1 |
On Mon, Sep 03, 2012 at 02:08:58PM +0000, Mark Bateman wrote: |
2 |
> Patrick Lauer <patrick <at> gentoo.org> writes: |
3 |
> |
4 |
> > |
5 |
> > On 06/23/12 21:21, Ciaran McCreesh wrote: |
6 |
> > > There's been a move towards using slots for "clever" things that don't |
7 |
> > > fit the traditional way of how slots worked. Examples include the new |
8 |
> > > gtk2 / gtk3 handling and Ruby gems virtuals. |
9 |
> > > |
10 |
> > > Aside from being abusive, |
11 |
> > No, it solves a real problem. |
12 |
> > > this screws things up for Paludis users. |
13 |
> > -EDONTCARE, use a supported package manager |
14 |
> |
15 |
> |
16 |
> |
17 |
> So if the packagemanager is struggling to resolve whether a package belongs in a |
18 |
> slot or not, would this be a case for encoding such metadata in the ebuild |
19 |
> filename. |
20 |
> |
21 |
> foo-slot2-3.2.1.ebuild |
22 |
> |
23 |
> This way the PM would be able to determine exactly what it has todo before it |
24 |
> starts to parse the ebuild |
25 |
|
26 |
No; the problem isn't getting the slot out of the metadata (moving it |
27 |
to the file name doesn't really do anything beyond make it slightly |
28 |
faster at the cost of being backwards incompatible for any existent PM |
29 |
that sees it); the problem is in the PMs resolver, and how it chooses |
30 |
to search the space. |
31 |
|
32 |
Basically, paludis does x, the rest do y; funky-slots was intended to |
33 |
make 'x' behave better at the cost of ebuild devs having to go mark |
34 |
shit up (leading to the retort 'do y instead'). |
35 |
|
36 |
~harring |