I won''t speak as to the other responses, however, what I do know is:

1) During the base design of the intel architecture, the researchers had to base the ''bit'' on the chipset output or the pins that fed data to the bus (this happened to be 8 pins at the onset).

2) The most efficient method for transposing the most data on 8 bits (pins) was by using the binary (Base2) system since each pin could only represent two states (on or off) and therefore the highest numerical range attainable in those 8 bits was 0-255.

3) The earliest assemblers were designed to represent information in only 1-2 characters of space (text). Since binary could take 8 digits and decimal could take 3 digits to represent the larges possible numbers on an 8 bit archetecture the hexidecimal system was used for convenience and practicallity reasons.

borelli33

The uses for BCD almost all revolve around perfect precision decimal math or conversion to and from decimal.

People who need to keep perfect financial quantities almost always use BCD ... because they must be stored in a format that quarantees every part of every cent ends up in one or the other persons back acount ... and doesn''t get rounded, truncated, or given to the bank ....

Also ... think of the step of goin the following ... given a list of numbers, add the first 2, multiply by the third, and accumulated the total (basic compounding interest problem) ... if the numbers are ENTERED in decimal ... then it will be quickest to perform an identical, MMX style per-digit conversion from ASCII to BCD ... and then process the math the way you learned in school .. then you actuall build a binary int from decimal string, just to do the math and then rebuild the result as a decimal string ...