HP3000-L Archives

August 1996, Week 5

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:
"F. Alfredo Rego" <[log in to unmask]>
Reply To:
F. Alfredo Rego
Date:
Sat, 31 Aug 1996 05:00:36 -0600
Content-Type:
text/plain
Parts/Attachments:
text/plain (52 lines)
Jeff Vance <[log in to unmask]> wrote:
 
>I was coming from the perspective that there are many non-system level UDCs
>setting HPSYSNAME (I have a few!) and there is probably a system level UDC
>that
>does the same.  The system UDC is executed first and then the acct/user UDCs
>overwrite the value of HPSYSNAME set by the system UDC.  I figure that we
>can't break any of these UDCs by causing the SETVAR HPSYSNAME to fail, but
>we could issue a CI warning "This system variable cannot be overwritten".
>The warning would most likely let all existing scripts continue to run as
>they do today, yet we have effectively changed HPSYSNAME from a job var to
>a system var that must be unique.
 
How about leaving HPSYSNAME as a "legacy" name, totally undisturbed, and
creating a NEW variable, called HPSYSTEMNAME (or whatever) with the
required new functionality?  I found that this process has served us well
while "extending" Adager's "classes" (to use Java terms), as we went (for
example) from:
 
        - supporting only Classic hardware to supporting classic AND RISC
          hardware;
 
        - from supporting only IMAGE/3000 to supporting IMAGE/3000 AND
          TurboIMAGE and then to supporting all of the above AND IMAGE/SQL;
 
        - from supporting only standard dataset files to supporting standard
          dataset files AND jumbo dataset files.
 
In general, I found that using this approach (leaving "legacy" stuff alone
and creating new "spaces" or "dimensions" for the new features) made our
lives easier in terms of coding, maintaining, and supporting BOTH the old
features and the new features.
 
For the mathematically inclined, you may think of the whole approach as
analogous to multidimensional vector-space algebra.  I am talking about a
"base" of linearly-independent vectors.  When you need a new feature, leave
the old dimensions alone and create a new, orthogonal vector in the base
that allows you to "span" such new feature.
 
 
 
+---------------+
|               |
|            r  |  Alfredo                     [log in to unmask]
|          e    |                           http://www.adager.com
|        g      |  F. Alfredo Rego               Tel 208 726-9100
|      a        |  Manager, Theoretical Group    Fax 208 726-2822
|    d          |  Adager Corporation
|  A            |  Sun Valley, Idaho 83353-3000            U.S.A.
|               |
+---------------+

ATOM RSS1 RSS2