[Home] [Groups] - Message: [Prev in Group] [Next in Group]
1955: Re: [MUD-Dev] **sigh** OOP wars and defining RDBMS
[Full Header] [Plain Text]
From: Ola Fosheim Grøstad <olag@ifi.uio.no>
Newsgroups: nu.kanga.list.mud-dev
Date: Tue, 2 Sep 1997 12:27:55 +0200 (MET DST)
References: [1]
Organization: Kanga.Nu
>>win. You may of course take another approach and throw distribution at
>>the problem and write off the hardware costs by an assumed gain in
>>development time (although efficient largescale distribution isn't all
>>that easy?).
>
>We've solved it. Efficiency is nto the issue for us, macheins are cheap.
>Scalability was the issue and our entire system architecture is a zero
>bottleneck almsot infinitly scalabale system.
Time will show then. I'm still not going to do convolutions in an
interpreter... Actually I would agree for any reasonable compiled
language, where you can control caching. I'm not talking assembler
efficiency. The problem with this thread is that it started with a
smallscale startup project using persistence libraries and gradually
got into fullfledged commercial systems. I'm not saying that you
cannot use hybrid systems for instance... In a startup you don't
want to do that, so that was the basic assumption, for a commercial
system things are different.
Ola.