I'm working with an mxj external right now that is outputting doubles to max that are really really small (like 1.1E-30) and max floats don't hold that many decimal places, however I want to compare these types of numbers in max?

Should I just continue to write the comparison part of the program in Java? Is there a way that max can retain lots and lots of decimal places?

Thanks, if anyone is interested, the patch involves hidden markov models and is still in the testing phase, but I hope to share it with the max community very soon!

java-doubles-to-max-floats

Max only represents 32 bits floating points values, so you might want to write your comparison in Java.

Java Doubles to Max Floats