IRCloggy #git 2019-04-06

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.

2019-04-06

osse It was just that it seemed you were looking for a very technical answer and were very confused.00:00
I was thinking "... but how it different from any other fetch?"00:00
nightshift That's what I thought, i've been sitting through this for the last several days trying to figure out why soni just doesn't use tags. But, of course, if it's a repo you don't control00:00
rafasc with default settings, anything else is fetched by walking the DAG using those references as starting points.00:00
→ learningc joined00:00
← de-facto left00:01
← bs left00:01
→ de-facto joined00:01
← moo left00:02
rafasc for --deepen, it basically sends a depth-request in the negotiation phase.00:03
Soni so, does deepen fetch everything you already have?00:03
rafasc What you already have is not transferred.00:03
if you --deepen=10, you only transfer 10 commits.00:04
Soni but how, if you're not allowed to fetch from a commit?00:04
→ altgray joined00:04
→ bs joined00:06
← bs left00:07
rafasc Because you are always fetching from a reference. With deepen=10, you say something like to the server: I have HEAD~123..HEAD, please give me more 10 commits.00:08
→ jelmer joined00:09
rafasc then the server goes ahead and sends you HEAD~112..HEAD~12200:09
Soni okay00:09
can I lie to the server00:10
rafasc No00:10
Soni why not00:10
← jelmer left00:10
← cbreak left00:10
→ jelmer joined00:10
→ cbreak joined00:11
rafasc because https://github.com/git/git/blob/master/Documentation/technical/pack-protocol.txt00:12
← xmate left00:13
Soni ah, because deltas00:13
fucking deltas00:14
any workarounds?00:14
→ bs joined00:14
rafasc workaround for what?00:14
Soni for needing to apply deltas onto what I have00:15
nightshift more disk space/memory, so you can do a standard fetch?00:15
Soni nightshift: can't do00:15
nightshift Yeah, I know, but, also kind of the point00:15
rafasc he it goes. Can't do, can't say...00:15
Soni can I just run the algorithms on the incoming network data and skip the disk entirely?00:16
I really don't care about bandwidth at this point00:16
I'm not paying for it anyway00:16
rafasc What's the question in the first place?00:16
!xy00:16
gitinfo You are asking about a solution, but we can help you a lot better if you could describe the actual problem you are trying to solve.00:16
Soni need to find a commit from a resource-constrained system00:17
that's literally my problem00:17
nightshift Soni, I think that was mentioned when you first started asking use. git is a VERY disk intensive processes, you either need enough space to write, or enough RAM to keep it all in a ram disk00:17
Soni nightshift: what if I don't have either?00:17
are there any workarounds?00:17
can I just parse the data as it comes in and skip what I don't care about?00:17
can I just keep enough RAM for the deflate or w/e?00:17
nightshift Then you need to find a completely different tool to solve your problem, especially since you refuse to tell us your ACTUAL use case00:18
← iam730 left00:18
→ orbyt_ joined00:18
nightshift No, you need to keep everything you fetch, until you get what you want. Once you've got what you want, you can toss everything newer, and older, but not until you have the commit you want00:18
← bs left00:19
Soni I don't need the files00:19
at all00:19
nightshift then, what DO you need?00:19
Soni just to check if the commit itself exists00:19
I don't care about the files in the repo00:20
nightshift why?00:20
nightshift headdesk00:20
Soni so I know it's a fork?00:20
← Sasazuka left00:20
nedbat Soni: you have the attention of experts willing to help you. Explain the entire problem. What are you building?00:20
→ iam730 joined00:21
Soni a thing that looks at forks00:21
nightshift why?00:21
Soni when they stop being forks, it stops looking at the ex-forks00:21
nedbat Soni: great. say more. What is it going to do?00:21
Soni that's it00:21
nedbat Soni: how would something stop being a fork?00:21
Soni with force-push00:21
→ bs joined00:21
Soni something that removes the specific commit I'm looking for00:21
or just make a new branch and kill the fork branch00:22
← mixaz left00:22
nedbat Soni: maybe i don't understand what you mean by fork? What commit are you looking for? Where did the commit come from?00:22
nightshift Are you ultimately checking to see if someone stole your work?00:22
nedbat Soni: and btw, why do you want to monitor repos for forkness?00:22
rafasc there is --negotiation-tip=, but I don't even know if this would help.00:22
Soni a fork is any repo that contains the commit but isn't run by me00:22
nightshift FINALLY00:23
nedbat Soni: can you say more about "the commit"? Is it really a single commit you are looking for?00:23
Soni nightshift: no, these are mail-order repos00:23
nedbat: it's an arbitrary commit specified by the person who runs my tool00:23
nedbat Soni: what is this person trying to achieve? Tell us the whole thing.00:23
Soni nightshift: i.e. person sends me an email saying they want me to look at their fork and I add it to the thing00:23
osse Git fetch runs another program under the hood to do the fetching (e.g. git receive-pack). I suppose it would be possible to invoke that directly, but you customize the arguments or data to pretend you have that commit and you want to deepen with 100:23
Soni and it autodetects if it stops being a fork00:23
nedbat Soni: why would this person do that?00:23
osse something like that00:23
Soni nedbat: because I'm anti-centralization00:24
