POS

[Marcus G. Daniels]

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

BS> This basically means _everything_. Now I don't know about you, but
BS> transactional updates on my stack are likely to be at least
BS> several orders of magnitude slower

Suppose you do _everything_, can't these tiny commits of the stack
just be roundoffs to set a bits in a dirty vector?  Then after N of
those things commit all the dirty pages?

BS> Then we can roll back to the last checkpoint by using these
BS> 'originally clean' blocks. Here a transaction is a 'commit-system'
BS> call. We can recover to a non-commited state however, if we can
BS> handle a (hopefully small) level of disjointness in data. 

Uh, sorry for being dense.  Where does the disjointedness sneak in?
If it sneaks in at all, that sounds to me like that is pretty much
hopeless lossage (I guess I don't understand what you mean by disjointedness).