I have tried source compiling of 1.55
and it was 4 passes 0.4 seconds 66048 bytes ( dos version )
the windows version 6 passes 0.9 seconds 104960 bytes
on my 933 Mhz ( 600Mhz ) celeron

however in the windows version there are a few differences between the binaries:
a few "6E 52 3E" blocks have changed to "28 10 37"
and
a few "ED 6C 02" blocks have changed to "62 F3 01"
but not interfering much with the workigness
MATRIX

It would be a good idea to try to run EXACTLY the same sources on many
different FASMs down to some pre 1.4x when some great changes were
made in FASM.
Drawing a graph would lead us to new problems/solutions and show the
developement of <speed> so far.

Matrix, isn't that just some header/PE structure data (like a timedatestamp), or perhaps uninitialized bytes?

i have total commander, and file compared is hexa mode, see for yourself.

MATRIX

Time stamp?

Yes, PE files contain the 32-bit time stamp - this value will be different each time you generate the executable.

lol ,
what a feature

so i can hold my data there? joy

many free variables :DDD

MATRIX