🚀 go-pugleaf

Rick DeNatale wrote:

> Now if Java, which is effectively a poorly implemented stripped down
> version of Smalltalk with a C++ flavored syntax, and lacking a few
> important features (like block closures), can do the job, good for it. I
> have my doubts however, due to it being let out into the cold cruel world
> without all of the support it needs.

I enjoyed you posting, Rick, and agreed with a lot of it, but I have to take 
exception to that last paragraph.

Java is not stripped down Smalltalk.  It differs fundamentally by being 
strongly typed.  In that is much closer to C++, and not just syntactically.

I would call Java "C++ specialized by addition of Smalltalk concepts", namely 
(a) late binding, (b) garbage collection, and (c) the accompanying loss of 
low-level control and performance penalties that these features cost.  I use 
the word "specialized" because Java, like Smalltalk, is a less general-purpose 
language than C++.  I could just as well have said "enriched", however, 
because Java is also more portable than C++, and "higher level" in that it 
provides garbage collection, packages, tighter encapsulation, late binding, 
etc.  Something is lost and something is gained.  Anyway, perhaps we could 
agree that Java is a conceptual child of Smalltalk and C++ -- whether bastard 
or legitimate is in the eye of the beholder and its ultimate social acceptance 
is yet to be determined.  Some bastard children have gone far.

In addition, Java adds something neither Smalltalk nor C++ ever had -- 
anti-virus security essential to the web environment.  I think this is its 
most essential feature vis. a vis. VB or any other contender for a web 
programming language -- it has the best security model to date, which isn't 
hard since no other widely used language has any security model at all, that I 
know of -- telescript being the other noteworthy recent attempt to create one, 
but it failed by being kept too proprietary.  Anyway, no other language will 
supplant Java for web development until it possesses this attribute, since 
security is today the Achilles heel of the web.

> It also is riding the bubble of the
> WEB environment, which while useful, looks like it is being overhyped as
> the universal UI so that companies can use  Netscape as the '90s version
> of a 3270 and get thin clients which they seem to believe will run on late
> '80s hardware that they don't want to upgrade. The backlash to that idea
> seems relatively inevitable, and besides when we're talking about
> browser/web server programming we're really talking about JavaScript which
> really isn't the same thing as Java at all.  I myself predict that Java
> will just be another holding action between the technical staff and the
> bosses.

Some truth to this, too, but they will upgrade all that late '80s hardware 
soon anyway, hardware isn't the issue for IS guys, it's complexity and 
deployment hassles and standards and the way the software industry changes so 
FAST that plague them.  Just as they had about made up their mind that yeah, 
Microsoft was it, they had beat IBM cold and had Novell on the run, and 
Exchange was maybe the safe bet to beat Notes eventually, everything changed! 
 Microsoft was caught napping by the Internet, but they can't be discounted 
yet, so everything is doubtful again -- who're you gonna follow?  The thing 
about those corporate IS guys is they want to FOLLOW someone.

So I don't know what you mean about backlash, here.  I admit I work at 
Netscape but I don't see them lashing back against the web browser concept.  I 
see them demanding much more from it, and if Microsoft does a better job of 
providing that than Netscape then they will eat our lunch and if we do then 
look out Bill, you just lost your platform monopoly for keeps.  MS does have 
awesome resources and we are the little guys.  Don't confuse our market cap 
with our size, there are way less than a thousand Netscape employees.  But we 
do have a lot of help because we're part of the open web standards world and 
we are really just building one or two little key components, and throwing 
those parts out there for use as fast as we can, where everyone else can build 
a whole coral reef that incorporates them.

You are right that Java is useless to anyone except a programmer.  But look at 
what can be built on it, in principle.  In Smalltalk, can you write an 
application that can potentially run on twenty million machines THIS YEAR, if 
it's popular enough?  I mean that twenty million people could be running a 
month from now, if they clicked on a link on your web page?

Come on, it's not the language, it's the substrate it's running on.  The web 
is the greatest distribution channel in the history of the computer industry, 
and for some reason (maybe luck) Java is there first as a plausible execution 
vehicle for applications that run ON THE WEB rather than ON YOUR PC.

It's just early.  There are no compelling Java applications yet.  The best is 
yet to come.  Hell, the first interesting Java program is yet to come, if you 
ask me.

I personally think the inventors of Java did a better language design job than 
the inventors of most of the other vastly successful programming languages, of 
which I would name: FORTRAN (ugh!), COBOL (aaaarg!), BASIC (oof), C (hmmm), 
and C++ (thanks, Barne, C needed help).  But like all of these, and unlike 
countless other fine languages that had much more limited success (Pascal, 
Smalltalk, Eiffel), Java benefits most from being first to market in a brand 
new niche.

So what's your idea for the first great Java app?

-- 
David Christie		Netscape Communications Corporation
dc@netscape.com		501 East Middlefield Road
415-937-4407		Mountain View, CA  94043