Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-01-20

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

All times shown according to UTC.

Time Nick Message
00:00 denete__ joined #git
00:04 worstadmin Is there a way to show the difference in commits between HEAD and origin/master:refs/remote/blah -- something like `git log HEAD..origin/master:refs/remote/1198` doesn't seem to work
00:06 Rodya_ joined #git
00:07 invisbl joined #git
00:08 jstimm joined #git
00:08 netj joined #git
00:10 worstadmin git-log doesn't seem to care about refspecs
00:11 dreiss joined #git
00:11 prg3 joined #git
00:14 canuk joined #git
00:17 mehola joined #git
00:19 durham joined #git
00:20 NaN joined #git
00:21 NaN I want to use vimdiff as difftool, I configure it has this post says > http://usevim.com/2012/03/21/git-and-vimdiff/
00:21 NaN but I still see all the diff like an output in cli
00:23 cyberz joined #git
00:23 NaN this is my global config >> https://ghostbin.com/paste/sevzv
00:24 parasite__ joined #git
00:24 durham joined #git
00:24 drewr joined #git
00:28 rewt NaN, are you using `git diff` or `git difftool` to diff?
00:28 MrLawrence joined #git
00:28 mischat joined #git
00:30 NaN rewt: git difftool
00:30 timvisher joined #git
00:31 MrLawrence joined #git
00:31 NaN rewt: I suspect it could be because vimdiff version or something
00:31 rewt that should bring up vimdiff... do you have a git config local to the repo?
00:31 NaN rewt: no local config
00:32 rewt and typing `vimdiff` on the command line brings up vimdiff?
00:32 NaN it does
00:35 NaN it wasn't vimdiff problem, I already update it :c
00:35 drewr joined #git
00:36 rominronin joined #git
00:39 rewt i just cleared my .gitconfig, and `git difftool` complained that it wasn't configured; then I did `git config --global diff.tool vimdiff`, and then `git difftool` used vimdiff
00:42 shoerain joined #git
00:42 sword```` joined #git
00:43 shoerain I get color codes like "ESC[33m", instead of the actual colors showing in `git diff` and `git log` in "Git Bash" on Windows. Any ideas on how I can fix this?
00:45 rewt shoerain, is the prompt coloured?
00:45 druonysus__ joined #git
00:45 shoerain yup, mintty (the terminal emulator) is
00:46 rewt what does it say for `echo $TERM`
00:46 gobbledygookstud joined #git
00:46 shoerain the ever venerable xterm-256color
00:46 rewt hmm
00:46 shoerain some things show color, like `git branch -a` and `git status`
00:47 rewt and `git diff` has no colours anywhere?
00:47 arescorpio joined #git
00:47 shoerain nope...
00:47 rewt weird
00:48 shoerain figured it's a setting i borked, but... not sure where to look. here's the [color] portion of my .gitconfig: http://sprunge.us/RcRD.
00:49 rewt try:  LESS=-R git diff ...
00:49 dhollinger joined #git
00:50 rewt btw, i didn't have to add any of that to my .gitconfig on windows
00:51 drewr joined #git
00:51 shoerain actually, i nuked my .bash_profile and it seems to render colors fine now
00:51 phanimah1sh joined #git
00:51 Anja joined #git
00:52 dreiss joined #git
00:52 jnewt joined #git
00:52 shoerain hmmm... I guess there's something in this .bashrc/.bash_profile that git doesn't seem to like: https://github.com/shurane/dotfiles/blob/master/.readable.bashrc
00:53 rewt maybe line 22
00:54 dezren39 joined #git
00:54 pity joined #git
00:55 shoerain yes it was! thanks rewt
00:55 shoerain guess I must've lowercased those env variables and flags
00:56 mentazoom joined #git
00:57 menip joined #git
00:58 Gloomy joined #git
00:58 gobbledygookstud hi
00:58 gitinfo gobbledygookstud: 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.
00:59 gobbledygookstud does anyone know a git command to make working area match staging area exactly as it is?
01:00 timvisher joined #git
01:03 gobbledygookstud does anyone know a git command to make working area match staging area exactly as it is?
01:03 Thorn joined #git
01:05 rewt maybe something in man git reset
01:05 gitinfo the git-reset manpage is available at http://jk.gs/git-reset.html
01:05 Sasazuka_ joined #git
01:08 gobbledygookstud I just went through it and non seem to please my needs
01:09 rewt have a look at the answers @ http://stackoverflow.com/questions/3107789/using-git-how-do-you-reset-the-working-tree-to-the-state-of-the-index
01:09 rewt few options presented there
01:09 gobbledygookstud what is the index? is it the same as  the staging area?
01:10 UnnamedPlayer joined #git
01:11 gobbledygookstud it's an old post rewt, it's a pitty it never got good attention
01:12 sword```` joined #git
01:12 gobbledygookstud excelent
01:13 gobbledygookstud git checkout-index
01:13 gobbledygookstud that does it
01:13 gobbledygookstud thank you rewt
01:14 denete_ joined #git
01:15 Gloomy joined #git
01:17 mvensky joined #git
01:19 gobbledygookstud left #git
01:21 inflames joined #git
01:25 nckpz joined #git
01:26 mvensky joined #git
01:29 mischat joined #git
01:30 safe joined #git
01:30 mvensky joined #git
01:33 dhollinger joined #git
01:33 umpc joined #git
01:34 dirtyroshi joined #git
01:35 timvisher joined #git
01:35 msonntag joined #git
01:37 wwweyewho joined #git
01:37 zong joined #git
01:41 Nilesh_ joined #git
01:46 EliasH joined #git
01:46 drbean joined #git
01:51 msonntag joined #git
01:52 pks joined #git
01:54 pity joined #git
02:02 SpeakerToMeat joined #git
02:03 Sasazuka joined #git
02:03 pity joined #git
02:03 nickabbey joined #git
02:04 pur3eval joined #git
02:07 parasite_ joined #git
02:07 baktrak joined #git
02:07 douglascorrea joined #git
02:07 rominronin joined #git
02:13 prg3 joined #git
02:14 aw1 joined #git
02:15 nckpz joined #git
02:17 lvh joined #git
02:17 Gloomy joined #git
02:18 timvisher joined #git
02:19 finalbeta joined #git
02:21 rkazak joined #git
02:23 anonymuse joined #git
02:26 Sasazuka joined #git
02:26 pity joined #git
02:30 mischat joined #git
02:34 EliasH joined #git
02:34 d^sh joined #git
02:36 pks joined #git
02:37 gugah joined #git
02:40 chipotle joined #git
02:40 sbeller joined #git
02:46 YuGiOhJCJ joined #git
02:46 dsantiago joined #git
02:49 roelmonnens joined #git
02:49 voldz1tic3n joined #git
02:51 fahadash joined #git
02:52 masuberu joined #git
02:53 drbean joined #git
02:54 aypea[3] joined #git
02:55 crayon joined #git
02:57 durham_ joined #git
02:57 Vampire0_ joined #git
02:58 Gsham joined #git
03:00 hodl joined #git
03:01 ikelso joined #git
03:01 Jonno_FTW joined #git
03:03 hodl I'm trying to replace a set of files with the content of another set of files that is also tracked by git. the files that I'm replacing have git histories that I wish I keep, and I don't care about the git histories of the new files. I keep encountering git detecting the files having been moved and the original git history of the files i'm replacing is
03:03 hodl getting superseded by the new history I don't care about. I knew this would be tricky before starting because git is good at detecting file moves and renames. anybody have any suggestion of how I might be able to accomplish this?
03:04 hodl to be more specific I want the replacements to be in a commit additional to the old histories I am trying to replace, and the history of the new files I don't care about. basically we switched scripting languages and want to retain the history.
03:04 bruce_lee joined #git
03:04 bruce_lee joined #git
03:05 hodl I did myself a disservice by git tracking the new transformed files in the same repository, or git wouldn't even be aware of them I think
03:06 hodl git blame is useless, but the per file history remains the same as the script code itself is the same, the syntax just changed
03:11 jstimm joined #git
03:12 timvisher joined #git
03:13 sunri5e joined #git
03:14 hexagoxel joined #git
03:16 jimi_ joined #git
03:18 Vortex34 joined #git
03:20 lvh joined #git
03:21 kyan joined #git
03:23 dreiss joined #git
03:25 MacWinner joined #git
03:26 pur3eval joined #git
03:28 aavrug joined #git
03:29 chachasmooth joined #git
03:33 mischat joined #git
03:34 midgetspy joined #git
03:36 fstd_ joined #git
03:40 rominronin joined #git
03:45 duderonomy joined #git
03:49 davimore joined #git
03:49 lvh joined #git
03:50 inflames joined #git
03:50 a_thakur joined #git
03:51 sadda joined #git
03:54 gfixler joined #git
03:59 pur3eval joined #git
04:00 timvisher joined #git
04:04 shgysk8zer0 joined #git
04:05 shgysk8zer0 left #git
04:12 sarbs joined #git
04:13 pity joined #git
04:15 meLon joined #git
04:15 ojacobson hodl: git's purely snapshot-based. "moved" and "renamed" detection is entirely post-facto.
04:15 ojacobson Create the tree you want to track and commit it.
04:18 mehola joined #git
04:19 sword```` joined #git
04:19 aypea[3] hi. is this scenario possible (does it even make sense)? http://pastebin.com/cwCU7hs1
04:19 Goplat joined #git
04:19 robotroll joined #git
04:20 niax joined #git
04:22 ojacobson aypea[3]: superficially - it looks possible, but deeply unusual unless loc master has unique changes
04:22 ojacobson Any DAG is _possible_, the ones that are good ideas are a somewhat more restricted set
04:23 pity joined #git
04:24 aypea[3] aypea[3]: unique changes?
04:24 aypea[3] sigh
04:24 * aypea[3] has gone mad
04:24 aypea[3] ojacobson: unique changes?
04:24 aypea[3] ojacobson: not sure what you mean
04:24 ojacobson commits which are not present on rem master or on any other branch in rem
04:25 finalbeta joined #git
04:25 aypea[3] ojacobson: loc master is just going to be my changes. very unlikely they'd wind up up-stream (remote).
04:25 ojacobson The normal fork workflow is to merge rem master periodically, not to merge random branches
04:26 ojacobson but you can do it your way if that makes sense in your environment
04:26 aypea[3] ojacobson: the loc master is kinda the one-true master. it would get fed from remote and be branched for use in production.
04:26 aypea[3] ojacobson: just not sure how merging stuff from remote master into it would work.
04:27 thiago joined #git
04:28 aypea[3] oh god. thinking about it breaks my brain a little.
