: if you haven't committed yet, you can do `git reset` to undo staged changes17:11
julius_ lets try17:11
ok apparently that worked, thanks17:11
← mokulus left17:12
← tiin57 left17:12
→ maroloccio joined17:13
rafasc julius_: maybe you used an extra space? like syncserver/. * ?17:13
because that shouldn't apply to the whole repo.17:14
→ Duikb00t joined17:14
← maroloccio left17:14
julius_ i may have done something more, git pull tells me that "the following untracked working tree files would be overwritten by merge" python/icap/ but i already git: git add python && git commit -am 'update'17:14
rafasc, yes with a extra space that would make sense...but there wasnt one17:15
rafasc astronavt: are you using -m ? Or one the similar options.17:15
astronavt rafasc: i am using `git read-tree -u --prefix=myrepo/foo-data/external remotes/something/master:assets/data`17:16
→ tiin57 joined17:16
→ royal_screwup21 joined17:17
astronavt if i use -m instead, will that solve the problem?17:17
rafasc astronavt: depends on the definition of "problem". I think either --reset or -m, so you get the chance to solve conflicts.17:18
astronavt it looks like --prefix and -m are incompatible17:18
rafasc probably smart to do a quick backup, if you have pending files.17:18
astronavt if there is a way to just "dump a specific tree from a remote into my work tree" i'd be happy to use that17:19
git-read-tree was the only way i knew how17:19
← rsx left17:19
astronavt the full problem statement would be: "dump a specific tree from a remote into a my work tree, with a prefix that i choose"17:20
the git repo happens to be hosted on github so as a workaround i can just use the "raw" github api17:20
→ iffraff_ joined17:20
astronavt i admittedly dont understand what an "unmerged entry" is in this context, as per the docs17:21
rafasc it essentially means conflicts.17:22
← Xiti left17:23
← hussam left17:23
→ Xiti joined17:23
→ justanotheruser joined17:24
→ hussam joined17:24
rafasc astronavt: try this:17:25
git read-tree --index-output=myindex <treeid>; GIT_INDEX_FILE=myindex git checkout-index --prefix=something/ -a17:26
← mozgalom left17:27
← guerra left17:27
rafasc you may want to need to add --force to overwrite existing files.17:27
backup and try I guess :)17:27
astronavt so this writes to a different index file, which you then checkout from?17:27
interesting17:27
→ fax joined17:28
rafasc astronavt: that's the idea, yes.17:28
← mns left17:29
→ JaseW joined17:29
astronavt alright, i ended up manually updating this time but i will try it. thanks17:31
→ hidedagger joined17:32
← fulsiram left17:34
→ fulsiram joined17:34
astronavt rafasc: do you know if i can read multiple trees to the same index file?17:34
e.g. git read-tree --index-output=myindex tree1 ; git read-tree --index-output=myindex tree217:35
→ tang^ joined17:35
astronavt or does read-tree let me write: git read-tree --index-output=myindex tree1 tree217:35
rafasc you should be able to, you would have to use a different --prefix for each subtree.17:35
← Ignacio left17:36
rafasc like git read-tree --prefix=subtree1/ tree1; git read-tree --prefix=subtree2/ tree2;17:36
astronavt in my case the subtrees go to the same prefix17:36
rafasc you want to overlay them?17:37
astronavt no, not that fancy. in the remote they are something like data/a, data/b. in my repo they are something like stuff/testdata/a and stuff/testdata/b17:37
oh i think i see17:38
i'll experiment17:38
rafasc git read-tree --prefix stuff/testdata/a <treeid of data/a>; same for b. Then you can just checkout from the index.17:39
git rev-parse <branch>:<path>; should get you the treeid you want.17:40
can't remember if read-tree resolves things like master:Documentation as tree id.17:41
← tiin57 left17:42
→ vicfred joined17:42
→ vicfred_ joined17:44
← amerlyq left17:45
→ mns joined17:45
→ tiin57 joined17:46
→ Tweth-U-PDS joined17:46
← royal_screwup21 left17:46
→ royal_screwup21 joined17:47
