1 |
Am 17.12.2010 17:37, schrieb Ciaran McCreesh: |
2 |
> On Fri, 17 Dec 2010 17:27:05 +0100 |
3 |
> Sebastian Luther <SebastianLuther@×××.de> wrote: |
4 |
>> Am 17.12.2010 16:25, schrieb Ciaran McCreesh: |
5 |
>>> So would anyone be especially opposed to making "best leftmost" an |
6 |
>>> explicit requirement, enforced by repoman where possible (at least |
7 |
>>> for the >= / < case)? |
8 |
>> |
9 |
>> Why can't the PM handle >= / < cases itself? |
10 |
> |
11 |
> Because things are almost never as simple as 'just' >= / <. You can add |
12 |
> in clever trickery to deal with very specific cases, but the second |
13 |
> someone throws things off by adding in a use dependency or a third |
14 |
> package, things get weird. |
15 |
|
16 |
I thought we were talking about the simplest case here, that is a list |
17 |
of atoms for the same cat/pkg. |
18 |
|
19 |
> |
20 |
> Consider a variation on the original case: || ( <a-2 >=a-2[x] ) where |
21 |
> the user has specified -x for a. What should happen then? |
22 |
> |
23 |
> What about || ( <a-2[x] b >=a-2 ) ? Should that be rewritten in the same |
24 |
> way? |
25 |
> |
26 |
> What about || ( <a-2[x] ( >=a-2 b ) ) ? Should the package mangler be |
27 |
> clever enough to figure that one out too? What if b isn't already |
28 |
> installed there? |
29 |
|
30 |
What would repoman enforce here? |
31 |
|
32 |
> |
33 |
> Which is really the problem: clever heuristics get extremely |
34 |
> complicated very quickly, and they're never enough. |
35 |
> |
36 |
|
37 |
Agreed. |
38 |
|
39 |
|
40 |
|
41 |
Sebastian |