[LEAPSECS] "real time" ?

[Mark Calabretta]

On Tue 2012/01/24 16:53:37 PDT, Rob Seaman wrote
in a message to: Leap Second Discussion List <leapsecs at leapsecond.com>


>Might I suggest that "real" is a poor descriptor here (no

>philosophy intended)?  There is a vast prior art of "real time"

>that means something entirely different.


In addition, there is nothing "unreal" about other time scales.
Likewise, "truetime_t" would incorrectly suggest something "untrue"
about the others.


>One facet of the

>idea is to use IEEE floating-point formats to express time

>values.  Whatever the desirability of that, is this really

>the essential concept?


Over the long term, POSIX time_t counts mean solar seconds, though
getting into trouble by counting atomic seconds in the short term.

This time scale counts atomic seconds over the longer term as well
as the short term.  As "taitime_t" or "iattime_t" would likely cause
howls of outrage (and anyway are not really appropriate), I suggest
"atomictime_t", or "atomtime_t".  Or maybe "attomtime_t" or just
"attotime_t" as a play on "atom" and the FP format.

Regards, Mark