I have to agree with this. Relational databases are relational for a reason. If you're in a situation of needing to make the same queries constantly I have to recommend memcache: http://www.danga.com/memcached/
We have this running between our php site and our postgres database. We can handle many times more traffic than we would be able to normally. It takes a little bit of dev work, but is worth it entirely. -Alex Alias™ wrote: > I have to say, I'm not sold on OO databases. > > Databases, ultimately, need to be two things - accurate, and fast. > > Generally, OO means trading performance for maintainability - in the > case of code, that's a substantial gain, but I'm not so sure that's > going to necessarily apply as much to data. Large databases need > constant care and maintenance, and I'm not sure that OO would add > significantly reduce the need for that - IANADBA, but I suspect the > opposite. Why do you think DBAs are among the most highly paid IT > workers? > > Obviously it would be nice to store data in a nice object format, but > why should the database under that have to conform to that if it's > going to be less efficient? A slow database is the kiss of death to > any web app. Using some kind of middleware to serialize the data into > table format seems like the best approach to me. > > Alias > > > On 11/09/06, *João Saleiro* <[EMAIL PROTECTED] > <mailto:[EMAIL PROTECTED]>> wrote: > > Sorry for this being slightly off-topic, but I think this is an > important subject. > > We use and code frameworks design patterns based, because of all that > things we know... improve maintainability, improve team work, create > better code, etc. > But also, and most important, to keep our focus more and more on the > objectives of our project - the business rules, the usability, etc -, > than the coding itself. > I felt that with ARP, AMFPHP/openAMF, etc my life became simpler; the > code got better; the total implementation time was reduced; etc; but > there is one thing that from my very first project still remains the > same: relational databases and the coding of the data layer. > With Flash Remoting, we are exchanging objects between the server and > the client. But those object are created from relational databases. I > feel that the code for querying and manipulating the database, and to > convert relational data to objects is BORING. I would prefer to > store my > objects, retrieve my objects, change my objects, and store them again, > and that's it. > There are solutions based on relational databases that "simulate" data > persistence, like Hybernate. I know little about Hybernate (i've read > some things, but never used it), but it does not remove the need for a > relational database, it only wraps it so we can use it as if the > database was OO. (am i right?.....). Wouldn't it be better if the > database was really OO? > Is there a proven working solution of Object Oriented databases? > Do you > know something about this? If there was, would you use it? > > João Saleiro > > _______________________________________________ > osflash mailing list > [email protected] <mailto:[email protected]> > http://osflash.org/mailman/listinfo/osflash_osflash.org > > > ------------------------------------------------------------------------ > > _______________________________________________ > osflash mailing list > [email protected] > http://osflash.org/mailman/listinfo/osflash_osflash.org > _______________________________________________ osflash mailing list [email protected] http://osflash.org/mailman/listinfo/osflash_osflash.org
