Java EE 6 - Gee, 6?

Java Enterprise Edition 6 was just released, even with Oracle's possible acquisition of Sun looming on the horizon. In the past I've seen J2EE and Java 5 EE releases come and go whilst I went on not caring. It wasn't because I didn't need anything more than Java's Standard Edition... I do... but the Java Enterprise Edition releases always seemed to be released with features I needed nine months ago and found in other frameworks.

I decided to take more of a look within this release and give it a fair shake. This release had a lot of nice stuff but was still playing catch-up with other tech: JPA starts catching up with the functionality offered by Hibernate, JAX-RS (RESTful Web Services) is finally produced as an answer to the myriad of other REST frameworks and dependency injection finally gets introduced into Java proper. All of these additions are features that were already offered by alternate libraries over a year ago (some of them more than five years ago) and are just now becoming standard.

Inversion of Control isn't the only elder pattern that finally debuts in Java EE 6. Asynchronous responses appear throughout the new Enterprise Edition, finally allowing threads to stop blocking subsequent operations. Asynchronous processing is finally available in the Servlet 3.0 spec so inbound HTTP requests don't have to hang forever while they await all the info to formulate a response. Session Beans can now be invoked asynchronously so EJB execution doesn't consume a client thread and eat up threadpools. JSF now has AJAX support for asynchronous webapp calls. REST-based Web Services allow for Web Service exposure with much less resource utilization and much higher concurrency through more streamlined asynchronous HTTP calls. All of these features not only have been present in alternate libraries for some time, but entire protologisms and design patterns have been honed to allow for just such asynchronicity. Comet and callback methods have been a predominant pattern for some time now, although Servlet 3.0 is the first time a Java Enterprise Edition has allowed for asynchronous Servlet processing. Now that WebSockets are becoming more prevalent Java Servlets are going to need to catch up to everyone else yet again and in short order.

Am I even tempted by this release of Java Enterprise Edition? Maybe kinda. I like annotation-based models, as I've seen how deep XML hell truly goes. JAX-RS annotations are a compelling alternative to JAX-RPC, and the new Bean Validation annotations would likely be very helpful. Asynchronous EJB invocation, along with a more minimalistic and annotation-based EJB spec, may even convince me to finally give Enterprise Java Beans a try once more.

I already use JPA, JAXB, JMS, JMX, JAAS, JAX-RPC, JavaSpaces and Servlets/JSPs/JSTL but in more of an a la carte sorta mechanism, intermingled with Spring, DWR, Hibernate and Camel. Now that I've typed the previous line... I realize that even though I've never intentionally downloaded the Java 5 EE package from Sun it seems I've been using Java Enterprise Edition all along. Weird. I guess it has slowly seeped into the crevasses of my frameworks, unintentionally filling out the rest of the libraries I use. Now that we have REST support and asynchronous context... I might walk the final yard and intentionally download it this time.

Someone to Give Me the Time

It's been really interesting to see the responses from Blitz, Fly Object Space and GigaSpaces concerning state management as well as Newton and Rio concerning service discovery. I'm definitely learning as I go, but the good thing is that it seems like there are many in the community eager to help.

Now I'm working on another issue with enterprise service development - scheduled services. There are some services out there who may want a have an event fire in 1000 milliseconds, or five minutes, or an hour, or somewhere in between. This would appear to be an easy thing to solve at first blush - until you consider volume, quality of service and scalability. It's a steep drop into complexity at that point.

Here's the thing: you could easily just do a scheduled executor in J2SE, but once your VM dies then your pending events die too. You could submit a scheduled job to something like clustered Quartz instances, but then you must have a reliable back-end database to write to (no native replication). You could use something like Moab Cluster Suite, but it seems to live outside the muuuuuuuuch more simple realm of event scheduling.

So let's think outside the box and use some replicated object store that isn't necessarily meant for scheduling. How about we slap a time to live (TTL) on a JMS message, throw it on a queue and wait for it to hit the dead letter queue? That might work at times, but TTLs are really intended for quality of service and not for scheduled events. Unless you have a consumer attached to the former queue constantly polling for messages you're not guaranteed to land in the latter dead letter queue.

How about using Camel's Delayer Enterprise Integration Pattern? Nope - that's just a Thread.sleep on the local VM. Doesn't do you much good once the VM dies. How about a delayed message using JBoss Messaging? I've heard tell that it exists, but I can't find much reference to it in the documentation.

This isn't a new problem - there's even JSR 236 that is intended to address this problem. But it's been hanging around since 2004 with very little activity of note, so I doubt it's going to have much hope of working by Monday.

Until JSR 236 is addressed I'll likely have to just find a way to deal with this on my own. Maybe create a JobStore for Quartz that's backed by a JMS topic? Or just suck it up and build a clustered Quartz instance with a fault-tolerant database?

Gah. Sticky wicket.

Empty Spaces

Yup, I'm still trying to work my way around Jini. This time it's JavaSpaces.

Apache River hasn't gone much farther from the last time I looked at it, but I liked the bare-bones reference implementation aspect. GigaSpaces seems a bit thick for my tastes and seems to be tightly coupled with their application server. I thought Blitz JavaSpaces might be a better fit, especially if I could use their fault tolerant edition.

