IRCloggy #git 2019-02-17

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2019-02-17

← fletom86 left00:01
→ fletom86 joined00:02
← finalbeta left00:04
← Karyon left00:05
→ Karyon joined00:07
→ donofrio joined00:11
← watabou_ left00:11
← alyptik left00:19
← mescalinum left00:26
→ justanotheruser joined00:26
← dege left00:29
→ scrptktty joined00:36
← VectorX left00:37
→ VectorX joined00:41
→ hofmann3900 joined00:46
← deltam left00:48
→ yohnnyjoe joined00:53
→ hotpancakes joined00:58
← Regon left01:16
→ alyptik joined01:17
→ kneeki joined01:18
→ pionen joined01:21
← Ignacy left01:24
← egovirus left01:29
← pionen left01:40
→ chris349 joined01:44
chris349 How can I bypass a everything is up to date message and force a push anyways?01:44
← hotpancakes left01:48
→ Dirak joined01:51
← yyy left01:54
Hello71 !xy01:55
gitinfo You are asking about a solution, but we can help you a lot better if you could describe the actual problem you are trying to solve.01:55
bremner re-running a hook, I bet01:55
chris349: you could always make an empty commit01:55
← yyy- left01:59
Hello71 bremner: think you're too optimistic02:00
← dre_ left02:06
→ pionen joined02:06
← blackmesa left02:07
→ watabou_ joined02:07
← Karyon left02:14
← pionen left02:14
→ pionen joined02:15
→ finalbeta joined02:16
← Dirak left02:16
← peepsalot left02:25
→ hotpancakes joined02:26
→ Dirak joined02:28
← hotpancakes left02:32
← F0rTex left02:36
→ F0rTex joined02:37
← watabou_ left02:39
← m0viefreak left02:47
→ stitched_unicorn joined02:47
→ boombatower joined02:51
← Achylles left02:51
← mven left02:51
← aidalgol left03:00
← pionen left03:01
→ hotpancakes joined03:01
→ pionen joined03:03
← courrier left03:10
← lagothrix left03:13
→ lagothrix joined03:13
← jcbitter left03:19
→ Karyon joined03:19
← hotpancakes left03:20
← veegee_ left03:21
→ watabou_ joined03:21
→ mven joined03:21
← mven left03:21
→ aidalgol joined03:26
→ veegee joined03:27
← causasui left03:29
→ jcbitter joined03:34
→ core7 joined03:35
core7 I am trying to learn more about git branches - Why does git clone not get all the branches? Why have two commands "git branch + git checkout" to get a branch? Even then I get to detached head state03:37
→ wxg joined03:39
← watabou_ left03:40
← jnewt left03:41
rewt git clone does get all branches. those 2 commands do different things; use the one that's for what you want to do.03:42
core7 rewt: I am trying to wrap my head around git clone + git fetch - why do i need a git fetch --all?03:43
why not have that functionality in git clone?03:44
rewt man git fetch03:44
gitinfo the git-fetch manpage is available at https://gitirc.eu/git-fetch.html03:44
rewt --all = Fetch all remotes.03:44
if you have multiple remotes, --all will fetch from all of them03:44
core7 rewt: why not fetch all branches when git clone happens?03:44
rewt it does03:45
core7 let me try something - brb03:45
rewt remotes are remote repos ... you're talking about so many different things at once03:45
`git clone` clones a repo; makes a local copy of it03:46
`git fetch` updates the local repo; fetches changes that are in the remote that you don't yet have locally03:46
`git branch` creates new branches03:46
`git checkout` checks out (switches to) another branch03:46
core7 rewt: I've a repo with multiple branches (git branch -a) - I clone - then I do a git checkout origin/mybranchname -> In git branch -a it's listed as remotes/origin/mybranchname -- the checkout leads me to a detached head state03:47
what exactly is happening?03:47
rewt you're checking out a remote branch -- you cannot commit to that, so you're in a detached head state03:48
core7 whats the point of checkoug out a remote branch if i can't commit to it?03:48
rewt checkout mybranchname to create a local branch that tracks the remote branch03:48
there is no point in that03:48
other than you want it checked out but don't want to create a local tracking branch for it03:48
core7 so i have to do "git checkout mybranchname"?03:49
rewt yes03:49
core7 rewt: thanks - that seems to work - so now i am on a local branch- when I commit and push, it goes to remote as well? or not?03:50
rewt yes, when you commit, your local branch is updated, and then you can push that branch to update the remote03:50
core7 rewt: why does git allow to checkout "git checkout origin/mybranchname" - and is there a use for that?03:52
