On Oct 25, 2007, at 10:33 AM, Nick Kew wrote:
On Thu, 25 Oct 2007 16:00:40 +0200
Plüm, Rüdiger, VF-Group <[EMAIL PROTECTED]> wrote:
However, on further reflection, I think it would be better to
fall back to old behaviour in this case. The difficulty is that
this redirection catches cases where something is unnecessarily
but perfectly legally URLencoded. So a perverse client might
re-encode the redirection, leaving us an infinite loop.
What does "old behaviour" mean? Using r->uri in this case?
Exactly.
Any objections?
Conceptionally, no objections here.
