IRCloggy #git 2024-11-01

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.

2024-11-01

cedd is this a typo in 'man git-rebase'? On line 124 it says the command 'git rebase --onto topicA~5 topicA~3 topicA' produces a rebase of 5-2-1-0. I was expecting it to produce 5-3-2-1-0. The 2nd commit in that list is excluded?00:02
gitinfo the git-rebase manpage is available at https://gitirc.eu/git-rebase.html00:02
← MIsAn left00:07
→ lacrymology joined00:09
larryv are you referring to this https://github.com/git/git/blob/master/Documentation/git-rebase.txt#L168-L18700:10
← lacrymology left00:10
→ crossby1004 joined00:17
← Nixkernal left00:29
cedd larryv: yea00:43
larryv ok well "git rebase ... topicA~3 topicA" selects commits that are in topicA but are not reachable from topicA~300:47
so topicA~3 itself is not rebased00:49
→ CycoCid joined00:49
cedd larryv: i see, thanks00:49
larryv np00:49
← Dreadnaught left00:51
→ Gary_Sinise joined00:52
← P1ro left00:52
→ P1ro joined00:53
← crossby1004 left00:57
← Turn_Left left01:01
← hdh left01:05
→ crossby1004 joined01:21
→ specing joined01:23
← w4 left01:38
→ Dreadnaught joined01:43
← pedahzur left01:44
→ thiago joined01:46
← CycoCid left01:46
→ w4 joined01:47
← gsi_ left01:50
→ gsi joined01:52
← ChmEarl left01:54
→ rainn92675560720 joined01:55
→ cabazon joined01:59
← aristaware443921 left02:00
← th3voic3 left02:00
← chiselfuse left02:00
← gildasio left02:00
← tolnanwo left02:00
→ th3voic3 joined02:00
→ tolnanwo joined02:00
→ gildasio joined02:00
→ chiselfuse joined02:00
→ aristaware443921 joined02:01
← ptrck5456 left02:02
→ ptrck5456 joined02:02
← n-st left02:03
→ n-st joined02:04
← Karyon left02:04
→ Karyon joined02:04
← insidesources left02:14
← crossby1004 left02:20
← mns left02:21
← davispuh left02:21
← TomyLobo left02:26
← hexnewbie left02:34
→ hexnewbie joined02:40
← lucasta left02:48
→ mns joined02:52
← gildasio left03:00
→ gildasio joined03:01
← cedd left03:02
→ Vonter joined03:03
← P1ro left03:16
→ P1ro joined03:17
← pks left03:20
→ pks joined03:21
← ferdna left03:25
← rainn92675560720 left03:43
→ hwpplayer1 joined03:44
← hwpplayer1 left03:47
→ hwpplayer1 joined03:51
→ uco3 joined03:51
← uco left03:52
uco3 → uco03:52
← kurahaupo left04:02
← hwpplayer1 left04:05
← bloody left04:09
← dlh404 left04:12
← bket left04:13
sybariten Nothing obvious that is wrong with this line in a gitignore file, i presume? weechat/.weechat_julius/script/plugins.xml.gz04:14
→ bket joined04:15
sybariten in short - the aforementioned file shows up as modified when i do git status -uall . I've been trying to write a wildcard for it, but the file doesn't go away from the git status output. Then i figured i'd copy the exact line verbatim as git shows it (no wildcards, just this specific file), but it still doesnt disappear from the git status output.04:16
→ yuljk joined04:16
← skapata left04:20
→ rosco joined04:24
→ hwpplayer1 joined04:27
larryv is it already under version control04:28
← cpackham left04:33
→ alamak joined04:36
pikapika_lunar → militantorc05:02
← mrpops2ko left05:18
← Gary_Sinise left05:18
← adamtajti left05:21
→ Guest66 joined05:22
← chuanshi left05:27
→ chuanshi joined05:28
← rosco left05:32
← chuanshi left05:35
→ chuanshi joined05:37
← g2anj left05:37
→ g2anj joined05:38
Guest66 I think here is dumb question that I just don't know the correct terminology to find the answer with google:05:45
I started with something like this:05:45
A----B----C----E----G05:45
\----D----F05:45
then I ran git checkout <commit F> and git rebase -i <commit G>, dealt with merge conflicts, and got to somewhere like this:05:45
A----B----C---ED----G----F05:45
(marked <commit ED> to signify merge conflicts were dealt with there)05:45
now, some other people have done work and I am ending up with something like this:05:45
A----B----C----E-----G----H----I05:45
A----B----C---ED----G----F05:45
is there some way to merge only the new commits (commits H and I in the example) to my branch which currently has most recent commit of F?05:45
when I try to rebase or merge I get all the merge conflicts that I already dealt with appearing again, which in this case is more than 6005:45
