Re: CVS update: subversion/subversion/libsvn_fs structure

Greg Hudson <ghudson@MIT.EDU> writes:
> Jim committed:
> > +NOTE: Copy nodes are pretty experimental. I'm not sure they're
> > +necessary at all. So don't take this stuff too seriously.
>
> Don't they also kill your relatedness heuristic?

Yes, you'd have to indirect through them. I hadn't thought of that.

I was going to write something longer, but I might as well say it
briefly now: I can't tell see what advantage copy nodes have over
simply setting a property on the copy that says where it was copied
from. I think this provides all the same information without
complicating the filesystem. It can be done entirely on the client.

(I'm annoyed I didn't notice this when they were first suggested. It
seems pretty obvious now.)
Received on Sat Oct 21 14:36:23 2006