Oh yes, I do remember those days at the end of last millennium when WAP was the latest innovation in mobile technology and the first phones equipped with a browser were hitting the market: Siemens S35, Nokia 7110, Ericsson R380, Motorola Timeport… I had an S35i and I loved it. Black and orange straight monochromatic display, just 3 lines by 14 chars worth of text, almost no graphics.
GPRS was yet to come, we had just CSD at 9600bps and we had to pay by air time, just like a normal phone call. It basically felt like going back to BBS age and use a modem over analogic PSTN, but with a textual browser that you could carry with you anywhere.
Phones didn’t come pre-configured and setting up one was tricky. For each carrier you had to enter a different dial-up number, user and password, gateway IP address and port, plus strange settings like connectionless. Of course people would never remember those parameters by heart and would look them up on the Internet.
Surfing the normal web with this technology was basically impossible. If you wanted to make a small page accessible from the phone you had to write in a special markup called WML which was not compatible with HTML and in fact represented a very different paradigm. Instead of pages you had decks and cards. For people used to the web, WML was real hard, first because it was XML based (which had been just released and mostly unknown – very few people could tell what well-formed and valid meant), second because it was a different set of tags, third because mobile browsing was a very different user experience. With no mouse, no qwerty and a very small display you had to design your decks very carefully, with usability in mind.
All of this was the brainchild of a company called Unwired Planet (nowadays known as Openwave) that was able to create a consortium together with the leading handset manufacturers. They called it WAP Forum (nowadays known as Open Mobile Alliance) and issued a complete suite of spec documents featuring a transmission protocol stack (WDP, WTP, WSP), a markup language (WML), an image format (WBMP), a binary compressed representation for XML (WBXML) and a middle-ware component architecture (the WAP gateway). A huge undertaking that also happened to be the topic for my master thesis, so I intimately knew every single bit of those documents.
Coding for WAP was very exciting, but experimenting with a real phone was expensive, slow and cumbersome. When something was wrong, the phone didn’t yield any meaningful error message, just a plain dumb 500 internal error. In order to easy development, a few emulators were developed by manufacturers, mostly as win32 native applications. None of them was satisfying to me, so I ended up developing my own.Usability was a big deal. Not only each phone had different display size, form factor and soft keys. On top of that, each manufacturer rendered some tags in a very different way. For example selection lists could be displayed similar to drop down lists or radio buttons.
There were a few wap sites available, some of them linked from the carrier’s WAP portal. There was no usable search engine, so you had to rely on word of mouth or list of links, just like the early days on the web. In the end none of those services became successful, mainly because the price you had to pay for a WAP session was so high.
I remember there were a good number of start-ups flourishing around WAP. In UK they were concentrated in London and used to gather in a periodic event they called the WAP Wednesday while in Italy they were spread over at least 4 different cities and had no pulse to gather together. Unfortunately I had just graduated and had no money to flight to London just to attend a meeting.
I remember many people abandoned WAP after the first navigation attempt, either because they were scared by navigation costs, or they found very bad services (bad for usability or for lack of interesting content). The industry was technologically ready, but the market was not and WAP became known as a big flop. One year later GPRS arrived and it was marketed as the next technology after WAP, when in fact it was just a substitute for CSD. Companies needed a way to make investors quickly forget the flop, who cared if WAP as a technology remained central for everything that came later?
