Invention of threads

[John Wexler <wex@xxxxxxxxxxxxxxxxx>]

I still have the source code of an operating system called BETS -
basic engineering test supervisor - which we wrote some time around 1970 to
investigate the behaviour of the rather untrustworthy System 4 implementation
of IBM's 360 architecture.  This provided for programs to divide into threads
and for everything that follows from that.  For what it's worth, it provided
a set of facilities for interprocess communication and synchronisation which
matched almost one-for-one the Occam features of channels, PAR and ALT.  At
the time, what we were doing was nothing surprising; not exactly commonplace,
but not revolutionary either.  Some of the ideas we picked up from IBM, and
some came from elsewhere.  So I don't think these ex-IBM "inventors" have got
a leg to stand on.  On the other hand, if I were called to give evidence in
support of Microsoft, I can't say how reliable my memory would be.
    John Wexler