UoM_Ontology_Standard draft deliverable preview session - Thu 22-Oct-2009    (22MH)

This is a preparation session for our upcoming 30-October-2009 workshop. We expect to preview the draft deliverables that the various champions have put together so far, surface specific issues that may be hampering the drafting work, and do some preliminary alignment, in preparation for a more in-depth discussion during the 30-Oct-2009 full-day workshop.    (22MI)

Archives    (22MR)

Conference Call Details    (22MX)

Resources    (22NY)

Attendees    (22OG)

Agenda & Proceedings    (22OW)

Agenda    (22OX)

1. Opening by session Chair -- FrankOlken / EdBarkmeyer    (22OY)

2. Briefings from deliverable champions -- EdBarkmeyer, RobRaskin / ChipMasters, PatHayes, HowardMason, JoelBender / PeterYim    (22OZ)

3. Q & A and Open Discussion (All) -- please refer to process above    (22P0)

4. Action Items and Conclusion -- FrankOlken / EdBarkmeyer    (22P1)

Proceedings    (22P2)

Please refer to the archives above    (22P3)

IM Chat Transcript captured during the session:    (22P4)

	FrankOlken: I expect we will use the same teleconferencing facilities as 					
	usual on Oct. 30. Our audio tech will call into the teleconference    (22TS)
	PeterYim: great, ... thanks, Frank    (22TT)
	PeterYim: Howard: note: maintenance agency ... registration authority    (22TU)
	FrankOlken: I think we want a maintenance agency (which has 					
	discretionary authority to approve new derived units).    (22TV)
	FrankOlken: I think the maintenance agency should be something like 					
	NIST.    (22TW)
	PatHayes: 1. Should we have a quantity model? Seems to be a majority 					
	agreeing that this is needed. 2. Explicit named units and Q. kinds? We 					
	can't do them all, how many will we do and what do we do about the 					
	others? Eg how do we relate to the UCM list? [Peter: maintenance issue 					
	arises when normative standard is formal. Who adds items to 					
	namespace/base?][Ed: yes, needs an authority or tool which can handle 					
	this. Not impossible and needs to be done. Certainly a scope 					
	issue.][Frank: exist 2 things. Extensions which are derived units, 					
	should be automatic. But also, new named units like 'watts'.] <<much 					
	discussion about mechanisms for administration and registration, etc.. 					
	ISO and OASIS both have various mechanisms...>>    (22TX)
	PatHayes: Should have said, I was scribing Ed's speech above.    (22TY)
	PeterYim: thanks, Pat    (22TZ)
	PatHayes: Ed: quantity kinds more promiscuous than units. Do we needs 					
	*kinds* of quantity kinds? Need to choose a level to stop publishing, 					
	this is unbounded in extent.    (22U0)
	JoelBender: Is 'scale' a 'derived unit'?"    (22U1)
	PatHayes: Ed: scales. Email discussions went nowhere. Do we have scales 					
	as part of the ontology? Unit conversions also an issue (4 so far). 					
	Other issues further out: uncertainty, tolerance. But these are v. 					
	complicated to do 'propoerly'. Do we need to handle these at all, if 					
	only crudely? Chips (eg) does, but its about measurement. Do we want to 					
	go this way?    (22U2)
	PatHayes: Joel: I don't think so.    (22U3)
	PatHayes: Ed: measurement is big area, need to bring other organizations 					
	in if we go there, eg UK NPL. Probably can't do it all, but also not 					
	completely omit.    (22U4)
	PatHayes: Howard: need to consider target market for our ontology, to 					
	focus discussion. XXphilosophy..    (22U5)
	PatHayes: Ed: priorities. We need to have a time limit, so some things 					
	are prunable. Final point: microtheories: not all extensions are 					
	mutually compatible, possibly. Up to us to provde guidelines for this 					
	situation.    (22U6)
	PeterYim: EdBarkmeyer: OMG QUDV    (22U7)
	SteveRay: Sorry, had to run to another meeting...    (22U8)
	PeterYim: Ed will make an attempt to invite the QUDV leads,					
	Roger Burkhart (John Deere) & NicolasRouquette JPL) to join us on Oct-30    (22U9)
	EdBarkmeyer: Scope: assume quantity, quantity kind, SI named units, 					
	scales, conversion are in, measurement and uncertainty, specification 					
	and tolerance are less clear, it depends on our target audience and use, 					
	available expertise and time.    (22UA)
	EdBarkmeyer: Big issue is extension: how to deal with the infinite list 					
	of quantity kinds and the very large list of derived units.    (22UB)
	EdBarkmeyer: Pat: UML may be all that many will understand, OWL will be 					
	definitive for the Sem Web folk, CLIF is really the hard form for people 					
	committed to axiomatic definition and FOL reasoners.    (22UC)
	EdBarkmeyer: Pat: there are other reasons to use CLIF -- it allows 					
	simpler consistent formulations of some complex notions    (22UD)
	EdBarkmeyer: Frank and Pat discussed concerns about the arithmetic, use 					
	of functions like sine or square root. Pat: we want to stay away from 					
	floating point and arithmetic error.    (22UE)
	EdBarkmeyer: The OWL folk have defined the datatype to be xsd:double, 					
	i.e., IEEE 754 numbers.    (22UF)
	EdBarkmeyer: Pat: we want to assume mathematically exact numbers and 					
	avoid computational issues. Yes, actual uses may have some problems if 					
	the values and computations exercise the fringes of the computational 					
	behaviors. (Ed: the common problem is just that decimal fractions (the 					
	heart of metric) don't have exact binary renditions.)    (22UG)
	PeterYim: Howard: sending our draft TC charter to Jamie and inviting him 					
	to do a 5 min. OASIS-101    (22UH)
	PeterYim: @Howard - please consider the proposal of "QUoM" (as an 					
	alternative to "QUOMOS")    (22UI)
	PeterYim: Joel: workplan slides - 					
	http://ontolog.cim3.net/file/work/UoM/UoM-standard-ontology_20091022/Trac-Motivation-and-Status--JoelBender_20091022.pdf    (22UJ)
	PatHayes: I also have to leave for another call. Peter, email me if you 					
	have any, er, tickets for me.    (22UK)
	JoelBender: 					
	http://ontolog.cim3.net/cgi-bin/wiki.pl?UoM_Ontology_Standard_OASIS_TC_Charter_Draft#nid22D2    (22UL)
	JoelBender: thank you    (22UM)
	PeterYim: ref. getting buy-in from the standard bearers (e.g. BIPM) - 					
	Steve & Howard making advances on those fronts; Ed feels we will be more 					
	ready when we have a draft standard; Pat suggest using the now popular 					
	notion of "Linked Data" to position ourselves    (22UN)
	PeterYim: great meeting, thanks, everyone    (22UO)
        -- session ended: 2009.10.22-12:27pm PDT --    (22UP)

For the record ...    (22PH)

How To Join (while the session is in progress)    (22PI)