The original promise of HTML was an abstract markup language for representing structured information in a common, display-independent format. Over the past few years, it has evolved into a rendering specification for a small number of desktop browsers. It's become practically impossible to build a browser which does not comply with visual formatting instructions suitable only for high-resolution PC displays.
For the Web to support the new generation of handheld devices, this paradigm will have to shift back. It's impossible to subtract from the power and control of modern HTML. This problem must be solved by extending the HTML infrastructure with new solutions: formats, protocols, and tools. A variety of such solutions are becoming available. Each offers a set of questions for the future of HTML:
Compact HTML is an HTML subset. WML is an XML DTD. Both have various advantages and disadvantages for various classes of small device. The crucial issue from our perspective, though, is compatibility. Will browsers for these languages be restricted to the narrow universe of content published specifically for wireless devices? Or will automatic tools enable them as gateways to the entire Web?
Style sheets are aiming at the right problem: disentangling information from presentation. CSS2 supports separate style sheets for palmtop devices. But, again: how many authors will do the work to define two styles for their documents? Can tools step in, intelligently translating from desktop to palmtop styles?
Web simplifiers are special proxies that filter images and/or complex HTML structures, reducing page download size and browser complexity. Split browsers go farther: they move most of the browser's parsing and rendering code onto the server, and the client processes a non-HTML stream.
What is the relationship between these classes of proxy and the other wireless solutions? Will split browser protocols be standardized, allowing clients and servers to be interchangeable? How does full XML browsing - which may require a proxy, due to the complexity of XML/DTD processing - fit this picture?
It's clear that no one technology will satisfy all needs, and it's important for us to understand and define the tradeoffs and architectural choices, identify potential protocols and formats, and standardize them as soon as possible. Geoworks sees the Web as crucial to the success of wireless network computers and smart phones, and we expect the W3C standards process to play a key role in the solution of this problem.