(1) The main reason that the company had hired me on was because they had a large survey scheduled in a few months, and needed some help from someone familiar with electronics.  This was a combination video and sonar survey of a 4.5 km long sewer, and the company had done neither of these before - though their partner in crime had apparently worked with sonar equipment before.  After many long hours of helping their jack-of-all-trades guru on the project, we had the finished result:  a sewer raft built out of inflatable boat fenders, a pelican case full of electronics, and a pan-and-tilt camera perched precariously on top.  All the data ran back through an overly complex fibre-optic setup back to mission control.

And that's how I ended up sitting in a small trailer outside of an open sewer, using a video camera to examine 50 years of built-up material on the walls of a half-charged sewer, for the entire night.  Not counting the 12 hours of setup or 4 hours of tear-down, of course.  Sometime in there I slept in my car for a few hours while someone else operated the camera, but otherwise it was a full-day affair.

(2)  For some reason, the company had taken on more work than they had engineers available.  I should point out that their main line of business is doing geophysical surveys, and I was their only full-time programmer (one of the partners in the company dabbled a bit in his spare time as well).  They needed someone qualified to do this particular survey, so they hired an old friend of the company's as a contractor, but he wasn't comfortable dealing with all of the computer equipment, backing up data, sending it back to the office, and so on.  That was where I came in on this one.

It was also how - about six months after starting a job as a "software developer" - I found myself living in a quonset hut at a 60-man exploration camp northeast of Yellowknife.  Oh yes, it was also late January.  Aside from frequent equipment failures, from custom geophysical equipment to the snowmobiles themselves, it wasn't a bad experience.  The food was decent and hearty, the tent didn't fall over (it wasn't one of those fancy "metal" huts), and we weren't asphyxiated the night that the stove pipe partially collapsed during a storm.  Walking 50 meters to the latrine in the middle of the night was less pleasant, however.  Now, whenever it gets cold in Vancouver, I just remind myself that unless it's -50C, it's not really that cold. 

(3)  As I've mentioned, the only other person to write any decent amount of software at this company was one of their engineers.  He was a good guy, not formally versed in development, but had dabbled enough in Visual Basic and Labview to make a few tools that he used occasionally.  I wasn't about to touch VB with someone else's ten-foot pole, but Labview's built-in signal processing seemed useful, so instead of rocking the boat I decided to learn how to use it.  My first major program to write with Labview was for doing over-water seismic surveys, gathering the data with a hydrophone and recording it with the boat's position and heading information (via GPS) so that it could later be processed.  While this had been field-tested before setting out on a survey, they wanted me along just in case anything went wrong.  It was probably a good idea, because every skipper has their boat wired slightly differently, and there is always going to be some stupid grounding issue on the next boat you deal with. 

There's nothing like sitting in the back of a small powerboat, bobbing in the waves of the Strait of Georgia, while trying to debug whether it's the software or hardware that's causing the problem.  That's when I decided that despite the lure of the sea, I would never EVER buy a boat.  Ever.  I think the most memorable part of that trip was leaning over the side of the boat, talking with the office on a cellphone (to try and fix the hardware problem we were having), and having to take brief breaks to feed the fishes.  After we fixed the triggering issue with the amplifier unit, I had them drop me off on a pier where I sat for the rest of the day while they ran the survey.

Amusingly enough, despite being my first major foray into programming with Labview, this software is still working today.  They've logged thousands of kilometres of data with it, and ran major surveys from Burrard Inlet to Lake Athabasca.  The processing tool is similarly bad, but it keeps on ticking.  As a tribute to its robustness (robustitude?), this was the first and last overwater survey that I had to do for the company.

(4)  The last entry here wasn't so much done for the company, as it was done for a side company that The Guru also worked for.  It was a geophysical instrumentation supplier that was getting complaints about two separate units they had shipped to a mine near Smithers, BC.  No one from that company had time to fly out there, so The Guru decided to try and suck someone else into it - namely me.  After being "trained" on the use and diagnostics for this particular piece of gear for about 30 minutes, they sent me on my way with one known working unit.

The flight into Smithers wasn't too bad, except for the wind shear when we started on our final approach.  No, it was the stall alarm going off constantly when the even smaller plane (probably a twin otter) was climbing out of the airport on its way to a dirt strip near the mine.  After a horrible flight (numb face + vomiting = not awesome), I felt like crap during the entire drive from the dirt strip to the mine.  The mine's doctor thought it may have been altitude sickness, as some people are more susceptible than others, but in retrospect I'm pretty sure that I had an anxiety attack which caused me to hyperventilate, combined with motion sickness.  Good times!

Being at the mine itself wasn't too bad, and I managed to diagnose and document the major problem with the mine's non-working unit.  I was then asked to train some of the management staff at the mine on how to properly use this particular equipment, which I remind you, I barely knew how to use myself.  While that went better than could be expected, I narrowly avoided being sent into the mine itself to show the miners how to use it.  Luckily they agreed that the additional safety orientation (necessary for me to go into the mine itself) was too much hassle, and so I did not travel into the bowels of the earth - but neither did I expose my massive ignorance of the actual field usage of said equipment.

In the end, the problem with the two broken units had been caused by another engineer skipping a step in the calibration procedure, because he thought it wasn't necessary.

Finally:  odd places you have worked?