XML size

[Matthew Vanecek]

--=-Ba7kZ3wRv5Cb1b77DrqN
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Thu, 2002-04-04 at 01:03, Cornel DIACONU wrote:

> I've NEVER (repet: NEVER) came to such a case of
> corruption of my DB file with either of the DBMSs I've
> worked so far (Oracle, Informix, DB2, PostgreSQL,=20
> MySQL, and you name anything in this world further).

Point of contention here--my company uses DB2.  While I agree mostly
with what you say, database corruption is a very common occurrence.  We
are, of course, protected by UPSes (where the local manager hasn't
requisitioned it for his/her personal use :( ), but are of course not
immune to the occasional 3-finger salute or hard power down by the
well-meaning-but-totally-clueless LAN admin.  This does cause the
database to get corrupt more often than we like (not always, mind
you--just often).  Other factors also cause the database to become
corrupt.  Fortunately, daily backups and real-time replication to our
mainframe make recovery much less painful.

The same problem can, of course, occur with ASCII text files--especially
with, oh, ext2 or FAT32 file systems (I've lost whole files, before I
got a UPS and switched to a journaled FS).

In any case, data corruption is not restricted to database files--it can
affect text files just as easily.

Regarding methods of access--well, it's what you're used to that
counts.  Me, I work in databases all day.  SQL is like a second language
for me--I even speak it sometimes (to the chagrin of co-workers!! =3DP).=20
However, I also spend quite a bit of time in text files--logs and such,
so I am well versed with various search methods (grep, find on OS/2 and
Windows, less, more, sed, eyeballing....).  For my time, it's much
easier and quicker to do a "select fields from table where key=3Dvalue",
than it is to do "/My[p \t]+Ac[c]*ount".  Especially since all I have to
type is "pgsql update gnucash_table set account =3D "My Account" where
account =3D "My Acount".

Anyway, I like the back-end/front-end separation.  If the front-end has
a standard method for calling the back-end, then you can just create new
plug-ins for the back-end, which go to databases, db3 files, XML files,
whatever.  But isn't that what the gnucash developers are trying to do
(or have done) anyhow?  In which case, this whole discussion is moot...
--=20
Matthew Vanecek
perl -e 'print
$i=3Dpack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);'
***************************************************************************=
*****
For 93 million miles, there is nothing between the sun and my shadow
except me.
I'm always getting in the way of something...

--=-Ba7kZ3wRv5Cb1b77DrqN
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: This is a digitally signed message part

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQA8ryPxOMmiB1jXEBsRAthkAJ4wxyoWK4Z/HihOowOCjzxLnS9fYwCfQxMF
luJgEmeLBLbRIhcKPa32LUI=
=GFZb
-----END PGP SIGNATURE-----

--=-Ba7kZ3wRv5Cb1b77DrqN--