I also know with the above example it would be easy to cherry-pick commits H and I, but in my actual case there is more than 20 commits that would need to be cherry picked05:45
← omero left05:47
← Notgnoshi left05:48
chuanshi → yxt05:49
→ omero joined05:49
yxt → chuanshi05:50
→ YuGiOhJCJ joined05:51
larryv git rebase --onto F G I05:51
← thiago left05:51
larryv where G is the "original" G, not your G.05:52
ZLima12 ^05:52
Guest66 thanks, seemed to do exactly what I wanted05:56
larryv great05:57
Guest66: hm, although that actually changes the branch leading to I. not sure if that's what you want06:00
→ cedd joined06:00
← Guest66 left06:05
← cedd left06:05
→ PKDrinker joined06:08
→ Guest66 joined06:09
Guest66 power outage06:11
but doesnt matter to me now which commit is at the head, just that H/I/F are all there06:12
← smallville7123 left06:13
Guest66 needed to fo git -f [branchname] [commit I] also06:13
← PKDrinker left06:22
LionSkeleton → LadyLioness06:23
← Guest66 left06:24
→ adamtajti joined06:31
→ kale joined06:32
→ smallville7123 joined06:40
← FH_thecat left06:41
← cabazon left06:41
← alamak left06:43
← newpy left06:44
← aqtn left06:45
← uco left06:57
→ uco joined06:59
→ cedd joined07:01
← larryv left07:02
← JamesMowery43 left07:05
← cedd left07:07
→ xx joined07:21
→ mrpops2ko joined07:22
← mrpops2ko left07:25
→ user_oreloznog joined07:28
→ jfsimon1981_b joined07:29
→ rosco joined07:34
← Carwash0324_ left07:35
→ RiFo joined07:39
← Vonter left07:44
→ Lord_of_Life_ joined07:44
← Lord_of_Life left07:46
Lord_of_Life_ → Lord_of_Life07:47
→ mrpops2ko joined07:50
→ obrien joined07:51
← rosco left07:53
← obrien left08:02
→ cedd joined08:03
← hwpplayer1 left08:07
← cedd left08:07
← memset left08:15
→ memset joined08:16
→ FH_thecat joined08:18
→ theyneversleep joined08:45
→ TinoDidriksen joined08:47
→ zloi_user joined08:50
← akibageek left08:55
→ akibageek joined08:56
→ jaredce joined09:03
← jaredce left09:07
→ iomari891 joined09:08
→ jaredce joined09:09
→ saavedro5 joined09:09
← saavedro left09:11
saavedro5 → saavedro09:11
← tolnanwo left09:13
→ warmana joined09:15
← Betal left09:16
→ tolnanwo joined09:16
← specing left09:26
← mikap left09:26
→ mika joined09:26
→ t0mm13b_ joined09:36
← t0mm13b left09:36
→ stenno joined09:41
← stenno left09:41
→ stenno joined09:41
→ onizu joined09:46
← t0mm13b_ left09:58
→ t0mm13b joined09:59
→ cedd joined10:05
← cedd left10:09
→ epicout joined10:10
→ zmt00 joined10:14
→ zmt01 joined10:15
← swamp_ left10:16
→ jmd joined10:17
← peirik_ left10:19
← zmt00 left10:19
→ peirik_ joined10:19
→ dlh404 joined10:20
← rossome left10:27
→ rossome joined10:28
← rossome left10:33
← gnoo left10:35
← epicout left10:37
← ZenWalke1 left10:39
→ ZenWalker joined10:41
→ lacrymology joined10:41
← ueberall left10:44
→ ueberall joined10:46
→ cedd joined10:46
← cedd left10:51
← peirik_ left11:03
→ peirik_ joined11:04
→ Vonter joined11:14
→ rossome joined11:21
← rossome left11:24
→ rossome joined11:27
← rossome left11:33
→ Betal joined11:33
→ gnoo joined11:34
→ rossome joined11:37
← rossome left11:38
→ rossome joined11:45
→ monkfish joined11:50
← rossome left11:51
→ rossome joined11:53
← jmd left12:00
Betal what can cause a git pull to do a merge commit? I just wanted it to fetch/merge remote (actually it created a merge commit even with fetch and merge)12:04
maybe because I have '[merge] ff = no' in git config12:05
yes, `git merge --ff-only` did works12:06
→ cedd joined12:07
← gnoo left12:07
← peirik_ left12:08
→ peirik_ joined12:08
BtbN If you have ff = no configured, it'll create a merge commit, yes12:10
My default has pretty much become to always pull --rebase. Don't think getting a merge commit is _ever_ what I want.12:10
← iomari891 left12:18
← m0viefreak left12:23
← akibageek left12:24
→ akibageek joined12:24
→ jmd joined12:25
← Non-ICE left12:30
→ gnoo joined12:32
→ blatant joined12:32
← lacrymology left12:41
→ Momentum joined12:43
→ davispuh joined12:43
Momentum Hi12:44
is it normal that git asks for passphrase in `git pull` before asking you to specify remote and branch?12:45
that seems a bit wrong12:46
bremner what command are you running?12:46
(exactly)12:46
Momentum `git pull`12:47
bremner if you check the man page, you will see other arguments are required12:48
