Re: Floats? (was Re: Wishful thinking)

• Date: Wed, 17 Jan 1996 11:48:59 PST
• From:
  Christopher Scott Delaney <cdelaney@indiana.edu>
• cc: moo-cows@parc.xerox.com
• Content-Type: TEXT/PLAIN; charset=US-ASCII
• In-Reply-To: <9601171915.AA29947@tom.eng.mvision.com>

On Wed, 17 Jan 1996, Tom Ritchford wrote:

> 
>     I'm very dubious about adding floating point numbers.
> 
> Knuth points out that floats are almost never useful unless
> 
> a -- You're dealing with really large or really small values (6.023 x 10^23)
> 
> b -- You have not taken the time to work out the reasonable
> range of output values.
> 
> Of course, if you had a MOO that represented the entire universe
> you might need it... but I'm sure that there would be a lot of lag
> if you had one object per star...
> 
> The problem is that floats are slow and can be large.

We've used floating point functions for over a year now on a MOO I'm 
helping to develop. (read: not yet open to the public).  They are, in 
fact, very useful and not slow at all if coded correctly.  And we aren't 
using them to represent the entire universe, just the area surrounding 
Jupiter and its MOOns :)

--Chris/Barrett

• Re: Floats? (was Re: Wishful thinking)
• From: tom@mvision.com (Tom Ritchford)

• Floats? (was Re: Wishful thinking)
• From: tom@mvision.com (Tom Ritchford)

• Prev: Floats? (was Re: Wishful thinking)
• Next: Re: Floats? (was Re: Wishful thinking)
• Index: MOO-cows
• Thread: MOO-cows