Path: utzoo!attcan!uunet!van-bc!ubc-cs!rick.cs.ubc.ca!b3300876
From: b3300876@rick.cs.ubc.ca (george kai yee chow)
Newsgroups: comp.sys.handhelds
Subject: Re: Two questions...
Message-ID: <1990Sep27.170631.6885@rick.cs.ubc.ca>
Date: 27 Sep 90 17:06:31 GMT
References: <1990Sep24.183934.575@rick.cs.ubc.ca> <36159@cc.usu.edu> <15643@shlump.nac.dec.com>
Sender: news@rick.cs.ubc.ca (Usenet News)
Organization: Rick Lab, CPSC, UBC, Vancouver, B.C., Canada
Lines: 24

In article <15643@shlump.nac.dec.com> edp@jareth.enet.dec.com (Eric Postpischil (Always mount a scratch monkey.)) writes:
>In article <1990Sep26.160424.12956@rick.cs.ubc.ca>, b3300876@rick.cs.ubc.ca
>(george kai yee chow) writes:
>
>>Yes, I'm aware of the problem of calculating with pi. But as I recall reading
>>in the 48sx's manual, the symbol modes was suppose to elminate that. On the 
>>48sx, you can ask to keep pi as a symbol to avoid the rounding error.
>
>There's a limit to how far the calculator can go with symbolic manipulation. 
>The calculator will preserve 'pi' as a symbolic in expressions.  When taking
>sines or cosines, it will recognize pi and pi/2 and return exact results.  In
>other cases, it does not have exact results prepared for fractions of pi.  So,
>to complete the calculation and return a number, it must convert the symbolic
>'pi' to a number and perform numeric calculations.

Another reason why I mentioned it is because both of my Casio's (fx-451M and
fx-8000G) return cos(pi/6) correctly and clearly they only use an approximation
of pi.

>
>
>				-- edp

George