On 08/15/11 06:13, Sandro Santilli wrote:
(just back from my mountaineering trip...)
> In order to reduce the impact of the hwaccel merge I'd
> like to propose the following procedure:
Why do you assume there will be any special impact of this branch
merge anyway ? I would only agree to this if this was applied to all
merging of branches by all developers, which I doubt you would agree to
either.
> 0. Reduce changes between hwaccel and master to those strictly
> related to hw acceleration. Commit to master or drop any
> unrelated change, as appropriate.
> 1. Create an hwaccel category on the bug tracker
> 2. Call for testing of the hwaccel branch
> 3. Give testers one week of time to test / file bugs
> 4. Fix the bugs in the hwaccel branch
> 5. Iterate between 3 and 4 till one week goes w/out new tickets
> 6. (ideally) Rebase non-master-merge changes to master
> 7. Merge hwaccel to master
>
> I'll be happy to do testing in step 3.
Almost nobody bothers to test branches or releases, so this is
entirely unrealistic. It's also too slow and inefficient, which would
make the merge never happen. I've been doing testing with a wide variety
of configurations and platform, that will suffice. Anyone that wants to
test the branch can do so anytime they want. Any bugs post merge will
get fixed like any other bugs, no big deal...
- rob -
_______________________________________________
Gnash-dev mailing list
Gnash-dev@gnu.org
https://lists.gnu.org/mailman/listinfo/gnash-dev
