Dr. Robert J. Harley harley at
Wed Sep 10 00:09:25 PDT 2003

Apparently SCO has a "team" of "experts" comparing Linux source code
with their code.  They claimed so a few weeks ago without saying who he
was.  Apparently ESR has "developed a utility" that "looks for common
code segments" (,4149,1257617,00.asp).

Anyhow, I wrote such a utility about ten years ago (based on an
undergrad project at Edinburgh U. a few years previously) and used it
e.g., to remove almost-duplicates from Adam's collection of notable
quotables.  I would be interested in using similar (warning: T-word
coming up) Technology to efficiently find matches (including close
rather than exact) between source-code bases, at least those that are
publicly available, such as the Linux kernel, *BSD sources, the early
Unixes etc.  I thought of suggesting same a while ago here on FoRK but
the main issue is the man-hours involved in management and interpretation
of the output and I don't see any way of generating revenue from it.
Kudos and all is cool but doesn't pay the bills.  Any ideas?


More information about the FoRK mailing list