[L2Ork-dev] New developer interested in doing a GSOC18 project

[Jonathan Wilkes]

> This is very much platform-dependent: platforms that do not have native support for 64bit floating point operations would surely suffer.

Turns out that *technically* they wouldn't suffer at all, or at least not
enough that I can notice on my Rockchip Chromebook.

This is because the compiler flags we currently use end up compiling
a binary that almost certainly doesn't leverage the hardware you listed below.

Thus current Purr Data performance ends up looking pretty much the same as
Pd-double (64-bit t_float precision) performance on my Chromebook.

But the point is taken-- if somebody

a) does the work to figure out
which compiler flags trigger the relevant optimizations across archs,
and

b) tests them to make sure they indeed do improve performance,

then those archs that don't have hardware acceleration for double-precision
floating point will have performance that looks worse by comparison
to those future speedups.

-Jonathan