04:29 dezren39 joined #git
04:32 mischat joined #git
04:32 timvisher joined #git
04:34 ketket joined #git
04:35 dez joined #git
04:36 jimi_ joined #git
04:38 dezren39 joined #git
04:39 t-mart joined #git
04:39 jimi_ joined #git
04:41 hobodave joined #git
04:42 ome joined #git
04:42 xall joined #git
04:43 dezren39 joined #git
04:44 Vgr joined #git
04:44 metachr0n joined #git
04:44 roelmonnens joined #git
04:44 NeverTired joined #git
04:44 Vgr joined #git
04:44 Vgr joined #git
04:44 ketket joined #git
04:45 dez joined #git
04:48 Darren_ joined #git
04:52 iamthad joined #git
04:54 arnsa joined #git
05:01 dermoth joined #git
05:11 dsdeiz joined #git
05:11 dsdeiz joined #git
05:11 rominronin joined #git
05:11 mehola joined #git
05:12 timvisher joined #git
05:12 lindenle joined #git
05:13 lindenle joined #git
05:13 pity joined #git
05:13 thierryp_ joined #git
05:15 mentazoom joined #git
05:18 cyan__ joined #git
05:20 miczac joined #git
05:23 aidalgol joined #git
05:26 ckruczek joined #git
05:30 armyriad joined #git
05:32 mohabaks joined #git
05:33 mischat joined #git
05:34 Garen joined #git
05:35 Cabanossi joined #git
05:36 tsunamie joined #git
05:36 armyriad joined #git
05:38 enleeten joined #git
05:40 simon816 joined #git
05:40 sarbs joined #git
05:41 giraffe joined #git
05:41 strugee joined #git
05:45 MrC joined #git
05:45 johnnyfive joined #git
05:46 sbulage joined #git
05:50 shgysk8zer0 joined #git
05:51 schneider joined #git
05:51 mahendratech joined #git
05:57 rominronin joined #git
05:57 svm_invictvs joined #git
05:59 pur3eval joined #git
06:01 dec0n joined #git
06:03 cyan__ joined #git
06:07 govg joined #git
06:10 MineCoins joined #git
06:12 thiago joined #git
06:12 pur3eval joined #git
06:19 lindenle joined #git
06:20 rscata joined #git
06:22 timvisher joined #git
06:22 thiago joined #git
06:26 fuzzybear3965 joined #git
06:26 fuzzybear3965 I have a bare repo on a remote that I want to clone using SSH.
06:26 fuzzybear3965 But, I'm getting a cryptic error: "fatal: protocol error: bad line length character: git@
06:26 fuzzybear3965 "
06:27 grawity what's the command you're using to clone
06:27 fuzzybear3965 Verbatim: 'git clone git@theline.design:website.git'.
06:28 fuzzybear3965 I've been here.
06:28 fuzzybear3965 http://stackoverflow.com/questions/8170436/git-remote-error-fatal-protocol-error-bad-line-length-character-unab
06:28 grawity and what do you get if you do `ssh git@theline.design git-upload-pack website.git`?
06:28 fuzzybear3965 I get a bunch of information, haha.
06:28 fuzzybear3965 It looks like a bunch of repo information.
06:28 grawity a list of refs, and then it hangs at 0000? that seems okay
06:28 fuzzybear3965 "00c9dd45844be1b12faae8a165fbef9a04671897701d HEADmulti_ack thin-pack side-band side-band-64k ofs-delta shallow no-progress include-tag multi_ack_detailed symref=HEAD:refs/heads/master agent=git/2.7.4
06:28 fuzzybear3965 003fdd45844be1b12faae8a165fbef9a04671897701d refs/heads/master
06:28 fuzzybear3965 0042e899f9822eb2dc9d87dfb76db06c365456354f4e refs/heads/printable"
06:29 fuzzybear3965 Yeah, exactly.
06:29 grawity hmm, and the output doesn't have 'git@' anywhere
06:29 fuzzybear3965 I'm not sure if it should hang, but it's not yielding an obvious error.
06:29 fuzzybear3965 I'm using pageant and Git for Windows.
06:29 fuzzybear3965 I can SSH in no problem.
06:29 grawity oh, so you're not using ssh – you're using plink
06:29 fuzzybear3965 ?
06:30 grawity openssh, rather
06:30 grawity if you're using pageant, then your (Git's) ssh client is plink
06:30 fuzzybear3965 Is that true?
06:30 fuzzybear3965 How can I see which SSH git is using?
06:30 grawity well, openssh does not support talking to pageant...
06:30 fuzzybear3965 That could be my problem, though.
06:30 diogenese joined #git
06:30 grawity try `env | grep GIT` – look for GIT_SSH
06:31 grawity also try `plink git@theline.design git-upload-pack website.git` to see what it outputs
06:31 fuzzybear3965 plink.
06:31 fuzzybear3965 Same thing.
06:31 fuzzybear3965 Yeah, it's using plink.
06:31 fuzzybear3965 The environment variable was exactly what you thought it would be.
06:31 grawity same thing but does it print 'git@' anywhere before the usual output?
06:31 fuzzybear3965 Uh, kind of.
06:32 xall joined #git
06:32 fuzzybear3965 I mean, I have to provide my password for the account git@theline.design.
06:32 grawity ah so it's not actually using a pubkey from pageant
06:32 fuzzybear3965 No.
06:32 fuzzybear3965 I'm password-authenticating.
06:32 fuzzybear3965 I'm not key-authing.
06:32 grawity well, that might be the problem then
06:32 fuzzybear3965 Is that a problem?
06:33 fuzzybear3965 I'll go ahead and add my public key into the known_hosts on the remote.
06:33 grawity IIRC plink didn't properly separate the password prompt from the output
06:33 grawity so either a) set up a pubkey, or b) unset GIT_SSH and clone using openssh as the client
06:33 fuzzybear3965 Does a) imply doing what I said above?
06:34 grawity if you meant "add my public key into the *authorized_keys* on the remote" then yes
06:34 phroa consider using !gitolite to manage repository storage and authorization
06:34 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.html#basic-use-case
06:34 overlord_tm joined #git
06:34 mischat joined #git
06:34 grawity (known_hosts is the reverse)
06:34 Mogget joined #git
06:36 fuzzybear3965 Oh, s***.
06:36 fuzzybear3965 Thanks.
06:36 fuzzybear3965 I just read what you wrote.
06:38 pur3eval joined #git
06:39 fuzzybear3965 Yo, I'm royally screwing this up.
06:39 fuzzybear3965 I copy-pasted my public key into the authorized_keys file in ~/.ssh for the git user on the remote.
06:40 fuzzybear3965 But, git@theline.design still prompts me for my password.
06:40 fuzzybear3965 Okay, this is strange.
06:40 fuzzybear3965 PuTTY authenticates using the public key just fine.
06:40 fuzzybear3965 But, Git Bash (using plink) doesn't authenticate using the public key just fine.
06:40 fuzzybear3965 My private key is loaded in pageant.
06:41 fuzzybear3965 Hey, grawity, I think we're good.
06:42 fuzzybear3965 I can git fetch and clone from the remote, now.
06:42 fuzzybear3965 Oh!!!!
06:42 EricBlade joined #git
06:42 fuzzybear3965 Within Git Bash I still couldn't git@theline.design without prompting a password.
06:42 fuzzybear3965 But, I realize why, notw.
06:42 fuzzybear3965 *now.
06:42 rominronin joined #git
06:43 fuzzybear3965 `ssh git@theline.design` uses MinGW's openssh binary.
06:43 fuzzybear3965 `git clone git@theline.design:website.git` uses plink (which has my private key loaded).
06:43 fuzzybear3965 That makes so much sense.
06:44 EricBlade hi all.  i've got a situation where i've got a public repo of code on github, but i want to also be able to develop pieces of it completely privately -- a subdirectory of the main, which will never be committed to github.  normally, i would just throw it in as a submodule, and call it good -- except i don't want to commit a reference to a private repo into the public repo.  is there some way to achieve this that I don't know of?
06:45 King_Hual joined #git
06:45 fuzzybear3965 Bam!
06:45 ThomasLocke joined #git
06:45 fuzzybear3965 ssh-agent and ssh-adding that key lets me SSH in without password.
06:45 fuzzybear3965 Thanks, grawity.
06:45 fuzzybear3965 You're the man!
06:46 fuzzybear3965 EricBlade, you're not cool with gitignoring the directory?
06:46 fuzzybear3965 You don't want anyone to see the folder name in the gitignore?
06:46 EricBlade i'm extending a project for a game engine, which requires everything in it be in one main directory, which is already part of the public repo ..
06:46 fuzzybear3965 Okay....
06:46 fuzzybear3965 So, are you cool with gitignoring the directory?
06:46 GodGinrai joined #git
06:47 fuzzybear3965 If no one else has the directory than the gitignore won't do anything good/bad for them.
06:47 EricBlade ... that would probably be acceptable.  can i just .gitignore the one subdirectory, and then clone a different repo into it?
06:47 ayogi joined #git
06:47 fuzzybear3965 Yep.
06:47 fuzzybear3965 Whatever's inside that directory won't be seen by git.
06:47 fuzzybear3965 You can put whatever you want in it.
06:47 fuzzybear3965 You can put a repo inside a repo inside a repo in that folder.
06:48 fuzzybear3965 And, strictly speaking, you don't *have* to gitignore.
06:48 fuzzybear3965 If you just don't add it to the repository then you're golden.
06:48 fuzzybear3965 But, if you're trying to avoid a git add -A messing up your entire life then, yeah, just gitignore it.
06:48 EricBlade yeah i don't like having my entire life messed up. i've already had that happen a few times in my time
06:49 fuzzybear3965 Haha, I feel that.
06:49 cyan__ joined #git
06:49 EricBlade hmm.  it does lose the ability to clone it in a single git command, but otherwise that would probably fit the bill.
06:49 dreiss joined #git
06:49 fuzzybear3965 What do you mean?
06:50 EricBlade with submodules, i (or my autobuilder) can clone --recursive and get the whole thing
06:50 fuzzybear3965 Oh, yeah.
06:50 EricBlade but i don't want to commit a reference to a submodule that isn't open, to github
06:51 fuzzybear3965 alias git clone A . && git clone B
06:53 spacelord_ joined #git
06:53 GodGinrai joined #git
06:53 GodGinrai whoa, it's weird seeing EricBlade talking in a different channel
06:53 GodGinrai XD
06:53 bayman joined #git
06:54 bayman hint: Updates were rejected because the remote contains work that you do
06:54 bayman hint: not have locally.
06:54 EricBlade or talking at all for that matter, right GodGinrai
06:55 bayman how do i push my local source as the de facto version on my repo?
06:55 GodGinrai EricBlade: lol yea XD
06:55 jameser joined #git
06:55 osse bayman: In other words forget about whatever is pushed already?  --force
06:55 GodGinrai bayman: you would use force
06:56 GodGinrai bayman: but beware... if anyone else is using it, it's basically a giant "fuck you"
06:56 bayman ok, i'll be the only one working on this
06:57 jim by forcing that, you're probably also forcing anyone who has the repo to abandon it (and possibly their uncommitted work) and reclone
06:58 osse jim: Only if they're noobs :P
06:58 GodGinrai jim: well, it's possible to rebase, depending on the situation... but either way, it creates a lot of needless effort
06:58 jim then you should consider whether to make a git repo public
06:59 Mo joined #git
06:59 brent__ joined #git
06:59 tvw joined #git
06:59 rominronin joined #git
06:59 Raging_Hog joined #git
07:00 qt-x joined #git
07:00 EricBlade oh.. i just figured something out.  I might be able to make the private repo have the public repo as a submodule, so long as the game engine respects all subdirectories of the "Assets" directory .. . . . .
07:00 pur3eval joined #git
07:00 jim socially speaking... making a repo public might be considered an invitation to help with something/work on it in some kind of shared way... it sounds to me like you don't want to extend that invitation
07:01 bayman left #git
07:04 Macaveli joined #git
07:04 GodGinrai jim: this is true, but many people extend the invitation by default because they want to make use of github's free hosting
07:05 EricBlade jim: just to clarify, you're responding to bayman and godginrai not me right?
