Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-03-13

| Channels | #git index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 alip_ left #git
00:01 dreiss joined #git
00:02 boboc joined #git
00:03 cdown joined #git
00:04 boboc joined #git
00:09 matoro joined #git
00:13 pfrench joined #git
00:14 chachasmooth joined #git
00:17 ocelotsloth joined #git
00:19 pfrench joined #git
00:19 chuchunaku joined #git
00:19 raijin joined #git
00:20 glowdemon1 joined #git
00:20 MrcRjs joined #git
00:22 cqi joined #git
00:24 mnaser left #git
00:24 cqi joined #git
00:27 venmx joined #git
00:28 pfrench joined #git
00:33 dsdeiz joined #git
00:36 fahadash joined #git
00:42 tanuki joined #git
00:43 livingst_ joined #git
00:50 peterbecich joined #git
00:51 chipotle joined #git
00:52 peterbecich joined #git
00:53 hasc joined #git
00:54 pfrench joined #git
00:56 hasc joined #git
00:56 peterbecich joined #git
00:57 DolpheenDream joined #git
00:57 mehola joined #git
00:59 pfrench joined #git
00:59 mehola joined #git
01:03 pfrench joined #git
01:06 bocaneri joined #git
01:07 tvw joined #git
01:07 pfrench joined #git
01:07 askb joined #git
01:07 peterbecich joined #git
01:08 jstimm joined #git
01:14 pfrench joined #git
01:15 chipotle joined #git
01:15 bocaneri joined #git
01:17 rkazak_ joined #git
01:18 pfrench joined #git
01:21 Spherical joined #git
01:21 A124 joined #git
01:22 bufferoverflow joined #git
01:22 Gsham joined #git
01:23 f3r70rr35f joined #git
01:28 pfrench joined #git
01:29 rchavik joined #git
01:32 acetakwas joined #git
01:32 wgrant joined #git
01:38 cdg joined #git
01:48 a_thakur joined #git
01:48 f3r70rr35f joined #git
01:51 pfrench joined #git
01:52 chitopunk joined #git
01:57 f3r70rr35f joined #git
01:59 pfrench joined #git
02:08 Rodya_ joined #git
02:08 cdg joined #git
02:08 pfrench joined #git
02:11 d0nn1e joined #git
02:14 pfrench joined #git
02:16 Gsham joined #git
02:17 finalbeta joined #git
02:18 Nilesh_ joined #git
02:21 xissburg joined #git
02:22 pfrench joined #git
02:22 pks joined #git
02:22 netj joined #git
02:23 dsdeiz joined #git
02:23 dsdeiz joined #git
02:23 hahuang65 joined #git
02:27 d^sh joined #git
02:30 souther joined #git
02:31 altendky joined #git
02:35 SantaClauze joined #git
02:35 xissburg joined #git
02:36 pfrench joined #git
02:38 chitopunk joined #git
02:38 ResidentBiscuit joined #git
02:39 svm_invictvs joined #git
02:39 ResidentBiscuit joined #git
02:39 marianina8 joined #git
02:41 pfrench joined #git
02:46 Goplat joined #git
02:46 Gsham joined #git
02:46 pfrench joined #git
02:48 hahuang65 joined #git
02:50 pfrench joined #git
02:51 BackEndCoder joined #git
02:52 hexagoxel joined #git
02:53 Gsham joined #git
02:55 sunri5e joined #git
02:55 mizu_no_oto joined #git
02:55 Cabanossi joined #git
03:04 BackEndCoder joined #git
03:05 Emperor_Earth joined #git
03:05 ojdo joined #git
03:06 Rotaerk is there a way to do the equivalent of "git branch -r --contains <commit ID>" *without* having a local clone
03:06 Rotaerk i.e. kind of like how ls-remote can list remote branches without having a local clone
03:06 preaction not to my knowledge
03:07 Rotaerk k, thanks
03:07 gfixler joined #git
03:09 gtristan joined #git
03:13 pfrench joined #git
03:14 frobnic joined #git
03:15 hexagoxel joined #git
03:19 elsevero joined #git
03:21 d^sh joined #git
03:22 Vortex35 joined #git
03:22 pfrench joined #git
03:23 kadoban joined #git
03:23 lagothrix joined #git
03:27 pfrench joined #git
03:28 xall joined #git
03:29 fstd joined #git
03:32 peterbecich joined #git
03:32 pfrench joined #git
03:35 kasual joined #git
03:37 durham joined #git
03:39 pfrench joined #git
03:39 peterbecich joined #git
03:42 andrew710 joined #git
03:48 pfrench joined #git
03:49 a_thakur joined #git
03:54 al-damiri joined #git
03:56 hexagoxel joined #git
03:56 Andrew_K joined #git
03:58 UniFreak joined #git
03:58 UniFreak should c-tag's .tags file be version by git?
04:01 chachasmooth joined #git
04:02 pabs3 joined #git
04:04 pabs3 is there a way to disable all fetches for a particular git remote? I want these to not fetch the foo remote: `git fetch foo`, `git fetch --all`, `git fetch`
04:08 Cabanossi joined #git
04:12 shgysk8zer0_ joined #git
04:15 pfrench joined #git
04:23 pfrench joined #git
04:23 A124 Is there a way to cherry-pick merge then commit without going to editor to edit commit message? in other words keep the commit message.
04:24 ShekharReddy joined #git
04:24 chachasmooth joined #git
04:25 sarri joined #git
04:25 sarri joined #git
04:29 robotroll joined #git
04:31 pfrench joined #git
04:31 dsdeiz joined #git
04:33 BackEndCoder joined #git
04:37 eb0t joined #git
04:38 Cabanossi joined #git
04:41 MrcRjs joined #git
04:41 cqi joined #git
04:42 pks joined #git
04:43 pfrench joined #git
04:44 hitman1_ joined #git
04:44 hitman1_ Hi I want to make a website and want to upload it on github. Can I do so ?
04:44 hitman1_ I will make it using django. Will it run if I host it on github?
04:44 hitman1_ and yes I want my own domain address - hitman1.com not hitman1.github.io
04:44 cdg_ joined #git
04:44 mrslaughter1775 joined #git
04:44 preaction hitman1_: for github questions, ask #github
04:45 MarioBranco joined #git
04:45 robattila256 joined #git
04:46 hitman1_ ok
04:48 preaction so... are you going to do that?
04:50 ayogi joined #git
04:50 BackEndCoder joined #git
04:52 memfrob joined #git
04:54 memfrob Hi all! Hey I'm using git am to apply a patch but the first part of the patch fails, so I fix the conflict in a text editor (modify the conflicting lines and apply the merge manually) but it doesn't apply the remaining portions of the patch if there's a conflict.
04:54 memfrob git am --continue && git add -A && git diff HEAD~ shows only the lines I did mysef.
04:54 memfrob *myself
04:55 hitman1_ preaction, yes
04:55 memfrob error: patch failed: drivers/misc/lkdtm.c:107 error: drivers/misc/lkdtm.c: patch does not apply ... so I fix the conflict at line 107 and that's the only change there.
04:56 memfrob Like wuuuuuhhhhhhhhhhhhh
04:58 hitman1_ btw, what is difference between git and github ?
04:59 memfrob github is a website, git is a client. Github is a git server.
04:59 preaction git is a version control system. github is a website that hosts git repositories
04:59 memfrob Your work goes into git, and git push sends the information to github which displays the changes.
05:00 memfrob You can use git without github but you cannot use github without git. browse sure, but not push changes.
05:02 memfrob Is this a git am bug, or does git am --continue not continue applying the patch once there's a conflict and gives up completely and lets you do all the work?
05:02 hitman1_ ok
05:02 memfrob not even putting all the rejecting lines in the .rej file ...
05:02 memfrob There's 15 other lines here that git completely ignored and says it's complete.
05:03 acetakwas joined #git
05:04 hhee joined #git
05:04 memfrob hitman1_, is there anything else you need help with? I'm glad to help.
05:05 preaction i'm waiting for them to join #github so i can answer their question about github pages
05:05 memfrob While you're waiting you got any insight on my issue? :P
05:05 preaction (and that it doesn't host django sites, it hosts static websites)
05:05 preaction nothing besides that you need to "git add" and then "git am --continue". not the other way.
05:06 preaction whenever git says "resolve this", it means "resolve this in the index by adding the right changes using 'git add'"
05:06 duderonomy joined #git
05:07 memfrob I do that but git am --continue after I run git add -A doesn't finish the rest of it, it tells me the patch is fully applied when it isn't.
05:08 duderono_ joined #git
05:08 pfrench joined #git
05:09 memfrob http://dpaste.com/1MTDWZ7
05:09 memfrob It says "Applying: lkdtm: add test for atomic_t underflow/overflow" but it isn't actually modifying any files after the first conflict is fixed.
05:11 memfrob http://dpaste.com/2743TPE The conflicting lines are #41 and #49 I fix those conflicts but lines 57 to 67 are completely forgotten by git.
05:13 memfrob git am --continue doesn't modify any files.
05:13 memfrob git diff HEAD~ only returns the lines I added in by hand.
05:14 pabs3 left #git
05:16 memfrob Now if I edit that file again by hand, finish applying it myself, run git add -A again, then git commit --amend, then git diff HEAD~ shows the output I was expecting. But git am is useless if it gives up after the first conflict.
05:17 preaction once it got messed up the first time, you had to reset before trying again
05:17 preaction but otherwise, i don't use am, so i don't know how to resolve conflicts in it other than how git makes you resolve conflicts in everything else
05:18 pfrench joined #git
05:21 stduolc joined #git
05:22 memfrob what do you use? patch?
05:22 preaction to do what?
05:23 pfrench joined #git
05:24 memfrob to patch stuff
05:24 memfrob I was told I'm stupid if I don't use git am here xD
05:25 memfrob The whole channel was like "don't use patch, git am is much more preferred"
05:25 preaction to patch stuff from what?
05:25 preaction why are you using files at all?
05:25 preaction what workflow is requiring that you use git am?
05:26 memfrob Merge a change from one tree to another and they're not closely related
05:26 preaction if they're related enough to use "git am", they're related enough to merge or cherry-pick
05:27 memfrob What's a good way to merge changes then from a branch that's ahead of your branch?
05:27 preaction "git merge"
05:27 memfrob That never works for me or does anything, I always have to use git pull.
05:27 preaction "pull" is "fetch" plus "merge"
05:28 preaction if you didn't fetch, then there's nothing to merge
05:28 memfrob Oh! That's why it never works!
05:28 preaction back up, what are you trying to accomplish?
05:28 preaction if you're using "pull", how are they "not closely related"?
05:28 preaction start at the beginning please, as i'm entirely confused
05:28 Taikibansei joined #git
05:29 memfrob I'm backporting security features from kernel 4.9 to 4.4 but the code changes DRASTICALLY between minor kernel releases. the subversion stuff isn't bad at all.
05:30 memfrob Yeah see git merge totally screwed everything up.
05:31 memfrob I want to merge this commit here: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/drivers/misc/lkdtm.c?h=linux-4.9.y&amp;id=50fbd977733d1cbafe75baf4b50424434dcfea8c so I run git merge 50fbd977733d1cbafe75baf4b50424434dcfea8c and it tries to apply all of 4.9's history into mine. If I use git am, I get 3 lines of changes, not a whole tree with a gazillion conflicts.
05:31 pfrench joined #git
05:31 memfrob I freakin HATE git merge.
05:32 memfrob Hasn't worked once for me, ever.
05:34 preaction so what you want is cherry-pick. you don't merge commits, you merge branches. you cherry-pick commits
05:34 preaction and git merge works completely fine on projects that aren't insane (and i don't think anyone would argue that the kernel is a sane project)
05:35 memfrob HAHAHHA!!!!!!!!!!!!
05:35 pfrench joined #git
05:35 Derperperd joined #git
05:36 memfrob Wait, your answer to my question though was git merge, now you're saying it's cherry-pick (which actually works btw)
05:36 memfrob thanks man! saves me from having to save patches to my drive all the time!
