Anyone understand how a double is written to file in binary mode??

I've examined it till blue in the face and can't figure where to place the
decimal when decoding when retrieving data as a byte array.

My goal is convert a bunch of data to byte array for sending via winsock.

I know it can be done by coercion with variants but I'd like to do the math
myself.

Thanks,

dj

IEEE doubles are complicated.  They're stored in 8 bytes, laid out like
this (the order of bytes is reversed here for clarity):

 SEEEEEE EEEEMMMM MMMMMMMM MMMMMMMM MMMMMMMM MMMMMMMM .....

...where S is the sign bit, EEE are exponent bits, and MMM are mantissa
bits.

The exponent is biased +1023, and the MSB of the mantissa is implied.

Decoding these is not trivial, and even less so in a language like VB
that has minimal support for masking and shifting.

                -- Jim

Thanks for the reply Jim.  At least now I understand why I see the
workarounds for this issue. (And why the decimal point kept moving
around on me).   I spent the last two weeks with my 8th grade Geometry
book (circa 1971), lord knows which one I'll have to dig out for this
<g>.  

Thanks for your time,

dj