"Telco" - "programming challenge"

[WMK]

IBM has placed out on one of their web-sites a "programming challenge". This is for an (unrealistically simple) program whose specifications are at:

http://www2.hursley.ibm.com/decimal/telco.html

I have created *ONE* COBOL solution (probably NOT optimal for performance) and have posted it at:

http://home.netcom.com/~wmklein/dox/TELCO.txt

That source should (I think) compile and run with any ANSI/ISO 85/89 conforming compiler. (It uses ONE intrinsic function - but that would be relatively easy to code around - if you had just an '85 Standard compiler).

I (and those posting the challenge) would be interested in other "solutions" and/or other timings.

NOTE WELL:
There are probably LOTS of things that I would do differenctly (particularly adding "error detecting and handling") if I were to make this into a "real" program.

I would also guess that using the (optional) binary file MIGHT be faster (at least on some platforms / with some compilers)

Bill Klein

 

[slade]

Hi Bill,

This is interesting; I've never come across something like this before; an industry specific benchmark. Is this a usual practice of IBM? Do you have any idea about how they would use this kind of info. I can guess, but I'm asking if you might know why.

Thanx, Jack.

 

[Crox]

Hi,

Computer Associates has a nice compare of COBOL to C. It can be used to compare compilers for COBOL and C. You can download COBVSC from their site.... just looking for it, cannot find it anymore at ftp://ftp.ca.com/pub/realia/Samples/
It was a file named COBVSC.ZIP.

If anyone is interested in this old file, mail me at COBOLCROX <at> hotmail.com

Perhaps someone even find a new place were you can download it.

Regards,

Crox

 

[Glenn9999]

I heard this post was out...I've provided solutions to the site holder in Pascal and Delphi (link coming soon hopefully to Usenet). While it is simplistic compared to real world application, I think the main purpose is that the site holder is trying to compare decimal implementations in different languages, which doesn't require a complex problem. In fact, the simplest most useful application for the determination would be beneficial to encourage people to code this in many different languages. It also provides a great cross section of languages and how different things get done.