I find it hard to imagine why I would ever want to use something like this when there are so so many free and open alternatives. Can someone give me an example where this would fit a need? Jt.
-----Original Message----- From: Tim Colson To: [EMAIL PROTECTED] Sent: 1/20/03 8:05 AM Subject: RE: [jug-discussion] prevalence layer (prevayler) > This is a persistence layer that, guess what, persists your objects > without a database. Very interesting. Has anyone heard of or > used this yet? Yes, I ran across it a few months ago and my workgroup had some discussion on the topic. IIRC we came to a few conclusions: 1) interesting idea for small to medium-ish datasets (but unclear how to scale to datasets larger than available RAM) 2) untested for clusters, and potentially resource wasteful (dupe datasets eating dupe memory) 3) project looks like the effort of one person 4) minimal documentation (I see that months later, the link to the Tutorial still hasn't been re-written) 5) features of a DB beyond simple persistence that might need to be re-created (ex. full-text searches, referential integrity - some folks didn't trust the java dev to maintain this in the object graph itself for various reasons) 6) not as many tools available (ex. reporting) Summary - it's an interesting idea, but we decided to stick with an RDBMS. Tim --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]