rewt you can checkout any ref, including remote branches, or even the hash of any commit in the repo03:53
core7 rewt: what is a good tutorial/video tutorial to understand these concepts?03:53
rewt sometimes you want to look at contents of the files at those points in history03:53
!tutorials03:54
gitinfo [!tutorial] A list of interactive git tutorials: http://try.github.com03:54
core7 Thanks03:55
← sgen left03:59
← Karyon left04:02
← rafasc left04:03
← inkbottle left04:04
→ mowcat joined04:12
→ dirtyroshi joined04:17
→ null1337 joined04:17
→ watabou_ joined04:20
← Cabanossi left04:20
→ rsrx joined04:26
→ Cabanossi joined04:27
← orbyt_ left04:28
← igemnace left04:30
→ _kk22kk_ joined04:41
→ Karyon joined04:41
← boombatower left04:45
metrixx → Guest2596204:49
→ metrixx_ joined04:49
← Guest25962 left04:53
← dviola left04:55
← pionen left04:56
→ khalella joined04:57
khalella i do not understand how to use initally pull git submodules. i add them to .gitmodules, and then git submodule update --init --recursive04:58
and git submodule foreach git pull origin master04:59
it will clone submodules that i already have pulled down, but will do nothing for new ones that i have added to .gitmodules04:59
so it needs to already exist05:01
→ pks_ joined05:01
← pks left05:03
pks_ → pks05:03
← Muzer left05:05
khalella pretty much none of the commands on this page will work: https://stackoverflow.com/questions/1030169/easy-way-to-pull-latest-of-all-git-submodules05:05
so i need to clone the repository manually first, as thought im committing it all into my repo05:05
→ igemnace joined05:05
khalella and then add it as a submodule, and submodule update --init ?05:05
i don't understand this at all. it doesn't seem right05:05
→ osa1 joined05:06
→ SkarmoutsosV joined05:07
khalella this means when i built it with jenkins or somethig its going to fail. i need to SSH into my jenkins box and manually clone easch of like 20 submodules into their little shiteating folders05:07
osa1 wwhen I have relative submodule paths and multiple remotes, how do I tell git to use remote1 when updating submodules instead of remote2 which is what git is using by default?05:07
→ hotpancakes joined05:08
→ Silenced joined05:08
← rsrx left05:10
→ Muzer joined05:11
← watabou_ left05:12
← hotpancakes left05:13
→ cd joined05:17
→ boombatower joined05:19
→ causasui joined05:26
← causasui left05:31
→ mven joined05:31
→ smokals joined05:38
← dermoth left05:39
→ keepLearning512 joined05:41
← xcm left05:43
→ xcm joined05:44
← donofrio left05:47
← stitched_unicorn left05:50
→ dermoth joined05:52
← Karyon left06:02
← dermoth left06:03
→ dermoth joined06:05
→ dege joined06:05
→ maryo joined06:12
← yohnnyjoe left06:14
→ hello joined06:15
← sjoshi left06:15
→ jaeckel joined06:18
← boombatower left06:19
← mven left06:23
→ sourcream joined06:28
→ time_ joined06:28
← hello left06:30
time_ → hello06:30
→ watabou_ joined06:31
→ iam730 joined06:31
← Envil left06:33
→ daegontaven joined06:33
→ hotpancakes joined06:33
→ Envil joined06:33
daegontaven i commited some code to branch i don't remember the name of that i deleted from origin..how do i search for the code?06:33
← Envil left06:33
daegontaven i'm sure there were no hard resets or anything06:34
← smokals left06:35
daegontaven nvm figured it out using git grep06:36
← hotpancakes left06:38
← craigo left06:40
→ TheSpectre joined06:42
← sourcream left06:43
← keepLearning512 left06:50
→ sjoshi joined06:58
← mowcat left07:02
← irqq left07:09
← thiago left07:11
→ causasui joined07:27
← jcbitter left07:31
← causasui left07:32
← igemnace left07:39
← _kk22kk_ left07:42
← Goplat left07:47
← quesker left07:53
→ tribrid joined07:55
→ mikecmpbll joined07:59
← Crocodillian left08:00
→ ComradeCrocodill joined08:00
← kadoban left08:01
← scrptktty left08:04
← rjsalts left08:07
→ rjsalts joined08:11
← mikecmpbll left08:13
→ mescalinum joined08:21
← watabou_ left08:26
→ kapil____ joined08:31
← mat001 left08:38
← mescalinum left08:43
→ mescalinum joined08:47
← dege left08:49
← sunri5e left08:54
→ sunri5e joined08:54
→ watabou_ joined08:54
→ hotpancakes joined09:01
← hotpancakes left09:06
→ sourcream joined09:11
← CarlFK left09:11
← random_yanek left09:14
→ blackmesa joined09:14
← im0nde left09:19
← SkarmoutsosV left09:21
→ YuGiOhJCJ joined09:21
← aidalgol left09:22
← pjboro left09:23
→ aidalgol joined09:23
