Richard Fateman schrieb: > > On 11/5/2017 4:21 AM, email@example.com wrote: > > I don't see why Roman numerals couldn't be handled in the same manner, > Roman numerals are not positional. So it wouldn't be "the same manner". > Also, is x*x x^2, 10*x or 100?
Using Roman numerals one has x*x = x^ii = c.
That "Roman numerals are not positional" will affect the conversion algorithms, but does not bear on when to interpret input or display output as Roman numerals.
A problem is the distinction of Roman numerals from variables, which also arises for non-Roman numerals with base greater ten. In Derive, a hexadecimal number like ace (= 10*16^2 + 12*16 + 14) must be written as 0ace to distinguish it from a variable. Thus, with input switched to hexadecimal, ace is interpreted as a variable and 0ice is interpreted as 0*ice (= 0), where ice is again a variable.
For simplicity, one might use the zero prefix for Roman numerals too; then all variable names would remain valid, and Roman numerals could even be mixed with decimal (or dual or octal) ones, as in your example expressions. Input like 0xxyy should become 0xx*yy (= 20*yy).
> > > but I doubt that it would it be worth the effort. > For sure. Unless a CAS is used for clock faces or tombstones.
PS: Derive actually allows to use any base from 2 to 36, by specifying InputBase := OutputBase := 36 for example. The limit of 36 arises because there are only ten decimal digits plus 26 ASCII letters.