Marcus G. Daniels marcus@cathcart.sysc.pdx.edu
12 May 1997 02:20:19 -0700

>>>>> "BS" == BRIAN SPILSBURY <zhivago@iglou.com> writes:

BS> We lose our measure of atomicity and imho that's
BS> the point of a transactional bd. 

Sorry for murdering the language here, what I want is something that has
executional persistence.

My point is that it could be possible to preserve the stack, etc. by
going fuzzy on granularity.  Also, I have a tendency to think in terms of
the likely nuts of bolts of the thing; if it is coded throughout as
transactions, then there's the option of having a mode where there is

BS> If you look at the solution I'm
BS> proposing its not hugely different from your example up there,
BS> except that its a more genearl case, and doesn't try to call page
BS> flushes transactions, it reserves that for world-global
BS> synhronization.

What is world-global synchronization and how does it occur?
Alternatively, how does one deal with disjointness in data?