8 Sep
2011
8 Sep
'11
12:16 p.m.
a week ago i created a new head jh/mtree, which is supposed to track
master:
git checkout --track -b mtree master
git push origin mtree:jh/mtree
now when i browse that head on gitweb page, i don't see any changes done
to master, such as the one i did today on lib/srdb1/schema/entities.xml.
what does "tracking master" mean? how can i see changes done on master
on jh/mtree head?
-- juha
8 Sep
8 Sep
1:19 p.m.
On Thursday 08 September 2011, Juha Heinanen wrote:
a week ago i created a new head jh/mtree, which is
supposed to track
master:
git checkout --track -b mtree master
git push origin mtree:jh/mtree
now when i browse that head on gitweb page, i don't see any changes done
to master, such as the one i did today on lib/srdb1/schema/entities.xml.
what does "tracking master" mean? how can i see changes done on master
on jh/mtree head?
It just means that master is the default upstream branch. So you
don't have to
enter any branch when pulling. When in you mtree branch, 'git pull' will pull
changes from master.
From git-branch manpage:
-t, --track
When creating a new branch, set up configuration to mark the start-point
branch as "upstream" from the new branch. This configuration will tell git to
show the relationship between the two branches in git status and git branch -
v. Furthermore, it directs git pull without arguments to pull from the
upstream when the new branch is checked out.
This behavior is the default when the start point is a remote-tracking
branch. Set the branch.autosetupmerge configuration variable to false if you
want git checkout and git branch to always behave as if --no-track were given.
Set it to always if you want this behavior when the start-point is either a
local or remote-tracking branch.
--
Greetings,
Alex Hermann
12:36 p.m.
Alex Hermann writes:
When in you mtree branch, 'git pull' will pull
changes from master.
alex,
thank you for your answer. i thought too that it works like than, but
it doesn't.
just to make sure, i just cloned sip-router again:
/usr/src/orig$ mkdir sip-router
/usr/src/orig$ git clone ssh://jh@git.sip-router.org/sip-router
Cloning into sip-router...
jh(a)git.sip-router.org's password:
Receiving objects: 24% (30355/123106), 14.38 MiB | 355 KiB/s
remote: Total 123106 (delta 86469), reused 117147 (delta 81991)
Receiving objects: 100% (123106/123106), 31.57 MiB | 295 KiB/s, done.
Resolving deltas: 100% (86469/86469), done.
then i created a local branch called 'jh' that tracks remote head
jh/mtree:
/usr/src/orig/sip-router$ git checkout --track -b jh remotes/origin/jh/mtree
Branch jh set up to track remote branch jh/mtree from origin.
Switched to a new branch 'jh'
then i do pull:
/usr/src/orig/sip-router$ eg pull
jh(a)git.sip-router.org's password:
Already up-to-date.
and after that lib/srdb1/schema/entities.xml does not contain the change
that i did to master this morning.
-- juha
1:40 p.m.
On Thursday 08 September 2011, Juha Heinanen wrote:
Alex Hermann writes:
Is local mtree branch supposed to track master?
When in you mtree branch, 'git pull' will
pull changes from master.
alex,
thank you for your answer. i thought too that it works like than, but
it doesn't. just to make sure, i just cloned sip-router again:
/usr/src/orig$ mkdir sip-router
/usr/src/orig$ git clone ssh://jh@git.sip-router.org/sip-router
Cloning into sip-router...
jh(a)git.sip-router.org's password:
Receiving objects: 24% (30355/123106), 14.38 MiB | 355 KiB/s
remote: Total 123106 (delta 86469), reused 117147 (delta 81991)
Receiving objects: 100% (123106/123106), 31.57 MiB | 295 KiB/s, done.
Resolving deltas: 100% (86469/86469), done.
then i created a local branch called 'jh' that tracks remote head
jh/mtree:
/usr/src/orig/sip-router$ git checkout --track -b jh
remotes/origin/jh/mtree Branch jh set up to track remote branch jh/mtree
from origin.
Switched to a new branch 'jh'
then i do pull:
/usr/src/orig/sip-router$ eg pull
jh(a)git.sip-router.org's password:
Already up-to-date.
and after that lib/srdb1/schema/entities.xml does not contain the change
that i did to master this morning.
From the above commands, I don't see why it should. Local branch jh tracks
origin/jh/mtree, not master. So why would a master change suddenly appear in
the jh branch? I would expect changes you push to origin/jh/mtree to appear in
your local jh branch after a pull, nothing else.
disclaimer: I always use EasyGit, but i expect these fundamentals to be the
same in bare git and EasyGit mentions no differences on this subject.