← vicfred left17:47
→ krofe joined17:48
→ mindCrime_ joined17:50
← Guest20 left17:50
← zebrag left17:52
← royal_screwup21 left17:52
→ zebrag joined17:52
→ gast0n joined17:54
← mns left17:57
astronavt rafasc: it does resolve names like that, i had it in my previous script (which is the one that had the "overlap" problem)17:57
rafasc you could probably do a "rm --cached -r <path>" before that, to empty whatever is there first.17:59
→ klonn joined18:00
rafasc could be easier than dealing with temporary index files.18:00
← jab416171 left18:00
← vicfred_ left18:00
→ Narrat joined18:04
astronavt oh thats a good poitn too18:05
point*18:05
thats a lot easier. i can move the files to an untracked backup location first18:05
← lbracher left18:06
rafasc or depending how you access the remote, doing a sparse clone and just move the files manually.18:07
many ways to go about it.18:08
→ rond_ joined18:08
← vorona1 left18:10
← tiin57 left18:12
→ ghoti joined18:12
← Tweth-U-PDS left18:14
→ Tweth-U-PDS joined18:14
← jetchisel left18:15
→ tiin57 joined18:16
→ madewokherd` joined18:17
← madewokherd left18:18
← sgn1 left18:18
→ royal_screwup21 joined18:20
→ lukedashjr joined18:23
← snowkidimd left18:24
← royal_screwup21 left18:24
← luke-jr left18:25
→ snowkidi_ joined18:26
← CryptoClub left18:27
→ gigasu_shida joined18:27
lukedashjr → luke-jr18:28
→ CryptoClub joined18:30
→ royal_screwup21 joined18:35
← tiin57 left18:42
→ drag0nius joined18:42
← drag0nius left18:42
→ grym joined18:44
← DevAntoi_ left18:45
grym i like using `git shortlog previous-tag..HEAD` to make changelog drafts, but I often want the entire commit message, not just the first line. how might i go about getting that format?18:45
s/changelog/release notes18:46
ikke --format "%b"18:46
→ DevAntoine joined18:46
← royal_screwup21 left18:46
→ tiin57 joined18:46
→ royal_screwup21 joined18:46
grym "fatal: ambiguous argument '%b': unknown revision or path not in the working tree."18:47
shortlog doesn't even have --format; log does, but produces that output18:48
ikke --format="%b"?18:48
man git short-log18:48
gitinfo the git manpage is available at https://gitirc.eu/git.html18:48
ikke man git shortlog18:48
gitinfo the git-shortlog manpage is available at https://gitirc.eu/git-shortlog.html18:48
ikke --format is listed there18:48
← gast0n left18:49
← c4017w_ left18:50
grym ok, that's _just_ the body... but not the first line?18:50
ikke %B18:51
should include both18:51
or %s %b18:51
← DevAntoine left18:51
ikke but you need 2 newlines in between18:51
%s%n%n%b18:51
grym ah ha!18:51
yes, perfect. cheers18:51
← royal_screwup21 left18:51
← zebrag left18:52
→ zebrag joined18:52
→ c4017w joined18:53
← grym left18:54
← hidedagger left18:55
← c4017w left18:58
← yoh left19:00
← thinkpad left19:02
← rcoote left19:04
→ thinkpad joined19:05
→ vicfred joined19:07
→ stdedos joined19:10
← kiweun left19:11
← tiin57 left19:12
→ tiin57 joined19:16
← riksteri left19:16
← stdedos left19:16
→ yoh joined19:17
← julius_ left19:22
→ royal_screwup21 joined19:22
← YellowSu1 left19:23
→ bsanford joined19:23
← olx69 left19:25
← bsanford left19:26
← royal_screwup21 left19:27
→ mns joined19:33
← cdown left19:34
→ gast0n joined19:36
← Duikb00t left19:36
→ YellowSuB joined19:37
← tiin57 left19:42
→ tiin57 joined19:46
→ c4017w joined19:47
→ Lord_of_Life_ joined19:48
→ gordonfish joined19:49
← fulsiram left19:49
← Lord_of_Life left19:50
Lord_of_Life_ → Lord_of_Life19:50
→ jetchisel joined19:54
← sord937 left19:58
← accesspoint left20:01
→ accesspoint joined20:02
siers → ij20:04
← justanotheruser left20:05
pikapika → pp20:06
pp → peepeee20:07
→ Duikb00t joined20:08
relipse how do I rm all untracked files ending in .rej, see paste: https://dpaste.org/jzkc20:08
peepeee → pikapika20:10
nkuttler relipse: do you have tracked files ending in rej?20:10
relipse nkuttler: no20:10
→ alfredb joined20:11
nkuttler just find . -name "*rej" -delete ?20:11
