Realigning Java EE 7 - A promise is a cloud; fulfillment is rain.

Markus Eisele
1
The news came out yesterday. Java EE 7 is going to be realigned and the PaaS enablement and multi-tenancy support will be moved to Java EE 8. While the email to the EG and the linked Aquarium post are only proposals the first EG members (including me) already responded positively. I am kind of relieved now and happy to share some of the reasons I see behind it.

re·a·lign  (/ËŒrēəˈlÄ«n/)
tr.v. re·a·lignedre·a·lign·ingre·a·ligns
1. To put back into proper order or alignment.
2. To make new groupings of or working arrangements between.

What happened? You can roughly split this up into two parts.

Emerging needs in the community
The significant enhancements in simplification, usability, and functionality in updated versions of major JSRs which came with Java EE 6 are still far from perfect. We have seen a lot of requests regarding ongoing efforts to complete what has been started with EE 6. To name but a few the integration of JSON-P, the simplified JMS 2.0 APIs, further Managed Bean alignment, including transactional interceptors, a JAX-RS 2.0 client API, support for method-level validation, a much more comprehensive expression language and many many more. We are in need of a version I like to call Java EE 6.5. A comprehensive and complete version of what we have as of today.

Slow progress on the cloud side of our agend
The second part is the fact that we didn't see enough progress with the PaaS or cloud parts of EE 7 so far. As Red Hat's Pete Muir said:
We've long advocated that we, the Java EE community are not ready to standardise cloud yet, and feel this is proven by OpenShift, our Java EE cloud offering, which is working well with Java EE 6. (Source: EG Mailinglist)
And this is pretty much to the point. Even if I am an advocate of innovation with standardization this might simply be too early and as I personally see this today, the EG is missing some significant members of the Java cloud market (e.g. Jelastic). Even if Oracle is working with them to optimize their GlassFish offerings and I believe that there are first best-practices around this simply isn't enough for starting to standardize it today.
The Arabian Proverb from the title does fit this very well. At the moment all we can do is to promise stuff but we have never actually seen enough of the rain to start extracting the essentials that need to make it into the platform.

Given those two parts it is a brave and honest decision of the EG to realign Java EE 7 to the community needs and make a second try on Cloud and PaaS with the Java EE 8 Platform release scheduled for the spring of 2015.

Post a Comment

1Comments

  1. I'm actually happy about this. It's too early to standardise; innovation is still going on strongly in the cloud field, and we need some more time to figure out things that would make to standardise. This has nothing to do with the EG being to slow or anything like that, it just doesn't make sense to standardise yet.

    Java EE 7 is going to be a bit boring release compared to Java EE 6, but it's only healthy to have some time of evolution and completion before having more revolutionary releases. There are some important things in there (CDI 2, JAX-RS 2.0, JMS 2.0) and this should not be delayed to wait for other, maybe more exciting, features.

    I hope the plans for EE 8 will include more plans around modularity. The delay of Jigsaw makes this discussion somewhat difficult, but it is one of the main areas that should be improved in my opinion. Support for OSGi would make most sense, most application servers already support this and there are several OSGi specifications being written for Java EE integration. If the Java EE specs acknowledge this too some real progress could be made in creating more maintainable enterprise applications.

    ReplyDelete
Post a Comment