07:05 pks joined #git
07:05 zeroed joined #git
07:05 zeroed joined #git
07:06 jameser joined #git
07:08 cdown joined #git
07:10 cdown joined #git
07:10 jameser joined #git
07:11 gottaname|wurk left #git
07:12 theoceaniscool joined #git
07:12 pur3eval joined #git
07:12 zeroed joined #git
07:12 zeroed joined #git
07:13 JeroenT joined #git
07:14 qt-x joined #git
07:15 dendazen joined #git
07:16 bocaneri joined #git
07:18 roelmonnens joined #git
07:18 GodGinrai EricBlade: you really know how to kill chat, huh? :P
07:19 EricBlade sorry :(
07:20 jeffspeff joined #git
07:21 GodGinrai well anyways, good luck w/ your git issues... I'm going to bed
07:21 pur3eval joined #git
07:25 zeroed joined #git
07:25 zeroed joined #git
07:26 timvisher joined #git
07:27 nidr0x joined #git
07:27 Spherical joined #git
07:27 mischat joined #git
07:28 jim EricBlade, I just thought it was socially misaligned (but not too much) that he (a) wanted to be the only one working on the code, and at the same time (b) made a mechanism for distributed work on the project (a git repo) public
07:28 mischat joined #git
07:29 jim EricBlade, so I was responding to him and not you
07:29 JeroenT joined #git
07:29 Sceorem joined #git
07:30 Pendrag0n joined #git
07:30 invisbl joined #git
07:32 lordjancso joined #git
07:33 blackwind_123 joined #git
07:33 Pendrag0n Hello Room, I've spent DAYS googling this (many hours multiple times, like probably over 40 hours without exagerating, so please no "go google it"), but with no luck, I cannot figure out a command to do the following,  I have a repo checked out, I type git status and it tells me that "Your branch is ahead of 'origin/master' by 1 commit." or 2 or 10.  How do I figure out which files have been commited but not pusehd?  After I can ge
07:33 MineCoins joined #git
07:35 phanimahesh joined #git
07:35 Pendrag0n no wait, that's wrong, git-diff doesn't show the differences if the file has already been commited.  Okay, so if I've already ran "git commit" but not "git push" I need to figure out how to A: see the list of files waiting to be pushed, then B: figure out how to do a diff from what I have and what is in the master branch.
07:36 jknetl joined #git
07:37 EricBlade jim: i thought so, just wanted to be sure, since we were both talking about not really public things right about the same time :)
07:38 Pendrag0n Also, does anyone in this room work for GitHub?
07:38 nostrora joined #git
07:38 jameser joined #git
07:40 worstadmin Why is it said git rebase can cause more merging conflicts than a git merge? I'm having trouble thinking about why
07:40 EricBlade Pendrag0n: try git show origin/master..HEAD ?
07:40 elsevero joined #git
07:40 redhedded1 joined #git
07:40 _ikke_ worstadmin: Each commit is applied individually, so each change all commits introduc can potentially conflict
07:41 spacelord_ joined #git
07:41 _ikke_ a normal merge only involves the end states of each branch
07:41 Pendrag0n OMG I just figured out the first part of my question, FINALLY.  it is "git diff --stat --cached origin/master"
07:41 worstadmin And git merge does what exactly? Builds a big patch of all the commits it will be applying and submits them all in one go?
07:41 Pendrag0n EricBlade, thank you
07:42 arnsa_ joined #git
07:42 freimatz joined #git
07:42 theoceaniscool joined #git
07:42 Pendrag0n and after running "git show -h" I found out that I can give it a -q flag, thus just getting a list of commits with notes, but without the individual diffs
07:43 Pendrag0n EricBlade, but your git show, well that answers my second question
07:43 elect__ joined #git
07:43 Pendrag0n thank you so, so much EricBlade.
07:44 Pendrag0n Okay, last question.  Does anyone here work for GitHub?  I had a meeting with a GitHub representative at an AWS Confernece in DC.  I represent BAH.  They promised a new feature, but unfortunately, I lost their contact info.
07:44 rishi` joined #git
07:44 EricBlade i don't know a lot of git commands, but it's useful to know that most git commands that accept a hash, will accept a range of just about anything you can identify a thing by .. ie, origin/master..HEAD
07:44 pdgwien- joined #git
07:44 tinanoo1 joined #git
07:45 draget_ joined #git
07:45 nephyrin` joined #git
07:45 Pie_Napple_ joined #git
07:45 heftig_ joined #git
07:45 VampiricPadraig_ joined #git
07:45 matth1askrgr joined #git
07:46 slikts joined #git
07:46 slikts joined #git
07:46 gsingh93- joined #git
07:46 AustinIncognito joined #git
07:46 JuliusN_ joined #git
07:47 Neo_ joined #git
07:47 thib_ joined #git
07:47 avar joined #git
07:47 avar joined #git
07:47 dsop joined #git
07:47 APic_ joined #git
07:47 bittyx joined #git
07:47 bttf_ joined #git
07:47 Chinorro_ joined #git
07:47 SwiftMatt joined #git
07:48 ghormoon_ joined #git
07:48 Pol joined #git
07:48 kba_ joined #git
07:48 tuor_ joined #git
07:48 AlexanderS_ joined #git
07:48 jhasse_ joined #git
07:48 jagob joined #git
07:49 overlord_tm joined #git
07:49 cmn_ joined #git
07:50 rideh- joined #git
07:50 barteks2- joined #git
07:50 daurn joined #git
07:51 unilovelight[m]1 joined #git
07:51 jokester_ joined #git
07:51 _ikke_ worstadmin: a 3-way merge between the head of branch1, head of branch2 and merge-base
07:51 _ikke_ so only 3 commits are involved
07:51 worstadmin Ah I see - so only between HEAD commits
07:51 _ikke_ !HEAD
07:51 _ikke_ HEAD != head
07:51 worstadmin I'm not sure how they differ
07:51 _ikke_ A repository only has one HEAD
07:51 worstadmin Ah okay
07:51 _ikke_ which is a symbolic ref pointing to the currently checked out branch or commit
07:51 worstadmin Im noticing a lot of commits in my rebase onto upstream (commits coming from my working directory) are conflicting with each other and I keep thinking that this is probably expected - which leads me to think rebase is better when there are fewer commits being applied to upstream
07:51 _ikke_ worstadmin: Note that the conflicts tend to be smaller (because the commits are smaller)
07:51 _ikke_ when rebasing
07:51 worstadmin True in most cases I imagine
07:51 Moggets joined #git
07:51 worstadmin thanks _ikke_
07:51 gitinfo joined #git
07:52 M-ou-se joined #git
07:52 peepsalot joined #git
07:52 leah2 joined #git
07:52 aviraldg joined #git
07:52 jason_ joined #git
07:52 enjunear[m] joined #git
07:52 up_the_irons joined #git
07:53 a_thakur joined #git
07:53 xall joined #git
07:53 Hobbyboy joined #git
07:53 qlkzy joined #git
07:53 a-l-e joined #git
07:53 sunny256 joined #git
07:54 Seveas joined #git
07:54 pur3eval joined #git
07:54 Takumo joined #git
07:54 Takumo joined #git
07:55 ghormoon joined #git
07:55 Macaveli joined #git
07:56 Alienpruts joined #git
07:57 a_thakur_ joined #git
07:57 mar77i_ joined #git
07:58 dikiaap joined #git
07:58 thiago joined #git
07:58 eduardas_m joined #git
07:58 fragMental joined #git
08:00 davimore_ joined #git
08:00 cdown joined #git
08:01 Derperperd joined #git
08:03 elsevero joined #git
08:03 cust joined #git
08:04 stickperson joined #git
08:05 Dev0n joined #git
08:10 Tobbi joined #git
08:11 UTAN_dev joined #git
08:13 johnny56_ joined #git
08:13 ihradek joined #git
08:14 hexa- joined #git
08:15 HcsOmot joined #git
08:15 jnavila joined #git
08:19 timvisher joined #git
08:20 mikecmpbll joined #git
08:20 m1m3-50 joined #git
08:24 johnny56_ joined #git
08:24 nettoweb joined #git
08:25 xberg joined #git
08:27 _geoff joined #git
08:27 Sceorem_ joined #git
08:27 invisbl joined #git
08:28 dersand joined #git
08:28 mar77i joined #git
08:29 ReenignE joined #git
08:29 ReenignE joined #git
08:30 chipotle joined #git
08:32 styx__ joined #git
08:32 jim EricBlade, no worries... I should probably do the nick: thing more often; I got out of the habit over the years
08:32 BinGOs joined #git
08:33 ac3takwas joined #git
08:33 misha joined #git
08:35 fuzzybear3965 joined #git
08:38 Alienpruts joined #git
08:38 elsevero joined #git
08:43 Alienpruts joined #git
08:44 JeroenT joined #git
08:45 garethdaine joined #git
08:46 Meteorhead joined #git
08:46 JeroenT_ joined #git
08:48 pur3eval joined #git
08:48 Tobbi joined #git
08:51 auzty joined #git
08:53 Balliad joined #git
08:54 JeroenT joined #git
08:57 mikecmpbll joined #git
09:02 e joined #git
09:02 kurkale6ka joined #git
09:02 djb-irc joined #git
09:04 matth1askrgr joined #git
09:04 matth1askrgr joined #git
09:04 bongjovi joined #git
09:05 SilverBe joined #git
09:06 UnnamedPlayer joined #git
09:06 elbeardmorez joined #git
09:06 jagob joined #git
09:08 cdown joined #git
09:09 hahuang61 joined #git
09:09 timvisher joined #git
09:10 monkeycat joined #git
09:10 SwiftMatt joined #git
09:13 Bloo[m] joined #git
09:14 leah2 joined #git
09:14 sunri5e joined #git
09:15 marcogmonteiro joined #git
09:15 peepsalot joined #git
09:16 jameser joined #git
09:17 buckowski joined #git
09:18 liberpater joined #git
09:19 Sri[m] joined #git
09:20 djb-irc joined #git
09:20 buckowski joined #git
09:21 timini joined #git
09:21 jameser joined #git
09:25 phanimahesh joined #git
09:25 clmsy joined #git
09:25 minikN joined #git
09:26 sea-gull joined #git
09:26 dhollinger joined #git
09:27 qlkzy joined #git
09:28 Tobbi joined #git
09:30 phanimah1sh joined #git
09:30 Guest93002 joined #git
09:31 GT4066 joined #git
09:31 Shashikant86 joined #git
09:31 zugzug joined #git
09:31 lb1c joined #git
09:32 cust joined #git
09:34 rivarun joined #git
09:37 mindfart joined #git
09:37 ramsub07 joined #git
09:39 hhee joined #git
09:41 djb-irc joined #git
09:42 rivarun joined #git
09:42 govg joined #git
09:45 cdown joined #git
09:45 chll_ joined #git
09:46 overlord_tm joined #git
09:46 lindenle joined #git
09:47 Shashikant86 joined #git
09:48 pur3eval joined #git
09:51 dminuoso joined #git
09:52 JeroenT joined #git
09:52 apotry joined #git
09:53 dminuoso I have a git sub-module "foo", and I added a remote "localfork". The entire history in localfork was rewritten. How can make "foo" to just be "localfork/feature" - I don't want to solve conflicts.
09:53 stefanha left #git
09:54 selckin don't know much about submodules, but sounds like removing foo and adding the new one could work?
09:54 timvisher joined #git
09:54 dminuoso Well, I just want to checkout that feature branch
09:55 dminuoso mmm
09:56 dminuoso Ohh this is so silly.
09:56 dminuoso I just recognized the mistake I was making.
09:56 dminuoso Nevermind.