nedbat Soni: "stops being a fork" still doesn't sound right to me. A repo is always a fork of the thing it started from, it just may have diverged far from it.00:24
← bs left00:24
Soni nedbat: in which case it should still have the commit that identifies it as a fork00:24
rafasc osse: as someone who tried messing with upload-pack and receive-pack, it's not that simple.00:24
nedbat Soni: how frequently do you think a repo will be force-pushed to no longer have anything in common with it's original repo? That sounds like a really unusual thing to happen.00:25
Soni nedbat: some caching will be done with the latest heads or so00:25
nedbat Soni: i don't know what that is a reply to.00:25
Soni i.e. on first run it'll scan the whole tree00:25
nedbat Soni: let's talk about the "unforking" idea.00:26
Soni: how frequently do you think a repo will be force-pushed to no longer have anything in common with it's original repo? That sounds like a really unusual thing to happen.00:26
Soni and then, if it finds the "fork" commit, it'll store a handful of the commit IDs for the newest commits00:26
if those disappear then it'll do a full rescan00:26
full rescans are quite rare tbh but should still fit within the resource constraints00:27
nedbat Soni: i don't think you're listening to my question00:27
Soni nedbat: you're asking me if I should even bother with this and allow anyone to hijack anything00:27
← gloomy_ left00:27
→ bs joined00:27
nedbat Soni: do you have an example of someone hijacking a repo like that?00:27
Soni nedbat: yeah, it happened a while back with what was it...00:28
it's been a while, sorry00:28
I think I still used windows back then so it's on a dead HDD :D00:28
nedbat Soni: you're building a tool to detect a thing that you know has happened once, in the distant past, it sounds like.00:29
Soni but yeah the project simply vanished, killed all commits and just pushed an empty repo00:29
nedbat: yeah, I am00:29
nedbat Soni: btw, why does it have to run on a resource-constrained machine?00:29
Soni not the main goal this is just a safety thing00:29
nedbat: because this is aimed at rpis and shitbox VPSes00:29
nedbat Soni: why?00:30
← bs left00:30
Soni nedbat: because that's my uh...00:30
nedbat Soni: it seems to be the main stumbling block.00:30
Soni: larger machines are not uncommon00:30
Soni because it's aimed at those ppl00:30
it's my what do you call it00:30
← cliluw left00:30
nightshift what people?00:30
Soni target demographic?00:30
nedbat Soni: to be blunt, it sounds like you are designing for poverty-stricken paranoids.00:30
Soni idk00:31
nedbat: what if I am?00:31
nedbat Soni: then you are making your life difficult, and we don't know how to help you build that.00:31
nightshift Then, as we've already told you, you've picked the wrong tool00:31
Soni what if I want ppl to self-host this on their phones or on a cheap pi?00:31
or on a shitbox VPS?00:32
nedbat Soni: it might not be possible.00:32
→ wootehfoot joined00:32
Soni nedbat: I do not care about bandwidth00:32
nightshift it's not a bandwidth issue, Soni, it's a diskspace issue00:32
nedbat Soni: great. but you've given yourself a difficult contraint00:32
→ cliluw joined00:32
Soni in fact for all I care I could just seek the git thingies all the time00:32
nedbat Soni: btw, why run it on one machine? These sound like the kind of people that would welcome a federated effort00:32
Soni and by "seek" I mean request everything all over until I get to the point I care about00:33
over. and over. and over.00:33
→ nurdie joined00:33
Soni no need to worry about bandwidth.00:33
nightshift btw, someone that is so poor the best they can manage is a pi with an 8gb card, isn't likely to have unlimited bandwidth00:33
nedbat Soni: until you are banned.00:33
Soni nightshift: shitbox VPS00:33
→ bs joined00:34
Soni nedbat: so what?00:34
nedbat Soni: what are you responding to?00:34
Soni "until you are banned"00:34
if there's no other way to do it then there's no other way to do it00:34
so you just do it the only way you can do it00:34
nedbat Soni: i just assumed you wanted this thing to keep monitoring.00:34
nightshift yeah, I can't even afford a shitbox VPS, technically I'm in your demographic, but, you don't what I was able to do? Get my pi a 500gb hdd.00:34
Soni nedbat: besides, this is just the initial scan00:35
nedbat Soni: we don't know how to build the thing you want.00:35
Soni the excessive bandwidth usage is very rare in normal operation00:35
nightshift but high disk/memory usage is to be expected, aka, what you want to do isn't possible00:36
Soni you don't need a lot of disk or RAM when you can just use a remote server as disk/RAM00:36
and all I'll be doing is an hourly shallow (1-commit) head clone00:37
← bs left00:37
nedbat Soni: if you have a remote server for disk, then why are you talking about limited disk?00:37
Soni nedbat: "disk" in this case means repeating the request as many times as needed00:37
← nurdie left00:38
← tm007 left00:38
Soni it's not that I have disk it's that I don't need it00:38
if I need a specific delta I just fetch the thing again and repeat until I have everything I need, without keeping any of it in RAM *or* disk00:39
nedbat Soni: you are talking as if you have a thing that works.00:40
Soni nedbat: I'm talking as if I wanna do an awful thing because nobody can come up with an alternative00:40
nedbat Soni: ok, you have a plan then.00:40
Soni nedbat: do you have an alternative00:40
nedbat Soni: no, i told you we don't know how to build the thing you want00:40
← jelmer left00:42
→ bs joined00:43
rafasc nedbat: can I contact you via DM?00:43
nedbat any time00:44
← wootehfoot left00:44
← bs left00:45
