[geeks] Misuse of Java

Joshua D Boyd jdboyd at cs.millersville.edu
Wed Nov 6 15:44:30 CST 2002


On Wed, Nov 06, 2002 at 01:51:21PM -0600, Jonathan C. Patschke wrote:
> On Wed, 6 Nov 2002, Greg A. Woods wrote:
> 
> > From the studies I've read (possibly biased studies -- I wouldn't know
> > how to tell the difference :-), Java code is inherently slower than
> > pretty much every other interpreted or byte-compiled (or object-
> > compiled, though I haven't a clue what the latter could really mean)
> > code.
> 
> It's pretty darned slow without JIT-compiling.  However, since JIT's so
> prevalent these days, classic VMs aren't really a consideration for most
> applications.
> 
> The bias goes both ways, though.  Someone (who -should- have known
> better) once told me that if Java didn't have garbage collection, it
> would actually be faster than a compiled binary!

Let us not forget the fact that Java doesn't have to be compiled to byte
code.  It can be compiled to machine code.
 
Unless things have changed dramatically recently, things like Java3D are
a real pain on platforms other than Windows and Solaris.  I never got
into liking the JavaGL package, since it had to be installed instead of
coming with Java. 
 
> [1] Honestly, I think Microsoft got this more right than anyone.
>     Layouts are done in "dialog units" which are derived from the height
>     of the system font, so you've got some amount of guarantee that your
>     widgets will scale sanely with the resolution of the display.

Personally, I like the GTK way of doing things with containers in
containers, etc, all sorting themselves out and scaling nicely, for the
most part.  But maybe that's just me.  I haven't really gotten into MFC.

-- 
Joshua D. Boyd



More information about the geeks mailing list