I'am confused about the future of J2ME. I understand J2ME is great for building applications on mobile devices. But I feel its use is limited to delivering personal applications or games etc. Is there a technology available to access databases or legacy systems from mobile devices like we do with J2EE? I think not yet and we are probably quite far from it. So then, J2ME would cater to a niche market only consisting of games, syndicated content or personal apps ( calorie calculator, diet menu planner :-) ). So probably not necessary for all the J2EE developers to jump the ship to J2ME.
Originally posted by archana:
Is there a technology available to access databases or legacy systems from mobile devices like we do with J2EE? I think not yet and we are probably quite far from it. So then, J2ME would cater to a niche market only consisting of games, syndicated content or personal apps ( calorie calculator, diet menu planner :-) ). So probably not necessary for all the J2EE developers to jump the ship to J2ME.
Of course we won't jump ship on J2EE. Different tools for different tasks. Regarding "niche market only consisting of games, syndicated content or personal apps", I think you forgot something, namely enterprise information applications. Some of the most beneficial applications of mobile technologies have been among field workers who don't need plenty of screen space but who save practically several hours a day by not needing to negotiate on the phone with a central in order to figure out where to go next.
I doubt J2ME developers will outnumber J2EE developers any time soon, possibly never. Personally, as a primarily J2EE developer I feel more than secured seeing Java succeed on handhelds as well.
I do NOT believe that many J2EE developer will "swtich" to J2ME anytime soon. However, I do believe that J2ME will add value to your skill set as many enterprises are going mobile these days. Also, with the number of J2ME devices exceeding PCs *this year* and the huge potential growth in the wireless market in the future, the number of J2ME developers will soon out number J2EE developers. The new J2ME developers are likely to come from C++ or VB or many other backgrounds.
The fact that J2ME devices outnumber Solaris boxes doesn't do it for me. There are millions of PCs out there and yet we have only some 10,000s developers writing commercial operating systems for them. (bad analogy?)
Another, perhaps better argument would be that for almost every enterprise J2ME application, there is the enterprise J2EE backend requiring on average more effort/developers than the client-side.
The only source for the large growth of the J2ME developer community that I can see are the game developers.
I'd love to hear people's visions about this subject. I think it's important to know where the industry is going--skill demographics included.
Devices will far out number PCs in the next several years. hence, the demand for J2ME developers will grow. Game is one area of growth. But you can do so much with today's powerful mobile devices (have you tried to run a Java servlet engine on a PDA or an Oracle database on a cell phone? They can be done TODAY). I do not believe game is the only growth sector.
For example, IBM has a huge division (more than 10K employees) on wireless/J2ME. Yet, IBM does no game nor consumer software.
Furthermore, even though you have one server per thousands of clients, it's still one server-side and one client-side application.
Somehow I think most of the "near-killer" consumer applications are already being distributed by default along with the handheld. The corporate world will always have the need for tailored applications on some sector before standardization kicks in and the infrastructure is commoditized.
Hmm. My thoughts are becoming more and more irrational so I'm just gonna go to sleep
VB makes it easy to program a PC, but I wouldn't want to try using it for building an enterprise system. In fact, the company I work for bought another (much smaller) company a while ago and there's a team currently re-writing the application that was the reason for buying them in the first place. It's currently a VB system, but it won't scale, it's brittle, etc. The re-write is J2EE based.
Currently, I'm not aware of a VB-ish development environment for cell-phones and the like, so for now it's between C/C++ and Java - and if you need to have more than something running on the phone (i.e. a server component) then I believe that J2ME is the way to go. Java has the edge because it'll do both the client and the server as well as run on multiple devices without lots of special code or rewrites. At least that's what Sun keeps telling us.
Originally posted by Burk Hufnagel:
Currently, I'm not aware of a VB-ish development environment for cell-phones and the like, so for now it's between C/C++ and Java
Microsoft has something called Embedded Visual Studio, which lets you write applications for PocketPC and Windows CE. It comes in two editions: one for C++ and one for VB.
Originally posted by Burk Hufnagel:
I guess those will help when WinCE phones are more popular, but for now (at least according to CNN) Java has the lead because it's on more phones.
PocketPC/WinCE have not reached phones (probably won't ever?) but they do have certain advantages on PDAs. For example, what is the biggest single issue with using a PDA as a true Personal Digital Assistant? The fact that it's just too difficult to keep your work calendar (Outlook) and your PDA (non-Microsoft) calendar in synch. This is where the Windows-Windows combination will have edge for some time, I think.
Thanks for enlightening me that MIDP can do HTTP. I was unaware of that previously. So mobile clients for enterprise applications is a possibility rather than a probability as I was saying previosuly. Thanks for taking the time to join this group and making us aware of the possibilities and advantages of J2ME.