Re: 'Theoretical' DB OS
From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Sat, 24 Mar 2007 19:09:06 GMT
Message-ID: <mHeNh.14650$PV3.151208_at_ursa-nb00s0.nbnet.nb.ca>
>>I'm working on a conceptual model (ie 'thinking about' / 'air code')
>>to implement a small operating system (extensible of course) that is
>>built completely out of a common set of database functions.
>>
>>The specific question I have now is an important one - how should I
>>store files?
>>
>>Heh. It's theory about what is the best physical implementation. ;)
>>
>>My first guess is that each attribute gets it's own file. Should I
>>make it where the user can select whether he wants it to be stored
>>fixed width or delimited?
Date: Sat, 24 Mar 2007 19:09:06 GMT
Message-ID: <mHeNh.14650$PV3.151208_at_ursa-nb00s0.nbnet.nb.ca>
David Portas wrote:
> On 24 Mar, 02:34, christianlo..._at_yahoo.com wrote: >
>>I'm working on a conceptual model (ie 'thinking about' / 'air code')
>>to implement a small operating system (extensible of course) that is
>>built completely out of a common set of database functions.
>>
>>The specific question I have now is an important one - how should I
>>store files?
>>
>>Heh. It's theory about what is the best physical implementation. ;)
>>
>>My first guess is that each attribute gets it's own file. Should I
>>make it where the user can select whether he wants it to be stored
>>fixed width or delimited?
> > Why do you want the database to be part of a file system at all? If > the database was not exposed to the file system then the security, > storage management and physical data integrity could be controlled > exclusively by the DBMS. Possibly that approach has some advantages.
David, I think he wants the dbms to be the file system. Received on Sat Mar 24 2007 - 20:09:06 CET