You don't mention for what version of GT this might take affect.  I
assume you're not talking about GT 4.2 (which doesn't seem to make
much sense at this late stage of the dev cycle).

Tom

On 9/17/07, Tom Howe <[EMAIL PROTECTED]> wrote:
> While looking into the solutions available for supporting database
> persistence for resources we discovered that the best solution was the
> Java Persistence API (JPA).  This approach requires the least demands
> on the service developer while providing a highly performant, standard
> solution that is acceptable under the apache license.  However, JPA
> requires java 5 because it is based on annotations.  As a result, we
> started to consider the costs and benefits of moving the java ws core
> to Java 5.  This email should outline these costs and benefits and,
> hopefully, justify moving the java ws core to Java 5.  Let's start
> with the costs of migrating to Java 5.
>
> Cost:
> - Upgrade to JDK 1.5. A version available for all platforms supported by
> Globus
> - Changes to language specification. Developers need to change use of enum
> as variable name.
>
> Benefits:
> - Allow use of JPA
> - Annotations and -javaagent may be leveraged by Java WS Core to improve
> service development usability through bytecode enhancement
> - Other features useful for cleaner code: Generics,
> java.util.concurrency, JAXP improvements, static imports, enhanced for
> loop
>
> A more in depth discussion of the costs and benefits is available
> here: http://www-unix.mcs.anl.gov/~ranantha/java5Upgrade.txt.
>
> Finally, since Java 6 has been out for quite a while and Java 7 is in
> the works, this seems like an appropriate time to upgrade to Java 5.
> It has been well tested and supported.  It could make developing
> services for the Java WS-Core a more pleasant experience.
>
> Please provide comments, concerns and questions.
>
> Thanks,
> Tom
>
>

Reply via email to