On 3 December 2011 13:59, Phil Holmes
<address@hidden> wrote:
This is the instructions I wrote to myself:
Before you start thinking about pushing a patch to staging, you
need to ensure you have the correct local branches up to date.
One time only, edit the .git/config file to look like this (there
will be other lines and sections, don't touch these):
@example
[remote "origin"]
fetch = +refs/heads/*:refs/remotes/origin/*
url = "" href="http://git.sv.gnu.org/srv/git/lilypond.git" target="_blank">git.sv.gnu.org/srv/git/lilypond.git
@end example
Now, each time you want to make a change and push to staging, do
the following:
@example
git fetch # (to be sure you have the current version of staging)
What is the difference between doing 'git fetch' and 'git pull -r' (which is what I do and lilygit.tcl does I believe)?
Otherwise I pretty much do what Phil does.
I don't work with branches but make my changes, make a patch, upload to Rietveld and the use lilygit to reset --hard and I then simply keep the patch file in a safe place and reapply it when I need to make a change or push it.
That way I work on many small patches by editing/amending and resetting and saving the updated patch file.
I find that much easier to understand conceptually :)