Nick Krempel (2014-01-18T02:13:03.000Z)
On 17 January 2014 19:25, Jens Nockert <jens at nockert.se> wrote:

> > On H26/01/17, at 20:12, Nick Krempel <ndkrempel at google.com> wrote:
> >
> > If you go with a representation-independent 'bitlen', I would suggest
> returning NaN for any negative number.
>
> infinity would make sense for a signed two's complement representation.


Yes! You win! +Infinity (*not* -Infinity) is now my top choice. (The
mathematical version of what you said is the 2-adic expansion. For example
the 2-adic expansion of -1 is ...1111111 - carrying on infinitely to the
left. First google hit:
http://www.cut-the-knot.org/blue/p-adicExpansion.shtml)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/es-discuss/attachments/20140118/562aee50/attachment.html>
domenic at domenicdenicola.com (2014-01-24T19:56:26.513Z)
On 17 January 2014 19:25, Jens Nockert <jens at nockert.se> wrote:

> infinity would make sense for a signed two's complement representation.


Yes! You win! +Infinity (*not* -Infinity) is now my top choice. (The
mathematical version of what you said is the 2-adic expansion. For example
the 2-adic expansion of -1 is ...1111111 - carrying on infinitely to the
left. First google hit:
http://www.cut-the-knot.org/blue/p-adicExpansion.shtml)