a less sevear SOS

I have a small algorythm problem, it should be a common algorythm. Fixed point multiplication, division and subtraction
i just cant seem to get them working, an example of this would be 12.000233 - 32.000212
the fractional part will always have 6 places to it
and the way it must be handled in these overload functions is as follows. A fixedpoint object is declaired with a integral member and a fractional member, both are ints, we cant use doubles or floats as the part to the left of the decimel (ugh).
if anyone has an algorythm for this handy please send it, if you are going to write it from scratch dont worry about it, too much work
i can attach what i have so far if someone wants to see it
btb project is due tomorrow at midnight, thats about 10 hours from now

So what's the question - apart from leaving it far too late for anyone to do anything practical in terms of helping....

I mean
http://www.google.com/search?q=fixed+point+arithmetic
has lots of info, and no doubt code samples

its hard to say, just looking for what help i can get, and considering how much i had to do for this project and how much free time i had i think i did pretty good

salem, you are awsome, that link had a link to a site that had the most obvious algorythm to do this stuff, never would have thought of it