IRCloggy #git 2020-04-15

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.

2020-04-15

← fission6 left00:00
← sidetrack left00:00
← g00s left00:00
→ przemoc joined00:03
refried_ nedbat: it looks like you'd have to `fetch` it from the server, and that many servers are configured to disallow asking for arbitrary commits, even known ones00:03
Eugene No, the git protocol does not permit you to fetch arbitrary objects. You can only fetch refs, and then hope that the object is included in the packfile you receive00:08
← Paradisee left00:09
Eugene I'm sure you could cobble together `git clone <URL> --mirror` and `git checkout <treeish>` into a one-liner without too much inventiveness ;-)00:09
← tiin57 left00:14
← m0viefreak left00:14
← perrier-jouet left00:16
← BlessJah left00:17
→ BlessJah joined00:17
← cdown left00:17
→ tiin57 joined00:17
→ cdown joined00:17
← dedowsdi left00:19
→ sidetrack joined00:20
← Wetmelon left00:21
← shush left00:21
← relipse left00:21
← expert975 left00:22
→ shush joined00:24
→ dedowsdi joined00:24
→ gnomadik joined00:24
← gnomadik left00:24
→ gnomadik joined00:25
← sidetrack left00:25
→ Wetmelon joined00:26
← cdown left00:30
→ cdown joined00:32
← shush left00:32
→ shush joined00:33
→ rjcarr joined00:35
← cdown left00:37
→ cdown joined00:37
rjcarr I had 5 modified files: A B C D E and wanted to commit only D and E. I staged them, but when I did a commit, I accidentally did a -am switch so all files were commited. Now, I’d like to backout those changes, but still leave the changes in my local workspace. I think I need to do something with “checkout”, but not sure. And if it matters, the changes were already pushed to master. Thanks!00:38
← cdown left00:39
→ cdown joined00:39
← cdown left00:42
← tiin57 left00:44
← troulouliou_div2 left00:44
← BPL left00:46
← shush left00:46
→ shush joined00:47
→ tiin57 joined00:47
→ cdown joined00:49
← rjcarr left00:49
→ sidetrack joined00:51
← yawkat left00:53
→ YuGiOhJCJ joined00:54
← shush left00:55
→ troulouliou_div2 joined00:57
Eugene !undo01:00
gitinfo [!fixup] So you lost or broke something or need to otherwise find, fix, or delete commits? Look at https://sukima.github.io/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!01:00
← gnomadik left01:02
→ gnomadik joined01:03
← cdown left01:03
→ cdown joined01:04
← Wetmelon left01:05
Hello71 they left fifteen minutes ago01:05
← AtumT left01:05
← sidetrack left01:05
← Sasazuka left01:06
← clime left01:06
rafasc maybe !logs01:07
gitinfo [!irclog] Public logs of #git are kept at: https://gitirc.eu/log01:07
← gnomadik left01:07
← emmanuelux left01:08
→ yawkat joined01:08
← topdownjimmy left01:11
← thiago left01:11
→ topdownjimmy joined01:12
← refried_ left01:12
→ thiago joined01:13
← SpeakerToMeat left01:14
← tiin57 left01:14
← ntqz left01:15
← akem left01:18
→ tiin57 joined01:18
→ inkbottle joined01:18
← zebrag left01:19
← pthomas left01:20
← dartmouthed left01:20
← xco left01:21
→ SpeakerToMeat joined01:23
← NODE left01:23
Hello71 someone who only waited 11 minutes for an answer is gonna go read the logs?01:24
← lagothrix left01:24
← Benett left01:25
→ refried_ joined01:25
← rafasc left01:26
→ cd joined01:26
→ Benett joined01:27
← SpeakerToMeat left01:28
→ shush joined01:28
→ lagothrix joined01:31
→ zacts joined01:37
← zacts left01:37
→ soft_concrete joined01:39
→ SpeakerToMeat joined01:42
← dmc left01:43
← tiin57 left01:44
→ jaziz joined01:45
← Metalsutton left01:47
← davispuh left01:47
→ tiin57 joined01:47
→ Onionnion joined01:48
← refried_ left01:51
brezanac Any recommendation for a standalone Git LFS server? One that is reliable and doesn't require too much tinkering to setup.01:53
andres_ → andres01:54
← g4570n left01:58
← cdown left02:02
→ dmc joined02:03
← shush left02:05
→ shush joined02:06
→ cdown joined02:07
← shush left02:08
← tiin57 left02:14
→ tiin57 joined02:18
← Cabanoss- left02:30
← thiago left02:34
← ewitz left02:34
← Brainium left02:38
← chachasmooth left02:38
→ chachasmooth joined02:40
→ Cabanossi joined02:40
→ Wetmelon joined02:43
← tiin57 left02:44
→ tiin57 joined02:48
→ sidetrack joined02:49
→ dqx_ joined02:49
→ Goplat joined02:52
← kupi left02:53
← sidetrack left02:54
← lagothrix left02:55
→ lagothrix joined02:55
← opal left03:00
← libertyprime left03:02
→ opal joined03:03
→ sidetrack joined03:05
→ libertyprime joined03:08
← sidetrack left03:10
← tiin57 left03:14
→ tiin57 joined03:18
← canadev left03:18
ChanServ set mode: +o03:19
Eugene No, I'm just an incompetent op who doesn't look at join/part messages :v03:19
Eugene changed the topic to: Welcome to #git, home of your favorite distributed blockchain database | First visit? https://gitirc.eu | Current stable version: 2.26.1 | Getting "cannot send to channel"? /msg gitinfo .voice | This channel is logged: https://gitirc.eu/log | git-outhouse: when even plumbing commands are too nice03:19
Eugene kicked Eugene (de-op)03:20
→ Eugene joined03:20
→ Sasazuka joined03:21
Hello71 is that also why you don't do /mode #git -o Eugene03:22
Eugene Close: somewhere, I have a pisg stats bot03:22
← veegee left03:22
← Sasazuka left03:22
Hello71 or are you just checking that your kick alias is working ;)03:22
Eugene It's apparently been broken since February, again03:22
Also it's a good conversation starter03:23
Hi, how are you?03:23
← Wetmelon left03:25
→ veegee joined03:27
→ MylesBorins joined03:28
← libertyprime left03:30
→ sidetrack joined03:32
→ onizu joined03:36
← Benett left03:39
← MylesBorins left03:42
← veegee left03:43
← sidetrack left03:44
← tiin57 left03:44
→ Benett joined03:44
→ veegee joined03:45
← Maxattax left03:45
→ tiin57 joined03:47
→ sidetrack joined03:51
← mowcat left03:52
← sidetrack left03:57
inkbottle How can I have a path or a reference to a local file but related to a specific commit?03:59
→ enriooooooo joined04:00
← lankanmon_ left04:01
→ lankanmon joined04:04
← tiin57 left04:14
→ fstd joined04:15
→ MylesBorins joined04:17
→ tiin57 joined04:18
→ Cabanoss- joined04:19
← Cabanossi left04:19
Cabanoss- → Cabanossi04:19
→ libertyprime joined04:19
← fstd left04:22
→ fstd joined04:22
_ikke_ inkbottle: <commit>:path/to/file04:30
→ blahboybaz joined04:31
← enriooooooo left04:34
← raven-au left04:35
→ raven-au joined04:35
inkbottle _ikke_: tx04:36
← retroj- left04:37
← Onionnion left04:38
→ NightQuest joined04:39
← libertyprime left04:40
← fitzgerald1337 left04:41
← tiin57 left04:44
→ libertyprime joined04:45
← cliluw left04:47
→ tiin57 joined04:47
→ clara97 joined04:53
→ sidetrack joined04:54
← sidetrack left04:58
→ th_ joined04:59
→ pks_ joined05:02
← umask077 left05:03
→ umask077 joined05:03
← pks left05:04
pks_ → pks05:04
→ sauvin joined05:06
← Heirlung left05:07
← blahboybaz left05:12
← tiin57 left05:14
→ Heirlung joined05:15
→ tiin57 joined05:18
→ TomTom joined05:25
← thomasross left05:27
← NightQuest left05:31
← hqdruxn08_ left05:35
→ hqdruxn08 joined05:38
→ Phylock joined05:38
→ mightyjens joined05:42
← tiin57 left05:44
← nabil left05:46
→ tiin57 joined05:47
→ feriman joined05:59