09:56 jozwior_ joined #git
09:58 ferr1 joined #git
09:58 djb-irc joined #git
10:01 Serpent7776 joined #git
10:01 Tobbi joined #git
10:05 e joined #git
10:06 phanimahesh joined #git
10:08 m4sk1n joined #git
10:11 aw1 joined #git
10:12 dodobrain joined #git
10:15 noidea joined #git
10:19 sven^ joined #git
10:19 sbulage joined #git
10:22 gabi_ joined #git
10:23 dodobrain joined #git
10:24 wid` joined #git
10:27 Snugglebash joined #git
10:28 borkr joined #git
10:29 omegak joined #git
10:29 afuentes joined #git
10:32 mahendratech joined #git
10:35 achadwick joined #git
10:39 ShalokShalom joined #git
10:40 timvisher joined #git
10:41 [0xAA] joined #git
10:42 manuelschneid3r joined #git
10:44 mablae joined #git
10:45 andschwa joined #git
10:46 SwiftMatt joined #git
10:47 notebox joined #git
10:48 magicturtle joined #git
10:49 wootehfoot joined #git
10:49 jameser joined #git
10:49 jozwior_ joined #git
10:49 magicturtle joined #git
10:52 ShalokShalom joined #git
10:53 Achylles joined #git
10:53 mehola joined #git
10:55 thierryp joined #git
10:55 nic__ joined #git
10:55 elbeardmorez joined #git
10:55 phanimahesh joined #git
10:55 douglascorrea joined #git
10:56 star_prone joined #git
10:56 xall joined #git
10:58 nanga joined #git
10:58 nanga joined #git
11:00 brent__ joined #git
11:01 msonntag joined #git
11:01 cdg joined #git
11:01 Tobbi joined #git
11:02 tinti joined #git
11:05 mentazoom joined #git
11:05 ShalokShalom joined #git
11:09 hahuang61 joined #git
11:10 bl0w3d_0ut joined #git
11:12 alwyn joined #git
11:12 alwyn Hi, is it possible to update a submodule without initializing/cloning it?
11:13 grawity you mean to repoint it to a newer commit?
11:13 alwyn grawity: yes.
11:13 rivarun joined #git
11:13 grawity probably with `git update-index`
11:13 alwyn Specifically, for simplicity, to its current master.
11:17 LBV_User joined #git
11:18 Snugglebash joined #git
11:20 timvisher joined #git
11:22 ssitani joined #git
11:26 Shashikant86 joined #git
11:28 davimore__ joined #git
11:29 drbean joined #git
11:31 Tobbi joined #git
11:33 bl0w3d_0ut joined #git
11:34 EricBlade joined #git
11:36 CussBot joined #git
11:36 koneko joined #git
11:40 Cavallari joined #git
11:41 ramsub07 joined #git
11:42 JeroenT joined #git
11:42 ElGrandeAmigo joined #git
11:45 aard_ joined #git
11:47 dendazen joined #git
11:50 ShalokShalom joined #git
11:51 JuliusN_ left #git
11:51 ShalokShalom joined #git
11:52 Tobbi joined #git
11:53 ertesx joined #git
11:54 pity joined #git
11:55 str joined #git
11:57 tribly joined #git
11:59 MrLawrence joined #git
12:03 Derperperd joined #git
12:03 JeroenT joined #git
12:07 Gloomy joined #git
12:10 timvisher joined #git
12:13 chipotle joined #git
12:15 YuGiOhJCJ joined #git
12:18 APic joined #git
12:20 al-damiri joined #git
12:21 jknetl joined #git
12:23 mentazoom joined #git
12:23 Cavallari1 joined #git
12:26 rnsanchez joined #git
12:27 rishi joined #git
12:27 jameser_ joined #git
12:29 jozwior_ joined #git
12:31 chele joined #git
12:32 xkr47 joined #git
12:33 jozwior_ joined #git
12:34 m1m3-50 joined #git
12:34 mizu_no_oto joined #git
12:34 aard_ joined #git
12:35 aard__ joined #git
12:35 rivarun joined #git
12:39 ispn joined #git
12:41 Ryanar joined #git
12:43 jozwior_ joined #git
12:46 Bizkit joined #git
12:46 ssitani joined #git
12:47 Darcidride joined #git
12:48 timvisher joined #git
12:49 Tobbi joined #git
12:52 Cavallari joined #git
12:53 mehola joined #git
12:55 phanimah1sh joined #git
12:56 Derperperd joined #git
12:57 Mimicize joined #git
12:57 JeroenT joined #git
12:58 jameser joined #git
12:58 mischat_ joined #git
12:58 Guest36461 joined #git
12:58 cust joined #git
12:59 RxMcDonald joined #git
12:59 RxMcDonald I added a file to the gitignore but git diff shows the diff on the file, why?
13:00 osse RxMcDonald: because it has already been added to the repo
13:00 osse RxMcDonald: !ignore_tr
13:00 gitinfo RxMcDonald: [!ignore_tracked] Git only applies ignore patterns to untracked files. You can't use ignore patterns to ignore changes to files that are already tracked by git. To remove files only from git, but keeping them on disk, use git rm --cached <file>. Still, see https://gist.github.com/1423106 for ways people have worked around the problem.
13:01 mischat_ joined #git
13:02 synthroid joined #git
13:03 ssitani joined #git
13:04 synthroi_ joined #git
13:05 jameser joined #git
13:08 sword```` joined #git
13:09 BSaboia joined #git
13:09 jameser joined #git
13:09 mikecmpbll seriously confused by http://nvie.com/posts/a-successful-git-branching-model/ . i've not been able to make it work as simply as it sounds. take a bugfix for example; i branch off master, fix a bug, merge that back in to master and dev with --no-ff, so i have a different merge commit in master and dev. now when i create a new bugfix branch off master and merge in to dev, it's trying to merge the merge commit from the previous bugfix as well...
13:10 star_prone joined #git
13:10 hahuang61 joined #git
13:10 Serious8 joined #git
13:11 treia joined #git
13:12 timini_ joined #git
13:17 osse mikecmpbll: sounds to me like you're not confused, but rather don't like that approach
13:17 mikecmpbll osse : might be a bit of both, but i'm definitely confused.
13:17 osse what's the confusing part?
13:17 anonymuse joined #git
13:17 mikecmpbll that doesn't seem right, yet it makes no mention of it?
13:18 osse hmm, the diagram never shows merging directly from master into dev
13:19 mikecmpbll i didn't suggest that either, just a bugfix branch from master gets merged in to dev.
13:19 osse i misunderstood
13:19 mikecmpbll master --branch--> bugfix branch --merge--> dev & master
13:19 nettoweb1 joined #git
13:19 osse by bugfix, do you mean what the diagram calls hotfix?
13:20 mikecmpbll apologies, yes I do.
13:20 ac3takwas joined #git
13:20 mikecmpbll and dev == develop.
13:20 osse it would be more apparent if the diagram showed more than one
13:20 mikecmpbll releases branches are a similar situation
13:21 marcoslater joined #git
13:21 Snugglebash joined #git
13:21 mikecmpbll actually no they're not because they branch from dev.
13:22 koneko joined #git
13:23 star_prone joined #git
13:23 Gunni joined #git
13:24 Gunni hi, does anyone here have expirience with gitweb with apache using the directoryindex method?
13:24 sdothum joined #git
13:24 osse mikecmpbll: http://imgur.com/a/iAFL7
13:25 osse mikecmpbll: if you merge into dev to create B then the commit A will appear in the log
13:25 _ikke_ why did you reupload that to imgur?
13:25 osse is that what you mean? because that's just how Git works
13:25 osse _ikke_: i desecrated it first
13:26 mikecmpbll osse : commit a and the merge commit, yes. i don't know, maybe it's just me but it seems odd on the merge request to see 2 commits, bugfix one & a merge commit from a previous bugfix.
13:26 mikecmpbll commit A is already in the dev branch because it was merged from the previous hotfix branch
13:27 osse Don't know what to say. I can understand it seems odd, but based on how Git works it simply isn't
13:27 osse the commit labeled A in my shitty paintjob not in dev, no
13:27 osse unless you merge that bugfix
13:28 mikecmpbll oh okay, i thought in the diagram that represents the same commit as the first red blob on the hotfixes!
13:28 osse each blob in that diagram is a separate commit
13:28 mikecmpbll ah okay so it's the merge commit
13:28 osse at least if you consistently use --no-ff
13:29 mikecmpbll fair enough, that helps put things in order in my mind at least. i think i'll do a fast-forward merge back into master from now on.
13:29 JeroenT joined #git
13:30 osse mikecmpbll: are you new to git? do you work in a team?
13:31 mekeor joined #git
13:31 mikecmpbll osse : heh, you might think so mightn't you, but nope, i'm not. and yes, a team.
13:31 osse ok
13:31 NeverDie_ joined #git
13:31 osse i didn't mean it that way.  I was just going to say that if you were and weren't, respectively, that branching model is a bit over the top
13:32 mikecmpbll since the team grew we moved to using a strategy with feature branches and all that; and since then i've never quite nailed down a flow that works smoothly.
13:33 mekeor whut? after running 'git clean -fg', git responded amongst other with "Removing ~/". wtf
13:33 Oatmeal joined #git
13:33 osse we average about 4 people. master is our "main" branch. we sometimes commit directly to it. work in a separate branch when it makes sense (longer, bigger features) and merge them in when they're done. Once we have something worth slapping a version label on, we do so. That's about it. What each dev does locally (rebase his work, merge from master) is up to him, but we mostly rebase
13:34 osse And that's about it. WOrks just fine
13:35 osse mekeor: what is -g ?
13:35 mikecmpbll osse : how do you handle emergency bug fixes or is that not applicable in your case? that's the only reason we decided to keep a master branch which reflects production, and use dev for features.
13:36 mikecmpbll then we can commit bugfixes on master and deploy right away, without deploying a load of stuff we don't want to yet.
13:36 osse mikecmpbll: not applicable. it's a mostly standard desktop application. no services or hosting or live DBs involved
13:36 mekeor osse: oops, i mant -d, typo
13:36 mikecmpbll i see 👍🏼
13:36 timvisher joined #git
13:36 Pulp joined #git
13:37 osse at least none that are on the internet :P
13:37 mikecmpbll :)
13:38 msonntag joined #git
13:40 NeverDie joined #git
13:42 Vgr joined #git
13:42 theoceaniscool joined #git
13:44 timini joined #git
13:44 timini_ joined #git
13:47 dendazen joined #git
13:47 Derperperd joined #git
13:48 prg318 joined #git
13:50 denete_ joined #git
13:50 lindenle joined #git
13:51 jameser joined #git
13:51 Khisanth joined #git
13:55 PioneerAxon joined #git
13:56 JeroenT_ joined #git
13:59 brent__ joined #git
13:59 NeverDie_ joined #git
14:00 ThomasLocke joined #git
14:01 anonymuse joined #git
14:02 inflames joined #git
14:03 mablae joined #git
14:06 jeffreylevesque joined #git
14:07 sword```` joined #git
14:10 cdg joined #git
14:11 timvisher joined #git
14:11 kpease joined #git
14:12 mischat joined #git
14:13 Gsham joined #git
14:13 star_prone joined #git
14:15 porsche944 joined #git
14:15 sword```` joined #git
14:15 rnsanchez joined #git
14:15 porsche944 I made a commit then did "git reset HEAD~2 --hard"
14:15 porsche944 how do I find the original commit
14:15 porsche944 if I do a git pull
14:16 porsche944 and git log
14:16 porsche944 I don't see it
14:16 osse porsche944: git reset --hard HEAD@{2}
14:16 osse do 'git show HEAD@{2}' to confirm it's the corret one
14:16 porsche944 what does that do
14:16 osse !reflog
14:16 gitinfo The git reflog (`git log -g`) records the SHAs of your HEADs for 2+ weeks. `git checkout -b myrestore OLDSHA` and `git reset --hard OLDSHA` will relink to that state via a new and current branch respectively, see http://sethrobertson.github.com/GitFixUm/ for full details.  WARNING: reset --hard will trash any uncommitted changes!  Visualize with: gitk --all --date-order `git log -g --pretty=%H`
14:16 porsche944 ah good idea
14:16 lektrik joined #git
14:17 mischat joined #git
14:17 porsche944 ok so I see it now
14:18 porsche944 so you're saying HEAD@{2} pulls my head forward 2 commits?
