[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Lucent Technologies & Sun Microsystems

You, tedk, wrote on Fri, Apr 04, 1997 at 09:54:49AM -0300:

> >The three agreements, announced at Sun's JavaOne developers' conference 
> >here, were:
> >
> >Lucent Technologies licensed the JavaTM technologies from Sun, setting 
> >the stage for the Inferno network operating system to run Java 
> >applications on information appliances and servers across many networks;
> >
> >Sun became a member of Lucent's Inferno Partners Program to help put 
> >Java-enabled end-to-end network solutions in the hands of Inferno 
> >customers;
> >
> >Lucent Technologies agreed to participate in Sun's industry-wide 
> >initiative to help define Sun's new PersonalJava and EmbeddedJava APIs, 
> >which were announced today. 
> I sort of predicted this a few months ago -- when I said the only problem
> with the Java OS was that  it didn't
> Sun has surrendered !!! 


I am not as enthusiastic about it. There's a _huge_ momentum currently
for Java, created by Sun and pushed by its alleged usefulness for Web
applets and the Net as a whole (although right now, after 2 years or
so, there _still_ aren't many convincing examples of useful Java applets
on the Net - even Javascript has done better!) and that momentum is only
going to grow. If Inferno runs Java programs perfectly, it may turn
to nothing more as a platform to running Java programs, effectively 
killing Limbo and creating a situation when there _will_ be JavaOS -
Inferno will be one! 

I don't want Java to kill Limbo. Limbo is much, much better language
than Java from where I stand (comparative statements about programming
languages are always dangerous and almost always a silly thing to say,
of course, but I'm pretty convinced at this one). Java is much easier
to learn, than, say, C++; however, it presents nothing new from a theoretical
OOPish point of view and its simplicity is often limiting. It's also
horrendously slow, and JIT does not fix that perfectly. Its success is
a success of buzzwords. In its own playing ground, it's inferiour to
Python (I believe there're also recent successful attempts to 
bytecode-ize Perl and provide it with its own "sandbox" no worse than 
that of Java). In our times, however, success of a programming language
is more dependent on advertizing and buzzwords than its real worth
as a programming tool. 

Limbo is interesting (I'm stepping on slippery ground here as I haven't
written any nontrivial programs in it yet due to severe lack of free
time, only read the guides, definitions and source code to some Inferno
programs). Its simplicity doesn't seem to limit it (well,
unions would be great). GC is useful and very well thought-out - it certainly
doesn't irritate the C-programmer part of my soul. It's very well 
integrated inside the OS, and the OS itself presents a number of new, very
elegant solutions to old problems (especially the namespace). 

I am certainly very far from being an expert, but from what I've seen
and felt, Limbo is one of the most important parts of Inferno for me. 
Inferno with Limbo is exciting and probably revolitionary; Inferno 
with Java (Limbo compiler being present, but not really used by anyone)
might remain revolutionary but isn't nearly as exciting.

What do you think?

Anatoly Vorobey,
mellon@pobox.com http://pobox.com/~mellon/
"Angels can fly because they take themselves lightly" - G.K.Chesterton