run without -delete first20:11
← tiin57 left20:12
← Duikb00t left20:12
→ royal_screwup21 joined20:13
canton7 or you could clean everything in src/local/portal20:13
→ nyxyn joined20:14
→ fulsiram joined20:15
← krofe left20:15
→ tiin57 joined20:16
→ c4017w_ joined20:16
← finn_elija left20:18
→ finn_elija joined20:18
→ m0viefreak joined20:19
← c4017w left20:20
→ justanotheruser joined20:20
← cim209 left20:23
→ workfrosty joined20:24
→ Natrick3 joined20:24
→ Duckle0 joined20:24
→ cim209 joined20:24
→ yourfate_ joined20:24
→ kshockey joined20:25
← kshockey left20:25
→ kshockey joined20:25
← Natrick left20:25
→ saidelike_ joined20:25
Natrick3 → Natrick20:25
→ moviuro_ joined20:25
← staafl_inv0022 left20:26
→ staafl_inv0022 joined20:26
← kyshoc left20:26
← saidelike left20:26
← marshmallow left20:26
← reda left20:26
← jetchisel left20:26
← segnior left20:26
← daidoji left20:26
← yourfate left20:26
← L1Cafe left20:26
← Duckle left20:26
← ddevault left20:26
← moviuro left20:26
Duckle0 → Duckle20:26
yourfate_ → yourfate20:26
→ ddevault_ joined20:27
→ marshmallow joined20:27
→ daidoji joined20:27
→ jetchisel joined20:27
→ segnior joined20:28
ddevault_ → ddevault20:28
→ finnb joined20:29
→ L1Cafe joined20:29
← led_belly left20:29
→ sgn joined20:31
→ led_belly joined20:32
→ reda joined20:32
← alfredb left20:32
→ jaziz1 joined20:38
← mindCrime_ left20:38
finnb Just playing around with git notes. Is it possible to have a note that will be available to read from any future commit. Let's say I'm on commit A, I then add commit B, then C. Is it possible to easily read the note from commit A once I'm on commit C? Like a silly TODO list, `git notes show --ref todo`20:41
← tiin57 left20:42
finnb In that example, I created the todo note on commit A. I haven't modified any notes on B or C20:42
← klonn left20:42
← jimmiehansson left20:43
← royal_screwup21 left20:44
→ royal_screwup21 joined20:44
→ tiin57 joined20:45
→ dpl_ joined20:49
→ DevAntoine joined20:49
← royal_screwup21 left20:49
rafasc finnb: No.20:51
→ mr_francois3 joined20:51
→ pcarphin joined20:51
→ sebas05 joined20:51
→ sagagnon40 joined20:51
rafasc finnb: For a todo list, maybe branch descriptions would be better.20:51
finnb Thought so :P Wanted to check if there was some weird thing I could do if I wanted to do that20:51
← zebrag left20:52
→ zebrag joined20:52
← dpl_ left20:53
finnb branch descriptions?20:53
rafasc git branch --edit-description;20:53
← DevAntoine left20:53
finnb man git branch20:53
gitinfo the git-branch manpage is available at https://gitirc.eu/git-branch.html20:53
finnb :D20:53
→ hqdruxn08_ joined20:54
← liefer left20:55
finnb Is there a way to attach general metadata to git repositories?20:55
← LuxuryMode left20:56
← rond_ left20:56
rafasc depends on what you mean by that.20:56
e.g. there is .git/description where you can store a description of the repo.20:57
← stats4647 left20:57
← hqdruxn08 left20:57
→ stats4647 joined20:58
← gpanders left20:59
rafasc you can push refs to non-standard namespaces... like git notes itself 'refs/notes/'..20:59
branches can have description, but they are stored along with configuration, which makes it harder to share.20:59
finnb I'm writing a small tool which will tell you which other repositories your repository depends on. I originally considered just storing this information in a file like ".other-repos" but thought I'd play around with other methods.21:00
rafasc if you want to stick with notes, you can always copy the commits from one to another.21:01
imMute "depends on" ?? like submodules?21:01
finnb not submodules21:01
It's just for fun21:01
→ Duikb00t joined21:02
rafasc orhpan branches are also an option: git switch --orphan21:02
you get a complete unrelated branch where you can store whatever metadata you want.21:02
finnb Just a way of storing information like, "my repo uses github.com/x github.com/y github.com/z " etc.21:02
← stats4647 left21:03
