- From: Paul Prescod <papresco@calum.csclub.uwaterloo.ca>
- Date: Sat, 10 Aug 1996 09:24:15 -0400
- To: "Keith M. Corbett" <kmc@harlequin.com>, Gavin Nicol <gtn@ebt.com>
- Cc: marc@ckm.ucsf.edu, www-style@w3.org, www-html@w3.org
At 06:06 PM 8/9/96 -0400, Keith M. Corbett wrote: >The only conceivable reason to prefer the attribute syntax to distinguish >elements is to work around the limitations of user agents that cannot (or >will not) allow some other means of extensibility, i.e. It is important to recognize that those user agents were a) designed according to the HTML specification of their day. b) in MASSIVE distribution. We're not talking about just Netscape and Microsoft, but Lynx, Lycos and everything in between. Even if all of the browsers were updated to support arbitrary GI's and CSS, we have only solved part of the problem. If I invent a new GI that I call "MyLink", how does a browser know to interpret that as if it were an <A> or <LINK>? We need a subclassing mechanism. Hopefully the SGML working group will allow this through architectural forms and DTDs. But in the meantime, we must stick with: <A CLASS="MyLink"> or <MYLINK HTMLFORM="A">. so that the semantics of the new element can be inferred. The former is compatible with the existing user agent base (and thus preferable). The latter is not. Paul Prescod
Received on Saturday, 10 August 1996 09:24:45 UTC