→ causasui joined09:28
→ random_yanek joined09:28
← random_yanek left09:28
← veegee left09:28
→ veegee joined09:29
→ random_yanek joined09:30
→ czart joined09:32
← causasui left09:32
→ im0nde joined09:33
← hello left09:38
← watabou_ left09:42
← Silenced left09:44
← cebor left09:46
→ cebor joined09:46
← maryo left09:49
← uebera|| left09:51
→ uebera|| joined09:52
← Puffball left09:53
→ igemnace joined09:57
→ plexigras joined09:58
← netj left10:03
→ netj joined10:03
← \void left10:04
← zulutango left10:08
→ Regon joined10:10
j416 daegontaven: man git-reflog10:10
gitinfo daegontaven: the git-reflog manpage is available at https://gitirc.eu/git-reflog.html10:10
← Dirak left10:14
← Tawonga left10:16
→ time_ joined10:16
← dpyro left10:21
→ sozuba joined10:22
→ yyy- joined10:23
← catsup left10:29
→ catsup joined10:29
← catsup left10:30
← justanotheruser left10:30
→ catsup joined10:30
→ Fernando-Basso joined10:32
→ mikecmpbll joined10:40
→ nemesit|znc joined10:44
← cd left10:46
→ cd joined10:47
→ saidi joined10:48
→ manuelschneid3r joined10:49
← wolfshappen left10:52
→ m0viefreak joined10:53
→ wolfshappen_ joined10:53
telling_ → telling10:54
← manuelschneid3r left10:54
→ manuelschneid3r joined10:55
→ Phylock joined10:58
time_ → hello10:59
← Bobdude left11:00
← lankanmon left11:00
→ hotpancakes joined11:02
← hotpancakes left11:07
← brandonkal left11:09
← Inline left11:16
→ Silenced joined11:19
← kjartan left11:22
← R2robot left11:23
→ kjartan joined11:27
→ causasui joined11:29
← sourcream left11:32
← causasui left11:33
→ sourcream joined11:36
← pyfgcr left11:37
→ nevodka joined11:38
← nevodka left11:38
→ nevodka joined11:39
← nevodka left11:39
→ watabou_ joined11:39
→ nevodka joined11:39
← nevodka left11:39
→ Achylles joined11:39
→ stitched_unicorn joined11:39
→ pyfgcr joined11:39
→ nevodka joined11:40
← nevodka left11:40
→ nevodka joined11:40
← nevodka left11:40
← mona_x left11:44
→ lankanmon joined11:52
← blackmesa left11:52
→ Inline joined11:57
← sourcream left12:00
← manuelschneid3r left12:01
← watabou_ left12:12
← Mattiaslndstrm left12:20
→ watabou_ joined12:31
→ gloomy joined12:31
← sjoshi left12:32
→ sahasatvik joined12:34
→ craigo joined12:38
← hello left12:43
→ tkore1 joined12:44
→ snowkidimd joined12:48
← wolfshappen_ left12:51
← tkore1 left12:51
→ egovirus joined12:52
→ AtumT joined12:52
← Silenced left12:52
→ Silenced joined12:53
← saidi left12:55
← Eryn_1983_FL left12:58
→ SkarmoutsosV joined12:58
← RoriconKnight left13:00
← watabou_ left13:03
→ wolfshappen joined13:05
← gloomy left13:06
→ Eryn_1983_FL joined13:08
→ saidi joined13:08
← snowkidimd left13:11
→ yohnnyjoe joined13:12
→ courrier joined13:12
→ dege joined13:12
→ snowkidimd joined13:13
← dege left13:15
→ hotpancakes joined13:16
← zampano left13:17
→ sjoshi joined13:19
← xcm left13:20
→ xcm joined13:21
→ leeN joined13:22
→ inkbottle joined13:24
→ quipa joined13:26
→ causasui joined13:30
← hofmann3900 left13:30
j416 does anyone know of a nice cli git-blame view? I'd like something like git gui blame, but which runs in the terminal. I currently use vim-fugitive's :Gblame but it'd be nice to have something stand-alone.13:30
osse standalone() { vim -c Gblame "$1"; }13:33
\o/13:33
j416 ...13:33
...thanks13:33
osse But seriously, tig maybe?13:33
j416 oh.. tig has a blame view?13:34
osse I don't know. Wouldn't surprise me13:34
j416 wow, it does.13:34
tig blame <path>13:34
← causasui left13:34
j416 not optimal though13:35
osse Looks nice.13:35
j416 maybe I'm just not used to it13:35
osse No syntax highlighting of the code though13:36
Not a big deal I guess13:36
j416 it seems you can press , to blame parent, then < to go back, but if you have pressed enter to view the commit, you can't blame parent (,) _or_ go back (<) anymore without first exiting the diff view, it seems13:37
→ maryo joined13:37
j416 (which is done with q)13:37
← sahasatvik left13:37
→ totte joined13:50
← mostlybadfly left13:53
→ gloomy joined14:05
← gloomy left14:08
→ mona_x joined14:13
← hotpancakes left14:16
→ lucasb joined14:19
→ keepLearning512 joined14:20
← snowkidimd left14:23
→ CryptoDavid joined14:23
→ snowkidimd joined14:23
← keepLearning512 left14:25
→ hotpancakes joined14:27
→ maryo_ joined14:30
← foul_owl_ left14:34
← maryo left14:34
← hotpancakes left14:34
→ gloomy joined14:35
← giuscri left14:36
← mns left14:36
→ mat001 joined14:36
