Rants And Ramblings About Mobile Technology
Anders Borg writing about the fun and crazy world of mobile and Internet service technologies.
You can also read the blog via Twitter or your phone via wap.abiro.com. See the left menu for more news.
Comments on blog entries are moderated, but I'm rather liberal as long as it's not blatant advertising.
For general comments, advertising and contribution queries, please use the feedback form.
Wednesday, May 30, 2007
JSR 232 / OSGi won't solve platform fragmentation
The Mobile Operational Management Spec (JSR 232) - about management, fragmentation, and what about MIDP? confirms what I wrote in OSGi to the rescue for mobile developers? that this new JSR won't solve the existing fragmentaton issues that are much more related to misunderstandings of sometimes too loose specifications, multiple KVM providers including phone manufacturers' internally developed/optimized implementations, system software (not the least multimedia support) affecting the Java behavior to a large degree, etc.
JSR 232 is mainly aimed at CDC implementations. As there are hardly any such implementations around, and market-share-wise CDC doesn't even show up as a titsy-bitsy fraction, it's clear JSR 232 won't do much of anything in the short term. As mentioned before, I believe much stronger in MIDP 3 than CDC for the foreseeable mobile future.
JSR 232 is mainly aimed at CDC implementations. As there are hardly any such implementations around, and market-share-wise CDC doesn't even show up as a titsy-bitsy fraction, it's clear JSR 232 won't do much of anything in the short term. As mentioned before, I believe much stronger in MIDP 3 than CDC for the foreseeable mobile future.
Comments:
Post a Comment
Links to this post:
<< Home
I think you make a common thinking error.
In the current model all changes to the platform are that need to be shared must be handled by the combination of the manufacturer + operator because applications are stand-alone.
In an OSGi model, you can download services that are shared between applications, these services can potentially be written by anyone, creating a middleware market for mobiles.
I think the reason the MIDP market is such a terrible mess is that normal competitive forces are not working to create a common standard, the developer is completely neutered. An OSGi model has the potential to break through this.
Kind regards,
Peter Kriens
In the current model all changes to the platform are that need to be shared must be handled by the combination of the manufacturer + operator because applications are stand-alone.
In an OSGi model, you can download services that are shared between applications, these services can potentially be written by anyone, creating a middleware market for mobiles.
I think the reason the MIDP market is such a terrible mess is that normal competitive forces are not working to create a common standard, the developer is completely neutered. An OSGi model has the potential to break through this.
Kind regards,
Peter Kriens
OK, so how does that solve platform fragmentation?
I agree there's potential for a middleware market, but that could have existed already using statically linked JARs licensed on a commercial basis.
I agree there's potential for a middleware market, but that could have existed already using statically linked JARs licensed on a commercial basis.
1. Size issues, statically linking is extremely expensive for mobile devices.
2. Sharing issues. Many services need to be shared to be interesting, the sandbox model forbids sharing.
3. Authorization. Interesting services require access to things not everybody should be able to touch. The OSGi security model enables this.
I think :-) This is of course all looking in the future, which is difficult at least. However, from a technical point of view I think the OSGi model is preferably over any alternative because it levels the playing field.
The mobile market has tremendous potential and since 1997, since I worked for Ericsson Research, I have been dreaming of a device like that is enabled by the OSGi/JSR 232 spec. CLDC and MIDP 3.0 seem, so, well, eh, ... nineties like? :-)
Kind regards,
Peter Kriens
2. Sharing issues. Many services need to be shared to be interesting, the sandbox model forbids sharing.
3. Authorization. Interesting services require access to things not everybody should be able to touch. The OSGi security model enables this.
I think :-) This is of course all looking in the future, which is difficult at least. However, from a technical point of view I think the OSGi model is preferably over any alternative because it levels the playing field.
The mobile market has tremendous potential and since 1997, since I worked for Ericsson Research, I have been dreaming of a device like that is enabled by the OSGi/JSR 232 spec. CLDC and MIDP 3.0 seem, so, well, eh, ... nineties like? :-)
Kind regards,
Peter Kriens
I have to agree on that. MIDP will probably have the same evolution as VHS: minor refinements over time, until it's deemed so outdated that it's completely ditched for something else.
I wouldn't complain if it's still Java (or better), as in comparison both C and C++ are technical dinosaurs.
I wouldn't complain if it's still Java (or better), as in comparison both C and C++ are technical dinosaurs.
Post a Comment
Links to this post:
<< Home