→ mbalmer joined06:01
← Phylock left06:02
→ Naptra joined06:03
← Betal left06:04
← mbalmer left06:05
← tiin57 left06:14
← ferdna left06:16
← donofrio left06:16
→ tiin57 joined06:17
→ g00s joined06:18
← Goplat left06:24
→ Abdullah joined06:24
→ IoannisJVM joined06:26
← IoannisJVM left06:27
← tpaper left06:28
→ Wetmelon joined06:29
→ tpaper joined06:30
← hussam left06:31
← dqx_ left06:31
→ nickodd joined06:31
→ crose joined06:34
← Jackneill left06:34
← gordonfish left06:41
← Wetmelon left06:44
← tiin57 left06:44
→ tiin57 joined06:47
→ rfuentess joined06:48
→ Aquazi joined06:48
→ dartmouthed joined06:48
← causasui left06:50
→ Jackneill joined06:53
→ fstd_ joined07:00
← clara97 left07:03
← fstd left07:04
fstd_ → fstd07:04
→ clime joined07:08
← soft_concrete left07:09
→ tonythomas joined07:12
→ mbalmer joined07:13
← tiin57 left07:14
→ wootehfoot joined07:14
→ calcul0n joined07:15
→ covid20 joined07:16
→ Borkr joined07:17
← mbalmer left07:18
→ tiin57 joined07:18
← ephemera_ left07:20
→ Paradisee joined07:22
→ SerialSmile joined07:27
→ ephemera_ joined07:28
wootehfoot Aloha, how much or little can be gained by07:31
by PGO+LTO when building git*07:31
tried to google, but git and 'compiling' is hitting way too many results, haven't found traces of 'lto' in git repo, but I figured some madman has already done this07:32
→ mobidrop joined07:38
← tiin57 left07:44
→ nabil joined07:45
→ mbalmer joined07:46
→ tiin57 joined07:48
→ xco joined07:48
← mbalmer left07:50
→ mbalmer joined07:50
← mat001__ left07:56
← Jackneill left07:58
→ akem joined08:00
← akem left08:00
→ akem joined08:01
→ wrksx_ joined08:01
→ wrksx__ joined08:03
← wrksx left08:04
→ dionysus69 joined08:05
← wrksx_ left08:06
← wrksx__ left08:08
→ wrksx__ joined08:10
→ Jackneill joined08:13
← tiin57 left08:14
→ tiin57 joined08:18
← Borkr left08:18
→ abraxxas joined08:24
← mbalmer left08:25
→ mbalmer joined08:25
← mbalmer left08:30
← xSke left08:31
→ xSke joined08:32
→ Daisae joined08:32
← tiin57 left08:44
→ tiin57 joined08:48
← libertyprime left08:49
→ libertyprime joined08:51
← xSke left08:55
← ExeciN left08:56
→ Squarism joined09:01
← Daisae left09:01
← YuGiOhJCJ left09:03
→ YuGiOhJCJ joined09:03
Squarism If you fork a repository "A" into a new one "B". Delete some files you dont want in B. Further along someone changes stuff files the files (deleted in B) of A. Someone pulls master from A into B. The deleted files will re-appear.09:04
Is there some trick to avoid this? Can you say yo're not interested in certain files when pulling?09:04
_ikke_ No, not really09:06
← YuGiOhJCJ left09:06
_ikke_ rebasing is your best option, but conflicts are inevitlbl09:08
→ currybullen joined09:09
← currybullen left09:09
→ currybullen joined09:11
→ selckin joined09:12
currybullen if i do a 'git diff master...HEAD', is there any way to include any current unstaged changes in the displayed diff?09:12
← tiin57 left09:14
← Qizzy left09:15
→ Qizzy joined09:15
← crater2150 left09:16
← SerialSmile left09:16
→ thefatbird joined09:16
→ crater2150 joined09:17
→ tiin57 joined09:17
→ ablackack joined09:17
→ hightower2 joined09:19
hightower2 Hey folks, I had two branches - master and dev which was branched from master. Because there were various commits to both, I first merged master into dev and resolved all conflicts. Now I would like to move the complete thing back to master. What's the right way to do this?09:21
→ mbalmer joined09:24
Habbie hightower2, merge dev into master - should be painless at this point09:24
hightower2 hmm strange, not sure why it gave me a ton of conflicts then.. let me check what the nature of the reported conflicts is then09:26
← mbalmer left09:30
cdunklau hightower2: did you _commit_ the conflict resolution?09:31
it would help to see your terminal session09:31
Habbie hightower2, it should be painless *now that master is merged into dev* as you said09:32
← svm_invictvs left09:34
→ svm_invictvs joined09:35
→ bolovanos joined09:38
→ mauz555 joined09:39
→ umask078 joined09:40
→ igemnace joined09:40
← umask078 left09:43
← tiin57 left09:44
← umask077 left09:44
→ umask077 joined09:45
→ tiin57 joined09:48
← libertyprime left09:55
← rustyshackleford left09:57
← nabil left09:59
→ mbalmer joined10:05
← mbalmer left10:07
→ cousteau joined10:11
cousteau Hi!10:11
← abraxxas left10:12
cousteau What's the difference between git ls-files --ignored and --others?10:13
← chiastre left10:14
cousteau git ls-files --ignored --exclude-standard doesn't seem to show all ignored files10:14
← tiin57 left10:14
→ chiastre joined10:15
osse it's not either-or10:15
--others show untracked files. you use --ignored *with* --others to filter the result to only ignored files10:16
cousteau OK so ignore others shows files that are ignored AND untracked10:17
→ calcul0n_ joined10:17
cousteau Whereas others by itself shows only untracked files10:17
And ignored by itself shows... uh... files that are tracked despite of being in .gitignore maybe?10:18
→ tiin57 joined10:18
osse yes10:19
tracked files that also match in .gitignore (ie. they would have been ignored)10:19
cousteau In general, if I want to list all files that are not part of the commit I'd do git ls-files --others, right?10:19
← akem left10:19
cousteau I see, thanks!10:19
osse yes10:19
→ akem joined10:20
← calcul0n left10:20
← akem left10:20
cousteau OK then that's exactly what I wanted10:20
osse you can also use --exclude to customize what you want to ignore10:20
→ akem joined10:20
osse might be useful for enumeration only certain types of ignored files10:20
*enumerating10:20
← calcul0n_ left10:21
cousteau Good, but for the time being I just wanted to know which files had been generated by this project10:21
→ nabil joined10:23
→ calcul0n joined10:23
cousteau Now I'm gonna see if I figure out the right git submodule foreach command10:23
→ hussam joined10:26
cousteau OK it was trivial, BUT the problem is that git submodule foreach shows paths relative to the repository home10:27
→ chloekek joined10:27
→ HumanGeek joined10:28
_ikke_ guardian: Might be a similar case to what you ran into: https://lore.kernel.org/git/87zhbdcflb.fsf@igel.home/T/#t10:29
chloekek Hi, I am looking for a command that would change the parent commit of a given commit, but no more than that (in contrast with rebase, which computes applies changes on top of something, rather than retain the exact snapshot of the tree).10:29
osse cousteau: within the foreach script snippet there are some vars you can use10:29
cousteau In fact I think this is more likely a task for find -newer10:29
osse: thanks, will look into them10:30
→ mbalmer joined10:30
← Human_G33k left10:31
chloekek It seems like what I’m looking for is grafts. I’ll try that.10:32
cousteau Anyway yes, I think I'll just use find, since I would also like to track WHEN files were created10:32
→ troulouliou_dev joined10:32
osse cousteau: it's exactly that you're after. git replaced --graft10:34
oops, that was for chloekek10:34
cousteau :)10:34
