Actually, the use of "code" as a count noun is quite old and well entrenched in the numerical modelling community. I can recall back in the early 1970s hearing Fortran programs routinely referred to as "codes".
Oh and this was on campus at CIT for a world leading RnD organisation in their math modelling section.
[0] http://people.sc.fsu.edu/~jburkardt/f_src/f_src.html
[1] http://www.unige.ch/~hairer/software.html
[2] http://www.mcs.anl.gov/project/openadf-automatic-differentia...