[16:58:46] <tdsmith> dstufft: what's your thinking on the `pip cache` PR? you still want to simplify `pip cache remove` to just accept a package name? and how are you feeling about --type?
[16:59:23] <tdsmith> it occurred to me that you could imagine a world where pip2 and pip3 and pypy -m pip see "different" wheel caches and only operate on the wheels they consider installable
[16:59:46] <tdsmith> i'm not sure if that would do more to meet or confound expectations
[17:00:54] <dstufft> tdsmith: I think we probably want to ditch --type yea, it feels a bit off to have a flag that can't be used for 1/2 the commands, and the other 1/2 the benefit of using it is fairly low
[17:02:55] <dstufft> I think simplfying remove to only take package names seems reasonable to me... I can't think of a super compelling reason to support more than that other than trying to only narrowly purge things from the cache, but I'm not sure that small use case is worth the trouble. We can awlays re-add the ability to do that in a future release if we get a bunch of folks asking for it.
[17:04:12] <dstufft> I think in that same vein, it might not make sense to have ``cache list`` because I struggle to think of a real use case for it that isn't error prone or tied to deleting files via cache remove... can you?
[17:04:40] <dstufft> theortically you could use it to see if pip install foobar==1.0 would be cached or not, but in reality you can't because the filename alone isn't enough for that
[17:04:49] <dstufft> you have to know the hash of the URL for the sdist too
[17:06:29] <dstufft> I don't think "different" caches is something we'd want to deal with, in theory it makes sense, but I think it will be confusing, particularly since some files they'll share a cache and some they won't