FoRK Classic: Was oSpace an April 1 joke?

I Find Karma (adam@milliways.cs.caltech.edu)
Mon, 25 Aug 1997 04:51:31 -0700 (PDT)


Rohit, I just now realized the date timestamp on this...
Was oSpace just a joke??? :) Adam

> From khare@xent.caltech.edu Sat Apr 1 10:44:29 1995
> Subject: Re: Multimedia Contacts
>
> We three could build oSpace -- Eve's object location stuff, your
> parallelizing corba dispatch w/dataflow, and I'd work on the object
> databases and constraint logic. We'd then be able to use it to implement
> Distributed Classroom, Distributed Documents, and Distributed
> Media. It'd be mobile-adaptive, too, by recognizing in the kernel
> mutable bandwith connections.
>
> The other half of the "transparent" magic:
> Caching object-expressions for nonmodified uses subsets.
>
> I.e. if I ask [pasteboard compatibleTypeTo:"foo"] should just be a hit
> and return YES or NO as long as pb isn't changed in the subset of
> variables that "compatibleTypeTo" and its sub-calls access.
>
> We shoot for Oleg Kiselyov's dream of an ultimate OS. How the MacOS can
> be further unifomalized: Deep down the OS is nothing but a manager of
> many databases: file system, process table, routing tables, list of
> known AppleShare servers, revision control system (projector) data,
> Think C projects. It would be much more efficient to have a single,
> really good distributed database manager implemented as a universal
> *core* program, than a multitude of "custom" database managers: a
> database system *instead* of the file system. Since everything is a
> database, a link between two records representing files is no different
> for the database manager than a link between a record in the Users
> table, a record in the "Processes" table, a couple of records in the
> "Files" table and a "Print jobs" table. No need for multiple (user,
> process, print job) IDs and messing with them, everything is
> uniform. Many-to-many links are possible. Hypertext documents and the
> process tree are the same thing deep down, and they should be the same
> for the OS.
>
> Call 4 details.

Maybe I should release your "cell" epiphany, too? :)

----
adam@cs.caltech.edu

Get your facts first, and then you can distort 'em as you please.
-- Mark Twain