oh, it depends on your configuration I guess12:48
Momentum yes, but it seems pointless to ask for passphrase before determining the branch to pull from, no?12:49
bremner Momentum: if you look in man git, there are various GIT_TRACE variables if you want to debug further12:49
gitinfo Momentum: the git manpage is available at https://gitirc.eu/git.html12:49
Momentum Thank you. I will look into that12:50
← YuGiOhJCJ left12:54
Momentum i guess it performs `git fetch` first12:58
so maybe that makes sense12:58
→ rsx joined13:01
→ FreeBDSM joined13:07
skered We can use ^ on a commit to ref the pervious commit. Is there something like that for tags?13:12
ex: git diff tag^..tag13:12
osse skered: no13:13
skered The diff between tag and last tag.13:13
Dang.13:13
← jaredce left13:13
← TinoDidriksen left13:14
→ hwpplayer1 joined13:21
← blatant left13:22
→ newpy joined13:22
→ blatant joined13:22
→ jaredce joined13:26
gsish there is no order in tags, how is git supposed to know what your idea of "the previous tag" is?13:28
the syntax above might be accepted, but most probably resolves the tags to commits and then uses the commit before that other commit that was tagged13:29
(have guessed there, haven't tried it myself)13:29
← jaredce left13:30
osse that would be true13:33
→ bloody joined13:35
→ Left_Turn joined13:41
skered gsish: I would guess by date and branch The example is accepts but it would be the previous commit to whatever tag refs.13:45
But given that it doesn't even exist I'm just talking about design13:45
But maybe if something like this existed maybe it would be called called something different.13:46
← itsuki left13:46
→ jensen1 joined13:46
skered maybe it's called a release where it's just a tag but specific to a branch.13:47
→ memset_ joined13:47
skered ... and ther is git-release ... /me reads13:48
gsish ask three persons about their idea of a software release or whatever they use tags for, and you'll get at least four answers13:48
skered Seems like it's just a wrapper around commit, tag, and push? Not really it13:49
→ itsuki joined13:49
gsish the date of creating a release need not relate to what you consider related tags13:49
imagine a team releases 1.0, 1.1, 2.0, 1.2, 2.1, etc -- what's their sort order?13:49
skered I would expect those to be in different branches.13:50
gsish listen to yourself :) this is _your_ idea, not everybody's13:50
← memset left13:51
skered is the ideas man too.13:51
skered So it kinda works out.13:52
gsish you might script something along the lines of "the nearest commit in the specified commit's ancestry that was tagged, too"13:58
have fun when you try this on graphs that are not strictly linear13:58
→ jaredce joined13:59
gsish it's much easier and more robust to expect the developer to just specify which points to address in the graph -- they should know, it's their project13:59
not the tool's job to guess, and outsmart its users13:59
another option could be to have a (user maintained) text file that lists references in the order that matches that user's idea of what their order is14:01
some may call this a ChangeLog, still doesn't exactly match the "stable releases", "customer release", "in house test versions, 'to get released' unless unfit", etc situations14:03
← memset_ left14:03
→ memset joined14:03
gsish chronological order of release dates is rather weak a condition, and most probably wrong anyway14:03
needs to reflect that a graph of potentially arbitrary complexity is involved14:04
as far as DAGs go, only loops are forbidden, other degrees of complexity are free IIUC14:05
← jaredce left14:05
gsish it might be easier to convert all of the software developing world to have _one_ scheme of how to name releases14:06
← kale left14:06
gsish (and see the complexity of your "version number ordering" libraries, their multitude, and their limitations)14:06
← mingdao left14:09
→ mingdao joined14:10
→ jaredce joined14:16
→ cousteau joined14:17
cousteau Hi! Just installed git on a new machine. Why is the default editor nano instead of vi? Which option / env var controls this? I don't have any core.editor setting enabled.14:18
(this is an Ubuntu-based distro so maybe they just went and modified git so that it uses nano by default)14:18
