An evergreen discussion among even the most seasoned programmers is when it's safe to use floating point numbers. Someone recently asserted that single-precision IEEE 754 binary float† was "good enough for money."
I don't recommend it. ...more
An evergreen discussion among even the most seasoned programmers is when it's safe to use floating point numbers. Someone recently asserted that single-precision IEEE 754 binary float† was "good enough for money."
I don't recommend it. ...more
As an exercise, my friend Andi suggested I re-implement the Unix utility yes(1) in assembly. Yes,
that thing that just prints y ad infinitum.
"You should re-implement
yes(1)," he says."That's easy!" says I.
"But," he countered, "with all the proper semantics of the C version!"
Let's get to it. ...more