discuss-gnustep
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re[2]: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for qualit


From: Helge Hess
Subject: Re: Re[2]: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control)
Date: Thu, 23 Oct 2003 22:46:25 +0200

On 23.10.2003, at 10:13, Manuel Guesdon wrote:
What do you mean by it "doesn't scale" ? Is is that problems came with
records counts, tables count or database complexity ? Do you have example
some examples ?

Well, the OmniGroup wo-dev list is *full* of examples. This is really too much to write down in a mail - but I would be happy to discuss the issue in person at next FOSDEM ;-)

Well, an issue in short:
EOF puts an object abstraction on top of a RDBMS and that abstraction basically disables almost any advantages a RDBMS is providing you - eg very much apparent with EOF doing its own constraint processing. Basically all reasons why someone would use a RDBMS database are just not valid in a "pure object" context like targetted by EOF. Partial selects are another very good example.

I'm not attempting to start a discussion on whether EOF is good or not. I *know* that this is very hard, having been a "EOF Zealot" ;-) myself for several years. If you have worked around EOF for the 101st time and jumped out of the window for the huge memory requirements imposed by its very nature of keeping everything as objects, you'll know why this is a concept sounding good in theory yet being impractical.

regards,
  Helge
--
OpenGroupware.org
http://www.opengroupware.org/





reply via email to

[Prev in Thread] Current Thread [Next in Thread]