On Sep 19, 2007, at 2:07 PM, James M Snell wrote:
During a discussion on Sam's blog [1] it was pointed out that a
requirement on the the self link was mysteriously dropped between two
draft versions of rfc4287 without any discussion or notice. This
change
was not caught at the time and has led to a situation in which a valid
self link can fail to meet the requirements that led to the
creation of
the self link in the first place. While it's a minor issue that is
easy
to work around using xml:base and software that fully implements
support
for relative uri's, it is somewhat disconcerting that the change was
made in the first place.
Editorial processes are conducted by humans. That's why we, and the
IETF, all had many MANY chances to review the spec. If there were
someone to whom this was important, presumably they would have
noticed it and called it out. -Tim
