Re: 12-21-95 INSERT draft

Foteos Macrides (MACRIDES@sci.wfbr.edu)
Sat, 20 Jan 1996 10:26:59 -0500 (EST)


Date: Sat, 20 Jan 1996 10:26:59 -0500 (EST)
From: Foteos Macrides <MACRIDES@sci.wfbr.edu>
Subject: Re: 12-21-95 INSERT draft
To: brian@organic.com
Cc: www-html@w3.org
Message-Id: <01I089UAK4AA0041EW@SCI.WFBR.EDU>

Brian Behlendorf <brian@organic.com> wrote:
>On Fri, 19 Jan 1996, Foteos Macrides wrote:
>> 	What this means, and really does need to be addressed explicitly,
>> is that clients which took the HTML 3.0 draft "seriously", implemented
>> it's DTD, including FIG, have been "left in the lurch" since last September,
>> and as of the INSERT draft release last month face the problem of deciding
>> with they also should ignore W3C drafts, and retain their implementations
>> of FIG as in the too long expired HTML 3.0 draft.  ;(  ;(  ;(
>
>This is the price of experimentation, and must be a price that ALL 
>browser authors MUST be willing to accept, as long as something is 
>classified "experimental".  You can keep on supporting it, sure, just as 
>Microsoft will probably always support BGSOUND and Netscape will 
>probably always support BLINK.  
>
>My $.02.

	I fear you missed the point, so at risk of beating a dead
horse, let me try to make it again.

	There is no W3C or IETF RFC or prelimnary draft whose DTD
breaks BGSOUND or BLINK.  BGSOUND, for example, is technically a
plug-in, and its equivalent, or more elaborate uses of audio,
can be achieved via INSERT -- but nothing in the INSERT draft
precludes any client from using the "historical" implementation
of BGSOUND.

	The plug-in tag originally was to be EMBED, but treated
as a container.  This would cause bothersome but not
insurmountable conficts with Netscape's empty implementation of
EMBED.  The change of the plug-in tag name to INSERT allows any
client to use the "historical" implementation of EMBED without
conflict or inconvenience.

	The INSERT draft is, in large part, a generalization and
embellishement of the APPLET markup -- but nothing in the INSERT
draft would cause a conflict for clients supporting the
"historical" markup for APPLET.

	In contrast, the DTD for FIG in the INSERT draft
eliminates the ability to use it as a client-side image map
handler which AUTOMATICALLY (of necessity) offers alternative
access to the actual information by, as Joe Meadows put it,
"everyone (be they blind, modem speed impaired, or whatever)."
In the INSERT draft, FIG is being used to specify blocking of
the plug-in, and as a vector for including CAPTION and / or
CREDIT --- BOTH of which could be achieved by other means,
WITHOUT breaking FIG as a client-side image map handler that
enforces automatic alternative access to the information.
(We're talking about enforced alternative access to the SAME
links that would be activated via the image map, not just to
a "Get foo now." advertisement for a client).

	This is the first time, ever, in the history of WWW
development, that the principle of "backward compatibility" has
not been respected.  It is a precedent -- specific to FIG in
the INSERT draft.

	I suspect that the bad link to the MAP proposal was
a form of Freudian slip by the in-house W3C editor for the
INSERT draft. -- It was his creativity which brought us FIG.

			 Fote

=========================================================================
 Foteos Macrides            Worcester Foundation for Biomedical Research
 MACRIDES@SCI.WFBR.EDU         222 Maple Avenue, Shrewsbury, MA 01545
=========================================================================