Before Openwave, before Phone.com, before Unwired Planet, was Libris. This company invented Handheld Device Markup Language and related browser, and launched it on the AT&T network. Keep in mind that the wireless connection was around 9600 baud, the phones displayed text, and displays were small. The browser embedded a handful of small graphics into the phone, which a site could call up to enhance its text. In 1998, the company (now Unwired Planet) co-founded the WAP Forum, and WML replaced HDML. UP was unable to leverage all of its excellent design work.
HDML (and its browser) downloaded decks rather than pages. Like WML, HDML was optimized for its very limited environment. But unlike WML, HDML had several features that actually enhanced the user experience.
On such a small screen with limited memory, cache control is very important. Thus HDML allowed the developer to designate a series of cards and decks that, when exited, would be removed from cache en mass. This didn't make its way into WML.
With such a large lag in connection (on most operators' networks, it took 30+ seconds to reconnect to the network; 10 seconds if already connected was typical if I remember correctly), the developer could indicate what content to pre-fetch so the user could get to it faster. This didn't make its way into WML.
The technology's primary drawback was the marketing assertion that it was "the web in your hand". Users took one glance at that, saw that the web could not be displayed on a 4-line text-only display, and "WAP is crap" was born.
The lessons learned 10 years ago have been forgotten. But what if they hadn't?
What if the WAP Forum had branched into a pair of services:??HTML-ish and thin-client-ish? If both sides had "won" that debate???If we now had an evolved mature standard to load scripted applets onto the phone, rather than the "15 versions of widget frameworks" we're now finding?