1 |
On Sun, Mar 2, 2014 at 2:44 PM, Mike Gilbert <floppym@g.o> wrote: |
2 |
> On Sun, Mar 2, 2014 at 1:09 PM, Jeroen Roovers <jer@g.o> wrote: |
3 |
>> Honestly, setting up a tracker and blocking it with bugs about packages |
4 |
>> which someones-sub-SLOT-checking-script has vetted to be involved could |
5 |
>> be done in less than a day (for the hundred or so packages that depend |
6 |
>> on dev-libs/libgcrypt). It doesn't need some QA team to study in depth |
7 |
>> -- it needs a couple of volunteers to do the checks and file the bug |
8 |
>> reports. |
9 |
>> |
10 |
> |
11 |
> Unless I have missed mgorny's point here, this isn't just about |
12 |
> libraries that have currently subslots. This is about every single |
13 |
> library in the tree. |
14 |
|
15 |
Correct - my script will spot packages that depend on libraries that |
16 |
are set up with subslots, but which don't contain a slot operator dep. |
17 |
|
18 |
It will not spot libraries that do not use subslots. |
19 |
|
20 |
However, I'm not sure why you'd need a tracker per library for that. |
21 |
Adding a subslot to a library impacts only that library. Once that is |
22 |
done then the script I wrote will detect all the other packages which |
23 |
should be modified to go along with it. You can add a subslot to a |
24 |
library without adding a corresponding slot operator dep to every |
25 |
package that uses it. |
26 |
|
27 |
I think it makes more sense to add subslots to libs and not revbump |
28 |
them, vs not doing anything for 18 months because it is too much work. |
29 |
Sure, the full benefit won't come up-front without all the work, but |
30 |
you'll at least get some benefit vs not ever getting any. The perfect |
31 |
is the enemy of the good... |
32 |
|
33 |
Rich |