[DotGNU]Re: Support Java for .GNU

[Norbert Bollow]

> > I know that you want to make it "generalized", but I think we should ellect
> > one primary language/framework/development platform.

Actually there are five areas where a decision of choice
of language needs to be made

1) Primary language for writing low-level, system dependant libraries
2) Primary language for implementing bytecode -> machine code
   compiler and bytecode interpreter
3) Also, we need a bytecode specification
4) Primary language for writing system independant code that
   will be compiled to bytecode
5) And then, the compiler that compiles the language (4) to bytecode
   needs to be written in some programming language.

> Only after the spec is built.  I think I speak for the rest of the core
> team when I say that we need to define where we're going before we
> decide what route we're taking to get there.  There will come a time to
> decide these things, now is not that time.

I agree.  It's not too early to start thinking about what the
options are, though.

For (1) I think C is the obvious choice, because
  - it's good for fast, low-level programming
  - many programmers know C

For (2) and (5) I think C is the obvious choice because that
will allow tight integration with gcc.

For (3) there are several options
a) Java bytecode, possibly with some modifications
b) Microsoft's new bytecode, possibly with some modifications
c) Create our own bytecode spec from scratch, possibly based on
   Knuth's ideas

For (4) I think Java is the obvious choice.

> That time will most likely come soon, though.

Yes.  First we need to know what we want from the bytecode.
Then choose a strategy for the bytecode.  Then make the other
language choices.

Greetings, Norbert.

-- 
Norbert Bollow, Weidlistr.18, CH-8624 Gruet  (near Zurich, Switzerland)
Your own domain with all your Mailman lists: $15/month http://cisto.com
Business Coaching for Internet Entrepreneurs ---> http://thinkcoach.com
Tel +41 1 972 20 59      Fax +41 1 972 20 69      address@hidden