IRCloggy #git 2020-03-26

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.

2020-03-26

tiin57 joined00:03
xelxebar left00:04
haasn` joined00:04
Dirak joined00:05
zanetti joined00:05
mat001 joined00:07
xco left00:09
xelxebar joined00:09
haasn` left00:10
PyDon joined00:12
PyDon hi00:12
mat001 left00:13
dartmouthed left00:15
nedbat PyDon: hey there00:15
PyDon hey nedbat, you are also everywhere :-D00:15
omnireq left00:18
omnireq joined00:18
PyDon can anyone help me? Given the following example: https://nopaste.linux-dev.org/?129538300:20
Why does $ git add link/file00:20
fail with: fatal: pathspec 'link/orig' is beyond a symbolic link00:20
but $ git add .00:20
is fine?00:20
omnireq_ joined00:21
haasn` joined00:21
Wetmelon joined00:22
dedowsdi left00:23
Dirak left00:23
omnireq left00:23
buckworst joined00:26
tiin57 left00:29
vdamewood left00:32
str rafasc, OK, if I git-filter-repo --path=myPath it will delete everything BUT myPath, which is OK, BUT, I want myPath to be the root folder of the repo00:32
What am I missing?00:33
tiin57 joined00:33
rafasc str: you are missing the subdirectory filter: https://github.com/newren/git-filter-repo#solving-this-with-filter-repo00:34
roadie joined00:36
str rafasc, AWESOME!00:37
rafasc, now, if I review the log in gitk, I see some blue-ish labels "replace/$hash" like here in this screenshot: https://i.imgur.com/2YiYUxJ.png00:38
What are those?00:38
rafasc man git replace00:39
gitinfo the git-replace manpage is available at https://gitirc.eu/git-replace.html00:39
rafasc git has a mechanism for object replacement, that's probably a map that link each commit to their source commit.00:39
PyDon left00:40
roadie left00:41
PyDon joined00:42
str ... it's like a log00:42
BUT I can get rid of them, right?00:42
Or I can even create the new repo without them, right?00:42
mat001 joined00:43
rafasc see --replace-refs00:46
str rafasc, MAN! YOU ARE THE BEST!00:48
mat001 left00:48
str Thank you a lot00:48
libertyprime joined00:50
jstimm joined00:52
dedowsdi joined00:52
cdown left00:54
cdown joined00:55
mescalinum left00:56
str left00:57
Brainium left00:59
tiin57 left00:59
omnireq_ left01:00
_Warl0ck left01:00
mescalinum joined01:00
dumb_moth left01:00
zanettiZanetti01:01
omnireq_ joined01:01
_Warl0ck joined01:01
Zanettizanetti01:02
tiin57 joined01:03
dumb_moth joined01:03
zanetti left01:03
zlogan left01:04
zanetti joined01:05
cbreak left01:07
jstimm left01:07
cbreak joined01:08
NODE left01:08
NODE joined01:10
zlogan joined01:12
oriba joined01:18
bitnoize left01:19
cliluw left01:24
is_null left01:27
m0viefreak left01:27
tiin57 left01:30
stenno joined01:30
roadie joined01:30
ferdna left01:32
tiin57 joined01:33
roadie left01:35
split_horizon joined01:38
tmz joined01:39
anderson joined01:41
bttf joined01:41
AppleGaming joined01:41
is_null joined01:43
Giant81 joined01:49
foul_owl left01:50
foul_owl joined01:52
PyDon left01:56
g4570n left01:58
rafasc left01:59
tiin57 left01:59
tafa2 left02:00
twb joined02:02
twb "git clone https://bitbucket.org/cowan/r7rs-wg1-infra/" ==> "remote: Mercurial (hg) is required to use this repository."02:03
Is it saying I can still use git, I just need some kind of hg thing like git-svn?02:03
Or is it saying "you have to use hg directly"02:03
tiin57 joined02:03
stenno left02:03
twb There's a "mercurial-git" package to let hg talk to git servers, but I can't see a tool to let git talk to hg servers.02:04
Ah, git-remote-hg02:05
kupi left02:06
zanetti left02:10
zanetti joined02:10
F0rTex left02:10
Es0teric left02:11
F0rTex joined02:11
cpplearner joined02:16
smichel17 left02:27
tiin57 left02:30
Sasazuka left02:31
tiin57 joined02:33
oriba left02:37
navidr left02:50
is_null left02:56
tiin57 left02:59
Narrat left03:00
is_null joined03:02
tiin57 joined03:03
thiago left03:09
caterfxo joined03:09
lagothrix left03:09
lagothrix joined03:09
Es0teric joined03:11
caterfxo left03:18
Beam_Inn joined03:23
bvdw left03:25
bvdw joined03:26
Beam_Inn left03:28
EdFletcher left03:29
tiin57 left03:30
EdFletcher joined03:30
FH_thecat joined03:31
libertyprime left03:32
tiin57 joined03:33
rzanetti joined03:34
zanetti left03:34
Goplat joined03:37
inkbottle left03:39
Es0teric left03:42
libertyprime joined03:44
zebrag joined03:44
Es0teric joined03:45
omnireq_ left03:46
omnireq_ joined03:49
Cabanossi left03:50
waveclaw joined03:54
yitz joined03:56
Cabanossi joined03:57
tiin57 left03:59
tiin57 joined04:03
AppleGaming left04:04
libertyprime left04:04
ojdo left04:06
libertyprime joined04:06
_0xDEADBEEF left04:08
_0xDEADBEEF joined04:08
FH_thecat left04:13
ojdo joined04:13
Khisanth left04:15
Wetmelon left04:18
omnireq_ left04:19
vicfred left04:20
omnireq joined04:23
