| 1 |
On 01/20/2015 01:11 AM, Alexis Ballier wrote: |
| 2 |
> On Tue, 20 Jan 2015 01:01:41 -0800 |
| 3 |
> Zac Medico <zmedico@g.o> wrote: |
| 4 |
> |
| 5 |
>> On 01/20/2015 12:13 AM, Alexis Ballier wrote: |
| 6 |
>>> On Mon, 19 Jan 2015 20:31:45 +0100 |
| 7 |
>>> Michał Górny <mgorny@g.o> wrote: |
| 8 |
>>>> 2. Subslots work correctly. Rebuilds are forced when the chosen |
| 9 |
>>>> library is upgraded. Moreover, USE flag change causes a rebuild |
| 10 |
>>>> when user decides to change the ffmpeg provider. |
| 11 |
>>> |
| 12 |
>>> |
| 13 |
>>> No offense, but this argument is complete crap. You should rather |
| 14 |
>>> fix portage bugs than propose to introduce tree-wide changes to |
| 15 |
>>> hide them... More precisely: || ( a:= b c:= d ) is perfectly |
| 16 |
>>> defined (in the "what it means" sense, not in PMS sense). When the |
| 17 |
>>> package is built, if 'a' is satisfied then a (and its subslot) is |
| 18 |
>>> added to the subslot list of the package; ditto for c. You end up |
| 19 |
>>> with a list of subslot deps, that you can store in vdb or whatever, |
| 20 |
>>> and use that to decide when to rebuild the package. |
| 21 |
>> |
| 22 |
>> That's an interesting proposal, but I immediately find myself |
| 23 |
>> questioning how closely it models reality. For example, maybe the |
| 24 |
>> package links to both the a:= package and c:= package, or maybe just |
| 25 |
>> to one of them. Shouldn't our model match reality as closely as |
| 26 |
>> possible, as long as it's practical? |
| 27 |
> |
| 28 |
> Do you have any such example ? |
| 29 |
|
| 30 |
Well, I think this demonstrates the sorts of ambiguities that may arise |
| 31 |
when using || deps to model reality. We may find that replacing || deps |
| 32 |
with USE conditionals and REQUIRED_USE constraints gives us a more |
| 33 |
accurate and practical model for some kinds of dependencies. |
| 34 |
|
| 35 |
> I think we can only make the safest assumption. Even without subslot, |
| 36 |
> if you consider this: || ( a b c d ), with a and c installed but |
| 37 |
> package automagically deciding to use only a, how can a PM decide |
| 38 |
> whether it is safe to remove a or not after the package has been |
| 39 |
> merged ? |
| 40 |
|
| 41 |
Right, this demonstrates that || deps are ambiguous. So, maybe we should |
| 42 |
look at alternatives that are not so ambiguous, such as USE conditionals |
| 43 |
and REQUIRED_USE constraints. |
| 44 |
-- |
| 45 |
Thanks, |
| 46 |
Zac |
Archives