Re: Grand Unified Theory of Date/Time modules

[Abhijit Menon-Sen]

On 2001-05-06 13:29:00, [EMAIL PROTECTED] wrote:
>
> > Given that the average case for the format I proposed is 128 bits[*]
> > [...]
> 
> Hmm? second-precision TAI is 64 bits, nanosecond is 64+32 and
> attosecond precision is 64+32+32. how is this ever going to average to
> 128?

Sorry, I meant the ASCII representation, not TAI. YYYYMMDDTHHMMSSZ is
128 bits long.

> In any case, I think building our base date/time module around libtai
> would be a good idea.

Perhaps so. I'll write Time::TAI and we'll see how useful it is, ok?

- ams