🚀 go-pugleaf

The arguments here have been invaluable-- I think you all have done a great
job of proving your points.  I have nothing left to argue against these
points (that Megginson letter was pretty damaging to my case).  I concede--
In the PWideChar/PSAXChar model, the start parameter should not be used and
the Pointer should be set to the start of the important characters within
the buffer.

So it  looks like the current count is

For Buffer
==============
Jeff Rafter
Vincent Parrett
Robert Loof
R�mon Sinnema

For WideString/SAXString
==================
Keith Wood

Not really on a side
=============
Mark Edington
Binh Ly

Keith didn't weigh in in the roundup-- but I was under the impression he was
opting for the SAXString earlier.  I think R�mon's point about this being
obscured in the Components and filters written is a safe bet-- there is a
large possibility that the majority of users won't actually implement the
interfaces-- they will use components that do it for them.

Also, there is still the mixed approach (which I have been biased against--
sorry!)  Does anyone want to open some discussion about that possibility?

Good Jobs!

--
Jeff Rafter
Defined Systems
http://www.defined.net
XML Development and Developer Web Hosting