| 1 |
Nikos Chantziaras <realnc@×××××.com> wrote: |
| 2 |
> It didn't look like it to me at all. From the man page: |
| 3 |
> |
| 4 |
> Examples: |
| 5 |
> |
| 6 |
> # match anything with a version containing 9999, which can be used in |
| 7 |
> # package.mask to prevent emerge --autounmask from selecting live |
| 8 |
> # ebuilds |
| 9 |
> =*/*-*9999* |
| 10 |
|
| 11 |
This has always been a horribly misleading example, because it |
| 12 |
does not follow the PMS rules of wildcard version matching: |
| 13 |
=foo/bar-9999* must mach foo/bar-9999p1 but not foo/bar-9999999 |
| 14 |
|
| 15 |
Maybe portage finally dumped this bad case, which was already reported |
| 16 |
https://bugs.gentoo.org/show_bug.cgi?id=572178 |
| 17 |
|
| 18 |
OTOH, it would appear reasonable that portage could accept |
| 19 |
=dev-qt/*-5.6* |
| 20 |
although there is of course the ambiguity that it is not clear |
| 21 |
that the -5.6* is the version specification. |
| 22 |
Perhaps one could open a portage bug/feature request about it. |
| 23 |
|
| 24 |
eix-0.31.7 accepts both (=dev-qt/*-5.6* with PMS rules |
| 25 |
as well as the non-PMS =dev-qt/*-*5.6* [with string instead |
| 26 |
of PMS matching rules]). |
Archives