Re: How would a relational operating system look like?

From: Brian Selzer <brian_at_selzer-software.com>
Date: Sat, 09 Jun 2007 03:41:48 GMT
Message-ID: <0kpai.13280$5j1.12278_at_newssvr21.news.prodigy.net>


"DBMS_Plumber" <paul_geoffrey_brown_at_yahoo.com> wrote in message news:1181330562.209696.115310_at_x35g2000prf.googlegroups.com...
> On Jun 4, 10:58 am, "Brian Selzer" <b..._at_selzer-software.com> wrote:
>> "DBMS_Plumber" <paul_geoffrey_br..._at_yahoo.com> wrote in message
>>
>> news:1180976599.740262.134510_at_k79g2000hse.googlegroups.com...> On Jun 3,
>> 7:57 pm, "Brian Selzer" <b..._at_selzer-software.com> wrote:
>> >> Are domains variable in a relational operating system?
>>
>> > Brian! Think bigger, man!
>>
>> > Not only are domains extensible, but even the set of operators in the
>> > algebra!
>>
>> I'm just wondering how each and every atomic value (file) in a relational
>> operating system can be located in one and only one place so that it can
>> be
>> shared by multiple users at the same time.
>
> This is gonna sound strange.
>
> There are no 'files'. A 'file' is an application level construct.
>

Strange indeed! A file is not an application level construct. It is simply a named allocation of persistent memory.

> If you wanted to create a 'domain' with the same properties as a
> 'file', then you'd need to also create a set of domain operators that
> did similar things to posix. But I would encourage people to avoid
> that notion.
>
>
Received on Sat Jun 09 2007 - 05:41:48 CEST

Original text of this message