IRCloggy #git 2022-04-08

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.

2022-04-08

← Jong left00:03
← donofrio_ left00:04
← duxsco left00:05
→ donofrio_ joined00:06
→ michigan joined00:06
→ wonko-the-sane joined00:10
← gurkenglas left00:13
→ Jong joined00:14
← jclsn0 left00:14
← DNH left00:16
→ jclsn0 joined00:19
← jclsn0 left00:27
→ tharkun joined00:27
→ jclsn0 joined00:33
→ slaughtr joined00:33
← _9pfs left00:34
← frobnic left00:34
→ frobnic joined00:34
→ _9pfs joined00:34
← Yruama left00:40
→ rgrinberg joined00:43
← jclsn0 left00:43
→ jclsn0 joined00:49
→ skered joined00:49
→ dvanloon0 joined00:51
skered I can look at my scrollback and see myself committing a change with the commit id post save/exiting editor, I can see myself changing branch, going back to the branch where the commit should be.. it's not there.00:51
← dvanloon left00:52
dvanloon0 → dvanloon00:52
skered I can `git diff commitid^..commitid` see the diff... and `git branch -a --contains commitid` returns nothing... and idea what's up?00:52
→ Guest69 joined01:00
← jclsn0 left01:06
← donofrio_ left01:06
→ vysn joined01:06
→ asocialblade joined01:07
← asocialblade left01:07
→ asocialblade joined01:07
→ jazzy2 joined01:09
→ jclsn0 joined01:09
← asocialblade left01:10
→ donofrio_ joined01:11
← jazzy left01:13
← jclsn0 left01:15
→ jclsn0 joined01:17
→ dviola joined01:22
← ano left01:23
→ sa0 joined01:25
← mjorgensen left01:26
← Guest69 left01:28
skered `git reflog` shows 6 goto branch from main, 5 commit, 4 goto main from branch, ... go back to branch.. not there :/01:30
← coaxing left01:30
← subopt left01:30
skered git checkout commitid and a couple other hoops and I'm back to what I would expect... it's like the commit didn't finish writing or something before I switched?01:30
← jclsn0 left01:32
skered The silver lining here... I now know about git-reflog.01:34
→ ano joined01:34
← igemnace left01:35
→ causasui joined01:36
→ jclsn0 joined01:37
← _9pfs left01:41
→ dtux joined01:41
← howdoi left01:42
→ subopt joined01:45
← Jacques left01:45
← pulse left01:45
← ChmEarl left01:50
← fflam left01:51
→ fflam joined01:52
→ Strom joined01:54
← jclsn0 left01:55
→ jclsn0 joined01:57
← sa0 left01:58
← sudoforge left02:00
→ llh joined02:01
→ kaiseisei joined02:02
← jamiejackson left02:05
← jclsn0 left02:07
→ jclsn0 joined02:10
→ R2robot joined02:12
→ geowiesnot joined02:21
← bloody left02:22
← jclsn0 left02:23
→ tyson2 joined02:26
← bertvvvs left02:28
→ jclsn0 joined02:28
← jclsn0 left02:35
→ jclsn0 joined02:40
← rgrinberg left02:43
← Rashad left02:48
← jclsn0 left02:51
→ jclsn0 joined02:51
← jclsn0 left02:59
→ rgrinberg joined03:00
→ jclsn0 joined03:04
← michigan left03:05
← fflam left03:06
→ fflam joined03:08
← causasui left03:10
→ mjorgensen joined03:16
← donofrio_ left03:16
← dviola left03:23
← cdown left03:28
→ Guest69 joined03:31
← Guest69 left03:31
← xkuru left03:37
→ coot joined03:39
← Thanatermesis left03:49
→ Thanatermesis joined03:49
← tyson2 left03:53
→ sa0 joined03:56
← fling left04:03
→ fling joined04:03
← zebrag left04:12
← rgrinberg left04:16
→ hbautista joined04:17
← gast0n left04:19
← sa0 left04:29
b0nn I'm looking for a way to remove a commit from a PR /without/ using rebase (if possible)04:30
ikke b0nn: you cannot remove commits without rewriting history04:30
b0nn I have a script that cherry-picks two commits onto the current branch for local development, but I don't want them to be committed to main04:30
ikke all hashes of following commits needs to change04:30
b0nn hmm, sure, but is there a way to do it with a single command (ie. not interactively)04:31
ikke if they're the last 2 commits, you can reset the branch back 2 commits04:31
b0nn They can be anywhere in the PR04:31
But, they will be together :)04:32
ikke ok, so rebase is good, but not interactive?04:32
b0nn hmm it's a pity that `git revert HASH` will create a new commit04:32
ikke yes04:32
b0nn Yeah - rebase might be doable if I can script it such that other people just click and collect04:33
ikke git rebase HEAD~ --onto HEAD~3 would skip HEAD~1 and HEAD~204:33
you can also set GIT_SEQUENCE_EDITOR env to a sed like command04:34
for interactive rebase04:34
b0nn hmm https://stackoverflow.com/questions/29094595/git-interactive-rebase-without-opening-the-editor04:36
