W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 2000

[media] Designers Suffer as Browser Makers Continue to Lag the W3C's Work

From: Kathleen Anderson <kathleen.anderson@po.state.ct.us>
Date: Fri, 16 Jun 2000 16:09:59 -0400
Message-ID: <002901bfd7ce$d90b6260$e924f79f@STATE.CT.US>
To: "'Web Accessibility Initiative'" <w3c-wai-ig@w3.org>
Read the article at:

Text follows:

InternetWorld In Print  Issue Date: June 15, 2000

I N T E R N E T   T E C H N O L O G Y
Browser Standards
Not On the Same Page
Designers Suffer as Browser Makers Continue to Lag the W3C's Work
By Larry Seltzer

There's never been a tougher time to write a Web page. Web coders, in their
struggle to get users to see a page the way its author meant them to, run
into conflicting implementations that make the job seem hopeless.
Theoretically, if you were to write your code to conform to standards, and
browsers supported that standard, you could rest easy in the knowledge that
users would have no trouble viewing your site.

This utopian notion runs headlong into two real-world problems, however.
First, none of the major browser releases fully supports any of the major
standards. Write your code to simply conform to standards, and you'll
guarantee that no one will see it as you intended. Second, browsers often
implement the features you want using proprietary syntax - Netscape's
proprietary layers model in Navigator 4 is a good example.

The leading organization in standards development is the World Wide Web
Consortium (W3C), which has moved to develop and improve standards with a
speed and independence that are both admirable and curious: The world of
standards has found itself several years ahead of the world of the browsers
that implement them. How relevant is version 3 of Cascading Style Sheets
(CSS3) when no shipping browsers implement CSS1 completely?

There's a boatload of blame to go around for this mess, and the first aboard
is Netscape. From version 1 through version 4, Netscape cavalierly created
new HTML tags, a scripting language, and an object model with no outside
input. The resulting mess is the most important "standard" to which
developers must adhere; the official standards for HTML and JavaScript are
largely documentation of the fait accompli presented in early versions of

Ironically, when Netscape decided to go all-CSS in Gecko, the open source
engine in Netscape 6, it abandoned its proprietary implementation from
Netscape 4.

IE3, Microsoft's first serious attempt at a browser, took a step toward
improving the usefulness of standards by partially implementing CSS1. Since
then, Microsoft has done a far better job of adhering to CSS and other
standards issued by the W3C than Netscape, yet it never seems to implement
any of them completely. The company is on the verge of releasing IE 5.5, but
it has never fully implemented CSS1, let alone any of its successors.

With each major browser release since IE3, Microsoft has told the press that
the next one would fully implement CSS and the Document Object Model (DOM -
an important specification for defining how scripting languages can access
browser features). Back when it had 3 percent of the browser market,
standards worked for Microsoft, but it seems that as IE's market share has
grown, the company has lost interest.

It's not that Microsoft is not working on implementing standards anymore;
when it comes to XML, there are few companies as enthusiastic as Microsoft.
Every copy of IE (and, by extension, Windows) has an XML parser built in,
and they have been adding support for SOAP, an XML-based messaging protocol,
into all their server programs.

When Netscape threw in the towel two and a half years ago and decided to
turn its browser development over to Mozilla.org, an open source group, it
essentially surrendered a good deal of decision-making to that body. Mozilla
has pledged to support HTML4, CSS1, DOM1, and XML (plus some support for
CSS2 and DOM2), but since Netscape 6 is still in its first beta release as
of this writing, it's too early to say whether Mozilla has fulfilled its

Conventional wisdom has it that software on the Web advances at tremendous
speed, but take a look at the work of the W3C and you learn just how
conservative the browser industry really is. It was obvious from the initial
release of CSS1 in late 1996 that it would be important, but to date not a
single browser fully supports it. CSS2 is now more than two years old and
has seen only spotty implementation in real-world browsers.

Browsers might have incorporated more complete CSS support if the W3C hadn't
fumbled the compliance-testing ball. The W3C has a CSS1 test available on
its site, but it is incomplete and difficult to run. There is no CSS2 test
on the site. One of the projects at Mozilla.org has been to build compliance
tests for HTML4, CSS1, DOM1, XML, and client-side JavaScript.

These standards give the coder much better control over the presentation of
a page than HTML alone, although it is in the essence of CSS and HTML that
the coder cannot completely control the presentation. CSS2 adds the ability
to define media-specific styles so that a page will appear properly on the
screen, in print, on a Web-enabled cell phone, or on a Palm Pilot. These
devices may breathe new life into both standards, as well as into the
also-rans of the browser market, such as Opera from Opera Software.

Support of Web standards doesn't free developers from considering the
specific limitations of specific devices, however. Even within computers
there are differences in pixel sizes, and standard Web pages that look great
on PCs may not work at all on other devices, such as TVs or Web phones.

The mission of CSS is in line with much of the rest of the mission of the
W3C: Separate content from presentation. The more we separate the process of
defining data from the definition of how to present it, the easier it will
be to access data from new devices, and the more control users will have
over how they want it presented.

Then there's the W3C's XML specification. XML is a data definition language.
Currently it means little in the context of everyday browsing, but it means
a lot in regard to server processes, and XML's momentum is undeniable.
Numerous other standards from the W3C and elsewhere build on XML as their
data definition system. Internet Explorer has had a built-in XML parser for
some time, and Netscape 6 promises to have one too.

A number of other W3C standards are becoming increasingly important. XSL
(eXtensible Stylesheet Language) is a style sheet system for XML data that
effectively allows XML to be transformed into HTML. Two different XML-based
standards, SMIL and HTML+Time, attempt to enable Web pages to synchronize
graphical and multimedia elements. RDF (Resource Description Format) is an
XML-based generalized system for defining metadata. MathML is an XML-based
mathematics definition language designed to enhance the presentation of
mathematical expressions on the Web and to facilitate machine-to-machine
mathematical communication. Even the bedrock HTML is not immune, as a shadow
is cast over it by XHTML, which is a recasting of HTML under XML rules.

Whatever happens with Netscape 6, it may prove to be just another platform
on which developers must test before sending a page from the staging server
to production. What developers need is fewer targets, not more. Should the
day come when W3C standards are fully and universally supported, Web
developers will finally have a single target. Of course, should the day come
when everyone else gives up and we all use Internet Explorer, the same would
be true. In any case, in the meantime a lot of old code will simply have to
fade away.

Copyright  2000 by Internet World Media, A Penton Media, Inc. Company.

Posted by:
Kathleen Anderson, Webmaster
State of Connecticut, Office of the State Comptroller
55 Elm Street, Room 101, Hartford, Connecticut  06106
voice: (860) 702-3355  fax: (860) 702-3634
email: kathleen.anderson@po.state.ct.us
URL: http://www.osc.state.ct.us
CMAC Access: http://www.cmac.state.ct.us/access
AWARE: http://aware.hwg.org/
Received on Friday, 16 June 2000 16:10:24 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 20:35:56 UTC