Each division was run like a separate company. One group used all IBM PS/2 PCs with DOS/Win. Another group used Macs. A third used some kind of proprietary CAD stations. They all used different software and so couldn't readily exchange any information. Oh, there was *no* LAN. If I needed to print something I put it on a *floppy* and carried it to the PC that had a printer. Which was old and battered and just managed to work well enough to print, most of the time. The IT guy (singular) was the one person there I felt sorry for - no one else there seemed to have any computing skills. When I started I shocked him by just asking for the media I needed so I could wipe my PC and set it up how I wanted it. He hadn't encountered someone who could do that within the company. So we spent a fair bit of time talking about tech since I was the only one there who spoke his language.
The PCs were, as I mentioned, IBM PS/2 units - 386s I think. They used WordPerfect for documentation, which was decent at the time - I liked it better than Word. They'd budgeted six months for the documentation project. They'd never had a real tech writer in house (TW is one of my degrees, and I'd just graduated), docs had always been done on a 'time available' basis by the engineers. Well, in less than 6 *weeks* I had all of the planned docs as drafts, and then some. Since the engineers didn't expect to have anything to review for months I ended up sitting on my thumbs because I needed the drafts reviewed before I could proceed. And that was after making work for myself. Since I lacked illustrations, and the engineers were too busy to help, I taught myself the CAD system the group used (AutoCad IIRC) and took the engineering drawings and stripped them down for illustrations. Management had *no clue* what to do with me, I ended up sitting in my office reading the newspaper half the time.
Oh, speaking of which. I sat in an internal, windowless office in the bowels of the building. I rarely saw another human being and *never* a customer. Yet I had to wear *suits* to work. To sit in an office and type docs. Because that was the tight-assed culture. Oh, and about the office. It was a cookie cutter, bland office with what looked like WWII Army surplus furniture. Big old metal desk, squeaky chair, etc. Very beige and gray, and bland. The office next door was *identical*, including the PS/2 on the desk.
Which makes the next bit all the more astounding. One day my boss came by and told me I had to move - to the office next door. Why? Because the office I was in belonged to sales and they had a new person coming in. Why not use the empty office? Because *this* one belonged to sales. Aren't they identical? "Yes, but *this* one belonged to sales." Isn't sales in the same direction as the other office - so isn't that office like 10 feet *closer* to the rest of sales? "Yes, but *this* one belonged to sales." OK, someone from sales must've pissed in the corner at some point and marked their territory.
OK, so I have to move - as pointless as it was. I'll just move my PC. "No, the PCs are tied to the offices. You need to install all your software and move your data on the other PC. Then setup this one so it is ready for the new person." *blink* They're *identical* shit-box PCs! There is no LAN, it isn't like the MAC address was tied to a port or something. Somewhere there was an inventory tracking spreadsheet that had each PC glued to a physical office. So I agreed.
And as soon as he was gone I swapped the fucking PCs. I imagine that at some point someone from facilities was doing an inventory and had a fucking coronary because the PCs didn't match their spreadsheet. Oh, I told the IT guy about it and he couldn't have cared less.
But that wasn't the best part. The best part was one of the products.
One of the IR products was an IR spectrometer housed in a NEBS level three compliant case, for industrial settings. Inside was a multiplexer for a number of fiber-optic lines that connected to the remote sensors. The case was internally baffled into two halves. One half had the multiplexer and the sensor head, which needed to be cooled, the other half had the electronics and the other half of the Peltier cooling system. So the sensor was the cold side, and the other was the 'hot' side - like an industrial McDLT. All of this stuff was mounted to the case, which had fins on the back - 'below' the gear - ostensibly for radiating away the heat.
Now, I'm 23, fresh out of college with a degree in Tech Writing and a degree in History. They have teams of engineers twice my age who design these things.
But I'm looking at this box on the assembly line and, well, it is WRONG. The sensor head is mounted, very firmly and with thermally conductive adhesive, directly to the inner wall of the case. All of the electronics, and most especially the radiator component of the cooler, were NOT mounted to the case. They were, in fact, mounted on standoffs and thermally isolated from the casing.
That mean the sensor was thermally coupled to the case, and thereby to the environment around the case. As it was chilled it would naturally draw in heat energy from the casing, which would chill, and pull heat in from the air - quite efficiently with all those fins.
Meanwhile the cooler would be dumping the waste heat into the sealed half of the box.
That's exactly the opposite of what you want.
So I asked the foreman about it, figuring I *must* be missing something. Because something *that* dumb couldn't be true, right?
Now, the foreman was a very nice guy, very down to earth, very laid back. A good craftsman with quiet confidence. So he just kind of smiled and shook his head, and told me that no, I was indeed correct. The system was flawed.
OK, I asked, so why doesn't anyone fix it?
Because none of the engineers wanted to deal with the volume of paperwork required to do an engineering change order. The cooler was powerful enough to still keep the sensor cold enough in most installations, even with the flaw, and the paperwork was so bad that they just left it as is rather than try to fix it. How it got approved that way in the first place, I have no idea.
I was only there six weeks, then my resume found its way to Xylogics and I jumped ship.