unexpected results for ==

dBASE PLUS 2.6.1.3   Win XP Pro SP2

in command window type

x = 0

? x            // returns 0 as expected

? type('0')  // returns 'N' as expected

? x == false // returns true  - is that WAD? I would expect an error 

Bruce,

Guess I will to from now on.

Thanks

Frank Polan

On Tue, 25 Mar 2008 21:04:47 +0000, Bruce Beacham
<bbeacham@beacham.no-spam.co.uk> wrote:

>Frank J. Polan wrote:
>> x = 0
>
>> ? x == false // returns true  - is that WAD? I would expect an error 
>
>There's a general feature of a number of languages which go back 
>decades, that false can be indicated by a 0 and true can be indicated by 
>a 1.    You can see why, of course - it's economical.
>
>I've noticed this in dBASE and so when I generate return values to x 
>that may be false or may be a count of some sort, I make x private and 
>test for type().
>
>
>Bruce Beacham

Frank J. Polan wrote:
> x = 0

> ? x == false // returns true  - is that WAD? I would expect an error 

There's a general feature of a number of languages which go back 
decades, that false can be indicated by a 0 and true can be indicated by 
a 1.    You can see why, of course - it's economical.

I've noticed this in dBASE and so when I generate return values to x 
that may be false or may be a count of some sort, I make x private and 
test for type().

In article <op6km6rjIHA.700@news-server>, bbeacham@beacham.no-spam.co.uk 
says...
> Frank J. Polan wrote:
> > x = 0
> 
> > ? x == false // returns true  - is that WAD? I would expect an error 
> 
> There's a general feature of a number of languages which go back 
> decades, that false can be indicated by a 0 and true can be indicated by 
> a 1.    You can see why, of course - it's economical.

In many cases, ANY non-zero value is "true", but ideally (or by
language 
definition), true = -1.

-- 
Remove the ns_ from if replying by e-mail (but keep posts in the 

ns_dkerber@ns_warrenrogersassociates.com wrote:

>> a 1.    You can see why, of course - it's economical.
> 
> In many cases, ANY non-zero value is "true", but ideally (or by
language 
> definition), true = -1.

That's interesting.   Needs an extra bit though, I imagine!