On Sat Jul 25 06:38:34 2015, RICHE wrote:
Show quoted text> From TODO:
>
> Can rmtree initial paths be sanitised before diving into _rmtree(), in
> a way that would reduce platform-specific workarounds therein?
This is worth exploring, but I bet it will be a heavy lift. Apart from explicit testing for VMS, there is implicit OS checking in use of the constants _FORCE_WRITABLE and _NEED_STAT_CHECK.
So I would not make this a high priority. Dealing with, e.g., the threads problem is, IMO, more important.
Thank you very much.
Jim Keenan