[unios] Re: Language Oriented System

[Srikant Sharma (Chiku)]

From: "Srikant Sharma (Chiku)" <srikants@wipinfo.soft.net>

On Fri, 18 Dec 1998, OJ Hickman wrote:

> From: OJ Hickman <hickman1@peak.org>
> 
> Language Oriented System (LOS) design.
> By Omer James Hickman - December 18, 1998
> 
> The current discussion centers around Object Oriented
> design, but I think that OO is an inadequate model for an
> overall computing system.
> 
> OO is still based on procedural programming, all it does is
> bind a set of procedures to a data structure. OOP means
> inanimate objects. The 'entities' that make up a computing
> system are NOT inanimate. They are 'alive' in they process
> in parallel and communicate in messages rather then
> invoking procedures in each other in a linear fashion.

why do I feel that this is just another twisted nomenclature 
for microkernel based design ?

Also can you be more specific on .. what you mean by
"The 'entities' that make up a computing system are NOT inanimate.
 They are 'alive" 

This could be a matter of perspective.  I feel a normal svr4 unix
to be 'very much alive' .... The preocesses are dynamic. These talk with 
each other using ipc. There is a parent ... there is a child ...
there is a group ....
for that matter ... a thread spawns a thread ...  .. if you go and
see the kernel implementation of solaris .. you will see all the threads
being referred as siblings ....
Is not this stuff alive ?

> 
> LOS is more like the way we (human beings) work together. We
> send messages, sentences, to each other. In its
> most basic form a sentence has a subject (noun) and an
> action (verb) this basic form can be expanded to convey very
> complex interactions. In the context of computing, of
> course, the syntax of such sentences would be relatively
> restrictive.

This is the basic idea of micro kernel architecture .....
The (micro)Kernel  is in fact a switch (or a kind of hub) where
all messaging transactions take place. I feel that the basic concept 
does not become a radically different concept if we give some (fancy)
nomenclature to it.

> 
> In the context of this paper I apply these conventions:
> 
> * 'Entity' a basic unit of the system - in practical terms a
> thread of execution.
> 
> * 'Vocabulary' - a definition of the 'words' of a language -
> the equivalent of a class in OOP languages.
> 
> * 'Jargon' - a specialized superset of a Vocabulary - like a
> derived class in OOP.
> 
> * Messages are assumed to be a pointer to client memory
> space.
> 
> 
> A LOS system would likely have a basic Vocabulary something
> like this:
> ________________________________
> RWvocab:   - Vocabulary name
> 
> OPEN [device]
> CLOSE [device]
> READ [buffer...]
> WRITE [buffer...]
> ________________________________
> 
> Service Entities then can extend the basic Vocabulary to
> create a Jargon to fit its specific needs. Maybe even adding
> logic and other macro capability.
> 
> For example:
> ________________________________
> FileSystemJargon:
> EXTENDS RWvocab
> 
> LOGON [filesystem]
> CD [directory name]
> IF [condition statment]
> ELSE
> RETURN [ret message]
> EXISTS [filename]
> FOPEN [file]
> FCLOSE [file]
> FSEEK [file_position]
> _________________________________
> 
> It should be possible to issue a complex command to the file
> system such as:
> 
> "LOGON DriveA CD somedir IF EXISTS file.ext FOPEN file.ext
> RETURN It_worked ELSE RETURN Error_file_not_found"
> 
> RETURN simply copies the error value to the
> beginning of the message buffer.
> 
> If it works then:
> 
> "READ ...data_buffer..."
> "WRITE ...data_buffer..."
> 
> The data_buffer would be space within the message.
> 
> Naturally the commands in such a language would be in binary
> format not a text message. It would almost resemble a
> virtual machine code.
> 
> __________________________________
> I know this is only a brief description of my LOS concept,
> but it seems to be the most natural and flexible model
> for interaction within a computing system.
> 
> _______________________________________________
> "Imagination is more important then knowledge."
> - Albert Einstein
> 
> Omer James Hickman - hickman1@peak.org - ojh@hotmail.com
> http://members.tripod.com/~OJ_Hickman
> 
> ------------------------------------------------------------------------
> To unsubscribe from this mailing list, or to change your subscription
> to digest, go to the ONElist web site, at http://www.onelist.com and
> select the User Center link from the menu bar on the left.
> ------------------------------------------------------------------------
> UniOS Group
> http://members.xoom.com/unios
> 


------------------------------------------------------------------------
To unsubscribe from this mailing list, or to change your subscription
to digest, go to the ONElist web site, at http://www.onelist.com and
select the User Center link from the menu bar on the left.
------------------------------------------------------------------------
UniOS Group
http://members.xoom.com/unios