osse cousteau: is EDITOR set ?14:20
cousteau neither VISUAL, EDITOR, GIT_EDITOR14:20
well, actually let me check14:20
yep, unset14:21
cousteau wanted to make sure they were not "set to the empty string"14:21
← jaredce left14:21
cousteau This is not the first thing I see this; I think it's an Ubuntu thing14:22
osse The hardcoded default editor can be set at compile time14:22
cousteau OTOH, that's a good thing because the default vi in Ubuntu is the "basic" vim which is rather unusable IME14:22
(but I already installed the "full" one)14:23
osse: makes sense. How can I check that? (Can I?)14:23
osse The source of the ubuntu package I suppose14:27
→ jaredce joined14:27
gsish distros invented the "alternatives" approach, check that14:28
when it's Ubuntu, then install vim-full or what's the package's name these days, to get something useful14:29
osse I don't know how to see how a debian package is built14:29
gsish some programs reflect compile time choices in --version or similar output14:29
would be surprised when git(1) assumes nano(1), quite the contrary14:30
cousteau osse: oh, was expecting there to be some option to make `git` spit out its default config14:30
osse cousteau: git config --edit I suppose will open it14:31
gsish would expect the distro to have "warped"/twisted the editor setup, because most users find nano more accessible14:31
cousteau gsish: that's what I was starting to suspect. A search for `nano` in the source for Ubuntu/Debian packages turned up nothing of interest. (Well, a search for nano(?!s|time), because damn, that was a lot of false matches...)14:32
← Echoz left14:32
cousteau osse: for example, gcc --version I think prints all the compile time options, showing all the --with and --disable and whatnot14:32
osse cousteau: found it. When Ubuntu packages Git they #define DEFAULT_EDITOR editor14:33
which Git uses14:33
gsish when in vim then :version tells you how unusable the "tiny" package is and why you want vim-full14:33
cousteau gsish: ok, so basically, I would understand that git on ubuntu defaults to EDITOR=the_default_editor, and the_default_editor defaults to nano14:33
osse and editor is that alternatives thing that gsish mentioned (I think)14:33
cousteau osse: cool!14:33
gsish check whether man -k alternative gives a hint14:34
these distros really have become pesky over time14:34
osse https://github.com/git/git/blob/master/editor.c#L17-L4614:34
gsish so either install vim and have the alternative updated in the process14:34
or when it was installed and the default was b0rked then fix it up again14:35
cousteau /debian/rules: DEFAULT_PAGER=pager DEFAULT_EDITOR=editor --> this?14:35
osse yes14:35
gsish update-alternatives14:36
that shall also cover other software that also follows those redirections14:37
cousteau I suppose I might as well just go and do that system-wide14:37
cousteau wondering whether he should do this system-wide or just change the core.editor option for git14:37
gsish and tell your distro how upset you are and don't like that feature which is not git's fault14:37
cousteau considering I'm the one and only user of this machine...14:38
gsish when you administer that machine then you probably use several accounts for that task (a guess)14:38
you may think of yourself as one person, but you will be wearing different hats at different times14:39
cousteau gsish: the reason I've installed git on this machine (well, aside from the fact that it's a new laptop and I would've eventually done that anyway) was that I wanted to tweak a repo with the source of a package that eventually "trickles down" to Ubuntu, because I was annoyed with the behavior14:39
it is QUITE an adventure14:39
so yeah I'm not sure they'll listen14:39
I also recently found a bug that was reported in 201914:39
plus... I'm willing to believe that nano > vi, for low enough values of vi14:40
(this distro comes with a vi, but it's a vi with a very low value)14:41
gsish that's the estimate of those who don't know vi, those who do know that vi > nano14:41
probably the crippled vim that was referred to above14:41
totally not representing what vim is about14:42
