My life as a sarariman

Languagehat has an interesting link up to a site on the origins of Japanese company names. Although they're all well known, perhaps the best-known Japanese company name isn't on the list, Sony. Your host's first job was with Sony, so he's going to use this as an excuse to talk about it.

But first, the name. Sony started out as Tokyo Tsushin Kogyo Kaisha, and was dedicated to the production of radio receivers (shades of the Walkman!) and electrical measurement equipment. Readers familiar with Sony only through their consumer electronics may not realize how huge Sony is today in the latter department. In any event, as radios and consumer electronics came to be the company's main claim to fame, they went for a name which both reflected that fact and was easier to sell to English speakers. Sony is claimed to be derived from both "sonus", Latin for sound, and "sonny", because they liked the suggestion of youth that it provided. SC's pet theory is that someone misspelled "Sonny" when silkscreening it onto a batch of parts, and the "sonus" justification was invented post hoc, to save money on having to make more.

In any event, in 1994, your host was just wrapping up high school, and wanted a job to keep him busy over the summer. Thanks to a friend of Dad SC's, who had just become legal counsel for Sony Imagesoft (a now-defunct subsidiary devoted to making video games), SC ended up with the job you all only wish you had -- as a video-game tester.

It wasn't actually a "sarariman" job -- that's the Japanese transliteration for "salaryman", and refers to a white-collar worker. SC had more of what they'd call an "arubaito" -- again a transliteration, from the German for work (arbeit). Arubaito implies part-time, but SC worked 40-hour weeks, albeit as an hourly independent contractor (in order to keep from having to provide benefits, testers were not officially employees of Sony Imagesoft).

Now SC is going to tell you why you're wrong, and you don't actually want to be a video-game tester. You probably think it means sitting around all day getting paid to play more-or-less finished product, having fun, and just making suggestions for how it might be tweaked to do better in the marketplace, much the way that focus groups are brought in to help edit films prior to release. You would be very, very wrong.

Like all software, video games have to be debugged. Unlike trivial things like the operating system that keeps all of your personal data, and the bank databases that keep all of your financial records, video games are held to very high standards for reliability. Crash bugs are simply not acceptable. So you have to go do the same thing over and over...and over and over...again, but with slightly different variations on all the possible key-pressing combinations until you get something to break. Early in a game's lifecycle, this is easy. Later, it's hard for the people at the development shop, but simplicity itself for the testing gods employed by the hardware makers. Your dedication to quality must be truly fanatical in order not to go screaming on your third consecutive day of testing menus when there's no actual game code to be played -- SC is glad he never found out what would happen to his mind on a fourth such consecutive day. Sony was not yet in the game hardware business when SC was there -- we had sneak peeks at the upcoming Playstation, but didn't have any development projects underway -- but Nintendo and Sega had some truly amazing people putting pressure on us. Despite the inclination of most people to think that playing games for money is stress-free, when an executive producer comes into your cubicle to let you know the release you swore was rock-solid actually is going to need another round of debugging, your day can be quite terrifying.

Your host was involved with the production of two games, done simultaneously on two platforms. Back then, Sony had the rights to produce games under the ESPN name, and so we worked on ESPN Speedworld and ESPN National Hockey Night (some better screenshots of Speedworld can be found here). Both were done in Super Nintendo and Sega Genesis variants. The release dates listed for both games on the linked site are wrong -- your host left Sony in August of 1994, and neither game was out on either system at that time, although Speedworld had been approved for production. Both made it to the stores in time for Christmas of that year.

You learn some interesting things working as a game tester. It's not commonly known, but there are five different versions of the Super Nintendo out there, distinguished by sound hardware alone. Sony provided the first-generation chips, but they and Nintendo had a falling out over the CD-ROM drive that ultimately became the Playstation, and so Nintendo had to find another supplier that could produce something backward-compatible without violating Sony's proprietary rights. This was problematic, and Nintendo ended up going with four different iterations of chips over the lifetime of the system. Games had to be seamlessly compatible with all of them, and they all had different quirks. The hardware makers provided different levels of access into the machines as well -- while Nintendo only allowed debugging information to come up through an expensive software package run on development machines (which meant the testers never got to use them), Sega had a nifty little test rig called an address checker, which would display the last memory address accessed (in hexadecimal) when the system crashed. This device plugged into the console, and EPROMs holding the games would be plugged into the address checker. Real cartridges were only ever put out at production time, and so we'd all stab ourselves frequently trying to pull the bare chips back out of the consoles after testing. There's an art to doing that right, but SC never got the hang of it.

We had a neat Easter egg in Speedworld that had to be pulled out because it caused the regular game to crash sometimes. It was a version of Pong featuring two clowns on Pogo sticks, who bounced a squid on a plate back and forth between them. Sending the squid flying back towards your opponent didn't occur automatically on contact; you had to press a button to make your clown whip the squid. It looked ridiculous, but none of the testers could get enough of it.

Like any creative work, there's a certain pride that comes in seeing your name in the credits when a game is finished. So although your host had already played ESPN Speedworld thousands of times, with all of the possible combinations of wheels, spoilers, engines, transmissions and racing season modes, he still bought a copy when it came out. Unfortunately, he didn't work on National Hockey Night long enough to be credited (they only kicked it over to the testing staff a month before he left), so he never bought a copy of that game. Since he never owned a Genesis, there are no copies of the Genesis versions of either game in Chez SC, either. But every now and then, he'll pull out his copy of Speedworld and remember when he had the job that everyone else wants.

Comments

[david]

I'm not a gamer, but your description of the "job" confirms what I've always suspected about what many young'uns believe to be a "dream job!"