b2676870@rick.cs.ubc.ca (wilson w ma) (05/22/91)
In article <1991May21.120147.4627@saaf.se> pausch@saaf.se (Paul Schlyter) writes: >Of course. THis is a limitation inherent in using BCD representaion for >floating-point numbers. 1/3 is 0.33333333....... infinetley, so it must >be rounded off somewhere. Suppose we round off after 12 digits: >0.333333333333 Now we want to multiply this with 3 - the answer to this >IS 0.999999999999 to 12 decimal places. [deleted...] Are you saying that the 48sx uses BCD? (I posted an article previously asking about floating point radix base and no one answered...) I need to know because I'm translating some code from Numerical Recipes. >Paul Schlyter, Swedish Amateur Astronomer's Society (SAAF) >InterNet: pausch@saaf.se >FidoNet: 2:201/600.2 George
bson@rice-chex.ai.mit.edu (Jan Brittenson) (05/22/91)
In a posting of [22 May 91 03:18:14 GMT] b2676870@rick.cs.ubc.ca (wilson w ma) writes: > Are you saying that the 48sx uses BCD? `Decimal' is probably a better term. 12 digits mantissa, one sign digit, and three digits 10's exponent (10's complement, i.e. 999 = -1, 998 = -2, etc). -- Jan Brittenson bson@ai.mit.edu