Khisanth joined04:29
tiin57 left04:30
tiin57 joined04:32
rzanetti left04:33
duderonomy left04:33
maxwilliamson left04:40
caterfxo joined04:40
maxwilliamson joined04:40
cruxeternus joined04:43
dedowsdi left04:44
nickodd joined04:49
wildlander_ left04:51
vicfred joined04:53
Industrial joined04:59
tiin57 left04:59
Phylock joined05:00
pks_ joined05:02
dedowsdi joined05:03
tiin57 joined05:03
pks left05:03
pks_pks05:03
clime left05:06
caterfxo left05:08
buckworst left05:10
xco joined05:13
caterfxo joined05:22
dmc left05:22
DarthRadar joined05:23
caterfxo left05:23
caterfxo joined05:24
tiin57 left05:29
tiin57 joined05:33
sauvin joined05:35
omnireq left05:38
omnireq joined05:38
g00s joined05:40
Technodrome left05:42
vicfred left05:48
diverger left05:51
causasui left05:54
diverger joined05:55
gschanuel joined05:58
sidetrack left05:58
tiin57 left06:00
tiin57 joined06:03
settermjd joined06:05
Betal left06:09
Wetmelon joined06:11
markopasha joined06:13
Industrial left06:15
nabil_ joined06:16
nabil_ left06:16
nabil_ joined06:17
Wetmelon left06:17
nabil left06:18
markopasha So I've got two files in the project which I edit everytime after pull. I edit them due to my computer settings then I create a different branch and work on it. These file can get updated so I cant completely ignore them. What is the better way to deal with this06:20
warthog9 left06:21
YuGiOhJCJ joined06:22
twb markopasha: split it into two files, one ignored, one edited?06:24
gschanuel left06:24
xco left06:25
gschanuel joined06:26
markopasha Would my changes be shown in history? Because one has couple of credentantials twb06:28
twb it sounds like you have file like ./config.php06:29
and it contains myhostname = "supercoolserver.com" and database_password = "swordfish"06:29
kjartan left06:29
tiin57 left06:29
twb So like split that into config.php with the default settings (version controlled), and config.local.php (not version controlled)06:30
kjartan joined06:30
osse markopasha: !config06:30
gitinfo markopasha: [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas06:30
warthog9 joined06:31
markopasha yea i have Gemfile and config/database.yml06:31
twb, osse gitinfo thank you06:32
gitinfo markopasha: you're welcome, but please note that I'm a bot. I'm not programmed to care.06:32
markopasha hahah:D06:33
twb Also if your database and app run on the same server, use socket auth and not f*cking passwords06:33
tiin57 joined06:33
twb passwords are the worst06:33
markopasha it is as environmental variables06:34
is that bad too06:35
twb that's just changing how you read the password06:35
but you're still using password-based authentication06:35
markopasha twb, yea true.06:36
twb With socket auth, the database just goes "hey kernel, what unix user connected to me? Oh, it was avahi-daemon? OK, then I'll use my GRANT rules for avahi-daemon"06:36
karab44 joined06:36
Goplat left06:37
_leb joined06:39
pandakekok9 joined06:39
cbreak left06:40
cbreak joined06:41
Forlorn_ joined06:42
warthog9 left06:43
warthog9 joined06:44
xco joined06:46
markopasha :D i see06:50
_leb left06:50
leb joined06:52
leb left06:52
omnireq left06:53
omnireq joined06:54
foul_owl left06:55
mat001 joined06:59
tiin57 left06:59
mat001 left06:59
mat001 joined07:00
tiin57 joined07:02
dmc joined07:04
dmc left07:04
foul_owl joined07:09
ephemera_ left07:17
NightStrike left07:19
Phylock left07:22
g00s left07:22
ephemera_ joined07:23
Noti joined07:25
noodle joined07:26
tiin57 left07:29
stitched_unicorn joined07:32
tiin57 joined07:33
funkatronixxx joined07:37
oxymoron93 joined07:46
Jackneill joined07:47
g00s joined07:48
loganlee joined07:55
feriman joined07:55
fstd_ joined07:55
loganlee https://git-scm.com/book/en/v2/Git-Branching-Rebasing hello on figure 47 it says, "If you run a git log when your history looks like this, you’ll see two commits that have the same author, date, and message, which will be confusing. " which two commits are those?07:55
Industrial joined07:56
mat001 left07:56
loganlee is it C4' and C6?07:57
osse I would guess C4 and C4'07:57
fstd left07:59
fstd_fstd07:59
osse this was an overly complicated explanation in my opinion'07:59
tiin57 left07:59
osse but i am fairly confident it's C4 and C4'08:00
dviola joined08:00
osse because it is the guy who merged and pushed who is doing the rewriting so that C4 appears after C5 instead of parallel to it08:01
Naptra joined08:01
tiin57 joined08:03
funkatronixxx left08:04
ilogger2 joined08:06
kent\n joined08:06
Nizumzen joined08:06
altendky joined08:06
mgsk joined08:06
Soni joined08:06
miklcct joined08:06
de-vri-es joined08:06
Tk55 joined08:06
T4cC0re joined08:07
ncdulo joined08:07
uebera|| joined08:07
dysfigured joined08:07