05:36 preaction because you said you wanted to merge. so i said to use merge
05:37 preaction i also exactly said: [00:26:42]  <preaction>if they're related enough to use "git am", they're related enough to merge or cherry-pick
05:37 memfrob oh there!
05:37 memfrob my bad
05:38 TbobbyZ joined #git
05:39 memfrob preaction, thank you
05:39 preaction np. good luck.
05:39 TbobbyZ joined #git
05:39 memfrob thanks it's going well! a lot easier now!
05:39 memfrob you're funny btw.
05:40 preaction i'm busy making cookies is why i suspect
05:41 memfrob now i'm craving sweets. i better have some ice cream left.
05:43 pfrench joined #git
05:43 memfrob yes i do!!! for the record i'm not a stereotypical fat computer guy or anything
05:44 memfrob and regardless of what richard hammond thinks, not just gay men eat ice cream.
05:44 bruce_lee joined #git
05:44 bruce_lee joined #git
05:44 korovkin25 joined #git
05:44 memfrob bruce_lee, welcome!
05:44 ashnur_ joined #git
05:45 bruce_lee \(^_^)/
05:47 ashnur_ left #git
05:49 pfrench joined #git
05:51 totemizer joined #git
05:51 totemizer hi
05:51 gitinfo totemizer: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.
05:51 memfrob gitinfo, fastest welcome I've ever seen.
05:51 memfrob totemizer, hi
05:52 totemizer any way to track people who clone a repo? it's a challenge and it would be helpful to see who clones the repository so they don't give up too early
05:53 totemizer like email or username from git config would be perfectly fine
05:53 memfrob it'd have to be a server that logs the IP of everyone who access the git server I believe, which would mean hosting it behind a firewall that you have access to and logs packets.
05:53 pfrench joined #git
05:53 memfrob I thought git was anonymous and used SSH public keys as security...
05:54 totemizer not sure i can touch the server that way :(
05:54 memfrob I highly doubt that's possible but I'd wait for an expert to comment on that before giving up.
05:55 marianina8 joined #git
05:55 marianina8 joined #git
05:56 pwndave joined #git
05:56 memfrob Very smart people hang out in #git
05:56 memfrob We even have philosophers and martial artists like bruce_lee
05:58 preaction totemizer: unless they use their ssh key to do it, no
05:58 memfrob heh that's what i thought!
05:58 preaction if they use their ssh key to do it, then yes if you have access to the server
05:58 preaction i remember "anonymous" ftp which would use email addresses for passwords
05:58 preaction memfrob: gitinfo is a !bot, is why
05:58 gitinfo memfrob: [!gitinfo] I am an IRC bot which responds to certain keywords to provide helpful(?) information to humans.  Please see http://jk.gs/git/bot for more information about how to (ab)use me.
05:58 totemizer i have some access, just not something like setting up something that reads the logs
05:58 totemizer or maps IP address
05:59 totemizer and everyone who can access this repo has to be already authenticated
05:59 preaction totemizer: unless you're making them authenticate, then no
05:59 memfrob totemizer's last comment clinched it.
05:59 memfrob such a clincher.
05:59 preaction do you control the account that hosts the git repos? (like, do you use gitolite?)
06:00 memfrob oohh he's onto something!!
06:00 hexagoxel joined #git
06:00 totemizer preaction: not personally, but i can maybe arrange simple stuff to be configured
06:01 totemizer with the it people, if i speak nicely :(
06:01 memfrob speak nicely, asking them to have very particular access so you can spy on people..
06:01 preaction gitolite's authentication works because it's a set of perl scripts in front of git commands. you could do something with (or similar to) that. but otherwise, frankly, none of this is worth the effort
06:01 totemizer it's not spying
06:02 dreiss joined #git
06:02 totemizer preaction: maybe you have a suggestion for solving the problem easier?
06:02 pfrench joined #git
06:02 memfrob I'm giving you a hard time xD
06:02 preaction my suggestion was: it's not worth the effort
06:03 preaction if you're using ssh, auth.log is a thing
06:03 totemizer :) i was hoping that maybe the same thing can be achieved a different way, with less effort, so it's worth it
06:03 totemizer hmm
06:03 preaction it's _worthless_ information, so the effort you've taken so far has already surpassed its worth
06:04 memfrob preaction, thank you very much for the help. totemizer is NSA so I'm leaving. (jk) but no seriously I've been messing with kernel code too long, I need a break. take care all!
06:04 totemizer btw, our interface is gitlab
06:04 preaction hosted gitlab? or local instal/l
06:05 totemizer well, hosted by us
06:05 totemizer :) not sure about the terminology you use, sorry
06:05 goTAN joined #git
06:06 preaction so local. then look at gitlab's docs and/or api to see if they make this information available
06:06 totemizer preaction: what you mean it's worthless? i definitely do not want to waste hours working on this, but last time we did something similar, we had no idea who tried to solve it and because of this those who tried and gave up fast .. we never heard of them
06:06 totemizer ok, thanks for the help
06:07 preaction why do you want the data again? what will this data provide that other data (pull requests, patch submissions) will not?
06:07 preaction you think people are downloading your git repo but then... what?
06:07 jstimm joined #git
06:07 dec0n joined #git
06:08 bwn joined #git
06:08 MineCoins joined #git
06:09 pfrench joined #git
06:10 totemizer preaction: this is a coding challenge inside our company for employees, the upper management wants me to push everyone to finish it. without knowing who tries, i can't do this. the user accounts i would be tracking are employee accounts not personal, and completely internal
06:10 preaction do they get to do this on work time?
06:10 preaction or is this personal time?
06:10 totemizer i think that's up to them and their managers :P
06:11 preaction so, personal time. tell upper management to fuck right off
06:11 totemizer :))
06:11 totemizer sure, i will just do that
06:11 preaction make them download a zip file of the git repo from a website after logging in
06:11 totemizer on an unrelated note, maybe someone is hiring, i might be looking for a job soon?
06:13 TbobbyZ joined #git
06:13 totemizer personally, i wanted to make a damned download button that asks for an email and gives a download link, but i was voted down because apparently that would stop people to participate
06:13 preaction what benefit do they get from doing this challenge?
06:13 totemizer who?
06:13 preaction the employee
06:13 rscata joined #git
06:13 preaction this reeks of fake team-building, which makes doing it on personal time even worse
06:14 totemizer oh, there is a small award, recognition, but ultimately the idea is that they can join the department and work on this stuff instead of what they are doing now
06:14 totemizer if they are interested
06:14 totemizer yeah, the personal time thing bothers me too
06:14 preaction okay, so they're applying for a job. then it doesn't matter if they're tracked
06:14 totemizer but frankly, i haven't had a job in the past 14 years where i have not had to make this compromise
06:15 Renter so everybody needs to finish it, or just the people applying ?
06:15 totemizer no one has to do anything
06:15 totemizer (apart from myself who should track those who try)
06:15 Renter ah
06:16 preaction doing work for your employer on personal time? because that's not personal time, that's work time. and i do 40 hours of work time per week (unless extenuating circumstances)
06:16 TbobbyZ_ joined #git
06:16 preaction my hobby time sometimes benefits work, but my boss doesn't get to tell me when, what, or how much to do
06:16 preaction so, my suggestion remains: don't bother. if this is a job application, treat it like any other job application
06:17 Renter ^
06:17 totemizer preaction, i think a smart employee can find the time in work hours to complete this challenge. most people i know at work are not pushed to do any kind of overtime or anything
06:17 Renter internal job applications are still just job applicatoins
06:17 jozwior joined #git
06:17 ispn joined #git
06:17 preaction and, ... sorry, but... don't your bosses have better things for you to be doing?
06:17 Renter what sort of 'challenge' are we talking about, again ?
06:17 totemizer they have a month to do something that takes several hours
06:17 totemizer preaction: yeah, but those things i don't need help from the #git channel :)
06:18 totemizer i wasn't even sure if it's possible, or maybe there is a hard or easy solution for it
06:18 preaction right, i'm suggesting to do those, and not this. tell them i said it wasn't possible and move on
06:18 Renter several hours' worth of entry chalelnge? that seems silly
06:18 preaction like, i'm willing to take one for the team here
06:18 eroux joined #git
06:18 preaction fall on that grenade
06:18 oskarkv joined #git
06:18 totemizer Renter: it's blockchain (ethereum) specific thing
06:18 preaction go over the top into no man's land
06:19 grawity speaking of blockchains
06:19 Renter totemizer: interesting, but anyhoo no, if you put it in a git repo, no way to track it unless you have an external authentication method with logging that blocks passage to the git repo
06:20 Renter like others have said
06:20 grawity linux.git just merged two tags signed by a key that's been revoked since 2011 :|
06:20 totemizer yeah, i understand, thanks :)
06:20 mda1 joined #git
06:20 grawity my commit log is all red and stuff now
06:20 MuffinMedic joined #git
06:20 Renter sorry to repeat
06:20 totemizer Renter: :)) better twice than not even once
06:21 Renter is true
06:21 Renter sometimes the obvious things are the ones that most need saying
06:21 Renter (incidentally, phone your mom and wife/husband/SO and tell them you love them now)
06:23 dmc /o\
06:25 pfrench joined #git
06:25 dreiss joined #git
06:30 jameser joined #git
06:33 qt-x joined #git
06:33 dec0n joined #git
06:34 raynold joined #git
06:35 pfrench joined #git
06:37 MuffinMedic joined #git
06:37 duderonomy joined #git
06:38 MrcRjs joined #git
06:39 dersand joined #git
06:40 Chinggis6 joined #git
06:40 overlord_tm joined #git
06:42 marianina8 joined #git
06:43 PHPanos joined #git
06:43 Chinggis6 joined #git
06:44 Chinggis6 joined #git
06:45 jstimm joined #git
06:46 pfrench joined #git
06:46 inflames joined #git
06:47 serialoverflow joined #git
06:48 darkbit joined #git
06:48 hexagoxel joined #git
06:50 aidalgol joined #git
06:50 ToxicFrog joined #git
06:51 pfrench joined #git
06:52 jay-m joined #git
06:54 ThomasLocke joined #git
06:55 pfrench joined #git
06:57 Raging_Hog joined #git
07:00 pfrench joined #git
07:01 Chinggis6 joined #git
07:03 alexandre9099 joined #git
07:06 freimatz joined #git
07:06 pfrench joined #git
07:08 chipotle joined #git
07:09 lordjancso joined #git
07:13 navidr joined #git
07:13 xall joined #git
07:15 diogenese joined #git
07:21 darkbit joined #git
07:22 theoceaniscool joined #git
07:24 _ng joined #git
07:26 pfrench joined #git
07:27 JeroenT joined #git
07:30 pfrench joined #git
07:33 xkr47_ left #git
07:35 xkr47 joined #git
07:36 zefferno joined #git
07:37 impliednude joined #git
07:38 cdleonard joined #git
07:38 Cabanossi joined #git
07:42 masuberu joined #git
07:43 roelmonnens joined #git
07:43 jnavila joined #git
07:46 andrew710 joined #git
07:47 elect__ joined #git
07:49 PtxDK joined #git
07:50 darkbit joined #git
07:50 a_thakur joined #git
07:52 govg joined #git
07:52 pfrench joined #git
07:57 pfrench joined #git
07:59 hexagoxel joined #git
08:01 chele joined #git
08:03 absinthe joined #git
08:03 pfrench joined #git
08:08 absinthe joined #git
08:08 Cabanossi joined #git
08:10 sostenuto joined #git
08:11 aspiers joined #git
08:13 pfrench joined #git
08:14 dvaske joined #git
08:14 trion joined #git
08:18 pfrench joined #git
08:19 a_thakur joined #git
08:21 adac joined #git
08:22 inflames joined #git
08:22 syg_ joined #git
08:26 murii_ joined #git
08:28 ebcz joined #git
08:31 tg joined #git
08:32 pfrench joined #git
08:32 eroux joined #git
08:33 clmsy joined #git
08:33 murii joined #git
08:34 plasticfish joined #git
08:34 a_thakur joined #git
08:34 tvl joined #git
08:34 NwS joined #git
08:34 plasticfish Question: what is the difference between a pull and a fetch?
