Note: www.cdegroot.com is in rebuild. Please accept my apologies for broken links, missing stuff, etcetera - more
  Home

One day in 1998, a bunch of engineers walks into the Sun corporate marketing department. They start rambling about some exciting piece of software they cooked up, and that the bosses told them to chat with the marketing types about its introduction.

After a short first presentation, a bewildered marketeer manages to summarize the product: "So this is something about services."
"Yeah, a network of services that spontaneously find each other and connect and can do all sorts of cool things!"
"What sort of cool things?"
"Well, you know - like Ken is hacking on this great implementation of Linda tuple spaces, the stuff they did at Yale. Really neat!"
"Tulip what?"
"Forget it. Oh yeah, and we do distributed transactions and stuff."
"Great. Now tell me how I'm going to put McNealy on stage at JavaOne demoing distributed transactions."
"Errr..."

10 minutes of miscommunication later, one of the engineers blurts: "Well, we could also make devices talk to each other with Jini."
"Devices?"
"Yeah, you know: like printers and cameras - plug them together and print pictures from the camera to the printer."
"Aah! NOW we're getting somewhere. We'll ring these Sony boys, they're crazy enough for such an idea, and THAT's something we can give to Scott for one of his stage acts..."

And so it was decided: Jini would be introduced by showing it off on devices - much more visual, and with the message that this technology would be so easy that even Sun's CEO could connect stuff together. Sadly enough, turnover at the marketing department was so high (Boo.com hired a massive amount away) that half a year later, none of the marketeers that were present at that meeting were still around, and the notion was that Jini was just for devices.

As you may have guessed, I made this up (even the bit about Boo.com ;-)). However, I assume that something like this happened. And it did quite a bit of damage to Jini, because the non-device people leave it alone, and the device people are complaining about the fact that Jini's requirements are too big for their limited processors (Jini currently needs a full Java2 platform to run, and not many devices can handle that).

It was and is wrong. Jini is for services. Services implemented by bits of software, bits of hardware, or a million monkeys hammering away on typewriters - it doesn't matter. If you can express your problem in terms of services, chances are that Jini can help you, whether you are building light switches or enterprise resource planning software.


 
Copyright (C)2000-2011 Cees de Groot -- All rights reserved.