I was able to get Blitz up and running then configured it to do unicast discovery to a pre-existing Jini registrar without a problem. I was having problems getting my client to connect in its security context, so I decided to dig a little deeper. As I did I also kept an eye towards fault-tolerance, but found that branch seemingly suspended. I later found a post from the author indicating he didn't really see a good motivation for moving forward with his fault-tolerance work:

In my spare moments I've been doing a re-implementation but the fact of the matter is that it's not a trivial problem to solve (though I believe I do have a solution). And here's the rub, this work doesn't pay the bills which means that it's going to take a long time to implement because I have to do a day's work first. For those who don't know, most of Blitz has been written during time between periods of employment - not over weekends and evenings as you might expect.
This presents me with a problem - users seem to want this feature but I'm struggling to see doing this as a good thing. Here's some of my reasons:

1. I'd be building a significant feature which will, judging by demand, make a lot of money for those who use it but zilch for me.
   
2. Not only do I earn nothing from this venture but I have to earn a significant amount of cash just to allow me time to develop the feature. Basically, I'd be financing everybody else's money making ventures.
   
3. One of GigaSpaces key value adds is the clustering/replication feature - they are fully commercial and need to earn a crust plus they're one of only a few credible companies that can provide corporate grade support for JavaSpaces. Were I to do this work for Blitz I'd maybe be damaging the market I've been helping to create.
   

Right now, I feel like the price of this piece of work is too high for me personally and for others in the commercial JINI world (and like it or not they are an important element in any future success for JINI). I can see why Blitz users might want this feature - they can avoid paying Gigaspaces a license for starters.

So... it seems like the development of an enterprise ready Blitz isn't in the cards. Casually strolling through Wikipedia's definition of a Tuple space brought up Fly Object Space, a tuple space that is not a JavaSpace implementation. While it doesn't fit into the Jini realm I know and love, it is a more minimalistic implementation of an object space that fits my desire of something smaller and to-the-point. It doesn't appear to support replication or fail-over on the non-commercial level, but I'm checking to see if there are plans to support it on a commercial level.

It's tough. I need an object space that has a minimalistic implementation, has a small footprint and can at least run active/passive for fault tolerance. Maybe I might have to dust off my old Terracotta instance and try out SemiSpace.

EDIT: Be sure and see Nati Shalom's comments following this post.

Jini in the Skull

Now that I have my uber-huge project out the door, I'm trying to think smarter about development in general. I kept thinking that being smarter about development meant thinking bigger - so I initially tried to get more involved in the infrastructure of things. But it wasn't a good fit in my brain - you can't lose yourself in a PowerEdge like you can lose yourself in a stream of bytecode.

This hit home when I was talking to a previously C++ programmer at work. Because I'm a) lazy and b) often muddle explanations, I often tell new Java developers... whilst stammering and mealy-mouthed... that Java passes objects by reference. This quickly helps people understand(ish) why setting a value on an Object inside a method doesn't necessitate a return value. However, when I say that I'm perpetuating a distinctly wrong concept. One I should be ashamed about.

I try to avoid the long conversation that ensues, but I should tell developers the truth. Java always passes by value, never by reference. In fact, it passes references by value. Most people look at me like a Buick is growing out of my head when I say "passes references by value." But the fact is that variables only hold primitives and references, and are always passed by value.

Take a look at Java is Pass-By-Value, Dammit! It all boils down to this fact: C and Java always pass by value... that's why you can't do the standard "swap" function. For example, in C++ you might do:

Object hello = new("hello");
Object world = new("world");

swap(hello, world);

printf("%s", hello);
printf("%s", world);

void swap(Object obj1, Object obj2) {
Object swap = obj1;
obj1 = obj2;
obj2 = swap;
}

This transposes the value in obj1 with the one in obj2. You should see "world hello" from running this craptastic pseudocode. Try this same code in Java and nothing happens - you've locally overwritten some values, but you didn't swap your references:

Object hello = new Object("hello");
Object world = new Object("world");

swap(hello, world);

System.out.println(hello);
System.out.println(world);

void swap(Object obj1, Object obj2) {
Object swap = obj1;
obj1 = obj2;
obj2 = swap;
}

Here you'll just get "hello world" out - your references remained intact, because you passed by value.

This is ultimately a cleaner way to develop... so it's a major plus for both C and Java. And my New Year's resolution is to stop telling people that Java passes by reference just so I can end the conversation sooner.

On the flip side of the complexity coin, I've been reading the the Jini specification by the Jini community as well as Jini and JavaSpaces Application Development by Robert Flenner. Jini has evidently been around forever, but I've only recently become interested in it. Remember the old pre-turn-of-the-century adage that Java would be running on everything in your house, from your kitchen toaster to the fridge? Evidently around 2000 or so Jini was sold as becoming the premier way Jini could allow your toaster to auto-discover your refrigerator and... er... do... heating and cooling... stuff. Who the hell knows. The idea of automagically connected and integrated micro-device clusters communicating across a mesh network is cool, but practical consumer applications are pretty much nil. Then once EJB's started incorporating RMI, Jini came back to the forefront as an easy way to do the heavy lifting of RMI without the thick refried JavaBean layer.

Once you get Jini up and running, it is wicked cool. Start up your Jini registrar and then poof, services get published across a network. Look for a remote service and poof you can discover it and invoke it - no need for stubs or manual marshalling. Once you get the Jini infrastructure up, you don't have to teach developers how to use it... they just implement an interface and the rest is done for them. You can have a mesh network of peer-to-peer nodes up and running within seconds, and the actual node developers don't even know they've done it.

No crappy WSDL's. No UDDI. No thick & slow SOAP XML transport over HTTP. Bytecode and serialized objects all the way. We're not just talking a faster mode of transport, we're talking about delving down two entire network layers.

The application for such technology is mind-boggling... which maybe is why people aren't using it as much as they could. Damn shame, too.