← YuGiOhJCJ left14:37
→ foul_owl_ joined14:38
→ mns joined14:39
osse j416: I think tig will be my main blame tool going forward. this is neat14:40
thanks to myself for the recommendation14:40
_ikke_ Yes, I use tig extensively as well14:41
I basically have a tmux screen with a terminal, tig and a live updated git status -sb14:41
s/terminal/shell14:41
osse _ikke_: how do you do the live thing?14:42
_ikke_ inotifywait14:42
osse ah14:43
do you rerun git status -sb whenever anything changes?14:43
_ikke_ https://gist.github.com/Ikke/64858046d15426aad19f14:44
yes14:44
Not sure if that's the latest version I use14:44
← Fernando-Basso left14:44
osse why capture the output?14:45
_ikke_ less flickering14:45
You capture the output before you reset the screen, then the rendering is much faster14:46
→ jungsubk joined14:48
osse makes sense14:49
_ikke_ kinda like double-buffer rendeirng14:49
not entirely, but comes close14:49
osse heh I fiddled a bit with inotifywait -m -r .14:52
my gitified prompt really makes it spew14:52
_ikke_ heh14:52
→ talx joined14:53
talx hello guys14:53
gitinfo talx: 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.14:53
talx repo sync14:53
error: in `sync`: [Errno 2] No such file or directory: '/home/nwertheizer/.repo/manifests/.git/HEAD'14:53
error: manifest missing or unreadable -- please run init14:53
I'm the newest person to git14:53
→ watabou_ joined14:54
talx I'm not sure what to do in google it says to delete a folder I don't have14:54
→ AtumT_ joined14:54
_ikke_ "manifest missing or unreadable -- please run init"14:55
It looks like you are running it somewhere that is not a repo-fied project14:55
← AtumT left14:55
talx what do you mea14:56
n14:56
_ikke_ nevermind14:56
j416 osse: :D14:56
talx repo init14:57
fatal: manifest url (-u) is required.14:57
→ AtumT joined14:57
_ikke_ I don't know repo that well, so cannot really help you14:57
← hexa- left14:57
← xcm left14:57
← jyc left14:57
→ xcm joined14:57
j416 _ikke_: you know that you can set status.short and status.branch and avoid -sb, right?14:58
→ jyc joined14:58
_ikke_ j416: I use an alias anyway14:58
j416: alias gst='git status -sb'14:58
j416 I see14:59
← watabou_ left14:59
← AtumT_ left14:59
talx hm14:59
j416 you could also tput cup 0 0 to remove all flickering (although it'd not redraw everything)14:59
talx I did repo init -u ssh://hsot:29418/software-repo15:00
_ikke_ right, I guess if the list became shorter, it would not remove later lines, right?15:00
talx I did repo init -u ssh://host:29418/software-repo15:00
it says :15:00
Permission denied (publickey).15:00
Permission denied (publickey).15:00
fatal: Could not read from remote repository.15:00
Please make sure you have the correct access rights15:00
and the repository exists.15:00
j416 yeah it only moves the cursor to 0,0 and whatever you don't draw will stay as before15:00
talx do anyone have any explanation for this ?15:00
j416 talx: you didn't set up your ssh key properly15:00
_ikke_ "Permission denied (publickey)"15:00
Celelibi Is there a way to provide my own merger when I perform a rebase?15:02
A tool that understands the semantics of the language could probably make a better job than any line-based one.15:02
_ikke_ You can provide your own merge driver15:03
osse Celelibi: yes15:03
_ikke_ check man gitattributes15:03
gitinfo the gitattributes manpage is available at https://gitirc.eu/gitattributes.html15:03
osse Celelibi: look up "Defining a custom merge driver"15:03
j416 are merge drives and diff drivers different things?15:04
Celelibi thanks, I'll look that up.15:04
j416 drivers*15:04
_ikke_ j416: I guess the same is the difference between mergetool and difftool15:04
s/is/as/15:04
j416 I mean, could you use a diff driver for a merge driver or vice versa?15:04
osse j416: the merge driver is expected to actually write the new merged version15:04
j416 ah.15:04
so it's diff driver plus extra?15:04
_ikke_ You need to be able to do an actual 3-way merge I guess15:05
Celelibi You can use more information when performing a merge.15:05
osse I mean technically a merge driver doesn't really have to diff anything15:05
j416 alright, so they're different animals15:05
osse It could pick one of the versions at random for example15:05
j416 ty15:05
osse But I would expect a merge driver to make sense of the differences15:05
talx j416: okay I'm doing it again15:06
