Proposal: removeDirectoryRecursive should not follow symlinks

[Edward Z. Yang]

Sure, I'm happy for someone to put this forward as a counterproposal.

My reasoning for changing the meaning of the identifier: I'm pretty
sure almost everyone using the current function doesn't actually
want to follow symlinks.  To actually verify this I'd have to do
some Hackage spelunking.

Edward

Excerpts from Greg Weber's message of 2015-01-05 14:33:41 -0800:
> How about being backwards-compatible friendly by adding a new function with
> the friendly behavior, adding a deprecation notice to the existing
> function, and putting the existing function under a new name that signifies
> the -rf? That would put me at a +1
> 
> On Mon, Jan 5, 2015 at 2:31 PM, David Feuer <david.feuer at gmail.com> wrote:
> 
> > +1. Following symlinks in such a potentially-destructive operation is
> > fundamentally wrong.
> >
> > On Mon, Jan 5, 2015 at 5:25 PM, Edward Z. Yang <ezyang at mit.edu> wrote:
> > > Discussion period: one month
> > >
> > > I propose to backwards incompatibly change the behavior of
> > > removeDirectoryRecursive to not follow symlinks.  We could optionally
> > > add a replacement function under a new name which does follow symlinks.
> > >
> > > This would bring its behavior inline with rm -rf.
> > >
> > > I also opened a ticket here:
> > https://ghc.haskell.org/trac/ghc/ticket/9266
> > >
> > > Thanks,
> > > Edward
> > > _______________________________________________
> > > Libraries mailing list
> > > Libraries at haskell.org
> > > http://www.haskell.org/mailman/listinfo/libraries
> > _______________________________________________
> > Libraries mailing list
> > Libraries at haskell.org
> > http://www.haskell.org/mailman/listinfo/libraries
> >