← fassl left00:46
→ bs joined00:50
← learningc left00:52
← bs left00:53
← omry left00:53
→ omry joined00:54
→ nurdie joined00:54
→ bs joined00:58
← nurdie left00:58
← bs left00:59
← moldybits left01:00
← stux|RC-only left01:01
→ gloomy joined01:02
Soni can I size-limit git?01:02
→ bs joined01:02
Soni also remember back when you could run git on 256MB of RAM and 20GB of disk01:03
or w/e it was back on the i38601:03
might've been less01:04
rafasc checkout git from that time?01:04
← lucasb left01:04
Soni can I use that old a git with something like github?01:04
rafasc yes, the protocol is really stable. But I doubt you'll gain from it.01:06
something that might interest you: https://www.eclipse.org/lists/jgit-dev/msg03073.html01:06
not directly related to your problem, but related with the motivation you seem to have.01:07
Soni I want ppl to selfhost and fork widely01:07
← bs left01:08
Soni I don't want consensus, I want competition01:08
rafasc But people can already do that?01:08
git is distributed. I can fetch from you without any 3rd parties involved01:08
Soni you clearly don't understand the motivation01:08
rafasc I do not. That's true.01:08
Soni I want a project without a BDFL. without an upstream.01:09
a project as diverse as the ppl who use it.01:09
I want a project driven by the community and its needs.01:10
no PRs, just community.01:11
nedbat Soni: that sounds great, but why do you need a hijack-detector to do that? The people will see it happen.01:11
rafasc like an anarchy repository?01:11
Soni nedbat: so you can run it on an unmonitored server01:11
← theoceaniscool left01:11
rafasc anyone can push anything they want?01:11
Soni rafasc: no01:11
these are forks01:11
each fork is on its own server01:12
← zonixo left01:12
Soni it just lists all of them in activity order01:12
→ finalbeta joined01:12
← ggSo left01:12
Soni gives all of them the same space01:12
it's definitely inspired by anarchy, but not the modern notion of anarchy (a chaotic mess) but the roots of anarchy, which is about trust and stuff (don't feel like explaining this right now)01:14
← dviola left01:14
Soni anarchy roots: someone makes something, they ask you to put an ad up for them, you do it or you don't based on whether you trust them or vice-versa01:15
→ nurdie joined01:15
Soni modern notion of anarchy: someone makes something, they torture you into putting an ad up for them and then you slice their throat later or something01:15
very different01:16
← Essadon left01:17
rafasc and you want that hijack-detector to make sure that 'ad' still points to what you accepted, and not to someone that might've be force pushed?01:17
→ misuto joined01:18
← nurdie left01:19
← yyy left01:20
→ sitaram joined01:21
← gloomy left01:21
Soni not really01:26
I just want the projects to still refer to the thing that they're supposed to refer to01:26
even if they diverge a lot01:26
all I care about is the commit01:26
→ bs joined01:27
Soni even if they push a commit on top of everything that clears everything and starts fresh01:27
nedbat Soni: but the projects could still have that commit, and yet have nothing to do with the original code.01:27
Soni nedbat: exactly01:27
and that's okay01:27
nedbat Soni: why is it ok? I could hijack your project and you wouldn't know it.01:27
Soni nedbat: that's fine01:28
I don't expect you to understand it01:28
I don't expect most ppl to understand it until I have it working01:28
nedbat Soni: if you can't explain it so we can understand it, why would code explain it?01:29
Soni nedbat: who said anything about code?01:29
← tang^ left01:29
nedbat Soni: "until I have it working".... that's code.01:30
Soni (altho code will also be available and it'll eat its own dog food)01:30
nedbat: that's a live demo01:30
← bs left01:30
nedbat Soni: ok, why will people understand it once you have it working?01:30
Soni nedbat: same way ppl understand git and github by using those things01:31
if you ask me if git explains itself... no, it definitely doesn't. unless you already know some things.01:31
nedbat Soni: the hijack detector is designed to detect something that you say is very very rare. how will i understand it?01:31
Soni you learn those things by using it.01:31
by playing with it you'll learn what a branch is, what a repo is, and so on01:31
nedbat: oh. well, it's rare. so you'll likely never see it happen.01:32
nedbat Soni: ok, we're very far from the topic of #git. Good luck with your efforts. I think figuring out how to explain it would be a useful exercise for you.01:32
Soni and that's okay because it's only supposed to happen rarely.01:32
nedbat: talking about git is off topic?01:33
rafasc if you want to refer to a commit, just tag the commit and gpg sign it.01:33
nedbat Soni: we haven't talked about git in a long time :)01:33
Soni <Soni> if you ask me if git explains itself... no, it definitely doesn't. unless you already know some things. <Soni> you learn those things by using it. <Soni> by playing with it you'll learn what a branch is, what a repo is, and so on01:33
anyway I'm gonna sleep, good night o/01:34
rafasc good night.01:34
← F0rTex left01:35
→ nurdie joined01:36
→ F0rTex joined01:36
← cfoch left01:40
← nurdie left01:41
nightshift Alright, now that the knives have been removed from my back, I'm having to dive into git-prompt.sh, cause, well, I kinda want to make the prompt just a bit more complicated when I'm in a repo01:42
rafasc nightshift: was that literal? o.O01:42
nightshift rafasc, no, there weren't literal knives in my back, just a feeling that there were, and that was what was keeping me from writing the function earlier01:43
