Here's more info from this year's JavaOne 2004 in San Francisco. The following link is a good summary of the J2ME technology (CLDC vs. CDC).

See: JavaOne 2004 Online

It's always been a common question: "What's the difference between J2ME CLDC vs. CDC?"

It's pretty straight-forward: CLDC is for currently shipping low-end cell phones. CDC is for higher-end smartphones and more capable consumer devices like TV set-top boxes, automobile dashboards (telematics), high-end PDAs, etc.

Time out for a JavaOne 2004 break. Lots of interesting stuff going on at JavaOne 2004 this year. Good opportunities for J2ME technology especially exist.

See: http://java.sun.com/javaone/general_sessions1.html

J2ME CDC technology on high-end smartphones will give an edge to next-generation wireless devices. Adding Swing and Java2D graphics from JSR-209 is a secret weapon that next-gen developers will use to differentiate themselves from the low-end MIDP phone market.

It'll be interesting to see the apps that develop on those next-gen smartphones...

OK. No votes. I get to stick with my first pick at a sample Web Service. Let's move to step #2: "Simulate":

1. Postulate - Take a guess what the market wants.
2. Simulate - Don't write the full functionality yet. Write a skeleton that emulates the product.
3. Demostrate - Show some potential customers and some non-potential customers and get their feedback.
4. Integrate (feedback) - Filter what comments are good and change your simulation.
5. Iterate - Go back to step 1 or step 3 and do it all over again.

My choice of example Web Service:

Locator service of family members - Where's dad? Where's mom? Where are the kids? (Maybe GPS coordinates or a map with icons)

Here's some skeleton code:

// Locations are stored as simple (x,y) coordinates for a map for each family member

public class FamilyLocation {
  String name;
  int x, y;

  // Web Service Methods
  String getName();
  Dimension getPosition();
}

The above is pretty straight-forward and is all we need to simulate the Web Service for now. The idea is to get a prototype interface done quickly and let it evolve instead of working on it for years to try to get it exactly right the first time. Software is a living, growing thing--especially Java software. ;-)

Lets look at the first step in the list: "Postulate":

1. Postulate - Take a guess what the market wants.
2. Simulate - Don't write the full functionality yet. Write a skeleton that emulates the product.
3. Demostrate - Show some potential customers and some non-potential customers and get their feedback.
4. Integrate (feedback) - Filter what comments are good and change your simulation.
5. Iterate - Go back to step 1 or step 3 and do it all over again.

Guessing what the market wants based on trends in the news, trade periodicals, and market studies are one way to postulate a new Web Service.

But, sometimes it's just a question of what would make sense as a product or service in your own daily life. I came up with these items for a possible Web Service:

1. Locator service of family members - Where's dad? Where's mom? Where are the kids? (Maybe GPS co-ordinates or a map with icons)
2. Price/info lookup service - You give a Bar Code # and you get back the price and product info.
3. Parking service - Give GPS co-ordinates or zip code and get empty parking spaces or available parking lots in your immediate area.
4. Stock price service - Self-explanatory.
5. News service - Get the latest headlines.

I listed the above in descending order of preference. But, jumping to step #3, I'd like to get some feedback. Post a comment to vote for one of the other items (other than "Locator service of family members") from my list above or add your own item.

I'll gather some feedback and pick which Web Service I'll use for this exercise.

Writing a Wireless Web Service is currently harder than it should be.

See: http://java.sun.com/developer/technicalArticles/J2EE/j2ee_ws/

The above is too low-level, too complex, and frankly too boring! Instead Web Service design should start with high-level concepts with an element of fun:

1. Postulate - Take a guess what the market wants.
2. Simulate - Don't write the full functionality yet. Write a skeleton that emulates the product.
3. Demostrate - Show some potential customers and some non-potential customers and get their feedback.
4. Integrate (feedback) - Filter what comments are good and change your simulation.
5. Iterate - Go back to step 1 or step 3 and do it all over again.

A better and faster way to design a Wireless Web Services is to simulate the Web services on an actual device with a demo program. This gets you Usability information right way and gets you feedback from people right way so that you can iterate on a design quickly. It lets you know what people think is fun and what is boring about your design, which is important these days in the wireless device market.

Wireless Web Services are not used much now because engineers are making it much too difficult.

I'll walk through an example in the next postings...

The Digital Home Working Group (now called the Digital Living Network Alliance (DLNA)) is standardizing the guideline for a home entertainment hub (gateway). There was an article in theregister.co.uk, "Digital home group touts convergence spec"

See: http://www.theregister.co.uk/2004/06/23/dlna_spec_launch/

Cool stuff! That's a market-driver: Home use of a "hub" that controls distribution of images, movies, and music around your house (probably wirelessly with 802.11a/b/g). J2ME CDC is in the Shell Home Genie gateway (click on the photo to above).

Good driver for J2ME CDC.

Also in the New York Times, they report on the next wave of high tech start-ups in Silicon Valley that are doing well and driving version 2.0 of Silicon Valley.

Look very closely at the list of companies:

• Google
• Six Apart
• LinkedIn
• HP
• Cisco
• Vontu
• Cassatt
• Agile Software
• eBay

See: http://www.nytimes.com/2004/06/22/technology/22VALL.html?pagewanted=1
(Free registration required)

I'm worried that if Scott and Jonathan aren't increasing Sun's compensation packages and benefits (especially health bennies) here, there will be a brain-drain away to these fun-sounding, up-and-coming start-ups with full parking lots, better potential (as seen by VCs), and better business models.

