Re: Concerning subscript bounds checks

[ok@goanna.cs.rmit.EDU.AU (Richard A. O'Keefe)]

rav@goanna.cs.rmit.EDU.AU (++           robin) writes:

>---I expect that the same observations could be made about
>re-writing the program in PL/I, or in Fortran 90.

Again, this imputes to me a claim I never made.
I did not claim that Ada was *better* in this respect than PL/I or F90.
All I claimed was that it is *good*.
There is a big difference!

Obviously, APL, PL/I, and Fortran 90 have array arithmetic (although
I used APL enough to be seriously confused by PL/I's differences there).
Equally obviously, a vector arithmetic expression all of the vectors in
which have the same declared bounds needs no subscript or length checks.

>   Why not post the original Fortran version, and the Ada
>translation, so we can see?

If there's enough interest, I'll put them on the Web.
The Fortran code has _already_ been published; it's algorithm
AS 47 from the Journal "Applied Statistics".

nelder_mead.ads		 18 non-blank-non-comment lines
nelder_mead.adb		216 non-blank-non-comment lines
nelmin.f		204 non-blank-non-comment lines (my F77 version)

(rav: look in goanna:~ok/Ada.d/ if you want to see them.)

The difference between the Ada and Fortran line counts is entirely due
to my Ada style being one declaration per line, with the Fortran code
still having several grouped declarations.

>	>Fifty years of programming language research, and we end up with C++ ???

>---50 years?  Wasn't Ada (after whom the language in which this
>posting also appeared first) the first programmer?  Wouldn't that
>make it ~130 years?

I did not write "fifty years of _programming_ research" by
"fifty years or programming _language_ research".
-- 
Fifty years of programming language research, and we end up with C++ ???
Richard A. O'Keefe; http://www.cs.rmit.edu.au/~ok; RMIT Comp.Sci.