08:35 _ikke_ plasticfish: pull = fetch + (merge|rebase)
08:35 djb-irc joined #git
08:35 dimi1947 joined #git
08:35 plasticfish _ikke_, thanks
08:36 pfrench joined #git
08:36 plasticfish by the way what is a rebase?
08:36 _ikke_ git rebase is a command that allows you to move / combine / reorder commits
08:37 plasticfish ok :-)
08:37 Murii joined #git
08:38 jast !rebase
08:38 gitinfo 'git rebase' takes away your local commits, updates your branch with new stuff from <upstream> (argument), then re-applies your local commits on top. This makes it look like your commits were created "after" the new stuff, and it can look cleaner than doing a !merge. Beware of !rewriting_public_history, though. Not to be confused with !interactive_rebase.
08:39 pijiu joined #git
08:40 pfrench joined #git
08:40 TikityTik joined #git
08:40 jay-m joined #git
08:42 Balliad joined #git
08:42 ij I've a commit object 0a88…. Shouldn't .git/objects/0a/88 exist then?
08:42 chachasmooth joined #git
08:42 metalraiden34 joined #git
08:43 MTecknology Does anyone like gitea?
08:43 jast ij: at first, yes... but git optimizes your storage if you run 'git gc' (or a subset of that automatically every now and then), combining many loose objects (the one in .git/objects/xx/xxxx...) into a packfile
08:47 math-alpha joined #git
08:50 ij jast, The object is in the pack(evidenced by `git show-index < .g/.idx`), so I ran `git unpack-objects < .git/object/pack/appropriate.pack`, but it still didn't show up.
08:51 cbreak unpack will only unpack objects that aren't already in the repository
08:51 jast it says so right in the manpage for unpack-objects :)
08:51 WeiJunLi joined #git
08:52 ij But… .git/0a/88… doesn't exist, so it isn't!
08:53 Seveas 'in a packfile' is also 'in the repository'
08:53 ij ak
08:53 Seveas mv the packfile out of the repository before trying to unpack it if you really want the loose objects.
08:55 cbreak or just unpack them into a fresh repo
08:55 Pieplay joined #git
08:57 pfrench joined #git
08:58 Dumblez0r joined #git
08:59 rwp joined #git
09:02 pfrench joined #git
09:02 roelmonn_ joined #git
09:04 mikecmpbll joined #git
09:06 ij Okay, I found a different commit and printed its zlib-compressed output. It's a plaintextish file with a "commit 279\x00" at the beginning. Does git know it's a commit just because of that prefix?
09:06 a_thakur joined #git
09:07 pfrench joined #git
09:07 osse yes
09:07 Derperperd joined #git
09:08 ij So if I commit a file with that inside, it will register as a commit, not blob?
09:08 osse no
09:08 ij Oh, blobs will have the same prefix, but with "blob"?
09:08 * ij checks.
09:08 osse ij: git prepends stuff like that. so if you have a file with that in it, then git will prepend "blob N\x00"
09:09 ij cool
09:09 ij Nice that commit format is plain text.
09:09 osse it is?
09:09 Seveas it's not
09:09 Seveas nullbytes make it definitely not plain text :)
09:09 osse i would expect it to be riddled with NUls
09:10 alexandre9099 joined #git
09:10 Seveas osse: neh, the actual commit object is rfc2822 like text iirc.
09:10 Seveas rfc2822-like
09:10 ij Well that — yes, but after the \x00. http://sprunge.us/hXQU
09:11 Seveas trees are binary gobbledygook
09:11 osse oh i was confusing commits and trees
09:12 PHPanos joined #git
09:14 ij Knowledge has been gained! \o/
09:14 Seveas That'll be $5,95
09:14 osse a comma. how european
09:15 ij I'm european, but never use commas.
09:15 Seveas you used one right there...
09:15 jast no true Europeanman, then
09:16 ij Seveas, For money I mean. :P
09:17 jast you won't use one for money, but when no money is being offered you will? that seems... not very profitable
09:17 osse it's a charity
09:18 cbreak ,,
09:18 jast we'll be swimming in commas before we know it
09:18 Dumblez0r joined #git
09:18 Chinggis6 joined #git
09:19 osse It's a CSV
09:19 osse Comma Swimming Venture
09:19 marcogmonteiro joined #git
09:20 ThomasLocke_ joined #git
09:21 math-alpha joined #git
09:21 j08nY joined #git
09:21 Seveas Comma, comma, comma, comma, comma, chameleon...
09:24 Zoxc joined #git
09:24 kexmex joined #git
09:24 ronny hi
09:24 Seveas \o
09:25 ronny anyone aware of a easy way to move patch hunks between commits on a branch?
09:26 trion left #git
09:28 Seveas git format-patch | git apply -n; git add -p
09:28 Seveas err, no -n there.
09:28 pfrench joined #git
09:28 ferr1 joined #git
09:30 ij Is what you really mean `git f-p > /tmp/p; git checkout another-commit; git apply < /tmp/p; ...`?
09:31 Seveas neh, you can f-p random commits while on another branch
09:31 Seveas but you get the idea. Can do it with git show | git apply  too
09:31 ronny i did a set of commits on that branch, and i really just want to move patch gunks around and remove some
09:32 ronny lowlevel poreclair commands are something i want to avoid, because then i#d have to fix up my branch after generating the new commits
09:32 Seveas ronny: another alternative: cherry-pick all the commits onto the other branch and use rebas -i to remove the onwanted hunks
09:32 WayToDoor joined #git
09:33 ronny Seveas: still not helpfull, please presume im not a shell or git internals magican
09:34 manuelschneid3r joined #git
09:34 Seveas ronny: sorry, we don't do prechewed dinners here. I pointed you in theright direction, manpages give you all the details.
09:35 permalac joined #git
09:36 plasticfish how do i revert to a previous version of my work?
09:37 Seveas !revert
09:37 gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?
09:37 grawity revert *to* a previous version?
09:37 grawity that's usually `git checkout <oldcommit> .`
09:37 grawity although now that I think of it, no
09:38 plasticfish so we shouldn't delete branches?
09:38 Cabanossi joined #git
09:39 Sound joined #git
09:39 grawity how did you come to that conclusion
09:39 biertie joined #git
09:40 pfrench joined #git
09:40 plasticfish if I do a merge, will the branch remains after the merge?
09:41 grawity after the merge, the commits will exist in both branches at the same time
09:41 grawity deleting the old branch would then only delete the name
09:43 jay-m joined #git
09:43 Seveas !soviet-git
09:43 Seveas hmm
09:43 Seveas !soviet_git
09:43 gitinfo In Soviet Git, commits are not on branches, branches point to a commit.
09:44 pijiu2 joined #git
09:46 pfrench joined #git
09:46 alcohol joined #git
09:49 plasticfish what's can be the purpose of forking if we have branches (when someone is working alone) ?
09:49 plasticfish *what
09:50 ij I make forks, because I don't want to stash everything and make the editor confused just to do a little patch and send it to the remote.
09:51 ij Not always, but it's happened. (I'm guessing a fork = local copy of repo)
09:51 cdg joined #git
09:53 TomyLobo joined #git
09:53 Chinggis6 joined #git
09:53 pfrench joined #git
09:54 JanC_ joined #git
09:54 cdg joined #git
09:55 sQVe joined #git
09:55 aard_ joined #git
09:56 DaveTaboola joined #git
09:58 Paraxial joined #git
10:00 roelmonnens joined #git
10:01 impliednude joined #git
10:02 pfrench joined #git
10:02 chll joined #git
10:03 lmatteis joined #git
10:04 lmatteis joined #git
10:05 sbasso joined #git
10:08 cdown joined #git
10:08 Cabanossi joined #git
10:08 mischat joined #git
10:14 Paraxial joined #git
10:14 pfrench joined #git
10:15 jay-m joined #git
10:17 acetakwas joined #git
10:19 MrcRjs joined #git
10:21 pfrench joined #git
10:21 impliednude joined #git
10:22 senaps joined #git
10:23 tmg joined #git
10:25 meLon joined #git
10:25 cdown_ joined #git
10:25 pfrench joined #git
10:26 mischat joined #git
10:29 tcorneli joined #git
10:32 cdg_ joined #git
10:33 Snugglebash joined #git
10:33 Snugglebash joined #git
10:34 mischat_ joined #git
10:35 muhannad_______ joined #git
10:35 pbrewczynski joined #git
10:36 irqq joined #git
10:38 Cabanossi joined #git
10:38 raijin joined #git
10:40 masuberu joined #git
10:41 Tobbi joined #git
10:41 finalbeta joined #git
10:42 Rodya_ joined #git
10:43 senaps joined #git
10:45 hexagoxel joined #git
10:47 pfrench joined #git
10:48 gtristan joined #git
10:50 Posh joined #git
10:51 jay-m joined #git
10:52 jameser joined #git
10:53 CaptainHeavy joined #git
10:53 King_Hual joined #git
10:56 pfrench joined #git
10:57 jameser joined #git
10:59 mischat joined #git
11:06 tjbp joined #git
11:08 tjbp joined #git
11:08 Cabanossi joined #git
11:08 rcn joined #git
11:11 tjbp joined #git
11:15 mizu_no_oto joined #git
11:16 stoopkid joined #git
11:19 nettoweb joined #git
11:20 kexmex joined #git
11:20 xall joined #git
11:22 ezzieyguywuf joined #git
11:25 Posh joined #git
11:25 theoceaniscool joined #git
11:28 pluszak joined #git
11:30 pluszak So I have a repo with 258k files which clocks at 8.5 GB. When I do git commit strace show git stats all the files, is there an option to disable that? It seems all it does is list those files on commit message prompt
11:30 rnsanchez joined #git
11:31 _ikke_ pluszak: If you know in advance that certain files / directories will not change, you can use git update-index assume-unchanged
11:32 pluszak _ikke_: that is not the case I'm afraid
11:32 jameser joined #git
11:34 grawity commit with -m to skip the prompty entirely? :D
11:35 _ikke_ pluszak: git config commit.status false
11:35 pluszak But why is git commit stating everything? If I do git commit -m will it behave the same?
11:37 mniip joined #git
11:38 chuchunaku joined #git
11:40 [Brain] joined #git
11:43 marianina8 joined #git
11:44 _ADN_ joined #git
11:45 as_g5pw joined #git
11:45 Rodya_ joined #git
11:46 amrits joined #git
11:46 aleb Why is there (a post-merge hook but) no pre-merge hook? I see the pre-commit-msg could be used to prevent merges which are not good, as it's being run even when using "git merge <branch> -m message". It states "The purpose of the hook is to edit the message file in place [...].  It should not be used as replacement for pre-commit hook." so then using it as a
11:46 aleb replacement for the non-existing pre-merge might not be a good idea. Any suggestions?
11:48 senaps joined #git
11:48 osse aleb: the pre-commit  is invoked by git merge
11:49 aleb osse: not from what I tested, let me retry.
11:49 mischat_ joined #git
11:50 pfrench joined #git
11:50 osse aleb: I'm wrong
11:50 mischat_ joined #git
11:50 osse I checked the source. Two functinos had the same name
11:51 aleb yes, that was another thing.. it would be useful if pre-commit could be used
11:51 DolpheenDream joined #git
11:51 acetakwas joined #git
11:53 aleb I wonder why post-merge exists, since there is already post-commit which could be used very well
11:53 YYF joined #git
11:53 Cabanossi joined #git
11:53 dsdeiz joined #git
11:53 dsdeiz joined #git
11:56 unenforcer joined #git
11:57 pfrench joined #git
11:57 aleb I could also use post-push instead of these.. but it's missing too
11:58 grawity post-push?
11:59 synthroid joined #git
11:59 aleb yes, I know there is pre/post-receive server-side, but I need to run it locally.