rafasc GIT_PS1_SHOWUPSTREAM="verbose name git"01:43
GIT_PS1_SHOWDIRTYSTATE=101:43
GIT_PS1_SHOWSTASHSTATE=101:43
GIT_PS1_SHOWUNTRACKEDFILES=101:43
GIT_PS1_SHOWCOLORHINTS=101:44
GIT_PS1_DESCRIBE_STYLE=branch01:44
GIT_PS1_STATESEPARATOR='|'01:44
complicated enough?01:44
nightshift rafasc, not quite. I want to switch from \w to \W in a repo01:46
and then add showupstream=verbose01:46
Not sure that's all that easy to do, but I'm looking at it.01:49
rafasc checking if git rev-parse --git-dir; then echo in repo; else echo not in repo; fi?01:53
nightshift rafasc, pretty much (just got there in the script),01:55
obviously, I'd change out the echo statements for the appropriate arg to __git_ps101:57
→ nurdie joined01:57
nightshift but, that still brings me back to using a variable for that argument. Should be ok as long as I use something other than PS1 though, I think01:58
rafasc yeah.01:59
→ hailochild joined01:59
→ dre_ joined02:00
nightshift This should be interesting02:00
← dre_ left02:01
← nurdie left02:01
→ dre_ joined02:02
→ boombatower joined02:05
← hailochild left02:06
← boombatower left02:06
→ supernov1h joined02:07
← kent\n left02:07
nightshift osse, when I write this function, do I need to set all the GIT_PS1_ variables out of the function, or can I set them in it?02:07
→ hailochild joined02:08
rafasc any will do I think.02:10
← n3wborn left02:12
← hailochild left02:12
nightshift thanks. Worst case, I just switch it. Final question, I'm exporting those in my bashrc, but, shouldn't have to, right?02:12
rafasc No need to export them.02:16
nightshift ok.02:17
→ nurdie joined02:18
← nurdie left02:23
→ pigsalsa joined02:23
← Lucas_Gray left02:26
← lagothrix left02:30
→ lagothrix joined02:30
← orbyt_ left02:32
→ moldybits joined02:33
→ stitched_unicorn joined02:36
→ bashfulshell joined02:39
→ nurdie joined02:39
→ zonixo joined02:40
→ ggSo joined02:40
← nurdie left02:44
← Primer left02:44
← dre_ left02:49
→ dre_ joined02:49
→ pd09041999 joined02:52
← dpyro_ left02:52
← thiago left02:52
← rafasc left02:52
→ dpyro joined02:52
→ stux|RC-only joined02:53
nightshift Alright, function finished (took WAY longer than it should have, I had a brain fart when it came to getting the variables set). Time to test02:56
dre_ → dre02:57
amosbird_ → amosbird02:58
nightshift Almost there, two problems I need to fix, the first one is nightshift@nightshift:gitviewer (master u=)$ what is the u=?03:00
→ nurdie joined03:00
nightshift oooooo, that's the upstream, which, for some reason is showing null03:02
← zonixo left03:03
← ggSo left03:04
→ Beam_Inn joined03:04
← nurdie left03:06
nightshift ok, that's actually not an issue, I don't think, it's doing what it's supposed to. rev_parse was probably the wrong choice for checking if I'm in a repo, or not. What I really want is is to take advantage of repo_info03:08
← Beam_Inn left03:09
nightshift but, that doesn't seem to be an option03:10
Oh well, I'll figure that out another day, maybe. I could just copy it into bashrc03:11
← mikecmpbll left03:16
→ nydel09 joined03:17
→ lalilulelo joined03:20
→ nurdie joined03:22
lalilulelo hello, I have a folder structure that includes "Assets/Editor" (i.e. it's Unity Engine). Editor was a git repository. Later on, I decided to make Assets a repository. I created a .gitignore that doesn't ignore the Editor directory, and then tried staging everything with git add ./*, commiting immediately afterwards. However, the Editor folder was ignored. Now I want to include the Editor folder again, and even after deleting its .git/ folder, no files03:25
within Editor are showing up in git status. What am I doing wrong?03:25
→ mven joined03:26
← nurdie left03:26
← kjartan left03:27
→ Goplat joined03:27
← kerframil left03:28
← Cabanossi left03:30
nightshift laliulelo, first problem is you had repos within repos, there's additional setup that covers that. Unfortunately, I can't help with that part. There is a way to force Editor to be tracked now, let me see if I can find that part for you03:32
→ kjartan joined03:32
lalilulelo nightshift, thanks for responding, I'll wait03:33
I might ask on StackOverflow as well because I can't for the life of me find much of anything about that on Google, at least in a way that I can phrase the problem in my query03:38
→ rsrx joined03:38
nightshift lalilulelo, try https://git-scm.com/docs/git-check-ignore first03:40
→ Cabanossi joined03:40
← justanotheruser left03:40
nightshift also https://stackoverflow.com/a/2329640403:41
← dre left03:41
→ dre_ joined03:42
→ nurdie joined03:43
← dre_ left03:44
lalilulelo thanks for those03:45
nightshift of course03:45
← pd09041999 left03:45
lalilulelo hmm, git check-ignore returns nothing03:47
nightshift then try the other one03:47
← nurdie left03:47
lalilulelo wooow lol03:48
so I tried it03:49
and the Editor directory still doesn't show up as staged in git status03:49
→ pd09041999 joined03:51
lalilulelo I appreciate your help, nightshift, I guess I'll ask in StackOverflow so the question can stay there and help others eventually and whatnot03:52
nightshift no idea, is it showing as untracked, or just not showing up at all?03:54
→ justanotheruser joined03:55
← mven left03:57
lalilulelo not showing up at all03:57
nightshift that's really weird03:58
lalilulelo I'm guessing it's somehow treating that folder as part of a submodule or something03:59
nightshift Sorry I'm not a little more experienced03:59
most likely03:59
I don't deal with those at all, so I don't have the slightest clue how to fix it03:59
lalilulelo it's alright, it beats me too hahah03:59
