What? I'm pretty sure Gates is the quintessential technical founder. He wrote the original Altair BASIC implementation himself in something like a week, and snuck into University of Washington's computer labs in the middle of the night in high school. I don't think I've ever seen anything that implied otherwise.
Bill Gates wrote that.
And was proud of it.
http://en.wikipedia.org/wiki/GORILLAS.BAS
(It was also QBasic, not GW-BASIC)
Donkey.bas, on the other hand, was co-authored by Gates, according to Wikipedia.
As far as never hearing anything about it, I based the statement on a couple personal accounts. But "Barbarians Led By Bill Gates: Microsoft From The Inside" has an account of at least one programmer calling Gates code terrible. The programmer in question co-wrote the book so you have to take it with a grain of salt but it seems true because the story he tells is one that paints Gates in a positive light (he doesn't fire the guy for calling his code junk right to his face)
This generally means that its not great, but it works. If time permits, it is great to tidy / refactor bad code, but this is not always possible. Reminds me of an item posted a couple of weeks ago - "confessions of a terrible programmer"
The year is 1974, and Linux, Mac, and Windows haven't been invented yet. There is no IDE, no SDK, and no documentation. Actually, there's no development environment at all. There are no APIs, either. Ruby, Python, Java, and friends are till 10 years or more from being invented. And yet, Gates wrote an interpreted language, on his own, in a week. Put differently, Gates wrote a language in a CAVE, with SCRAPS, in a WEEK.
It's easy to think that now we'd just Google it and scrape it together using copypasta if we really needed to. Hell, not only was there not Google, but there was no technical community to ask questions. Don't forget what he was really working with there.
It doesn't?? Then what does? Quick and usable in the business world is practically the holy grail of programming!!!