osse chloekek: note that this will only work in the local repository. you can use filter-branch or equivalent to actually (re)write the object in properly, which necessitates force push and all the rest of it10:34
← mbalmer left10:35
chloekek osse: Thanks. Actually I just found a third-party command git-reparent which does automates this. :)10:38
← JanC left10:41
→ JanC joined10:42
← tiin57 left10:44
→ tiin57 joined10:47
← chloekek left10:49
← Anty0 left10:50
← g00s left10:56
→ pepa joined10:58
pepa to do the equivalent of svn update would be git pull right?10:58
_ikke_ pepa: Not quite11:00
in the sense that they are not directly equivalent11:00
pepa I want to get code from remote11:00
which git command should I use11:01
→ mbalmer joined11:02
_ikke_ git pull would work, but if you run that arbitrarily, you will also get a lot of merges in you history, which would make it hard to follow11:02
pepa: did you create commits locally already?11:02
pepa my real issue is11:04
→ mowcat joined11:04
pepa I do "git pull" and I get "failed to connect... timed out"11:04
any suggestions on how to debug this?11:04
like is there any git command built for debugging stuff like this?11:05
_ikke_ GIT_TRACE=1 git pull11:06
but that goes low level11:06
what does git remote -v return?11:06
← ryzokuken left11:08
→ ryzokuken joined11:08
pepa GIT_TRACE=1 git pull ------------> GIT_TRACE is not recognized as an internal or external command11:09
← mbalmer left11:10
← ryzokuken left11:10
_ikke_ oh, you are on windows11:10
→ ryzokuken joined11:10
← tiin57 left11:14
→ tiin57 joined11:18
← fstd left11:19
→ fstd joined11:20
lineos The repository has the structure: develop branch and branches for features. If feature2 depends on feature1, but feature1 has not been merged to develop yet, what would be a good strategy to start working on feature2 before this happens? Make feature2 branch from feature1? When doing this in the past I have always had a lot of merge conflicts. Is there a better strategy to avoid this? Maybe branch from develop and pick11:24
changes from feature1?11:24
← [Brain] left11:24
← shicks2525 left11:25
→ shicks2525 joined11:25
→ fermion joined11:27
→ fitzgerald1337 joined11:27
→ LiENUS joined11:27
→ nickmass joined11:28
→ holdsworth joined11:28
← Technodrome left11:28
→ thevisitor joined11:29
← bket left11:30
← osse left11:30
→ bket joined11:30
→ osse joined11:31
lineos or it's the same anyway11:32
bkhl Branching from feature1 seems most straight forward based on what you describe.11:32
lineos and after feature1 is merged, merge develop into feature2 or merge feature1 into feature2?11:33
→ c0san0stra joined11:33
← fitzgerald1337 left11:34
_ikke_ You could just merge feature2 into develop when feature2 is ready11:34
lineos No need to update feature2 before merging?11:35
_ikke_ merging feature1 into feature2 makes no sense, as feature2 is based on feature111:35
bkhl _ikke_: except if there are minor changes to feature1, or develop is merged into the feature1 branch11:36
→ fitzgerald1337 joined11:37
→ thiago joined11:37
bkhl Anyway even if you do, it should then be no problem to either first merge feature1 and then feature2, or just feature2 directly (which will also include feature1)11:37
… to develop I mean.11:37
lineos Thank you, I'll try that then.11:38
← feriman left11:39
→ thomas_25 joined11:40
← tiin57 left11:44
→ tiin57 joined11:48
→ xSke joined11:48
_ikke_ For merge conflicts it does not matter anyway what you merge into what11:48
→ silviof joined11:50
→ caterfxo joined11:52
→ mescalinum joined11:54
← Jackneill left11:55
→ aliss joined11:59
← mescalinum left12:00
← fitzgerald1337 left12:01
→ mescalinum joined12:03
← mescalinum left12:07
→ Janni joined12:09
Janni Hello there! I have some questions concerninge "automatic merge strategies". A bit of background...12:09
In our company we have decided to go with a testing/branching pattern where tests are done (mostly) in a single shared environment.12:10
→ Jackneill joined12:11
Janni (I'm not here to discuss this particular choice, actually it's working out quite fine so far.)12:11
nkuttler what do you mean by tests?12:11
like, a feature branch?12:11
→ bsjd joined12:12
Janni Alright, I'll explain.12:12
Developers will work on their own feature branches. However, the (entire) application is not that easy to deploy so we want to avoid too many versions being deployed at the same time.12:13
← tiin57 left12:14
Janni Also, the mental capacities of the testers are... limited. They tend to be confused by having to look at different versions of the application in order to test different features.12:14
So we have decided to keep feature branches for the developers, but to deploy a wild mixture of untested features together in a single application.12:15
So we have a branch "test" into which everyone can merge their features. This branch is then automatically deployed (after each change).12:15
← dedowsdi left12:16
Janni This approach is working out quite well so far, except for the obvious problem... conflicts.12:16
We were thinking a solution as follows: we merge features automatically into the testing branch.12:17
Let's say we take all open PR on our github repository (potentially only those tagged with a certain label).12:17
→ tiin57 joined12:18
Janni Then each time there is a new PR (or a PR changes) we recreate the "test" branch, by resetting it to master and then merging the PRs one-by-one, skipping over any PR that causes a conflict, triggering some sort of notification.12:19
I was just wondering whether anyone here already had experience with that sort of approach and whether they could recommend a strategy (or even a tool).12:19
_ikke_ Sounds a bit like what Junio (the git maintainer) does for the pu branch12:20
Janni I'm not too worried btw, whether the approach will work or not. The non-automatic approach we're currently using is working alright, and the automisation will makes things even more practical.12:20
_ikke_ look also in man git rerere to let git automatically resolve conflicts that you solved before12:21
gitinfo the git-rerere manpage is available at https://gitirc.eu/git-rerere.html12:21
Janni ikke: *looking into it*12:21
Thanks.12:21
→ Brainium joined12:22
→ fission6 joined12:23
→ cruxeternus joined12:30
→ fassl joined12:35
← Jackneill left12:35
← karoshi left12:36
→ Wetmelon joined12:39
← tiin57 left12:44
→ ar1el joined12:45
← theorangeone left12:46
→ tiin57 joined12:48
→ mbalmer joined12:50
→ Maxattax joined12:50
→ Jackneill joined12:50
← FFY00 left12:51
→ theorangeone joined12:51
→ FFY00 joined12:52
← mbalmer left12:56
← Wetmelon left12:56
→ CodeSlingerPaul joined12:59
→ kristofferR joined13:00
← thiago left13:01
← xcm left13:02
→ xcm joined13:02
→ thiago joined13:03
→ g4570n joined13:04
← FFY00 left13:04
→ FFY00 joined13:05
← jaziz left13:07
→ c0c0 joined13:09
→ mat001 joined13:10
→ dostoyevsky joined13:11
dostoyevsky is there a good way to completely remove files from the git repo? Sometimes I check in large blobs to make sure e.g. some code runs reliably, but later I might find a way to do it without the blob... and would like to slim the repo down again, at the price of breaking older versions in git13:12
