JavaWorld
addict
Reged: 06/20/03
Posts: 482
|
|
Regarding application development in the Java space, what are your predictions?
|
Denis Robert
Unregistered
|
|
Main Prediction: Almost everything Mr Zeichnich said will NOT happen.
Ajax is overrated for enterprise apps. I expect to see it growing in the B2C space, but for internal apps, it's likely to meet stiff resistance. 2 major issues: performance/latency, which is atrocious, and stability, which is even worse.
Ruby is also very much overrated, as are most pure scripting languages. The complexities of writing large "scripted" applications are not well understood, and it'll be some time yet before best practices emerge. When working on large applications, you can't be spending 80% of your time managing the code base. None of the C-based scripting languages (Ruby, Python, Perl are all C-based) have anything resembling Java's hierarchical class-loading scheme, which is one of Java's most successful features for dealing with large apps.
JavaScript is horrible; it's near impossible to debug, and should be relegated to the museum of computing horrors. Unfortunately, it's here to stay. But to ENCOURAGE its spread is like encouraging plague carriers to run about freely.
On the other hand, something happened in 2006 which this article completely ignores: Java6 now directly supports scripting language integration (it was a separate JSR before). And Groovy, the only official scripting language for the JVM, has just reached 1.0. This will have a significant, if slow, impact on the Java community, where we're more likely to see the gradual move to hybrid applications.
I'm skeptical of any predictions of massive, sudden shifts. SOA is really nothing buy a buzzword covering an array of ill-defined technologies. So how could any company even contemplate "going SOA"??? No. Companies will adopt SOME of the technologies covered in the SOA space, one at a time. Slowly. Creakingly slowly. In 5 years, we may be in a position to state what SOA really looks like.
So please, bring back SOME level of reason into the debate; not just the wishful thinking of a few script-kiddies.
|
Unregistered
|
|
Predictions:
-Most companies won't understand SOA in 2007, but will implement what they think is a SOA solution, fail, and then claim that SOA is a failure -Most companies that don't "get" SOA will end up in the same boat that they were in with EAI or distributed computing: vendor lock-in or a distributed rat's nest architecture -Any company that doesn't have drilled in security, change management, or agile development policies will complain that SOA is hell on wheels -Ruby buzz will start to die. Scripting languages running on virtual machines (JVM, CLR) will start to work their way into the enterprise instead -The Spring framework will continue to quietly work its way into every developer's toolbox -- creating a big bang in 2008
|
|