> >> And, there is at least one Ada compiler available which compiles to
> >> J-Code, so making Ada as platform-independent as Java.
> >More so. Can anybody name one CPU with a JVM that doesn't already
> >have an Ada compiler?
> >Now how many CPUs have Ada but not JVM?
> This is a bit ingenuous. If you were to have a stored program in a database
> (e.g. replace PL/SQL's stored code method (p-code?) with J-code) then you
> really wouldn't want it to be hardware specific (J-Code is a very compact
> representation). The whole point of Java's byte code is it's cross platform
> capability, who would want to squander this in _every_ situation?
I was responding to the much-exaggerated claim of Java's portability.
Sure, it's nice to have the same "object code" run on multiple platforms.
But the Java folks are trying to make us think that the Java _language_ is
somehow more portable than any other. It is not.
Nor is it true (as they try to mislead us to think) that the Java
_language_ is somehow needed to get the benefits of J-Code. Not only are
there at least two Ada to J-Code compilers, but there are ways to generate
J-Code from Cobol, C++, Eiffel, Smalltalk, BASIC, and a host of others. I
forget the URL and NT problems prevent me calling up Netscape right now,
but do a Web search for the page titles "Languages for the Java Virtual
Finally, as the poster prior to me pointed out, Microsoft is applying
their usual tactics to screwing up even the portability that Java does