There are a number of specific differences between LEAP and WAP, but for my expository value, I’m going to focus on what I feel to be the three key issues:
- WAP was designed with web browsing in mind as the main application of wireless, whereas LEAP was designed to implement 2-Way Messaging.
- WAP is booby-trapped with patents, whereas LEAP is a patent-free standard, published in RFC form.
- IP-convergence: LEAP is based on IP whereas WAP has to re-invent the wheel.
The Killer App
The killer app issue is important because "necessity is the mother of invention." The fact is, no one really wants to use their clunky and slow phone interface to navigate websites; the need is just not there. Some applications, like movie listings, would be handy if they could be executed smoothly, however the current wireless environment just doesn’t lend itself to web applications very well. The whole web has been driving for more bandwidth, more multimedia, fiber to the home and WAP is supposed to translate all this to a cell phone’s tiny display and bandwidth? Seems unlikely at best.
LEAP, on the other hand, has taken a cue from Europe and Asia (especially Japan, NTT DoCoMo’s home turf) and first looked at the killer app in places where cell phone use is ubiquitous. The killer app is of course Short Messaging Service (also referred to as SMS, 2-Way Paging or Messaging). The first application level protocol developed by the LEAP forum is EMSD which is designed to provide this service. The LEAP forum is filling in a vacuum where there is demand for a service, and that’s just good business sense.
The Patent Issue
Patents are another area WAP stumbles. Imagine TCP/IP was controlled by 300 different companies, each with varying degrees of control over different parts of the system. To implement TCP/IP programs or hardware you would have to keep track of licenses for different companies and any work you were to do with TCP/IP would require a large legal exploration before even going ahead. Part of the reason for TCP/IP’s success and wide implementation is because it is free of such restrictions.
To date there have been at least two Patent infringement claims filed from members of the WAP forum relating to patents they have on parts of WAP:
I won’t get ahead of myself by making calls as to whether technology patents such as these are the right way to go, or doomed to failure, because that debate is going on right now by people well versed in both law and technology (Lawrence Lessig) and I doubt I could add anything meaningful to that argument. However, I will say that this patent issue is a hurdle that WAP will have to deal with. It will impede it’s progress; whether it will stop it entirely remains to be seen.
LEAP is of course patent-free and RFC published and will not have to overcome this hurdle. LEAP follows more closely the development formula of successful protocols like TCP/IP.
Let’s say a server on the Internet wants to send some data to a WAP phone. Where does it start? How does it address the phone? I don’t know, but not by IP address. Although, I guess it’s conceivable that would could setup your WAP gateway with some dummy IP addresses that it could translate into WAP-device addresses. I’m sure that will go over about as well as LANE did.
LEAP is IP-based. LEAP devices have IP addresses, they run IP and they run UDP. These are familiar and successful protocols and they are ubiquitous. There is a wide base of people who know how to support and troubleshoot these protocols and there is a huge number of developers already familiar with these technologies.
I should point out a few things:
- 3G wireless offers IP support directly to wireless devices
- WAP 2.0 specifications were recently released by the WAP forum that are incorporating IP support, a PUSH model, and about a dozen other bells and whistles
- LEAP had IP support and a Push system from day 1.
So there are changes going on that might change this difference between WAP and LEAP, but the support is slightly different:
WAP 2.0 is offering a modified version of TCP (Wireless Profile TCP, the WAP 2.0 gateway will translate between Internet TCP and WP-TCP) running over top of IP. LEAP is offering ESRO, over UDP, over IP. It’s tough to say at this point what the operational difference between the two architectures will be since they are essentially trying to accomplish the same thing by two different methods: WAP 2.0 is starting with TCP and whittling it down, and LEAP starts with UDP and adds some functionality via ESRO. Both methods are trying to accommodate IP to the wireless environment.
After reading the preceding arguments for the LEAP protocol, it seems like the right solution. However, that doesn’t necessarily mean it will become a de facto standard.
The WAP forum has over 300 members comprised of major companies such as Microsoft, Nokia and NTT DoCoMo and LEAP is backed almost solely by Neda Communications; WAP also has a head-start in implementation. The LEAP forum has many elegant arguments for it’s technology and has followed the standards process in the true spirit of the original reasons that standards bodies like the IETF and RFCs were first created. The fact remains though, that Neda is like a protagonist from an Ayn Rand novel: an individual with a good idea working against a collective with political and economic power.
Efficiency to compensate for the wireless environment was Neda’s main goal when designing LEAP. Efficiency is a big weapon against the current high-latency, low-bandwidth wireless environment. For the current situation I think implementing LEAP sounds like a really good idea, but unfortunately I have yet to see that kind of real-world comparison between WAP and LEAP.
However, I have to wonder how long high latency, high bit error rate and low overall bandwidth will be an issue. Wireless is still very new, but a lot of big companies are putting tons of money into it. Originally the Internet had bandwidth constraints and was largely text-based because of it; today, video and huge media-rich applications are moving over it.
I don’t think anyone is sitting on their hands for wireless. What happens if/when the wireless environment becomes as rich as the current wired networks? Obviously LEAP’s efficiency won’t work against it, but it may not be as big an advantage over WAP.