Re: Quantified Expressions: "some"

["Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>]

On Wed, 17 Nov 2010 19:12:53 -0600, Randy Brukardt wrote:

> "Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de> wrote in message 
> news:puezn7v9naom.1nrc5rh8wzm7p.dlg@40tude.net...
>> On Wed, 17 Nov 2010 17:14:27 +0100, J-P. Rosen wrote:
> ...
>>> Actually, It's been decided recently to change "|" to "=>", so it will 
>>> read:
>>> ( for all X in 1 | 3..5 | F'Range => P(X) )
>>
>> Not much better: "=>" means "implies." P is neither implied nor selected.
>> Actually it is antecedent, on the left side. IF kind of X, such that P(X)
>> AND X in the set, THEN something [ ELSE something ].
> 
> "=>" is what SPARK uses.
> 
> I personally agreed with you and voted for neither.

At least they could use "when" instead of "=>".

> But I was the only one.

I'd prefer a set-measure notation:

   P'Min (1 | 3..5 | F'Range)  -- or  P'Inf
   P'Max (1 | 3..5 | F'Range)  -- or  P'Sup

Of course first-class functions with operations on them and sets were more
suitable for this stuff, i.e. "apply f to S, then measure the result with
m:

   Min (P (1 | 3..5 | F'Range))

But it is difficult to do.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de