[lug] Doing your own transactions

[Michael J. Pedersen]

On Thu, Jan 18, 2001 at 11:24:02PM -0500, John Karns wrote:
> Have any thoughts on a comparison to Postgres?  

Absolutely none. I've only marginally looked at Postgres, mainly because I've
not had a need for a database of that complexity in the past few years. MySQL
and flat files have been more than enough (sorry, my apps of later have only
had a few hundred records, not the thousands of records with ACID requirements
which would cause me to use interbase, postgresql, or any of the other,
larger, rdbmses).

> I understand that ibase evolved from a versioning system engine, and thus
> records are somehow stored differentially, if that's the right term.  
> IOW, for two records with similar content, rather than duplicating data,
> differences are stored.  This would of course save on disk space and tend
> to shrink the size of the DB which would tend to improve efficiency
> concerning disk access.  But it would also seem that there would be some
> penalty to reconstruct a record from it's component parts.  I'm curious
> about what net effect this would have on performance: in context of a
> large DB with a significant amount of patterned data vs a large DB with
> more randomized data.

I've not heard of that being done before. Of course, that means precisely
nothing, as I've not dug through the interbase files themselves to see what
they looked like. Still, you're right in that those would make for some
interesting questions.

-- 
Michael J. Pedersen
My GnuPG KeyID: 4E724A60        My Public Key Available At: wwwkeys.pgp.net
My GnuPG Key Fingerprint: C31C 7E90 5992 9E5E 9A02 233D D8DD 985E 4E72 4A60
GnuPG available at http://www.gnupg.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 240 bytes
Desc: not available
URL: <http://lists.lug.boulder.co.us/pipermail/lug/attachments/20010118/f84d3cd3/attachment.pgp>