I might be biting off a bit much04:36
ikke I'd say using --onto is easier04:36
b0nn The catch with the GIT_SEQUENCE_EDITOR is the script will need to know where in the PR the commits to drop are04:36
ikke yes04:36
you need to know the commit message or abbreviated hashes04:37
or the line number04:37
b0nn Hmmm I can get that, because when I cherry pick them onto the branch I can grab the last two commits04:37
OR store the number of commits in the branch04:37
ikke if you know the abbreviated commit hashes of the cherry-picked commits, it would be something like GIT_SEQUENCE_EDITOR='sed /^pick (hash1|hash2)/d' git rebase -i <target-branch>04:40
← great_taste left04:41
← fling left04:45
→ fling joined04:45
→ _9pfs joined04:52
← Jong left04:53
→ saroy joined04:53
← _9pfs left04:54
→ madewokherd joined04:58
→ w_lehinsun__ joined04:59
← madewokherd` left05:01
← w_lehinsun left05:01
← w_lehinsun__ left05:08
← dtux left05:20
→ _9pfs joined05:24
→ FH_thecat joined05:25
← Gustavo6046 left05:29
← _9pfs left05:30
→ _9pfs joined05:31
← andypandy left05:42
→ Jong joined05:47
→ andypandy joined05:47
→ mncheck joined05:49
← _9pfs left05:57
← hbautista left06:01
→ dionysus69 joined06:03
→ ikwyl6 joined06:06
→ bkircher joined06:07
→ srinidhi joined06:13
← mjorgensen left06:14
← Misotauros left06:14
jclsn0 → jclsn06:18
→ zeenk joined06:22
jazzy2 → jazzy06:22
← thiago left06:22
osse you can always do GIT_SEQUENCE_EDITOR=myprogram and write something more than a one-liner06:23
might keep you sane in the long run :p06:23
→ dtux joined06:23
← dionysus69 left06:26
→ mexen joined06:28
→ TheCatBehemoth joined06:33
→ dviola joined06:34
→ enkeyz joined06:34
← Betal left06:36
← geowiesnot left06:38
→ epolanski joined06:39
→ palasso joined06:42
← dviola left06:42
→ dviola joined06:44
← dviola left06:45
→ dviola joined06:46
→ rLapz joined06:51
→ EvilDMP joined06:55
→ sa0 joined07:05
→ rfuentess joined07:07
→ MiniontobyPI joined07:14
← Akiva left07:20
← ano left07:20
← mat001_ left07:21
← mat001 left07:21
← vysn left07:22
→ mat001 joined07:25
→ mat001_ joined07:27
→ gnoo joined07:28
→ sv0 joined07:30
→ john_johnk joined07:31
→ ano joined07:36
← dviola left07:38
→ dviola joined07:40
← dviola left07:45
→ dviola joined07:46
→ kenanmarasli joined07:48
← sv0 left07:49
← Thanatermesis left07:53
→ a1eaiactaest joined07:57
→ nyah joined07:58
→ donofrio joined08:00
→ Jacques joined08:01
← skapata left08:01
← dviola left08:05
← Murr left08:05
→ Murr joined08:05
→ dviola joined08:06
← donofrio left08:07
→ Haohmaru joined08:08
→ mncheckm joined08:08
← dviola left08:12
→ dviola joined08:13
→ vitali64 joined08:15
← ano left08:19
← dviola left08:20
→ ano joined08:23
→ dionysus69 joined08:29
← littlebobeep left08:35
← coot left08:36
→ MajorBiscuit joined08:44
→ tsdh joined08:47
→ DNH joined08:52
← EvilDMP left08:56
→ dviola joined08:58
→ EvilDMP joined08:59
→ duxsco joined09:01
→ benoliver999 joined09:03
benoliver999 Say I have two branches, A and B. I deleted a file from branch A and commit. On branch B, I made a update to the file. When I pull in changes from branch A into branch B, what happens to the deleted file?09:05
ikke You'd get a conflict09:05
delete/modify conflict09:05
benoliver999 Makes sense, thanks!09:06
hendry Can I avoid this explicit fetch tracking stuff? i.e. avoid `git branch --set-upstream-to=<remote>/<branch> branch-i-am-working-on`?09:07
← benoliver999 left09:07
ikke hendry: do you create these branches yourself locally, or do you get them from the rmeote?09:08
hendry ikke: i pushed to remote of the branch of the same name, and I think it would be intuitive to `git pull` from the same remote branch name. I'm not sure why git is doing this to me.09:09
behaviour i think was different before right?09:09
← iffraff left09:11
→ tulpa_ joined09:15
→ pretty_dumm_guy joined09:16
← bodiccea left09:19
ikke pull always needed to know what remote to fetch from09:20
either by explicitly specifying it, or by way of tracking information09:20
← dviola left09:21
← kaiseisei left09:22
→ dviola joined09:22
← a1eaiactaest left09:23
← rLapz left09:24
← dviola left09:26
→ gurkenglas joined09:28
← chenqisu1 left09:28
→ A_Dragon joined09:29
Jong What's the word for resolving a warning so it no longer blows up the terminal with garbage text? I trying to writing a commit message09:31
