11 Nov 2008

See also: IRC log


jar, alanr, dbooth
Jonathan Rees


<jar> scribe: nobody

<jar> scribenick: nobody

RFC 2616

<alanr> happening?

<jar> maybe. waiting for harry. everything hinges on harry

<jar> davidb is here

<alanr> ok. Not sure how much I can be there as I'm heading to dr. but we'll see

<jar> silent for now. we'll give ia few more minutes. david didn't want to just stare at rfc2616 as i suggested, which is reasonable...

<alanr> :)

<alanr> what are you looking for?

<jar> but my idea was that if we're not doing any work outside of calls, we should at least do work during the time reserved for the call

<jar> alanr, david and I are both looking at rfc2616

<alanr> to what end?

<jar> looking for classes and relations. remember the idea was to have on hand, for reference at least, an ontology (or rdfs) that captured important aspects of rfc2616

<dbooth> dbooth: Actually, it isn't that i don't want to stare at 2616 -- i think that is a good suggestion -- my suggestion of canceling today was made before I saw jar's agenda suggestion, when i was feeling guilty of not personally making any progress since the last call. :(

<jar> with an eye to superclasses and superproperties, etc

<jar> or else a refutation of the idea that you can derive an ontology from rfc2616

<jar> i think you should be able to do an rdf schema, but rfc2616 is not meant as, and shouldn't be taken to direct, any ontology

<jar> because it's a software thing, not a philosophical thing

<alanr> software things are in the realm of the IAO, so I'm not sure that would be justification.

<alanr> and rdf schema has exactly the same sort of semantics as OWL, just a different selection wrt to expressivity

<jar> no, by "software thing" I mean a formalism drawn without any interest in truth or real world denotation, but only with regard to effect. That is, things are not what they seem

<jar> an rfc2616 entity is an information artifact - that's fine. but what does it mean? that's outside the scope of rfc2616.

<jar> or a better example might be content negotiation

<alanr> content negotiation is a process

<jar> content negotiation is a mechanism, the syntactic form of which is given. but what governs its correctness? nothing really (in rfc2616)

<alanr> information artifacts are the results of decisions by sentients

<alanr> if cn can't be represented in IAO it will have failed

<jar> so the statement "x is a representation of y" is not falsifiable under the terms of rfc2616. it's up to the server, and there's no way the server can be wrong - what it says goes.

<alanr> not that I'm volunteering at the moment

<jar> unless you look at awww, which is not rfc2616.

<jar> (that's "falsifiable" in the Popper sense, David)

<alanr> that's a bug

<jar> i'm not sure. the protocol is all syntax, and almost no semantics. what semantics there is is all in the caching arena.

<jar> a book on english grammar isn't obligated to tell you what sentences mean

<jar> so i don't think it's a bug. in any case it doesn't matter whether it's a bug or not

<dbooth> What is "IAO"?

<jar> information artifact ontology

<alanr> information artifact ontology

<dbooth> thanks

<jar> http://groups.google.com/group/information-ontology

<dbooth> I'm not convinced that non-falsifiability is a bug.

<dbooth> "Popper sense"?

<jar> right that's what I was saying

<jar> hang on i'll get you a url

<jar> http://en.wikipedia.org/wiki/Falsifiability (surprise)

<alanr> in any case all that means is either a) that in the description of the process, there is some process where the output is always labeled a "representation". or b) "representation" is effectively a synonym for whatever the class of things that happens to allowable outputs of the process.

<jar> actually i think 'refutable' might work better for audiences who don't know about Popper. what do you think?

<dbooth> so IAO is for things like journal articles?

<alanr> yes

<dbooth> good

<dbooth> sounds quite useful

<alanr> http://neurocommons.org/page/Information_Artifact_Ontology

<jar> although the current focus is on what happens in a lab - measurements and so on

<alanr> http://groups.google.com/group/information-ontology

<alanr> current, yes. But we also have narrative object in there.

<jar> well i think we need to distinguish carefully between an rfc2616-representation and an awww-representation, because they are different classes (or roles).

<dbooth> Me browsing through http://neurocommons.org/page/Information_Artifact_Ontology -- looks like nice work!

<jar> representation = "entity subject to content negotiation"... this is so vague as to be useless.

<jar> the work has barely begun. it is very difficult.

<alanr> alright - then representation is used as a name for the class of things that are part of that process.

<alanr> that class is presumable otherwise constrained.

<dbooth> collecting requirements, jar?

<jar> yes. this stance (what alanr just said) is what we need to capture (assuming we want to take on the task of accounting for rfc2616)

<dbooth> jar, i agree with "well i think we need to distinguish carefully between an rfc2616-representation and an awww-representation, because they are different classes (or roles)."

<jar> requirements... ok, not sure where to put it

<jar> new wiki page i guess

<dbooth> i don't have a requirement to suggest, i was just asking where you are in the work.

<dbooth> ... the IAO work, that is.

<jar> oh.

<jar> i think everything there is to know is on that wiki page and in the archives of the google group

<dbooth> ok

<jar> i'm not taking any management role. not sure who is, maybe alan

<dbooth> well on first glance it looks like a great start. Probably just needs white knight to swoop in and do a lot of work. ;)

<alanr> Barry and I

<jar> the effort is heavily BFO-oriented so may not be to everyone's taste (I often wonder if BFO is worth it)

<dbooth> BFO?

<jar> google it

<dbooth> Basic Formal Ontology

<jar> Trying to capture some of what we've talked about today in this wiki page:

<jar> http://esw.w3.org/topic/AwwswRfc2616

<jar> refresh

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/11/11 16:03:51 $