← UTAN_dev left15:07
→ AtumT_ joined15:08
← yohnnyjoe left15:09
← AtumT left15:11
talx well I think I did set it right15:14
→ hotpancakes joined15:15
← wxg left15:15
→ WakiMiko joined15:19
← hotpancakes left15:20
→ blipz joined15:21
← stfn left15:23
→ stfn joined15:24
→ hexa- joined15:24
→ orbyt_ joined15:27
→ hotpancakes joined15:30
← hotpancakes left15:30
← SkarmoutsosV left15:30
→ causasui joined15:30
← courrier left15:32
→ hussam joined15:34
← causasui left15:35
← tombusby left15:36
← greatgatsby left15:37
→ greatgatsby joined15:38
← gloomy left15:38
→ tombusby joined15:39
← foul_owl_ left15:40
→ manuelschneid3r joined15:40
→ foul_owl_ joined15:44
talx hello guys15:45
how come I get this error15:45
error: in `sync`: [Errno 2] No such file or directory: '/home/nwertheizer/trees/npu-stack/.repo/manifests/.git/HEAD'15:45
error: manifest missing or unreadable -- please run init15:45
but when I did it in ~/.ssh15:45
it worked and I manged to initiate and sync15:45
I want to get all the directories in /home/nwertheizer/trees/npu-stack/15:45
and I get the error above15:46
any idea how to issue it ?15:46
→ blackmesa joined15:49
talx 939 people here15:50
all bots ?15:50
← quipa left15:51
_ikke_ talx: lot of people idling15:52
← orbyt_ left15:52
← blipz left15:53
RichiH _ikke_: who you callin' idle mate?15:54
→ peacememories joined15:55
→ igemnace_ joined15:55
_ikke_ not you apparently ;-)15:55
← igemnace left15:55
← blackmesa left15:56
→ blackandblue joined15:57
← talx left15:57
→ kadoban joined15:58
→ mven_ joined16:06
← stitched_unicorn left16:08
→ stitched_unicorn joined16:10
→ zampano joined16:12
← stitched_unicorn left16:13
→ keepLearning512 joined16:13
← jelmer left16:14
→ jelmer joined16:14
← keepLearning512 left16:17
→ gloomy joined16:18
← troyt left16:20
→ troyt joined16:20
← mooasaurus left16:25
→ mooasaurus joined16:26
← peacememories left16:28
← mooasaurus left16:34
→ mooasaurus joined16:34
→ peacememories joined16:41
→ totte_ joined16:42
← totte left16:43
totte_ → totte16:45
→ thiago joined16:50
→ metrixx joined16:50
→ dpyro joined16:51
→ watabou_ joined16:51
→ basteh joined16:53
← metrixx_ left16:54
← watabou_ left16:55
← jungsubk left16:55
← gloomy left16:58
→ venmx joined17:06
→ R2robot joined17:09
← peacememories left17:09
→ watabou_ joined17:14
← dfaught left17:15
→ ferdna joined17:17
→ WhiskerBiscuit joined17:19
← watabou_ left17:19
→ blackmesa joined17:19
→ blackmesa1 joined17:22
← blackmesa left17:24
→ yohnnyjoe joined17:26
← blackmesa1 left17:27
→ krrishg joined17:28
→ causasui joined17:31
← causasui left17:36
→ watabou_ joined17:36
watabou_ → watabou17:37
← renich left17:45
← Achylles left17:49
← kapil____ left17:50
→ jcbitter joined18:03
→ blackmesa1 joined18:06
← maryo_ left18:08
← krrishg left18:08
← mven_ left18:11
← yyy- left18:15
→ dege joined18:17
← Silenced left18:18
→ gloomy joined18:20
← WhiskerBiscuit left18:20
← greatgatsby left18:20
← neunon left18:26
→ neunon joined18:28
Celelibi How do I get the list of commit ids between two commits?18:34
(in order)18:34
→ govg joined18:35
Celelibi The problem I have is when there are multiple path.18:35
Basically, the shortest path should be ok.18:36
Or any path.18:36
kadoban Celelibi: Does git rev-list do close enough to what you want?18:36
← kp__ left18:36
Celelibi I used to use git log, I don't know what rev-list do.18:36
← netj left18:38
→ netj joined18:39
→ treehug88 joined18:39
← mitxela left18:40
← manuelschneid3r left18:40
Celelibi Hm.18:40
rewt Celelibi, why do you want the list of commits? there may be another way to solve the real problem18:40
→ mitxela joined18:41
→ pulec joined18:41
Celelibi Here's the deal: I found a patch that apply to a pretty old commit. And I would like to make it work with the latest revision.18:42
So I made a branch out of this patch and now would want to rebase it to master.18:42
But there are many merge conflicts on the way.18:42
So I'd rather deal with them one at a time, by rebasing that branch to every commit on the way to master.18:43
→ kp__ joined18:45
Celelibi There might be smarter ways to identify which commit create the conflicts without rebasing to every commit.18:45
But it works.18:45
rewt ok, so `git merge-base branch1 branch2` gets you the last common ancestor of the 2 branches; then from there you can do something like `git log --reverse --ancestry-path $(that hash)..master`18:46
