| 1 |
On Sat, 19 May 2007 07:00:58 +0400, Kent Fredric <kentfredric@×××××.com> |
| 2 |
wrote: |
| 3 |
|
| 4 |
> ... |
| 5 |
> |
| 6 |
> Imo, the cyclic dep problem could be solved as thus, |
| 7 |
> |
| 8 |
> A depends B |
| 9 |
> B depends C||A |
| 10 |
> |
| 11 |
> Where C is a minimalist subset of A required for building B, which is |
| 12 |
> only depended on if A is not present. |
| 13 |
> A is also a replacement for C. |
| 14 |
> |
| 15 |
> So the flow would go like such. |
| 16 |
> |
| 17 |
> Emerge A: |
| 18 |
> * depends on b |
| 19 |
> * A is missing, so depend on C |
| 20 |
> *emerges C* |
| 21 |
> *emerges B* |
| 22 |
> *removes C* <-- otherwise A & C containing the same files = headache |
| 23 |
> *emerges A* |
| 24 |
> |
| 25 |
> Yes, indeed I agree that we could just do this by hand by changing a |
| 26 |
> USE flag, but we should at least be open to the idea of looking for a |
| 27 |
> way to automatically resolve the problem. Computers exist to make our |
| 28 |
> life easier, not the other way around :) |
| 29 |
> |
| 30 |
|
| 31 |
Just curious: why nobody suggests to allow Portage to use the preferred |
| 32 |
method of binary distros: emerge several interdependent packages in one |
| 33 |
transaction. Just prepare the source for A and B and compile both in any |
| 34 |
order. IMHO this is what the ideal system should do. |
| 35 |
|
| 36 |
-- |
| 37 |
Andrei Gerasimenko |
| 38 |
-- |
| 39 |
gentoo-user@g.o mailing list |
Archives