Re: HDML

Tim Hyland (hyland@uplanet.com)
Wed, 21 May 1997 16:56:05 -0700


Message-Id: <3.0.32.19970521165603.00cf0eac@vortex.uplanet.com>
Date: Wed, 21 May 1997 16:56:05 -0700
To: Paul Prescod <papresco@calum.csclub.uwaterloo.ca>, www-html@w3.org
From: Tim Hyland <hyland@uplanet.com>
Subject: Re: HDML

Paul, thanks for your message.  Your comments are valid and timely.  I have
included replies below to your individual points. 

At 09:03 AM 5/14/97 -0400, Paul Prescod wrote:
>HDML is described at
>http://www.w3.org/pub/WWW/Submission/1997/5/Overview.html
>
>HDML may well be a good idea. I do not know that market and do not know.
>The current specification is a good starting point but before it is
>finalized some major things must be changed:
>
>#1. It must be recast as XML. The authors have doubts as to whether XML
>support the features HDML needs. I have no doubts: it can. XML should be
>the basis for all new non-HTML W3C markup languages, and it was in fact
>designed *precisely* for domain-specific markup languages for the Web.

Jon Bosak, chair of the W3C's SGML work group, has offered to request an
evaluation from the experts involved with the definition of XML.  We hope
that HDML is found to be definable as XML, but our concern is that some
non-display elements of HDML may not work as XML.   

>#2. It must be renamed, reoriented and the changes the new name an
>orientation imply must ripple through the entire document. The days of
>designing markup languages for particular devices is *over*. If we
>accept this proposal we will soon have "webtv markup language", "large
>monitor markup language" and other device-specific languages. The
>authors admit that HDML may be useful in more generalized settings for
>some important classes of data. For instance they quite often mention
>stock quotes and weather reports. Well why wouldn't I use HDML to
>display those in a little corner of my 14" monitor, or on my TV with its
>brutal resolution??
>
>What HDML is *really* about is card-oriented transactional user
>interfaces. Therefore it is best called the transaction markup language,
>or the card markup language or some such. The language should be
>reorganized in that light. Rethinking it in this way will make the other
>applications more obvious and produce a better language that recognizes
>the needs of those other applications. The resulting language will be
>less device-oriented, more powerful and this will actually result in
>MORE USAGE which will benefit handheld devices.

This is a very good point.  I must admit that our focus on hand-held
devices has colored (even limited) the definition and specification.  

>There should thus be a comprehensive review of the needs of other
>communities and a review BY members of other communities -- e.g.
>PointCast and the PUSH crew. Once this is done it may be found that all
>that is necessary is a name change (I would be surprised, but it may
>happen). If so the name should be changed, the language should be
>brought into conformance with XML and the revised language should be
>resubmitted.

We are in the process of setting up a mailing list to discuss these issues.
 Not sure if it will be maintained by W3C or independently, since this is
not a formal W3C activity.  If anyone wishes to participate in that list,
let me know.

>One other issue: I do not understand why Unwired Planet wants to
>maintain copyright ownership over this specification.

This was an oversight and will be corrected on the submission request
template.

> Paul Prescod

-- Tim



Tim Hyland
Product Manager              E-mail:  hyland@uplanet.com
Unwired Planet               Phone:   +1.415.596.5289