"Resolve verbose warning" ?09:31
→ dviola joined09:32
davve quiet, then silent for nothing09:33
Jong " Resolved warning clogging the logs" ?09:33
davve "Made warnings more quiet"09:34
"Made warnings less verbose"09:34
Jong but I did what the warnings suggested to make the warning go away. I didn't suppress the warning or bypass it09:34
davve oh ok09:35
"Fixed warning(s)"09:35
imo09:35
can mention which it were to be more specific09:35
Jong Thanks!09:35
davve or what types09:35
np. thank you for putting thought into your commit messages :)09:36
Jong Fixed warning clogging terminal09:36
how about that09:36
davve perfect09:36
← dviola left09:37
Jong How does this sound? "Stop validation samples from leaking into training set"09:38
→ iomari891 joined09:39
→ dviola joined09:39
← DNH left09:39
davve it might make sense, but I dont know the context09:40
would prefer your previous comment09:40
i think its kind of important to stay consistent with your messages though, so make it "yours"09:41
Jong not sure if you know python but a colleague accidentally used itertools.chain to chain validation samples into training set.09:41
davve aha09:42
← john_johnk left09:42
← jazzy left09:44
Jong "Remove code causing validation samples to be trained" ?09:44
← dviola left09:45
← Jacques left09:46
→ dviola joined09:47
Jong what do you when you have a series of issues. Like 10, all 5 liner fixes.09:51
Do you really make 10 PRs in such a case?09:51
→ john_johnk joined09:52
→ Guest3354 joined09:53
← dviola left09:56
→ dviola joined09:57
ikke Single PR, multiple commits, or maybe even a single commit09:58
It depends09:58
What these things change09:58
if they have some common goal09:58
If they are completely unrelated, then I would make separate PRs09:58
← Guest3354 left09:59
→ DNH joined10:02
→ coldmountain joined10:03
hendry i have a branch "foo" that i want to merge in, but i need to rewrite the commits. When i `git rebase -i foo` from main, I get totally different commit to what I'm expecting. Really confused.10:03
→ donofrio joined10:03
nedbat hendry: what do you mean by totally different commit? rebasing will rewrite the commits.10:06
hendry: if you are trying to rewrite the commits on foo before the merge, then you probably want `git checkout foo; git rebase -i main`10:06
← epolanski left10:09
→ jimklimov joined10:09
hendry nedbat: ah! yes, i need to rebase from branch10:09
← dviola left10:16
Jong Suppose you have 10 3-liner code changes. All not relevant to each other. Would you make 10 PRs back to back? Wouldn't that appear a bit spammy?10:16
oh okay ikke10:16
good answer10:16
→ dviola joined10:17
→ jimklimov1 joined10:17
nedbat Jong: some maintainers might prefer one PR10:17
← john_johnk left10:20
← jimklimov left10:21
→ lelux joined10:22
← EvilDMP left10:35
→ EvilDMP joined10:40
← dviola left10:41
← fling left10:41
→ dviola joined10:42
→ fling joined10:43
← vishal left10:46
→ vishal joined10:48
← dviola left10:49
→ dviola joined10:50
← hololeap_ left11:00
→ hololeap_ joined11:02
← donofrio left11:06
→ donofrio joined11:06
→ michigan joined11:08
→ igemnace joined11:09
→ tyson2 joined11:16
→ bertvvvs joined11:16
→ john_johnk joined11:23
→ cdown joined11:25
← dviola left11:33
← TheCatBehemoth left11:34
→ dviola joined11:34
← dviola left11:35
→ dviola joined11:37
→ Yruama joined11:45
← fling left11:45
← wonko-the-sane left11:46
→ wonko-the-sane joined11:46
→ bin4ry joined11:47
→ fling joined11:47
→ qn joined11:47
← bin4ry left11:48
→ bin4ry joined11:48
← bin4ry left11:56
→ bin4ry joined11:56
→ littlebobeep joined12:00
← FH_thecat left12:00
← DNH left12:02
← MajorBiscuit left12:04
→ bertiger joined12:04
→ DNH joined12:05
→ MajorBiscuit joined12:05
→ mannequin joined12:07
→ TheCatBehemoth joined12:10
→ coaxing joined12:12
← TheCatBehemoth left12:15
← EvilDMP left12:19
← MajorBiscuit left12:24
→ kaiseisei joined12:25
→ constxd joined12:26
→ EvilDMP joined12:26
→ TheCatBehemoth joined12:27
→ coot joined12:27
← cloaker left12:29
→ cloaker joined12:30
← funhouse left12:44
← Xenguy left12:45
→ gh34 joined12:45
← fling left12:49
→ fling joined12:49
← lelux left12:51
→ bloody joined12:52
→ pulse joined12:58
← dviola left13:00
→ dviola joined13:02
→ xkuru joined13:07
← kaiseisei left13:15
→ Guest24 joined13:20
→ zebrag joined13:31
