[Omake] Re: [CUCS Bug 545] "or" and "and" take each argument to be a _sequence_ of values.

[Jason Hickey]

Whew.  In any case, if I understand, your proposal is to rename and/or, 
and then replace them with new definitions.  If so, I would suggest the 
following.

     and(e1, ..., en)  == /\_i. [[e_i]]
     or(e1, ..., en)   == \/_i. [[e_i]]
     all(e1 ... en)    == /\_i. [[e_i]]    ("all" is the current "and")
     exists(e1 ... en) == \/_i. [[e_i]]    ("exists" is the current "or")

where [[e_i]] is the value of e_i, interpreted as a bool; it is 
basically $(if e_i, true, false).

Identities:

     - and() = true
     - or()  = false
     - all($(EMPTY)) = true
     - exists($(EMPTY)) = false

Some other cases:

     - and(true, $(EMPTY), true) = false

     - all(true $(EMPTY) true) = true
       (the $(EMPTY) is not significant in array formation)

     - Forcing the $(EMPTY) to be significant:
       X[] =
          true
          $(EMPTY)
          true
       all($X) = false

The main issue (otherwise I believe I would be in agreement), is 
replacing and/or with backwards-incompatible definitions.

Jason

-- 
Jason Hickey                  http://www.cs.caltech.edu/~jyh
Caltech Computer Science      Tel: 626-395-6568 FAX: 626-792-4257