From the post: > Why am I not using float64 time already??? [fp used ot be slow and was not standardized, but it got faster and after a while you could count

From the post: > Why am I not using float64 time already??? [fp used ot be slow and was not standardized, but it got faster and after a while you could count on having IEEE-754] This is not the reason. Until IEEE-754 fp was terrible. As in buggy terrible. It took years of pain to learn how to get it right. And you can see this in the standard: it's a very short standard, super easy to understand, trivial, really...if you ignore the enormous amount explanations, corner cases, special cases, surprising symmetries, and other copious testament to the well-meaning Mistakes that Were Made in the years before. Intel provided a ton of trade secret understanding from the mistakes in the 8087, and UCB maths prof was intimately involved. It's an incredible achievement. The history of the standard is quite exciting but has been covered on HN several times before.