[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: MARS rise & set times

Margaret Leber wrote:

> Tsk,tsk,tsk. Modern Java runtimes (the only kind there are on OSX) often 
>  compare very favorably with handcoded C for many operations, apples to 
> apples.

Here we go again. Some Java afficianado stridently defends the language 
saying it doesn't *have* to be so slow. Yet I keep on seeing 
applications like Mars24 that burn up my CPU doing hardly any real work 
at all.

> See http://kano.net/javabench/index (warning: rude page heading)

I note the comparison was between Java and C++, not Java and C. I won't 
defend C++ on performance either. My few forays into C++ ended when, 
after very aggressive optimization, I realized I couldn't get more than 
25% of the speed I was already getting with the same algorithm (a 
Reed-Solomon decoder) in C.

> Of course, these numbers would surely be different if C were used; 
> there'd probably be one somewhat better-performing version for whatever 
> platform the developer had at-hand (SPARC?), and everybody else 
> (including you and me, Phil) would see 0% utilization because there 
> wouldn't be a version for their Windows, OSX and Linux machines, nor 
> anything runnable in web browsers. :-)

No, if the application is released as open source for one platform, and 
if it has any general appeal at all, ports to all the major platforms 
would appear very quickly. Some ports are especially straightforward; 
for example, Mac OS X is all UNIX underneath, and it even supports 
fullblown X11. That makes source ports between UNIX, Linux, BSD and Mac 
OS X remarkably easy.

That leaves M$FT Windows as the oddball, but it should die a horrible 
death for the sake of the human race anyway.


Sent via amsat-bb@amsat.org. Opinions expressed are those of the author.
Not an AMSAT member? Join now to support the amateur satellite program!
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org