orgg joined08:07
kashike joined08:07
fructose joined08:07
nikivi joined08:07
hexnewbie joined08:07
NitrousPG joined08:07
FFY00 joined08:07
Habbie joined08:07
shabius joined08:07
TomyLobo joined08:07
POQDavid joined08:08
n-st joined08:08
iloop joined08:08
NODE joined08:08
Kamilion joined08:08
causasui joined08:08
Dirak joined08:09
net_wayfarer joined08:09
loganlee osse, i agree that it is C4 and C4'08:09
saidelike joined08:10
arooni joined08:11
dmc joined08:15
Celelibi joined08:16
HumanGeek joined08:19
loganlee when i was out of caffeine and i read a sentence it made no sense to me. then i just drank coffee mixed in cold milk, and now i read again and it makes sense now08:22
cpplearner I'm trying to rebase my forked 'develop' branch to the original 'devleop'08:24
branch, but thers' a conflict.08:25
So, I'm trying to add another branch, and pull updates there, and rebase mine to this branch.08:25
Does it sound good?08:25
Or, is there any better way?08:25
osse You will get conflicts either way08:26
cpplearner Oh, yeah. Sorry.08:26
osse unless you mean you want to rebase your branch on something other than develop08:26
say, master or stable or whatever08:26
cpplearner I want to first apply the original repo's updates, and rebase mine to this one, and fianlly make a PR.08:28
As you said, of course I should resolve the conflict.08:28
But, I'm not sure it's the desirable way.08:28
Maybe, another good solution?08:28
osse without knowing more specifics it's hard to say08:29
Why don't you think that's desirable?08:29
cpplearner Hmm, I might break something, so I want to be careful. ='(08:30
osse how can you break anything?08:31
cpplearner Uhh, I'm a total newbie, so I'm just saying. But, can I ask you one more quesiton?08:33
baako joined08:33
cpplearner Is it possible to pull som specific branch from the original repo to my local branch whose name is different from the former?08:33
tiin57 joined08:33
dedowsdi joined08:34
funkatronixxx joined08:37
wootehfoot joined08:37
funkatronixxx left08:39
osse yes08:39
ryzokuken joined08:40
TomyWork joined08:41
mobidrop joined08:43
osse git pull origin branchname08:43
cpplearner osse: Thanks you! I got it working by git pull origin source-branch:destination-branch! =)08:44
chele joined08:45
clime joined08:46
Borkr joined08:50
makaveli7 joined08:51
TomyLobo left08:54
shored joined08:57
tiin57 left09:00
tiin57 joined09:03
nikivi left09:14
nikivi joined09:14
sgn joined09:18
funkatronixxx joined09:24
bvdw joined09:25
fitzgerald1337 joined09:29
tiin57 left09:29
funkatronixxx left09:31
tiin57 joined09:33
bitnoize joined09:38
bdeshi joined09:42
funkatronixxx joined09:45
funkatronixxx left09:46
maczor joined09:48
maczor Hi I've run into an issue when checking out a branch that contains a submodule with update=none while submodule.recurse is set to true that I was hoping to get some opinions on.09:51
When I do this I get a fatal error as git detects that there is no submodule repository present (due to the update=none) however because of the recurse=true it tries to checkout the referenced commit in this submodule.09:52
With GIT_TRACE=1 I get this output:09:53
19:20:18.129343 run-command.c:663 trace: run_command: cd09:53
no-update-submodule; unset GIT_PREFIX; GIT_DIR=.git git09:53
--super-prefix= no-update-submodule/ read-tree -u --reset09:53
4b825dc642cb6eb9a060e54bf8d69288fbee490409:53
fatal: not a git repository: ../.git/modules/no-update-submodule09:53
fatal: could not reset submodule index09:53
It's possible to workaround this issue by deiniting the submodules before checking out the branch. However this is non-ideal09:55
dartmouthed joined09:55
maczor However this seems like a bug?09:56
Phylock joined09:57
scarcry joined09:59
tiin57 left09:59
tiin57 joined10:03
clime left10:03
clime joined10:03
funkatronixxx joined10:03
funkatronixxx left10:07
funkatronixxx joined10:11
fassl joined10:11
libertyprime joined10:17
funkatronixxx left10:18
causasui left10:19
giorgian joined10:22
Dirak left10:23
albel727 joined10:23
raven-au_ joined10:28
osse maczor: if there's a submodule don't you need to run submodule init regardless of whether update=none or not?10:28
tiin57 left10:29
maczor osse: the full reproduction steps are:10:31
The reproduction steps are to:10:31
- Clone the repo10:31
- Checkout a branch containing submodules that are set with10:31
update=none (git checkout branch-name)10:31
- Update and init submodules (git submodule update --init)10:31
- Attempt to switch to a different branch (git checkout other-branch)10:31
Uploaded file: https://uploads.kiwiirc.com/files/1ef06a0d080695398ece32fdd5837ed3/pasted.txt10:31