← aliss left13:13
→ thebope joined13:13
selckin !paste13:14
gitinfo Please use a pastebin for snippets longer than one line. Persistent and can be used with a GitHub account: https://gist.github.com/ - one hour auto-delete: https://upaste.de/13:14
← tiin57 left13:14
selckin dostoyevsky: depends what you mean with good, this is a sript i keep around https://upaste.de/oGE13:15
but probably 'no'13:15
dostoyevsky selckin: my go to solution in the past has always been: create a new repo... based on the state without the large blobs... that seems to be cleanest... so I'd have repos like: project, project2 ... and the highest number is the one that's currently active13:17
→ tiin57 joined13:18
dostoyevsky selckin: thanks for the script!13:19
→ rustyshackleford joined13:19
← cousteau left13:26
← kristofferR left13:27
→ Phylock joined13:30
← roadie left13:30
→ refried_ joined13:35
→ Borkr joined13:43
← tiin57 left13:44
→ coet-i3 joined13:46
coet-i3 Hi everyone! How can I know why is 'git push' asking for a user and password of github.com instead of using my ssh key? I thought I had everything setup properly, but it's still asking me for user/passwd13:46
osse coet-i3: what does git remote -v print?13:47
→ tiin57 joined13:48
coet-i3 osse: origin https://github.com/myuser/myrepo.git (fetch)13:48
→ FH_thecat joined13:48
→ chandoo joined13:48
coet-i3 and the same but with (push)13:48
Hello71 what's the point of censoring your username13:48
osse coet-i3: git remote set-url origin [email@hidden.address] and try again13:48
→ vdamewood joined13:48
coet-i3 osse: yay! that worked, thanks!13:49
← JanC left13:50
coet-i3 I didn't know that the usage of ssh keys depends on the way you clone the repository :S13:50
→ JanC joined13:50
→ relipse joined13:51
osse well... technically it doesn't since we were able to fix it :p13:52
the URL matters.13:52
leah2 how can i resolve REVISION:PATH to a BLOB id?13:53
git show will display it, but i need the hash13:53
osse rev-parse13:54
leah2 can i do that for many such strings at once?13:54
osse yes13:55
Hello71 coet-i3: well you need to use ssh to use ssh13:55
→ amahl_ joined13:56
coet-i3 Hello71: yes, that makes sense :P13:57
leah2 osse: but just via cmdline, not stdin?13:58
osse yes13:58
leah2 why >.>13:58
osse huh?13:59
leah2 name-rev etc all have --stdin13:59
osse oh13:59
→ aliss joined13:59
osse i misunderstood13:59
→ cdown_ joined14:01
osse Would be useful to read stdin14:01
leah2 i want to find all blobs that belong to a filename over the history14:02
perhaps parsing git-log --raw is better14:02
osse yeah that or rev-list14:02
leah2 how does rev-list help?14:02
osse it's basically log without the prettiness14:02
also it has a --stdin flag14:03
leah2 and without the diff-tree14:03
so i dont get the blob ids?14:03
osse diffing is not related14:03
leah2 or print the tree14:03
← cdown left14:03
cdown_ → cdown14:03
osse try git rev-list --objects master -- filename | grep filename14:04
← Janni left14:05
leah2 aah14:05
osse awk -v f=filename '$2 == f { print 41 }'14:05
leah2 perfect14:05
← vdamewood left14:05
→ gordonfish joined14:05
← bsjd left14:05
← theorangeone left14:06
osse awk -v f=filename '$2 == f { print $1 }'14:06
← FH_thecat left14:07
→ theorangeone joined14:07
leah2 ofc14:07
i didnt see --objects at first14:07
osse take a look at --filter too14:08
leah2 yep, but here i just need all blobs14:08
osse might be able to get away without any grepping or awking14:09
leah2 this is a perl script anyway14:09
← mowcat left14:11
← dmc left14:12
← sgn left14:12
← tiin57 left14:14
→ mowcat joined14:14
→ sgn joined14:14
→ theoceaniscool joined14:15
pepa can you please look at this stack trace? https://pastebin.com/BvEfqq3C14:17
→ tiin57 joined14:17
pepa it seems the problem only happens from the console, not from a web browser for example14:18
Habbie is something like telnet able to connect to the port from the console?14:18
_ikke_ Is there a proxy involved?14:19
tango_ g2214:20
ehm sorry14:20
Habbie _ikke_, i put in five bucks14:21
pepa telnet is not recognized as an internal command14:21
wtf14:21
windows doesn't have telnet anymore?14:21
Habbie you have to install it14:21
some pkg command14:21
pepa in windows14:21
Habbie yes14:21
coet-i3 why does _ikke_ have an '@' in front of the nickname?14:21
Habbie but any other tool that can connect to a port would work too14:21
like an ftp client or ssh or curl or wget14:21
pepa, dism /online /Enable-Feature /FeatureName:TelnetClient14:22
pepa but anyway I am also connecting to the same url and port from a web browser14:22
Habbie yes14:22
we read that :)14:22
pepa so that works that there is a connection14:22
so that shows that there is a connection14:22
if that was your question14:22
Habbie no14:23
the question was whether any other software *in the console* can make the connection14:23
_ikke_ pepa: set | findstr http14:24
→ donofrio joined14:25
pepa so it's telnet host:port14:25
right?14:25
Habbie it's 'telnet host port' on my linux14:26
pepa I'm on windows14:26
Habbie just telling you what i know - it's probably a space for you too14:27
← coet-i3 left14:28
pepa well my partner can do the git pull14:28
and the telnet does not work for him either14:28
Habbie how does it fail?14:28
can he also look at 'set | findstr http' and compare the output with yours?14:29
pepa could not open connection to the host, on port 443: connect failed14:29
what does 'set | findstr http' do?14:30
Habbie it looks for environment variables with http in their name, is my guess14:30
← Eduard_Munteanu left14:31
pepa it doesn't show anything14:31
result is empty14:31
Habbie on both systems?14:32
pepa yes14:33
Habbie then i'm out of ideas14:34
pepa it says14:35
"Couldn't find host yyyyyyyyyyyyyyyyyy in the _netrc file"14:35
what is the _netrc file?14:35
→ royal_screwup21 joined14:35
← tanuki left14:36
→ tanuki joined14:38
→ mbalmer joined14:39
Habbie i don't know14:40
weird host name btw ;)14:40
→ roadie joined14:41
→ Newami joined14:41
→ BPL joined14:41
← Newami left14:42
→ sidetrack joined14:43
_ikke_ pepa: curl (which git uses in the background) uses a netrc file to specify some rules per host14:43
that it cannot find it is not an issue14:43
→ feriman joined14:43
Habbie ah14:43
so we can get curl -v?14:43
← fraktor left14:43
Habbie from both systems?14:43
that should be useful14:43
_ikke_ libcurl does not met curl as a command is installed14:44
← mbalmer left14:44
← tiin57 left14:44
Habbie right14:45
then can we get useful debugging out of the working system perhaps?14:45
← figuinhas left14:47
→ tiin57 joined14:48
pepa so it's curl -v host:44314:48
→ malmalmal joined14:48
_ikke_ curl -v http://hos14:49
curl -v https://hos14:49
pepa so I did git remote -v14:52
in the working system14:52
this returned a url ending in .git14:52
← HumanGeek left14:52
pepa then I did14:53
curl -v https://.....git14:53
and it says "failed to connect to xxxx: timed out"14:53
