On Jan 23, 2011, at 8:28 PM, pbreit wrote: > Is there any chance that is correct behavior?
"Correct in what sense?" is the necessary question. Trailing slashes on URLs tend to be ignored. For args, it could be used as a means of asking for empty strings, but it could get pretty confusing. Do we want to interpret URLs like http://a/c/f////something// ? We could, of course, but it does seem confusing to insist that http://a/c/f is different from http://a/c/f/ and different still from http://a/c/f//, etc. I don't have a strong opinion, since one could argue that the ability to have empty strings in args is a good thing. I'm not sure that either version of rewriting would do the right thing, but we could write some unit tests to find out. The important thing is to define what we want and be consistent about it.