If I do this when submodule.recurse is set to true, git tries to switch to a specific commit for that submodule despite it not being there and throws the fatal error.10:33
giorgian Hi. If I do `git stash -u` where there are only untracked files, git stash show gives no output, even with -v. Is this a bug?10:33
tiin57 joined10:33
osse maczor: and this only happens if update=none?10:34
sounds like it should happen always10:34
maczor osse: it only happens when update=none and submodules.recurse=true. There are no issues for submodules without update=none. I also tried without submodules.recurse=false and I don't see the fatal error10:35
osse I don't know enough about submodule config to be sure, but it sounds like a bug10:36
antoine joined10:38
POQDavid left10:46
POQDavid joined10:47
DelPhyinf joined10:47
makaveli7 left10:48
ghost43 joined10:48
makaveli7 joined10:53
Habbie left10:53
Habbie joined10:55
PacKetSlayer joined10:56
NODE left10:56
fitzgerald1337 left10:58
maczor Thanks osse. Me too.10:59
NODE joined10:59
tiin57 left10:59
staafl_r19 joined11:02
tiin57 joined11:03
ghost43 left11:04
Brainium joined11:04
baako left11:04
Hello71 joined11:05
opal joined11:06
andreas31 joined11:07
chiastre joined11:07
xelxebar joined11:07
gxt joined11:08
tryte joined11:10
maczor left11:10
POQDavid left11:17
POQDavid joined11:17
iNs joined11:20
ghost43 joined11:20
justanotheruser joined11:23
nikivi left11:23
nikivi joined11:23
ghost43 left11:25
bitnoize left11:26
ghost43 joined11:26
bitnoize joined11:27
giorgian left11:28
tiin57 left11:29
funkatronixxx joined11:31
thiago joined11:32
tiin57 joined11:33
def_jam joined11:35
eblip joined11:35
funkatronixxx left11:36
stenno joined11:38
mescalinum joined11:38
dedowsdi left11:38
rosco_y joined11:40
dedowsdi joined11:41
rosco_y I'm currently in my master, and have made a number of changes and realize that the changes would have fit nicely in a new branch. Is it safe for me to create a new branch at this time, and check my changes into the new branch?11:41
pandem yes11:42
libertyprime left11:42
jast I do that all the time11:42
rosco_y thank you pandem :)11:42
jast though I often forget to create and switch to the new branch before committing, so then I have the extra step of cleaning up master11:43
chiastre left11:44
chiastre joined11:44
_0xDEADBEEF joined11:45
wootehfoot left11:45
iNs left11:45
Kaivo joined11:45
iNs joined11:46
scorphus @rosco_y and @jast, I have a handy git command for exactly that: https://github.com/scorphus/scorphiles/blob/master/bin/rtomayko/git-cut-branch11:49
jast cool... I think it can be simplified a lot, though :)11:50
nabil__ joined11:51
jast the bare essentials of it is this: git branch <new name>; git reset --hard '@{u}'; git checkout <new name>11:51
(as always, --hard is a potentially destructive operation, so don't use this unless you know what you're doing :))11:52
hyperair joined11:52
nikivi left11:56
nikivi joined11:56
jast a safer alternative that avoids destroying uncommitted changes: old=$(git symbolic-ref --short HEAD); git checkout -b $new; git push -f . $old@{u}:$old11:57
add some error handling and sanity checks and you end up with a ~15 line script11:57
Technodrome joined11:58
buckworst joined11:59
tiin57 left11:59
jast for a friendlier reflog entry on the source branch: old="$(git symbolic-ref HEAD)"; git checkout -b "$new"; git update-ref -m "cut branch $new" "$old" "${old#refs/heads/}@{u}"12:01
ghost43 left12:02
tiin57 joined12:03
ghost43 joined12:03
dedowsdi left12:03
POQDavid left12:04
POQDavid joined12:04
aw1 joined12:05
jast some more hints:12:05
to unambiguously abbreviate an object ID: git rev-parse --abrev-ref <sha1>12:06
CodeSlingerPaul joined12:06
jast to strip the refs prefix from a shell variable: current="${ref#refs/heads/}" (avoids an echo/sed pipe)12:07
thiago left12:09
Lucas_Gray joined12:09
mescalinum left12:09
mowcat joined12:11
jast to look up a branch's upstream: git for-each-ref --format='%(upstream)' refs/heads/foo12:11
mescalinum joined12:12
jast that's all I can think of for now12:12
ChrisLane joined12:14
thiago joined12:21
ghost43 left12:22
nikivi left12:22
ghost43 joined12:22
nikivi joined12:22
snakeice joined12:22
NODE left12:23
ChrisLane left12:24
NODE joined12:25
ChrisLane joined12:25
Lucas_Gray left12:26
NODE left12:26
NODE joined12:28
Maxattax joined12:29
graingert joined12:29
tiin57 left12:29
NODE left12:29
graingert how does $XDG_CONFIG_HOME/git/ignore work on windows? https://git-scm.com/docs/gitignore12:30
NODE joined12:31
jast from the manpage: If $XDG_CONFIG_HOME is either not set or empty, $HOME/.config/git/ignore is used instead.12:32
tiin57 joined12:32
graingert jast: does windows have $HOME?12:33
osse graingert: echo $HOME in Git Bash12:34
