commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From S├ębastien Brisard <sebastien.bris...@m4x.org>
Subject Re: [Math] Recursive Iterative approach on FFT
Date Tue, 14 Feb 2012 06:25:15 GMT
Leandro,
you should certainly not feel so bad. Trying to speed up some of the
already implemented algorithms is a noble task; it's true that
benchmarking is sometimes tricky [1]. You seem to be bursting with
ideas about new developments for Commons-Math. These new developments
would certainly benefit the whole community of users. You also seem to
be very (very, very) enthusiastic, which is great! However, enthusiasm
should not be confused with haste!
To cite but one example. My first contribution to CM was iterative
linear solvers. Before the first line of code was actually committed,
several months of discussion elapsed. That could seem frustrating at
first, but it was in fact very, very exciting, and instructive for me.
I learned a lot (and still do) from those discussions. These long
discussions also ensured that everyone kind of agreed about the API
design, which improved a lot in the course of a few months.
To close this message, please allow me to pass over to you the advice
I recieved when I joined: start with small, simple patches. As your
knowledge of the library and the CM practices, you will also gain
confidence, and will be able to send more complex patches.
I am sure we are all looking forward to reading one of yours. Good
luck for that, and a warm welcome!
S├ębastien

[1] Part 1 of this paper
http://www.ibm.com/developerworks/java/library/j-benchmark1/index.html
is quite good (part 2 is not so instructive).


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message