so something like: git log --reverse --ancestry-path $(git merge-base patchbranch master)..master18:47
(not tested)18:47
Celelibi if merge-base does what I think it does, then it sounds useless there.18:48
← snowkidimd left18:48
← virmin left18:48
rewt you need to find the commit where master and your branch diverges... that's what merge-base gives you18:49
Celelibi Ah. It's patchbranch~18:49
→ plutonas joined18:49
rewt it's a single commit?18:49
Celelibi The patchbranch has a single commit.18:49
rewt ah18:49
plutonas i have a merge conflict in some file and would like to make version of the file the actual one. Is there a way to do this without editing the << and == and >>> in the actual file? (I use command line)18:49
rewt then yeah, you don't really need merge-base there18:50
Celelibi rewt, I don't really understand the effect of --ancestry-path. Isn't that what git log would do anyway with a range of commits?18:52
→ snowkidimd joined18:52
→ moei joined18:52
rewt you can try with and without, see if there's a difference18:53
← snowkidimd left18:53
rewt whenever doing anything like this, I like to have gitk open to see the git history at each command to make sure I'm on the right track18:57
→ SkarmoutsosV joined18:57
← watabou left18:59
→ manuelschneid3r joined18:59
Celelibi rewt, I use an alias git lol. Which is basically gitk but in a terminal. ^^19:02
← benharri left19:03
Celelibi But there are really too many commits to check on every commit. And since 99% of the commits rebase without conflicts...19:03
→ benharri joined19:04
← blackandblue left19:04
Celelibi (near 40000 commits)19:04
→ blackandblue joined19:04
→ wolfshappen_ joined19:05
← wolfshappen left19:05
Celelibi BTW, I'm pretty glad reflog was there when a mistake happened and I lost my commit. ^^19:05
Maybe I should put a tag every now and then.19:06
rewt you have 40k commits to rebase this patch against?19:06
Celelibi yup19:06
→ n3wborn joined19:07
← plutonas left19:07
← neunon left19:08
← mona_x left19:09
→ neunon joined19:10
← SkarmoutsosV left19:14
→ SkarmoutsosV joined19:14
← SkarmoutsosV left19:15
davidfetter hi19:15
how do i find out at what commit a file went away?19:15
j416 davidfetter: git log -- file19:16
davidfetter j416, thanks!19:16
j416 yw!19:16
Celelibi Is there a way to find the latest tag earlier than a given date?19:17
Hm... with 2 commits per seconds, it's 5 hours to rebase against 40k commits. ^^19:20
xcm → Guest8882519:21
→ xcm joined19:21
← Guest88825 left19:21
← blackmesa1 left19:21
j416 keep in mind that you would preferably need to verify that the patch applied cleanly (semantically) for each one19:21
why do you need 40,000 patches?19:22
← czart left19:25
← benharri left19:26
← kjartan left19:26
→ benharri joined19:26
← lucasb left19:28
→ horribleprogram joined19:31
→ kjartan joined19:31
→ causasui joined19:32
← egovirus left19:33
rewt Celelibi, you could maybe use git bisect to skip the first 20k, then the next 10k, then the next 5k, etc19:35
→ \void joined19:35
j416 ah, read the backlog now. sorry.19:36
bisect seems like it'd do the trick19:37
← causasui left19:37
j416 mark the commits where patch application fails as bad, the others as good19:37
should be able to automate it with git bisec trun19:37
→ causasui joined19:38
→ watabou joined19:38
j416 result will be the first commit where the patch does not apply19:38
worth a shot at least.19:38
← zampano left19:38
→ dviola joined19:39
→ courrier joined19:39
← bebbet left19:42
→ Achylles joined19:42
→ bebbet joined19:43
→ zampano joined19:48
→ SkarmoutsosV joined19:50
→ rafasc joined19:50
← osa1 left19:50
← gloomy left19:51
← n3wborn left19:53
→ gloomy joined19:54
→ n3wborn joined19:55
→ CarlFK joined19:57
CarlFK is there a 1 line version of: if -d foo; cd foo; git pull; else git clone foo ?19:59
_ikke_ no19:59
CarlFK boo.19:59
← crtcji left19:59
← watabou left20:06
→ orbyt_ joined20:08
osse Looks like one line to mee20:09
← mattcen left20:13
→ mattcen joined20:18
← totte left20:18
→ totte joined20:19
j416 CarlFK: [ -d foo ] && (cd foo && git pull) || git clone foo # ?20:20
not that you couldn't use if on one line too.20:20
CarlFK I was hoping for something that just took a url as a parameter20:22
j416 you could easily script it?20:23
CarlFK and built into git or distributed with git so that I can assume it exists20:23