Betal joined12:34
buckworst left12:34
jast seems to use Windows's $USERPROFILE12:34
graingert you mean %USERPROFILE%12:35
jast that's just syntax differences depending on which shell you use12:35
graingert so if I put ignore lines in here they will be ignored globally? %USERPROFILE%\.config\git\ignore12:36
jast well, it seems likely :)12:36
AppleGaming joined12:37
osse ttps://github.com/git/git/blob/2f268890c2cd2f115424936bbee27f8269080e5e/compat/mingw.c#L2524-L254612:39
goddammit12:39
https://github.com/git/git/blob/2f268890c2cd2f115424936bbee27f8269080e5e/compat/mingw.c#L2524-L254612:39
I wonder why this code is in git itself and not the git-for-windows repo12:39
jast hysterical raisins I bet :)12:42
_ikke_ yum12:43
jast Welcome to #git, home of the best breakfast foods12:46
Lucas_Gray joined12:46
_0xDEADBEEF left12:48
andreas31 left12:50
libertyprime joined12:52
andreas31 joined12:53
blowry joined12:54
jayr joined12:57
maxwilliamson joined12:57
tiin57 left12:59
mauz555 joined13:01
mauz555 left13:01
mauz555 joined13:02
tiin57 joined13:03
crose joined13:04
wootehfoot joined13:05
mescalinum left13:08
mescalinum joined13:08
mauz555 left13:09
mauz555 joined13:10
cdown_ joined13:10
cyberpear joined13:12
cdown_cdown13:13
silverwhitefish joined13:14
POQDavid left13:25
POQDavid joined13:25
irrgit joined13:26
gxt left13:26
gxt joined13:27
tiin57 left13:30
POQDavid left13:31
POQDavid joined13:32
mat001 joined13:32
oriba joined13:32
tiin57 joined13:33
POQDavid left13:40
POQDavid joined13:41
lucasb joined13:42
_0xDEADBEEF joined13:43
duderonomy joined13:44
libertyprime left13:46
g4570n joined13:47
nikivi left13:49
nikivi joined13:49
bvdw left13:50
bvdw joined13:51
Fernando-Basso[m joined13:53
Fernando-Basso[m In an Ubuntu system, ‘git diff -R’ display whitespace changes. On another computer running Arch Linux, ‘git diff -W’ (not ‘-R’) displays whitespace changes. Reading the help, I couldn't see any thing that would hint that either ‘-R’ or ‘-W’ would display whitespace changes.13:53
_ikke_ diff -R is reverse diff13:55
Fernando-Basso[m -R: Swap two inputs; that is, show differences from index or on-disk file to tree contents.13:55
-W, --function-context: Show whole surrounding functions of changes.13:55
From git diff --help13:55
“Show whole surrounding functions of changes.” <-- what does that mean?13:56
Wetmelon joined13:56
thiago left13:56
_ikke_ It means that it tries to show the complete function, not just 3 lines before and after13:56
Fernando-Basso[m yeah, but what does it have to do with visually displaying whitespace changes? What does “function” mean here?13:57
mowcat left13:57
_ikke_ it should not13:57
(have an influence)13:58
pupsikov joined13:59
tiin57 left13:59
tiin57 joined14:03
thiago joined14:04
POQDavid left14:11
POQDavid joined14:12
Lucas_Gray left14:13
scde joined14:17
Guy1524 joined14:18
chele left14:21
nickodd joined14:24
nabil_ joined14:27
akem_ joined14:27
katie_- joined14:29
chele joined14:29
Aaeriele joined14:29
tiin57 left14:29
mooasaurus joined14:29
diverger joined14:30
nabil__ left14:30
Guy1524 left14:30
doug16k joined14:30
cjohnson joined14:30
berndj joined14:30
luis joined14:31
th_ joined14:31
timewalker joined14:31
jetchisel joined14:31
sjohnson joined14:32
mat001 left14:32
mat001_ joined14:32
sbeyer joined14:32
fitzgerald1337 joined14:33
tiin57 joined14:33
fahadash joined14:33
Dougie187 joined14:33
ExeciN joined14:34
_0xDEADBEEF left14:37
cfoch joined14:39
ojdo joined14:40
jayr Fernando-Basso[m: are the git-configs different on the two systems?14:40
rustyshackleford joined14:40
Kobaz joined14:43
Kobaz is there a way to have fully automatic remote tracking branches, like subversion nested repos sort of thing14:43
ie: when you git clone, the whole thing comes down14:43
osse when you clone the whole thing does come down14:44
all the things you see when you run git branch -r are in fact local14:44
however, updating them aftewards isn't automatic. git fetch does that.14:45
Kobaz it's not a branch in the same repo, it's another repo14:45
like nested repos14:45
that's what i meant actually14:45
i kind of set it up a while ago but i don't think it's quite right14:46
it involved a lot of messy merging14:46
ah submodules14:46
rhinojunior joined14:46
ribasushi joined14:46
Kobaz and subtree merging14:47
so here's the situation... i want to split code from updates... so we have an update/deploy system that tracks changes pushed to each customer, and i don't think that belongs in the main repo14:48
so someone could choose to check out only the code, and not the built-updates based on the code14:48
because i have 500 megs of deployment files right now and there's only a few dozen releases tracked right now14:49