→ masber joined13:33
← Guest24 left13:33
→ haha joined13:34
← haha left13:34
→ haha231 joined13:34
→ geowiesnot joined13:36
← roadie left13:39
→ thiago joined13:45
→ skapata joined13:46
← skapata left13:46
→ skapata joined13:46
← haha231 left13:50
→ cdown_ joined13:50
← cdown left13:53
cdown_ → cdown13:53
← fling left13:54
→ fling joined13:55
← cloaker left14:00
→ The_Blode joined14:03
← The_Blode_ left14:04
→ gast0n joined14:05
← Murr left14:05
→ Murr joined14:05
→ MajorBiscuit joined14:15
→ wakeup joined14:16
→ Akiva joined14:17
← masber left14:26
→ ChmEarl joined14:26
hololeap_ → hololeap14:27
← tyson2 left14:29
→ tyson2 joined14:33
→ masber joined14:34
→ madewokherd` joined14:38
← mncheckm left14:40
← mncheck left14:40
← madewokherd left14:41
→ sobkas joined14:44
← wonko-the-sane left14:50
← liefer left14:53
→ liefer joined14:55
← pretty_dumm_guy left14:56
→ EvilDMP_ joined14:57
→ pretty_dumm_guy joined14:57
← dviola left14:58
← EvilDMP left14:59
→ Guest73 joined15:01
Guest73 hi15:01
← Guest73 left15:02
→ dviola joined15:02
← dionysus69 left15:03
← liefer left15:03
← pretty_dumm_guy left15:06
← dviola left15:07
→ dviola joined15:12
→ Thanatermesis joined15:12
→ Guest32 joined15:21
← Guest32 left15:21
← dviola left15:23
→ dviola joined15:27
← DNH left15:27
→ LuxuryMode joined15:28
→ dhch joined15:30
→ Gustavo6046 joined15:31
→ DNH joined15:32
→ tchan1 joined15:32
← dviola left15:33
→ ___nick___ joined15:33
← dhch left15:33
→ dviola joined15:33
← masber left15:34
← iomari891 left15:34
→ masber joined15:36
← tchan left15:36
← tyson2 left15:37
→ great_taste joined15:39
→ oncall-pokemon joined15:39
← analoq left15:41
→ wonko-the-sane joined15:41
→ analoq joined15:42
← rfuentess left15:44
← michigan left15:48
→ JanC joined15:50
→ jinsun joined15:51
wakeup I have 2 commits in my repo15:52
← coot left15:52
wakeup I am trying to merge them into 115:52
so I am using git reset --soft HEAD~215:52
then git commit15:52
then git push -f15:53
but the first command doesn't work15:53
"fatal: ambiguous argument 'HEAD~2': unknown revision or path not in the working tree.15:53
Use '--' to separate paths from revisions, like this:15:53
'git <command> [<revision>...] -- [<file>...]'15:53
"15:53
git reset --soft HEAD~1 does nothing15:53
→ impermanence joined15:54
← sobkas left15:55
→ iffraff joined15:55
wakeup well, I just did this: https://stackoverflow.com/a/9254257/392411815:55
→ sobkas joined15:56
← dviola left15:56
← EvilDMP_ left15:57
wakeup but I would appreciate any explanation of why the other command didn't work15:57
→ dviola joined15:57
← dviola left16:02
← madewokherd` left16:02
→ dviola joined16:04
→ foul_owl joined16:04
← ___nick___ left16:05
→ ___nick___ joined16:07
← ___nick___ left16:08
→ ___nick___ joined16:10
→ iomari891 joined16:10
thiago wakeup: you can't reset to before the first commit16:11
but if you want to merge the two, you don't have to. Reset to the first, then commit --amend16:11
← geowiesnot left16:13
→ EvilDMP joined16:15
← masber left16:17
← vicecea left16:17
→ vicecea joined16:18
→ Guest69 joined16:23
→ madewokherd joined16:25
→ rgrinberg joined16:27
← DNH left16:32
→ bodiccea joined16:33
← duxsco left16:35
← saroy left16:36
← wakeup left16:36
→ treefrob joined16:38
← Haohmaru left16:39
→ duxsco joined16:40
→ geowiesnot joined16:41
← MajorBiscuit left16:44
→ rsx joined16:44
→ odoood joined16:49
← dviola left16:53
→ menace joined16:53
← menace left16:53
→ menace joined16:53
menace → Deknos16:53
→ glik22_ joined16:56
→ dviola joined16:57
← rsx left16:57
glik22_ i have branch A that was created off master, commits to A, then created branch B off of A, a single commit to B and more commits to master. i want to rebase B onto master, so from B i run `git rebase --onto master A B`. this succeeds but then my B commit is wiped out. i can get it back with `git reflog` and `git revert`. what's the correct rebase command?16:57
→ hqdruxn08 joined17:00
rewt by "B commit" you mean the commit only referenced by branch B that you moved when you rebased? that's expected... "correct rebase command" would depend on what you're expecting17:01
you could make another branch (C) at B so when you rebase B, C still points to the original commit17:01
thiago man git-rebase has nice graphs17:01
