ad hoc sub "rm-cache"

[finanalyst]

I found the sub rm-cache was added.
Is there a need for rm-cache, other than in the tests?
If yes, then it should be re-written as a method on the class. Otherwise we have a single ad-hoc sub.

In other words, if I remove rm-cache from Pod-To-Cached, is it going to affect a downstream module.

It was not in the original API and it is not documented.

[JJ]

It was added to avoid the external dependency with rmtree, which seems kind of abandoned

[finanalyst]

I use File::Directory::Tree extensively in my tests.
It is used in all the other tests. I added it and removed rm-cache

[finanalyst]

Also File::DIrectory::Tree is mentioned in perl6 documentation explicitly, see rmdir.
So, if the module has been abandoned, that's a big thing, I think. Not just for this module.

[finanalyst]

I've removed rm-cache and replaced with File::Directory::Tree rmtree in the tests.
If this is a major concern, we can revert the change.

[JJ]

rm-cache adds a functionality, encapsulating the directory where the cache is actually hosted. Even if you want to use File::Directory::Tree in rm-cache, rm-cache should stay. Once a functionality is added, you really need a good reason to eliminate it, and if you do it, use deprecation to eliminate it gracefully. So please put it back there.

[JJ]

If you don't want to use it in the tests, that's fine. But I don't see a good reason for removing a tested and released functionality, even more so when we were still discussing what needed to be done with it.

[labster]

If you need me to do something for File::Directory::Tree, just let me know. No issues are open now so I assume it's all good until I learn otherwise.