| 1 |
On 04/27/2012 12:57 PM, David Leverton wrote: |
| 2 |
> Zac Medico wrote: |
| 3 |
>> So, here's a description of the whole algorithm that I'd use: |
| 4 |
>> [snip] |
| 5 |
> |
| 6 |
> I think the following is equivalent, but simpler and more general since |
| 7 |
> it doesn't have to mention details like ** and friends that aren't |
| 8 |
> currently in PMS, and doesn't assume that the rule for handling KEYWORDS |
| 9 |
> is simply "does it contain at least one of the accepted values? (plus |
| 10 |
> handling of ** etc)". (For example, I can imagine something like |
| 11 |
> "accept the package if it has amd64, or if it has both ~amd64 and x86" |
| 12 |
> being potentially useful for some people, although I don't think it's |
| 13 |
> implemented anywhere at the moment.) |
| 14 |
> |
| 15 |
> 1) Pretend that all stable keywords in the package's KEYWORDS are |
| 16 |
> replaced with the corresponding ~arch ones |
| 17 |
> 2) If this would result in the package being masked by keywords (I |
| 18 |
> forget the exact terminology Portage uses, but I'm sure you know what I |
| 19 |
> mean), then apply the masks/forces from package.use.stable.* |
| 20 |
|
| 21 |
Yeah, that appears to be equivalent. |
| 22 |
-- |
| 23 |
Thanks, |
| 24 |
Zac |
Archives