Hi, On an unrelated note, there's often no need to change implementations to run 1.5 features in a 1.4 VM. I've been using http://retrotranslator.sourceforge.net/ for years, enabling the use of generics, enumerations, concurrency and other 1.5 features by automatically replacing calls in the bytecode to their provided 1.4 compatible counterparts. Although the process isn't wholly faultless (there are some parts of 1.5 with too much reliance on native hooks to be optimally supported by this method, sometimes compile time weaving will be confused by the class changes - e.g. SCR plugin generating bind methods on the wrong compile time class) and there may be circumstances precluding its use (disk space, the runtime component adds 700kb) I haven't had any significant problems using it to the extent that, in my own APIs I don't regularly consider 1.4 limitations even though the code frequently ends up running in such environments.
Best, Dan. On 12 Mar 2014, at 20:26, BJ Hargrave <[email protected]> wrote: > I am working on the Promise API and Tim Ward is working on the Async API. Tim > has the overall "pen" on the RFC. > > I have just recently released some changes to the Promise API and am > investigating some other improvements. Tim needs to refresh the RFC doc to > pick up my Promise changes. > > With respect to the Promise API and Java 1.4 compatibility, we are using > generics in the API but they can be elided by -target jsr14 using a compiler > that supports that. In the implementation of Promise (PromiseImpl), we are > using java.util.concurrent types from Java 5. But someone wishing to use the > Promise impl in a pre-Java 5 environment can modify the implementation to use > synchronized instead. > > -- > BJ Hargrave > Senior Technical Staff Member, IBM > OSGi Fellow and CTO of the OSGi Alliance > [email protected] > > office: +1 386 848 1781 > mobile: +1 386 848 3788 > > > > > From: Mike Wilson <[email protected]> > To: "'OSGi Developer Mail List'" <[email protected]> > Date: 2014/03/12 14:22 > Subject: Re: [osgi-dev] asynchronous service calls and/or messaging > Sent by: [email protected] > > > > Yes, will check the Github page. Do you have any outstanding changes that you > will publish the next few days that could be worth waiting for? > > Btw, does this spec still target Java 1.4 for compatibility? > > Best regards > Mike > > BJ Hargrave wrote: > > From: Mike Wilson <[email protected]> > > > I have read through the spec a first time. I like that it keeps the > > separation of concerns between API and distribution, just like > > Peter's blog post. (I have a few comments on the spec, but I'll > > leave them for another mail thread later this week.) > > Great. Make sure to check the link for the latest updates. > > > > > Do you have a target date for this specification's release? > > We hope in 2H2014 but nothing nailed down yet. > > > Will there be any reference implementation? > > Of course. All OSGi specs have a reference implementation. > > > > > Also, is there parallel work being done on an update of the Remote > > Services specification to add these async features, or will this be > > done in a new iteration after this specification is finalized? > > This I am not sure. I suspect it will be serialized. There is already RFC 203 > which is updating Remote Service Admin. I am not sure how RFC 206 will > influence Remote Services in the next spec release at this time. > > -- > BJ Hargrave > Senior Technical Staff Member, IBM > OSGi Fellow and CTO of the OSGi Alliance > [email protected] > > office: +1 386 848 1781 > mobile: +1 386 848 3788 > > _______________________________________________ > OSGi Developer Mail List > [email protected] > https://mail.osgi.org/mailman/listinfo/osgi-dev > _______________________________________________ > OSGi Developer Mail List > [email protected] > https://mail.osgi.org/mailman/listinfo/osgi-dev
_______________________________________________ OSGi Developer Mail List [email protected] https://mail.osgi.org/mailman/listinfo/osgi-dev