14:18 osse fuck
14:18 osse no
14:18 osse i meant HEAD@{1}
14:18 osse it fines what the previous value of HEAD was
14:18 osse *finds
14:18 overlord_tm joined #git
14:18 porsche944 ok
14:19 m1m3-50 joined #git
14:19 miczac joined #git
14:19 porsche944 git show HEAD@{1}
14:20 porsche944 fatal: ambiguous argument 'HEAD@': unknown revision or path not in the working tree.
14:20 ShekharReddy joined #git
14:20 osse your shell is weird.quote it.  git show 'HEAD@{1}'
14:21 porsche944 ah
14:21 porsche944 yeah I'm using powershell
14:23 sword```` joined #git
14:23 GodGinrai joined #git
14:24 roko joined #git
14:24 jimi_ joined #git
14:24 GodGinrai joined #git
14:24 porsche944 weird
14:24 porsche944 when I did git show HEAD@{2}
14:24 porsche944 it was the right one
14:25 osse could very well be
14:25 porsche944 but when I did git reset --hard HEAD@{2} no luck
14:25 roko When I merge a branch with my master, why is the branch not deleted? What is the reason for keeping it?
14:25 osse ahh, probably because of the pull
14:25 porsche944 now I'm at some other commit
14:25 osse porsche944: 2 because 3 whenever HEAD changes, etc.
14:25 osse whatever. use 'git reflog HEAD' to find the right one
14:25 porsche944 yep I have the commit hash
14:25 osse or better yet 'git reflog branchname'
14:26 porsche944 so I have the commit hash
14:26 porsche944 how do I reset to there
14:26 GodGinrai roko: Because git doesn't presume things for you
14:26 clmsy joined #git
14:27 osse porsche944: git reset --hard abcd1234
14:27 Oatmeal joined #git
14:28 annoymouse joined #git
14:28 porsche944 that was it thanks
14:28 m4sk1n joined #git
14:28 porsche944 roko: periodically merging a branch into another is a valid workflow
14:28 porsche944 some people prefer to do that rather than one big merge at the end
14:28 star_prone joined #git
14:28 porsche944 there are other reasons as well
14:29 porsche944 maybe someone needs part of what you've done now
14:29 sword```` joined #git
14:29 roko porsche944: So it is a valid workflow that even after merging I can continue with the same branch rather than creating a new one
14:29 porsche944 right
14:30 todd_dsm joined #git
14:30 Bizkit joined #git
14:31 _geoff_ joined #git
14:32 HardlySeen joined #git
14:33 Ryanar joined #git
14:34 Snugglebash joined #git
14:34 ski7777 !subprojects
14:34 gitinfo 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.
14:35 boombatower joined #git
14:35 RxMcDonald leave
14:35 HardlySeen joined #git
14:35 RxMcDonald left #git
14:36 juslintek joined #git
14:36 juslintek Hey, I'm having problem, have two bitbucket accounts
14:36 juslintek I have added two hosts with different files to be used for private keys
14:37 juslintek but when I do with git a clone I still get fatal: Could not read from remote repository.
14:37 juslintek key is added to target bitbucket account
14:37 dendazen joined #git
14:37 juslintek Maybe I need somehow to relogin to different git
14:37 nanga joined #git
14:38 canuk joined #git
14:38 tobiasvl ssh keys?
14:38 juslintek like here: http://stackoverflow.com/questions/3225862/multiple-github-accounts-ssh-config
14:38 HcsOmot joined #git
14:41 nettoweb joined #git
14:42 HcsOmot left #git
14:44 BlessJah juslintek: try cloning with <username>@bitbucket.org instead of git@bitbucket.org
14:44 crose joined #git
14:44 juslintek <BlessJah>, works
14:45 NeverDie joined #git
14:45 juslintek How to make dynamic username on npm?
14:45 HcsOmot joined #git
14:45 juslintek git+ssh://<username>@bitbucket.org
14:45 BlessJah juslintek: what are you trying to do?
14:45 BlessJah because !xy
14:45 gitinfo Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...
14:47 HcsOmot left #git
14:47 juslintek BlessJah, I'm trying to add private github repo, to node.js package
14:47 timvisher joined #git
14:47 juslintek so on intall it would automatically check existing ssh keys per hostname
14:47 juslintek and install it
14:48 jbitdrop joined #git
14:48 Derperperd joined #git
14:48 BlessJah why two accounts then?
14:48 ocbtec joined #git
14:48 belak BlessJah: I'm glad some people noticed that feature. :)
14:48 netj joined #git
14:48 juslintek one is for me as freelnacer
14:48 juslintek other is for company worker
14:49 BlessJah belak: are you bb developer?
14:50 fahadash joined #git
14:50 BlessJah juslintek: and dynamic username in npm?
14:50 belak Yep. One of the two main devs who have worked on the ssh server.
14:50 anonymuse joined #git
14:50 HcsOmot joined #git
14:52 BlessJah belak: very nice feature to have, I guess you have own ssh implementation?
14:52 BlessJah or did you hack openssh?
14:52 timini joined #git
14:53 cyan__ joined #git
14:53 belak It was originally an openssh patch. We redid it past year as a custom server in go, using the semi-official ssh library
14:54 belak *last year
14:55 BlessJah I've heard once that on-premises bitbucked and hosted one are two different bitbuckets, is that true?
14:56 belak Yes. On premise is H
14:56 belak Whoops
14:56 belak On premise is almost entirely java
14:56 belak BB cloud is mostly Python
14:57 ski7777 I think gitinfo cant help me
14:57 ski7777 I have two repos and I want a subfolder of repo 1 to be a folder of repo 2. Is this possible?
14:58 _ikke_ ski7777: Did you look at any of the options that gitinfo returned?
14:58 sword```` joined #git
14:58 GodGinrai 09:56 < belak> Yes. On premise is H <-- That's... uh... <.<
14:58 GodGinrai :P
14:58 ssitani joined #git
14:58 ski7777 _ikke_: yes, but they didnt seemed to be good for that
14:58 BlessJah juslintek: in case I loose connection (I'm on train), with questions Iwant to figure if deploy keys would solve the issue for you
14:58 belak GodGinrai: I love typing on phones >_<
14:59 GodGinrai lol
14:59 GodGinrai That's why I have a phone w/ a real keyboard
14:59 * GodGinrai <3's his Pre3
14:59 * BlessJah nokia E5 :D
14:59 _ikke_ ski7777: Can you explain a bit why those options don't fit? (I cannot guess)
14:59 ski7777 _ikke_: any ideas?
14:59 juslintek BlessJah, deploy keys, you mean public ssh-keygen generated key?
15:00 belak I'm gonna get a little more sleep, but feel free to send me a message if you've got any more questions and I'll respond when I get into work
15:00 ski7777 _ikke_: They seemed to bind a entire repo as a subfolder and I need a sufolder of repo 1 as subfolder of repo 2
15:00 mizu_no_oto joined #git
15:01 BlessJah juslintek: kinda, keys you add to repository to allow read-only access
15:01 BlessJah i.e. for deployments and tests
15:02 _ikke_ ski7777: iirc, there is no existing solution to do that
15:02 Dougie187 joined #git
15:02 ski7777 _ikke_: :(
15:02 roelmonn_ joined #git
15:02 * ski7777 is sad
15:03 mischat joined #git
15:04 Es0teric joined #git
15:05 juslintek BlessJah, so guess these are the ssh-keygen generated pub keys. :-D
15:05 Furai joined #git
15:05 ski7777 I have two repos and I want a subfolder of repo 1 to be a folder of repo 2. Is this possible?
15:05 juslintek Well when I do ssh-add <directory_of_the_public_key>
15:05 ski7777 Anybody else an idea?
15:05 juslintek and enter password
15:05 juslintek it allows me to clone
15:06 juslintek via ssh
15:06 juslintek but when I use yarn or npm it doesn't
15:06 koneko joined #git
15:08 sarri joined #git
15:08 sarri joined #git
15:09 BlessJah it could be done with subtree-like script but it'd require some git internals knoowledge
15:09 BlessJah juslintek: are those keys password-protected?
15:10 juslintek BlessJah, yes
15:10 dsfdsfdsf joined #git
15:11 anjen joined #git
15:11 Shashikant86 joined #git
15:11 * BlessJah doesn't know much about password protected keys, authentication agent and how does npm interact with them
15:12 miczac joined #git
15:13 BlessJah ski7777: could you explain why you want one repos subdir to be others dir? (and what's the difference between subdir and dir?)
15:14 LeBlaaanc joined #git
15:14 robotroll joined #git
15:15 netj joined #git
15:15 ssitani joined #git
15:15 timvisher joined #git
15:16 sword```` joined #git
15:16 ski7777 BlessJah: I am developing a python module, which I want to integrate in another project. Moving the files in the root of the module repo is not possible because of examples, which should not be in the main project. There are no differences between dir and subdir
15:18 Derperperd joined #git
15:19 nickabbey joined #git
15:21 lvh joined #git
15:22 Shashikant86 joined #git
15:22 JeroenT joined #git
15:22 sword```` joined #git
15:24 ResidentBiscuit joined #git
15:24 ayogi joined #git
15:24 mischat_ joined #git
15:24 EricBlade joined #git
15:24 MadLamb joined #git
15:24 oskarkv joined #git
15:25 BlessJah ski7777: is it viable to use setuptools or pip to install the module?
15:25 ski7777 On this system no
15:25 ski7777 But Maybe I could make my system downloading the lib in other ways
15:26 RxMcDonald joined #git
15:26 BlessJah will you continue to develop the other module in the repo in which it is now?
15:27 ski7777 yes
15:27 RxMcDonald how do you remove cached files matching a patern? I.e. all *.js files?
15:28 BlessJah do you want to import once, continously (i.e. always master branch) or only releases?
15:28 BSAlb joined #git
15:29 BlessJah RxMcDonald: find . -name '*.js' -exec git reset -- {} +
15:29 MadLamb I know this is not a Github support channel, but maybe someone can help me to figure it out. I have a pull request on github and I was checking the remote references for it and refs/pull/18/merge points to a different commit than the branch the pull request is open for. Do someone know if its normal to have a different reference there and why this might be happening?
15:29 jost_ joined #git
15:29 hasc joined #git
15:30 cyan__ joined #git
15:30 BlessJah MadLamb: just an idea what might have happened, is the different commit a merge?
15:30 sword```` joined #git
15:31 RxMcDonald BlessJah: didn't work
15:32 grilled-cheese joined #git
15:32 BlessJah it removes thosse from cache, now you may remove those files
15:32 MadLamb BlessJah, in the pull request reference yes, but in the branch reference no. The branch was rebased.
