For data recovery purposes (more precisely : to sort out files recovered through “raw file carving” which are in fact parts of valid files successfully recovered through filesystem based recovery — or similar situations), I would need a command line program, or a PowerShell code, that would compare two files, starting from a specified offset, usually different for each file (usually those are small files contained within bigger files — for instance a MPG file which starts as part of a valid VOB file, but which continues with fragments of other files, and may contain fragments not belonging to any file that was still registered in filesystem structures at the time of the recovery), until the first difference is found, and simply report the offset of the first difference, reliably and quickly. I've already asked this at SuperUser, but didn't get a usable solution. The only existing tool I found which apparently does exactly what I need, “cmp”, is part of (as I understand it) an obscure IBM industrial-level operating system, and doesn't exist as a standalone executable.

An alternative method would be to calculate checksums of partial areas in files of a specified size, then report matching areas ; for instance, set the area to 1MB, calculate the MD5 for each MB of each processed file, store them in a temporary list, then report all areas that have been identified as identical. (WinHex has a feature called “block-wise hashing and matching” which I hoped would work for this task, but it turned out to be totally impractical — details in the aforementioned SuperUser thread.)

If not here, where could I ask something like this with a good hope of getting satisfying replies ?