Re: What makes an OODBMS different from an RDBMS?
Date: Wed, 28 Nov 2001 14:00:15 -0000
Message-ID: <9u2qjf$9k7$1_at_pea.uk.research.att.com>
"Carl Rosenberger" <carl_at_db4o.com> wrote in message
news:9u2mps$dsm$05$1_at_news.t-online.com...
> Tim Mills wrote:
> > This sounds like a shadow-object algorithm, similar to shadow-paging but
at
> > an object level. By what method do you ensure that allocation of free
space
> > and the pointer update is atomic and durable?
>
> The pointer update currently is not atomic.
> Are you working on something similar?
I've been developing an ODMG/C++ OODB for about 6 years. I tend to follow the advice in Grey & Reuter's Transaction Processing Concepts and Techniques. That, together with C. Mohan's work at IBM on the ARIES recovery scheme, and Leonidas Fegaras work on implementing OQL provide a fairly complete basis for implementing an OODB with standard ACID transactions and a decent query language.
Cheers,
Tim Received on Wed Nov 28 2001 - 15:00:15 CET