gitinfo the git-rebase manpage is available at https://gitirc.eu/git-rebase.html17:01
thiago always refer to them when using --onto17:01
three-branch --onto is confusing17:01
glik22_ rewt: i'd like my brach B to have everything in master and then only the single commit that is on the B branch. no commits from A17:03
rewt then what you did seems right17:04
thiago yep17:04
are you sure the rebase succeeded?17:04
try this: git checkout master && git cherry-pick B17:04
does that succeed?17:04
← dviola left17:05
→ bambanxx joined17:05
glik22_ the rebase succeeded but i still have changed from A. git checkout master && git cherry-pick B worked17:05
thiago the command looks right and should do the same thing as that cherry-pick17:05
so it's likely the problem is in some part of the situation that you didn't describe17:06
glik22_ i think i can get around this by going to master, creating a new branch B', and cherry picking the commit i want from B.17:07
thiago yes17:07
glik22_ i must have done something weird that i can't recall. this should fix it though.17:07
→ dviola joined17:07
glik22_ ok, cool. thanks for brainstorming with me17:08
← dviola left17:13
→ dviola joined17:14
← mexen left17:15
← hqdruxn08 left17:16
→ pretty_dumm_guy joined17:18
← zeenk left17:19
← dviola left17:19
→ hqdruxn08 joined17:19
→ dviola joined17:26
← hqdruxn08 left17:26
← odoood left17:28
→ gildasio1 joined17:31
← OnkelTem left17:34
→ OnkelTem joined17:37
→ han-solo joined17:39
← dviola left17:41
→ orbyt joined17:42
← OnkelTem left17:44
→ cvmn joined17:45
→ tyson2 joined17:45
→ dviola joined17:46
→ OnkelTem joined17:46
← dviola left17:50
→ hqdruxn08 joined17:51
→ dviola joined17:52
← tchan1 left17:54
→ tchan joined17:59
← donofrio left18:01
← GNUmoon left18:03
→ funhouse joined18:05
← gildasio1 left18:09
→ donofrio joined18:11
← AnapodoPsalidaki left18:14
→ GNUmoon joined18:16
→ jianwu joined18:16
← EvilDMP left18:21
→ structoha joined18:21
→ EvilDMP joined18:21
← EvilDMP left18:21
→ EvilDMP joined18:25
→ junebug joined18:27
← LuxuryMode left18:28
← EvilDMP left18:29
← geowiesnot left18:31
→ coot joined18:32
→ Misotauros joined18:32
← dviola left18:33
→ Jacques joined18:33
→ dviola joined18:34
→ DNH joined18:36
→ sudoforge joined18:37
→ structoha1 joined18:39
← dviola left18:40
← structoha left18:40
← gildasio left18:45
→ dviola joined18:45
→ gildasio joined18:45
→ cliluw joined18:47
← cliluw left18:47
→ cliluw joined18:48
← gildasio left18:48
→ gildasio joined18:48
← rgrinberg left18:49
← iomari891 left18:53
← Jacques left18:54
→ beencubed joined18:55
→ ThorMojito joined18:56
→ hxztnxt joined18:57
← orbyt left19:00
→ reset joined19:02
→ dionysus69 joined19:04
→ epolanski joined19:12
← iffraff left19:12
← bambanxx left19:13
→ bambanxx joined19:14
← qn left19:17
← duxsco left19:18
← vitali64 left19:22
← dviola left19:23
→ Betal joined19:23
← bambanxx left19:24
→ dviola joined19:25
→ texasmynsted joined19:25
texasmynsted If I have say five commits on a new feature branch and want to squash them to a single commit, would this work `git rebase -i --root`?19:27
← dviola left19:27
← hxztnxt left19:32
→ _9pfs joined19:32
ikke texasmynsted: You probably don't want to use --root19:33
← _9pfs left19:33
→ _9pfs63 joined19:34
ikke unless your repository only has 5 commits19:34
texasmynsted yeaaahhh. That was NOT what I wwanted19:34
→ skapate joined19:35
← skapate left19:35
→ skapate joined19:35
← sobkas left19:35
ikke git rebase -i <base-branch>19:35
texasmynsted base-branch being the commit where my feature branch split from?19:36
_9pfs63 → _9pfs19:36
ikke does not necessarily have to be exactly that commit19:36
→ duxsco joined19:36
ikke if you created the feature branch from master, than specifying 'master' suffices19:37
texasmynsted Is there not a way to automagically identify that point? The branch my feature branch split from has MANY more commits on it now.19:37
_9pfs How do I remove files from the first commit? I did it once with `git filter-branch`, but is there a better way? I accidentally committed a file too big for Github, and had to remove it.19:37
texasmynsted I will just try and see what happens19:37
I am sure I can abort or reset19:37
ikke _9pfs: git rm --cached && git commit --amend19:37
_9pfs I removed the file already, just wondering what to do for the future.19:37
← skapata left19:37
