can we decide on ap_r* finally?

[rb...@covalent.net]

So far, there have been multiple conversations about how to finish the
ap_r* patch, but nothing has come of it.  I have a series of patches I
REALLY want to apply, but there is some disagreement on the best way to
proceed.

There are basically two options for how we proceed.

1)  Fix OLD_WRITE_FILTER so that it uses the ap_f* functions.
2)  Remove OLD_WRITE_FILTER, and replace the ap_r* functions with macros
that map to ap_f*.

Each model has advantages and disadvantages that have been covered in
great depth. I will cover them again if anybody would like me to (as
impartially as I can).  We have  heard from Greg, OtherBill, and myself,
and have two votes for #2, and one for #1.

Would some other people review this and please send opinions/votes.  As I
see it we are in a stalemate, which doesn't do any of us any good.  I have
a couple of ideas as to how to break the stalemate.  This message is the
first.  If this doesn't work, then I would suggest that we come up with
some objective measurements, such as performance, and make the decision
based on those measurements.

Ryan
_______________________________________________________________________________
Ryan Bloom                        	rbb@apache.org
406 29th St.
San Francisco, CA 94131
-------------------------------------------------------------------------------