this is in the working system14:53
← SirD_P left14:57
pepa any ideas?14:57
← random_yanek left15:01
← mowcat left15:01
_ikke_ No, sorry15:02
amahl_ for some personal projects I've used git in a silly way15:04
where master branch has hundreds of tiny commits. each is a small change15:05
← thansen left15:05
amahl_ now I wanted to squash it a little15:06
→ SirD_P joined15:06
amahl_ and all web tutorials say I should use merge. its supposed to be fast forward. one branch15:06
and yet I'm getting merge conflicts15:06
I first created a new branch from the very first commit15:07
→ thansen joined15:07
amahl_ and tried to do "git merge xxx" where xxx is a dozen commits forward15:07
actually git merge --squash xxx15:07
→ random_yanek joined15:08
amahl_ merge conflicts out of nowhere15:09
→ zhuoz joined15:09
← theoceaniscool left15:09
← feriman left15:10
amahl_ am I supposed to use rebase instead? if I want a new squashed master15:10
→ feriman joined15:10
amahl_ shouldn't merge and rebase achieve the same goal when dealing with just one branch?15:11
osse amahl_: to squash the N latest commit you can do: git reset --soft HEAD~N; git commit15:11
using merge for this seems strange15:11
to squash some commits between now and M commits ago you can git rebase -i HEAD~M15:12
and then change pick to fixup/squash as you please15:12
→ ferdna joined15:12
amahl_ I see15:13
pepa how is it possible that telnet host 443 does not work ...... but git pull does work?15:13
← tiin57 left15:14
Habbie pepa, you'll have to find out what git is doing differently. maybe he does have a netrc?15:14
pepa this is in the pc that works "fine"15:15
also....15:15
the pc working wrong "sometimes" works15:15
so it's more of an intermitent issue15:16
Habbie 'fun'15:16
← BPL left15:18
→ tiin57 joined15:18
← akem left15:19
pepa actually15:19
everything is working15:19
because15:19
"git remote -v"15:19
is working15:19
that means the connection to the server is successful right?15:20
_ikke_ no15:20
It just lists what is configured locally15:20
pepa what git command can I try that's light and short to test connection15:20
_ikke_ git ls-remote15:20
git remote show origin as well15:21
pepa so that connects with the server15:22
_ikke_ yes15:22
pepa I mean git ls-remote15:22
_ikke_ it does15:22
pepa so if that works the connection works as well15:22
_ikke_ yes15:23
→ akem joined15:26
← refried_ left15:28
amahl_ rebasing going nicely15:28
but still got some "BOTH ADDED: blahblabh" nonsense15:28
like.. theres just one master. whats the second whatever?15:29
← ShapeShifter499 left15:30
amahl_ doing git rebase -i HEAD~M15:30
to squash one branch onto itself15:30
→ ShapeShifter499 joined15:30
← justanotheruser left15:31
→ refried_ joined15:31
→ mbalmer joined15:33
← malmalmal left15:34
→ HumanG33k joined15:37
← mightyjens left15:37
osse amahl_: Did you reorder lines? Do you have any preexisting merges in that range?15:37
← mbalmer left15:38
→ Prestige joined15:38
amahl_ didn't reorder anything. just sequences like pssssss rsssss p p pssss15:38
not sure of merges15:39
← HumanG33k left15:39
→ HumanG33k joined15:40
Prestige I had a branch merged with master that was reverted automatically (github hook), and now I'm wanting to merge the same branch with master again. Since it was merged before, the commits exist on master but are covered by the commit that reverted the merge previously. How should I go about setting this up for a merge?15:40
osse amahl_: git log --merges HEAD~M..HEAD (but not while a rebase is in progress)15:40
Prestige: !revert_merge15:41
gitinfo Prestige: [!faulty_merge] If you have published a merge commit that turned out to be faulty and you'd like to get rid of it, you can use 'git revert' but there are gotchas to be aware of. Please read: https://raw.github.com/git/git/master/Documentation/howto/revert-a-faulty-merge.txt15:41
osse hmm, maybe this only covers what you what the hook already does for you15:41
anyway, it's dirty bidness15:41
→ chloekek joined15:42
← cdown left15:42
Prestige I was thinking of creating a diff patch by comparing HEAD and HEAD~3 and applying that on top, and creating a new commit15:42
not sure if that would work15:43
osse sounds reasonable15:43
but you might want to try git cherry-pick instead15:43
or maybe git rebase --onto15:43
(which essentially do what you described under the hood)15:43
→ cdown joined15:44
← tiin57 left15:44
→ justanotheruser joined15:47
→ calcul0n_ joined15:47
Prestige oh cherry pick is exactly what I needed, thanks15:47
→ tiin57 joined15:48
pepa git bring-me-an-ice-cream doesn't seem to be working15:48
← calcul0n left15:50
← archpc left15:51
← royal_screwup21 left15:53
→ bsjd joined15:56
← bsjd left15:56
→ john2496 joined15:56
→ thomasross joined15:57
← john2496 left15:58
← SilentGhost left15:58
→ john2496 joined15:59
← john2496 left15:59
← rfuentess left16:00
→ SilentGhost joined16:00
Hello71 you need to run it with sudo16:04
→ fitzgerald1337 joined16:04
← sidetrack left16:08
← fission6 left16:09
→ fission6 joined16:13
← tiin57 left16:14
← irrgit left16:15
→ tiin57 joined16:17
← Aquazi left16:19
← SilentGhost left16:19
→ SilentGhost joined16:20
→ sidetrack joined16:23
← sidetrack left16:28
fuchstronaut after cloning it I got a "WARNING: 'new-name' is an alias for 'old-name'", is there anything I need to know about the alias?16:28
→ irrgit joined16:30
← tonythomas left16:30
→ fannagoganna joined16:30
→ causasui joined16:31
j416 how do I check what commits (actual changes) were added for v2.26.1 since v2.26.0, in git.git? I don't get the branching model. "git log ^v2.26.0 v2.26.1" lists a bunch of tagged commits, back to v2.17.4 -- I'm not sure why.16:32
I suppose it's the right list, but there are a bunch of merges; --no-merges makes it a bit cleaner16:34
→ chuffer joined16:34
j416 I guess I'll just go read their branching model docs, I'm sure they've documented it.16:35
→ spaceone joined16:35
← chuffer left16:36
spaceone how can i show all commits in branch A which are not part if branch B?16:36
if → in16:36
_ikke_ git log B..A16:36
spaceone no, this shows only when then branch has been split16:36
but i cherry-picked from A to B some commits, which are also in the log16:37
→ chuffer joined16:37
→ dmc joined16:38
_ikke_ man git cherry then16:38
gitinfo the git-cherry manpage is available at https://gitirc.eu/git-cherry.html16:38
← nickodd left16:39
spaceone ah, this looks nicer!16:40
← chuffer left16:42
→ Quellus joined16:44
← tiin57 left16:44
→ nickodd joined16:44
→ tiin57 joined16:48
← pepa left16:49
← Abbott left16:51
← roadie left16:56
→ roadie joined16:56
→ karlding joined16:57
← Maxdamantus left17:01
→ Maxdamantus joined17:01
← maxwilliamson left17:03
→ maxwilliamson joined17:04
← fission6 left17:05
→ fission6 joined17:07
→ FH_thecat joined17:07
← FH_thecat left17:07
→ sidetrack joined17:10
← fission6 left17:12
→ mowcat joined17:12
← chloekek left17:12
← tiin57 left17:14
→ fission6 joined17:14
→ tiin57 joined17:17
← diogenese left17:19
← Prestige left17:19
