Robert A. Rosenberg has rehearsed the obvious, focusing on the details
of inventory management.

My point was a different one.  It was to illustrate in a homely way
that serialization requirements are ineluctable.

It is of course possible to avoid the explicit use of the word
'serialization', say by using the mnemonic CS[D], Compare and Swap
[Double], instead.  It is not possible to avoid its substantive use to
resolve 'race' conditions satisfactorily.

This topic has always been a vexed one that, from time to time, still
gives rise to Luddite denials and Lysenkoist excursions like
Dijkstra's semaphores.  Its pursuit in fora like this one is, I
suppose, all but hopeless.   I should have known better.

John Gilmore, Ashland, MA 01721 - USA

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to