Re: FFT proposal for gsoc 2021

From: Stefan Seefeld (stefan_at_[hidden])
Date: 2021-03-30 15:20:47

Hi Eduardo,

On 2021-03-30 6:58 a.m., Eduardo Quintana via Boost wrote:
> I've written and uploaded a first draft proposal to GSOC-2021 for the
> Boost.Math.FFT library.
> https://github.com/Lagrang3/gsoc-2021/releases/download/v1.0/gsoc.pdf
> I realize now that maybe I should have given more details about the proposal itself,
> like user's interface and implementation ideas.
> That can be fixed in version 2.0.

Your schedule above looks very ambitious, and, I fear, quite unrealistic:

You plan a single week (June 7 - June 13) for

* defining the API for your FFT library
* implementing it
* writing tests for it
* writing benchmarks for it
* documenting (part of) it

Unless the bulk of this is already done, and the tasks merely consist of 
polishing the code and integrating it into a new special-purpose git 
repo, I don't think this will work.

For that reason, as well as for reasons I already outlined in a previous 
post, I strongly suggest you change the ordering (and timing) of your 
milestones, to first focus on an API that can be implemented as a 
wrapper around FFTW (as I suspect that will be the implementation most 
developers will use until you have an alternative that actually 
out-performs FFTW), and only then start to re-implement your API with 
other backends (including your own).

That will allow you to realize that you won't be able to meet all your 
goals, and still make this project a success, i.e. have something 
actually usable. If you end up with a library that works, but doesn't 
perform well, I'm afraid it won't attract any users, which would be a 
pity, wouldn't it !?

Stefan
-- 
       ...ich hab' noch einen Koffer in Berlin...