im tired of scripting it :p20:23
j416 maybe you don't need it20:24
→ mowcat joined20:33
← jcbitter left20:35
← netj left20:45
→ netj joined20:45
← gloomy left20:48
→ greatgatsby joined20:49
← igemnace_ left20:50
→ mona_x joined20:51
← cd left20:52
← orbyt_ left20:57
← dviola left21:02
→ orbyt_ joined21:07
→ jcbitter joined21:10
→ dviola joined21:13
→ AtumT joined21:13
← AtumT_ left21:16
rafasc CarlFK: git clone https://github.com/user/repo $repo || git -C $repo git pull21:17
delete the last 'git' in there.21:18
CarlFK and add a cd repo i think21:19
rafasc there's no need to cd.21:19
_ikke_ -C is cd21:19
CarlFK oh neet21:19
rafasc and clone url dir; clones into the specified dir.21:19
→ Mattiaslndstrm joined21:20
CarlFK so this does that part of what I want? git -C $dir pull21:20
_ikke_ yes21:21
Rh0nda → Rhonda21:21
← manuelschneid3r left21:21
→ mven_ joined21:22
← horribleprogram left21:22
rafasc if using what I posted, you probably want to redirect stderr to null.21:22
because git complains when it isn't able to clone.21:22
CarlFK I don't mind the complains output as long as the || eats the error state so my bash -e script doesn't abort21:23
rafasc if pull fails I guess.21:25
|| true at the end.21:25
CarlFK what would cause it to fail?21:25
rafasc no network? not sure.21:25
→ laerte joined21:25
← kneeki left21:26
CarlFK that's a fail I want to catch so good.21:26
_ikke_ no permissions, disk full21:26
→ spacesuitdiver joined21:26
rafasc when git pull fails because of no-ff, does it exit with error?21:26
→ AtumT_ joined21:27
_ikke_ are you confusing git push?21:27
Or are you assuming merge.ff == only?21:28
← AtumT left21:28
rafasc the latter, or --ff-only21:28
← andrybak left21:29
_ikke_ I would assume so21:29
or rather, expect21:29
supernovah in what case would git log --graph give a ')' instead of a '|' in a branch diagram21:30
rafasc when you're using --decorate21:31
supernovah but it only appears on one commit in the same branch21:31
rafasc also, piping log output is very misleading cause the output varies if on a tty.21:32
try git log ... | cat21:32
and check the differences21:32
or explicitly pass an argument to decorate (--decorate[=short|full|auto|no])21:33
supernovah log --graph --abbrev-commit --decorate --format=format:'%C(bold blue)%h%C(reset) - %C(bold green)(%ar)%C(reset) %C(white)%s%C(reset) %C(dim white)- %an%C(reset)%C(bold yellow)%d%C(reset)' --all21:35
and its a tty obviously21:35
and if I were piping to a file I'd use an pty with an ioctl to lie about it being a tty anyway because unformtted text is gay21:36
omg I know what it is lol, it's the closing bracket for the (HEAD -> master, origin/master, origin/HEAD bit21:37
← pyfgcr left21:37
← mitxela left21:38
→ tsprlng joined21:39
_ikke_ and parsing git log output is not recommended anyway21:39
supernovah who by and who suggested that was being done anyway21:39
rafasc you're strongly implying you are doing it.21:40
_ikke_ supernovah: The question kind of implies your are trying to parse it21:40
supernovah don't make assumptions then _ikke_21:40
← xcm left21:40
rafasc supernovah: then tell us what you're doing.21:40
supernovah rafasc: reading the freaking output of git log wondering why I saw a ) instead of |21:41
_ikke_ supernovah: I have to make assumptions all the time21:41
→ xcm joined21:44
← plexigras left21:45
rafasc did the result of %d word-wrap on you?21:47
→ nowhere_man joined21:48
→ blackmesa1 joined21:49
supernovah rafasc: yea21:49
→ kneeki joined21:50
← GreenJello left21:50
← xcm left21:51
→ xcm joined21:51
→ GreenJello joined21:53
rafasc _ikke_: half of the time here, we make assumptions on what people are doing vs what they're saying they doing.21:53
_ikke_ yes21:53
rafasc I feel you. 👊21:54
did you figure out the rebase thing you were looking into yet?21:54
supernovah rafasc: the opposite of x:y21:55
← wildlander left21:55
_ikke_ Rebaing thing?21:55
← Anthaas left21:55
rafasc _ikke_: weren't you trying to figure out why some tests relating to rebase failed?21:56
some encoding issue, or something like that?21:56
_ikke_ ah, yes21:56
Not yet, need to spend a bit more time21:56
it has more to do with commit encoding than rebasing21:57
rafasc supernovah: it works 90% of the time. And the time saved by it make it worth the 10% it doesn't.21:58
← SkarmoutsosV left22:00
