Could Python supplant Java?

James J. Besemer jb at cascade-sys.com
Fri Aug 23 08:40:58 EDT 2002


Paul Rubin wrote:

> "James J. Besemer" <jb at cascade-sys.com> writes:
> > How about:
> >
> >     def fn( arg : type ): pass
>
> I prefer
>      def fn (arg(type)): pass
>
> since this is closer to how class declarations work.  Local and global
> variables could also be declared:
>
>    global x(type)
>    local y(type)
>
> Right now there are no local declarations at all, which I think is a
> shortcoming.

Excellent suggestions.  I hadn't figured out how to make the colon syntax work
with non-argument variables.

> I don't know about Python vs. C++ line counts.

I don't either.  Somebody argued a big savings and I explained why don't see
it.

> I do know that my last
> company did a number of big C++ projects before my time.  Then they did
> a Java project.  They had no real Java experience before, and the Java
> development environment sucked compared to the C++ environment.  Lines
> of code was maybe comparable between Java and C++, and coding time
> was probably also comparable.  But as one guy put it, the debugging
> time with Java was at least a factor of 3 lower, despite the crappy
> debugging environment, because there was almost no time spent chasing
> down weird pointer errors or malloc/free errors.  It was all just logic
> errors in the app that were easy to isolate and fix.  Python has similar
> benefits.

I think you make a good point.

I suppose my experience is analagous to some of the Pythonites who say "I
simply don't make THOSE kind of errors."  I have used C and C++ for so long
that I generally don't make pointer errors or malloc errors -- the kind that
can take a long time to fix.  Generally, in C++ I too generally am only
debugging the "logic" errors.

But given your average college graduate or a person who's only used C for a
few years -- yeah, your point is quite valid.  Since they aren't properly
disciplined to avoid the problems, they spend more time fighting them.  And
Java's or Python's garbage collection would be a big win.

Hell, I've already said GC is probably Python's biggest win even for me.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






More information about the Python-list mailing list