jayr left14:50
Kobaz oooooh, i have a better idea... put the deployment on nfs and don't git that. Everyone would use the same deployment mount so we keep versioning in sync14:50
deployment is heavily human-coordinated so there's no locking issues or that sort of thing14:51
osse: thanks, you're a genius!14:51
makaveli7 left14:52
nikivi left14:52
nikivi joined14:52
sidetrack joined14:55
osse i find that hard to believe14:55
NODE left14:56
libertyprime joined14:58
NODE joined14:58
tiin57 left14:59
sidetrack left14:59
tiin57 joined15:03
thiago left15:04
thiago joined15:04
dhollinger joined15:07
mat001_ left15:09
yonder joined15:09
xcm joined15:10
mat001 joined15:11
nikivi left15:11
nikivi joined15:12
R2robot joined15:17
libertyprime left15:20
paul424 joined15:20
roadie joined15:22
libertyprime joined15:26
scde left15:26
makaveli7 joined15:27
mowcat joined15:27
tiin57 left15:29
POQDavid left15:32
POQDavid joined15:33
tiin57 joined15:33
vdl joined15:35
nuala joined15:36
causasui joined15:38
Kobaz osse: hah, i solve half my problems just talking through them to my dog15:39
justanotheruser left15:39
osse dog, eh15:40
Kobaz he helps me code15:40
osse i guess it's like a rubber dock, except it's got four legs, no wings, and is made of flesh15:40
*duck15:40
POQDavid left15:44
POQDavid joined15:44
hexnewbie left15:46
hexnewbie joined15:46
nikivi left15:49
nikivi joined15:49
mowcat left15:50
parsnip was the solution a frisbee?15:54
mescalinum left15:55
justanotheruser joined15:55
ghost43 left15:55
thiago left15:55
ghost43 joined15:55
thiago joined15:56
Kobaz yes15:57
NightStrike joined15:57
xcm left15:59
tiin57 left15:59
xcm joined16:00
tiin57 joined16:03
mobidrop left16:07
syr joined16:10
opal left16:12
nuala left16:12
ghost43 left16:15
ghost43 joined16:16
HumanGeek left16:19
HumanGeek joined16:20
chele left16:28
tiin57 left16:29
feriman joined16:32
tiin57 joined16:33
multiproc joined16:36
tonyj joined16:37
thiago left16:38
eb0t joined16:40
eblip left16:40
eb0t_ joined16:41
def_jam left16:41
cek joined16:46
thiago joined16:49
troulouliou_div2 joined16:54
faceface_ joined16:59
tiin57 left16:59
faceface_ this may be dumb, but my server is running on git controlled files, I want to ff a different branch, and then rebase my current branch on top of that17:00
git branch # REL1_33 / * REL1_33-gero17:00
troulouliou_div2 left17:01
faceface_ I want to ff REL1_33 because it's behind origin, then rebase REL1_33-gero on REL1_33 without switching to REL1_33... impossible?17:01
tiin57 joined17:03
axisys joined17:08
cdown left17:08
g00s joined17:09
andreas31 left17:10
caterfxo joined17:11
mingdao joined17:14
troulouliou_div2 joined17:14
andreas31 joined17:15
pinkisntwell joined17:16
pinkisntwell Sometimes when doing merges on yaml files I get conflicts even though what happened is someone added a bunch of lines, I added another bunch of lines but those two sets of lines have common subsets. Git thinks that those common subsets means they are somehow related, even though they are not.17:17
How do I get git to just put one bunch above the other and move on?17:17
tango_ joined17:19
libertyprime left17:21
karoshi joined17:23
karab44 joined17:23
scde joined17:25
karab44 left17:25
rafasc joined17:27
tiin57 left17:29
tiin57 joined17:33
thiago left17:34
thiago joined17:35
celphi How do i remove all unmerged files17:36
oriba left17:42
fflam joined17:42
nikivi left17:45
nikivi joined17:45
omnireq joined17:48
thiago left17:48
thiago joined17:49
Fernando-Basso[m jayr: Besides different git versions on ubuntu and arch linux, nothing else was explicitly changed than the defaults, which I can't check right now.17:50
mauz555 left17:51
nikivi left17:58
nikivi joined17:59
tiin57 left17:59
Jackneill joined18:01
nikivi left18:02
tiin57 joined18:02
nikivi joined18:02
mescalinum joined18:03
meejah joined18:04
nickodd left18:07
nikivi left18:08
nikivi joined18:08
Giant81 joined18:08
diogenese joined18:09
TomyWork left18:09
omnireq left18:10
omnireq_ joined18:10
funkatronixxx joined18:10
xco joined18:11
troulouliou_div2 left18:11
meejah celphi: what was your git question from #python?18:12
tango_ left18:14
tango_ joined18:15
vicfred joined18:16
Jck_true joined18:16
multiproc left18:17
nikivi left18:23
nikivi joined18:24
m0viefreak joined18:24
scde left18:27
tiin57 left18:29
omnireq_ left18:31
omnireq_ joined18:31
tiin57 joined18:33
shush joined18:42
shush Hello, so I fucked up on something. I created a copy of a remote branch locally, merged (with fast-forward), then pushed.18:43
I need to to undo that18:43
jfindlay joined18:43
