[FoRK] "Super Tuesday" only "semi-great," Texans claim

[Jeff Bone]

On Feb 13, 2008, at 2:26 PM, Eugen Leitl wrote:

> On Wed, Feb 13, 2008 at 01:16:17PM -0600, Jeff Bone wrote:
>
>> Hmm, I dunno if I'll ever be able to consider a stack language truly
>> functional...
>
> Stack machines and threaded languages have a lot things going for  
> them.
> Message-passing OOP is also easy there.

Oh, no doubt.

My first "serious" language (I won't consider TurboPascal dabbling  
"serious" ;-) was a graphical Forth called GraForth that had  
primitives for 3-D graphics --- back on the old, original PC.  Could  
do wireframes at really respectable frame rates for the time.  Native  
3D cartesian coordinates, transforms, etc.  I built an animated 3D  
star mapper in it, my first serious program really.

PostScript / NeWS --- all very respectable.

By "functional" in my comment above I meant in the specific sense of  
functional languages.  I'm not sure I buy the handwaving of e.g. Joy's  
advocates, who will say (nb, cribbing some Joy thing I read today on  
progeddit) that e.g. 3 + is actually two functions from stacks to  
stacks.

I just don't think you can argue away the statefulness of stack  
languages;  i.e., I don't think you can legitimately call any stack  
language a "functional" programming language in the same sense as e.g.  
Haskell (or Clean, etc... --- i.e. that class of languages  
specifically developed in response to Backus' Turing Award lecture.)   
I.e., not declarative / applicative / doesn't have the same  
referential properties, etc.

jb