can only serve as a bad example how not to do it (the distro packaging)14:42
cousteau gsish: have you ever tried "vim tiny"?14:42
→ crossby1004 joined14:43
gsish unvoluntarily, because you gotta do some editing before you get to install the full version that should have been there from the start14:43
have you tried that :version thing?14:43
← dsimic left14:44
gsish notice that this is not a voice to eliminate nano, I feel that everybody should be free to use nano, assuming they use the nano(1) command to start it14:44
cousteau if you run `vi --version` it'll probably say it's the "Huge version", but the default in Ubuntu is the "Tiny version" (vim.tiny)14:44
the "crippled" one you mentioned, I guess14:44
gsish the trouble comes from presenting something that is not vim to those who do invoke it, and to not ship a full vim because "it's too big" (getting this from those who have the space to ship huuuuge browsers and office suites)14:45
cousteau and... I'm no expert but the little experience I've had with vim.tiny before going "wtf is this crap" and installing the complete one leaves much to be desired14:45
→ dsimic joined14:45
gsish feels like installing "a different VCS" and calling it "a git alike" :) (to get back on topic)14:46
cousteau so even if I have some basic knowledge of vim, I've found THAT version to be less usable (or less easy to use) than nano. In that regard, I'd concede that nano may be > vi, for low enough values of vi (vi = vim.tiny)14:46
← mingdao left14:46
cousteau but, dunno, maybe installing the full vi should've also updated editor to vi, just as it did update vi to vim.basic (the huge one -- it's called "basic" for some reason)14:47
gsish see above, this is so not vim, it's a terrible abomination14:47
→ mingdao joined14:47
cousteau gsish: I'm pretty sure this abomination IS vim14:48
just a very trimmed down version14:48
osse Fitter, happier. More productive.14:48
cousteau with lots of compile time flags disabled14:48
it's just not what one thinks of when they say "vim", but it's still technically vim14:49
It's maybe like expecting sh to be bash and then getting mad because it's dash or some other basic POSIX shell14:50
And not like expecting sh to be bash or dash or another POSIX compilant shell, and encountering tcsh, and damn, that shell IS annoying (not POSIX compliant)14:50
gsish I wish you to spend a day in busybox, and then tell me "it's technically a Linux system" (as you are accustomed to by desktop use with lots of software installed)14:51
cousteau Anyway, you know what, editor is gonna be vim in this machine. I say so.14:51
← mingdao left14:51
gsish there is a difference, and you don't want to accept it when there is no real reason for the crippling14:51
→ Chaser joined14:52
gsish (busybox has its purpose, but everyday desktop use is not it)14:52
→ mingdao joined14:52
cousteau gsish: yeah, that's my point. Busybox may not be representative of what an average Linux is, but it IS still Linux (...it is, right?). As in "an element belonging to the set of Linux, but not a good representative of the average element of the set of Linux"14:53
← limon left14:54
osse update-alternatives --display editor14:54
cousteau so vim.tiny is not "not Vim" in the sense that it's a program that somehow emulates vim and was named "vim" by someone who hates users; it's just a very trimmed down version of the standard Vim, with lots of options disabled.14:54
osse: oh, that is super interesting, once I managed to parse that monstrosity beyond my eyes14:55
← Xlusive left14:56
gsish maybe I was talking nonsense above, the package name appears to be "vim" and not "vim-full", and the installed program is named "vim.basic" :-O14:56
cousteau priority: nano: 40, vim.basic: 30, vim.tiny: 15, ed: -10014:56
gsish and +1 to whoever picked that choice does hate his users14:56
cousteau so yeah... I stand corrected. This seems to be an issue of "someone thinks that nano > full vim", as you suggested14:57
osse pretty sure the alternatives are updated if you explicitly install a package14:57
gsish there may be a non-technical reason, running nano by default results in fewer support requests at the distributor's14:57
osse that's what happened here, anyway14:57
