Re: Various notes about the TI 92
[Prev][Next][Index][Thread]
Paul Pollack (paulp@televault.com) wrote:
: Aaron Bergman wrote:
: > My guess is that the 92 probably tries substitution first and it
: > comes up with 0^0 which is often defined to be 1 to make certain
: > formula prettier.
: I think that may be it as well - if so, that just means that the 92
: evaluates limits in a horrible way.
Well, substitution works in all other cases for continuous
functions, so it's not horrid. They just forgot about this
special case for 0^x which isn't continuous.
:
: > But 0^2 is a positive zero, you see. It makes a weird sort of
: > sense, really. Lim x->0 of 1/x^2 = +oo, for example.
: I'm not saying I don't understand why it did that; I thought of the
: above example myself (and it fits with the fact that the 92 gives 0^3,
: 1/ans(1)=undef), but no matter how you cut it, the 92 is still wrong
: on this.
: It doesn't make sense for the 92 to say that 1/0 is undefined and then,
: two or three lines later, to say that it's equal to infinity. The 92's
: symbolic manipulation routines leave much to be desired when it comes to
: dealing with infinity.
What I find interesting here is how this reflects on how the calc
stores the number. Obviously, it must be storing 0 differently
from 0^2. I'm not quite sure why they would do that. They must be
storing 0^2 as 0+, I guess.
:
: > Why not implement RSA? :)
:
: Good idea. You do the implementation and I'll beta test it. :)
Well, the theory's easy. What's the maximum integer size on the
92? It'd be pretty easy to come up with something that would work
somewhat slower than molasses in winter.
Aaron
<pre>
--
Aaron Bergman -- abergman@minerva.cis.yale.edu
<http://pantheon.yale.edu/~abergman/abergman.html>
The quote left intentionally blank.
</pre>
References: