help-octave
[Top][All Lists]
Advanced

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

Re: uint64 typecast can't be initialised with more than 15 decimal posit


From: Jordi Gutiérrez Hermoso
Subject: Re: uint64 typecast can't be initialised with more than 15 decimal positions
Date: Mon, 4 Mar 2013 13:17:39 -0500

On 4 March 2013 13:10, Jordi Gutiérrez Hermoso <address@hidden> wrote:
> When your computations exceed this precision, you'll get wrong
> results that if I recall correctly, as undefined behaviour according
> to the C++ standard (perhaps they're implementation defined instead
> of undefined, can't remember).

And in the Matlab and hence Octave languages, if you exceed the
integral precision during a computation you get the fantastically
bizarre saturation semantics that are actually really tricky to
implement in C++. Did you know that these saturation semantics are why
Octave had 64-bit integer arithmetic implemented before Matlab? Octave
won a pointless race with Matlab here that nobody knows about and
nobody cares about.

- Jordi G. H.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]