15:32 grilled-cheese joined #git
15:33 hahuang61 joined #git
15:33 RxMcDonald shouldn't it be find . -name '*.js' -exec git rm --cached -- {} +
15:33 MadLamb BlessJah, it seems that somehow github didn't update the pull request reference when I force pushed the branch rebased.
15:33 BlessJah MadLamb: I'm out of ideas
15:33 MadLamb BlessJah, the branch looks fine, the github page that shows the pull request looks fine, but the reference seems to be wrong
15:34 MadLamb BlessJah, since travis uses the PR reference to checkout and test the code, the wrong commit is being tested
15:34 mns joined #git
15:34 cdown joined #git
15:35 BlessJah RxMcDonald: if you want to create a commit that removes all *.js files, yes
15:35 grilled-cheese joined #git
15:37 Tobbi joined #git
15:38 pity joined #git
15:39 grilled-cheese joined #git
15:39 Tsutsukakushi joined #git
15:39 jaafar joined #git
15:40 leeN joined #git
15:42 nidr0x joined #git
15:42 MadLamb I know this is not a Github support channel, but maybe someone can help me to figure it out. I have a pull request on github and I was checking the remote references for it and refs/pull/18/merge points to a different commit than the branch the pull request is open for. Do someone know if its normal to have a different reference there and why this might be happening?
15:45 mgedmin MadLamb: githubs PR refs point to a synthetic merge commit that merges your branch into whatever base branch that PR was aimed at
15:46 thebope joined #git
15:47 grilled-cheese joined #git
15:48 timvisher joined #git
15:51 nic__ left #git
15:51 BlessJah hh
15:52 star_prone joined #git
15:53 MadLamb mgedmin, so if I merge my branch to the base branch I should expect to not have the same commit reference, but the same result, right?
15:53 masuberu joined #git
15:53 m1m3-50 joined #git
15:53 anonymuse joined #git
15:53 anjen joined #git
15:53 mgedmin essentially, yes
15:54 mgedmin (I think you might even get the same commit hash if you explicitly specify a commit time for the merge commit, and make the message identical)
15:55 Endarked joined #git
15:56 Es0teric joined #git
15:57 ash_workz joined #git
15:57 Snugglebash joined #git
15:57 MadLamb mgedmin, is weird because my build is passing on the base branch, and when I run the tests on the branch the results are different from the results obtained by this gitgub PR reference
15:59 nickabbey joined #git
15:59 f0rks joined #git
15:59 mgedmin well, that's the reason why GH does this: so you can test merges, not just branch tips
16:01 finalbeta joined #git
16:03 m1m3-50 joined #git
16:03 rivarun joined #git
16:03 Guest95139 joined #git
16:05 MadLamb mgedmin, yes, I understand but the merge seems to point to a commit which doesn't make sense
16:06 MadLamb mgedmin, the merge commit produces an error that was solved in the branch that is being merged
16:06 hobodave joined #git
16:06 cdg_ joined #git
16:07 MadLamb mgedmin, its like it didn't update the merge when I force pushed my rebased branch
16:07 MadLamb mgedmin, it is keeping the reference of a merge that happened before
16:07 fragMental joined #git
16:07 mgedmin I've a cron script that periodically does git pull on a bunch of git clone --mirror'ed github repos
16:08 mgedmin I noticed that it force-updates PR refs with a lag of several days sometimes
16:08 dreiss joined #git
16:08 mgedmin or maybe you need a git fetch the right refspec
16:09 sarri joined #git
16:09 sarri joined #git
16:10 nd_ joined #git
16:11 sweatsuit joined #git
16:13 zivester joined #git
16:13 Ranmaru joined #git
16:14 Snugglebash joined #git
16:14 brent__ joined #git
16:15 Ryanar joined #git
16:16 Darren_ joined #git
16:21 goTAN joined #git
16:21 eduardas_ joined #git
16:23 eduardas_m joined #git
16:23 PettanShoutaKun joined #git
16:24 timvisher joined #git
16:24 MadLamb mgedmin, yes, but who fetches it is travis, and they always fetch refs/pull/<pr_number>/merge
16:24 MadLamb mgedmin, and after fetching then checkout FETCH_HEAD
16:24 mgedmin oh, you're seeing mismerged results on travis?  interesting
16:24 sword```` joined #git
16:25 MadLamb mgedmin, yes, but the travis behavior seems fine, what seems to be wrong is the remote merge reference from GH
16:25 mgedmin you probably need to talk to github/travis people about this ...
16:25 MadLamb mgedmin, do you know if refs/pull/<pr_number>/head should reference the same hash as the branch?
16:26 nickabbey joined #git
16:26 MadLamb mgedmin, /head, not /merge
16:26 mgedmin wait, there's more than one ref for each pull?  disregard anything I said before, I clearly know less than you do :)
16:26 MadLamb mgedmin, there are two refs for each pr, /merge and /head
16:27 MadLamb mgedmin, travis uses /merge. I'm just curious if /head should point to the same hash as the branch, because then it would explain a lot (since they are different)
16:28 mgedmin that would make sense
16:29 nowhere_man joined #git
16:29 jccn joined #git
16:30 rkazak joined #git
16:30 sword```` joined #git
16:31 govg joined #git
16:31 pbrewczynski joined #git
16:33 isysd joined #git
16:34 snowkidind joined #git
16:35 kaldoran joined #git
16:37 jmpp joined #git
16:38 Rodya_ joined #git
16:38 fstd joined #git
16:39 roelmonnens joined #git
16:39 shgysk8zer0 joined #git
16:39 CalimeroTeknik joined #git
16:41 NeverDie joined #git
16:42 CalimeroTeknik hello, I have some diff (no newline at end of file) between two branches, but that much isn't enough for git to accept merging one into the other. how can I remove the cumbersomeness?
16:44 m1m3-50 joined #git
16:46 e14 joined #git
16:46 ResidentBiscuit wat
16:46 RxMcDonald left #git
16:47 finalbeta joined #git
16:49 svm_invictvs joined #git
16:50 EliasH joined #git
16:51 tang^ joined #git
16:52 altendky joined #git
16:52 gajus joined #git
16:52 emg joined #git
16:53 jmpp joined #git
16:53 jmpp joined #git
16:55 CheckDavid joined #git
16:55 mehola joined #git
16:58 Learath2 joined #git
16:59 gugah joined #git
16:59 LionsMane joined #git
17:01 finalbeta joined #git
17:02 durham joined #git
17:04 imachuchu joined #git
17:04 hasc_ joined #git
17:05 cheater left #git
17:06 rnsanchez_ joined #git
17:07 Rodya_ joined #git
17:07 timvisher joined #git
17:15 thiago joined #git
17:16 anddam I rebased and ended up with a set of duplicate commits, how do I revert this given that I pushed?
17:17 dminuoso !rewrite
17:17 gitinfo Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum
17:17 chardan joined #git
17:17 anddam the push target is my own github fork so i have no problem rewriting history or even deleting it
17:17 anddam nobody has pulled it
17:17 dminuoso anddam: I suppose you could just use the reflog
17:17 moritz reset to an older version, force-push
17:17 dminuoso And then push force once you have readjusted your HEAD
17:23 nicoulaj joined #git
17:23 stickperson joined #git
17:23 mischat joined #git
17:24 whitby adambrenecki: http://stackoverflow.com/a/30927009/1267663
17:25 whitby If I may share my own answer
17:25 whitby anddam: ^
17:25 whitby (I mistyped that nick)
17:26 e14 joined #git
17:27 jmpp joined #git
17:27 jmpp joined #git
17:27 shgysk8zer0 joined #git
17:28 miczac joined #git
17:29 Derperperd joined #git
17:29 duderonomy joined #git
17:30 einnjo joined #git
17:30 lindenle joined #git
17:32 MadLamb mgedmin, FYI, I duplicated the branch and created another PR (you can't create two PR's for same branch). The /head reference in this new PR has the same has as the duplicated branch.
17:32 lindenle joined #git
17:32 MadLamb mgedmin, that kinda says that the merge on the old PR happened from the wrong HEAD
17:32 MadLamb mgedmin, :)
17:34 t0by joined #git
17:34 t0by joined #git
17:34 hahuang61 joined #git
17:35 MadLamb mgedmin, same hash as the duplicated branch*
17:36 anddam moritz: how do I quickly found the first non-duplicated changeset that I need to reset to?
17:36 anddam whitby: got it, reading now
17:36 pandeiro joined #git
17:37 anddam dminuoso: reading the man page, but I never heard it so unlikely I'll be able to use it without guidance
17:37 jozwior joined #git
17:37 dminuoso anddam: It's very very simple. Type "git reflog"
17:38 dminuoso anddam: Every time git does something, it stores the entire state inside the reflog, very much like a commit.
17:38 mizu_no_oto joined #git
17:38 dminuoso anddam: And with that reflog you can undo any crazy damage you might have done.
17:39 dminuoso anddam: And as it happens, you can just simply use "git reset" with those hashes that you can see in the reflog
17:39 dminuoso (Usually you will want to git reset --hard)
17:40 ash_workz how do I show all remote branches again? (ie: pushed from another machine, but you want to see it in a list)
17:40 anddam oh that's handy
17:40 dminuoso anddam: The reflog is why things such as rebase can be used lightly, because you cannot damage anything.
17:40 anddam dminuoso: that's the show subcommand, I see
17:40 inflames joined #git
17:41 anddam dminuoso: well I beg to differ
17:41 anddam but nice tool indeed
17:41 ash_workz like, git branch wont show it
17:41 doener joined #git
17:41 dminuoso !reflog
17:41 gitinfo The git reflog (`git log -g`) records the SHAs of your HEADs for 2+ weeks. `git checkout -b myrestore OLDSHA` and `git reset --hard OLDSHA` will relink to that state via a new and current branch respectively, see http://sethrobertson.github.com/GitFixUm/ for full details.  WARNING: reset --hard will trash any uncommitted changes!  Visualize with: gitk --all --date-order `git log -g --pretty=%H`
17:41 manuelschneid3r joined #git
17:41 ash_workz I thought it'd be git branch -r, which indeed does show many branches, but I don't see what I'm looking for in that list
17:42 dminuoso ash_workz: git ls-remote ?
17:42 ash_workz that just shows 'origin'
17:42 dminuoso ash_workz: git ls-remote --heads origin
17:42 PaulCapestany joined #git
17:43 anddam dminuoso: I'm not sure what the proper hash is https://gist.github.com/anddam/c6a5091c6dd7f977d9bff53cefe78b2b
17:43 dminuoso anddam: The first column
17:43 anddam this is the about 740 commits I duplicated
17:43 dminuoso anddam: You can also use HEAD@{744}
17:43 ash_workz okay, I don't get it... I switched to master locally, created a new branch, made changes, committed, then pushed, and I don't see the branch remote or locally with `git ls-remote --heads origin`
17:43 anddam I mean how to pick the right row
17:43 dminuoso anddam: Either will work, but SHAs are likely better because you can't do things wrong with them.
17:44 timvisher joined #git
17:44 dminuoso anddam: This is where I leave someone else to answer the question, because I don't know how you can easily retrieve the correct entry.
17:44 dminuoso (I have never had the issue where the entry wasnt in the recent reflog)
17:44 anddam I think the one I screwed up with is     fedca06 HEAD@{749}
17:44 dminuoso anddam: Ohh, nevermind.
17:46 thierryp joined #git
17:46 ash_workz I had to specify when I pushed `git push origin <branch>`
17:46 raspado joined #git
17:46 theoceaniscool joined #git
17:47 ash_workz but now it shows up in `git ls-remote` but when I go to check it out it says :error: pathspec '<test>' did not match any file(s) known to git.
