This draft paper is a rather substantial rewrite of
Items 19, 27 and
I immodestly consider this the most important as well as the best paper I have written.
I wrote this report as my response to the work of a
colleague at TUV.
I simply could not understand that work,
and I was not quite sure whether the described language cum system had a semantics.
Also the "interpretation" of the (two phase-commit) transaction protocol seemed problematic.
So in this technical report I present, probably for the first time,
a formal description of that famous protocol.
This technical report also gives formal specifications of Window systems,
and a rather free-wheeling (but still precise) form of SQL-like relational data bases.
Papers [93,94,95] are "near identical"