The chance of this error would be reduced, I believe, if "order of
evaluation" is described in terms of (implicit) parenthesization.

Ken's treatise Conventions Governing the Order of
Evaluation<http://www.jsoftware.com/papers/EvalOrder.htm> from
1966 is worth reading in this regard.


On Tue, May 21, 2013 at 2:55 PM, <[email protected]> wrote:

> RM>     We sometimes say that J evaluates in a right-to-left order.
> RM>     And that's fairly accurate.
>
> Caveat instructor.  In the last workshop I gave, I learned that
> that is actually fairly ambiguous.
>
> At least two or three participants concluded that
>
>         10 + 6 - 9
>
> yields 13: nine minus six plus ten.
>
> Fair enough, I'd say, given that R->L prescription.  I'll be certainly
> more cautious/explicit when doing the next workshop.
>
>                                                                 Martin
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to