[LEAPSECS] drift of TAI

[Adi Stav]

(Hi, I'm Adi, long-time lurker, first-time caller)

On Mon, Sep 15, 2008 at 01:29:13AM -0700, Rob Seaman wrote:

>> In the absence of days and years, though, calculations that involve  

>> only metric units are a lot easier than what we presently put up with.

>

> Hence the prevalence of radians in everyday use :-)


That's a separate issue. You can base your angle units on perimeter or 
radius length, and then you can divide them into decimal or sexagesimal
subunits as a completely independent decision. Several systems dividing 
the circle into decimal units have been propsed and used in practice, 
some still today; on the other hand, it's easy to divide the radian into
sexagesimal units. (I'm surprised that I'd never encountered
such a system -- dividing the radian into 60 "degrees", each divided
into turn into 60 minutes and 3600 seconds, would not only have given
"degrees" that are only slighly (5%) smaller than perimetral degrees,
but also be more consistent with regards to division by 60.)

Actually, our radians don't require any particular radix at all, we use 
them as whole units. A computer might display a radian in decimal
notation, yet stores them in binary floating-point representation, or as
nominator-denominator pair, or whatever. Nothing decimal about them.