--
Greetings,
Alex Hermann
12:58 p.m.
Alex Hermann writes:
Is local mtree branch supposed to track master?
...
From the above commands, I don't see why it
should. Local branch jh tracks
origin/jh/mtree, not master. So why would a master change suddenly appear in
the jh branch? I would expect changes you push to origin/jh/mtree to
appear in your local jh branch after a pull, nothing else.
if jh/mtree tracks master (as it according to the wiki page is supposed
to do) and when jh tracks remote branch jh/mtree, then according to
normal logic it follows that also jh tracks master.
-- juha
1:58 p.m.
On Thursday 08 September 2011, Juha Heinanen wrote:
Alex Hermann writes:
It does, but requires a bit more work. Git doesn't do recursive pull's
(automatically).
You'd first have to pull master in jh/mtree, push that out to origin, and then
pull that into local jh.
The --track option really doesn't do anything more than setting a default, it
does not imply automatic behavior in any way.
--
Greetings,
Alex Hermann
Is local mtree branch supposed to track master?
...
From the above commands, I don't see why it
should. Local branch jh
tracks origin/jh/mtree, not master. So why would a master change
suddenly appear in the jh branch? I would expect changes you push to
origin/jh/mtree to appear in your local jh branch after a pull, nothing
else.
if jh/mtree tracks master (as it according to the wiki page is supposed
to do) and when jh tracks remote branch jh/mtree, then according to
normal logic it follows that also jh tracks master.
2:45 p.m.
Alex Hermann writes:
You'd first have to pull master in jh/mtree, push
that out to origin,
and then pull that into local jh.
alex,
what would that mean in terms of eg or git commands?
in 'jh', i tried 'eg pull --branch master', but it failed miserably:
$ eg pull --branch master
jh(a)git.sip-router.org's password:
*
branch master -> FETCH_HEAD
CONFLICT (directory/file): There is a directory with name pkg/kamailio/gentoo in HEAD.
Adding pkg/kamailio/gentoo as
pkg/kamailio/gentoo~5ecd7deddc7a6fb23f8d0bb530be3e68672c495d
Removing pkg/kamailio/gentoo/kamailio-1.2.0.ebuild
Removing pkg/kamailio/gentoo/kamailio.init
Removing pkg/ser/gentoo/ser-0.8.10.ebuild
Removing pkg/ser/gentoo/ser-0.8.11.ebuild
Removing pkg/ser/gentoo/ser-0.8.12.ebuild
Removing pkg/ser/gentoo/ser-0.8.14.ebuild
Removing pkg/ser/gentoo/ser-0.8.9.ebuild
Removing pkg/ser/gentoo/ser.init
Automatic merge failed; fix conflicts and then commit the result.
-- juha
2:53 p.m.
On Thursday 08 September 2011, Juha Heinanen wrote:
Alex Hermann writes:
I think something like this:
with jh/mtree tracking master (local)
and jh tracking origin/jh/mtree (remote)
eg switch jh/mtree
eg pull --branch master
eg push
eg switch jh
eg pull
You'd first have to pull master in jh/mtree,
push that out to origin,
and then pull that into local jh.
alex,
what would that mean in terms of eg or git commands? in 'jh', i tried 'eg pull --branch
master', but it failed miserably:
$ eg pull --branch master
Automatic merge failed; fix conflicts and then commit
the result.
I think this is doing the right thing, but stumbles upon conflicting changes.
--
Greetings,
Alex Hermann
3:11 p.m.
Alex Hermann writes:
me too, but question is why there are conflicting changes? i have not
pushed any changes to jh/mtree that would conflict with something done by
others.
-- juha
I think something like this:
with jh/mtree tracking master (local)
and jh tracking origin/jh/mtree (remote)
eg switch jh/mtree
eg pull --branch master
eg push
eg switch jh
eg pull
eg does not like switching to jh/mtree, but suggest that a local branch
is created that tracks jh/mtree. that is why i first did:
git checkout --track -b jh remotes/origin/jh/mtree
Branch jh set up to track remote branch jh/mtree from origin.
Switched to a new branch 'jh'
$ eg pull
--branch master
Automatic merge failed; fix conflicts and then
commit the result.
I think this is doing the right thing, but stumbles upon conflicting
changes.
