Infinity and infinities -- effects of Motion 3
(A discussion is ongoing in the ER subgroup, but parts of this are
relevant to the main group, hence the CC:)
The main thing I get from Motion 3 is that it resolves the ambiguity
inherent in the 754 treatment of infinity: exact inifinity in R* as
opposed to overflow within R due to the representation range.
Motion 3 makes it clear that only the second interpretation applies
(in IA, which differs from FPA in many ways anyway).
Dan Zuras also offered this amusing item, in reaction to a comment
from Mike Cowlishaw that DFP can represent many infinities:
I joked that we could have them stand for Aleph0, Aleph1,
Aleph2, ..., Aleph(Emax) but that nobody would have any
use for any but the first 3.
Well, many many more than that: 2*Emax/3 * 10**(p-1)
Perhaps the combination field could encode ordinal levels...
Michel.
---Sent: 2009-04-21 14:17:36 UTC