cousteau like, literally >, since 40 > 3014:57
→ Xlusive joined14:58
cousteau osse: or rather, what DIDN'T happen because nano was still the greatest editor14:58
(or DID happen, but had no effect)14:58
bremner wonders what the git connection is here14:58
gsish anybody remembers SuSE's bragging about "the Nuremberg Windows" (inspired by feedback from one customer who liked how migrating didn't change a lot for him)14:58
cousteau bremner: I ran `git commit` and got nano and was like wtf14:59
then I asked here how to know what git thinks is the default editor, to see if I could figure out why it was nano in my distro14:59
bremner you should talk to your distro14:59
or configure core.editor14:59
cousteau obviously, this being a subject on favorite editors, it spiraled into an off-topic discussion on various vim flavors, nano, and poor choices by distros15:00
and that's why we're discussing vim and distros in #git15:00
gsish well alternatives were pointed to in one of the earliest answers15:00
bremner so, asked and answered. and now just venting?15:00
cousteau bremner: I only claim like 40% of responsibility in this discussion15:01
jensen1 → asze15:01
gsish took the OP some time to accept that answer, so ... no, appears to not have been answered by then when the venting started15:02
← berkay left15:02
→ lacrymology joined15:02
→ berkay joined15:05
← AnrDaemon left15:06
← afdee1c left15:06
cousteau gsish: no, you mentioned alternatives, which is kind of incomplete without explaining WHICH alternative I'm looking for (in this case, "editor"), then somehow suggested that installing the full version of vim would've automagically fixed this (which it could have, but it did not), and then quite a while later osse did find the original point where Ubuntu/Debian overwrite this default. Shortly after, you mentioned I should complain to my distro, and that15:07
opened a Pandora box of opinions, me trying to "play devil's advocate" in my distro's name (but also kind of thinking you're right), and lots of speculation on what would be the "right" way to fix this.15:07
it's all your fault, man :)15:07
In any case, since osse eventually found the actual culprit (that someone at Ubuntu/Debian considers that nano is a 33% greater than vim.basic), I'll correct that unforgivable mistake locally and move on15:10
→ skapata joined15:13
← lacrymology left15:18
→ lacrymology joined15:18
← RiFo left15:19
← peirik_ left15:21
→ peirik_ joined15:21
→ afdee1c joined15:26
→ Non-ICE joined15:26
→ AnrDaemon joined15:26
→ wolfdale8335 joined15:29
← Chaser left15:29
← dostoyevsky2 left15:30
← wolfdale833 left15:31
wolfdale8335 → wolfdale83315:31
→ dostoyevsky2 joined15:32
← antranigv left15:34
→ Turn_Left joined15:35
← jmd left15:39
← Left_Turn left15:39
→ thiago joined15:39
→ ChmEarl joined15:48
→ specing joined15:48
← cedd left15:50
→ peirik__ joined16:02
→ pedahzur joined16:04
← peirik_ left16:04
→ Echoz joined16:05
→ antranigv joined16:07
← theyneversleep left16:10
→ epicout joined16:11
→ spare joined16:13
← thiago left16:16
→ thiago joined16:16
← thiago left16:17
← hbautista_ left16:23
← peirik__ left16:24
→ peirik_ joined16:24
→ thiago joined16:26
→ Chaser joined16:32
← blatant left16:33
→ blatant joined16:34
→ cluelessperson joined16:43
← blatant left16:44
→ blatant joined16:45
← cptaffe left16:46
← blatant left16:46
→ cptaffe joined16:46
→ blatant joined16:47
← stats4647 left17:00
← blatant left17:00
← gryffus_ left17:00
→ stats4647 joined17:00
← akibageek left17:01
← jaredce left17:02
← Chaser left17:05
→ ryuu9 joined17:06
← stats4647 left17:07
← militantorc left17:09
→ pikapika_lunar joined17:09
← ryuu left17:09
ryuu9 → ryuu17:09
→ gryffus_ joined17:09
pikapika_lunar → militantorc17:12
← militantorc left17:16
→ pikapika_lunar joined17:17
→ Left_Turn joined17:19
← warmana left17:19
pikapika_lunar → militantorc17:20
→ stats4647 joined17:21
← Turn_Left left17:22
jfsimon1981_b → jfsimon17:23
→ fling joined17:24
← thiago left17:24
→ thiago joined17:25
→ blatant joined17:25
← blatant left17:29
← cluelessperson left17:29
→ Chaser joined17:30
→ blatant joined17:35
← blatant left17:36
→ blatant joined17:39
← Chaser left17:43
← rsx left17:43
→ akibageek joined17:47