17:47 ash_workz s/test/branch/
17:47 dminuoso anddam: Which line started the "bad business" ?
17:47 dminuoso anddam: Pick the one below it, and git reset --hard (or git checkout if you want to inspect it first) to it
17:47 dreiss joined #git
17:50 dsdeiz joined #git
17:51 porsche944 joined #git
17:52 MineCoins joined #git
17:54 anddam dminuoso: point is I'm not sure, since I didn't understand why the issue happened
17:54 anddam dminuoso: I could as well wipe my fork, clone again and go with that, but I'd like to try and use the tool to fix this
17:54 dminuoso anddam: It's better to understand what happened, otherwise you will be poking around hoping to fix it
17:54 dminuoso And by doing that you might actually do more damage :P
17:56 ihradek joined #git
17:56 svm_invictvs joined #git
17:56 mischat joined #git
17:57 mischat joined #git
17:57 systemovich joined #git
17:57 mischat joined #git
17:58 overlord_tm joined #git
17:58 ferticidio joined #git
17:58 dreiss joined #git
18:00 PaulCapestany joined #git
18:00 nickabbey joined #git
18:03 hh2010 joined #git
18:05 sea-gull joined #git
18:05 e14 joined #git
18:13 aw1 joined #git
18:13 timvisher joined #git
18:16 b1tchcakes joined #git
18:16 publio joined #git
18:18 hahuang61 joined #git
18:18 thiago joined #git
18:18 m0viefreak joined #git
18:19 jccn joined #git
18:19 kaldoran joined #git
18:20 MineCoins joined #git
18:22 karstensrage can you do pull requests with gitolite?
18:23 thallada joined #git
18:23 bremner there is some pull request extension iirc. I dunno what it does.
18:23 thiago do you mean "send someone an email to ask them to pull"?
18:23 thiago or do you mean have a web review tool for the code?
18:24 karstensrage i mean more like there is someone with a repo locally, and someone else that can commit to gitolite
18:25 karstensrage so can the pull request go to gitolite rather than have the someone else with push rights  pull from the repo with no push rights
18:25 thiago yeah, you can send them an email to ask them to merge
18:25 irco joined #git
18:25 karstensrage merge a branch on gitolite?
18:26 anddam dminuoso: I didn't get the mechanics that results in "duplicated" commits
18:26 nonconvergent joined #git
18:27 garethdaine joined #git
18:28 nonconvergent If I want to check and see if master has commits from release/vX.y.z (ie, was that release merged to master) is that git diff origin/master...origin/release/vX.y.z or the other way around?
18:28 nonconvergent or ..?
18:31 karstensrage http://gitolite.com/gitolite/user.html#pers
18:31 karstensrage that looks like what im after?
18:31 xissburg joined #git
18:31 ankit01ojha joined #git
18:32 ayogi joined #git
18:34 stickperson joined #git
18:37 jstimm joined #git
18:38 dsdeiz joined #git
18:38 dsdeiz joined #git
18:40 jozwior joined #git
18:41 blackwind_123 joined #git
18:41 hahuang65 joined #git
18:43 diogenese joined #git
18:44 timvisher joined #git
18:45 voldz1tic3n joined #git
18:46 _ikke_ karstensrage: gitolite is mainly an authorization layer
18:46 jnavila joined #git
18:48 Rodya_ joined #git
18:49 hahuang61 joined #git
18:50 kaldoran joined #git
18:50 hecliunyx joined #git
18:51 sim590 joined #git
18:51 Murii joined #git
18:54 jagob joined #git
18:55 sbeller joined #git
18:56 Es0teric joined #git
18:57 chardan joined #git
18:59 ikelso joined #git
19:02 roelmonnens joined #git
19:03 sim590 joined #git
19:06 Eugene karstensrage - you can grant branch-level permissions with a prefix, eg pull/. The user/group that you assign the permission to will already need to exist
19:07 duderonomy joined #git
19:07 cliluw joined #git
19:07 Eugene The idea of a "Pull Request" doesn't actually exist as a git object. `git-request-pull` will create an email for you, and Github's feature is just a fancy wrapper for their Issues system andd merging manually
19:08 Eugene The generic way to make a PR is to send an email with either a public URL(eg, git:// or https://).... or by attaching a git-format-patch
19:09 davimore_ joined #git
19:09 Eugene And yes, the personal branches feature in gitolite is probably closest to what you want. YMMV ;-)
19:10 timvisher joined #git
19:12 Derperpe1 joined #git
19:13 Anja joined #git
19:14 denisMone joined #git
19:15 nostrora joined #git
19:15 TFrag joined #git
19:15 danlovespipes_ joined #git
19:16 ramsub07 joined #git
19:16 divyanshu joined #git
19:19 Lac3rat3d joined #git
19:19 ihradek joined #git
19:19 Lac3rat3d anyone able to help me out? I've added some Git aliases, which work fine when referenced. However, since adding them whenever I open a new bash terminal I get the following output:
19:20 Lac3rat3d alias.alias1: command not found
19:20 Lac3rat3d alias.alias2: command not found
19:20 Lac3rat3d fatal: Not a git repository (or any of the parent directories): .git
19:20 Lac3rat3d alias1 and alias2 are not actually the names of the aliases, I've just genericized it here for anonymity :D
19:21 wrkrcoop joined #git
19:21 wrkrcoop how can i clone a repo from github using curl? curl —user “” repo_name doesn’t work
19:21 tobiasvl Lac3rat3d: whenever you open a bash terminal?? weird. do you have some bash prompt integration or something?
19:22 Lac3rat3d What do you mean by 'bash prompt integration'?
19:22 tobiasvl well why would it run the aliases every time you source .bashrc or similar files
19:22 tobiasvl if you don't have something that integrates bash and git
19:22 Lac3rat3d i'm not sure, that's why I'm asking
19:22 tobiasvl like git working dir status on the prompt or something
19:22 Lac3rat3d I use hub to wrap git
19:22 Lac3rat3d if that's what you mean
19:22 einnjo joined #git
19:22 tobiasvl no, I mean something like https://github.com/magicmonty/bash-git-prompt or similar
19:23 Rodya_ joined #git
19:23 Lac3rat3d no i don't have anything like that
19:24 _ikke_ wrkrcoop: I don't think you can clone with just curl
19:24 _ikke_ wrkrcoop: most you can do is download an archive
19:24 wrkrcoop _ikke_: ok i think that’s enough right?
19:25 _ikke_ wrkrcoop: an archive of a single revision
19:26 btree joined #git
19:27 tobiasvl ah a !crosspost
19:27 gitinfo Note: The above question was posted in both #git and #github
19:27 chardan joined #git
19:28 tyreld joined #git
19:29 stickperson joined #git
19:29 mizu_no_oto joined #git
19:30 druonysus__ joined #git
19:32 btree left #git
19:32 Sasazuka joined #git
19:32 e14 joined #git
19:33 canuk joined #git
19:33 wizardedit joined #git
19:35 Lac3rat3d tobiasvl: any other ideas for me?
19:38 tobiasvl Lac3rat3d: oh maybe you have git bash completion turned on, that would explain why it happens then
19:38 tobiasvl Lac3rat3d: but I don't know why bash would fail when loading it without knowing what the alias is
19:39 iampigman joined #git
19:41 sim590 joined #git
19:42 ubry joined #git
19:44 AlineGomes joined #git
19:44 fahadash joined #git
19:44 lvh joined #git
19:45 Lac3rat3d tobiasvl: i do have hub wrapping git, which provides tab completion
19:45 tobiasvl ah ok
19:46 Lac3rat3d it's not like the aliases don't work, it's just i get that strange output every time i open a terminal
19:46 tobiasvl well in that case that mystery is cleared up
19:46 tobiasvl well something doesn't work
19:46 tobiasvl hub can't define the alias
19:46 tobiasvl can you tab complete the aliases?
19:46 sim590 joined #git
19:47 Lac3rat3d yes, tab complete works and the alias works
19:47 Sasazuka when resolving conflict, the change on the bottom is "mine"  and the top is "theirs" ?
19:47 nickabbey joined #git
19:47 Lac3rat3d oh hmm
19:48 timvisher joined #git
19:48 Lac3rat3d it works with tab completion only when referencing it with git
19:48 Lac3rat3d ie, git alias1
19:48 Lac3rat3d tab complete works then
19:48 Lac3rat3d and alias works
19:48 Lac3rat3d but if i do hub alias1
19:48 Lac3rat3d tab completion doesn't work
19:48 Lac3rat3d but if i type it manually the alias still works
19:48 pandeiro joined #git
19:49 jccn joined #git
19:51 kaldoran joined #git
19:52 glebihan joined #git
19:52 Lac3rat3d anything i can do to fix it?
19:52 jnavila joined #git
19:52 EricBlade joined #git
19:53 _ikke_ You need to tell bash to use the completion for hub too
19:53 _ikke_ (I don't know how it works for bash)
19:54 Lac3rat3d afaik, the completion is provided by hub, not hte other way around
19:54 Lac3rat3d *tab completion
19:54 axisys joined #git
19:54 Lac3rat3d i could be wrong though
19:54 ihradek joined #git
19:55 Lac3rat3d it's strange though, i have 5 or 6 git aliases set up, and i only get the message about two of them
19:55 Lac3rat3d followed by the 'fatal' message
19:57 multi_io joined #git
19:58 kaldoran joined #git
19:58 druonysus__ joined #git
19:59 chardan joined #git
20:00 noecc left #git
20:01 invisbl_ joined #git
20:01 pouge joined #git
20:02 nostrora joined #git
20:02 invisbl_ joined #git
20:04 chipotle joined #git
20:07 Derperperd joined #git
20:12 hasc_ joined #git
20:15 Learath2 joined #git
20:15 satifant joined #git
20:16 miczac_ joined #git
20:18 Es0teric joined #git
20:19 timvisher joined #git
20:20 hecliunyx joined #git
20:20 spacelord_ joined #git
20:21 Lac3rat3d left #git
20:22 miczac_ joined #git
20:22 nickabbey joined #git
20:25 iampigman joined #git
20:27 svm_invictvs joined #git
20:30 Es0teric joined #git
20:30 kyan joined #git
20:30 krsr joined #git
20:33 PettanShouta-Kun joined #git
20:34 hecliunyx joined #git
20:38 theoceaniscool joined #git
20:38 mvensky joined #git
20:41 elsevero joined #git
20:44 Atm0spher1c joined #git
20:49 timvisher joined #git
20:52 voldz1tic3n joined #git
20:53 phox joined #git
20:53 nickabbey joined #git
20:54 phox hm, do I break anything by having two branches point to the same commit?  specifically I am wanting to create a new branch in origin but there is nothing new to commit so for now there would be two branches pointing to the same commit (which will then diverge and occasionally have a merge from one to the other).  any special caveats here?
20:57 tobiasvl phox: no, branches point to the same commit all the time
20:57 tobiasvl what you describe sounds perfectly normal
20:58 mehola joined #git
20:58 phox ok.  just something I have not encountered (or noticed encountering) so I wanted to make sure I wasn't setting myself up the bomb ;)  thanks!
20:58 phox oh also, kinda related, is there a way to "dry run" a push, i.e. see what it _would_ do in case I mis-brained which commits are where and how it would likely decide to fast-forward or not?
