Re: ismap functionality...

Tony Sanders <sanders@BSDI.COM>
Message-id: <199312302246.QAA18175@austin.BSDI.COM>
Subject: Re: ismap functionality... 
In-Reply-To: Joe Germuska's message of Thu, 30 Dec 1993 15:47:01 CST.
Organization: Berkeley Software Design, Inc.
Date: Thu, 30 Dec 1993 16:46:14 -0600
From: Tony Sanders <sanders@BSDI.COM>
Content-Length: 482
> So is this idea counter to philosophy, or too hard for client authors to
> implement? :-) (my suggestions usually violate philosophies :-)

The problem is that client side decoding is inflexible.  You have to define
a spec and stick with it.  Witness <FIGA>, it only supports polygons.  As
soon as someone implements it someone else will want some other feature
and then we will have incompatible versions floating around.  This doesn't
happen with the current scheme.