→ ___nick___ joined17:52
← blatant left17:56
→ blatant joined18:02
← thiago left18:02
← ___nick___ left18:02
→ thiago joined18:02
← blatant left18:04
→ ___nick___ joined18:04
← MyNetAz left18:06
→ MyNetAz joined18:07
← ___nick___ left18:08
→ ___nick___ joined18:10
→ afdee1c_ joined18:12
← monkfish left18:15
← afdee1c_ left18:23
← SuperLag left18:23
→ SuperLag joined18:24
← th3voic3 left18:24
← techtruth left18:25
→ lucasta joined18:25
→ th3voic3 joined18:25
← lacrymology left18:26
→ wootehfoot joined18:28
→ afdee1c_ joined18:30
← thiago left18:31
→ thiago joined18:31
→ rainn92675560720 joined18:35
← afdee1c_ left18:39
← zloi_user left18:39
→ JordiGHost joined18:45
JordiGHost Can git annotate (blame) each line with the commit message?18:45
ChatGPT made up a `--show-summary` option when I asked it.18:47
I love it when the machines give me false hope.18:47
Hrm, when I told it that it made it up, it gave me a short but convoluted shell script.18:48
Ah, c'est la guitte.18:48
← dermoth left18:48
← rainn92675560720 left18:53
→ dermoth joined18:54
→ Narrat joined18:55
specing JordiGHost: install tig, it does this18:57
(shows commit when you press enter on the line)18:57
JordiGHost I wanted to give someone a quick command to do it, but git sucks too much for what I consider basic functionality.18:58
→ cedd joined18:58
→ jaredce joined19:00
← cedd left19:03
← cbreak left19:04
← jaredce left19:04
→ cedd joined19:05
→ Turn_Left joined19:05
→ Sampersand joined19:06
← JordiGHost left19:08
← Left_Turn left19:09
osse lol19:09
→ cbreak joined19:10
→ techtruth joined19:10
→ iomari891 joined19:11
Sampersand So, i'm trying to write a `git-branch-refs` command which prints out what `@{-1}` , `@{-2}` etc would refer to if i wrote `git switch @{-###}` . Currently, I'm getting my output like `git rev-parse --symbolic-full-name @{-{1..$max}}` , however, I want to stop early if fewer than `$max` entries are available. Is it possible to get the possible19:16
number for which `@{-###}` is valid?19:16
← cousteau left19:16
← stats4647 left19:16
Sampersand I tried checking if the output equaled `@{-###}` (ie the input equals the output) but for some reason my history doesn't have `@{-71}` or `@{-72}`, but it does have 73 onwards19:17
→ cousteau joined19:17
Sampersand ideally i'd do `local gitmax=$(git somehow-find-max-value) ; [[ $max -ge $gitmax ]] && max=$gitmax` and then the rest of my logic is hunky dory19:19
← lucasta left19:20
← Jck_true left19:22
→ Jck_true joined19:22
→ Everything joined19:24
→ stats4647 joined19:28
→ theoceaniscool joined19:29
→ ferdna joined19:33
← Everything left19:34
← Sampersand left19:34
→ Sampersand joined19:34
→ Everything joined19:35
→ Sampersand26 joined19:35
← Sampersand left19:35
← stats4647 left19:40
← theoceaniscool left19:43
→ rolo joined19:44
← Sampersand26 left19:51
→ Sampersand joined19:52
→ theoceaniscool joined19:55
→ JordiGH joined19:56
→ limon joined19:57
← BigShip left19:59
→ Sampersand48 joined20:01
Sampersand Hi all, is there a way to get the amount of refs in git history? IE what's the maximum number I can do `git switch @{-NNN}`20:02
→ stats4647 joined20:02
← Sampersand48 left20:03
→ m0viefreak joined20:04
JordiGH Those aren't refs.20:05
Branches are refs, but commits are commits.20:05
Sampersand oh what are they20:05
JordiGH Oh, wait, you want the depth of the reflog. I needed to check gitrevisions(7) again.20:06
I guess that's `git reflog | wc -l`?20:07
Sampersand ah wow20:07
JordiGH I'm not sure that's accurate, does `git reflog` show everything? Mine seems to stop at 300.20:08
ikke git log --oneline -g --grep-reflog="checkout:"20:08
JordiGH Why omit resets, pulls, merges?20:09
Oh, I see, because that duplicates some hashes.20:10
→ BigShip joined20:10
JordiGH jordi@eris:~/vcs/grist/saas$ git reflog | cut -f 1 -d' ' | sort -u | wc -l20:11
14420:11
jordi@eris:~/vcs/grist/saas$ git log --oneline -g --grep-reflog="checkout:" | wc -l20:11
9720:11
Shouldn't those match?20:11
ikke Sampersand was refering to @{-NNN}20:12
with -NNN (negative), it refers only to checkouts20:12