→ jazzy joined19:37
_9pfs ikke: I'll try that in the future.19:37
texasmynsted ah. :-) `git rebase -i main` was exactly what I needed. Thank you19:38
← junebug left19:38
→ wakeup joined19:38
→ dviola joined19:39
→ bambanxx joined19:40
→ vicfred joined19:42
← coldmountain left19:42
← vicfred left19:43
← dviola left19:44
→ vicfred joined19:45
→ dviola joined19:47
← wakeup left19:47
← cvmn left19:49
← jianwu left19:50
texasmynsted sigh. I guess there is no way to git rebase --no-verify --continue19:50
← dviola left19:50
texasmynsted shouldn't `git rebase --no-verify -i main` prevent the `prepare-commit-msg` hook from firing?19:55
ikke I think only the pre-commit hook19:56
→ dviola joined19:56
← skapate left19:57
→ AbleBacon joined20:01
← dviola left20:02
← duxsco left20:02
→ dviola joined20:03
← structoha1 left20:03
← dviola left20:03
→ structoha1 joined20:04
← ___nick___ left20:04
→ dviola joined20:04
← Murr left20:04
texasmynsted This was a life saver -> https://gist.github.com/MrSwed/4bb4e0a8c2a2f35827e6d9b41064258220:04
→ Murr joined20:05
→ rgrinberg joined20:09
← structoha1 left20:16
← han-solo left20:18
→ Lord_of_Life_ joined20:19
← Lord_of_Life left20:21
Lord_of_Life_ → Lord_of_Life20:22
← bambanxx left20:27
→ euouae joined20:31
euouae Hello if I have a shallow clone with `git clone --depth 1 ...` and then later (say after a lot of commits) I do a `git feth --depth 1`, do I get the latest snapshot of the project?20:32
s/feth/fetch20:32
rewt man git fetch says "If fetching to a shallow repository created by git clone with --depth=<depth> option (see git-clone[1]), deepen or shorten the history to the specified number of commits."20:33
gitinfo the git-fetch manpage is available at https://gitirc.eu/git-fetch.html20:33
BtbN Keep in mind that fetch alone won't get you anything20:34
you still need to merge/checkout20:34
euouae I don't understand the manpage though20:34
what is it saying?20:34
rewt which part?20:34
euouae the shortening of the history20:34
I get the latest snapshot but the history is incomplete, right?20:35
rewt it shortens the history20:35
the history is incomplete to begin with because of the shallow clone20:35
shorten history means chop off older commits to make the history shorter20:35
euouae so if I'm in commit A with a shallow copy and there's A -> B -> C -> D and I fetch, do I grab commit D's snapshot?20:35
and my history looks like A -> D ?20:35
rewt C -> D20:36
euouae C -> D? not just D?20:36
rewt maybe just D20:36
tias20:36
euouae rewt: I don't care so much for the history, but is it true that I grab the latest snapshot of the project?20:36
BtbN I'm not sure what you mean by "snapshot". You are aware that fetch does not modify the working tree?20:37
rewt but not A -> D for sure, because A is not D's parent20:37
euouae BtbN yes, let's say I merge20:37
oh, I need to merge manually?20:37
no, `git pull --depth 1` works. Nevermind the fetch, let's say I do the pull20:38
rewt can you even merge if there is no merge-base?20:38
euouae pull has a --depth20:38
rewt pull is fetch + merge20:38
euouae oh, does `git pull --depth 1` upgrade from A to A -> B?20:38
is that what it means by "deepening the history"?20:38
← rgrinberg left20:39
rewt `git pull -depth 1` means `git fetch --depth 1; git merge`20:39
pull is just a shortcut for typing fetch then merge20:39
euouae I agree20:39
I'm only curious about the behavior of --fetch 120:39
I can test it, `git clone -b ` allows to clone an early commit20:40
Hm, after `git clone <repo> --depth 1 -b <early_tag>` if I try to pull I get `git pull --depth 1` -> already up to date20:42
and if I try `--depth 2`, I just obtain more of past history20:43
→ bambanxx joined20:50
→ rgrinberg joined20:51
rewt well, yeah, git clone --depth 1, followed by git fetch --depth 1 has nothing to do because history is already 1 deep20:52
← dviola left20:52
euouae I was hoping to obtain shallow copies of the most recent stuff every time20:52
BtbN The manpage of fetch sounds like that's exactly what it does?20:53
euouae no, on a shallow copy it just grabs more history20:53
from the past20:53
→ dviola joined20:54
→ orbyt joined20:55
BtbN Doesn't look like it does to me20:55
→ hbautista joined20:55
BtbN for a "shallow update" you probably want to fetch --depth 1 + reset anyway20:55