→ diogenese joined17:19
→ mbalmer joined17:22
→ stitched_unicorn joined17:23
← cbreak left17:26
← refried_ left17:26
→ refried_ joined17:27
→ cbreak joined17:27
← mbalmer left17:35
osse j416: when you've understood it, be kind and write an "ELI5" type summary because it makes no sense to me17:38
eggbean I've found an old file. Is it possible to add a first commit on a repository that already has other commits?17:39
osse j416: i think the older tags are backports (or "backmerges"), so they're newer in the time sense17:39
← mobidrop left17:40
→ archpc joined17:41
j416 osse: :D thanks for making me feel less stupid.17:41
eggbean: you can have multiple roots for a repo, yes. not sure if that's what you're asking. if so, man git-checkout look for --orphan.17:42
gitinfo eggbean: the git-checkout manpage is available at https://gitirc.eu/git-checkout.html17:42
eggbean j416: Thanks, I'll have a look17:43
← tiin57 left17:44
← sidetrack left17:45
← Paradisee left17:46
→ tiin57 joined17:48
→ mikecmpbll joined17:50
→ sidetrack joined17:50
→ royal_screwup21 joined17:52
fuchstronaut Is "WARNING: 'new-name' is an alias for 'old-name'" while cloning a repo something to be concerned about?17:52
← kleisli_ left17:53
→ kleisli_ joined17:53
← sidetrack left17:55
→ cliluw joined17:56
→ shush joined17:58
← archpc left17:59
← royal_screwup21 left18:00
← squirrel left18:01
← kleisli_ left18:01
← cliluw left18:03
→ cliluw joined18:03
→ m0viefreak joined18:06
← mat001 left18:12
← bolovanos left18:13
← tiin57 left18:14
← sauvin left18:14
→ royal_screwup21 joined18:15
→ archpc joined18:17
j416 fuchstronaut: sounds like something github would say18:17
→ tiin57 joined18:17
j416 fuchstronaut: you might want to clone from the renamed one instead, or at least update your remote to match that18:17
→ mbalmer joined18:18
→ mat001 joined18:20
→ Betal joined18:21
← royal_screwup21 left18:22
→ Lucas_Gray joined18:23
← nickodd left18:25
← refried_ left18:25
← mbalmer left18:26
→ akem_ joined18:27
← akem_ left18:27
← Tobbi_ left18:27
→ refried_ joined18:29
← akem left18:30
→ sidetrack joined18:31
oprypin i have many commits and merges and whatnot. i had run `git merge origin/master` a few times, including as the last operation18:33
i want to turn the entire diff against origin/master into one commit, reusinng the commit message of one particular commit among those18:33
what's a non-messy way to do that?18:34
← sidetrack left18:36
→ squirrel joined18:36
oprypin oh. self-answer:18:37
git merge origin/master && h=$(git rev-parse HEAD) && git reset --hard origin/master && git reset --soft $h && git commit --reuse-message $particularcommit18:37
no that's wrong18:38
git merge origin/master && h=$(git rev-parse HEAD) && git reset --hard origin/master && git checkout $h . && git commit --reuse-message $particularcommit18:38
← stitched_unicorn left18:42
→ mns_ joined18:43
← mns left18:43
← tiin57 left18:44
→ dqx_ joined18:45
← dqx_ left18:46
→ royal_screwup21 joined18:47
→ akem joined18:48
→ tiin57 joined18:48
← akem left18:48
← mns_ left18:48
→ akem joined18:48
← qqx left18:49
→ sidetrack joined18:50
← teratorn left18:53
→ qqx joined18:55
← calcul0n_ left18:56
→ calcul0n joined18:57
← shush left18:58
→ shush joined18:59
← currybullen left19:00
← shush left19:00
→ shush joined19:01
Kobaz how would i show a specific file in the stash?19:03
git stash show stash@{0}:LIB/any/perl/WebServices.pm 'stash@{0}:LIB/any/perl/WebServices.pm' is not a stash-like commit19:03
tried -- and .. as well19:04
→ foodz joined19:08
← xcm left19:10
→ stealth[] joined19:11
→ xcm joined19:11
→ mikecmpb_ joined19:12
← mikecmpbll left19:12
← kjartan left19:13
← maxel left19:14
← tiin57 left19:14
→ kjartan joined19:14
→ cagomez joined19:14
→ maxel joined19:14
→ mns joined19:15
cagomez Hi, I branched off of `master` and created branch `feature_1`. Then I branched off of `feature_1` to create `feature_2`. I did some work and now have commits from both branches. how can I undo the changes from `feature_1` branch while I'm on `feature_2`?19:16
→ mikecmpbll joined19:16
← sidetrack left19:16
selckin create a new branch, and cherry-pick the commits, or interactive rebase to remove feature_1 commits19:17
← mikecmpb_ left19:17
→ tiin57 joined19:18
→ dqx_ joined19:18
← royal_screwup21 left19:20
← troulouliou_div2 left19:24
→ mnglssrspndr joined19:28
← Soliton left19:31
← dqx_ left19:31
cagomez okay, tried something else. I pulled `feature_1` locally and got the list of files changed: `git diff --name-only $(git merge-base master HEAD)`. Then I checked out `feature_2` and checked out those files from `master`: `git checkout master -- [list_of_files]`. Then `git add .` and pushed19:31
← stealth[] left19:31
→ Soliton joined19:31
← meaningless left19:32
selckin i hope you don't have coworkers19:32
→ royal_screwup21 joined19:34
→ mescalinum joined19:35
→ stealth[] joined19:36
Kobaz cagomez: that sounds really bad. you should use merge properly19:36
your method 'works' but, other people working on the repo will hate you19:37
and if you ever do that on someone elses repo. they will hate you19:37
→ mikami joined19:37
cagomez is it bad because of the git history?19:38
← mikami left19:38
Kobaz yes19:38
cagomez what if I `get reset --soft` and re-do the commits, then `force push`?19:38
not sure if that's even worse19:39
Kobaz you'll have to fix it on the remote repo, you've already pushed19:39
cagomez yup, it's been pushed.19:39
Kobaz you can undo the commits on the remote repo as long as no one else has checked them out yet19:39
cagomez people rarely check out commits. we QA the branch, then merge it into master19:39
* checkout branches19:40
← xcm left19:40
Kobaz https://devconnected.com/how-to-undo-last-git-commit/19:41
good luck, make backups19:41
→ xcm joined19:41
Kobaz you'll need to do this on the remote repo19:41
← fitzgerald1337 left19:43
← tiin57 left19:44
→ azwieg103 joined19:44
→ soft_concrete joined19:45
→ fitzgerald1337 joined19:46
→ tiin57 joined19:48
← azwieg103 left19:51
→ karoshi joined19:53
→ nessonic joined19:54
← aliss left19:55
cagomez I `reset soft`, redid commits, and force pushed. the commit history is good now19:56
→ no_gravity joined19:58
→ clara97 joined19:58
no_gravity Can "git status" show the overall number of added/deleted lines?19:59
← fassl left19:59
_ikke_ No, git diff can19:59
git diff --stat19:59
or git show -s --stat19:59
no_gravity _ikke_: Nice, thanks!19:59
_ikke_ just git show --stat, not with -s20:00
no_gravity _ikke_: That shows me the info for the last commit20:00
_ikke_ no sorry, git show shows HEAD20:00
→ mbalmer joined20:00
_ikke_ yea20:00
git diff --stat then20:00
no_gravity Yup20:00
osse y u no --shortstat20:01
ironic that "--shortstat" is longer than "--stat"20:01
_ikke_ yea20:01
was about to answer that20:01
no_gravity It is so satisfying when a commit has more deletions then insertions.20:01
The day ends on a high note then.20:01
_ikke_ heh20:01
Boss without a clue: "I don't pay you to remove code, I pay you to write code" :P20:02
