P1788 M0020.01 -- vote NO
I vote NO on Motion 20.01 -- BRA framework for interval comparisons.
(BRA = Binary Relation Algebra)
This is a beautiful mathematical arrangement, but in my opinion it
belongs in the 1788 standard as a Mona Lisa reproduction would fit
in an Algebra textbook.
(Minor beauty blemish, btw: The definition of atomic relations (middle
of page 3 of the Lohez paper) should say "for any *distinct* i and j".)
It might also have been useful if one or two useful examples of BRA
generator sets had been presented, instead of just deferring the choice
to a future motion -- though I don't think it would have swayed my vote.
I actually have some sympathy for the "IntervalComparison" datatype and
the operations defined on it -- but I would (a) keep it as a generic BRA
type (applicable to comparisons, but to other relations as well), and
(b) have it available as a separate package. I just don't think P1788
is the place for it, except as a reference in a programming note.
Michel.
---Sent: 2010-10-03 20:49:17 UTC