12:00 _Cyclone_ joined #git
12:02 synthroi_ joined #git
12:03 scottjl joined #git
12:05 Dumblez0r joined #git
12:06 setuid left #git
12:10 PHPanos joined #git
12:11 jknetl joined #git
12:11 pluszak _ikke_: Yep, I guess that's what I wanted
12:11 gugah_ joined #git
12:14 Ardethian\work joined #git
12:15 aielima joined #git
12:16 leitao joined #git
12:16 dragoonis joined #git
12:18 eclecticjohny joined #git
12:19 YuGiOhJCJ joined #git
12:19 manuelschneid3r joined #git
12:19 mischat joined #git
12:25 xall joined #git
12:25 tcorneli joined #git
12:25 Calinou joined #git
12:26 DuncanT joined #git
12:27 dad27 joined #git
12:28 star_prone joined #git
12:29 dad27 joined #git
12:29 Calinou is it possible to automatically "git pull" a branch when a commit is made to it? how is this usually done? using a daemon/CI system?
12:30 Calinou (for deploying a Symfony project)
12:30 Calinou and then run a few commands (to compile Sass, for example)
12:30 osse !deploy
12:30 gitinfo Git is not a deployment tool, but you can build one around it (in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/deploy.html
12:31 Calinou thanks
12:32 erebel55 joined #git
12:34 erebel55 does anyone know how I can go back to the latest commit here?
12:34 erebel55 http://i.imgur.com/WX2lakA.png
12:34 erebel55 I did a git reset --soft HEAD~1
12:35 erebel55 and then I tried to get back by doing a git reset 1d67ae562faf6c87389ba6653ad960654378ff1c
12:35 erebel55 but my files aren't the same as that commit still
12:35 osse if you did reset HEAD~1 then 1d67ae562faf6c87389ba6653ad960654378ff1c is the new HEAD
12:35 osse try HEAD@{1}
12:35 osse do a git show HEAD@{1} and see if that's the correct commit
12:36 erebel55 okay
12:37 kpease joined #git
12:38 kpease_ joined #git
12:38 erebel55 @osse hmm that is giving me this
12:38 erebel55 http://i.imgur.com/aZvpw7p.png
12:38 Cabanossi joined #git
12:38 osse what does that tell you? it doesn't tell me anything
12:39 osse if you did git reset --soft HEAD~1 and nothing else since that then HEAD@{1} is the commit you want
12:39 erebel55 I don't even see that commit in the git log, so I'm confused
12:39 erebel55 I did
12:39 erebel55 git reset --soft HEAD~1
12:39 erebel55 git reset 1d67ae562faf6c87389ba6653ad960654378ff1
12:40 IntruderSRB joined #git
12:40 osse erebel55: that's because 'git log' lists the log from HEAD and backwards
12:42 erebel55 hmm okay
12:42 shinnya joined #git
12:43 osse so 1d67ae562faf6c87389ba6653ad960654378ff1 is the current HEAD
12:43 onmeac joined #git
12:43 osse (check that by doing 'git rev-parse HEAD')
12:44 rominronin joined #git
12:44 ti2 joined #git
12:45 Rodya_ joined #git
12:46 gugah joined #git
12:46 erebel55 osse, yeah that returns the id you pasted
12:47 erebel55 http://i.imgur.com/rS3ZOGs.png
12:48 GodGinrai joined #git
12:49 erebel55 osse, but my changes aren't there, so I don't understand
12:49 osse erebel55: do 'git log HEAD@{1}'. is 1d67ae... the second commit ?
12:49 erebel55 changes from that commit
12:49 freimatz joined #git
12:50 erebel55 osse,
12:50 erebel55 http://i.imgur.com/adfkHmd.png
12:50 jay-m joined #git
12:51 osse a "yes" would suffice
12:51 osse so, git reset HEAD@{1} will do what you initially asked for
12:51 erebel55 ahh sorry
12:52 erebel55 okay cool, I'll try that :)
12:53 whitby joined #git
12:53 pfrench joined #git
12:54 HardlySeen joined #git
12:56 sbasso joined #git
12:57 acetakwas joined #git
12:58 cdown joined #git
12:58 Tobbi joined #git
12:59 acetakwas joined #git
13:00 Achylles joined #git
13:02 leitao joined #git
13:03 EV3RYDAYPR0GRESS joined #git
13:05 cookie_mr joined #git
13:06 roelmonn_ joined #git
13:06 star_prone joined #git
13:07 masuberu joined #git
13:08 erebel55 that worked, thank you osse. Apparently there were some unstaged files that got reset, but I'll have to deal with that
13:08 pfrench joined #git
13:08 nettoweb joined #git
13:09 Derperperd joined #git
13:09 durham joined #git
13:10 Gsham joined #git
13:11 rominronin joined #git
13:11 clmsy joined #git
13:13 Greench joined #git
13:13 pfrench joined #git
13:13 porsche944 joined #git
13:15 tsdev joined #git
13:16 rnsanchez joined #git
13:18 Greench hi there. I'm a bit lost with my git, I did a mess. Should I git reset?
13:18 btree joined #git
13:18 pfrench joined #git
13:18 rudi_s Greench: First !backup, then work on a copy.
13:18 gitinfo Greench: Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
13:19 rudi_s And then describe the problem in detail.
13:19 Greench cheers
13:19 chipotle joined #git
13:19 btree left #git
13:19 xall joined #git
13:20 Greench I worked on a branch, I pushed the commits. The commits have been merged. I deleted my branch. Then I saw I missed a few things so I worked on the files, but forgot to recreate a branch, so I'm working on master. I don't care about my changes. I just want to dismiss all local changes and pull from master and recreate my branch.
13:21 Greench I did a git pull but it opened a merge request...did not understand that.
13:22 nettoweb joined #git
13:22 Dumblez0r joined #git
13:22 kbs joined #git
13:22 Greench rudi_s: should I git reset then or delete the whole folder and clone again, or simply delete the .git folder?
13:23 crose joined #git
13:24 flaviodesousa joined #git
13:25 rudi_s Greench: If you don't have any local commits or uncommitted changes (git diff to check), then running git reset --hard origin/master (or the proper name of the upstream branch you want to reset to) should work fine.
13:26 Greench all good, thanks rudi_s :)
13:26 Karazhan joined #git
13:28 zeroed joined #git
13:28 root4 joined #git
13:29 rudi_s np
13:29 darkbit joined #git
13:29 jay-m joined #git
13:30 dvaske joined #git
13:30 mmlb joined #git
13:33 ti2 I'd like to store annotated tag commits on a git server where I may only push branch refs to.  Is there a possibility to prevent tag commits from being purged by 'gc' without destroying the possibility to clone the entire repo?
13:34 ti2 (I tried to link a tag commit as a 'pseudo' file, committed the tree and pushed that; but it breaks 'git clone' on that repo.)
13:36 kbs joined #git
13:38 ExoUNX joined #git
13:38 eroux joined #git
13:39 shtrb joined #git
13:41 shtrb I followed a workaround attempt in git-svn to set core.fscache and core.ignoreSet to true to solve very long commits, but I got myself into a problem that some files no matter if I change them or not are not set as modified . even calling git diff filea does not see differences (git checkout -f . does revert to the correct state)
13:42 shtrb tried to undo both actions but getting the same result
13:42 shtrb any help would be appricated
13:42 xall joined #git
13:44 jeffreylevesque joined #git
13:46 Rodya_ joined #git
13:46 amosbird hi
13:47 amosbird how can I make git ignore these deleted files that were tracked https://paste.wentropy.com/truA
13:48 kadoban joined #git
13:48 roelmonnens joined #git
13:49 osse echo 'thirdparty/snappy-1.0.5/' >> .gitignore
13:49 roelmonnens joined #git
13:49 gugah_ joined #git
13:50 bitmod joined #git
13:50 koneko joined #git
13:50 davidfetter_ge joined #git
13:50 livingstn joined #git
13:51 netj joined #git
13:51 gugah__ joined #git
13:51 mikecmpb_ joined #git
13:52 rokups joined #git
13:53 gugah joined #git
13:55 rokups suppose i fork a project and then want to submit some changes back upstream. merging from upstream will cause conflicts. is there some "right way" of handling fork or it just boils down to manually resolving those conflicts when merging from upstream? not sure what i could search for, any pointers would be appreciated.
13:55 psprint left #git
13:55 gugah_ joined #git
13:55 qqx shtrb: Neither of those the settings you reference is recognized by git.
13:56 zivester joined #git
13:56 sbasso joined #git
13:56 IntruderSRB joined #git
13:56 cdown joined #git
13:56 jast rokups: if both sides add new things, conflicts are simply part of life sometimes...
13:57 amosbird osse: well it's in gitignore already
13:57 amosbird osse: it's in the .git commit tree and i just deleted them
13:57 osse amosbird: then make a commit
13:58 rokups jast: i understand that. im referring to compatible changes though. suppose i submit a single commit to upstream and it gets merged. now fork and upstream have same changes with different commit ids. any way to properly handle that collision or its just usual conflict resolution?
13:58 Gsham joined #git
13:59 masuberu joined #git
13:59 Greench so I'm working on a git, and each time I pull from master, I have to change the database path and other few things for development, and the site admin has to change these parameters back when merging my commits for production. How should I proceed to prevent this? I would like to avoid ignoring the whole files that contain these settings, if possible.
13:59 cdg joined #git
13:59 amosbird osse: well i remember there is a way to ignore such files
13:59 osse amosbird: yes and you got it. but you have to delete them first
14:00 amosbird assume-unchanged
14:00 grawity Greench: I'd say don't track the configs in Git in the first place
14:00 grawity since they're really server-specific
14:00 qqx Greench: !configs
14:00 gitinfo Greench: [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas
14:00 grawity hell, I hope you're not putting the database password in a Git repo
14:00 Greench I'll read that thanks
14:00 shtrb qqx , then I don't even have an idea what could make a gile to be ignored by git diff (not in ignorelist)
14:01 e14 joined #git
14:01 shtrb I would appriate ideas (some files are identified by git as modified after a modification)
14:01 qqx shtrb: Did you use `git update-index --assume-unchanged`?
14:02 Dougie187 joined #git
14:02 osse amosbird: that won't help when the files are gone
14:02 shtrb no
14:02 noecc joined #git
14:03 shtrb but maybe the client have done that , is it possible to undo such an action ?
14:03 pluszak Calinou: you want a CI server hooked to your repo
14:03 Brski__ joined #git
14:03 amosbird ah, it's  git ls-files --deleted -z | git update-index --assume-unchanged -z --stdin
14:03 qqx shtrb: You can do `git update-index --no-assume-unchanged` on each file. Or a single `git update-index --really-refresh` might take care of it.
14:04 osse huh
14:04 osse very well then
14:04 Brski__ left #git
14:04 Brski joined #git
14:04 shtrb wait , I can see one file have a smaller h in when doing git ls-files -v
14:04 Armin` joined #git
14:05 shtrb Thank you ! you brought to the right track
14:06 osse amosbird: Btw, it's not fool proof. !assume_unchanged
14:06 gitinfo amosbird: git update-index --assume-unchanged is sometimes suggested for keeping yourself from committing changes to a file, but it's actually a promise to Git that the file is identical to the committed version, to reduce unnecessary disk reads. Not only will Git feel free to discard the flag, it will also happily overwrite your changes whenever a checkout/merge/rebase/... updates it.
14:07 shtrb yes , but I didn't run it the only thing I did was to use add two core settings as a workaround . It did make the commits much faster but broke other stuff
14:08 shtrb git --version give me 2.8.4.windows.1 if anything
14:08 Brski Hey. I have to set up a dev and production environment for a website project. I gueess I'll have to create a "dev" branch for dev and use "master" for production. There are some files on the production server that are not in the git repo (as specified by .gitignore). My concern is that the files that are on the production server will be removed once I do a git pull. Am I wrong? Doesn't it work that way?
