My Blog List

"Brand-X" Browsers -- Non-HTML Browsers

A few of the entries in my list don't actually display HTML documents. One can thus question whether they're truly "Web browsers," since HTML is the normal format for Web documents. However, the Web is more than just HTML; it is a system to allow many kinds of media to be served and browsed via many different protocols, with the common thread being the standards (especially URL syntax) by which these media are referenced and accessed. Thus, non-HTML browsers can still be seen as part of the Web, though the use of variant formats does have the potential to Balkanize the Web by creating disjoint sets of incompatible documents.
WML, the Wireless Markup Language, is the main non-HTML contender in this area, being heavily promoted for use on cell-phone browsers. While I question whether a different markup language is needed for this application (rather than the use of logically-structured HTML that would be compatible both with cell-phone browsers and normal browsers), the developers of WML at least went about the development of a new markup language and media type in the correct way, creating an open-standard XML DTD, properly registering a MIME type (text/vnd.wap.wml), and using the correct HTTP headers in communications between servers and WML browsers to allow the use of content-type-acceptance negotiation (so a server can serve WML or HTML for the same URL as appropriate). Thus, I give them an A+ for correct, standards-based implementation, but a grade of "incomplete" when it comes to proving their case that a new markup language was needed at all. I suspect a large part of the impetus for a separate wireless markup language came from the misconception that HTML is about graphical-design layout for standard-size PC screens, rather than about logical structure as it truly is when used correctly. And the presence of WML, as well-developed and well-implemented as it may be, only helps perpetuate and extend this misconception, while the development of a handheld browser that gave a good rendering of cleanly-structured standard HTML would help encourage the increase of logical Web design.
But it could be much worse; some of the competitors to WML, such as "HDML" and some "web clipping" applications, are proprietary systems that offer no way for developers to find out how to create sites using them unless they pay for special developer programs or training kits. That's totally alien to how the Internet developed through open cooperation.
Salvaging the broad accessibility of the Web despite the balkanization of variant data formats are proxy servers which translate normal HTML into WML "on the fly" to allow WML-browser-users to browse standard Web sites in addition to ones designed specifically for them. In fact, the Google search engine has a WML-enabled version which includes a proxy server to translate all sites found in searches, and all sites linked from them, to WML. Also, PyWeb lets webmasters enable WML access to their sites via a server that translates the code in real time and also recognizes a few embedded tags as HTML comments letting the site developer suggest things like where to break a large page up into smaller chunks for WML use. These proxies work pretty well... on Web sites designed to be text-browser-friendly.
More information about WML (and the WAP protocol used by WML-capable cellphone browsers) can be found at:
http://www.wapforum.org/
A critical review of WAP and WML is at:
http://www.osopinion.com/Opinions/MikeBanahan/MikeBanahan1.html

No comments:

Post a Comment

Related Posts Plugin for WordPress, Blogger...

dg3