Craddock, Chris wrote:
RS said
1. It has little to do. There is something which we can call "IT
culture". PC environment (I mean human env) is more likely to
"restart-like", while mainframe environment is more likely "tight
controlled".
Of course, YMMV, this is generalization, etc. etc.
[<CLC>] Funny you should mention that. I will assert that today there is
almost nothing significant that you can "fix" (i.e. restore service)
faster by rooting around trying to find the problem, than by just
restarting the application or the server it is running on.
And yes, that's a generalization. Problems do eventually have to be
diagnosed, but your chances of doing that well enough during a critical
situation are basically zip. Have been for many years now.
In my experience, PC programmers simply cannot or will not perform any
kind of post-mortem dump analysis. And, though Micro$oft operating
systems appear to have the ability to take a "dump", I have never met
anyone that knew how to, or cared to, read one. The only thing they know
how to do is reproduce a problem under a debugger. If that can't be
done, they chalk everything up to problems with the underlying OS,
drivers, or services.
All seasoned mainframe software developers know that a fair number of
problems occur only under "special" circumstances -- often related to
timing, asynchronous activity, or other hard-to-control variables. Such
problems often can't be reproduced -- even when you *know* what's wrong.
Without post-mortem analysis to identify the cause of these problems,
such bugs would never be fixed. And, on PCs, they never are.
Case in point: I was talking with one of our tech writers a couple of
weeks ago, discussing the concepts of supported vs unsupported software
for a matrix she was to publish on our web site. I tried to use
Micro$oft examples because she's familiar with their products. She told
me there are scores of universally-known bugs in supposedly "fully
supported" versions of M$ Word that are decades old and still not fixed.
The tech writing user community deals with these problems by avoiding
them. (Doctor: It hurts when I do that. Then don't do it!)
That evening, I pulled up to a gas pump that had one of those flat panel
TV screens above. The thing was "stuck" on a Windows BSOD. Somebody was
paying to advertise a trap screen! :-D
These anecdotes illustrate a *huge* cultural difference between our
platform and others. This sort of thing would simply not be tolerated on
a mainframe! PC programmers don't have the tools they need to make their
software bullet-proof because they just don't care...
--
Edward E Jaffe
Phoenix Software International, Inc
5200 W Century Blvd, Suite 800
Los Angeles, CA 90045
310-338-0400 x318
[EMAIL PROTECTED]
http://www.phoenixsoftware.com/
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
