[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: Object Creation in ozone
On Fri, 11 Aug 2000, Knapp, Robert (CAP, CMC) wrote:
> >%_I just start checking this stuff out
> a few days ago. Don't worry, in a few
> months you'll flinch when you see my name. :)
Nah, test it out. We'll see. ;)
Falko
>
> The other 1/2 of the questions are still sort
> of formless thoughts regarding scalabilty, running
> on outdated hardware(486), differing platforms(issues
> with the macos threading) and the like.
>
> RobK
>
> -----Original Message-----
> From: Falko Braeutigam [mailto:falko@smb-tec.com]
> Sent: Friday, August 11, 2000 12:48 PM
> To: ozone-users@ozone-db.org
> Subject: RE: Object Creation in ozone
>
>
> On Fri, 11 Aug 2000, Knapp, Robert (CAP, CMC) wrote:
> > >%_Thanks! Your explanation answers about 1/2 the questions floating
> around
> > my head. :)
>
> Hmmm... 30 lines and it answers "1/2 the questions"... ?!?! :( :) Not that
> good, I think. What are the other 1/2 questions? ;)
>
>
> Falko
>
> > RobK
> >
> >
> > --------
> > Good questions! ;)
> >
> > IMO it is not possible to swap between different ODMG databases without
> > changing to much code no matter if they "fully ODMG compliant" or not. For
> > example Objectivity, the only commercial OODBMS that I have a license for,
> > claims to support ODMG. In fact, the database classes have to extend a
> > special
> > base class an all method have to call a fetch() method. Both is
> explicitely
> > ruled out by ODMG.
> >
> > And, the actual API is just one part of the ODMG spec. Another (very
> > important) part is the definition of the underlying object model. This
> part
> > is
> > way to unrestricted in the current version 3.0. For example, the ODMG
> object
> > model assumes that objects are activated on the client. (for historical
> > reasons, I think) but it's not strickly defined somewhere. ozone is based
> on
> > a
> > single instance architecture where objects live inside the server. This
> > difference is very important for the application architecture because it
> > influences performance and transactional behaviour but ODMG is completely
> > unaware of this.
> >
> > However, even if it does not allow to re-use 100% of the code, programming
> > against the ODMG interface makes it easier to port applications (even more
> > if
> > you keep in mind the above mentioned restrictions of ODMG) and to re-use
> the
> > knowledge of the programmers.
> >
> > The ozone ODMG interface is there and it works. It tries to be as ODMG
> > compliant
> > as possible. I don't know how it works for real world applications. So any
> > experiences are welcome. ;)
> >
> >
> > Falko
> > --
> > ______________________________________________________________________
> > Falko Braeutigam mailto:falko@smb-tec.com
> > SMB GmbH http://www.smb-tec.com
> >
>
> ----------------------------------------
> Content-Type: text/html; name="unnamed"
> Content-Transfer-Encoding: quoted-printable
> Content-Description:
> ----------------------------------------
>
> --
> ______________________________________________________________________
> Falko Braeutigam mailto:falko@smb-tec.com
> SMB GmbH http://www.smb-tec.com
>
----------------------------------------
Content-Type: text/html; name="unnamed"
Content-Transfer-Encoding: 7bit
Content-Description:
----------------------------------------
--
______________________________________________________________________
Falko Braeutigam mailto:falko@smb-tec.com
SMB GmbH http://www.smb-tec.com