1 |
On Tue, Apr 23, 2013 at 3:25 PM, Ian Stakenvicius <axs@g.o> wrote: |
2 |
> Alternatively, we could enforce repoman checks on any commit or push |
3 |
> operation in master, and leave branches to their own devices. Of |
4 |
> course, I haven't seen (or looked for, tbh) how tree development will |
5 |
> be implemented/suggested to be done in git so I've no idea what role |
6 |
> branches might play.. |
7 |
|
8 |
Keep in mind that the following are all roughly equivalent in the world of git: |
9 |
1. The master branch on anything other than the copy of the |
10 |
repository hosted by infra. |
11 |
2. Some other named branch anywhere. |
12 |
3. An overlay. |
13 |
|
14 |
Just because you're committing to master doesn't mean that anybody |
15 |
will ever see it. |
16 |
|
17 |
As far as how things will go - this seems very much TBD. There have |
18 |
been various random thoughts on this list, and in IRC discussions. |
19 |
|
20 |
My two cents: |
21 |
|
22 |
1. For most routine work, just do it all in master, rebasing commits, |
23 |
so that in the end it looks just like the current CVS (well, assuming |
24 |
CVS had cross-file atomic commits). |
25 |
|
26 |
2. For big projects, do the work in an overlay first. Then for |
27 |
staging purposes consider a short-term branch in the main repository |
28 |
with frequent rebasing during any QA checks, then do one big |
29 |
fast-forward-merge at the end. That lets you do big atomic tree-wide |
30 |
changes without blocking commits, and having the final merge in a |
31 |
single commit (think KDE stabilization, big eclass changes, etc). |
32 |
|
33 |
We already have overlays, so I'm not sure we want to end up with lots |
34 |
of long-term branches in the main portage repository. |
35 |
|
36 |
All of this is of course talking about the portage tree. More |
37 |
traditional software development projects like portage (the software), |
38 |
eudev, etc should of course use more traditional use of git. |
39 |
|
40 |
Rich |