rafasc this will sound like I am mocking you, but I am not: README.md21:04
finnb no... no I get you21:04
I'm writing a tool for fun which will do some other things with the metadata21:04
← mellernoia left21:04
finnb At that point, I'd store it in a file like ".my-fun-tool"21:05
Betal don't builders usually have this info? like go modules, java gradle/maven, python requeriments ... ?21:05
finnb bazel21:05
yeah21:05
← magpi left21:05
rafasc I think I would go with an an orphan branch personally, if you're not using a dependency manager.21:05
it's super easy to store information with git config.21:05
finnb Can you push orphan branches?21:05
← jim87 left21:06
→ jim87_ joined21:06
finnb I'd like the metadata to be available to anyone who clones the repo21:06
→ magpi joined21:06
finnb > you can push refs to non-standard namespaces... like git notes itself 'refs/notes/'..21:06
← Duikb00t left21:07
finnb But would that still be tied to a specific commit?21:07
I guess I can test this now quickly :P21:07
rafasc you can push orphan branches just as regular branches. They work the same.21:08
For non-standard namespaces, you'll have to clone/push manually to the namespace.21:08
finnb Could you add it to your git config....?21:09
to push/fetch each time or do you literally have to write "git push refs/notes/weird-thing-im-doing"21:09
rafasc you could add it to the config, but it is something you'll have to do manually.21:09
finnb I see21:10
→ DevAntoine joined21:10
rafasc remember that configs are not shared across clones.21:10
finnb I noticed with notes you have to add something to ensure you fetch them anyway21:10
→ royal_screwup21 joined21:10
rafasc you need to manually configure the fetch spec, yeah.21:11
→ liefer joined21:11
rafasc I have used git config to store metadata, e.g. with --file you could do: git config --file=mysuperduper --add dependency.url github/a21:11
→ kiweun joined21:11
← tiin57 left21:12
finnb interesting21:12
rafasc and fetch it the same way you fetch things from the regular gitconfig, but passing --file.21:12
but for anything more complex than that I think a proper dep maintenance tool would be better.21:14
finnb I agree a proper tool is better, I'm just playing around :)21:14
git notes --ref refs/notes/weird add -m "hello there" >> so that's how you'd use a non-std namespace?21:14
← DevAntoine left21:14
rafasc you can use namespaces with notes, but they're still per commit.21:15
→ tiin57 joined21:15
rafasc when I mentioned non-std namespace was for branches.21:15
Like pushing a branch into refs/metadata21:15
finnb What's the syntax for that?21:15
I tried and failed21:15
rafasc git push <local>:<remote>21:16
git push master:refs/metadata21:16
finnb ahh21:16
← kiweun left21:16
rafasc 'master' could be anything.21:17
finnb Right so it's a special branch for metadata21:19
choices21:19
well, I learnt something new today21:19
so thankyou :)21:19
rafasc no problem. :) Messing with git is fun.21:20
I did a git rss reader for a podcast I liked.. So yeah tell me about weird-tools21:21
← accesspoint left21:22
← nyxyn left21:23
← amahl left21:23
← jim87_ left21:25
finnb https://github.com/git-depend/git-depend21:28
That's what I was playing around with21:28
← crose left21:28
rafasc let me see If I understand this correctly.21:29
finnb Excuse the readme21:29
rafasc What you're trying to do is to syncronize two repos that have the same structure, but are different.21:29
?21:29
finnb Trying to commit some feature to two repositories atomically21:29
rafasc that's a hard problem.21:30
finnb s/hard/interesting21:30
magpi → Guest2731021:30
→ magpi_ joined21:30
← Guest27310 left21:30
magpi_ → magpi21:30
→ DevAntoine joined21:31
→ stats4647 joined21:31
rafasc push has --atomic, so you could have it like a monorepo.21:31
git has something called namespaces (it's a different one from the one we recently mentioned).21:32
git help namespaces21:32
finnb man git namespaces21:32
