Re: scons check breaks with latest version of gpsd

[Greg Troxel]

"Gary E. Miller" <gem@rellim.com> writes:

>> I would expect that all the rounding is specified between IEE754 and
>> POSIX.  So was this a glibc bug fix?
>
> C99 and ISO/IEC 9899:TC3 (Draft C 2007) say of printf("%f"):
>
>     The value is rounded to the appropriate number of digits.
>
> That does not explicitly explain 0.00 and -0.00.

This is tricky, because C99 doesn't (I'm pretty sure) require IEEE754,
but on the other hand the only computer I'm aware of people using that
doesn't do IEEE754 is a Vax.

https://en.wikipedia.org/wiki/Signed_zero

> POSIX 2007 says it include C99 definition of pritnf() by reference.

Interesting.  So I'd say that on a computer using IEEE754 the printf
rounding has to follow the 754 rules.

> There has been at least one other RasPi report of rounding issues:
>
> https://gitlab.com/gpsd/gpsd/-/issues/65
>
>     "lat":41.575030049
>
> vs:
>
>     "lat":41.575030048
>
> But that was on Debian Buster Pi that no one else could duplicate.

Interesting, and I wonder if tests should just have some sort of
epsilson/close-enough in numerical comparison vs bit-for-bit ascii
compare -- but I realize that's a big change.

signature.asc
Description: PGP signature