2038: only 21 years away

Posted Mar 26, 2017 19:03 UTC (Sun) by fest3er (guest, #60379)
Parent article: 2038: only 21 years away

I think time should be represented as a signed 128-bit value in picoseconds that uses the time of the beginning of the multiverse as estimated as of 1/1/2000 as the epoch. That should allow a computer to store any time from the beginning of the universe to almost any conceivable date in the future (5 exayears is almost inconceivable). Being signed, it would allow the time of the beginning of the universe to be moved backward almost as far as is needed.

2038: only 21 years away

Posted Mar 26, 2017 19:42 UTC (Sun) by liw (subscriber, #6379) [Link]

Can we stop treating time in a simplistic linear fashion, please? Given the general relativity theory, time should be expressed as a (potentially infinitely long) vector of "stretch" factors for different points in the timeline, and a hashmap of such vectors, to allow different points of view, as well as a scrambling function to represent the route of a Tardis as a complicated directed, cyclical graph of points of view of stretching vectors, combined with a seed value for scrambling to represent the Doctor who was using the Tardis at any given, if you pardon the expression, time.

And this is still a simplistic model, since it doesn't take into account retconning for narrative purposes at all. But adding the retcon compressor is left as an exercise for the reader.

After that, we can finally have some peace of mind as far as time representation goes, and tackle the three fundamnetal unsolved problems in computer science, namely cache validation, naming of things, and counting beyond two.