| 1 |
On Tue, Nov 23, 2010 at 08:30:34PM -0500, Mark Loeser wrote: |
| 2 |
> Markos Chandras <hwoarang@g.o> said: |
| 3 |
> > Hi there, |
| 4 |
> > |
| 5 |
> > The official policy for live ebuilds is the following one: |
| 6 |
> > |
| 7 |
> > http://devmanual.gentoo.org/ebuild-writing/functions/src_unpack/cvs-sources/index.html |
| 8 |
> > |
| 9 |
> > I don't quite agree with this policy and I guess most of you don't agree |
| 10 |
> > either looking at the number of live ebuilds/package.mask entries. |
| 11 |
> > |
| 12 |
> > My proposal is to keep empty keywords on live ebuilds without masking |
| 13 |
> > them via package.mask |
| 14 |
> > |
| 15 |
> > Users interpret this as a 'double masking' which in fact it is since |
| 16 |
> > they need to touch two files before they are able to use the package. |
| 17 |
> > |
| 18 |
> > I also know that we can use overlays for that, but distribute the |
| 19 |
> > ebuilds among dev/proj overlays is not always a solution. |
| 20 |
> |
| 21 |
> I'm personally against such a change and would infact like to see all |
| 22 |
> live packages nuked from the tree and moved to some experimental tree. |
| 23 |
> If you move them there, I don't care what policies you apply, but we |
| 24 |
> should try to maintain a solid set of working packages in the main tree, |
| 25 |
> which no one can guarantee with a live ebuild. I know most people |
| 26 |
> aren't going to agree with me, but I felt the need to say it anyway. |
| 27 |
> |
| 28 |
|
| 29 |
Well I think it might be a good idea to have an experimental tree with live ebuilds. |
| 30 |
It could make it far more simple to run an ~arch system and differentiate with what you need in live ebuilds. |
| 31 |
So while I'm not a dev or anything else within Gentoo, I think it's an idea worth considering. |
| 32 |
And after all, it would make this entire debate on what to do with ebuilds void as you would be specifically getting the applications from a different tree. |
| 33 |
|
| 34 |
-- |
| 35 |
Zeerak Waseem |
Archives