I've thought about this a bit. While people who really want to experiment
with attributes in content-type returns *can* go through .asis or .cgi, it's
a kludge and they really shouldn't have to go through those kinds of hoops.
What I *really* don't want is for someone to ask "why did this break?", and
for me to have to tell them that we deliberately broke it in the 0.6.2 ->
0.6.3 transition.
However, if the code is #ifdef'ed, I can tell the guy instead to just compile
with -DW3O_SUCKS, or some such, until such time as the W3O sees fit to
release code which complies with their own standards. (Gee, do I sound
peeved?) So (he asks with trepidation), who has a problem with this?
rst
with attributes in content-type returns *can* go through .asis or .cgi, it's
a kludge and they really shouldn't have to go through those kinds of hoops.
What I *really* don't want is for someone to ask "why did this break?", and
for me to have to tell them that we deliberately broke it in the 0.6.2 ->
0.6.3 transition.
However, if the code is #ifdef'ed, I can tell the guy instead to just compile
with -DW3O_SUCKS, or some such, until such time as the W3O sees fit to
release code which complies with their own standards. (Gee, do I sound
peeved?) So (he asks with trepidation), who has a problem with this?
rst