20:58 _ikke_ phox: when you create a new branch, it almost always points to the same commit as an existing branch
20:58 danlovespipes__ joined #git
20:59 phox _ikke_: makes sense, yeah, but typically I have only done that locally so I wasn't sure if I was going to stub my toe on something doing it remotely :)
21:01 hahuang61 joined #git
21:02 Tobbi joined #git
21:03 manuels joined #git
21:03 manuels_ joined #git
21:03 hobodave_ joined #git
21:04 ramsub07 joined #git
21:05 eliasp joined #git
21:08 Ekno joined #git
21:08 cdg joined #git
21:09 miczac_ joined #git
21:09 jimi_ joined #git
21:09 gigq joined #git
21:11 pity joined #git
21:16 timvisher joined #git
21:17 systemovich joined #git
21:17 Sasazuka_ joined #git
21:17 a-l-e joined #git
21:18 hashpuppy joined #git
21:18 e14 joined #git
21:18 systemovich joined #git
21:22 systemovich joined #git
21:22 jeffreylevesque joined #git
21:24 invisbl joined #git
21:24 NaN joined #git
21:24 miczac_ joined #git
21:25 systemovich joined #git
21:27 j416 phox: man git-push
21:27 gitinfo phox: the git-push manpage is available at http://jk.gs/git-push.html
21:27 j416 phox: look for dry-run ?
21:28 CheckDavid joined #git
21:29 phox j416: ah-hah, thanks
21:29 phox sorry, having kind of a dumb day, this has been a hell of a week
21:29 j416 don't be :)
21:32 suttonwilliamd joined #git
21:34 thiago joined #git
21:35 Kovalsky joined #git
21:36 pdgwien joined #git
21:37 phox cheers
21:37 * phox goes to try git push -n
21:38 phox oh that's why I didn't brain that there is a manpage for git-push, our dev environments are missing it :l  ah-HAH!
21:39 Hello71 wat
21:39 dendazen joined #git
21:40 dikiaap_ joined #git
21:40 phox alright, further is there a way to make it tell me what commits it's operating on when I'm pushing/whatever?  like, a verbose flag, except -v shows me other extra stuff and not that? :)
21:40 umpc joined #git
21:42 tobiasvl what do you want exactly
21:42 tobiasvl when pushing you want a list of all the commits it's pushing?
21:42 Bizkit joined #git
21:43 phox well moreso if I push say "origin foo:bar" I'd like to know what commit that's operating on, without having to look at 'git log'
21:44 peepsalot joined #git
21:44 matsaman joined #git
21:44 TFrag joined #git
21:44 mikecmpbll joined #git
21:44 pdgwien joined #git
21:45 Hello71 git push does a lot of stuff
21:45 Hello71 including but not limited to pushing *multiple commits*
21:45 phox right, but it could give me a (truncated) commit hash for each, that would be informative for me :)
21:46 phox doesn't mean it's always necessary, but as someone who is not yet always 100% sure what is going to happen due to e.g. server configuration (skip/allow/force fast-forward, etc.) I'd like to see what exactly is going to happen, emphasis on exactly :)
21:47 Derperperd joined #git
21:47 ManiacTwister joined #git
21:47 mloy joined #git
21:48 timvisher joined #git
21:49 askb joined #git
21:50 tobiasvl well you can never really know what server config will do
21:50 tobiasvl best not to think about that
21:50 tobiasvl a push is just you telling the server what you got
21:50 tobiasvl SHOW ME WHAT YOU GOT
21:50 tobiasvl and then the server decides what to do with it
21:51 tobiasvl anyway, if you push foo:bar you should ideally know what foo is already
21:52 Anja joined #git
21:53 Hello71 can't know for sure, race conditions
21:53 jstimm joined #git
21:54 thiago explain
21:54 thiago why wouldn't you know?
21:55 Hello71 I'm saying that you can't know for sure, because git does not offer a mechanism to take a lock on the remote side
21:55 Hello71 you can't know in advance.
21:55 thiago yeah, except that foo doesn't depend on the remote
21:55 thiago it's what the local side wants to push
21:55 thiago if you meant knowing bar, I'd understand
21:56 Hello71 knowing bar is only half the battle
21:56 Hello71 but right, what I meant was "you can't say for sure in advance what git push will do"
21:57 minikN joined #git
21:58 thiago right, we know what it'll send, but we can't know how the server will react
21:59 Hello71 oh, right, yes. for some reason I thought git push needed information from the remote
21:59 TomTomTo1 joined #git
22:00 glebihan joined #git
22:00 phox well, races aside I would hope dry-run checks what the server would do given the present state of things
22:01 glebihan_ joined #git
22:03 dendazen joined #git
22:04 Tobbi joined #git
22:06 thebope joined #git
22:08 XenophonF joined #git
22:08 LionsMane joined #git
22:08 PettanShoutaKun joined #git
22:11 PettanShoutaKun joined #git
22:12 glebihan_ joined #git
22:12 Mobutils joined #git
22:13 glebihan joined #git
22:14 timvisher joined #git
22:16 pre42 joined #git
22:17 chipotle joined #git
22:18 salamanderrake joined #git
22:19 phox thanks again for the input folks, we are using a (slightly wrong) modified git-flow and I was un-wronging it and wanted to make sure I didn't crap in everyone's Cheerios while doing so :)
22:20 matsaman s/cheerios/oats/,s/cow milk/plant milk/
22:21 phox you'd think it would be camel milk, all things considered.
22:21 matsaman heh
22:22 matsaman that reminds me, friend was reading from this book 'Eight Flavors' about flavors/spices
22:22 matsaman and there was one recipe mentioned for something to do with chili con carne that utilized milk "from a male goat"
22:22 matsaman which the author said "hopefully was a poor transcription"
22:22 phox lol
22:22 matsaman (paraphrasing)
22:23 stick-shift joined #git
22:23 phox I wouldn't put it past carnies.
22:23 phox "the travelling underbelly of the nation"
22:23 matsaman hahah
22:23 matsaman that is an excellent pun
22:24 matsaman I'll probably remember that for my entire life, gj
22:25 pre42 joined #git
22:26 pre42 joined #git
22:26 danlovespipes_ joined #git
22:27 PettanShoutaKun joined #git
22:27 jstimm joined #git
22:29 PettanShouta-Kun joined #git
22:30 lyxus joined #git
22:33 Anja joined #git
22:33 sea-gull joined #git
22:35 denete__ joined #git
22:37 lyxus Folks, what would be the best strategy  to keep a mirror a list of mirror up to date ?  I was thinking git clone mirror and then pull push ?
22:38 lubg joined #git
22:41 mizu_no_oto joined #git
22:43 d^sh joined #git
22:44 eIks joined #git
22:46 NaN joined #git
22:46 NaN why >> git difftool --tool=vimdiff --no-prompt
22:46 jimi_ joined #git
22:47 NaN shows diff like git diff and don't open vimdiff (working)?
22:47 ZeZu joined #git
22:49 timvisher joined #git
22:49 Remram joined #git
22:49 Remram joined #git
22:59 mehola joined #git
23:00 Dougie187 left #git
23:01 fmeerkoetter joined #git
23:05 mki joined #git
23:08 ramsub07 joined #git
23:08 kpease joined #git
23:09 BSab joined #git
23:09 ispnsn joined #git
23:11 dirtyroshi joined #git
23:12 pity joined #git
23:15 lacour joined #git
23:16 mentazoom joined #git
23:17 timvisher joined #git
23:17 salamanderrake joined #git
23:19 Noldorin joined #git
23:24 MadLamb joined #git
23:27 kyan joined #git
23:29 fahadash joined #git
23:29 memfrob joined #git
23:29 kadoban joined #git
23:30 memfrob Hi all, WOW lots of people! Awesome! I thought there'd be like 50 people in here and I'd have to wait a week to get an answer
23:31 Zarthus We're all about bakin git just right for you.
23:32 zivester joined #git
23:33 gugah joined #git
23:33 memfrob I'm trying to merge a branch from one git repo to another but not the entire branch and it's history, just the commits related to what the branch is about.
23:34 iampigman2 joined #git
23:34 matsaman memfrob: individually, or would as a single commit suffice?
23:36 memfrob either is fine. I tried looking at a branch where I only needed like the 5 past commits of the tree, but --depth 5 resulted in a fatal error, unable to merge unrelated histories, so I dropped in the --allow-unrelated-histories and it pulled in the entire branch history which had thousands of conflicts.
23:36 systemovich joined #git
23:38 SwiftMatt joined #git
23:38 memfrob I've tried countless combos of git fetch / git merge but the only way I could accomplish anything was to use git cherry-pick on all the commits I wanted. Problem with that is that it's ridiculously time consuming to fix all the commits of those patches because I need to fix all those conflicts PER PATCH versus being able to just fix all of them at once, plus it's a real pain to have to manually cherry-pick 50-100 commits.
23:39 memfrob If I could somehow use git pull but ignore the history of everything I don't need, that'd be awesome.
23:41 robattila256 joined #git
23:41 douglascorrea joined #git
23:42 adymitruk memfrob: if all you need is a single commit, then all you need to do is copy the files.
23:43 adymitruk you don't need to link the repos at all
23:44 memfrob What do you mean by copy the files? I'm not simply adding files to a git tree from a repo, it's C source, lines being added and removed from files already in my tree.
23:45 phox bah, now my upstream repo is perfect except for one extra commit that happened because something didn't get rebased right before pushing :x
23:46 memfrob phox, deleting commits easy. just use git revert or delete the commit and it's history entirely with a git push -f
23:47 ihradek joined #git
23:47 memfrob Nobody will ever know it was there unless they already cloned it with that commit in there.
23:47 adymitruk so you want patches
23:47 thib joined #git
23:47 adymitruk is there a common point where in the history they look identical?
23:48 timvisher joined #git
23:48 madewokherd joined #git
23:48 memfrob The patches I need are based off a commit where virtually everything has changed, that's why I'm trying to avoid those.
23:48 rivarun joined #git
23:49 memfrob Now before the commits that I need from the branch, there was a point in time where the trees were identical yes.
23:49 adymitruk but you can limit to the file
23:50 adymitruk or are there changes in that file too?
23:50 memfrob Can I post an example here?
23:50 adymitruk if you cherry-pick --no-commit, you can reset the other changes so you only get the ones you want
23:50 adymitruk sure. is it a public repo?
23:50 SwiftMatt joined #git
23:51 memfrob yes it's all public. I'm working backporting the KSPP (kernel self protection project) patches but I'm not the best at git, I never did this exact thing in git before so I'm new to this very thing.
23:52 garethdaine joined #git
23:53 adymitruk ok sure. post the points in both histories and I'll tell you what to do
23:54 memfrob https://git.kernel.org/cgit/linux/kernel/git/kees/linux.git/log/?h=kspp/gcc-plugin/randstruct Now I'm on kernel 4.4 right now, the changes between 4.9-rc2 and 4.4 are enormous. Now I only need the first 6 commits in that particular example, but whenever I try to do this, out of all the combos I tried, it keeps wanting to bump everything to kernel 4.9 and there's countless commits (as expected) and I don't want to update to 4.9 either. I want the
23:54 memfrob randstruct commits separate from the vanilla kernel source, message "Linux 4.9-rc2" from 2016-10-24
23:56 adymitruk so you want the first 6 commits off 4.9 for 4.4 but restricted to a set of files in a dir?
23:57 memfrob Well no I'd also like to have git try to apply the "use designated initializers for function pointer structs" commit as well, and let me solve the conflicts (which I'm sure they're a lot of)
23:58 memfrob And that commit touches 42 different files.
23:59 adymitruk at least the histories are common. I'll clone them now
23:59 Remram joined #git

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