me neither, guess I'll have to try and read their docs04:00
I was hopeful I would be able to fix this without dealing with that, lol04:00
nightshift hahahaha, git's docs are better than some, though.04:01
← Copenhagen_Bram left04:01
lalilulelo ah, I hope they'll be helpful enough04:02
nightshift me too04:02
It's time for me to shut down for the day, good luck04:04
→ nurdie joined04:04
← nightshift left04:04
lalilulelo aaaa04:07
this actually did it: https://stackoverflow.com/a/1789374/357934604:07
and thanks :304:08
→ Copenhagen_Bram joined04:08
← fatalhalt left04:09
← nurdie left04:09
← stux|RC-only left04:15
→ shanks8080 joined04:24
→ nurdie joined04:25
→ stux|RC-only joined04:29
← nurdie left04:29
← freeman42x left04:32
← stux|RC-only left04:35
→ assemblyman joined04:46
→ nurdie joined04:46
← shanks8080 left04:46
← veegee left04:49
← nurdie left04:50
← inkbottle left04:52
← bashfulshell left04:54
→ pks_ joined05:02
← rsrx left05:02
→ kapil____ joined05:02
← pks left05:04
pks_ → pks05:04
← zalt left05:06
→ nurdie joined05:07
← nurdie left05:12
← m0viefreak left05:12
→ Envil joined05:13
← KnoP left05:24
← jnewt_ left05:26
→ sauvin joined05:27
→ KnoP joined05:27
→ nurdie joined05:28
→ rsrx joined05:31
← rsrx left05:32
← justanotheruser left05:32
→ rsrx joined05:32
← nurdie left05:32
← rsrx left05:34
→ rsrx joined05:34
← rsrx left05:36
← KnoP left05:37
→ shanks8080 joined05:45
→ justanotheruser joined05:46
→ sternmull joined05:48
→ nurdie joined05:49
→ SkarmoutsosV joined05:50
→ dre_ joined05:53
← nurdie left05:54
← assemblyman left06:06
→ clime joined06:10
→ nurdie joined06:10
← pd09041999 left06:13
→ pd09041999 joined06:14
← nurdie left06:15
← niax left06:17
← GreenJello left06:19
← irqq left06:22
← pd09041999 left06:23
→ pd09041999 joined06:23
→ xmate joined06:23
→ nurdie joined06:31
← nurdie left06:36
← pR0Ps left06:37
→ niax joined06:39
→ mixaz joined06:40
→ nurdie joined06:52
→ zalt joined06:53
→ made2591 joined06:55
→ pR0Ps joined06:55
← nurdie left06:58
→ Raging_Hog joined06:58
→ RoriconKnight joined07:00
→ zonixo joined07:04
→ ggSo joined07:04
← dpyro left07:04
→ ricekrispie2 joined07:05
← ricekrispie left07:07
← made2591 left07:09
→ nurdie joined07:13
← SkarmoutsosV left07:18
← nurdie left07:18
← RoriconKnight left07:26
← wildlander left07:30
← sphalerite left07:33
→ nurdie joined07:34
← Goplat left07:36
→ sphalerite joined07:36
→ opalepatrick joined07:38
← opalepatrick left07:38
← zulutango left07:39
← nurdie left07:40
← kjartan left07:40
→ hussam joined07:42
→ kjartan joined07:45
→ czart joined07:50
→ in1t3r joined07:51
← altgray left07:52
→ spacesuitdiver joined07:55
→ nurdie joined07:56
← R2robot left07:57
→ keepLearning512 joined07:59
← nurdie left08:00
→ YuGiOhJCJ joined08:01
← keepLearning512 left08:03
→ nowhereman joined08:10
← dre_ left08:13
Marmotte_ → Marmotte08:15
← zonixo left08:16
← ggSo left08:16
→ nurdie joined08:17
→ Lucas_Gray joined08:19
← spacesuitdiver left08:21
← Lucas_Gray left08:21
← nurdie left08:21
→ Lucas_Gray joined08:22
← Lucas_Gray left08:24
→ irqq joined08:27
← al-damiri left08:27
→ Lucas_Gray joined08:28
→ mikecmpbll joined08:35
← Lucas_Gray left08:38
→ nurdie joined08:38
← nurdie left08:42
→ crumbking joined08:44
→ learningc joined08:45
← crumbking left08:45
→ Lucas_Gray joined08:46
→ crumbking joined08:47
→ zulutango joined08:51
← osse left08:55
→ osse joined08:56
← yonder left08:56
→ altgray joined08:57
← zulutango left08:58
← xmate left08:59
→ nurdie joined08:59
→ dhollin3 joined09:01
← pd09041999 left09:03
← dhollinger left09:03
← nurdie left09:04
→ zulutango joined09:04
← nowhereman left09:05
→ martin-_-_ joined09:11
← Novae left09:11
→ Novae joined09:11
martin-_-_ hi I wanted to go back to a specific commit and remove all commits which happend before09:11
I did git reset --hard f234f283d4c41017e086e0e4708c8955cc18074009:11
git status tells me now HEAD detached at f234f2809:12
how can I attach it again?09:12
→ nurdie joined09:12
→ keepLearning512 joined09:14
→ pd09041999 joined09:15
→ libertyprime joined09:15
→ n3wborn joined09:16
→ nowhereman joined09:17
← nurdie left09:17
← keepLearning512 left09:19
← catsup left09:20
→ catsup joined09:20
← catsup left09:21
→ catsup joined09:21
← cd left09:22
← stitched_unicorn left09:25
→ powerbit joined09:31
ndimitrij → ndim09:32
→ leeN joined09:33
→ cd joined09:37
← Novae left09:38
→ Novae joined09:41
← sunri5e_ left09:42
← zalt left09:44
← ricorambo left09:45
→ jottr joined09:45
→ made2591 joined09:46
← made2591 left09:47
→ made2591 joined09:47
→ sunri5e joined09:49
← xcm left09:51
→ xcm joined09:51
← Novae left09:55
→ Novae joined09:55
← czart left09:56
← leeN left09:58
← made2591 left09:59
← mikecmpbll left09:59
← jottr left09:59
← lankanmon left10:00
j416 check your branch out before you reset.10:01
also, you cannot remove commits before another commit by resetting; you need some more for that. Interactive rebase might be of use.10:02
not quite sure what you mean.10:03
martin-_-_: ^10:03
← nowhereman left10:05
→ R2robot joined10:08
→ plexigras joined10:09
→ zonixo joined10:09
→ ggSo joined10:10
← sunri5e left10:14
→ sunri5e joined10:14
← Muzer left10:15
→ SkarmoutsosV joined10:20
→ m0viefreak joined10:22
→ GreenJello joined10:22
→ Muzer joined10:27
← hofmann3900 left10:28
← igemnace left10:28
→ theoceaniscool joined10:33
← GreenJello left10:34
→ keepLearning512 joined10:35
← Novae left10:37
→ Novae joined10:38
→ Regon joined10:41
← ggSo left10:42
← zonixo left10:43
→ zonixo joined10:43
→ ggSo joined10:43
→ dviola joined10:50
→ lankanmon joined10:50
← n3wborn left10:50
← nixdork left10:51
→ nixdork joined10:51
← martin-_-_ left10:51
