search

ian-ross / arb-fft

Pure Haskell arbitrary length FFT library
https://www.skybluetrades.net/haskell-fft-index.html
BSD 3-Clause "New" or "Revised" License
11 stars 1 forks source link

BSDify (perhaps?) #1

Closed cartazio closed 10 years ago

cartazio commented 10 years ago

Just throwing it out there :)

either way, this is a lovely lovely code base that is probably one of the nicest i've seen. Also really captures a lof of the ideas clearly

cartazio commented 10 years ago

(reading the fftw papers etc for ideas and then completely paraphrasing them as you've done is enough to make it a nonderivative work, but i'm not a lawyer, though I can ask one)

ian-ross commented 10 years ago

Thanks for the kind words!

As far as licenses go, I'm actually kind of on the fence about things. I quite strongly believe in the Free Software mission, and I'm also quite a fan of dual licensing for open-source vs. commercial users, but I can also see the point of making things less burdensome for commercial users. Like you, I'm still trying to figure out the best way to make a living from this sort of thing! I did the FFT stuff (and I'll be doing more like this) partially as marketing for my mad Haskell and math skillz. GPLing it for the moment is kind of a way of reserving the right to try to make some money from it more directly. I don't know whether that's a good way to go or not. We'll see. I'm quite open to loosening the licensing in the future, but while I'm working this stuff out, I wanted to start with the GPL as a default.

As for papers and paraphrasing, a lot of the stuff I've written I just made up myself. I figured out the step from the standard 2^N to the mixed-radix case myself while out walking the dog; the same goes for most of the implementation of Rader's algorithm; and the same goes for most of the combinatoric stuff for planning. (I think people around where I live are used to seeing me wandering through the forest with my dog with a distracted look on my face as I do maths in my head...) I took a look at the FFTW paper to get some idea of what they did, then stepped back from it, thinking "OK, too clever -- let's start with something dumb". In terms of IP, the code is definitely a cleanroom implementation.

I also think there's a lot to be said for writing big chunks of exposition. In many ways, I think the blog articles are more valuable than the code. There just isn't very much of this sort of thing out there, showing the development of a medium-sized piece of relatively complex code from beginning to something-like-end.

On 26 January 2014 18:20, Carter Tazio Schonwald notifications@github.comwrote:

(reading the fftw papers etc for ideas and then completely paraphrasing them as you've done is enough to make it a nonderivative work, but i'm not a lawyer, though I can ask one)

— Reply to this email directly or view it on GitHubhttps://github.com/ian-ross/arb-fft/issues/1#issuecomment-33323749 .

Ian Ross Tel: +43(0)6804451378 ian@skybluetrades.net www.skybluetrades.net

cartazio commented 10 years ago

Valid points all. And yes the blog posts are stupendous!

ian-ross commented 10 years ago

OK. BSD it is.

cartazio commented 10 years ago

woot! :+1: :)