→ Anthaas joined22:01
→ watabou joined22:03
← Anthaas left22:05
→ Anthaas joined22:08
← Anthaas left22:09
→ Anthaas joined22:09
← yohnnyjoe left22:13
← thiago left22:15
← Mattiaslndstrm left22:23
ComradeCrocodill → Crocodillian22:25
← spacesuitdiver left22:29
→ spacesuitdiver joined22:31
← mescalinum left22:35
→ yyy- joined22:35
← watabou left22:36
→ mescalinum joined22:40
← fstd left22:41
→ fstd joined22:43
← fstd left22:45
→ fstd_ joined22:45
fstd_ → fstd22:46
← venmx left22:52
→ libertyprime joined22:57
← leeN left22:59
← HelloShitty left23:00
→ watabou joined23:04
← xcm left23:04
→ xcm joined23:05
← mescalinum left23:10
→ novum joined23:10
novum anyone have a good idea? I want just code (w no comments) in one repo with just a minimal repo. However, I really want to have good documentation as well. However, I am afraid that the comments will make the script slower (it's not a great language don't ask) and I also want to have very thorough documentation with pictures, but I don't want to add overhead to the repo. Should I just make a separate repo?23:13
* and just a minimal readme23:13
← blackandblue left23:13
← Achylles left23:14
→ matheustavares joined23:19
← metrixx left23:20
rafasc novum: comments that make a script slower? weird.23:20
novum: just keep the script+comments in your repo and use it as is.23:20
novum it's not python or shell or ruby or anything like that. For loading definitely slower. for execution maybe. somewhat unclear.23:20
rafasc then strip the comments as part of your build process.23:21
→ Karyon joined23:21
novum hm. it's a vimscript. there is no building23:21
rafasc something like some javascript minimized libs work.23:21
novum vimscripts23:21
there is no compilation. I could write my own vimL minimizer I suppose but..23:22
rafasc sed '/^"/d'23:22
novum I want people to be able to use plug 'novum/whatever' and be done. no building23:22
rafasc you can distribute the minimized version as well.23:23
novum in a separate repository?23:23
rafasc I'd say in the same repo.23:23
either in a directory, or in another branch.23:23
← fstd left23:24
→ mescalinum joined23:24
novum that can't be how it works, though.23:24
→ fstd joined23:24
novum you have plugin/ autoload/ etc in the top directory23:24
a separate branch .. could work and I think plug uses --bare or something and I know you can set the branch but unsure if it will only clone the branch you say. I suppose that'd work.23:25
rafasc for example: keep master as your 'built' files.23:26
← mescalinum left23:26
→ Achylles joined23:27
novum what does that mean? master is the stripped version and dev or some other branch is the commented version?23:27
→ zetneteork joined23:27
rafasc in another branch, could be something like an orphan branch, do your regular development with comments and all... Then build it and deploy it to the master branch.23:27
novum: yes, similar how github deals with project pages.23:28
← treehug88 left23:28
novum this'll probably be gitlab with github as a mirror23:28
but point taken. I suppose I will give it a whirl. thanks23:29
rafasc they're not branches meant to be merged together. They're just independent branches that live in the same repo.23:29
← zetneteork left23:29
rafasc novum: orphan branches are host agnostic.23:29
→ oojacoboo joined23:30
rafasc novum: how other projects deal with this problem?23:30
→ TheRuckus joined23:30
rafasc (if it is indeed a problem)23:30
novum any specific thoughts on how to "build" to master other than something like git checkout master; git checkout dev -- myfiles then manually fix with a script or by editing23:30
other projects do not have comments23:32
they are apparently a luxury23:32
or they just keep them in presumably. most I have seen have no comments23:32
rafasc I would probably abuse git archive.23:34
novum never used it will read man kthx23:34
rafasc and specify a custom tar."deploy".command that includes the "buiding process"23:35
you'll have some overhead because it will compress and uncompress... but meh.23:35
← TheRuckus left23:36
← nowhere_man left23:36
← watabou left23:37
← courrier left23:38
← kjartan left23:40
→ kjartan joined23:44
← g2anj left23:47
← n3wborn left23:47
→ g2anj joined23:51
← n000g left23:53
→ horribleprogram joined23:53
→ n000g joined23:56

Logs Search ←Prev date Next date→ Channels Documentation