gitinfo the git manpage is available at https://gitirc.eu/git.html21:32
j416 finnb: what about if there's a merge conflict? would that abort everything?21:32
rafasc man gitnamespaces21:32
gitinfo the gitnamespaces manpage is available at https://gitirc.eu/gitnamespaces.html21:32
rafasc I messed with this once, can't remember what for.21:33
finnb j416 - not solving that just yet but that's a hard problem. At the moment, yes, abort everything.21:33
j416 I see.21:33
→ blyat-73 joined21:34
j416 don't forget that you can have a conflict even though git thinks everything went fine.21:34
finnb j416, example?21:34
rafasc and it would only work if you had control of the server.21:34
finnb The idea is to "lock" the repositories with a git note. If you can write a git note to every repo21:34
j416 the merge may have worked syntactically, but not semantically21:34
e.g., a function was renamed in one branch and a function call to the old name was added elsewhere in the other branch21:35
finnb The idea is to "lock" the repositories with a git note. If you can write a git note to every repo, then you're ok to merge.21:35
That's my basic idea which you can probably now tell me why that isn't going to work :P21:35
← DevAntoine left21:35
finnb j416 - should CI normally catch this?21:36
I'm trying to palm off as much as I can to CI21:36
j416 depends on your language and test coverage I suppose21:36
I wouldn'21:36
I wouldn't bet on it21:36
I always check my merges manually21:36
finnb I see what you're saying21:37
rafasc this would require server side modification or at least hooks, that denies pushes if you don't have the lock, whatever it is.21:37
j416 you probably got this already but I'll say it anyway; this feels like a lot of effort for something that would be better handled by a proper dependency manager or a monorepo, or even submodules21:37
finnb Again, just for fun :)21:38
j416 have fun, I guess21:38
← royal_screwup21 left21:38
rafasc another thing you can toy with is signed pushes.21:38
→ royal_screwup21 joined21:38
finnb Signed pushes?21:38
rafasc git can sign the push transaction.21:38
and reject everything that isn't signed by the person holding the lock.21:39
← arcatech left21:39
finnb That would require a server modification right?21:39
← tiin57 left21:42
finnb > this would require server side modification or at least hooks, that denies pushes if you don't have the lock, whatever it is.21:42
If you wanted to stop bad actors or silly people, yes. But if the client could detect if it got the lock or not, then I believe it will be OK and you wouldn't have to modify the server.21:42
rafasc I think you can do it with server-side hooks. But I'm not sure.21:42
finnb I'll add it to be reading list :)21:42
*my reading21:42
← gearnode left21:43
← royal_screwup21 left21:43
rafasc git lfs has a mechanism for file locking, not sure if it would help you.21:43
finnb writing it down21:44
It would be good to at least understand how they did that21:45
→ tiin57 joined21:45
← is_null left21:46
→ nyxyn joined21:46
→ cliluw joined21:47
finnb I like the idea of signing the pushes though. Saying this user obtained the lock and now they can push would be cool.21:48
← skered left21:48
finnb Thanks for the advice anyway! Going afk now :)21:48
rafasc same o/21:48
ARoxdale It is normal when using "git subtree" that "git log" doesn't follow files which have been added in a subtree past the tree merge commit?21:50
→ jim87_ joined21:51
→ DevAntoine joined21:51
← sgn left21:52
→ sgn joined21:52
→ xgfvc_ joined21:52
← LetoThe2nd left21:53
ARoxdale This article suggest a manual merge instead of using subtree. https://saintgimp.org/2013/01/22/merging-two-git-repositories-into-one-repository-without-losing-file-history/21:53
I will try that instead21:53
← xgfvc_ left21:54
← xgfvc left21:55
→ lawt joined21:55
→ gpanders joined21:56
→ Duikb00t joined21:56
← DevAntoine left21:56
← theoceaniscool left21:57
