FYI: Discussion with OASIS UnitsML TC by David A. Wheeler (CONVERT)

["David A. Wheeler" <dwheeler@dwheeler.com>]

FYI - I thought it'd be a good idea for the office TC to know that we (the office formula subcommittee) are briefly coordinating with the OASIS UnitsML TC. Here's more info about that.

--------------------------------------------------------------------

One of the functions we've had to spec in the office formula subcommittee is CONVERT.  This is a function that converts between various unit systems.  Its specification is related a number of really important international standards, it turns out to be subtler and more complicated than it might appear at first, and we wanted to make sure we "got it right".

As part of wrapping up the definition of a function named CONVERT, I attended an OASIS UnitsML TC teleconference on Feb. 14, 2007.  I think it's a VERY good idea for standards groups to coordinate when it makes sense, and this is one of those times.  This coordination was noted in the office-formula mailing list, but I thought that it might be wise for the office TC to also be aware of this since it's an external interaction.  Let me know of any problems with this!

Below you can see what I requested, and that I got VERY POSITIVE responses from the UnitsML TC.  I'm REALLY HAPPY with how things went!!!  In particular, the chair of the OASIS UnitsML TC has agreed to personally review our CONVERT function spec (to make sure it's correct and to make suggestions), he will make some hard-to-get information available to me, and he will make sure that their UnitsML specification allows cross-references to the OpenFormula work (particularly our unit names).

In my mind, this is really great, and it's EXACTLY how things are supposed to work.  In this case, two different standards groups have work that "touch" each other, and so they are briefly working together to make sure they're really working TOGETHER instead of in conflict.  It's this sort of interaction with the broader community of standards that shows the powerful advantage of OpenDocument (and good standards work in general); OpenDocument development has always worked fervently to reuse and work with other standards.

We're also showing the advantages of widespread review.  Here we have an international expert who's agreed to review a function that touches his area of expertise.  Indeed, just in last few days we've had other experts who have been reviewing the draft and noted issues in other areas such as multi-value XOR and ZTEST.  This is great stuff folks!  My thanks to all.  We can produce really great results with this kind of widespread expert review.

Details below. I've intentionally written the text below in a formal way, just in case this winds up in some official records.  To understand it, note that the OASIS UnitML TC is developing an XML schema for describing measurement units (called UnitsML). The expectation is that other projects will use this XML schema to actually describe units. NIST has a draft units database called UnitsDB that is a likely user of UnitsML, and that UnitsDB draft could be very useful to us.

Anyway, I think this coordination is a good thing.  Hooray!

--- David A. Wheeler 

=========================================

David A. Wheeler is the chair of the OASIS office formula subcommittee, and attended the UnitsML TC teleconference on Feb. 14, 2007.  He wanted to coordinate the office formula ("OpenFormula") subcommittee with the UnitsML TC, and in particular make sure that the results from these two groups do not conflict.

Mr. Wheeler made three requests; here are those requests, and his understanding of the responses from the UnitsML members:

1. Would someone from the UnitsML TC be willing to review OpenFormula's CONVERT function, to make sure that it's accurate and give suggestions?
Robert Dragoset (UnitsML chair and who works at NIST) said he'd be willing to do that review.

2. Could we (the OpenFormula subcommittee) gain access to related measurement information, in particular NIST's UnitsDB and ISO 31?
Robert Dragoset will send Mr. Wheeler information on how to view the draft of UnitsDB. Mr. Dragoset warned that UnitsDB is still a draft, and has some known errors. ISO 31 is not an open access document, and is extremely expensive.  Robert Dragoset has a personal copy of ISO 31, which he received from a previous user. All agreed that Mr. Dragoset could not simply send a photocopy to Mr. Wheeler.   Since both Mr. Dragoset and Mr. Wheeler live in the Washington, DC area, Mr. Dragoset said he'd be willing to let Mr. Wheeler visit to look at Mr. Dragoset's copy of the ISO 31 document at NIST.

3. Would the UnitsML group be willing to add a place in its scheme so that the OpenFormula symbol could be recorded in a UnitsML-compliant file?
Robert Dragoset wasn't sure if the current schema allowed this yet, but stated that they very much intend to have the ability to include maps to other code lists.  Mr. Wheeler got the impression that Mr. Dragoset would ensure that the final UnitsML schema would have this kind of capability.

Mr. Wheeler thanked everyone in the UnitsML TC for their time.