Hi, All ! I would like to have 48 bits integers and 48 bits floats (like in the TI FPP). These will solve the precision problems in 99%. Also I think that TI floating point format is more optimal then ieee standard (Of course I do not know is the TI FPP library code is as good as the number format :).
Also do we really need to follow any standard in such a microcontroller like MSP430 ? What should we get from having standard floating point format ? (I do not think that one needs to process any "not a number" situations in MSP430 project) Actually I use inline assembler to work the 48 bits integers for the now, but it looks very ugly :(. Thanks, Oleg..
