TMG-L Archives
Archiver > TMG > 2004-01 > 1072982692
From: Lee Hoffman <>
Subject: Re: [TMG] Re: Problem 3: corrupted pdf file for Individual detail report of whole project
Date: Thu, 01 Jan 2004 13:46:31 -0500
References: <CLEDKMONADFEACGECFGJMEMBEAAA.lorna.henderson@paradise.net.nz><001501c3d08c$f42c03f0$cc289c18@corneroffice>
In-Reply-To: <001501c3d08c$f42c03f0$cc289c18@corneroffice>
Ronald J. Emery wrote:
>You have brought back old memories.
>
>The first computer I ever programmed was the IBM 650 computer which we
>installed while waiting for the UNIVAC II computer to arrive. It was 1957.
>The program was on a deck of punched cards, each card containing one
>"instruction" and the "word" in memory that that instruction was to be
>stored.
>
>When the big machine arrived, we went off to progam it and one of our
>replacements was a keen young fellow who thought that we had been pretty
>sloppy and who decided that he was going to document things properly.
>
>So he took each program deck of punched cards, sorted them in order of
>memory location, produced a printout of the program and made a file folder
>for each program. You had to admit, it was pretty.
>
>But what the poor fellow hadn't realized was that when we were debugging
>these programs and found an error, we had simply punched up a new card for
>that address and placed it at the back of the deck. So there were multiple
>cards aimed at the same address but the last card read in would replace
>anything that was there before.
>
>These multiple cards of course ended up in random order after his sort of
>the program deck. All of a sudden none of the programs worked. We were
>called back to fix up the mess, and we did. But I don't think I ever lost my
>disdain for documentation.
I was like you for many years until I retired from the Navy and got a job
with a small OEM computer company back in 1980. We had a project that
required full documentation and as I was project leader, it was my job to
ensure that the document was complete. About a year after this, the
company went out of business, and I went on to another company. But I
happened to keep a copy of the documentation which came in handy when the
buyer of the project called me and asked me to troubleshoot a problem and
make some modifications. That documentation (from the beginning) sure
reduced the time that it would have taken me to find the problem and make
the other requested modifications.
Incidentally, while in the Navy, we had a very nice sort program and I
wanted to see how it was programmed. So I disassembled it, researched the
system calls and then wrote up the documentation to help me understand
it. That was a fun project and I learned a lot. I was even able to create
an updated version that reduced one sub-routine by about three lines. When
we ran it on our local machine, it cut run time by about three minutes (out
of 17 minutes) with the same results. I never did submit the changes for
an update and we went back to using the regular program just in case I had
missed something that my small tests didn't find.
As I mentioned, I learned a lot from this and was able to include a number
of things I learned from that documentation effort in later projects. I
also learned that later documentation must be checked to make sure that it
agrees with what was actually done like your replacement found. His effort
was good, except that he didn't check for possible changes. Thus
documentation can be good, but if it isn't done from the beginning it must
be checked against the final result to make sure it is accurate.
Also as Lorna mentioned dropping card decks were always a problem. We
always numbered our decks, but that didn't always work when a new card or
two were inserted. <g> I saw a friend drop a card tray for a 15,000 line
program -- un-numbered at the time. He just sat down and cried. The rest
of us just tried to console him. Our supervisor nearly joined him as the
project was already behind in the schedule (not unusual <g>).
Lee Hoffman/KY
TMG Tips: <http://www.tmgtips.com>;
My website: <http://www.tmgtips.com/lhoffman>;
A user of the best genealogy program, The Master Genealogist (TMG)
This thread: