peer-to-peer Amirouche Boubekki (05 Oct 2019 12:24 UTC)
We need a pre-SRFI list hga@xxxxxx (05 Oct 2019 12:41 UTC)
Re: We need a pre-SRFI list Arthur A. Gleckler (05 Oct 2019 19:14 UTC)
Re: We need a pre-SRFI list hga@xxxxxx (05 Oct 2019 20:20 UTC)
Re: We need a pre-SRFI list Duy Nguyen (06 Oct 2019 01:47 UTC)
Re: We need a pre-SRFI list elf (06 Oct 2019 01:51 UTC)
Re: We need a pre-SRFI list hga@xxxxxx (06 Oct 2019 02:18 UTC)
Re: We need a pre-SRFI list elf (06 Oct 2019 02:33 UTC)
Re: We need a pre-SRFI list Arthur A. Gleckler (06 Oct 2019 04:57 UTC)
Re: We need a pre-SRFI list hga@xxxxxx (06 Oct 2019 11:42 UTC)
Re: We need a pre-SRFI list Amirouche Boubekki (06 Oct 2019 06:09 UTC)
Re: We need a pre-SRFI list Arthur A. Gleckler (06 Oct 2019 17:30 UTC)
Planning how to organize Scheme discussion Lassi Kortela (06 Oct 2019 17:48 UTC)
Re: Planning how to organize Scheme discussion hga@xxxxxx (06 Oct 2019 19:41 UTC)
Re: We need a pre-SRFI list Arthur A. Gleckler (06 Oct 2019 18:30 UTC)
Re: We need a pre-SRFI list Lassi Kortela (06 Oct 2019 19:31 UTC)
Re: We need a pre-SRFI list Amirouche Boubekki (06 Oct 2019 19:48 UTC)
Re: We need a pre-SRFI list Amirouche Boubekki (06 Oct 2019 19:56 UTC)
Re: We need a pre-SRFI list elf (06 Oct 2019 01:53 UTC)
Re: We need a pre-SRFI list Vladimir Nikishkin (06 Oct 2019 03:06 UTC)
Re: We need a pre-SRFI list Duy Nguyen (06 Oct 2019 04:13 UTC)
Matrix libraries Lassi Kortela (06 Oct 2019 14:51 UTC)
Re: Matrix libraries John Cowan (06 Oct 2019 17:55 UTC)
Who's working on what? Lassi Kortela (06 Oct 2019 19:39 UTC)
Re: Who's working on what? Amirouche Boubekki (06 Oct 2019 20:19 UTC)
Re: Who's working on what? Amirouche Boubekki (06 Oct 2019 20:26 UTC)
Re: Who's working on what? John Cowan (06 Oct 2019 20:40 UTC)
Re: peer-to-peer Amirouche Boubekki (05 Oct 2019 14:43 UTC)
Re: peer-to-peer Arthur A. Gleckler (06 Oct 2019 05:14 UTC)
Peer-to-peer, sockets and binary s-expressions Lassi Kortela (06 Oct 2019 12:41 UTC)
Re: Peer-to-peer, sockets and binary s-expressions Amirouche Boubekki (06 Oct 2019 13:46 UTC)
Re: Peer-to-peer, sockets and binary s-expressions John Cowan (06 Oct 2019 20:35 UTC)
Re: Peer-to-peer, sockets and binary s-expressions Vladimir Nikishkin (07 Oct 2019 02:42 UTC)
WebSockets Lassi Kortela (06 Oct 2019 12:47 UTC)
Re: WebSockets Per Bothner (06 Oct 2019 14:40 UTC)
Re: WebSockets Amirouche Boubekki (06 Oct 2019 19:53 UTC)

Matrix libraries Lassi Kortela 06 Oct 2019 14:51 UTC

> Doesn't slib include pure-scheme matrix algebra and fft already?
>
> To do anything fast nowadays you need gpu acceleration, and I'm not sure
> it's easy to rewrite from scratch. So the potential srfi would need to not
> just include all the functions, but possibly support different backends,
> i.e. pure scheme, blas, gpu-blas, etc.
>
> I wonder if that would be really a good thing to include in the language
> itself (even if -large). But I'm not an expert by no means.

Fully agreed.

For workaday array/vector support in Scheme: Per, Marc, John and
possibly others have written several SRFIs.

For the heavy lifting and fancy math, gobs of brainpower and R&D dollars
have gone into things like CUDA, BLAS, Torch, TensorFlow, R and
Matlab/Octave. There is no conceivable way we can approach that level of
features and performance from scratch. "If you can't beat them, join
them" would be the winning move here :)

For prior art on multi-backend matrix stuff that integrates into Lisp,
look into Clojure's core.matrix and Incanter.

One thing that's easy to forget in these discussions is storage formats.
It's not that useful if you can't ship matrices from those common
packages into Scheme and out again. People who are accomplished with
this stuff won't want to use Scheme as their only tool. We should have
effortless compatibility with established formats.

At some point, standard network protocols will probably emerge for
shipping ML/statistics batch jobs to GPU clusters and getting out the
results. When they do, we should jump in.