Re: Looking for Steelman

Magnus Gustavsson <Magnus.Gustavsson@bill.forsmark.uu.se> writes:
>I'm very interested in getting hold of the so called STEELMAN document
>(the origin of the ADA programming language).


preston@dawn.cs.rice.edu (Preston Briggs) writes:
>>I'm not sure I'd call STEELMAN the origin of Ada; there were a whole
>series of men, STRAW though STONE, with STEEL occuring near the end.
>These documents were requirements, not language specifications.


Actually, according to an article by Edsger Dijkstra (published in SIGPLAN
Notices at about the time when the selection process for Ada was
culminating) this was one of the biggest problems. The *MAN documents
*were* in fact imposing a certain amount of specifications on the solution
space... too many such requirements (and far too vague ones) in fact.


Dijkstra noted that one of the requirements was that "Ada shall be a strongly
typed language" (or words to that effect).


He then went on to point out that this "requirement" was really rather silly.
What is a "strongly typed" language anyway?


(He also pointed out that the easiest way to have a strongly-typed language
was to provide only one universal data type. :-)


--


-- Ron Guilmette, Sunnyvale, CA ---------- RG Consulting -------------------
---- domain addr: rfg@netcom.com
---- uucp addr: ...!uunet!netcom!rfg
--

Return to the comp.compilers page.
Search the comp.compilers archives again.