Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1a 7/7/83; site rlgvax.UUCP
Path: utzoo!linus!philabs!seismo!rlgvax!dave
From: dave@rlgvax.UUCP (Dave Maxey)
Newsgroups: net.jokes
Subject: Re: YASEM (yet another strange error message)
Message-ID: <1291@rlgvax.UUCP>
Date: Sun, 16-Oct-83 23:08:46 EDT
Article-I.D.: rlgvax.1291
Posted: Sun Oct 16 23:08:46 1983
Date-Received: Mon, 17-Oct-83 23:44:43 EDT
References: <267@rocks34.UUCP>, <1288@rlgvax.UUCP> <1508@gatech.UUCP>
Organization: CCI Office Systems Group, Reston, VA
Lines: 49

Note: YASEM is the term I first used (I'm a fan of YA... names). Since I've
noticed a lot of YASEM comments here, I thought that maybe a good consultant
story is in order. I didn't actually start the YASEM trend, but I have
contributed quite a bit, since I was once a university computing center
consultant (sometimes referred to as insultant, we took a lot of abuse).
That was over three years ago. I still remember an oriental programmer
plopping the listing of his program in front of me and saying ",Ahhh,
computer not work... you fix?" After I picked my jaw up off the floor I
began flipping through the listing. It was a huge Fortran program, uncommented
with gotos going everywhere. A friend of mine likes to call it spaghetti code.
After looking at the dump at the end, and the load map, I began looking at the
section of code in question. Whenever I tried to flip to it he told me that
that part of the code was fine and tried to get me to look at something else.
(Now here's the part that all you consultants out there should remember,
because it can get to the meat of the problem real quick). I asked him why he
thought there was nothing wrong with his program thus leading him to believe
that the computer was broken (he didn't seem to believe me when I told him
that except in very strange circumstances, if the computer doesn't work it
probably won't work at all. Otherwise it will screw up everyone's program the
same way. Usually...). He said that he hadn't changed anything and that it
had worked before. I had heard that story, so I was ready for him. Very
rarely will someone run a program a second time just to get a copy, so I
asked him ",If it already worked, why did you want to run it again?" "Well,"
he replied ",I have different data..." "Ahhhh, sooo (pardon the pun), the
data changed!" He looked somewhat surprised at my reply, and staunchly
stated "Program work fine, computer not work... you fix." What did he think I
was going to do? Pull out a Phillips screwdriver, go into the computer room
and...? I'd just about had enough. I began a discussion of the difficulties
in locating computer hardware problems and the time constraints involved while
I casually flipped through the pages of his listing again. He was so intent on
grasping my forty dollar computer terms that he didn't notice what I was doing.
I found it! A divide with the divisor coming from a recently read in variable.
He had a 0 in his new data! How simple! Computers just don't like indeterminate
mathematics. I was so pleased that I got a little cocky. I told him that we
could fix the problem, but that we wanted to satisfy our customers and I could
show him a little trick to get around the computer malfunction, until such time
as it was properly dealt with. I discovered his program was the type that was
supposed to ignore 0 values in the input (at least the divisor values), and
put in a check to make it fetch another data set (more spaghetti). He looked
as if he doubted me, but went off to try it anyway. He came back and said that
his program ran fine now and that we could fix the computer when ever we felt
like it! He was beaming. Another satisfied customer...

	I told the operations shift supervisor later that day that he had had
a computer malfunction that morning, but that I had fixed it. He gave me the
strangest look...

			- Dave Maxey (alias tbm)
			{seismo,mcnc,brl-bmd,allegra}!rlgvax!dave