[PATCH] Re: New Bug Reports/Feature Requests

[Linus Torvalds]

On Wed, Mar 2, 2016 at 9:44 AM, Linus Torvalds
<torvalds at linux-foundation.org> wrote:
>
> .. to some degree, Lubomirs least-square polynomial would actually be
> superior: it is the same kind of approximation, but it's an
> approximation that has been specialized for one particular gas and
> pressure/temperature range we happen to care about.

Ok, so I just sent out a two-patch series that I think it the
reasonable thing to do for now, and which basically just makes us use
Lubomir's polynomial.

That second patch that switches over to Lubomir's polynomial also has
a suggestion in the footnotes for perhaps getting reasonable values
for common gas mixes.

If somebody calculates the least-squares polynomials for the
individual gases (O2/N2/He) the same way we have for air right now, it
looks like we could likely just do a stupid linear mixing of those for
gas mixes.

Yeah, I have absolutely zero physics to back that linear mixing up
with, and I only tested it for a single case, but it sounds like it
should give reasonable values to me.

But I may well be full of shit.

              Linus