Hal Murray writes: > I assume the answer depends on the context. For something like POSIX, I'd > prefer something longer than 292 years but might accept that if it solved the > problem and seemed better/cleaner than the alternatives. > > For usage internal to a project like this, I'd be happy with 100 years. > That's assuming it will go through various major changes over time. > > Has anybody done studies of project lifetimes? How many of them stay around > for 10 or 25 years without major changes?
As a corollary, planned obsolescence never goes according to the plan. Equipment using NTP in one way or another could easily outlive more than the timeframe you've set above. For these kinds of fundamental things that get baked into equipment that may not get updated, my personal rule-of-thumb would be whatever time you can reasonably dream up, then multiply by ten. The 292 years above would be smack dab in the middle of that, so it seems to be OK. :-) Dave Mills had it summed up pretty nicely when he was settling for the 128bit format: "the 64 bit value for the fraction is enough to resolve the amount of time it takes a photon to pass an electron at the speed of light. The 64 bit second value is enough to provide unambiguous time representation until the universe goes dim." (he seems to refer to the completion of the nucleon decay and the beginning of the Dark Era in the Heat Death theory). But if you really wanted to go to fundamental limits, even a 256bit time value wouldn't be enough: the fractional resolution would still be short of the Planck time (by roughly 15 bit) and the seconds part shorter than the various theories about the ultimate death of the universe (by an ungodly amount of bits, depending on which theory you subscribe to). Regards, Achim. -- +<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+ Factory and User Sound Singles for Waldorf Q+, Q and microQ: http://Synth.Stromeko.net/Downloads.html#WaldorfSounds _______________________________________________ devel mailing list [email protected] http://lists.ntpsec.org/mailman/listinfo/devel