It will be a fight again to keep the best talent at the high tech companies here.

The New York Times also sees that cell phones should be cell phones (voice) first above all else. They also used my Star Trek Communicator comparsion (hmmmmm...)

"Manufacturers are not only making smaller, lighter, slicker phones, but they are also adding features that make yesterday's models look like the clunky communicators on old Star Trek episodes."

See: http://www.nytimes.com/2004/06/20/business/yourmoney/20cell.html
(Free registration required)

I liked this quote and feel the same:

"But intriguing as those features are, will the average person really use many of them? Probably not. Besides, they drain batteries, and too many demands on the computer chips inside a phone can sometimes weaken reception."

This is cool. Using a J2ME cell phone, you can tag locations of things, like your favorite restaurant, or a gas station selling gas under $2/gal (are there still any???) :-) or a police speed-trap, then broadcast it for others to see the important info on a map on their cell phones.

See: http://wavemarket.com/products.html

Good use of J2ME tech.

But, kids will probably wind up using it to keep track of the vice principal I bet! :-) Another reason to keep cell phones out of schools.

A breaking technology news story today is that scientist can now teleport an atom from one location to another.

See: http://www.abc.net.au/science/news/stories/s1133890.htm

Of course, Star Trek came up with the idea first. It seems lots of good ideas start out in popular science fiction.

Java and J2ME technologies would do well by studying what works and what doesn't work in the Star Trek world. The Star Trek communicator is a good example. It's simple and intuitive to use. Tap once to have a voice-activated device to talk to anyone, anywhere on the starship Enterprise (enterprise, get it? a metaphor for the Enterprise :-) ).

But, what everyone doesn't get is that the best use of a device like that is audio-only. No video, no games, no applets, just simply audio. The smartphone of today is slow to be adopted because it tries to do too much in a small device. It's like the crosstrainer athletic shoe: tries to do everything (be a running shoe and a tennis shoe and a basketball shoe), but winds up doing everything poorly, since it can't concentrate on a core compentency.

Important lesson from Star Trek when desinging a communicator. J2ME technology really belongs on a tri-corder, not a communicator.

Nokia is going after the wireless Enterprise market.

It'll be an uphill fight, since the Enterprise market is showing signs of very slow adoption of cell phone usage to access corporate data.

See: http://wireless.newsfactor.com/story.xhtml?story_id=25412&category=hndhld

Most wireless Enterprise workers use a laptop PC, not their cell phone to access Enterprise data. The RIM pager and PalmOne Treo 600 smartphones are the exceptions, but are a niche market and are used primarily for corporate e-mail access.

Nokia will find that the Consumer market will continue to dominate the revenue stream for cell phones, while the wireless Enterprise market will flounder for a while, since Enterprise workers can continue to use their wireless laptop PCs with faster 802.11b/g connections and are able wait until they get to their office PCs with also much faster and more secure connections to their Enterprise networks.

Anytime, anywhere Enterprise access sounds nice in theory, but in practice not every Enterprise worker wants to be connected to work 24/7. There are other things to life, ya know!

Here's an example of a cool use of Java technology. It reminds me of Harry Potter's Marauder's Map from the 3rd book/movie, Harry Potter and the Prisoner of Azkaban.

It's an interactive map of the arriving and departing flights in and around San Jose Airport. Follow this link and click on the airplane to see more info about it.

See: http://www4.passur.com/sjc.html

The cool part is that it is a Java applet and J2ME CDC/Personal Profile can run the same bytecodes on a PDA or high-end smartphone.

Convergence of the Internet multimedia (movies, music, & content) on computers and in the living room is happening today. The wireless Enterprise market is sitting back and slowly forming. It looks like the consumer market will drive the networked small device market first, with corporate wireless usage trailing until a better understanding is formed of how mobile devices help the Enterprise worker.

The consumer market knows that flashy and fun drive its users. So, that is easier to understand.

See: http://www.businessweek.com/magazine/content/04_25/b3888601.htm

How is J2ME technology going to play in this space? J2ME CDC will be the driving tech used to tie together all the consumer products under one open standard. It'll be a fast evolving space... But, the winners will be nimble and quick and use J2ME CDC right off the bat.

RFID tags are getting more and more popular in pilot studies and prototypes for the next networked way of keeping track of things from pets to ranch animals to Walmart products on the shelf.

See: http://www.skyetek.com/demokits.html

So, what do you do with RFID tags?

With a kit like the above, a 8-page business plan, and a trip to the venture capitalists on Sand Hill Road in Menlo Park, you'd make it big in about 2-5 years. Of course, you should use J2ME technology for your prototype... but, that's just a technicality. :-)

It's been said that the 3G in 3G cell phone networks stand for: Girls, Games, and Gambling.

See: http://www.techcentralstation.com/122001A.html

So, what's going to be the killer J2ME app?

Who knows? Maybe a poker/blackjack MIDlet where you play for Fry Electronic's gift certificates against a virtual Pamela Anderson... Hey... I wonder if I can copyright that..? :-)