Date: February 26, 1997 Time: 12:51
From: Jerry Lawson
The Hartford Insurance Group
(860) 547-2960 jlawson AT itthartford DOT com
-----------------------------------------------------------------------------
IBM -
IBM -
Is it just a plot to drive me crazy, or are you after everyone :-)>
I tried to restore a directory structure the other day using Restore by
Sub-directory branch on my OS/2 machine (running L6). Was I surprised when it
didn't do what I expected! Here's what I did (and I had been doing for
years).....
1. Highlighted the appropriate drive letter on the GUI main menu
2. Selected Restore by sub... on the appropriate pull down
3. Filled in the subsequent panel with:
Source \Gold
Target G:\
Replace option of my choice
Was I surprised when everything got dumped into the root! It didn't use to
work this way - it used to go into the same named directory - that is - all
the data would have been placed in a directory called Gold, subdirectories
would have been built under that.
Sooooo - I went to the Readme files, but couldn't find any specific mention of
the change, although I did find the infamous "internal" APAR to fix things
that no one else thought "broke" (Sorry - couldn't control myself!) I was
about to write a poison pen note when I thought "Better check out the manual"
(I know that's an odd thought but then again it was a Monday......:-)
GUESS WHAT! My OS/2 client was now working as documented! Guess I better put
away the Pen and the Curare! But then on Tuesday, a Win95 user, who had just
installed a new copy of the Win95 client at L6, stopped by my office and
wanted to know how to do a subdirectory level restore. So guess what I told
him?
Source \Gold
Target g:\gold
Replace option as appropriate
Guess what he got! A directory structure of G:\Gold\Gold Have I been
watching too many "Time Warp" movies on TV? I got out the Windows manual,
and it is documented the same way as the OS/2 manual. This client is
obviously working as coded, but not as documented.
So - IBM - The jig is up! Come Clean!!!!!
My questions....
1. When did the OS/2 client really change to work like this?
2. When will the rest of the clients change to work like this?
It seems to me that on a function like this, all clients should behave the
same way, since customers are want to switch between platforms without notice.
Now my last question - While I was comfortable with the original way to make
this work, it perhaps is not intuitively obvious, and obviously not working as
documented. On reflection, I think the documented approach makes sense, and
is more intuitive. The question to everyone else is - "Does anyone object if
all clients work as documented, or does the pub need to change?
This last one is an important question, as I can recall many times where IBM
programs did not work as documented for years and years, and when finally
called on it, the selected approach was to change the code, instead of change
the documentation. The problem, of course, was that since everyone else knew
what the code really did, changing it caused grief and aggravation for many
people, just because it was easier to do a PTF that a Pub Change.
*****************************************************************************
Jerry Lawson
The Hartford Insurance Group
jlawson AT thehartford DOT com
Any idiot can face a crisis. It's the day to day stuff that really wears you
down.
Anton Chekov
|
