- From: Shane P. McCarron <shane@aptest.com>
- Date: Wed, 05 Jul 2000 22:49:41 -0500
- To: Gerald Oskoboiny <gerald@w3.org>
- CC: Paul McGarry <paulm@opentec.com.au>, www-validator@w3.org
Gerald Oskoboiny wrote:
> > Seriously, its conformance indications leave a lot to be desired.
>
> How's that? If it said 'must', it would be incorrect.
>
> I think the HTML 4 spec(s) are exemplary as far as technical
> specifications go.
Speaking as someone who develops formal test suites for a living, I
think that HTML 4 is a huge improvement over HTML 3.2, but is still
terrible from a conformance perspective. It uses phrases like "many
browsers do..." or "often this means...". Those are not phrases you
expect to find in a formal standard. Also, the fact that HTML 4 bent
over backward to accommodate non-visual browsers means that it has
almost no guidance for visual browser creators. This means that they
are free to implement things pretty much however they wish. That is not
really a good way to promote interoperability and application
portability (where, in this case, an application is an HTML 4 conforming
document).
--
Shane P. McCarron phone: +1 763 786-8160
ApTest fax: +1 763 786-8180
mobile: +1 612 799-6942
e-mail: shane@aptest.com
Received on Wednesday, 5 July 2000 23:49:44 UTC