[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
Re: [xml-dev] Re: [ubl-dev] Top 10 uses of XML in 2007
- From: Elliotte Harold <elharo@metalab.unc.edu>
- To: Stephen Green <stephen.green@bristol.gov.uk>
- Date: Mon, 19 Feb 2007 14:24:36 -0500
Stephen Green wrote:
> It's interesting then to see how Open Office implements
> its Base database. Internally it's just an SQL insert statement
> with some XML metadata. But of course it's zipped. At runtime
> it is made to look and act as a relational database but it is just
> text serialised and deserialised, it seems. Not sure what means
> in terms of the discussion. If there were nothing to it more than
> that it's a wonder all RDBMS databases don't do this. Surely
> they use binary for a reason other than to be fashionable.
>
I suspect OpenOffice's database is relatively low performance for small
data sets. Real heavyweight databases like Oracle and DB2 play all sorts
of tricks with file layouts to optimize queries and indexes, up to and
including writing their own file systems and perhaps even disk drivers.
Storing terabytes of data is not nearly as hard a problem as searching
terabytes of data. The performance needs of search completely swamp any
concerns with small size for your typical oracle or DB2 or Mark Logic
installation.
Of course not everyone has such large amounts of data, which is why we
have products like Cloudscape or Derby for smaller uses. I'm not
familiar with OpenOffice's database but it sounds like it's suitable for
small projects only, which is fine. I certainly wouldn't want to have to
install Oracle to manage my checkbook. :-)
--
Elliotte Rusty Harold elharo@metalab.unc.edu
Java I/O 2nd Edition Just Published!
http://www.cafeaulait.org/books/javaio2/
http://www.amazon.com/exec/obidos/ISBN=0596527500/ref=nosim/cafeaulaitA/
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]