From: Dave (dfisher_at_pophost.com)
Date: 2002-02-27 13:42:52
On Wed, 27 Feb 2002 01:44:06 -0600, Doug Sutherland wrote:
>Jim already has it set up nicely. It's trivial to install
>SDL and FFTW. I personally think these should remain as
I think that it would be extremely important to keep the other open source
libraries separate, also. That's what makes the entire open source framework
work and takes advantage of everyone's work effort. If we folded the source
into OpenEEG, then we are left with maintaining and/or updating the libraries
ourselves. :-[
John--what are your thoughts about the work you have been doing with the FFT
filters? Are you modifying the FFTW source? I got the sense that you were
really digging into the math behind the analysis, and were coming up with some
of your own implementations. Is that correct?
>separate packages, not integrated into OpenEEG. If all
>people included support libs there would be a lot of
>redundancy and compatibility issues. It's not hard at all
>to install these (on linux anyways). Jim mentioned that
>he might extract the needed parts from SDL and FFTW, I
>think it's not worth the effort and is generally a bad
>idea anyways. People may have good reasons to use both
>SDL and FFTW for other software. If the OpenEEG stuff
>was set up with automake, it would be as simple as
>./configure, make, make install to build all three of
>these peices.
That would be the best way to package our own stuff, but automake and the
configuration tools are a bear to use. I started to climb that learning curve
mountain late last year, and then realized that I did not need it until I had
something to distribute, so stopped the climb. But they do a *great* job at
identifying what is and is not installed on a Linux system. And, as you said,
it makes the whole installation/compilation process a snap.
Dave,
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:39 BST