14:08 chipotle joined #git
14:08 selckin !deploy
14:08 gitinfo Git is not a deployment tool, but you can build one around it (in simple environments) or use it as an object store(for complex ones). Here are some options/ideas to get you started: http://gitolite.com/deploy.html
14:09 shtrb and thank you qqx
14:10 IntruderSRB joined #git
14:11 jast rokups: if it's done using a normal 'git merge' upstream, there are no duplicate commits and you won't get any conflict. if they use something like cherry-pick and you merge, Git will still be able to merge without needing manual conflict resolution if you don't have new changes close by.
14:11 rokups great, thank you jast
14:12 jast basically if Git sees that both sides of a merge make the exact same change somewhere, it's smart enough to just use that
14:12 _chrispop_ joined #git
14:17 eclecticjohny joined #git
14:17 shtrb oh and the workaround was from https://github.com/msysgit/msysgit/wiki/Diagnosing-why-Git-is-so-slow
14:17 zeroed joined #git
14:18 ShekharReddy joined #git
14:18 alexandre9099 joined #git
14:18 Timvde If I rebase commits A and B onto another branch and squash them into one, does the squashing happen first, or does git cherry-pick A and then amend the changes in B?
14:19 cdown joined #git
14:20 lowercaseman joined #git
14:21 _ikke_ Timvde: I think the latter
14:22 _ikke_ Timvde: rebase works very sequentally
14:22 zeroed joined #git
14:22 zeroed joined #git
14:22 Timvde _ikke_: Okay, then I should do the squashing myself first, otherwise I'm going to get really bad merge conflicts :P
14:23 nanga joined #git
14:23 nanga joined #git
14:23 miklcct joined #git
14:23 qqx shtrb: Ahh, looks like those are settings are unique to git for Windows. I was looking at the non-Windows source.
14:24 IntruderSRB joined #git
14:24 skelterjohn joined #git
14:25 shtrb qqx, yes the slow commit only exists on windows (booted on the same machine form linux and the problem does not exist at all)
14:25 Guest27155 left #git
14:25 Es0teric joined #git
14:26 lb1d joined #git
14:27 whaley joined #git
14:28 raijin joined #git
14:29 LeBlaaanc joined #git
14:30 Derperperd joined #git
14:32 j08nY joined #git
14:32 kbs joined #git
14:32 AaronMT joined #git
14:33 jameser joined #git
14:33 Tobbi joined #git
14:34 DarkPsydeLord joined #git
14:35 IntruderSRB joined #git
14:36 nettoweb joined #git
14:38 Cabanossi joined #git
14:42 acetakwas joined #git
14:42 dhollinger joined #git
14:42 IntruderSRB joined #git
14:43 eroux joined #git
14:46 Intruder_ joined #git
14:47 Rodya_ joined #git
14:52 tvw joined #git
14:53 Murii joined #git
14:54 theoceaniscool joined #git
14:55 troulouliou_div2 joined #git
14:55 andrew710 joined #git
14:56 AaronMT joined #git
14:58 hexagoxel joined #git
15:00 thiago joined #git
15:00 Eugen joined #git
15:02 bitmod what's the best mental model for how git works?
15:02 a_thakur joined #git
15:02 canton7 bitmod, a graph of commits! !bottomup
15:02 gitinfo bitmod: 'Git from the bottom up' starts with explaining the building blocks of git and proceeds to tell you how they fit together. http://ftp.newartisans.com/pub/git.from.bottom.up.pdf (http://ikke.info/git.from.bottom.up.pdf)
15:02 jast !gcs
15:02 gitinfo [!concepts] "Git Concepts Simplified" presents all the important structures in Git, step by step, including selected internal details so it becomes much easier to understand what's going on whenever something's going on. http://gitolite.com/gcs.html
15:02 canton7 also !learn-git-branching
15:02 publio joined #git
15:03 bitmod thanks
15:03 canton7 err, !learn_git_branching
15:03 gitinfo A nice tutorial visualizing git branching: http://pcottle.github.io/learnGitBranching/
15:03 perlpilot bitmod: I like the metaphor used in "git for ages 4 and up" personally.
15:03 canton7 jast, any chance of automatically turning _ into -, and vice versa? triggers have a horrible mix of the two currently
15:03 cdown joined #git
15:04 perlpilot bitmod: https://www.youtube.com/watch?v=1ffBJ4sVUb4
15:04 jast canton7: not sure if there isn't a case where the difference is important...
15:04 canton7 if there's a case where the difference is important, I'd call that an issue with the trigger in question :P
15:04 canton7 since who's going to remember which one is which?
15:05 jast I was thinking about symbol triggers like '--'... but I guess there doesn't need to be a '_' or '__' trigger
15:06 canton7 yeah, or special-case if the _ or - in question is the first char of the trigger
15:06 rmbeer joined #git
15:06 rmbeer hello
15:06 worktoner joined #git
15:06 durham joined #git
15:06 jast well, I can't do it in the query phase without a major headache, so it would involve rewriting the existing data
15:06 impliednude joined #git
15:07 worktoner I'm moving 10+ repositories from github onto a private gitlab server. Anything I should look out for?
15:07 leitao joined #git
15:08 canton7 or do something ultra-hacky like "if no results { if string contains '-' { replace '-' with '_' } else if string contains '_' { replace '_' with '-'} if managed to do a replacement { search again } }
15:08 hexa- use gitlabs importer feature if you want to import more than the tree itself, make sure you are on a recent gitlab version
15:09 canton7 ... since you don't tend to see _ and - mixed
15:09 jast worktoner: as long as your gitlab version is >=8.7, most data should easily migrate over. a few things might be left behind, such as projects/boards.
15:09 TbobbyZ joined #git
15:10 tang^ joined #git
15:11 worktoner jast: Ok I will take a look. I've run the gitlab importer on one repository as a test first.
15:12 worktoner hexa-: Yes the importer seems to be the best way to go
15:12 bitmod perlpilot: thanks, looking at it now
15:13 oskarkv joined #git
15:14 gtristan joined #git
15:15 mischat_ joined #git
15:15 rmbeer other day i suggest make a new protocol for git, and this idea is discarded... can known now why?
15:16 clemf joined #git
15:17 moritz where did you suggest it, and who discarded it?
15:17 worktoner jast: It appears I'm using the 8.17.2 version of gitlab
15:17 IntruderSRB joined #git
15:18 rmbeer moritz, here, something users
15:18 rmbeer and anyone admit
15:19 rmbeer s/admit/accept/
15:19 MuffinMedic joined #git
15:19 roelmonn_ joined #git
15:19 jast what are you trying to achieve by making a new protocol?
15:19 moritz rmbeer: Open Source generally works like this: either you do something yourself, or you convince somebody else to do it for
15:20 moritz rmbeer: so it seems you didn't convince anybody to do it for you; now you can consider doing it yourself
15:20 corentin moritz: you could say this of anything basically
15:20 roelmonn_ joined #git
15:20 rmbeer jast, a proper protocol for git is best...
15:20 corentin not just opensource
15:20 jast what do you mean by "proper"?
15:20 moritz corentin: well, there are cases where people have contractual obligations to do stuff
15:20 moritz corentin: but yes, it's pretty general :-)
15:20 rmbeer jast, proper, owner
15:20 chachasmooth joined #git
15:20 corentin moritz: ^^
15:21 rmbeer moritz, yes, i can make it... but no if anyone want this....
15:21 jast rmbeer: but what exactly gets better if you do that?
15:22 rmbeer moritz, also need more info before to proceed
15:22 boombatower joined #git
15:23 govg joined #git
15:23 shtrb left #git
15:23 rmbeer jast, like use the own git protocol for work in your system directory, but by TCP/IP. Can be a other alternative/choice to use openssh and http
15:24 MuffinMedic joined #git
15:25 moritz rmbeer: it's still not obvious which problem you try to solve
15:25 Shatnerz0 joined #git
15:25 rmbeer but also can take more work in the future for git with more code of this.... i see the source code and seem hard for add...
15:25 jast I'm still not seeing the advantage, really...
15:25 jast this would actually make things like authentication *more* complicated because then suddenly Git needs its own authentication framework
15:26 chipotle_ joined #git
15:26 jast by using SSH or HTTP we can let well-established third-party components take care of authentication, and administrators can re-use their existing configuration
15:27 jast I mean, Git already has its own native protocol (Git TCP, port assignment 9418) but it's anonymous
15:27 bitmod what's the difference between "writing a file" to the disk vs the repo?
15:27 rmbeer moritz, is best, http is only for view a site web page for info, ftp is for transfer files, ssh is for open a remote account shell. Why git can't have a own protocol for work with files? Also why git make a protocol with openssl and http, and only this two and not other like ftp?
15:27 sz0 joined #git
15:28 moritz rmbeer: git supports writing over HTTPS
15:28 jast bitmod: that's a bit apples vs. oranges, since the repo typically lives on the disk, too :)
15:28 moritz rmbeer: and the question "why not?" can easily be answered with "because there's extra effort, and extra security risks involved"
15:28 jast or, shall we say, cars vs. automotive vehicles
15:28 brent__ joined #git
15:28 Abbott Can you sync soft links with git? I want a softlink in bin that points to "../src/executable"
15:28 Abbott or is there a better way to do that?
15:29 jast Abbott: Git preserves symbolic links
15:29 marianina8 joined #git
15:29 bitmod jast: so when someone says "adding is writing to the disk", vs "adding is writing to the repo", what does this mean?
15:29 jast bitmod: I'm not sure what exactly they mean because these terms are a bit vague
15:29 rmbeer moritz, sure...
15:29 jast basically you have your working tree, where the files you're directly editing/looking at are living
15:30 jast and then there's the .git subdir which contains metadata, history etc.
15:30 Abbott is that how software developers setup their makefiles? I have a makefile that will build the executable in the src directory and then I was hoping to just sync a symlink that points into the src directory so I don't have to keep copying or linking or whatever
15:30 Abbott but that seems unorthadox. How do people normally handle this?
15:30 Abbott unorthodox
15:30 jast when you use 'git add' on a file in the working tree, that version of the file is recorded in Git's metadata in preparation for committing, and 'git commit' will create a reference to that piece of metadata
15:30 bitmod jast: isn't "adding" adding a file to the repo though?
15:31 rmbeer moritz, but also is bad idea, why you need install openssh, or apache for work with this... And unknown a method more easy for this...
15:31 bitmod jast: here's the context btw: https://youtu.be/1ffBJ4sVUb4?t=1066
15:31 jast I can't watch videos right now, sorry
15:31 rmbeer install, configure, test and error....
15:31 j08nY joined #git
15:31 jast rmbeer: and if you make your own protocol, you'll probably end up having to install an SASL provider, anyway
15:31 kbs jast: yes, adding a file also adds it to the object database within the .git directory
15:31 jast kbs: I'm not the one you need to tell about it ;)
15:32 kbs s/jast/bitmod
15:33 bitmod kbs: is the object database in some way distinct from the repo?
15:33 jast bitmod: well, in standard git terminology 'git add' *stages* file contents, meaning they're added to the staging area for inclusion in the next commit. "adding to the repo" is more informal and *probably* refers to that, but who knows :)
15:33 jast the object database is part of the repo
15:33 dhollinger joined #git
15:33 xissburg joined #git
15:33 jast for the most part, the repo consists of the object database and refs (named references to objects). refs are used to implement branches and tags, for instance.
15:33 Abbott also I have "src/binary" "bin/*" and "bin/!binary" in my .gitignore file, where bin/binary is a symlink to src/binary, but git status is showing that git is ignoring bin/binary even though I have "bin/!binary" in the .gitignore file. I have "bin/!otherfile" in the .gitignore file and it seems to see those just fine
15:34 cdg joined #git
15:34 jast Abbott: '!' only works at the very start of a pattern
15:34 jast so, the pattern should be: !bin/binary
15:34 gitinfo [!binary] Storing binary files in git causes repo balloon, because they do not compress/diff well. In other words, each time you change a file the repo will grow by the size of the file. See !annex for some solutions
15:34 bitmod jast: are ok, so a staged file is not permanent on the git tree until it is committed?
