RE: oBIX Discovery Service

Title: Message

Actually the way I look it is that Obix is a "dynamic schema 
language".  The most important aspect of Obix is to normalize a few basic 
types that everyone can agree upon, and then provide the mechanisms to 
"discover" how to interrogate vendor specific data structures using those data 
types.  That is where we have been going with the "point 
service", although maybe that isn't the best term.  It is also the impetus 
for the inheritance by prototype designs that Dave Robins and I proposed early 
in the cycle.

  
-----Original Message-----
From: Considine, Toby (Facilities 
  Technology Office) [mailto:]
Sent: Friday, 
  July 16, 2004 5:38 PM
To: 
Cc: 
  Brian Frank; Anno Scholten; ; Peter Manolescue; Aaron 
  Hansen
Subject: RE: oBIX Discovery Service

  
As we name these scenarios, we are creating a new 
  verb known the the oBIX language - in the same way that each object and method 
  known in an OO environment creates a new verb.  This, is the essense of 
  the Domain Specific Language (DSL) I was talking about last week.  You 
  may recall that I suggested that a Web Service done right might be 
  indistinguishable from a DSL.

  
 

  
http://blogs.msdn.com/keith_short/archive/2004/06/18/159710.aspx

  
 

  
While I am not sure myself what I mean by that, it do 
  feel that it is a useful lense with which to look at oBIX, or the development 
  of any language.

  
 

  
tc

  

  

  
  From: Aaron Hansen 
  [mailto:] 
Sent: Friday, July 16, 2004 1:59 
  PM
To: Considine, Toby (Facilities Technology Office); Peter 
  Manolescue
Cc: Brian Frank; Anno Scholten; 
  
Subject: RE: oBIX Discovery 
  Service

  

  
I do 
  wonder if this belongs in the point service.  What we are talking about 
  (scenes/shed levels) are nothing more than a group of set points.  It 
  would certainly be a distinguishing feature between oBIX and all other XML 
  point services.

  
 

  
BTW 
  - We should be having these discussions on the mailing 
  lists...

  
    
-----Original Message-----
From: Considine, Toby 
    (Facilities Technology Office) 
    [mailto:]
Sent: Thursday, July 15, 2004 
    9:43 AM
To: Aaron Hansen; Peter Manolescue
Cc: Brian 
    Frank; Anno Scholten; 
Subject: RE: oBIX Discovery 
    Service

    
I think this, and what doug spoke about, are the 
    *reason* for oBIX.  We have control systems, lots of control systems, 
    now, and they generally do what they are supposed to.  What we do not 
    have is standards-based a way to orchestrate them to meet the needs of the 
    enterprise.

    
 

    
There are two purposes to oBIX

    
- Revealing properties if the networked control-system 
    object to the enterprise to enable information-based decision 
    making to the enterprise

    
- Proviiding a platform upon which hundreds of 
    enterprise-based applications can be written using the interface to the 
    control system.

    
 

    
This mioght be saying that Doug and Premisses might be 
    ones for the Roadmap.

    
 

    
tc

    

    
    From: Aaron Hansen 
    [mailto:] 
Sent: Thursday, July 15, 2004 9:20 
    AM
To: Peter Manolescue
Cc: Brian Frank; Anno Scholten; 
    Considine, Toby (Facilities Technology Office); 
    
Subject: RE: oBIX Discovery 
    Service

    

    

    
My 
    initial thought is that this is not a function of the discovery 
    service.  However, the discovery service should expose the node(s) 
    and service binding(s) so that this can be 
    done.

    
 

    
This sounds 
    similiar to what Doug Ransom was talking about in the last cc.  Would this be something for the Power 
    Management sig?  
    

    
 

    
Best,

    
Aaron

    
 

    
      
-----Original Message-----
From: Peter Manolescue 
      [mailto:]
Sent: Thursday, July 15, 2004 
      3:48 AM
To: Aaron Hansen
Cc: Brian Frank; 'Anno 
      Scholten'; 'Considine, Toby (Facilities Technology Office)'; 
      
Subject: oBIX Discovery 
      Service

      
Aaron,

      
 

      
I got to 
      thinking about the oBIX discovery service.  As I understand it, a 
      user (human being or application) can go to the facility, invoke the 
      discovery service and get back a tree structure describing its 
      phyiscal configuration and the addressable points (sensors, actuators, 
      controllers etc) with their associated types.  It can then (if it has 
      permission) extract data from those points and ultimately set certain 
      parameters.  

      
 

      
It struck me 
      that Premise Systems (http://www.premisesystems.com/ - 
      recently bought by Motorola) has developed a concept that matches this 
      idea very closely.  Although their primary target is residential, I 
      know that their technology is being successfully used in commercial 
      applications.  A very powerful 'tool' they use is the idea of a 
      'scene'.  This means that a pre-defined configuration of a building 
      can be demanded.  This addresses one of the needs that Toby 
      expressed: if you want to put the sports hall into 'event-mode' at 9pm for 
      an unforseen basketball match then a simple (and obviously secure) command 
      will set the heating, lighting, access, public address etc to the 
      required set-points.

      
 

      
You may want 
      to take a look at their software (free to download).  I'm thinking 
      that it might also be a good idea to invite them to join our merry 
      band.  Not only would we get Motorola into the fold but Dan Quigley, 
      the president, is a well-connected ex-Microsoft guy who could bring quite 
      a lot to the party.

      
 

      
Best 
      regards,

      
Peter Manolescue 
      

      
securityXML Ltd 
      
Bleuckeveldlaan 
      12 
3080 
      Tervuren 
Belgium 

      
Tel +32 2 768 
      1784 
Mobile +32 
      485 574878 
Fax 
      +32 2 306 0969 
www.securityxml.com