_. will not be removed and will not be a spelling error.  Aside from 
other issues that have been raised, there is the question of what 
to display when data does contain a NaN (from external sources). 
Answer? _. .



----- Original Message -----
From: Raul Miller <[EMAIL PROTECTED]>
Date: Tuesday, March 18, 2008 8:36
Subject: Re: [Jprogramming] Handling NaN error with #:_
To: Programming forum <[email protected]>

> On 3/18/08, Mark D. Niemiec <[EMAIL PROTECTED]> wrote:
> > "Raul Miller" <[EMAIL PROTECTED]> wrote:
> > > Wouldn't this make testing difficult?
> >
> > For test suites, one can still always hand-craft a NaN:
> >   [ nan =: 3!:2 a.{~225 0 0 0 8 0 0 0 1 0 0 0 0 0 0 
> 0 0 0 0 0 0 0 248 255
> > _.
> >
> > However, it's much harder to create one inadvertently if one 
> can't spell it.
> > It's much easier to weed out old code if a spelling error (or 
> similar)> is signalled (much like how x. y. u. v. m. n. were 
> changed in 6.01)
> 
> This motivation seems contrary to the design of J.
> 
> For example, I can already type 2+3. when I meant to type 2+.3
> 
> More generally, when working with mathematics, it has generally
> been important that I check my work.  And getting rid of _.
> would not change that.
> 
> I have seen many languages introduce mechanisms which make
> lesser used forms take more typing than other expressions, but
> in the general case longer more round-about code does not
> equate to correct code.  It just means that you have to figure
> out what has gone wrong in something rather verbose.
> 
> Put differently, this sounds like a solution looking for a problem.

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to