15:34 jast gah, shut up, gitinfo
15:35 Abbott jast: that did it. Thank you :)
15:35 IntruderSRB joined #git
15:35 jast bitmod: in a sense, yes. if you decide to stage a different version, the one you staged first will no longer be *referenced* by anything but the object containing the data will still exist until a later cleanup job gets rid of it
15:36 bitmod thanks
15:36 jast Abbott: I'd probably have the Makefile copy or symlink the file after building it. better than having a symlink pointing to nothing before the first run of the Makefile, IMO
15:36 rmbeer jast, what about of gitolite and garry? not is this a make all the code with permission and encryption method?
15:36 xall joined #git
15:37 rmbeer or recycle any code from openssh
15:37 pfrench joined #git
15:37 jast rmbeer: I don't know garry but gitolite essentially just helps you set things up and uses an SSH or HTTP server in the background. it doesn't handle authentication itself, either, it just puts keyfiles in the right place. the only thing it *does* handle is authorization.
15:37 IntruderSRB joined #git
15:38 Cabanossi joined #git
15:40 TheLawyer joined #git
15:40 al-damiri joined #git
15:40 TheLawyer hello everyone .. can someone tell me how to find requirements of opensource projects ... e.g. https://github.com/sugarcrm/sugarcrm_dev
15:40 jast TheLawyer: you mean dependencies?
15:40 synthroid joined #git
15:40 chachasmooth joined #git
15:41 pippijn left #git
15:41 netj joined #git
15:41 h12o joined #git
15:41 TheLawyer jast: I mean how and where developers document the requested project features by consumers or by themselves
15:42 TheLawyer it dpeends on who is deciding what to do .. I'm asking if there is a common convention to document that so I can read any open source project requirements
15:42 jast TheLawyer: varies a lot based on available tools and individual preferences. many projects on GitHub use the "Issues" function to collect both bug reports and feature requests.
15:42 impliedn joined #git
15:42 h12o joined #git
15:42 rominronin joined #git
15:42 durham joined #git
15:42 jast in the development of Git itself, on the other hand, everything is basically discussions on the mailing list
15:42 sixtyten joined #git
15:43 thiago TheLawyer: there is no common convention
15:43 h12o joined #git
15:43 kyan joined #git
15:44 jast and in some projects requirements/plans are simply not documented/recorded anywhere
15:44 thiago s/some/most/
15:44 jast :)
15:45 Chinggis6 joined #git
15:45 jast I guess I should know, I made a NIH tool this weekend, with 0 bytes of spec documents
15:46 bitmod jast: regarding branches, when you create a new branch, is the contents of that branch basically a copy of master?
15:46 bitmod (and then you add new files etc etc)
15:46 thiago bitmod: correct
15:47 mischat joined #git
15:47 kbs (I've always assumed most open source projects start because someone has an itch to scratch, rather than a formal requirements stage :)
15:47 Gustavo6046 joined #git
15:47 thiago bitmod: when you create a branch, the old and new branches are identical and even indistinguishable.
15:48 thiago bitmod: to tell them apart, you need information that isn't recorded in the branch, like your recollection of what you've done
15:48 Rodya_ joined #git
15:50 Random832 joined #git
15:50 Rodya_ joined #git
15:51 kbs Having started to watch the 'ages 4 and up' video, I can see bitmod's question about the difference between objects and the repository...
15:52 bitmod thiago: thanks
15:53 CaptainHeavy joined #git
15:53 bitmod kbs: yeah, still not 100% sure i understand the nuances yet
15:54 jast that's what we're here for
15:54 kbs bitmod: fwiw (in case it wasn't clear) both commits and 'binary objects' all go into the same object database. The analogy is less worse if rather than twisting the binary object to form a commit, he created a new object which pointed to the file he's adding
15:54 kbs I think the problem is that the analogy sort of confuses adding a file with creating a commit
15:55 kbs the label-is-branch analogy is pretty nice though
15:55 maurer a/go 24
15:55 chachasmooth joined #git
15:57 d0nn1e joined #git
15:57 jagob joined #git
15:58 obiwahn joined #git
15:58 obiwahn What are the best practices when working with submodules?
15:58 obiwahn how do you switch branches if you have a few submodules that might be different between branches
15:59 obiwahn how do you do a bisect when submodules are involved?
15:59 thiago you still bisect
15:59 thiago I don't see a problem
15:59 TheLawyer thanks guys
15:59 TheLawyer left #git
15:59 thiago if the submodule link is updated, you go into the submodule, ensure that it has the right commit, then build it
16:00 obiwahn but you need to update the submodules so it matches the commit
16:00 thiago yes
16:00 obiwahn so the bisect command does not take care of the submodules?
16:00 thiago correct, it does not
16:01 obiwahn so best you have a script that checks the correctness of your submodules?
16:01 thiago yes. That's "git submodule update"
16:01 dreiss joined #git
16:02 obiwahn and well it could be recursive
16:02 qqx There's currently work on adjusting `git checkout` to handle updating submodules, that may take care of bisect as well.
16:02 qqx But currently that needs to be handled somewhat manually.
16:06 sehnny joined #git
16:08 Cabanossi joined #git
16:09 cqi joined #git
16:10 muhannad_______ joined #git
16:10 kbs joined #git
16:11 obiwahn I think submodules are the way to go. And we introduced them lately in the company i work for, but there are people complaining and do really understand their pain. Switching branches is a basic operation. If that does not work, then something is wrong. I am not sure if i am able to convince them to use hooks or scripts that fix the issues they have. So i am really glad to hear that there might be improvements
16:11 sgen joined #git
16:11 obiwahn in the foreseeable future:)
16:11 obiwahn and I do really
16:13 impliednude joined #git
16:14 sarri joined #git
16:14 sarri joined #git
16:14 h12o joined #git
16:17 jimklo joined #git
16:17 Vampire0 obiwahn, in my personal opinion, submodules are not the right tool if you have multiple repos that you develop in parallel. They are fine for integrating code of a 3rd party that is only upgraded from time to time to a newer version. For composing multiple repos that you develop concurrently in, I'd probably choose another method like git-slave or repo (!subre)
16:17 gitinfo obiwahn: [!subprojects] So, you want to add git repositories inside of other git repositories? Well, you have four main options. First is to just do it, add the repo to the outer project's .gitignore, and treat them entirely separately. Best if they are entirely separate. Otherwise your best options are "!submodule", "!gitslave", and "!subtree". Try those commands in this channel, or in a PM to avoid flooding.
16:17 durham joined #git
16:19 madewokherd joined #git
16:23 iota_pi joined #git
16:28 acetakwas joined #git
16:28 eroux joined #git
16:29 orbyt_ joined #git
16:32 Noldorin joined #git
16:33 ResidentBiscuit joined #git
16:35 leitao joined #git
16:39 h12o joined #git
16:42 Eugene canton7 - This is a broken link https://github.com/canton7/SyncTrayzor#buy-me-a-beer
16:42 rominronin joined #git
16:42 canton7 Eugene, thanks - there's an update coming in the next release
16:42 Eugene With more beer?
16:43 canton7 switching it to point to a charity fundraiser page. Donations to me + income tax got too painful to be worth it
16:43 rominron_ joined #git
16:43 thiago .t
16:43 chachasmooth_ joined #git
16:43 Eugene I usually link to eff.org/donate
16:43 thiago oops
16:45 BackEndCoder joined #git
16:45 TikityTik joined #git
16:45 canton7 cool. I've set up JustGiving page for MSF, which will get pushed out in the next release
16:46 canton7 I want to get some funky little network usage graphs in first, though
16:47 synthroid joined #git
16:47 Dumblez0r joined #git
16:48 rmbeer left #git
16:48 Eugene I'm just reading about it now, looking into it for work("alternatives to dropbox")
16:48 vamiry joined #git
16:55 WeiJunLi joined #git
16:55 chachasmooth joined #git
16:56 ToxicFrog Oh sweet
16:56 ToxicFrog This will be useful for me
16:56 ToxicFrog I've been quite happy with Syncthing but it's definitely a second class citizen on windows.
16:58 Eugene We've historically used a Windows file server(and I've done the same personally), usually with a GPO to force My Documents onto it and a VPN so it works outside of the office
16:58 Eugene But..... yeah
17:00 ToxicFrog (I don't use it at work but I use it a whole lot at home)
17:00 chardan joined #git
17:02 MuffinMedic joined #git
17:03 Derperperd joined #git
17:03 erebel55 joined #git
17:05 chitopunk joined #git
17:05 fahadash joined #git
17:05 koneko joined #git
17:07 obiwahn Vampire0: we have 3rd party things like v8, rocksdb, boost, icu
17:08 obiwahn and we update them just form time to time
17:09 metalraiden34 joined #git
17:10 ChadTaljaardt joined #git
17:12 kfoonamalik joined #git
17:13 obiwahn !subtree
17:13 gitinfo The subtree merge method is great for incorporating a subsidiary git repo into your current one with "unified" history. Read http://git-scm.com/book/en/v2/Git-Tools-Advanced-Merging#_subtree_merge for more info, see also !git-subtree and !git-stitch-repo.
17:14 multi_io joined #git
17:15 zxd joined #git
17:15 zxd what does the .. mean in $ git push origin master ... To john@githost:simplegit.git    fbff5bc..72bbc59  master -> master
17:18 manuelschneid3r joined #git
17:19 Naan joined #git
17:19 govg joined #git
17:20 acetakwas joined #git
17:22 metalraiden34 joined #git
17:22 chachasmooth joined #git
17:23 xissburg joined #git
17:29 Gsham joined #git
17:29 kbs zxd: I believe that's the old and new values for the master label on origin.
17:30 xall_ joined #git
17:30 Jellyg00se joined #git
17:30 akkad joined #git
17:33 Dumblez0r joined #git
17:34 a_thakur joined #git
17:34 chachasmooth joined #git
17:34 mrslaughter1775 joined #git
17:34 nd joined #git
17:35 gunnaro joined #git
17:35 a3Dman joined #git
17:38 ToxicFrog zxd: what kbs said; additionally, it's presented like that so you can copy-paste it into stuff like 'git diff' and 'git log'.
17:38 Cabanossi joined #git
17:39 synthroid joined #git
17:40 cdg_ joined #git
17:40 zxd ToxicFrog: in git diff   it provides ranges or just diff from these two commits
17:40 synthroi_ joined #git
17:41 marianina8 joined #git
17:42 fedepad joined #git
17:42 aielima joined #git
17:43 Dumblez0r joined #git
17:43 chipotle joined #git
17:43 Rodya_ joined #git
17:45 a3Dman joined #git
17:45 Greench left #git
17:45 marianina8 joined #git
17:48 marianina8 joined #git
17:49 leitao joined #git
17:52 erebel55 are git lfs questions allowed?
17:52 kexmex joined #git
17:53 chachasmooth joined #git
17:55 mustmodify joined #git
17:56 mustmodify Ahhh! I just did `git merge master` rather than `git checkout master` ... those are not the same.
17:56 JrCs joined #git
17:56 mustmodify Should I uh... `git reset HEAD~1` ?
17:56 porsche944 you're done
17:56 mustmodify forever?
17:56 porsche944 blow away the repo
17:56 mustmodify Just start over?
17:56 mustmodify ok
17:56 porsche944 lol
17:56 porsche944 no
17:56 mustmodify sad.
17:56 porsche944 no
17:56 mustmodify But I guess that happens.
17:56 porsche944 yes git reset HEAD~1
17:56 ayogi joined #git
17:57 mustmodify ok, just wanted to confirm before I made it worse. :)
17:57 marianina8 joined #git
18:00 chachasmooth joined #git
18:00 matoro joined #git
18:00 h12o joined #git
18:00 rezrov joined #git
18:01 dreiss joined #git
18:02 duderonomy joined #git
18:02 dstolfa joined #git
18:03 dstolfa Hello, I'm curious about a particular case of git blame
18:03 dstolfa For example, when I mistyped a query yesterday, namely doing git blame <file> --since=2017, instead of --since=1.1.2017
18:03 dstolfa I got random lines of the file blamed on my local user, with the time of my last commit
18:03 dstolfa Except, I've never seen this code in my life
18:03 dstolfa Is this expected behaviour, and if so, why?
