29 Apr
2016
29 Apr
'16
12:35 p.m.
On Apr 29, 2016, at 3:25 PM, Diane Bruce <db at
db.net> wrote:
On Fri, Apr 29, 2016 at 09:13:42PM +0200, Raymond Wiker wrote:
>
>> On 29 Apr 2016, at 21:10 , Paul Koning <paulkoning at comcast.net> wrote:
>>
>>
>>> On Apr 29, 2016, at 3:01 PM, ben <bfranchuk at jetnet.ab.ca> wrote:
>>>
>>> I liked Forth when it was still threaded.
>>
>> ???
>>
>> Base FORTH is not, in and of itself, threaded. PolyFORTH was if memory serves.
Then again, creating a thread scheduler (cooperative scheduler) for FORTH is just a modest
exercise for the programmer.
>
> I'm guessing Ben means threaded as in "Threaded Interpretative
Language", and not a concurrent programming language.
>
> There are still plenty of Forth implementations based on threading (of words).
Oh. Sure. FIGForth is (PDP-11 Forth is an example of that). It's very easy indeed
to make a threaded-code Forth implementation. In any case, whether a particular
implementation is or isn't shouldn't matter so long as the language works
correctly, right?
Sure. Adobe postscript is a thread interpretative
language (TIL). It looks
very much like FORTH if you squint real hard.
I don't know if PS is threaded. It could be. PS claims not to be inspired by Forth,
something that's hard to credit. It does have some crucial differences, among which
is runtime (usually) rather than compile time name binding.
Also, old DEC Fortran looks very TILish if you
didn't use the option
to generate native PDP-11 code. ;)
Right, PDP-11 Fortran IV (as opposed to IV-Plus) generated threaded code. And P-code
(such as in DEC Basic-PLUS "push/pop code") is essentially the same thing, give
or take indirection through a jump table. Same for TUTOR and the first ALGOL compiler
(for the EL-X1) -- those are actually hybrids of threaded or P-code and machine code.
paul