On Thu, 14 Apr 2016 22:55:34 +0200, Eric Barnhill wrote:
I don't see commons-math as legacy code.
It depends on the meaning of "legacy code":
https://en.wikipedia.org/wiki/Legacy_code
The first definition in the above reference could very apply to Commons
Math, in particular with respect to development not following the
evolution
of the Java language.[1]
Regards,
Gilles
[1] I know that the JVM will always be able to run old bytecode. The
question
is whether anyone will want to run old (byte)code if doing so
result in
loss of performance or the necessity to comply with outdated API.
Just to look around my
corner of the world, ImageJ has a solid place in the biomedical
imaging world and commons-math is its workhorse. Second, I think Java
is, or should be anyway, a natural step outside of scientific
computing environments like Matlab or Octave for when the benefits of
compiled code are important. I think it's a lot easier to deal with
in-Matlab Java than .mex files, with their silly interface and their
planned obsolescence. I think the only reason Matlab-Java is used so
much less is because the Matlab interface is so buggy and lousy. I'm
planning to start contributing to Octave this year and see if we
can't
start outperforming Matlab on Java integration pretty quickly.
On 14/04/16 15:45, Gilles wrote:
Hi.
I think that we could list quite a few things there:
https://helpwanted.apache.org/
But that would require that we know what we want in the first place.
Where is this project heading to?
Is this "legacy" code?
...
Regards,
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
