[OT] RE: M$ banging nails into java's coffin?

Richard A. O'Keefe ok at atlas.otago.ac.nz
Mon Jul 23 01:12:40 UTC 2001


Edmund Ronald <eronald at rome.polytechnique.fr> wrote:
	
	M$ is doing a Netscape Act II here. They are killing Java's run-anywhere
	distribution model. After this it becomes just-another-C-type-language. It
	may be very succesful as a programming language. But it will not threaten
	M$ with the prospect of application programs running on screens that have
	not paid the M$ tax. Ultimately, however, the Java people can only blame
	themselves - their language was too cumbersome to prosper in its niche
	despite all the help it had.
	
If you knew how many W3C "standards" were strapped tightly into a Java
straitjacket, you might not be quite so sure that Java isn't prospering
in its niche.  Xalan and Xerces are not written in Visual Basic...

The point of my previous message is that people delivering serious applications
in Java on M$ platforms have long had alternatives to M$'s Java implementation.
Yes, people using IE will notice that applets in Java don't work.
Yes, that will be a problem.
No, that won't kill Java.
Heck, browsers never did support the current version of Java (whatever the
current version might have been), and I've still got 1.1.something on my
Mac while I've got 1.3.something on my Solaris box.  Run everywhere never
_was_ true for Java, but that didn't stop it.

Mind you, there are a lot of clueless people out there.
The official web site developer/maintainer for a certain well regarded
computer science department in a certain well regarded univerisity (not
this one, thank goodness, not even this country, a richer one) didn't
believe me when I complained about the characters being teeny tiny on
my 90dpi resolution screen:  this expect seriously thought in 2001 that
the current state of technology didn't allow anything better than 72dpi.
(I happen to know that his own department has many screens finer than that.)

	This should be a lesson for the Squeak crowd:
	1. Make a language which is REALLY run-everywhere (In-browser Squeak demos 
	   run very badly on my OS X TiBook). RELEASE A RELIABLE PRODUCT!
	
Visual Basic doesn't run everywhere.  Doesn't seem to be hurting it.

	2. Make a language which does not iterate through 10 incompatible
	   versions (after release). RELEASE A MATURE PRODUCT !
	
You think that all those Windows versions, are like, maybe, compatible
or something?  Doesn't seem to have hurt M$.

	3  Make a language which does not iterate through 10 different GUI models
	   (remember AWT, SWING ?) RELEASE A FULL-FEATURED PRODUCT!
	
AWT is still there.  Swing (not SWING) requires it.

As for releasing a full-featured product, no can do.
There's no such animal.  There is _always_ a new "essential" feature.

	4. Make a language WITH a GUI builder at release so people can do their
	   little office dev jobs _fast_ . RELEASE A USEFUL PRODUCT !
	
If that's aimed at Squeak, Squeak _has_ at least two GUI builders
(Morphic does most of what a GUI builder does and then some).
I have a GUI builder for Java on my Solaris box; if I did much GUI work
in Java I'd use it.  I've got another couple for Windows sitting on CD-ROMs
that are no use to me because I haven't got a Windows box.  Indeed,
GUI building is at least part of what Java Beans is all about.

	5. Do not threaten Microsoft. DO NOT BREATHE OR ATTEMPT LIFE :)
	
Ah yes.  Let us now praise famous men, and judges that may save us.




More information about the Squeak-dev mailing list