On Sun, Mar 2, 2014 at 2:44 PM, Mike Gilbert <[email protected]> wrote:
> On Sun, Mar 2, 2014 at 1:09 PM, Jeroen Roovers <[email protected]> wrote:
>> Honestly, setting up a tracker and blocking it with bugs about packages
>> which someones-sub-SLOT-checking-script has vetted to be involved could
>> be done in less than a day (for the hundred or so packages that depend
>> on dev-libs/libgcrypt). It doesn't need some QA team to study in depth
>> -- it needs a couple of volunteers to do the checks and file the bug
>> reports.
>>
>
> Unless I have missed mgorny's point here, this isn't just about
> libraries that have currently subslots. This is about every single
> library in the tree.

Correct - my script will spot packages that depend on libraries that
are set up with subslots, but which don't contain a slot operator dep.

It will not spot libraries that do not use subslots.

However, I'm not sure why you'd need a tracker per library for that.
Adding a subslot to a library impacts only that library.  Once that is
done then the script I wrote will detect all the other packages which
should be modified to go along with it.  You can add a subslot to a
library without adding a corresponding slot operator dep to every
package that uses it.

I think it makes more sense to add subslots to libs and not revbump
them, vs not doing anything for 18 months because it is too much work.
 Sure, the full benefit won't come up-front without all the work, but
you'll at least get some benefit vs not ever getting any.  The perfect
is the enemy of the good...

Rich

Reply via email to