sio-dev
[Top] [All Lists]

Re: [sio-dev] Definition of the SIO project

To: sio-dev@xxxxxxxxxxxxxxxx
From: Cameron Ross <cross@xxxxxxxxxxxx>
Date: Tue, 1 Jun 2010 14:24:40 -0400
Message-id: <AANLkTik7bBCpTFwLrwm8rzK7j9v5uxzOBgEYR_1DXzbj@xxxxxxxxxxxxxx>
Hi John,

Comments inline below...

On Tue, Jun 1, 2010 at 11:19 AM, John F. Sowa <sowa@xxxxxxxxxxx> wrote:
Cameron,

Those are very promising developments.  I had some questions:

CR> The Common Logic IDE is based on the Samian Platform being
 > developed by Kojeware.  The objective of the Samian Platform
 > is to provide tools and software infrastructure to support
 > the construction of semantic systems based on Common Logic.

I couldn't find any information about that.  The Kojeware site
says that they are developing the Samian Platform, but there
is no further info about it.  I found many hits for "Samian
Platform" on Google, but most of them are for dating pottery
in the Roman Empire.
 
Kojeware is a start-up and up till now our focus has been on baselining our platform architecture and constructing our "minimum viable product"... essentially, something we can get into the market quickly to solicit real feedback from the user community (for those interested in such things, we're roughly following the Customer Development Methodology promoted by Steve Blank: http://steveblank.com/category/customer-development-manifesto).  Anyway, the Kojeware website is in development and I expect it to launch July 1st.  A description of the Samain Platform will be included on the website.  "Samian" is a branding name we use for our platform and internal product development projects.  It has nothing to do with ancient pottery ;-).
 
Is that platform open source?  The plans for the OOR are to
make it and and its supporting software open source.  Eclipse,
for example, is free and open source, but the Eclipse license
permits any software, commercial or free, to run on it.
 
Since Eclipse is very widely used in the IT industry, it could
promote commercial applications of Common Logic.  But we also
want to support research and education, and a lightweight
platform for AI software would also be desirable.  But open
source would be a high priority for that as well.

The OOR is logically partitioned according to "technology" and "content".  Technology being the software infrastructure supporting the OOR and content being the actual ontologies.  These are clearly two very different things and I believe the term "open" needs to be defined differently for each.  One definition for OOR openness w.r.t. ontology content is that the content may used and disseminated unencumbered.  One definition for OOR openness w.r.t. the supporting software infrastructure is open source.  However, I believe that to realize this vision of OOR and SIO we will have to establish a marketplace that embraces both open source and proprietary software models.  The Eclipse ecosystem has done a good job with this and we can look to it for guide (we can also look to Eclipse for baseline open source infrastructure).  Therefore, IMHO we should define openness w.r.t. OOR support infrastructure as an "open system" environment.  Note that the definition I'm using for "open system" may be found here: http://www.acq.osd.mil/osjtf/whatisos.html.  This implies that we should focus on defining interfaces etc. and let the market develop tools and infrastructure, according to these interfaces, as a loosely coupled collection of both open and closed source modules.  Both the FMF and the Samian platforms promote this concept of an open ended collection of loosely coupled components. 

We shouldn't view the current implementation of the BioPortal software as a panacea for OOR.  It is my understanding that the BioPortal will not support Common Logic ontologies at its very core.  This means that repositories such as COLORE will need to look elsewhere.  This is not to say that an instance of BioPortal shouldn't be able to integrate with an instance of some CL supporting repository hosting the COLORE content.  Different implementations will need to interoperate. 

The Samian Platform is not open source.  However, Samian bundles are intended to be deployed within OSGi which can support an open systems environment.  Samian services will interoperate via an open communication protocol such as FIPA (or maybe the FMF messaging format if its appropriate).  Messages passed between Samian services (agents) are encoded according in Common Logic and the ontology describing the content of these messages is passed with the message and accessible by all agents in the system.  Modules from other platforms such as FMF, or whatever, will be able to intetroperate as long as we get the "interfaces" right.
 
CR> However,  I would also consider the FMF message format
 > outlined in http://www.jfsowa.com/pubs/paradigm.pdf if you
 > could make the details of the format available.

We've implemented a lot more since the original paper from 2002
( http://www.jfsowa.com/pubs/arch.htm ).  We want to make the
spec freely available, but it's still undergoing some changes.
I'll discuss the issues with my colleagues to see how soon we
can make anything available.

It would be great if I could be the guinea pig for a public release of the FMF message format.

Cameron.

_________________________________________________________________ 
Msg Archives: http://ontolog.cim3.net/forum/sio-dev/   
Join Community: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J 
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/sio-dev/  
Unsubscribe: mailto:sio-dev-leave@xxxxxxxxxxxxxxxx 
Community Shared Files: http://ontolog.cim3.net/file/work/SIO/ 
Community Wiki: 
http://ontolog.cim3.net/cgi-bin/wiki.pl?SharingIntegratingOntologies     (01)
<Prev in Thread] Current Thread [Next in Thread>