→ made2591 joined10:54
→ bs joined11:00
← bs left11:00
← libertyprime left11:01
→ yyy joined11:01
→ freeman42x joined11:03
Soni how do I check if a repo can fetch from arbitrary commits?11:10
→ mikecmpbll joined11:20
← SkarmoutsosV left11:21
← mixaz left11:24
j416 you cannot fetch from a commit, you fetch from a remote11:25
→ zalt joined11:27
→ basiclaser joined11:28
→ ferdna joined11:28
← made2591 left11:28
→ made2591 joined11:29
← made2591 left11:30
→ made2591 joined11:30
→ czart joined11:32
← yyy left11:39
→ czart_ joined11:42
→ dudek joined11:43
← czart left11:43
→ Makaveli7 joined11:45
osse you just have to try11:48
→ Achylles joined11:48
← irqq left11:49
→ Fernando-Basso joined11:55
→ dre_ joined11:59
→ opalepatrick joined12:06
→ bs joined12:08
← Shentino left12:09
→ Shentino joined12:10
← bs left12:12
← Fernando-Basso left12:12
→ bs joined12:14
→ z|bandito joined12:15
← opalepatrick left12:15
← bs left12:17
zampano_ → zampano12:17
→ opalepatrick joined12:19
→ nowhereman joined12:19
← jstimm left12:24
→ bs joined12:24
← thejunkjon left12:27
← keepLearning512 left12:27
← opalepatrick left12:28
← ferdna left12:28
→ GreenJello joined12:31
← zalt left12:32
← bs left12:33
→ bs joined12:34
← Shentino left12:34
→ Shentino joined12:35
← GreenJello left12:36
→ greatgatsby joined12:37
← dre_ left12:38
← VLetrmx left12:40
→ frederfred joined12:41
Soni how do I check if a remote allows me to fetch from arbitrary commits?12:43
→ _kk22kk_ joined12:44
→ Essadon joined12:50
← Essadon left12:50
→ Essadon joined12:50
← JanC left12:54
→ JanC joined12:54
nedbat Soni: osse said, "you just have to try"12:56
→ gloomy joined12:56
gloomy → crestfallen12:58
crestfallen → Crestfallen12:58
← Crestfallen left12:59
← altgray left13:01
→ gloomy joined13:03
← gloomy left13:04
← greatgatsby left13:12
j416 and you can still not fetch from commits13:13
← kjartan left13:18
→ Fernando-Basso joined13:21
→ kjartan joined13:23
← _kk22kk_ left13:24
← Muzer left13:25
← dudek left13:30
→ Muzer joined13:32
← dviola left13:34
← in1t3r left13:35
→ inkbottle joined13:39
→ lucasb joined13:43
← Anthaas left13:51
→ mixaz joined13:51
→ Anthaas joined13:52
osse j416: you can if the server allows it13:54
see the uploadpack.allow* config options13:54
unless I misunderstand those, of course13:55
← random_yanek left13:59
← theoceaniscool left14:02
→ igemnace joined14:04
← Makaveli7 left14:04
→ random_yanek joined14:06
→ bashfulshell joined14:10
j416 osse: yes, but you still fetch from a remote.14:18
→ fatalhalt joined14:18
j416 just trying to get the actual question clear14:18
← impermanence left14:19
j416 in any case; seems it's been answered a couple of times now. :)14:19
→ zalt joined14:23
← Lucas_Gray left14:24
→ Xeago_ joined14:30
→ jelmer joined14:31
← jelmer left14:31
← Xeago left14:31
Xeago_ → Xeago14:31
→ GreenJello joined14:33
→ irqq joined14:34
→ spacesuitdiver joined14:37
← made2591 left14:38
← spacesuitdiver left14:38
→ spacesuitdiver joined14:38
→ made2591 joined14:39
→ Lucas_Gray joined14:39
← made2591 left14:42
dx_ob → Neel14:44
← GreenJello left14:46
← zalt left14:46
← donofrio left14:47
→ donofrio joined14:49
← justanotheruser left14:50
→ made2591 joined14:53
→ Xeago_ joined14:54
← Xeago left14:55
Xeago_ → Xeago14:55
→ qu4ku joined14:56
→ boZipper joined14:57
← supernov1h left14:58
← boZipper left14:58
→ chiasmata joined14:58
← inkbottle left15:02
← chiasmata left15:04
← nowhereman left15:04
← ayjay_t left15:06
→ ayjay_t joined15:06
→ dpyro joined15:07
← hussam left15:12
→ jamiejackson joined15:13
→ leeN joined15:17
← YuGiOhJCJ left15:17
dhollin3 → dhollinger15:19
← tralfaz left15:20
→ Wryhder joined15:25
← Lucas_Gray left15:25
→ mowcat joined15:26
Wryhder → Lucas_Gray15:26
← mowcat left15:26
→ theoceaniscool joined15:26
→ yonder joined15:27
→ mowcat joined15:27
→ justanotheruser joined15:30
→ wildlander joined15:40
→ wreed joined15:44
→ wootehfoot joined15:46
← jamiejackson left15:46
→ opalepatrick joined15:48
← opalepatrick left15:49
← learningc left15:49
→ learningc joined15:50
→ jamiejackson joined15:51
← wreed left16:01
→ Phylock joined16:02
← SleePy left16:03
→ opalepatrick joined16:06
← leeN left16:06
→ SleePy joined16:07
← opalepatrick left16:08
→ opalepatrick joined16:10
→ jelmer joined16:11
← theoceaniscool left16:12
→ thiago joined16:14
→ GreenJello joined16:18
→ stux|RC-only joined16:24
← jamiejackson left16:27
← Lucas_Gray left16:28
← spacesuitdiver left16:29
→ jamiejackson joined16:32
← qu4ku left16:34
← Fernando-Basso left16:34
→ dviola joined16:36
← jamiejackson left16:38
→ theoceaniscool joined16:38
→ zalt joined16:46
→ jamiejackson joined16:47
→ chiasmata joined16:50
← mikecmpbll left16:51
← R2robot left17:01
→ altgray joined17:01
→ R2robot joined17:03
→ pd_09041999 joined17:08
→ hussam joined17:08
← pd09041999 left17:11
→ syn0 joined17:12
← mixaz left17:15
← opalepatrick left17:22
clime does anyone know in which version of git, -C switch was introduced?17:22
mmattice_ → mmattic17:23
mmattic → mmattice17:23
_ikke_ 1.8.517:24
← dviola left17:24
clime _ikke_: thanks is there a good way to find that out?17:25
i guess Documentation and RelNotes17:26
_ikke_ Indeed17:26
clime was trying to find in git log before, thanks17:26
_ikke_ I happen to have a list (outdated) which lists what features were introdoced when17:26
clime cool17:27
_ikke_: is it public somewhere?17:27
_ikke_ http://tinyurl.com/git-changes17:27
← pd_09041999 left17:28
clime _ikke_: thanks!17:28
← kjartan left17:31
← jcbitter left17:35
→ kjartan joined17:36
→ tarkus joined17:37
→ jcbitter joined17:38
← learningc left17:38
→ learningc joined17:38
← bket left17:43