no_gravity _ikke_: My userbase is my boss. They don't give a shit about code. They want features and performance :)20:03
Kartagis hello. remind me please? wasn't it git format-patch <hash> for a single commit?20:05
← mbalmer left20:05
_ikke_ no20:06
← g4570n left20:06
_ikke_ It can be used for a range20:06
or limited to one commit20:06
Kartagis _ikke_: I got nothing when I used the above.20:07
← m0ltar left20:07
_ikke_ Kartagis: try git format-patch -1 <commit>20:08
and note that it's directly writing to a file20:08
→ m0ltar joined20:09
← Hello71 left20:13
← refried_ left20:13
Frogging101 How do I fast-forward a branch that I'm not on? I've been using e.g. "git fetch origin master:master", but I'd be surprised if there weren't a more general way that doesn't require a remote20:13
← tiin57 left20:14
Frogging101 There's `git branch -f`, but that will move the branch head regardless of whether it's a fast-forward20:14
→ Hello71 joined20:14
← no_gravity left20:14
_ikke_ the only thing that has a check for only doing fast-forwards is merge20:15
(and pull by extension)20:15
Frogging101 fetch does too with that form20:15
I've seen it reject non-ff20:15
_ikke_ yes, the only other thing I should have mentioned20:15
← royal_screwup21 left20:15
Frogging101 ah20:15
And merge/pull require the target branch to be checked out?20:15
_ikke_ yes20:15
→ refried_ joined20:16
Frogging101 I wonder if there's a scriptable way to check if a commit is a direct ancestor (or whatever the exact criteria for a fast-forward is)20:16
so then I could implement a `git branch -f` wrapper that checks first20:16
_ikke_ merge-base20:17
→ royal_screwup21 joined20:17
_ikke_ git merge-base --is-ancestor A B20:17
Frogging101 ah. I'll look into it. Thanks :)20:18
→ tiin57 joined20:18
Frogging101 Fwiw, my use case for this is that I've been maintaining patch sets (feature branches?) on many different repos, and every day I fetch from upstream and then rebase. I could just rebase onto origin/master, but I also want to keep my local master updated so I don't forget that it's behind20:20
And it's nice if I can avoid the extra steps of checkout master, pull master, checkout feature20:20
_ikke_ I just use git checkout -B / branch -f in that case20:21
But that's because I know I have nothing on the branch that I care about20:21
(checkout -B in case I have the branch checked out, branch -f otherwise)20:22
Frogging101 I generally don't, but I like to be defensive. I can be forgetful20:22
_ikke_ Then I cound on the reflog :)20:22
count*20:22
← troulouliou_dev left20:22
→ troulouliou_dev joined20:23
Frogging101 That works if I know that I've clobbered something. Otherwise I find out later and then get confused and waste time retracing my steps :P20:23
→ Brutus joined20:23
gitinfo set mode: +v20:24
Frogging101 So I try to set up my environment and tools in a way that helps to avert the novel ways I find to screw myself20:24
_ikke_ sure20:24
Just make an alias / script which does that check before doing the action20:25
Frogging101 yup. I do love how Git subcommands are actually just scripts with a naming convention ^^20:25
very extensible20:25
_ikke_ yes, though most git commands are compiled built-ins nowadays20:25
Frogging101 indeed20:26
_ikke_ but that git foo will look for git-foo is still supported20:26
← Maxattax left20:26
Frogging101 Git does so many things right that I often look at its code as a guide for how to design things... :p20:27
Brutus does anyone know a good simple guide to follow to set up and use a repo on github? i'm trying to set up a way to store my dotfiles on it20:27
_ikke_ Brutus: github itself has good docs20:27
Brutus _ikke_: yeah they talk about a lot of things, but not what i'm looking for20:27
or at least, i can't find it in there20:27
_ikke_ Brutus: https://help.github.com/en/github/creating-cloning-and-archiving-repositories/creating-a-repository-on-github20:27
Brutus yeah was already there20:28
← roadie left20:28
_ikke_ Ok, but what are you missing then?20:28
Brutus it's mostly talking about the website stuff20:28
→ cdown_ joined20:28
Brutus so i set up git config --global with user and name20:29
trying to do a commit20:29
https://medium.com/toutsbrasil/how-to-manage-your-dotfiles-with-git-f7aeed8adf8b20:29
trying to follow that guide20:29
_ikke_ https://help.github.com/en/github/creating-cloning-and-archiving-repositories/cloning-a-repository20:29
Brutus the part i'm missing is how to get the git client to actually do that. i get errors like fatal: No configured push destination.20:29
i'm not trying to clone a repo20:29
_ikke_ aha20:29
→ sidetrack joined20:30
Frogging101 Brutus: git init; git remote add origin <github>; git push -u origin master20:30
_ikke_ git push --set-upstream <remote> <branch>20:30
Brutus yeah the remote branch, that's below that. i don't know what that means20:30
Frogging101 (skip git init if you already did it)20:30
← cdown_ left20:30
Frogging101 The remote is the "origin" you added with "git remote add"20:30
but you could call it anything. origin is just the usual name20:31
Brutus ok one moment20:31
← cdown left20:31
→ cdown joined20:31
Frogging101 and if you already have an "origin" remote: `git remote set-url origin <repo url>`20:31
(this is from memory; hope I didn't get anything wrong)20:31
_ikke_ that's correct20:32
Frogging101 good ^^20:32
_ikke_ Brutus: for what it's worth, I can recommend using something like chezmoi if you want to manage dotfiles with git20:32
→ sudomake joined20:33
Brutus i hoped this would explain github a bit to me20:33
i'm already storing them at a cloud provider, trying to do version control with git might be worthwile20:34
_ikke_ well, part of what you are doing it not unique to working with github20:34
It's covered by general guides on how to use git20:34
Frogging101 none of these steps are unique to github, actually20:34
_ikke_ ie, !book20:34
gitinfo There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable20:34
Brutus i know, but it's the setup part that's confusing me. i used to work with subversion software20:34
→ dviola joined20:35
_ikke_ Getting a good grasp of the git basics helps with that20:35
Brutus yeah that's what's confusing me. the first steps. every tutorial i found skips right over that20:35
gitinfo set mode: +v20:36
Brutus let me check out that book20:36
gitinfo set mode: +v20:36
Frogging101 (oh, that's what that does.)20:36
← sidetrack left20:37
Xiti` → Xiti20:42
→ vdamewood joined20:44
← tiin57 left20:44
→ kleisli_ joined20:44
→ tiin57 joined20:48
→ BPL joined20:51
← stealth[] left20:52
← lightstalker left20:52
← sudomake left20:53
→ sudomake joined20:53
→ mbalmer joined20:55
→ pako8128 joined20:55
← mbalmer left20:59
→ Sasazuka joined21:00
← sudomake left21:02
← dhollinger left21:03
→ rafasc joined21:04
→ dhollinger joined21:05
Brutus i got it working21:05
had to create a new github repository, but it's now working with the same commands i used before, but now without error messages21:05
← nessonic left21:08
