Goal #1: prevent fear, uncertainty, and doubt:
> most vendors hewed closely to early versions such as HTML
> 2.0, lately Microsoft and Netscape have virtually ignored
> the follow-on 3.0 draft under development by the Internet
> Engineering Task Force.
> "For the moment, like it or not, I feel like I have to
> choose the lowest common denominator of HTML in most
> cases," says Monson.
> "I suspect that when we get to HTML 4.0, Netscape will
> still be doing its own proprietary tags."
Goal #2: plug the good work we're doing!:
> Perhaps recognizing that key vendors had lost confidence
> or interest in its efforts, leaders of the IETF working
> group recently recommended that authority for HTML
> standards shift to the World Wide Web Consortium, which
> boasts heavy support from Microsoft, Netscape and other
> major Web vendors. Some electronic commerce companies,
> including Time Warner, are optimistic that the consortium
> will do a better job of keeping official standards in
> sync with the browser implementations of the key vendors.
> In fact, the consortium recently released a proposed HTML
> 2.0 follow-on, called HTML 3.2, that splits the difference
> between some of the key extensions from Microsoft and
> Netscape.
[But we have to keep the facts straight, too. Reportage like this will only
incense the Klensins... people will assume W3C is pushing this party line that
IETF rolled over for us!]
Goal #3: support/attract new members:
> Jonathan Hirshman, a senior producer in New York with
> Time Warner Inc.'s Pathfinder series of Web sites. "The
> only way we're going to be able to do that is to get some
> agreement among the vendors."
Goal #4: encourage rationality in the web market
[mainly by bringing to light existing irrationalities -- which then motivate
our own work on style sheets, etc]
> Monson, for example, has developed a three-part checklist
> to decide if he should use a particular proprietary tag
> in a given application. First, he asks, is the new tag
> essential, or are there standard HTML conventions that
> could achieve the same layout effect?
>
> If the answer to question one is yes, Monson asks if the
> new tag is sure not to cause any browsers to crash. If
> the answer to that is yes, he asks if the new tag will
> bring to the site some new capabilities that will increase
> its value to users.
> [Pathfinder] supports HTML extensions such as tables in
> an attempt to jazz up the look and feel of its Web sites.
> Hirschman claims the move--implemented last month--has
> helped, boosting page hits, although Pathfinder has to
> support two interfaces.
---------------------------------------------------------------
July 8, 1996 * PCWEEK (Ziff-Davis)
HTML hodgepodge
Stung by browser incompatibilities and crashes, Webmasters are treading
lightly when it comes to piecing together proprietary extensions.
By _Jeff Moad_
Eric Monson knows as well as anyone the dangers of jumping too soon on the
jumble of HTML extensions that regularly come down the pike from such vendors
as Microsoft Corp. and Netscape Communications Corp. Not so long ago, the CIO
and Webmaster at Mainstream Online took advantage of new HTML tags from
Netscape for creating tables and adjusting font sizes. Only later did he
realize that they weren't supported in many older Web browsers and, in fact,
caused some to crash.
"We got virtual bags full of angry Email," recalls Monson, whose Stamford,
Conn., company runs job placement and career planning sites.
As a result, Monson is now taking a much more conservative approach to
adopting new HTML tags and extensions. Recently, for example, he reversed a
decision to implement scrolling windows called frames supported in Netscape's
Navigator browser. The reason: He was afraid Microsoft's Explorer and other
browsers wouldn't support the enhancement smoothly.
"For the moment, like it or not, I feel like I have to choose the lowest
common denominator of HTML in most cases," says Monson.
He's not alone. These days, a lot of Webmasters are treading carefully when
it comes to patching together new HTML extensions to make their Web sites
sing. They're realizing that one byproduct of the increasingly intense
competition between Web technology vendors is that not all extensions to the
bedrock HTML technology are created or implemented equally. That means the
HTML extensions one vendor's product recognizes may make another's draw a
blank or display nonsense.
"We want to be able to put out a product that delivers the richest possible
experience for as many users as possible," says Jonathan Hirshman, a senior
producer in New York with Time Warner Inc.'s Pathfinder series of Web sites.
"The only way we're going to be able to do that is to get some agreement among
the vendors."
Consensus building
It won't be easy to win these concessions, however. Whereas most vendors
hewed closely to early versions such as HTML 2.0, lately Microsoft and
Netscape have virtually ignored the follow-on 3.0 draft under development by
the Internet Engineering Task Force.
The implications are particularly acute for companies attempting to build
Web-based electronic commerce businesses appealing to a broad audience of
customers. Unlike IS managers building intranet applications, they can't
control the range of browsers being run by users or the way users access their
sites. For most users, upgrading to the latest version of Navigator or
Explorer is no more complicated that clicking on an icon, yet many are still
running antiquated browsers.
As a result, many IS managers are opting for the conservative approach to
using new HTML extensions. At San Francisco-based Charles Schwab & Co. Inc.,
for example, the decision is based on which extensions are supported by the
browsers most in use on online services such as CompuServe Inc. and America
Online Inc. Because it only recently started supporting the latest versions of
Navigator and Explorer, Schwab has stayed away from proprietary tags such as
frames.
Critical questions
Others are taking a similar approach with a few well-considered exceptions.
Mainstream's Monson, for example, has developed a three-part checklist to
decide if he should use a particular proprietary tag in a given application.
First, he asks, is the new tag essential, or are there standard HTML
conventions that could achieve the same layout effect?
If the answer to question one is yes, Monson asks if the new tag is sure not
to cause any browsers to crash. If the answer to that is yes, he asks if the
new tag will bring to the site some new capabilities that will increase its
value to users. "Just because we can do it doesn't mean that we should," he
says.
In a few cases, the answer to all three questions has been yes. Mainstream,
for example, has used the Netscape support for frames to allow users to scroll
through job listings using a tool bar. They can drill down for more
information on each listing via a CGI script that accesses a database.
Mainstream must, however, provide users of browsers that don't support frames
access to subsets of the same information on separately maintained, static
HTML pages.
Time Warner's Pathfinder is making similar trade-offs. The company recently
decided to support HTML extensions such as tables in an attempt to jazz up the
look and feel of its Web sites. Hirschman claims the move--implemented last
month--has helped, boosting page hits, although Pathfinder has to support two
interfaces.
If they're lucky, Hirschman and other Webmasters may soon get some relief.
Perhaps recognizing that key vendors had lost confidence or interest in its
efforts, leaders of the IETF working group recently recommended that authority
for HTML standards shift to the World Wide Web Consortium, which boasts heavy
support from Microsoft, Netscape and other major Web vendors. Some electronic
commerce companies, including Time Warner, are optimistic that the consortium
will do a better job of keeping official standards in sync with the browser
implementations of the key vendors. In fact, the consortium recently released
a proposed HTML 2.0 follow-on, called HTML 3.2, that splits the difference
between some of the key extensions from Microsoft and Netscape.
Other IS managers, however, will continue to play it safe. "The vendors have
done a lot of work toward creating brand name recognition for their browsers
based on proprietary tags," says Monson. "I suspect that when we get to HTML
4.0, Netscape will still be doing its own proprietary tags."
Senior Editor Jeff Moad can be reached at _jeff_moad@zd.com._