jfindlay another developer has a branch with a submodule update, but that submodule update is not part of any commit on that branch. It seems to be treated in a floating unstaged state attached to the branch head. He pushed the branch to the remote and I fetched it from the remote and see the same behavior18:46
silverwhitefish left18:46
jfindlay checking out the branch results in an unstaged change to the submodule hash18:46
if I `git submodule update --init`, the submodule update is gone, but git thinks there is no difference between the local and remote copies of the branch18:47
silverwhitefish joined18:47
cliluw joined18:49
cliluw left18:49
cliluw joined18:49
EdFletcher shush: you can do that with `reset`, but since you've pushed already you must force-push the `reset`ed remote branch now18:51
shush How do I know which commit the previous one is?18:51
omnireq_ left18:52
omnireq_ joined18:52
mescalinum left18:52
osse git reflog branchname18:53
or git reflog origin/branchname since that's the thing you need to push --forcev18:53
paul424 left18:55
jfindlay so the question is, where is the submodule update being stored if not in a commit?18:55
osse jfindlay: nowhere. it's stored in commits18:56
jfindlay it is not in any commit18:56
cliluw left18:56
osse jfindlay: when you checkout a different branch submodules aren't automatically updated18:56
jfindlay it appears only as an unstaged change when you checkout the branch18:56
osse exactly18:56
because the submodule is still in its old state (before the checkout)18:57
paul424 joined18:57
osse so you need to run git submodule update after checkout18:57
cliluw joined18:59
jfindlay so the submodule is holding onto the state from the last time it was updated, and the only time there is a discrepancy is because you switch between branches with different submodule states?18:59
tiin57 left18:59
jfindlay assuming you don't have some kind of hook to auto update submodules on branch change18:59
osse there can also be a discrepancy if you go in there and make changes yourself18:59
jfindlay right18:59
osse but yes, that's right19:00
there is the possibility of a post-checkout hook in which you can run 'git submodule update'19:00
i'm not sure why that's not just the default19:00
jfindlay so how do I update the remote branch? It is in pull request and shows the submodule hash change19:01
osse push, I guess?19:01
tiin57 joined19:03
paul424 left19:03
duderonomy left19:04
jfindlay if I checkout the branch locally and then `git submodule update ; git push origin the_branch`, I get `Everything up-to-date`19:04
osse then the remote branch is up to date19:04
mescalinum joined19:04
osse you have to clarify what you want19:04
duderonomy joined19:05
Dirak joined19:12
jfindlay so the branch is in pull request on the remote. The submodule change appears in the pull request for a different reason: the branch and the target branch have different submodule hashes19:12
thanks for the help19:13
aw1_ joined19:13
osse ne nada19:13
sounds like things are as they should be19:14
jfindlay yes19:14
Apple joined19:14
aw1 left19:14
Narrat joined19:15
AppleGuest3843019:15
AppleGaming left19:16
Noti joined19:19
multiproc joined19:20
funkatronixxx left19:21
xSke joined19:25
aw1_ left19:25
tiin57 left19:29
funkatronixxx joined19:30
Codaraxis joined19:33
tiin57 joined19:33
sidetrack joined19:33
qgTG joined19:33
g00s_ joined19:36
g00s left19:39
shush left19:43
manj-gnome joined19:49
bvdw left19:50
bvdw joined19:51
smichel17 joined19:52
fahadash left19:53
troulouliou_dev joined19:54
dre_ joined19:55
chandoo joined19:57
midnight joined19:59
tiin57 left19:59
tiin57 joined20:03
Borkr left20:07
cliluw left20:10
cliluw joined20:10
cdown joined20:11
multiproc left20:12
Sasazuka joined20:13
cdown left20:18
shush joined20:19
shush left20:19
shush joined20:19
shush osse: I did the reflog command, and I see something like this:...20:20
aebdff80c accolades-redesign-components@{1}: branch: Created from HEAD20:21
Is that hash (aebdff80c) the commit?20:21
osse yes20:21
shush So I just run `git reset --hard accolades-redesign-components@{1}` and that should move it back for me?20:22
Then I do push --force20:22
caterfxo left20:22
osse eys20:23
makaveli7 left20:24
scde joined20:24
manj-gnome left20:24
mauz555 joined20:25
jetchisel left20:25
mescalinum left20:25
jetchisel joined20:26
Maxattax left20:28
silverwhitefish left20:28
tiin57 left20:29
inkbottle joined20:30
tiin57 joined20:33
caterfxo joined20:33
Tawonga joined20:37
Noti left20:37
jim joined20:39
silverwhitefish joined20:40
Ploppz joined20:52
shush What does @{1} mean?20:53
g00s_g00s20:53
_ikke_ 2nd reflog entry of HEAD20:54
j416 shush: the knowledge you seek is buried somewhere deep at man gitrevisions20:54