← mikecmpbll left21:13
j416 osse: ELIF: git log --oneline v2.26.0..v2.26.1 of git.git: https://termbin.com/20zs ; bugs are fixed in the earliest version they need to be fixed in, and then merged into later versions. A bug was fixed based on 2.17.3, creating 2.17.4, and that was later merged into each maint branch for 2.18, 2.19, and so on. The last one being maint for 2.26, creating 2.26.1. Hence all the merges.21:14
← tiin57 left21:14
→ mikecmpbll joined21:14
j416 osse: this gives a good overview of the workflow; turns out it's not that complex afterall https://git-scm.com/docs/gitworkflows21:14
s/ELIF/ELI5/21:15
>.>21:15
← mikecmpbll left21:16
→ mikecmpbll joined21:17
← Xiti left21:17
rafasc j416: --simplify-by-decoration helps a lot to understand "workflows"21:17
j416 osse: related announcement https://public-inbox.org/git/xmqqy2qy7xn8.fsf@gitster.c.googlers.com/21:17
← Lucas_Gray left21:18
→ tiin57 joined21:18
j416 rafasc: trying that, it seems like it trims away anything that can't be decorated?21:18
[...] by omitting commits that are not referenced by tags.21:19
huh.21:19
rafasc More or less, I think it still keeps some undercoated commits, it it determines they're necessary to show what happened.21:19
j416 interesting. In this particular case it didn't help a lot.21:19
rafasc undercotated, ayy21:19
j416 viewing all of the history of git.git though, it becomes clearer with --simplify-by-decoration. Might come in handy at some point. Thanks.21:21
→ Xiti joined21:21
← Naptra left21:21
j416 especially in combination with --graph21:21
rafasc git log --graph --oneline --tags=v* --simplify-by-decoration21:21
the cascade up becomes quite apparent.21:22
j416 yep21:22
sans --tags it's perhaps even clearer21:22
← mikecmpbll left21:23
→ Lucas_Gray joined21:24
← Nyle left21:25
rafasc yeah, assuming your HEAD contains what you are looking for.21:25
← ncdulo left21:26
j416 nod21:26
rafasc j416: even better: git log --graph --oneline --simplify-by-decoration --decorate-refs=refs/tags/*21:27
j416 haha.21:27
it's nice to have the branch names there, though21:27
→ dqx_ joined21:27
j416 but this is clean.21:27
← zhuoz left21:28
rafasc It's easier to see the "cascade up" events over time.21:28
j416 looks pretty much the same here21:29
→ felix_221986 joined21:29
j416 starting from master21:29
→ ncdulo joined21:29
← Lucas_Gray left21:29
→ Nyle joined21:29
rafasc I have a couple of extra remotes, From time to time, I fetch from gitgitgadget, and also have Junio's personal feature branches.21:32
*topic branches21:32
← Borkr left21:33
j416 I see21:33
rafasc The ones he creates from applying the patches sent to the mailing list. So the extra refs make more noise.21:33
j416 nod21:33
rafasc git rev-list --tags --count --no-walk = 162121:34
also a reason why I used the --tags=v* instead of just --tags21:35
j416 check21:35
← feriman left21:35
j416 are those your own tags?21:35
872 extra tags o_O21:36
← dqx_ left21:36
← royal_screwup21 left21:37
rafasc No, I think the bulk of them are from gitgitgadget. It tags the state of the PR when it sends it to the mailing list.21:38
j416 ah.21:38
rafasc I also have fetched from some git devs, so I have some of their tags.21:39
To be honest, I wish git namespaced tags as well. Kinda difficult to understand the origin of some tags.21:40
j416 :)21:42
→ LuxuryMode joined21:42
→ tgamblin_ joined21:43
← tiin57 left21:44
→ royal_screwup21 joined21:44
ukleinek rafasc: the obvious workaround is to only fetch from remotes the stuff you really want.21:45
← shicks2525 left21:45
← tgamblin left21:45
rafasc it's not that simple. If you add a remote because you want a certain patch, fetch from it and delete it right away, you still keep the tags.21:46
→ tiin57 joined21:48
→ shicks2525 joined21:48
rafasc Performing tag cleanup is not easy, or requires you set some things in advanced. E.g. configure git to avoid fetching tags, and fetch them with a custom refspec into a namespace of your choice.21:48
→ Abbott joined21:48
rafasc But that has other issues as well.21:49
→ mbalmer joined21:49
→ sidetrack joined21:50
Abbott I have a repo with two branches that are both just a single commit. I want to consolidate them into one branch where one commit comes after the other. I know that rebase is appropriate here, but which do I need to be checked out in and which order do I enter the arguments?21:50
I always mess up rebase syntax21:50
← felix_221986 left21:50
rafasc Abbott: the rule of thumb is that git commands almost always affect the branch you are currently at.21:50
so if you want commit from branch A to be moved to the top of branch B, you need to checkout A; rebase B.21:51
→ g4570n joined21:51
← covid20 left21:52
→ rsrx joined21:52
← amahl_ left21:53
rafasc (and after that, you may want to checkout B; and merge --ff-only B; to update B to the same place)21:53
← shicks2525 left21:53
rafasc merge --ff-only A;21:53
← mbalmer left21:54
→ greatgatsby joined21:54
greatgatsby Hi. Is there any tool or technique for working with a directory of git repos? For example, if all my local repos are in ~/git/ and I want to know modified or staged files in each repo, is there any easy way to do that?21:57
j416 greatgatsby: there's that android thing called... is it repo?21:58
I suppose it's this https://source.android.com/setup/develop/repo21:59
never tried it.21:59
osse j416: makes sense21:59
j416 \^^/21:59
← fannagoganna left22:00
→ Lucas_Gray joined22:00
greatgatsby j416, check it out now, thanks22:01
j416 \o22:01
→ Es0teric joined22:02
← soft_concrete left22:04
→ jaziz joined22:05
← sidetrack left22:06
← g4570n left22:06
← clara97 left22:08
← BPL left22:10
← iam730 left22:10
← tiin57 left22:14
→ tiin57 joined22:18
→ g00s joined22:18
→ sidetrack joined22:20
← sidetrack left22:24
← ferdna left22:25
→ wrksx_ joined22:25
→ dedowsdi joined22:26
→ g4570n joined22:26
← wrksx__ left22:30
← TomTom left22:35
← CodeSlingerPaul left22:39
→ perrier-jouet joined22:40
→ mbalmer joined22:43
← tiin57 left22:44
← justanotheruser left22:44
← Phylock left22:46
→ tiin57 joined22:47
← mbalmer left22:48
→ expert975 joined22:49
← fission6 left22:50
← tpaper left22:51
← crose left22:55
← refried_ left22:57
← m0viefreak left22:58
← rsrx left23:04
→ justanotheruser joined23:06
← troulouliou_dev left23:06
← dionysus69 left23:10
← infernix left23:11
→ troulouliou_dev joined23:13
← tiin57 left23:14
← troulouliou_dev left23:16
→ tiin57 joined23:17
→ NightQuest joined23:18
← mescalinum left23:22
← Lucas_Gray left23:23
← onizu left23:27
← ar1el left23:27
→ mescalinum joined23:29
← gxt left23:29
→ gxt_ joined23:29
← calcul0n left23:31
← vdamewood left23:31
← royal_screwup21 left23:31
← mescalinum left23:35
→ mbalmer joined23:38
← mbalmer left23:43
← tiin57 left23:44
→ rosco_y joined23:45
→ tiin57 joined23:48
← mauz555 left23:48
→ Apple joined23:49
Apple → Guest821823:49
← tgunr left23:52
← AppleGaming left23:52
→ tgunr joined23:53
← Guest8218 left23:54
→ sidetrack joined23:57

Logs Search ←Prev date Next date→ Channels Documentation