On 08/01/2007, at 12:33 AM, Ruslan Zasukhin wrote:
From: Andy Dent <[EMAIL PROTECTED]>
Date: Sun, 7 Jan 2007 19:39:51 +0800
My first use of c-tree Plus in shared-file mode was with a IIci as
server, using LocalTalk to Classics. Even with that hardware, it had
sub-second lookup times.
For what size of dbs 0.5 - 5MB?
and what kinds of queries?
:-)
Obviously that was an extreme low-end example, back in about 1992. As
I recall, those were string or number key lookups and a database of
some MB.
However, you'd be surprised at just how efficient such engines can be
- I wouldn't waste your time getting into arguments with such engine
users - I know of databases in the 100's of MB used in shared file
situations in the education and mining sectors. Sorry, can't point to
any examples other than KIDMAP from http://www.mercator.com.au/ as
the FoxPro stories were anecdotes from friends here in WA.
search by indexes do WHO? Right, the CPU of client computer.
This means that N pages of index must travel by network to client
computer,
Yes, but when an engine has been specifically written for such a task
and optimised through many years, the index storage can be optimised
for such use.
So my point is that EVEN IF a DB engine is _optimized_ for shared
access,
this technique still a lots behind to client/server technique.
I never argued that it was faster than client/server, just that
people radically underestimate just how fast it can be :-)
Andy Dent BSc MACS http://www.oofile.com.au/
OOFILE - Database, Reports, Graphs, GUI for c++ on multiple platforms
REALbasic, C++, Python, Mac and Windows development and porting
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