merge just makes no sense20:55
→ junktext_ joined20:56
_9pfs → _20:56
_ → _9pfs20:56
euouae I don't know what you mean by + reset20:56
You can test the behavior on a repo with `git clone <repo> -b <old_tag> --depth 1` followed by `git fetch --depth 1` for example20:57
or pull20:57
BtbN reset instead of merge20:59
← dviola left21:00
→ dviola joined21:01
euouae hm... git fetch only goes backwards on a shallow clone I checked it21:01
BtbN Works fine for me21:01
fetch + reset --hard21:01
euouae it does? doesn't it just retrieve older commits when you fetch?21:02
BtbN Why would it?21:02
euouae that's what it does here21:02
BtbN Hmm tell you what. Let's just both be wiser and continue without solving this mystery :P it will probably be a better use of our time21:02
I'll agree that you're right and I'm probably the super noob who can't get anything right anymore :P21:03
BtbN You have to understand that a merge makes no sense without history21:03
euouae I agree21:03
← donofrio left21:04
← bambanxx left21:04
BtbN I'm literally looking at my git output right now, and see no issue. So I don't understand what issue you are seeing.21:04
git pull won't work21:04
you have to manually fetch and then hard reset to what you fetched21:04
euouae when I fetch it only goes backwards21:05
e.g. if I'm on 0.6.1. and I do a `--depth 10` (say) I might end up in 0.6.021:05
that's what I see on `git log`, history from 0.6.0 to 0.6.121:05
it doesn't go "up" to 0.6.2. Does that make sense?21:06
→ donofrio joined21:06
BtbN https://bpa.st/G4JA21:06
It doesn't seem to particularily care if it goes up or down to me21:06
← dviola left21:06
→ bambanxx joined21:07
bn_work hi, I have created a new branch via `git branch foo`, committed some changes & pushed but when I pull from another machine it doesn't show up when I do `git branch --list -a`, why? Did I need to explicitly set the branch I created to track remote? If so, how do I do that after the fact?21:08
← bambanxx left21:08
→ dviola joined21:08
→ skapata joined21:09
← skapata left21:09
→ skapata joined21:09
bn_work Do I need to use `git branch -u` while on the original machine and re-push?21:09
BtbN bn_work, "git branch" just creates the branch, but does not check it out or otherwise does anything until you check it out21:09
euouae BtbN right but you're using the extra command line argument `v5.17`... that's why.21:09
If you tried `git fetch --depth 1` you'd get what I'm talking about21:09
BtbN euouae, you can also put "origin master" there if you just want latest21:09
bn_work BtbN: I checked it out (or `git switch foo`), sorry I forgot to mention that21:10
before committing and pushing changes21:10
BtbN But you need to give it a reference that points to a specific commit. Otherwise the whole undertaking makes no sense.21:10
bn_work BtbN: what do you mean?21:11
I thought when one "branches" it takes current HEAD of current branch21:11
BtbN If you don't tell it what to fetch, while being on a pretty much detached-head state... what's it supposed to do?21:11
wait, wrong person21:11
← dionysus69 left21:11
BtbN bn_work, yeah, that's what it does by default21:12
euouae I told you what it does :P BtbN21:12
BtbN Well, you got your solution then21:12
bn_work BtbN: so what am I doing wrong? why wasn't the branch pushed?21:13
BtbN How did you push it?21:13
parsnip --depth N goes backwards. so specify the branch that is forwards?21:13
bn_work BtbN: just `git push`21:13
parsnip --depth 1 will go forward if the branch has moved forward21:13
← dviola left21:13
euouae that makes sense21:13
bn_work when I do a `git log --all -graph`, I don't see an `origin/foo` so I'm assuming it never got pushed21:14
BtbN Yeah, if you make a new branch, it has no upstream, so git push won't do much21:14
→ structoha1 joined21:14
→ fillx joined21:14
BtbN you initially need to git push -u origin bla21:14
← ThorMojito left21:16
→ dviola joined21:16
→ ThorMojito joined21:16
→ bambanxx joined21:16
← kenanmarasli left21:16
bn_work oh wait, I switched back to `develop` prior to pushing, I thought `push` pushes all...21:17
ok, that seems to have worked, seems like a pain21:19
is there any way to make that process more efficient?21:19
ie: default to all branches pushed + new branches are automatically `git push -u origin ...` ?21:20