Sampersand I see! So my idea is to have an equiv of `dirs` for git, so i can see all my recent checkouts. the problem is that `@{-N}` doesn't remove duplicates, so i was making my own function to remove dups. However, it's extremely slow for some because it needs to check for uniqueness, and i was hoping there was something that was already extant20:12
here's what i have now: https://pastebin.com/5QRpr6E220:14
(in zsh)20:14
sco1_ → sco120:17
Sampersand is there something more efficient or already builtin?20:17
JordiGH I mean... the reflog?20:18
Sampersand what about it?20:19
JordiGH That shows you what you want.20:19
You could --grep-reflog="checkout:" to limit it like ikke suggested.20:20
→ john_johnk joined20:21
Sampersand ah, so `@{-n}` is not some "stack" that's maintained somewhere in `.git`, but it actually dynamically determines it upon each invocation?20:21
← rolo left20:23
JordiGH Not really a stack, just a list of where you were. You don't really pop it.20:23
→ ferdna_ joined20:23
JordiGH It's in .git/logs/HEAD if you just care about HEAD (other refs are under .git/logs/refs)20:23
dirs is a poor comparison because that is a stack.20:24
← _val__ left20:25
Sampersand i see20:25
← ferdna left20:25
Sampersand 'cause i frequently bounce between like three or four branches with long names, and `@{-N}` is a godsend. But sometimes i forget which one is which20:26
but yeah, ok so getting a "unique" version of it is not really integral20:27
← memset left20:31
→ memset joined20:32
← afdee1c left20:35
→ zloi_user joined20:36
→ rolo joined20:39
→ afdee1c joined20:40
→ Juesto joined20:46
← rolo left20:46
← Juest left20:47
Juesto → Juest20:47
← spare left20:52
→ Juesto joined20:53
← Juest left20:56
Juesto → Juest20:56
← afdee1c left20:59
→ afdee1c joined21:01
← afdee1c left21:03
← ___nick___ left21:04
→ afdee1c joined21:08
← Narrat left21:11
→ afdee1c_ joined21:14
← thiago left21:24
← theoceaniscool left21:24
osse dare i suggest tsb-completing the one youcwa21:30
← Traneptora left21:30
osse God dammit21:31
Tab-completing the one you want21:31
← hwpplayer1 left21:35
← adamtajti left21:36
→ Traneptora joined21:38
← john_johnk left21:38
→ hwpplayer1 joined21:38
b0nn lol, tab complete in IRC to stop typos :P21:40
→ kale joined21:40
← mika left21:42
→ mika joined21:42
← peirik_ left21:43
→ peirik_ joined21:43
→ marc2377 joined21:43
→ john_johnk joined21:48
← Everything left21:49
← marc2377 left21:51
← iomari891 left21:55
Lore osse, I have connected ASPell to WeeChat IRC client here, and configured it in a way it highlights typos.21:55
→ peirik__ joined21:55
Lore i.e. https://linux.die.net/man/1/aspell21:56
osse, for example, a typoo and it there are now 5 suggestions at the top right corner configured to be shown: https://i.imgur.com/jV1mNyI.png21:59
← peirik_ left21:59
Lore s/it //21:59
It doesn't help with logical mistakes though ^^"22:00
← zloi_user left22:01
← Sampersand left22:03
→ lucasta joined22:03
→ forgotmynick joined22:05
→ thiago joined22:06
b0nn Lore: I have a thing in irssi where I can <Ctrl-z> to review what I am about to send and check for typos, - I still typo the hang out everything I send22:07
← tolnanwo left22:13
← limon left22:17
→ limon joined22:19
← khrbtxyz left22:26
→ tolnanwo joined22:26
← kale left22:26
→ Left_Turn joined22:28
← hwpplayer1 left22:28
← Turn_Left left22:31
← memset left22:31
→ memset joined22:32
← hexnewbie left22:34
← mrpops2ko left22:35
← stats4647 left22:36
← benaldo left22:45
→ benaldo joined22:46
→ stats4647 joined22:55
→ mrpops2ko joined22:56
← user_oreloznog left22:59
← stats4647 left22:59
→ Gary_Sinise joined23:00
← warthog9 left23:02
→ warthog9 joined23:05
← mrpops2ko left23:05
→ gryffus joined23:08
← gryffus_ left23:08
← xx left23:12
→ zloi_user joined23:12
← sentriz left23:13
→ Juesto joined23:17
← Juest left23:19
Juesto → Juest23:19
→ stats4647 joined23:19
→ mrpops2ko joined23:22
← peirik__ left23:22
→ peirik_ joined23:22
← stats4647 left23:24
← stenno left23:34
← bn_work left23:34
→ hbautista_ joined23:35
← hbautista_ left23:36
← afdee1c_ left23:36
← th3voic3 left23:36
→ th3voic3 joined23:38
← mrpops2ko left23:49
→ stats4647 joined23:51
← stats4647 left23:55
→ rosco joined23:56
← FreeBDSM left23:58
← seanh left23:58
→ cdown joined23:58
→ seanh1 joined23:59

Logs Search ←Prev date Next date→ Channels Documentation