Alexandre, What numerical techniques are important to you? How large are the typical data sets, grids, etc? I do a lot of processing on 500,000 sample signals, so Smalltalk is out for most of that number crunching (FFT, DWT, norms, etc.) and visualization. So far, I could probably get away with Smalltalk curve fitting, but that won't always be the case. One option would be to tell NR and GSL "thanks but no thanks" and find or start an MIT numerics library. In my mind, it should be C-callable C++; the trick to that is proper use of extern "C".
Just to make it clear, I have large parts of GSL mapped into Pharo. My concern now is how to release that w/o being adversely affected by GPL; help with understanding "GPL compatible licenses" would be appreciated. Like I said, I don't care what anyone else does with the code, but I won't have years of my work rendered useless to me for commercial interests. Nothing in GSL is worth taking that risk. There should not be any problem releasing the function definitions; the question is over how much, if any, of the wrappers I can release w/o losing it all to GPL. Bill ________________________________________ From: [email protected] [[email protected]] On Behalf Of Alexandre Bergel [[email protected]] Sent: Friday, November 05, 2010 7:24 AM To: [email protected] Subject: Re: [Pharo-project] Pharo and Astronomy? Thanks Bill. I am at the stage of convincing Astronomers to use Pharo to do their soft, but this is not easy Cheers, Alexandre
