Re: Patent

Kai Henningsen (kaih@khms.westfalen.de)
07 Nov 1999 09:10:00 +0200


htw6966@htw-dresden.de (Konrad Rosenbaum) wrote on 06.11.99 in <Pine.LNX.4.10.9911061715290.2310-100000@pinguin.local>:

> On Wed, 3 Nov 1999, Edward S. Marshall wrote:
>
> > On Wed, 3 Nov 1999, Joe wrote:
> > > Does the linux kernel use a database? (retorical q here) I think
> > > not .. thus this patent is irrelevant to what is being done in
> > > kernel code.
> >
> > What would you call a filesystem, if not a database? :-)
>
> Just a filesystem or Pre-Database :-)
>
> Databases have mandatory transaction management. Means if process A
> accesses data fields process B has to wait till A finished (except both do
> reads only).

So any database without multi-user support doesn't count as a database?
Whereas a simple index like modern libdb2 does?

Somehow I don't think so.

>On some better databases you may work around transaction
> management - as on better filesystems you may activate transaction
> management (by locking files).

My definition says that some better databases *have* transaction
management (which, incidentally, is *not* primarily about multi-user
capabilities, but about commit and rollback).

> Not that I ever wanted mandatory t.m.... but I remember an angry professor
> who didn't like the idea of filesystems compared with databases - don't
> run into the same trouble :-)

The filesystem *is* a database - a *specialized* database. A typical Unix
fs, for example, allows for exactly one type of key (a pathname) and has
only BLOBs for non-key data fields (the file data). OTOH, an AS/400 file
system is just a view onto a relational database. (Even the active
processes are in that database.)

MfG Kai

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/