| 1 |
>>>>> On Sun, 22 Nov 2015, Michał Górny wrote: |
| 2 |
|
| 3 |
> On Sat, 21 Nov 2015 18:49:03 -0500 |
| 4 |
> Ian Stakenvicius <axs@g.o> wrote: |
| 5 |
|
| 6 |
>> All told, I think i'm in favour of banning the variables, and |
| 7 |
>> potentially providing getter functions that would output the path |
| 8 |
>> of these files if they need to be accessed -- '$(get_eclasspath |
| 9 |
>> [name])' or $(get_licensepath [name]) or the like. I don't know if |
| 10 |
>> these could be implemented in-eclass or if they would be something |
| 11 |
>> that would have to be added to EAPI7..? |
| 12 |
|
| 13 |
The more I think about the idea of having a LICENSEDIR containing |
| 14 |
symlinks, the less I like it. It would add the overhead of creating |
| 15 |
the dir to all ebuilds, whereas it would be used in a handful of cases |
| 16 |
only. Also the symlinks won't be permanent, so postinst messages like |
| 17 |
"read the license at ${LICENSEDIR}/GPL-2" cannot be used, unless the |
| 18 |
ebuild would follow the link (e.g. using readlink) which seems clumsy. |
| 19 |
|
| 20 |
So IMO a function like get_license_path is the better solution here. |
| 21 |
|
| 22 |
> And if I understand your idea correctly, get_licensepath would still |
| 23 |
> not work for binary packages. |
| 24 |
|
| 25 |
Is there a scenario where LICENSEDIR would work, but the function |
| 26 |
would not? |
| 27 |
|
| 28 |
Ulrich |
Archives