Re: Digit strings

[David Cortesi <cbmvax!uunet!INFORMIX.COM!cuvmb.cc.columbia.edu!cortesi>]

After others described problems with expressing big numbers,

On Tue, 4 Feb 1992 08:39:54 -0500, Bob LeChevalier wrote:

> ...we designed the language
> to make it easier to clearly say simple numbers with few
> significant figures.... When you want to give many sig.figs.,
> you will have to be cumbersome, or
> do a slight variation on your approach and say 1x10**24+111 using the full
> MEX grammar....

The gismu list contains a set of words comparable to "million" etc.,
and they all have rafsi.  The following is probably not grammatical
(I haven't checked the ebnf) but could (should?) there not be some
way to use these in a way analagous to common English usage?

        * pareci gig vomuxa meg zebiso
        123 (american billions) 456 millions 789 ?

Not precisely that way no doubt, but...?

In looking into this possibility I note that the rafsi for kilto (1E3)
is ki'o, which is also the number comma.  Although the uses are related
they are distinct,  which leads to a potential problem.  Suppose
you wanted a term for what Americans call three billion and the British
call three thousand million: wouldn't it be "ci ki'omegdo" ?

Now magically this means the same no matter how you parse it
(as 3,000 million or as 3 thousand-million) -- and granted, you
could always stick in "boi" if you were determined to say 3,000 megdo
-- but there is syntactic ambiguity when you don't. Sooner or later
there will be a time when the distinction will matter (I just
can't think of one now...)