← gpanders left22:00
← Duikb00t left22:00
← chachasmooth left22:02
→ chachasmooth joined22:03
→ arcatech joined22:04
→ xgfvc joined22:06
← fulsiram left22:06
→ fulsiram joined22:09
cYmen why does a shared repo prevent non-fast-forward?22:11
← tiin57 left22:12
→ DevAntoine joined22:12
← tkob left22:13
← nyxyn left22:13
→ tkob1 joined22:13
→ royal_screwup21 joined22:15
→ tiin57 joined22:16
← DevAntoine left22:17
→ krofe joined22:19
← Narrat left22:19
← royal_screwup21 left22:20
→ chupawan_ joined22:20
← chupawan left22:24
→ royal_screwup21 joined22:24
→ Kaisyu joined22:25
← CodeSlingerPaul left22:28
→ theoceaniscool joined22:29
nedbat cYmen: that's likely a server-specific policy, and possibly one you can change. non-fast-forward updates might be difficult for collaborators to use, and if they've built work on commits you've abandoned, the merge could be very difficult.22:29
→ SleePy joined22:31
← indy left22:31
→ DevAntoine joined22:33
← sudoforge left22:34
→ skered joined22:37
← fax left22:37
← Tk55 left22:38
← DevAntoine left22:38
← CryptoClub left22:40
← tiin57 left22:42
→ Tk55 joined22:43
→ CryptoClub joined22:43
→ indy joined22:43
→ tiin57 joined22:45
cYmen right so it is totally possible, I guess22:47
but if you set a repo to shared it will automatically be set22:47
and I want to understand the issues you re describing22:47
but why is it different from remote work?22:47
i assumed it would have to be something weird that involves file permissions22:47
or maybe you can just explain your concerns in a bit more detail...?22:48
→ Duikb00t joined22:50
← iffraff_ left22:50
← jim87_ left22:50
← zebrag left22:52
→ zebrag joined22:52
→ DevAntoine joined22:54
← Duikb00t left22:54
→ twiclo joined22:56
← john_johnk left22:56
twiclo I did a git reset --hard before committing and lost all my work. I'm pretty sure the answer is no but I figured I should ask if that can be done undone22:56
← DevAntoine left22:59
→ cads2 joined23:03
→ lightstalker joined23:04
← royal_screwup21 left23:04
→ royal_screwup21 joined23:05
→ Xenguy joined23:05
← sgn left23:08
→ sgn joined23:09
← royal_screwup21 left23:10
← sgn left23:10
→ sgn joined23:11
← tiin57 left23:12
rafasc twiclo: you said before committing, did you ever run git add?23:12
→ kiweun joined23:12
rafasc If you staged the changes, it can be retrieved. If otherwise, I'm sorry for your loss.23:12
← bolovanos left23:14
→ DevAntoine joined23:15
→ tiin57 joined23:16
← kiweun left23:17
→ jess joined23:18
→ gearnode joined23:19
← workfrosty left23:19
← DevAntoine left23:20
→ djapo joined23:23
djapo how do I print what remote all branches are tracking and how do I change them?23:23
→ mindCrime_ joined23:25
rafasc djapo: git branch -vv shows them inside []23:26
← xgfvc left23:27
rafasc you can change them with git branch --set-upstream-to=<something>23:27
djapo thanks23:28
→ sudoforge joined23:28
rafasc git for-each-ref refs/heads --format="%(refname)%09%(color:red)%(upstream:short)"23:30
djapo: you can play with formatting if you want it in a neat format.23:30
→ zrhoffman joined23:31
← djapo left23:32
ARoxdale For the record, the method recommended in this link did preserve file history, but only when git log --follow is used. https://saintgimp.org/2013/01/22/merging-two-git-repositories-into-one-repository-without-losing-file-history/23:36
→ DevAntoine joined23:36
→ cdown joined23:36
ARoxdale I think this is less related to the merge than 'git log' having poor performance in tracking moved/renamed files23:36
→ royal_screwup21 joined23:39
← DevAntoine left23:40
← tiin57 left23:42
← royal_screwup21 left23:44
← LionsMane left23:44
→ Duikb00t joined23:44
← blyat-73 left23:45
→ tiin57 joined23:46
→ royal_screwup21 joined23:46
← Duikb00t left23:49
← tejr left23:51
← sgn left23:53
→ sgn joined23:53
→ DevAntoine joined23:57
→ djapo joined23:58
← computeiro left23:58

Logs Search ←Prev date Next date→ Channels Documentation