[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

Re: [ba-ohs-talk] Concept: Typed Versioning

On Thu, 2002-03-21 at 17:08, Eric Armstrong wrote:
> Sandy Klausner wrote:
> > So now two mechanisms need development. One to represent the dynamic linking
> > and separate meta logic for determining the syntax/semantic update behavior.
> > Seems that the dynamic linking mechanism should be generic, while provisions
> > for user control to declare domain specific update behavior would drive the
> > generic linking.
> Thinking in terms of the great picture that Lee gave me at dinner one night,
> a link to a node (in Nodal) always points to the top of a chain. Below that
> the chain is the progression of revisions, where the top most revision in the
> chain is the most recent.
> I'm pretty sure I agree with Eugene that link typing must be an integral part
> of the system. How that gets added to Nodal, I'm not sure, but let's say it
> is.    (01)

It's still an open question as to what a "link type" really is.  If we
stick to the data-knowledge-context modularization that I've argued for,
then "semantic" links could be held in the "data" system, but would need
some sort of external maintenance.    (02)

Assuming the addressability issue is somehow handled by having virtually
every aspect of addressing into the node/version tree be translatable to
a URI, then I don't think it is too difficult to get there.  As far as 
the notification issue goes, that it is a pretty fundamental aspect of
what has come to be called "dynamic databases".  I certainly see NODAL
as being one of those.  At what granularity (esp. in time) is a hard
design choice.    (03)

> I see link types as something the user controls. (I think Eugene and I
> different concepts here. What I'm presenting now is my view.) The user
> sets a link type as one of (for example) response, counter-argument,
> inline inclusion, simple reference, like that. The way the system displays
> linked material is primarily determined by those types, and also controlled
> by the user.    (04)

These are clearly "semantic" categories and would have to reference some
sort of external ontological structure...  There is a possibility of
building these on top of more structural link types: dated vs. undated, 
direct vs. relative, etc.    (05)

Lee Iverson                    
leei@ece.ubc.ca                Dept of ECE, 2356 Main Mall
http://www.ai.sri.com/~leei/   Vancouver BC Canada V6T 1Z4
Office: (604) 822-3381    (06)