← learningc left17:45
← mowcat left17:46
→ learningc joined17:46
← swift1 left17:48
← theoceaniscool left17:52
← learningc left17:52
→ learningc joined18:10
→ Industrial joined18:11
Industrial Hi. Say I have a repo with MANY different little projects inside (https://github.com/helm/charts.git). How do I search through the git history (file changes, commit messages, etc) for a subtree of the repo?18:12
I want to check matches for "password" or "Password" inside the PostgreSQL Helm Chart, because there is currently a bug with it.18:13
So I'd like to see what happened specificly concerning passwords :-)18:13
specifically*18:13
→ mikecmpbll joined18:13
→ orbyt_ joined18:16
← made2591 left18:16
→ blackandblue joined18:19
→ dege joined18:19
clime Industrial: git log -S "(Password|password)" --pickaxe-regex .18:19
Industrial: cd to the subtree you want to search in first18:20
← kapil____ left18:22
Industrial clime: Thanks ! :-)18:25
→ nurdie joined18:26
← syn0 left18:27
→ nurdie_ joined18:29
clime Industrial: you can also do git log --grep="password" --grep="Password" .18:29
that actually searches in the commit messages. Idk if there is a command that combines both18:29
_ikke_ or git log --grep=[pP]assword18:30
"18:30
← nurdie left18:30
→ gaia joined18:31
→ nic-hartley joined18:32
→ bergercookie joined18:33
clime Industrial: also -G will probably give you more results then -S, -S only displays changes that affected number of occurrences - i needed to browse man pages18:33
gaia hi. i need to undo a few changes from a patch that i created with git add --patch. is there something like"git remove --patch" or so ?18:35
← learningc left18:36
_ikke_ gaia: git reset --patch18:36
→ learningc joined18:36
gaia ah... testing :)18:37
← ggSo left18:40
← zonixo left18:40
gaia _ikke_, thanks for this tip :) That saved me a lot of headache18:44
→ pd_09041999 joined18:44
→ TheZ joined18:45
_ikke_ :)18:54
→ Goplat joined18:54
← nurdie_ left18:56
← wyre left18:59
→ mroutis joined18:59
→ APic joined19:03
→ TJ- joined19:04
mroutis I often use `git log --since=... --no-merges --patch` to get an overview of what happened on a project, however, the project maintainer doesn't "squash & merge" when integrating branches to the main one, I often see commits that don't make sense because they depend on another commit, is there a way to `squash` them when reading them with git log?19:05
osse mroutis: maybe you want --first-parent, but I'm not sure how that affects the diff output19:06
i don't think there's an option to do exactly what you describe19:06
← Achylles left19:07
→ hseg joined19:07
osse mroutis: ah, you can try --first-parent -m *without* --no-merges19:08
i think you will see the merge commits with a patch that is the entirety of the branch merged in19:08
mroutis osse: thanks for the quick response, I'll take a look19:08
← igemnace left19:11
mroutis osse: awesome, dude, the `-m` flag did the work!19:11
→ ferdna joined19:11
← ayjay_t left19:12
osse \o/19:12
→ ayjay_t joined19:12
→ anddam joined19:12
→ wyre joined19:14
anddam howdy, I need a bit of git-wisdom. I forked https://github.com/micropython/micropython on my github account to have an repo of mine to push to. I then would like to fork https://github.com/dmazzella/circuitpython/tree/dmazzella-patch-119:15
mroutis osse: so, what the command is doing is removing commits that are part of a merge by using `--first-parent` and in those cases, by using `-m` it kind of squash the whole thing into the merge commit, right?19:15
anddam this is a branch of a repo that has a pending fork on circuitpython, in turn a fork of micropython19:15
← bergercookie left19:17
osse mroutis: the -m flags makes git log output one diff per parent for merge commits19:17
mroutis: the --first-parent "censors" the second on19:18
one19:18
anddam if I try to fork https://github.com/dmazzella on github it warns me that I have already forked micropython/micropython . Now I'm not sure if I should keep the one repo micropython/micropython and add to it the dmazzella/circuitpython branch or I should have two repos, micropython/micropython AND circuitpython/circuitpyhon (or dmazzella/circuitpython)19:18
wow I confused even myself while writing that…19:18
mroutis osse: that's genius, thanks again :)19:18
→ AtumT joined19:20
← eraserpencil left19:20
← hseg left19:20
← snsapiens left19:21
→ eraserpencil joined19:23
anddam mmm, I did 'pull origin patch-1' while I was on master that is tracking upstream/master, this resulted in an automated merge commit19:23
how do I delete this last commit?19:23
← chiasmata left19:25
← swalladge left19:25
anddam no, actually I can find that on the web. What I want to understand is why that commit was auto-generated, and why issuing :q! at the commit message didn't stop the commit19:25
→ swalladge joined19:26
anddam I see the merge is the default pull action, pull meaning 'fetch+merge' but I still expected to manually perform the commit19:26
ah --no-commit, this documentation is actually useful. Now I only have to learn to read it *beforehand*19:27
mm a nice 'git reset --hard HEAD' seems to have done the trick, can anyone confirm that is appropriate ?19:30
← d^sh left19:37
→ sns joined19:38
← meLon left19:38
→ d^sh joined19:39
clime i basically have this command to get remote branch fetch url: git config remote."$(git config branch."$(git symbolic-ref --short HEAD)".remote)".url19:41
seems a little bit complex...is there anything shorter?19:41
it also does not account for origin fallback that git fetch does19:42
← lucasb left19:42
→ nurdie joined19:43
→ meLon joined19:51
