| 1 |
On 01/20/2015 12:13 AM, Alexis Ballier wrote: |
| 2 |
> On Mon, 19 Jan 2015 20:31:45 +0100 |
| 3 |
> Michał Górny <mgorny@g.o> wrote: |
| 4 |
>> 2. Subslots work correctly. Rebuilds are forced when the chosen |
| 5 |
>> library is upgraded. Moreover, USE flag change causes a rebuild when |
| 6 |
>> user decides to change the ffmpeg provider. |
| 7 |
> |
| 8 |
> |
| 9 |
> No offense, but this argument is complete crap. You should rather fix |
| 10 |
> portage bugs than propose to introduce tree-wide changes to hide them... |
| 11 |
> More precisely: || ( a:= b c:= d ) is perfectly defined (in the "what |
| 12 |
> it means" sense, not in PMS sense). When the package is built, if 'a' is |
| 13 |
> satisfied then a (and its subslot) is added to the subslot list of the |
| 14 |
> package; ditto for c. You end up with a list of subslot deps, that you |
| 15 |
> can store in vdb or whatever, and use that to decide when to rebuild |
| 16 |
> the package. |
| 17 |
|
| 18 |
That's an interesting proposal, but I immediately find myself |
| 19 |
questioning how closely it models reality. For example, maybe the |
| 20 |
package links to both the a:= package and c:= package, or maybe just to |
| 21 |
one of them. Shouldn't our model match reality as closely as possible, |
| 22 |
as long as it's practical? |
| 23 |
-- |
| 24 |
Thanks, |
| 25 |
Zac |
Archives