Conference turris::languages

    Which Algol?
    
    Algol-60, either as itself or disguised as CORAL-66 or Simula, is still
    in use.
    
    One big problem with Algol-60 was that I/O wasn't defined as part of
    the language spec, so it was preety tough to write useful portable
    code.
    
    The direct descendant of Algol-60, Algol-68 cured this, but was much
    more than a simple revision of Algol-60.  It's still by far the
    cleanest "power" language that I've used (I'd class Algol-68, Ada and
    C++ as some "power" languages).  Unfortunately it suffered greatly from
    the spec being written in an obscure notation, and by the incorrect
    implication often drawn that it was just a modernized Algol-60.  I
    believe that a variant of Algol-68 was used internally for a long time
    by ICL for O/S work.  After his proposed successor to Algol-60 was
    rejected in favor of van Wijngaarden's language which eventually became
    Algol-68, Wirth produced the series of languages Algol-W, Pascal and
    then Modula and Modula-2.  The current Modula-3 language borrows
    heavily from a number of Algol-68 concepts.
    
    So I don't think Algol's really dead.  Certainly the concepts
    introduced by Algol-60 and -68 are still very much alive, even if
    compilers for pure Algol-60 or -68 are a little hard to find these
    days.

    Because it was a European language, and most users of computers at the
    time were American.  Fortran got there slightly sooner, definitely
    uglier, but with better numerically crunching performance.  Exit
    Algol.
    
    /Bevin

  Thanks all.

  Re: .1

	; Algol-60, either as itself or disguised as CORAL-66 or
	; Simula, is still in use. 

  So THAT'S what CORAL-66 is !  Is CORAL-66 still a supported product
  and still a standard in the UK defense industry ?

	; One big problem with Algol-60 was that I/O wasn't defined as
	; part of the language spec, so it was preety tough to write
	; useful portable code. 

  But you could write really portable useless programs ? ;-)

	; After his proposed successor to Algol-60 was rejected in
	; favor of van Wijngaarden's language which eventually became
	; Algol-68, Wirth produced the series of languages Algol-W,
	; Pascal and then Modula and Modula-2.  The current Modula-3
	; language borrows heavily from a number of Algol-68 concepts.

  How much do Pascal and Modula-n borrow from Algol-60 ?  -68 ?
  Would Ada fit into that continuum somehow ?

	; So I don't think Algol's really dead.  Certainly the
	; concepts introduced by Algol-60 and -68 are still very much

  No contest... that's why I'm curious.

	; alive, even if compilers for pure Algol-60 or -68 are a
	; little hard to find these days. 

  Know where I can find one for a VAX or PDP-11 ?

  Regards,
  rcs

	Re: .2

	; Fortran got there slightly sooner, definitely uglier, but with 
	; better numerically crunching performance.  Exit Algol.

	...i.e. 1957 vs. 1958 ?  When did we have "programmers" per se ?
	With some exceptions, computing at this point is by and large 
	all scientists with hyperactive calculators ?

	Regards,
	rcs

>	; Algol-60, either as itself or disguised as CORAL-66 or
>	; Simula, is still in use. 
>
>  So THAT'S what CORAL-66 is !  Is CORAL-66 still a supported product
>  and still a standard in the UK defense industry ?

    CORAL-66 is a stripped-down Algol-60, with a few additions.  Basically
    take out call by name (which I believe found its way into Algol-60 by
    accident) and add TABLEs, and some strange rules about storage layout
    to let you do pointer arithmetic.  These storage layout rules result,
    among other things, in the address of a 2-byte word (expressed as an
    integer - all addresses are integers in CORAL-66) being exactly half of
    the address of the first byte of the word (!)

    As far as I know it's still a UK standard, although Ada is replacing
    it.

>  How much do Pascal and Modula-n borrow from Algol-60 ?  -68 ?
>  Would Ada fit into that continuum somehow ?

    Pascal is a block structured procedural language.  It's lexically
    similar to Algol-60 (e.g. "begin" and "end", ":=" for assignments and
    "=" for comparisons), and the syntax of its IF construct has the same
    dangling-ELSE problem as Algol-60.

    I'm not too familiar with the Modulas.  I've never programmed in any of
    them, just read about them, but Modula-2 looks like Pascal with a few
    extra basic types and a defined type-safe separate compilation
    mechanism.

    Modula-3 adopts Algol-68's structural type-equivalence (abandoning
    Pascal and the earlier Modulas' equivalence-by-name).  Like Algol-68,
    it assumes that programmers have better things to do than tracking down
    memory leaks, so it provides automatic storage reclamation (unlike
    Algol-68, Modula-3 allows the programmer to turn storage reclamation
    off in certain "unsafe" regions of the program).

    Ada is sort-of based on Pascal, but has an enormous quantity of stuff
    that's new, or at least of uncertain parentage :-)  Once you get used
    to it, though, it's not too bad.  It doesn't deserve the "big and
    clumsy" image that it's acquired.
    
>	; alive, even if compilers for pure Algol-60 or -68 are a
>	; little hard to find these days. 
>
>  Know where I can find one for a VAX or PDP-11 ?

    Well, Systems Designers in the UK market a VMS compiler for Algol-68
    RS.  At least they used, to, but I have a feeling that they were
    bought up a couple of years ago, so I don't know if the compiler's
    still available.
    
    John

    The person to contact re Algol-68 is Charles Linsey at Manchester
    University, U.K.
    
    He has just (last month) announced� a Algol-68S (*minor* subset)
    compiler for various platforms (mainly Unix I think) which is available
    for a very resonable free (couple of hundred pounds < $500)
    
    	Rik
    
    �In comp.newproducts as few weeks ago. I think he is [email protected] or
    something like that