The Duodecimal Bulletin, Vol. 55, No. 1, Year 1209 [pdf]
dozenal.org
dozenal.org
With the advent of modern AI tools, this question has never been more important.
I bet that annual meeting they held in that wee room back in 1983 was riveting.
(Of course any squabbling is instantly forgotten the moment they have to act against their common arch enemy, the Hexadecimal Society)
When the exact representation of frequently used rational numbers is irrelevant, base 2 has no competition.
If you want to represent exactly more rational numbers than with bases 2 or 10, than either base 30 shall be used (= 2 * 3 * 5) or bases that are multiples of 30, like the traditional 60 or like 240, which fits well in a byte.
> the problem is that Latin uses base ten, so bases larger than ten end up with names that put a bit too much of an emphasis on their relationship with decimal: undecimal, duodecimal, tridecimal, etc. people who like base twelve like to call it "dozenal" instead of "duodecimal" for this exact reason. these names are simply too biased in decimal's favor. ideally, every base should have a unique name that reflects its properties, rather than trivial information about its size.
More like the need for workers and that is a problem.
The article on page 38 is really funny to anyone not in the US:
Fahrenheit temperature usually ranges from about 0° (cold) to about 100°
(hot). On the other hand, those who use the awkward Celsius scale usually range from
about 18° to about 38°! Interesting.
And apparently only the metric system was imposed by tyrannical governments. Maybe someone could ask the people in metric countries today if they would like to go back to the "natural" measurements that were in use before that happened? And maybe also switch to counting everything in dozen and gross at the same time.
Even if that really were objectively a better system, I think few would make that change if it wasn't forced on them.
(or are upside-down digits their way to mark icky base-10 numbers if they have to write them?)
Edit: ah, they explain it on page 23.
31 Oct is 25 Dec
Seems obviously wrong, or is that yet another dozenal notation, where what looks like the digit three is really a one? Because it should have been real easy to avoid mistakes like that for an entire decade by just remembering that 1190 = 1980 decimal (next time the decades and dozen-years align like that will be in 2040).