According to offical history, Java Card turns 10 years old this month. In October 1996, Sun released the Java Card 1.0 specification. This was not a very big event, as only Schlumberger were actively working on it. And at Sun, not many people knew what smart cards were.
The Java Card 1.0 specification was a very short document, only a few pages long. There were a few mistakes and typos, and it was hard to understand what to do with it, but it was there. And the first product was there too, the first Cyberflex. In this first release, the card remained based on a classical ISO file system, and Java cardlets (a Schlumberger trademark) were only used as scripts to perform actions on these files and automate some treatments.
Nevertheless, the important thingwas done: contrarily to what most people thought at the time, it was possible to program a card using a high-level programming language, and to run a standard virtual machine on a card. At that time, with 256 bytes of memory and a few kilobytes of EEPROM and ROM, few people believed in this, and the specification, together with the Cyberflex card, was the proof that it was indeed possible.
This move to Java was not immediately followed by the industry, and Schlumberger’s Austin team remained a bunch of visionaries for a while. A few months later, the Java hype and Sun’s marketing moved things around, and the rest of the industry followed. But that’s another story.
One sad remark: this anniversary comes almost exactly at the time when the Schlumberger/Axalto/Gemalto Austin team moves a bit away from Java Card. Bertrand Ducastel has gone back to Schlumberger, and Ksheerabdhi Krishna and Mike Montgomery may be pursuing other goals (although we would love to see them again at a JCF meeting). The technical lead of this work, Scott Guthery, left Schlumberger a long time ago, but he is still around the smart card indsutry, and he is in particular active around ETSI/3GPP.
No Comments