gitinfo shush: the gitrevisions manpage is available at https://gitirc.eu/gitrevisions.html20:54
j416 in fact not so deep.20:55
scde left20:56
j416 shush: to phrase it like the docs do, the nth prior value of HEAD, in this case 1 prior, so what HEAD was before what it is now20:56
shush So not _one commit_ behind, but rather _one ref log_ behind?20:57
j416 yes20:57
tafa2 joined20:58
kleisli__ joined20:59
j416 trying this out it looks very much like I'm wrong though.20:59
tiin57 left20:59
j416 @{1} means currentbranch@{1}, not HEAD@{1}.21:00
_ikke_: ^21:00
so, if you were on master, @{1} would be where master pointed to before what it points to now21:01
whereas if HEAD is detached (not on any branch), @{1} would I think mean HEAD@{1}.21:01
caterfxo left21:02
j416 shush: ^21:02
osse I'm 83% sure you're wrong21:02
_ikke_ I'm 97.3% sure21:02
j416 from the man page:21:02
> You can use the @ construct with an empty ref part to get at a reflog entry of the current branch. For example, if you are on branch blabla then @{1} means the same as blabla@{1}.21:02
nedbat shush: i interpreted shush's question as, "what does the suffix @{1} mean?"21:02
tiin57 joined21:03
caterfxo joined21:03
shush 102% unsure of 42% of this21:04
j416 I'd be happy to learn if I got this wrong; please do enlighten.21:04
also, what nedbat says is relevent. Since shush has the ref prefix, it's what accolades-redesign-components was one reflog entry prior.21:05
just to get that clear.21:05
osse Looks like you're right21:05
j416 \^^/21:06
at least 17% right, I would like to think.21:06
osse How have I not noticed.21:06
chandoo left21:06
j416 I was also quite sure it was HEAD, until today.21:07
but not 100%21:07
osse Maybe it changed ten minutes ago21:07
shush I would hope you guys would use fractions for your confidence estimates 1/37th of the time21:07
j416 I think so21:07
fractions are fake21:07
shush They're only fake by a factor of 0.13421:08
chandoo joined21:10
warthog9 joined21:11
pupsikov left21:14
chandoo left21:22
cek left21:26
andrybak joined21:26
tiin57 left21:29
multiproc joined21:29
tiin57 joined21:33
chandoo joined21:36
sunrunner20 joined21:37
fitzgerald1337_ joined21:38
zulutango joined21:38
Dougie187 left21:38
tafa2 left21:39
shored left21:39
fitzgerald1337 left21:40
shored joined21:40
irc_user joined21:48
tafa2 joined21:48
silverwhitefish left21:51
tiin57 left21:59
tafa2 left22:00
tafa2 joined22:01
tiin57 joined22:03
dre_ left22:04
Ploppz left22:07
Jackneill left22:07
albel727 left22:10
dmc left22:11
ferdna joined22:11
kantlivelong joined22:13
CodeSlingerPaul left22:13
caterfxo left22:13
POQDavid left22:14
nikivi left22:14
pikapika joined22:15
devster joined22:15
funkatronixxx left22:16
POQDavid joined22:16
nikivi joined22:16
rafasc HEAD@{n} should be equivalent to @{-n}22:16
er, maybe not.22:16
caterfxo joined22:17
j416 I think that's checkouts22:17
i.e. not counting resets or such22:17
could be wrong22:17
pikapika left22:17
j416 hm wait, isn't "-" short for @{-1} or something22:17
meejah left22:17
j416 I use - all the time but I forget what it's short for22:17
rafasc I think you're right, only for checkouts.22:17
orbitex joined22:18
j416 oh, yeah, - for checkouts of course (I was thinking, it works for detached HEAD, so it must be also for resets, but obviously I'm too tired to even be typing this)22:18
pikapika joined22:19
tafa2 left22:22
stenno left22:22
magic_ninja_work joined22:22
snakeice left22:23
fitzgerald1337_fitzgerald133722:25
DelPhyinf left22:25
multiproc left22:26
wootehfoot left22:29
multiproc joined22:29
tiin57 left22:29
mauz555 left22:30
mauz555 joined22:30
tiin57 joined22:33
tafa2 joined22:34
mauz555 left22:35
feriman left22:35
cbreak joined22:40
chandoo left22:41
tafa2 left22:41
shush left22:44
shush joined22:44
vdamewood joined22:45
shush left22:49
ephemer0l_ left22:49
shush joined22:50
ephemer0l joined22:51
butterthebuddha joined22:55
mauz555 joined22:56
tiin57 left22:59
mauz555 left23:01
lucasb left23:02
tiin57 joined23:02
scde joined23:05
Dirak left23:09
omnireq_ left23:09
jstein joined23:10
Dirak joined23:13
omnireq joined23:13
multiproc left23:15
dmc joined23:22
omnireq left23:23
sbeyer left23:23
cliluw left23:27
cliluw joined23:28
justanotheruser left23:28
tiin57 left23:29
tiin57 joined23:33
CryptoDavid joined23:40
justanotheruser joined23:45
prak2 joined23:45
oriba joined23:46
fitzgerald1337 left23:47
Phylock left23:48
omnireq joined23:48
Guest38430 left23:53
buckworst joined23:54
andrybak left23:55
mescalinum joined23:55
scde left23:56
omnireq_ joined23:57
NightStrike left23:57
omnireq left23:57
tiin57 left23:59

Logs Search ←Prev date Next date→ Channels Documentation