In 1993 I was in college, but there was that brief period of inactivity in the summer that we referred to as "summer break" or something like that. Your options for summer break were (a) do nothing or (b) get a job for three months.
I chose not to do nothing. I chose something different.
Well, actually, I indulged in nepotism en masse. Firstly I got a job with a company my brother worked for; they had some computers and noone in-house to tend them, and they had a few things they wanted done on said computers. So I spent 48 hours trying to force-feed myself COBOL so that I could play with The Mighty Accounting System, then discovered that since all they wanted was reports, I could generate standard reports from the TMAS and use some C to coerce, cross-reference, collate and coordinate the data into the reports they wanted. Quelle pushover. After a fortnight there I'd even started decoding the fileformats that TMAS used and dispensed with the report-generating for some of the reports.
Then John called. John is an odd character who appears every so often, for "often" greater than six months, and gives me a job proposition. Everyone needs a guy called John, I think. John basically sweet-talked the company he worked for in London into hiring me, sans interview, for "a small Visual Basic project".
Stop me if you've heard this one before.
It turned out to be a small job; write a phone book for the company employees to use. Features: comprehensive, fast, easy to use.
Simple.
Er, not quite. I was hired, it transpired, partly as a pawn in the political games in the office. I belonged to IT, sort of, but worked for/in/near Software Dev. Or some such nonsense. Which made life sort of, well, interesting for most of the time I spent in the office. The application itself was subject to manoeuvering; for example, Software Dev wanted me to use xBase-format files. "Here's the library", they said, "and here's the documentation". I argued for flat files, since speed was of the essence and we were talking about a VB3 app running on Win3.1 on 386 and 486 machines with less memory than most palmtops have these days. So I did it both ways. The results? xBase files took *ages* to load and the app held onto the machine while the files loaded. Flat files, on the other hand, took about 30 seconds to load, and the user got control back after *8 seconds*.
Flat files stayed in the app.
The project went on and on, and being that I was partially an IT resource I occasionally got to do fun stuff like recreating a bootptab file (after the company moved offices) one computer at a time for a 500-computer network, so I didn't have 100% devotion to The Phone Book From Hell. And it was from hell, by the time everyone had their say - phone numbers, mobile numbers, email addresses, pager numbers, nearest photocopier, nearest fax, administrivia. In the end, I threw together some doco and some test plans and dumped the whole thing on Software Dev.
I subsequently learned that it was never rolled out. Software Dev decided to redo it from scratch themselves, and canned the project unfinished after 18 months. I've heard of second-system effect, but really. A *team* in *18 months* can't even *replicate* my *one-man*, *three-month* effort? I can't possibly be that good, can I?