18:04 dstolfa I would expect to get the wrong output in the sense that I'm not getting the expected dates and what not, but not an outright wrong blame
18:05 vedranm joined #git
18:06 robotroll joined #git
18:07 dsteplight__ joined #git
18:11 ertesx joined #git
18:11 raijin joined #git
18:11 Sasazuka joined #git
18:12 Gsham joined #git
18:12 rezrov joined #git
18:14 rominron_ joined #git
18:15 Rodya_ joined #git
18:16 mknod joined #git
18:16 erebel55 does anyone know how I can debug a git push that just keeps restarting at around 300MB/1.3GB? It's pushing the lfs files
18:19 Droolio joined #git
18:19 cdg joined #git
18:20 marianina8 joined #git
18:20 l2y what should I do if I have staged changes on the branch X but I decided I want to apply them as a commit on the branch Y?
18:21 h12o joined #git
18:21 nellicus_ joined #git
18:21 grawity !float
18:21 gitinfo If you have made a change in your working directory and have NOT YET COMMITTED, you may "float" that change over to another (`git checkout oldbranch`) or new (`git checkout -b newbranch`) branch and commit it there.  If the files you changed differ between branches, the checkout will fail.  In that case, `git stash` then checkout, and `git stash apply` and go through normal conflict resolution.
18:21 a3Dman joined #git
18:21 robnester joined #git
18:21 l2y ah, of course! stash
18:22 l2y thanks, I forgot about this feature
18:22 rezrov joined #git
18:23 rezrov joined #git
18:23 eh3 joined #git
18:24 durham_ joined #git
18:24 dreiss joined #git
18:25 raijin joined #git
18:26 dandaman joined #git
18:26 svm_invictvs joined #git
18:27 mustmodify left #git
18:28 roelmonnens joined #git
18:28 apotry joined #git
18:29 crayon joined #git
18:29 absinthe joined #git
18:29 crayon joined #git
18:31 MrcRjs joined #git
18:31 csd_ joined #git
18:31 rominronin joined #git
18:34 eroux joined #git
18:35 vamiry joined #git
18:37 chachasmooth joined #git
18:38 rezrov joined #git
18:41 h12o joined #git
18:42 jnewt joined #git
18:42 durham joined #git
18:42 rezrov joined #git
18:42 a3Dman joined #git
18:47 hasc joined #git
18:48 hasc joined #git
18:48 Guma joined #git
18:49 Es0teric joined #git
18:49 MineCoins joined #git
18:49 rezrov joined #git
18:50 rezrov joined #git
18:51 Darcidride joined #git
18:52 tmg joined #git
18:52 TbobbyZ_ joined #git
18:53 csd__ joined #git
18:54 Levex joined #git
18:54 csd__ I'm having trouble authing with github. I'm trying to set up having a secondary account. I added the account name to my git config --local credential.username and i get asked for my pw when i try to push to a new repo for the first time. But it's not accepting my github pw--even though i can use it to log in to the website. Wondering whether i'm doing something wrong or whether i locked myself out or something
18:56 Junior`` joined #git
18:56 Junior`` left #git
18:56 Junior joined #git
18:56 Rodya_ joined #git
18:56 Soni joined #git
18:56 csd__ any idea what i'm doing wrong? it does prompt me using the correct username
18:57 Soni so I have 2 branches, `master` and `1.10`
18:57 Soni I implemented a feature in `master` but also fixed a bug in `master`
18:57 Soni they're separate commits
18:57 Soni is there any way to merge the bugfix into `1.10` without pulling in the feature?
18:57 hhee joined #git
18:58 csd__ nm the username was wrong :-/
19:00 Eugene Soni - man git-cherry-pick
19:00 gitinfo Soni: the git-cherry-pick manpage is available at http://jk.gs/git-cherry-pick.html
19:01 h12o joined #git
19:05 Soni Eugene: uh I did it wrong how do I cancel/undo?
19:05 Soni error: could not apply aa3cbac... PowerCrops 1.1.1 | hint: after resolving the conflicts, mark the corrected paths | [...]
19:05 t0by joined #git
19:05 t0by joined #git
19:06 Rodya_ joined #git
19:06 erebel55 is there anyway to ignore a certain file for specific users? or will they all have to create their own user specific gitignore?
19:08 Eugene Soni - !fixup
19:08 gitinfo Soni: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!
19:09 rwp Soni, That doesn't look wrong. It is just a !conflict.
19:09 gitinfo Soni: [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice! A nice video explaining merge conflicts: https://www.youtube.com/watch?v=zz7NuSCH6II
19:09 ahrs joined #git
19:10 Soni except I didn't want the PowerCrops 1.1.1 commit, I wanted the fix server commit
19:10 madewokherd usually whatever command you used --abort
19:10 Soni ok
19:10 Soni thanks
19:11 marianina8 joined #git
19:11 rwp erebel55, Ignores in a local working copy are in the repository. But the user could set their own global ignore for all of their projects. But something about this request feels wrong to me. Why would it need to be user specific?
19:13 Rodya_ joined #git
19:14 erebel55 rwp, I have a few huge files (500MB) that I only need changes from one user for. I don't need any changes made by other users and they don't need changes made by me to these files.
19:15 TbobbyZ joined #git
19:15 eclecticjohny joined #git
19:15 madewokherd I don't think ignores affect already-tracked files
19:15 diogenese joined #git
19:15 rwp madewokherd, I thought of that too. But that is still for all clones for all users.
19:15 sathed joined #git
19:16 erebel55 oh okay hmm
19:16 jstimm joined #git
19:18 rwp I am also shuddering at 500MB files checked into git coupled with tracking changes to them. I'm guessing they are probably binary. Which means the repo will grow about that size with every new commit of them.
19:18 jnavila joined #git
19:18 grawity that's not because they're binary
19:19 grawity but because most types of binary files don't delta-compress well
19:20 erebel55 I'm using git lfs
19:20 erebel55 for the big binary files
19:20 jnavila joined #git
19:22 h12o joined #git
19:25 marianina8 joined #git
19:26 nellicus_ joined #git
19:28 valize joined #git
19:30 livingstn joined #git
19:32 govg joined #git
19:32 dvaske joined #git
19:32 aidalgol joined #git
19:32 dvaske_ joined #git
19:36 grayjoc joined #git
19:39 fuzzybear3965 joined #git
19:41 Derperperd joined #git
19:42 h12o joined #git
19:42 gopar joined #git
19:44 irqq joined #git
19:44 livingst_ joined #git
19:45 theoceaniscool joined #git
19:45 rominron_ joined #git
19:45 cdown joined #git
19:45 penanze joined #git
19:46 nowhereman joined #git
19:47 sysanthrope joined #git
19:48 jstimm joined #git
19:48 marc_v92 joined #git
19:49 a_thakur joined #git
19:49 penanze joined #git
19:52 Levex joined #git
19:52 dersand joined #git
19:53 Cabanossi joined #git
19:54 fedepad joined #git
19:54 finalbeta joined #git
19:57 synthroid joined #git
19:58 cdg joined #git
19:58 gopar joined #git
19:58 fbwnd joined #git
19:59 satifant joined #git
20:00 dendazen joined #git
20:00 Rept joined #git
20:00 jmpp joined #git
20:00 * jmpp greets!
20:00 jmpp I'm in the middle of a failed rebase, with a commit that failed to apply due to a confligt
20:00 jmpp conflict
20:01 jmpp how do I see that commit's id?
20:01 jmpp 'git status' doesn't show it to me
20:02 Murii joined #git
20:02 _ikke_ jmpp: It should have mentioned it when it applied the commit
20:02 jmpp only gave me the commit log
20:02 h12o joined #git
20:02 jmpp and I know I can find it from that
20:03 jmpp but wouldn't it be 10000% better to give me the commit id right away?
20:04 adac joined #git
20:05 _ikke_ jmpp: "Could not apply 5bdf04978ad14bbc0554a2d77e275499da044095... a"
20:05 _ikke_ This is what it says for me
20:06 _ikke_ cat .git/rebase-merge/done
20:07 lucido-cl joined #git
20:08 PaperShuttle joined #git
20:09 mrcrjs_ joined #git
20:12 rossome joined #git
20:13 Perspicaciosity joined #git
20:13 Murii joined #git
20:13 maestrojed joined #git
20:13 gugah_ joined #git
20:15 miczac joined #git
20:17 cdown joined #git
20:18 maestrojed I have been trying to figure this out myself but am failing. I see a large chunk of code removed from a file in my 'prod' branch. I would like to know what commit removed this code. It is likely that the commit was made in another branch and merged in. I have tried 'git log -s "a substr of the removed code" prod -- path/to/file' but that didn't find anything. Any help is appreciated
20:18 thiago joined #git
20:19 _ikke_ maestrojed: right, because git log by default only looks at one side from the merge
20:19 maestrojed _ikke_ Ok, so how do I go about this?
20:19 _ikke_ maestrojed: try to add -c
20:20 maestrojed I tried that too. Got nothing but maybe I have that argument in the wrong order or something? 'git log -c -s "a substr of the removed code" prod -- path/to/file'
20:21 MrWoohoo joined #git
20:22 _ikke_ maestrojed: It could be you also need to add --full-history
20:22 maestrojed ok, thanks
20:23 h12o joined #git
20:23 maestrojed hmm, still nothing :(
20:26 _ikke_ not sure if typo, but it's -S, not -s
20:27 jmpp _ikke_: I don't have a .git/rebase-merge directory
20:27 jmpp odd
20:28 maestrojed yeah, typo, sorry. But still can't find this. Thanks for the tips though!
20:30 fuzzybear3965 joined #git
20:30 orbyt_ joined #git
20:34 sunrunner20 Can you practically or at all implement the pull request model without a git server like github or bitbucket?
20:35 fuzzybear joined #git
20:35 grawity sure
20:35 root4 joined #git
20:35 belak Open source projects have been doing that by mailing patches for ages
20:35 grawity the kernel devs have been doing so for a long time
20:35 belak :P
20:35 grawity I mean, you don't have to be a github or a bitbucket to run a simple Git server
20:35 sunrunner20 Ha ha
20:35 preaction sunrunner20: git format-patch and git am
20:36 preaction or !gitolite
20:36 gitinfo Gitolite is a tool to host git repos on a server. It features fine-grained access control, custom hooks, and can be installed without root. Download: https://github.com/sitaramc/gitolite Docs: http://gitolite.com/gitolite/ Quick example: http://gitolite.com/gitolite/overview/#basic-use-case
20:36 grawity just git-daemon (or gitolite if you want to get fancy), and git request-pull
20:36 hakan joined #git
20:36 belak Alternatively you can just do it with branches and controlling access to master
20:37 Atm0spher1c joined #git
20:37 sunrunner20 Googling just pointed me to gitorious, any opinions on that here?
20:37 grawity Gitorious was borged by GitLab last year or so
20:37 sunrunner20 So it's dead, k
20:37 sunrunner20 :/
20:38 grawity here's a typical linux.git pull request https://lkml.org/lkml/2017/2/20/331 https://lkml.org/lkml/2017/2/23/562 https://lkml.org/lkml/2017/3/12/29
20:38 grawity (with your typical Linus commentary)
20:38 sunrunner20 Unless I can convince money to pay for bitbucket
20:40 ochorocho joined #git
20:41 syg_ joined #git
20:41 inflames joined #git
20:41 kbs my 0.02 - mostly depends on the sort of workflow that makes sense for the people that will be working on the codebase.
20:41 Eugene !host_gui
20:41 gitinfo There are several options for self-hosting git repositories with a web GUI: !gitblit !gitlab !rhode_code !gerrit and !gogs. Keep in mind that these all take far more work than a (non-GUI) !gitolite install. An overview of nearly all GUIs is available at: https://git.wiki.kernel.org/index.php/Interfaces,_frontends,_and_tools
20:41 Eugene I use Gitblit at $DAYJOB; I like it.
20:42 Eugene If paying money is at all an option, Github organizations are pretty cheap and I like that more
20:42 fuzzybear3965 joined #git
20:43 sunrunner20 Needs to be locally hosted
20:43 sunrunner20 I think
20:43 h12o joined #git
20:44 kbs how many people will be working on itL
20:44 Ardethian joined #git
20:45 kbs ...it? and someone willing to do sysad work to keep it runnig?
20:45 sunrunner20 <25people
20:45 sunrunner20 And sysadmin duties will fall on me
20:45 sm_ joined #git
20:45 sunrunner20 So less sysadmining the better
20:45 hashtagstrashtag joined #git
20:45 sunrunner20 I'm proposing a hosted solution now
20:46 Levex joined #git
20:47 Silenced joined #git
20:48 kbs that sounds a sound strategy (private github maybe?) You'll give everyone a relatively well-documented workflow, and let you actually do work instead of hand-holding.
20:49 _ikke_ sunrunner20: gitlab community edition can be run locally
20:49 metachr0n joined #git
20:50 irqq_ joined #git
20:51 hahuang65 joined #git
20:52 originalslon joined #git
20:53 jeffreylevesque joined #git
20:53 Es0teric joined #git
20:53 Cabanossi joined #git
20:54 originalslon Is it possible to add additional information to branches?  I want to add some indication to a branch about which server it has been deployed to.
20:55 cjohnson originalslon: in the README is how I do that
20:55 cjohnson explain what each branch is for and does
20:55 chardan joined #git
20:56 rmbeer joined #git
20:57 shakalaka joined #git
20:57 porsche944 or there are tags
20:57 originalslon @cjohnson, yes, just a some meta data explaining which server the code has been deployed to
20:57 fuzzybear3965 joined #git
20:58 raspado joined #git
20:58 originalslon @porsche944 good idea with the tags
21:00 originalslon @porsche944 not sure it meets my use case though.  I don't want another branch/commit, ideally it is just a piece of meta data on the branch that shows to which server the code has been deployed.
21:04 h12o joined #git
21:04 fuzzybear3965 joined #git
21:04 xissburg joined #git
21:04 Snugglebash joined #git
21:04 porsche944 yeah
21:05 cluelessperson joined #git
21:06 sunrunner20 Eugene Fitbit looks promising
21:06 cluelessperson joined #git
21:06 Eugene I can attest to its minimal maintenance requirements - I ain't done shit in months.
21:07 PaperShuttle joined #git
21:09 m0viefreak joined #git
21:09 ChadTaljaardt left #git
21:10 Tobbi joined #git
21:10 cluelessperson joined #git
21:11 fedepad joined #git
21:11 Xethron joined #git
21:11 RalphJr45 joined #git
21:12 cluelessperson joined #git
21:12 jost__ joined #git
21:13 fuzzybear3965 joined #git
21:13 cluelessperson joined #git
21:14 floppyears joined #git
21:14 floppyears joined #git
21:14 _ADN_ joined #git
21:15 cluelessperson joined #git
21:16 rominron_ joined #git
21:18 dtrott joined #git
21:18 blackwind_123 joined #git
21:19 Soltis joined #git
21:19 ertes is there a way to make a repository more shallow by deleting objects?  i'm operating on a huge repo that changes quite often, and it's taking a lot of disk space
21:20 Soltis Is there a way to pull the last N commits off a different branch and rebase them onto the current one?
21:20 dsdeiz joined #git
21:21 fuzzybear3965 joined #git
21:22 Soltis Basically like an apply-patch type deal so I can get the work off this other branch when I don't need (or want) the commit IDs to remain the same.
21:22 rwp Soltis, man git-cherry-pick
21:22 gitinfo Soltis: the git-cherry-pick manpage is available at http://jk.gs/git-cherry-pick.html
21:23 rwp ertes, AFAIK one can only clone it again and at that time use --depth 1 to make it shallow. But if your repo is continuously growing then you would need to do that every time you wanted to trim. AFAIK.
21:24 ertes rwp: i see…  can/does git use reflinks (btrfs) for local cloning?
21:24 h12o joined #git
21:24 ojacobson it uses hard links for clones on the same FS, and it can use shared object stores using a git-specific mechanism
21:24 Eugene ertes - file:// URLs will be cloned using hardlinks if possible; man git-clone and see also --shared
21:24 gitinfo ertes: the git-clone manpage is available at http://jk.gs/git-clone.html
21:25 ojacobson Eugene: other way round
21:25 ojacobson file:// URLs are fresh clones, paths are cloned using hard links
21:25 Eugene Well I was close
21:25 Eugene Not bad for not readingf
21:25 rwp ertes, You can definitely use the -l, --local option when cloning. Uses hard links. Other than that I don't know. And don't think there is a btrfs specific option but I don't konw.
21:25 ertes thanks, people…  hard links are good enough
21:25 Soltis rwp: I was afraid you'd say that.
21:25 Soltis rwp: Another thing to memorize; joy.
21:26 ertes shared objects are indexed by hash?  that is: i can deduplicate between multiple repos?
21:26 rwp Soltis, Well... Another option is to branch, then use git rebase -i on the branch and modify it as you want. Then you can rebase against it. (shrug)
21:27 eclecticjohny joined #git
21:27 ertes oh, nevermind
21:27 rwp All objects, shared or otherwise, are indexed by hash. But off the top of my head I don't know if the new objects are hardlinked into the upstream repo. I don't think so but would need to test.
21:27 ertes it's something different…  i'll prefer hard links in this case =)
21:28 cjohnson I have a master, qa, and several features. How can I see which branches have been merged into qa that haven't been merged into master?
21:28 cjohnson Even "has ever been merged once" is enough for me
21:28 ertes i thought it would set up kind of a shared object repository that multiple repos could use, but it just uses the objects from the source repo
21:28 Eugene That's alternate worktree
21:28 Eugene man git-worktree
21:28 gitinfo the git-worktree manpage is available at http://jk.gs/git-worktree.html
21:29 mitzip joined #git
21:31 rwp cjohnson, I think that is the function of 'git cherry', to report which refs have not yet been picked. But possibly 'git rev-list' too.
21:31 ertes worktrees sound like they could solve a different problem for me (not this one though)…  thanks =)
21:32 eclecticjohny joined #git
21:33 impliednude joined #git
21:33 fedepad joined #git
21:33 eclecticjohny joined #git
21:36 kyan Hi! I deleted a bunch of files (~90mb), committed, pushed, made a new commit that undid the deletion, and pushed again. Why did it upload ~90mb again? The remote should have already had those items in its history, I'd think...
21:36 cjohnson thanks rwp
21:37 ojacobson kyan: pack negotiation does not guarantee that the pack is minimal
21:37 kyan ojacobson: Ahh, cool, I won't worry about it then. Thanks :)
21:37 ojacobson dup objects are harmless (to git, though perhaps not to your ISP bill) so git doesn't try too hard to eliminate them
21:38 raijin joined #git
21:38 gopar joined #git
21:38 kyan (my concern was if somehow the remote had lost objects, or my local objects differed from the remote's for some reason, is why I asked)
21:38 kyan Thanks!
21:43 tmsmith joined #git
21:44 cjohnson If I tag my branch, then delete the underlying branch, then push a new branch with the same name as the old branch, can I still get back to the previous state using that tag?
21:44 h12o joined #git
21:46 m0viefreak you don't tag branches, you tag commits
21:46 m0viefreak so yes
21:46 rwp cjohnson, Yes. Because the tag will associate with the hash and forever that hash won't change.
21:46 rkazak_ joined #git
21:47 cjohnson beautiful, thanks
21:48 svm_invictvs joined #git
21:48 nowhereman joined #git
21:58 jfelchner joined #git
22:00 dminuoso joined #git
22:00 dminuoso Hi. Why cant I git reset --hard a pathspec?
22:00 guardian left #git
22:00 kadoban dminuoso: Because what would that mean?
22:00 fuzzybear3965 joined #git
22:01 dminuoso kadoban: Now that you actually mention it..
22:01 dminuoso Yeah it makes absolutely no sense heh.
22:01 kadoban :)
22:02 Impaloo joined #git
22:03 h12o joined #git
22:04 kyan joined #git
22:04 impliednude joined #git
22:07 blackwind_123 joined #git
22:07 sbasso joined #git
22:08 Whiskey how do i revery all changes in files and get latest git?
22:08 Whiskey "revert"
22:08 Whiskey whit other words, put it back to the state from last commit
22:10 Whiskey git revert HEAD maybe
22:12 ojacobson !revert
22:12 gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?
22:12 Lunatrius joined #git
22:13 fedepad joined #git
22:14 Whiskey git reset --hard HEAD <- did work fine for me
22:20 dandaman joined #git
22:21 Wind0r joined #git
22:22 BackEndCoder joined #git
22:27 h12o joined #git
22:28 tekniq joined #git
22:28 CaptainHeavy joined #git
22:29 madewokherd joined #git
22:30 Impaloo joined #git
22:32 NeXTSUN joined #git
22:33 star_prone joined #git
22:35 Gsham joined #git
22:36 askb joined #git
22:38 Cabanossi joined #git
22:40 tang^ I'm drawing a blank here. I'm trying to find commits from a specific author on a specific (non-master) branch
22:41 tang^ nm, I'll add a date for when it was branched from master
22:44 star_prone joined #git
22:44 dsdeiz joined #git
22:44 dsdeiz joined #git
22:46 doublebit joined #git
22:47 DolpheenDream joined #git
22:48 rominronin joined #git
22:49 jimklo_ joined #git
22:50 bwn joined #git
22:50 Tobbi joined #git
22:53 impliednude joined #git
22:57 fbwnd joined #git
23:01 rkazak_ joined #git
23:01 whaley joined #git
23:01 matoro joined #git
23:02 tang^_ joined #git
23:04 phtes anyone around to give a second opinion on something
23:05 cagedwisdom joined #git
23:06 Rotaerk joined #git
23:07 Gustavo6046 joined #git
23:08 Cabanossi joined #git
23:08 xaviergmail joined #git
23:09 Ruzzy joined #git
23:11 Ruzzy my computer is origin and the VPS I deploy to is a remote. got a new person in who has to make changes to the application.
23:11 Ruzzy does he clone the repo and push to the same remote or do we do something else?
23:11 sircomun joined #git
23:11 dsdeiz_ joined #git
23:11 paws_ joined #git
23:12 floppyears Hi, I have a code base that relies on 2+ directories for the build. Each directory is a separate git repo. I've been learning Jenkins piepline and I know that I could specify multiple git urls. How do I configure the scm polling to trigger the job if any of the git repos receive commits?
23:14 Aeterne joined #git
23:15 a_thakur joined #git
23:15 bwn joined #git
23:15 steven joined #git
23:17 Ardethian\ joined #git
23:17 ok91 joined #git
23:19 jfelchner_ joined #git
23:20 guampa joined #git
23:22 Es0teric joined #git
23:28 chardan joined #git
23:28 whaley joined #git
23:29 jdelreal joined #git
23:29 safe joined #git
23:35 rkazak_ joined #git
23:36 aspiers joined #git
23:37 BackEndCoder joined #git
23:37 menip joined #git
23:37 Gustavo6046 joined #git
23:38 mizu_no_oto joined #git
23:39 thiago floppyears: not a Git question. You should ask in #jenkins or whatever forums they use.
23:39 _rgn joined #git
23:39 _rgn joined #git
23:40 hhee joined #git
23:42 venmx joined #git
23:43 floppyears thiago: thanks for pointing that out. I was distracted when I typed that ;)
23:44 Orphis joined #git
23:45 chachasmooth joined #git
23:48 jstimm joined #git
23:49 durham joined #git
23:53 Sasazuka_ joined #git
23:53 Cabanossi joined #git
23:54 livingstn joined #git
23:54 matoro joined #git
23:58 Gustavo6046 joined #git
23:59 damccull joined #git

| Channels | #git index | Today | | Search | Google Search | Plain-Text | summary