Carl Zetie discusses the unresolved issues and hurdles for Web services in mobile application with his article "Don't Hold Your Breath For Mobile Web Services."
I'm glad to finally see this recognized somewhere in the mainstream press. Its been on my mind for months now.
I differ with Zetie in only a few areas:
> I am more pessimistic to whether WSDL and UDDI have a place at all in mobile applications and WANs. Given the low powered nature of most devices in the foreseeable future and the issues of cost and network latency, I believe a less dynamic and more tightly coupled approach to mobile applications will be more appropriate. Dealing with mobile application updates will be marginal given the miniscule file size and the ability to install applications over the air. The ability to process WSDL and UDDI consumes processing power and battery life (a key concerned often overlooked) in addition to adding to the size of the application code.
> I am not as convinced that 3G networks are going to resolve as many issues as many such as Zeite suggest. Perhaps in theory 3G networks will. However carriers are charging subscribers based on bandwidth consumed thereby putting the onus bandwidth conservation on developers. Consideration also needs to be given to the potential worse case scenario of saturated mobile networks.
> I also think more attention could have been paid to the need for asynchronous Web services (messages queued and sent once a connection is available) versus synchronous Web services ("real time" request and response) that are commonly thought of today.
My concern is that the optimism in 3G networks and the more powerful devices will yield less streamlined or even bloated specifications that do not make economic sense and are simply not useful. Stay tuned.
