[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Exactness

This page is part of the web mail archives of SRFI 77 from before July 7th, 2015. The new archives for SRFI 77 contain all messages, not just those from before July 7th, 2015.

bear <bear@xxxxxxxxx> writes:

> Instead of saying "rely on inexact numbers being floats", can
> you enumerate the qualities of floats that you find important?

I already listed them among the 10 properties. Of course most numeric
programs don't need all of them, but I don't want to care about each
aspect separately each time. I trust IEEE designers, I know that
common pitfalls are taught, that methods for estimating errors are
known, that the knowledge about which algorithms are numerically
stable exists independently of the language, and that people who
implement numeric algorithms know how to apply floating point to them.
I don't want each language to introduce a distinct set of pitfalls.
It's a too delicate territory to be creative without a good reason.

> I think that the problem here is that a lot of people (myself
> included) do *NOT* wish to discourage implementors from doing
> better (producing exact results more often) than the floating-
> point format would allow.

They have been encouraged for many years. Why didn't do they?

   __("<         Marcin Kowalczyk
   \__/       qrczak@xxxxxxxxxx
    ^^     http://qrnik.knm.org.pl/~qrczak/