osse anddam: :q! doesn't vim exit with a non-zero exit code19:59
you can use :cq for that20:00
← tarkus left20:02
anddam osse: or I should have deleted the auto-populated first line and let git raise an error for that20:03
osse: I'm not using vim (anymore) btw20:03
← nurdie left20:04
→ nurdie joined20:04
anddam I was using a minimum set of features, and I did not like its default command shortcuts, so I went for something smaller20:04
osse: and one bonus is there is no vimL20:04
← learningc left20:05
← basiclaser left20:06
osse hey I was just answering why :q! didn't work20:07
← mroutis left20:07
← nurdie left20:09
anddam osse: other than exiting with 0 I think the issue is that merge creates a valid commit message20:09
osse: sorry, I misread your last line20:09
← Neel left20:10
← justanotheruser left20:20
→ hofmann3900 joined20:25
→ ezio joined20:25
← tmz left20:26
→ lurkless joined20:30
← sternmull left20:34
← orbyt_ left20:35
← czart_ left20:35
← blackandblue left20:38
osse anddam: Afaik any non-empty message is valid20:39
← sweatsuit left20:41
→ orbyt_ joined20:42
→ sweatsuit joined20:43
← wootehfoot left20:43
← Envil left20:43
→ fstd_ joined20:43
→ sweatsuit_ joined20:44
→ nurdie joined20:45
← sweatsuit left20:47
← fstd left20:47
fstd_ → fstd20:47
← sweatsuit_ left20:49
→ sweatsuit joined20:49
→ Olipro joined20:50
← nurdie left20:50
← eraserpencil left20:53
→ eraserpencil joined20:53
← orbyt_ left20:54
→ tmz joined20:59
← ayjay_t left21:03
→ ayjay_t joined21:04
→ theoceaniscool joined21:05
nezZario I have no idea what one of my devs did here. They did a merge commit, so it appears in the log, ... but then a whole bunch of work from another dev -- in another commit -- is shown as being deleted/removed in their commit... What the heck are they lying to me about?21:05
they said they were just merging their feature branch in21:05
← clime left21:06
← cbreak left21:08
→ cbreak joined21:09
→ justanotheruser joined21:09
→ rafasc joined21:17
→ wootehfoot joined21:19
← ayjay_t left21:21
→ made2591 joined21:21
→ ayjay_t joined21:22
→ bket joined21:24
→ nurdie joined21:27
→ Puffball joined21:27
→ goodafternoon joined21:30
← thiago left21:30
← nurdie left21:31
← Industrial left21:33
→ inkbottle joined21:36
← justanotheruser left21:38
→ kent\n joined21:41
→ perrier-jouet joined21:43
← kjartan left21:44
← ixian left21:45
→ ixian joined21:46
→ orbyt_ joined21:47
→ nurdie joined21:48
→ kjartan joined21:49
← staafl_ left21:50
← nurdie left21:52
→ gaia_ joined21:54
← gaia left21:58
→ moei joined21:59
← skered left22:00
j416 nezZario: sounds like you have a bigger problem than your git repo22:03
← orbyt_ left22:07
→ nurdie joined22:08
→ veegee joined22:09
→ [Brain] joined22:11
nezZario Ha. Yes I'm pretty sure22:11
← nurdie left22:12
→ orbyt_ joined22:14
rafasc nezZario: could be accidental. Like mistakes during conflict resolution, etc.22:16
but if you have the history, you can check who did what.22:17
unless one of them is forging commits.. But if that's the case, you have way bigger problems.22:17
← lalilulelo left22:19
nezZario No the guy just doesn't know how to use git. Is there a way to know if there was a conflict?... I bet he just clicked 'mine'. He is using some kind of UI. I mean he brought it to my attention that he was worried he did something wrong. I'm trying to figure out what he could've done.22:21
← orbyt_ left22:21
nezZario A way to find out - assuming nothing malicious or any intentional tampering22:21
I don't see it in git log22:22
j416 nezZario: get someone on your team who does know to help out?22:24
← zulutango left22:25
j416 if you have no one on your team who knows git well, give them a couple of days to learn22:25
!book22:25
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 !parable22:25
→ justanotheruser joined22:25
→ zulutango joined22:27
rafasc nezZario: git show <merge> --cc; can show some information on how the merge was resolved, or you can just checkout both parents and merge it yourself again.22:30
git checkout --detach <merge>^1; git merge <merge>^2; you can use the detached head to avoid creating unnecessary branches.22:31
← justanotheruser left22:32
→ gxt joined22:36
→ dre_ joined22:36
→ ggSo joined22:36
→ zonixo joined22:36
← gaia_ left22:41
← made2591 left22:42
← theoceaniscool left22:43
→ nowhereman joined22:44
← mikecmpbll left22:44
← Phylock left22:45
→ nurdie joined22:49
← dege left22:53
← nurdie left22:54
energizer I have two stashed items which are incompatible but I want to merge them and apply them as a single commit on top of the current master. What can I do?22:57
rafasc stash apply one, stash apply two. ?22:58
← freeman42x left22:58
← zonixo left23:00
energizer error: Your local changes to the following files would be overwritten by merge23:00
← ggSo left23:00
← finalbeta left23:04
energizer uh oh i lost them23:06
rafasc stash apply doesn't remove the stash.23:09
← plexigras left23:15
→ orbyt_ joined23:18
← dre_ left23:19
→ igemnace joined23:19
← TJ- left23:19
→ gambeta joined23:21
energizer got it now23:23
thanks23:23
→ nurdie joined23:26
→ mroutis joined23:26
→ victorqueiroz joined23:26
← mroutis left23:30
← nurdie left23:30
← silverballz left23:32
← ferdna left23:36
→ jbowen joined23:46
→ nurdie joined23:47
← wootehfoot left23:47
← cd left23:49
← causasui left23:49
→ cd joined23:50
→ silverballz joined23:50
→ SuperL4g joined23:50
← nurdie left23:51
d1b_ → d1b23:52
← jbowen left23:52
→ jbowen joined23:52
← SuperLag left23:53
→ SuperLag joined23:55
← SuperL4g left23:56
→ cruzo66342 joined23:59
← gambeta left23:59

Logs Search ←Prev date Next date→ Channels Documentation