1 |
On Sun, 12 Jun 2011 16:49:19 +0200 |
2 |
Ulrich Mueller <ulm@g.o> wrote: |
3 |
> >> It would be more accurate to say that it's guaranteed except for |
4 |
> >> cases in which circular dependencies make it impossible to |
5 |
> >> guarantee. |
6 |
> |
7 |
> > This or other words, this makes this an unreliable feature. So, |
8 |
> > right now, users can't assume anything specific about dependencies |
9 |
> > being installed in pkg_*? |
10 |
> |
11 |
> Currently, PMS only says: "These must be installed and usable before |
12 |
> the ebuild is treated as usable." It doesn't say anything about |
13 |
> phases. |
14 |
|
15 |
The intention with the "usable" stuff is this that purely RDEPEND |
16 |
cycles are resolvable, but any such cycles must be resolved before any |
17 |
package which has a DEPEND upon anything in the cycle is resolved. So |
18 |
if you've got this: |
19 |
|
20 |
first <-- rdepend --- second <-- depend --- third |
21 |
--- rdepend --> |
22 |
|
23 |
Then (first, second, third) and (second, first, third) are the only |
24 |
legal orderings. But if either RDEPEND became a DEPEND (and if we're |
25 |
not dealing with binary packages) then there would be no legal ordering. |
26 |
|
27 |
Whether or not what PMS says matches the intention is another matter... |
28 |
|
29 |
-- |
30 |
Ciaran McCreesh |