7

A comment on the question Why did 8-bit Basic use 40-bit floating point? says the following (emphasis added):

Re, "Nowadays, floating point is usually either 32 or 64 bits." More specifically, floating point is usually IEEE 754 floating point. Pior to the publication of that standard in 1985, it was the Wild West: Pretty much every system had its own internal representation for floating point numbers. I saw a book once that was basically a step-by-step instruction manual for how to construct your own floating point system. It was on the desk of my co-worker who was implementing the floating point arithmetic for a Lisp interpreter.

Some keyword searching turned up only Jean-Michel Muller's Handbook of Floating-Point Arithmetic, which was published in 2009; presumably the comment is discussing something older.

What is this book? It sounds quite interesting.

texdr.aft
  • 3,495
  • 1
  • 19
  • 42
  • 1
    What exactly is the question? Are you asking about the linked book (Handbook of Floating-Point Arithmetic) or the IEEE standard book? If the first, you may need to acuirre the book. For the later, simply follow the Wiki link as it lists all relevant documents. – Raffzahn Feb 19 '21 at 11:33
  • @Raffzahn The one that the commenter's co-worker was reading, referred to in the bolded part of the comment. – texdr.aft Feb 19 '21 at 11:36
  • 2
    I suspect it would be more effective to make this a comment on the relevant post; that way you could tag Solomon Slow directly to notify them. Only they can answer the question, others can only guess... – Stephen Kitt Feb 19 '21 at 11:59
  • In that case, opening a question isn't much helpful. Post a comment to that answer addressing the one who wrote this ( @SolomonSlow), as this will ping him on the issue. – Raffzahn Feb 19 '21 at 11:59
  • 3
    I agree that the right way to find out is by posting a comment over at the original question. This question doesn't have enough information to pin it down. If you are interested in books going back to the 1970s or 1980s, start by researching Don Knuth or Jerry Sussman. – Walter Mitty Feb 19 '21 at 12:20
  • 4
    Asking a new question rather than commenting on the existing question is a valid approach. A comment could be added to the other question linking to here, which would ping the original poster. – Chenmunka Feb 19 '21 at 14:01
  • @Chenmunka Really? Spreading out a personal conversation over a series of questions? – Raffzahn Feb 19 '21 at 19:15
  • 1
    @Raffzahn It's how Stack Exchange is designed. :shrug: – wizzwizz4 Feb 20 '21 at 16:25
  • 2
    Were the description in the comment a little more detailed, it might as well be someone else who is able to find the book in question. With the one we have, it might be harder, but still not impossible in principle; it might have in fact been a very distinctive book. – user3840170 Feb 21 '21 at 06:19
  • 3
    As the author of the comment in question, I will repeat here that it was many decades ago. I've lost touch with the person who owned the book, and when I tried to search for something similar a few decades later, I turned up nothing. "The Handbook of Floating Point Arithmetic" cited above looks like it might contain some useful information, but it also looks like more of an academic work--unlike the "how-to" manual from my memory. – Solomon Slow Feb 21 '21 at 17:15
  • 1
    @SolomonSlow Would it happen to be Pat H. Sterbenz's Floating-point Computation from 1974, readable here? "This book is directed toward two different types of readers. First, it is addressed to the obvious audience of those who are interested in using higher-level languages to write programs which will perform floating-point computation. Second, it is also directed toward the compiler designers and machine designers who are concerned with floating-point operations." – texdr.aft Feb 22 '21 at 03:35
  • @texdr.aft, IDK, but nice find. I think I'm going to read that one. – Solomon Slow Feb 22 '21 at 12:58

0 Answers0