HP3000-L Archives

April 1998, Week 4

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Glenn Cole <[log in to unmask]>
Reply To:
Date:
Thu, 23 Apr 1998 17:37:28 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (64 lines)
Item Subject: cc:Mail Text
The May issue of Doctor Dobb's Journal has several interesting articles,
with a few specifically on strategies for "fixing" the y2k problem.

-----

The articles with the most application to the 3000 are "Date Compression
and Year 2000 Challenges (page 20) and "Strategies for Solving the Y2K
Problem" (page 26).  The latter article focuses on date expansion and date
windowing, but appears fairly thorough given the limited focus.

The Compression article looks briefly at date expansion and date windowing,
but it takes the stance that date windowing is too prone to error, and date
expansion affects record layouts too hard.

Assuming that the date is stored in six bytes (a HUGE assumption, IMHO),
it covers a myriad of alternate formats that fit in the same six bytes.


The thing that I have *never* seen addressed in any y2k discussion is how
best to handle the screen fields.  Maybe that's just not considered to be
a problem, and they will be handled just as before.  (I don't really
believe that, though.)

-----

The other article (page 115) is an interview with Bob Bemer.
I never heard of him either, but he has some pretty impressive credentials:
he worked with FORTRAN creator John Backus *and* COBOL creator Grace Hopper
*and* IBM S/360 designer Fred Brooks.  He says his major contribution to
COBOL is the PICTURE clause.

Mr. Bemer claims:

- source code is not available for about *30 percent* of the apps
        that need to be modified

- there are "a lot of mismatches" between the source code that DOES exist
        and the actual running apps

- "Date-field expansion [yy -> yyyy] will never make it."

- date windowing is "Also flawed.  For one thing it's evanescent."
        (I had to look that up.)

Mr. Bemer's solution, though, is a product that examines the *object*
code.  Also, near as I can tell from the DDJ article, the product is
heavily dependent on packed decimal operations.  (I could be mistaken,
but that's how it appears to me.)  Not surprisingly, the product runs
(only?) on IBM mainframes, specifically MVS and perhaps OS/VS.

For more info, besides the DDJ article, see his company's web site at
        < http://www.bmrsoftware.com/ >

-----

Anyway, the magazine is an interesting read.

FWIW.

--Glenn Cole
  Software al dente, Inc.
  [log in to unmask]

ATOM RSS1 RSS2