Re: [] \= '[]' (was Re: Ann: SWI-Prolog 7.1.0)

On 28/11/2013, at 4:09 AM, Jan Wielemaker wrote:

> 
> Not really. There have been attributed variables and strings for a long
> time.

They don't break the traditional hierarchy at all.
Attributed variables satisfy var/1.  There is more to know
about them, but usually you don't care to know it.
strings are just constants; Prolog systems have long had
a range of things that are atomic/1 but do not satisfy
atom/1 or number/1.

> At some point, it is likely that there will be rational numbers as
> a real type rather than a compound rdiv(N,D).

That will actually *improve* the fit to the traditional hierarchy.
> 
> 
> What I'm after is more runtime type information,

Well, we had better introduce objects and classes then.

>  I would have been happy to
> introduce a real type `list'.

It would have made Prolog a more complicated language.
Many Prolog systems had a special *run-time* representation
for dotted pairs without this needing to be surfaced in the
*semantics* of the language.

You know, it's the first thing I look for in Prolog source code.
Documentation of the intended data structures.
Too often I don't find it.