Re: All languages are equally fit

DKleinecke wrote:
On Nov 10, 2:06 pm, hru...@xxxxxxxxxxxxxxxxxxxx (Herman Rubin) wrote:
In article<cfef2492-e489-42dd-aeb9-765c3e7ce...@xxxxxxxxxxxxxxxxxxxxxxxxxxx>,

DKleinecke<dkleine...@xxxxxxxxx> wrote:
On Nov 9, 9:46=A0am, hru...@xxxxxxxxxxxxxxxxxxxx (Herman Rubin) wrote:
Also, it may be that the higher order assemblers of the
type I suggest may make it possible to produce more
efficient Linux kernels in less time. =A0They provide
access to the operations which the current high level
languages do not, with many of the natural computer
operations no longer present in the language, so they
dropped out of the computers.
Can you offer any examples of natural computer operations that have
dropped out of computers because they are not accessible to current
high level languages? And any examples of how older high level
languages were better (the word "current' in paragraph aboive)?

High level languages rarely if ever, to my knowledge, had
these operations. Many of the early computers had division
giving an integer quotient and a remainder in one operation;
I have not seen that lately. Also, the separation of a
floating point number into exponent and mantissa, and the
converse production of floats. Fixed point (not integer)
arithmetic. Multiplication giving both the most and least
significant parts, both for fixed point and floating point
numbers. Unnormalized floating arithmetic. Boolean operations
on fixed point numbers, or on floating point.
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Department of Statistics, Purdue University
hru...@xxxxxxxxxxxxxxx Phone: (765)494-6054 FAX: (765)494-0558

Thank you the list.

As I read Herman, I think he was saying 'current' as opposed to 'future hypothetical', rather than to 'old'.

It seem to me that all of your examples are too hardware dependent to
be accomodated in a general purpose higher order language. The authors
of C would tell you that these are, because of the machine dependence,
more suitable for handling as subroutines than as operations of the
language. I would agree. You will find suitable subroutines in C
subroutine libraries.

I would disagree. It's awful that there's no way to both know the quotient and remainder with a single statement - not only clumsy, but it forces you to do the division twice (or do some other costly computation). And there is no way this can be adequately pushed to a library - unless, perhaps, the library is ASM, then you at least don't get extra penalties.

C relegated input-output to subroutines for the same reason and
thereby offended a great many people at the time it came out.

So far as I know all modern higher-order languages are descended from
C and share its attributes.

What?? Lisp, Smalltalk, Eiffel, Haskell, are descended from C?? (Rebol? Ruby? Javascript, even?)
.

Relevant Pages

  • Re: All languages are equally fit
    ... efficient Linux kernels in less time. ... languages do not, with many of the natural computer ... dropped out of the computers. ... Also, the separation of a floating point number into exponent and mantissa, and the ...
    (sci.lang)
  • Re: OS X.4.X update...
    ... back" some languages you have removed, unless they are smart enough not ... has very little in the way of multiple language support. ... applications which support Unicode - AppleWorks is not one of them). ... It doesn't even touch on the other things you can do with computers now ...
    (comp.sys.mac.system)
  • Re: All languages are equally fit
    ... access to the operations which the current high level ... languages do not, with many of the natural computer ... dropped out of the computers. ... both for fixed point and floating point ...
    (sci.lang)
  • Re: Meyer(s) gets it wrong
    ... but not in the same way that many, supposed, OO languages support them. ... I have two programs running on seperate computers. ... TBH I find the idea of seperate processes which have defined interfaces, ... be described as being 'just' good SP practice. ...
    (comp.programming)
  • Re: All languages are equally fit
    ... languages do not, with many of the natural computer ... dropped out of the computers. ... both for fixed point and floating point ... You will find suitable subroutines in C ...
    (sci.lang)