EPANET uses a MAGICNUMBER to "identify" the binary output file as having been written by itself. EPANET and EPANET-MSX both use the same MAGICNUMBER and write it to the header and footer of the binary output file.

Since they format the binary data differently, it would be useful to use a slightly different MAGICNUMBER in MSX so that it would be easy to distinguish between binary files written by EPANET from those written by EPANETMSX after the fact.

This would allow us to provide more meaningful error messages to users or write a library that supports both formats.

I wanted to recommend that in the instructions/documentation for compiling using the makefile that users check for machine-specific architecture flags. In the past, I compiled epanetmsx on a supercomputing cluster, and had initially had very slow run times because I was not aware of the modifications I needed to make to the makefile. After updating the makefile with the correct architecture flag, the runtime of epanetmsx was as I expected (much faster).

When fixing this issue, here are resources I found helpful:

https://stackoverflow.com/questions/45933732/how-to-specify-a-compiler-in-cmake
https://stackoverflow.com/questions/11783932/how-do-i-add-a-linker-or-compile-flag-in-a-cmake-file

And here is the modification I made to the makefile in my case:

SET(CMAKE_C_COMPILER "icc")
SET(CMAKE_C_FLAGS "-xCORE-AVX2 -axCORE-AVX512,MIC-AVX512")