3:17 p.m.
is this conflict due to some git push done to master recently:
$ eg pull --branch master
jh(a)git.sip-router.org's password:
*
branch master -> FETCH_HEAD
CONFLICT (directory/file): There is a directory with name pkg/kamailio/gentoo in HEAD.
Adding pkg/kamailio/gentoo as
pkg/kamailio/gentoo~5ecd7deddc7a6fb23f8d0bb530be3e68672c495d
Removing pkg/kamailio/gentoo/kamailio-1.2.0.ebuild
Removing pkg/kamailio/gentoo/kamailio.init
Removing pkg/ser/gentoo/ser-0.8.10.ebuild
Removing pkg/ser/gentoo/ser-0.8.11.ebuild
Removing pkg/ser/gentoo/ser-0.8.12.ebuild
Removing pkg/ser/gentoo/ser-0.8.14.ebuild
Removing pkg/ser/gentoo/ser-0.8.9.ebuild
Removing pkg/ser/gentoo/ser.init
Automatic merge failed; fix conflicts and then commit the result.
if so, how to fix the conflict?
-- juha
3:21 p.m.
Juha Heinanen writes:
is this conflict due to some git push done to master
recently:
looks like the conflict is this:
$ eg log --merge
commit a89e2b1a2f71c932f38c0bfdcceb2fa1880431f5
Author: Henning Westerholt <henning.westerholt(a)1und1.de>
Date: Mon Sep 5 21:50:23 2011 +0200
pgk: add symlink to kamailio directory for new gentoo ebuild
commit d7b0e82c14caa1809bb6f0d96646354eabf2c253
Author: Henning Westerholt <henning.westerholt(a)1und1.de>
Date: Mon Sep 5 18:17:40 2011 +0200
pkg: remove obselete gentoo ebuild, add new generic ebuild from Claudio Fur
henning, is there anything you can do about it?
-- juha
4:10 p.m.
On Thursday 08 September 2011, Juha Heinanen wrote:
Juha Heinanen writes:
is this conflict due to some git push done to
master recently:
looks like the conflict is this:
$ eg log --merge
commit a89e2b1a2f71c932f38c0bfdcceb2fa1880431f5
[..]
commit d7b0e82c14caa1809bb6f0d96646354eabf2c253
Author: Henning Westerholt <henning.westerholt(a)1und1.de>
Date: Mon Sep 5 18:17:40 2011 +0200 henning, is there anything you can do about it?
Hi Juha,
I also saw this during my cherry-pick for the 3.1 branch, i don't know the
reason. I not did something special, just a bunch of git rm and git mv
commands, but maybe git has some difficulties with this?
There is a gui 'git mergetool', but its also possible to do this by hand with
'git rm' and 'git add'. If you run 'git status' it should show you
the files/
directories in question. I guess in your case
'git rm pkg/kamailio/gentoo -r' should help to resolve the conflict.
Best regards,
Henning
5:16 p.m.
On Thursday 08 September 2011, Juha Heinanen wrote:
Henning Westerholt writes:
Hi Juha,
yes, its strange. Not sure if had some local changes in the 3.1 tree and
therefore git was not able to remove that directory completely, but this still
would not explain your problem.
Best regards,
Henning
'git rm pkg/kamailio/gentoo -r' should
help to resolve the conflict.
henning,
that removed the conflict, although i don't understand how the conflict
was possible in the first place, because i have never touched that
directory in any branch. 
1:27 p.m.
On 08.09.2011 10:16, Juha Heinanen wrote:
a week ago i created a new head jh/mtree, which is
supposed to track
master:
git checkout --track -b mtree master
git push origin mtree:jh/mtree
now when i browse that head on gitweb page, i don't see any changes done
to master, such as the one i did today on lib/srdb1/schema/entities.xml.
Do you possibly need -u (--set-upstream) when pushing your branch to the
repo?
http://stackoverflow.com/questions/520650/how-do-you-make-an-existing-git-b…
HTH,
--Timo
12:42 p.m.
Timo Reimann writes:
Do you possibly need -u (--set-upstream) when pushing
your branch to the
repo?
i don't know about that. when i created jh/mtree, i followed
instructions on sr wiki page:
http://sip-router.org/wiki/git/crash-course
if those instructions are invalid, i would appreciate if someone
familiar with git would fix them.
-- juha
4814
days inactive
4814
days old
15 comments
4 participants
participants (4)
-
Alex Hermann -
Henning Westerholt -
Juha Heinanen -
Timo Reimann