| 1 |
>>>>> On Thu, 4 May 2017, Jeroen Roovers wrote: |
| 2 |
|
| 3 |
> Atom Prefix Operators [> >= = <= <] |
| 4 |
> Sometimes you want to be able to depend on general |
| 5 |
> versions rather than specifying exact versions all the time. |
| 6 |
> Hence we provide standard boolean operators: |
| 7 |
|
| 8 |
> Examples: |
| 9 |
> >media-libs/libgd-1.6 |
| 10 |
> >=media-libs/libgd-1.6 |
| 11 |
> =media-libs/libgd-1.6 |
| 12 |
> <=media-libs/libgd-1.6 |
| 13 |
> <media-libs/libgd-1.6 |
| 14 |
|
| 15 |
Irrelevant here. The package list in bugs is not a dependency |
| 16 |
specification, but just a list of qualified package names and |
| 17 |
versions. |
| 18 |
|
| 19 |
> PMS does not define what a valid atom looks like, except that it |
| 20 |
> somehow doesn't describe atoms to begin with, apparently because |
| 21 |
> that has yet to be replaced with something better, or is perhaps up |
| 22 |
> to the PM to define. [...] |
| 23 |
|
| 24 |
PMS got rid of the ambiguous "atom" term 10 years ago, in favour of |
| 25 |
"package dependency specification": https://bugs.gentoo.org/174322 |
| 26 |
That makes it much more clear where it is supposed to be used, namely |
| 27 |
in dependencies. (Imagine that! :-) |
| 28 |
|
| 29 |
> PS: It might be a week before I feel like reading the "don't touch |
| 30 |
> my stuff" cabal again, so please don't mind if I happen to touch |
| 31 |
> your stuff while you work out what's wrong with your attitude. |
| 32 |
|
| 33 |
Not going to comment on this. |
| 34 |
|
| 35 |
Ulrich |
Archives