This thread I opened has triggered a major campaign in the REST vs. WS wars that rage across this Group sporadically. This time I feel that we are seeing some very constructive debate with a genuine interest to engage with what the other side offers (well at least in one direction :-)). So what provisional conclusions can we reach? I would suggest:
* In contrast to 3 years or so ago when we had the great CORBA vs. WS wars, I think we can conclude that CORBA has entered the legacy stage of its lifecycle. I would be most interested to hear from any dissenters of that opinion! * WS has overwhelming support from the vendor community, particularly the big boys. * REST looks intriguing and has inspired some firebrand evangelists. However it seems unlikely to blow WS into the ditch anytime soon. * There is probably an interesting market in educating developers/consultants/analysts/architects etc. about the the virtues and utility of REST. o From that, assuming some sort of success in conversion there should develop a market in training the same people in REST. * Thanks to apostles like Gregg keeping the flame burning and igniting forests of paper with examples of mobile code, Jini is still a potentially significant player in the SOA market. Sun, to be frank, has done sod all to promote Jini or JavaSpaces to anything like their full potential, IMHO. As I recently mentioned to my J/JS Group, the least they can do is offer a Gregg a fat salary and a chauffeur-driven RR Phantom. o Quite a few of you are J/JS practitioners as far as I know, e.g. Gregg, Daniel C, Patrick M, Mark P, David F (of CORBA wars memory!) etc. How do you see J/JS fitting into the SOA scenario? Do you see it as an adjunct to Java EE 5, WS or just POJO? * Do any of you see any other serious rivals to the ones mentioned in this message for implementing SOA? Gervas
