I was thrown in feet first working on Links LS 1999 Edition. This was the first iteration of Links to allow St. Andrew's-style pit bunkers, as well as the first iteration to support 3D objects. I was brought in near the tail-end of the development cycle, so the development team was really starting to crack down on bug submissions.
Let me see how I can best sum up the Access way of doing things: ship it and patch it until it was right. Access had their cost structure set up in such a way that support calls weren't the death knell that they are to the rest of the industry. They also had a bit of a secret margin on every unit (more on that later).
As a result of the Access way, testing was pretty much an afterthought. The test plan was to throw the test department and the tech support staff at the game until it broke, fix it, and repeat. The test department received the manual for a tech review three days after it went to press. Needless to say, it was an interesting experience.
There were a few things that surprised me, however. Very few people worked on Sundays. Admittedly, this is a very religious state, but in other local game companies, Sunday was just another day. The only real pressure in the office was pressure that you placed on yourself to do great work. I remember Les Oswald re-editing video clips of the courses over and over again because they just didn't feel right to him.
My most vivid memory of that product was the first bug I had to fight for. Near the first tee box on St. Andrews Old Course is a little shack. Due to a bug in the 3D model, you could see through the undersides of the shack roof. (The normals were facing wrong on the polygons.) It basically looked like total ass. I bugged it, and it came back "Won't Fix." It took me nearly two weeks of arguing up the chain of command to get it fixed. The fix took less than 20 minutes to find and fix. This series of events is one that becomes all-too-common for a tester.
So, what was the secret margin on each box of Links LS 1999 Edition? Would you believe that it was the box itself? Of course you wouldn't. That's why it was a secret.
Near the end of his service at Microsoft, he may have become a raging egomaniac because of PC Gamer calling him a "Game God," but you can say one thing for Bruce Carver...he was a hell of a negotiator. Due to his price negotiation skills, the price for the box was about one-fifth the price that other companies paid for their boxes. On Microsoft Golf 2001 Edition, a plain, non-embossed box with a paper sleeve to hold the disk and no manual had a total materials cost of goods of nearly $10, whereas the LS99 packaging, with multiple CD's, Arnold Palmer embossed on the package, a real jewel case, a full manual, etc., and a flap on the box was sub-$5.
In an industry where every dollar extra on a franchise can amount to boucoup bucks, Bruce's packaging helped ensure a steady revenue source...especially since the packaging for the courses was even cheaper.
He bought the disk cases in bulk from cheap manufacturers. He bought the boxes from local companies that wanted the recurring business and were willing to offer discounts for repeat customers. His shipping and packaging staff was a skeleton crew. He only cello-wrapped the course boxes, not the main game. He took care of the disk replication locally. Finally, in order to get the game packed and shipped to stores in time, he called in on his secret weapon: every employee at Access.
Any other time, his skeleton crew could handle the packaging and shipping, but not on day one. Most of the developers were salaried. Everyone else was pretty cheap. So for one day, he sent everyone in the company over to the warehouse, and the entire company packed up the game. I shifted between slipping the liner art into the front cover of the jewel case and putting disk #3 in the case.
Remember: if you already have a sunk cost in staffing, and it would cost an inordinately large amount to bring on temp staff to handle a job for you, screw the temp staff. Use your currently sunk cost.
It was an interesting experience that led me into my first lead assignment, Links Extreme, which is where I learned my first bit of what test should not do. It was also the beginning of the end for many things. More on that later.
By the way, if any old Access/Microsoft cronies read this and see mistakes, please post them in the comments so I can correct them. Thanks.