← palasso left21:20
euouae which program are you using? git?21:20
← dviola left21:21
euouae git has aliases you can write21:21
_9pfs → hello-smile4821:21
← DNH left21:21
→ hxztnxt joined21:22
euouae git config --global alias.pushall '!git remote | xargs -l git push --all'21:22
← hxztnxt left21:23
← coot left21:23
→ distantsail joined21:25
euouae BtbN thanks for the help. I was a bit slow but I get it now.21:26
← euouae left21:26
→ dviola joined21:27
→ DNH joined21:27
← fillx left21:29
bn_work yes, I'm aware of aliases, can I use variables in it though?21:32
← dviola left21:33
bn_work re. `git config --global alias.pushall '!git remote | xargs -l git push --all'`: why is the `git remote |` part necessary? is `origin` not always guaranteed to be the "name"?21:33
← madewokherd left21:35
→ dviola joined21:36
← dviola left21:41
→ dviola joined21:44
← gh34 left21:44
← TheCatBehemoth left21:46
j416 you can name your remotes whatever you like.21:49
← structoha1 left21:52
← rgrinberg left21:52
← distantsail left21:52
← pretty_dumm_guy left21:53
cbreak (I recommend not calling your remote "master", or "HEAD", that'd be confusing :P )21:53
→ feriman joined21:59
← tulpa_ left22:00
← feriman left22:01
rewt bn_work, that alias looks like a push to all remotes... if you have more than 1, they can't all be named origin22:02
→ feriman joined22:02
→ madewokherd joined22:03
← dviola left22:03
← ThorMojito left22:04
→ dviola joined22:04
← Murr left22:04
→ Murr joined22:05
← hello-smile48 left22:08
← feriman left22:09
← dviola left22:09
← orbyt left22:10
→ Xenguy joined22:11
← donofrio left22:17
→ dviola joined22:18
→ _9pfs joined22:21
← srinidhi left22:22
← _9pfs left22:22
→ _9pfs22 joined22:22
→ feriman joined22:23
← feriman left22:25
_9pfs22 → _9pfs-libera22:25
mat001_ Hey. I have this command in travis.yaml file: git init && git pull https://github.com/company/app.git. This will pull master branch.22:25
How can I pull a remote branch?22:25
→ feriman joined22:26
OnlineCop `git submodule sync path_to_submodule; git submodule update --init path_to_submodule` perhaps?22:27
← _9pfs-libera left22:27
→ ThorMojito joined22:28
OnlineCop Since my project has 4 submodules (and I only want 3 of them), mine looks like `git submodule sync path1/submod1 path2/submod2 path3/submod3` ...22:28
→ _9pfs-libera joined22:28
← _9pfs-libera left22:29
→ _9pfs-libera82 joined22:29
→ Unknown joined22:33
← dviola left22:33
← DNH left22:33
← feriman left22:37
OnlineCop Sorry, you just wanted a different branch and not a submodule? I misunderstood.22:37
← bambanxx left22:38
→ feriman joined22:39
rewt mat001_, you probably want `git clone` instead of `git init && git fetch && git merge`22:40
→ michigan joined22:41
rewt and that will also get all remote branches22:41
→ bambanxx joined22:43
← eaii^ left22:44
← vdamewood left22:44
← bambanxx left22:45
→ humanBird joined22:46
← mat001 left22:48
← ThorMojito left22:48
→ Guest70 joined22:48
→ dviola joined22:49
← feriman left22:49
→ feriman joined22:49
→ mat001 joined22:52
bn_work rewt: ah, ok22:53
← Guest70 left22:54
bn_work rewt: so any idea how to use a variable to create an alias for `git push -u origin <new_branch_name>` ?22:54
← dviola left22:55
→ dviola joined22:56
→ cation21 joined23:01
← w0ng left23:02
← dviola left23:02
→ dviola joined23:04
← Unknown left23:04
→ Unknown joined23:04
cation21 → cation23:04
→ Shentino joined23:08
← enkeyz left23:09
→ w0ng joined23:09
← sa0 left23:09
← dviola left23:10
← feriman left23:10
→ sa0 joined23:10
→ causasui joined23:10
→ dviola joined23:11
Unknown test23:11
daddy icles23:12
Unknown selci23:14
xD23:14
← sa0 left23:15
← Shentino left23:16
← dviola left23:17
→ Shentino joined23:18
→ zumba_addict joined23:21
→ lgc joined23:22
← lgc left23:22
← john_johnk left23:26
→ john_johnk joined23:27
← _9pfs-libera82 left23:28
→ dansan joined23:36
rewt bn_work, if you have the branch checked out, you can just `git push -u`23:36
← Akiva left23:37
→ structoha1 joined23:38
rewt otherwise you're looking at something like `git push -u branch:branch`23:39
→ dviola joined23:43
→ blaklistd joined23:47
← dviola left23:51
→ _9pfs joined23:53
← zumba_addict left23:53
← blaklistd left23:53
→ dviola joined23:54
→ ThorMojito joined23:57

Logs Search ←Prev date Next date→ Channels Documentation