Re: [Ltru] ISO 639-6

[Peter Constable <[e-mail]>]

I’ll provide this reply and then drop the discussion.

Gerard appears to suppose that, since I refer to “ROI”, I must be speaking on behalf of Microsoft. That is simply not the case. I was speaking as a language technologist expressing my thoughts on what benefits, in practical terms, I think would accrue to language communities around the world. Benefits are not realized simply by having a tagging system that comes with tens of thousands of already-sanctioned elements. Rather, benefits are realized by having a tagging system that accommodates their needs *and* that works in practice and is widely supported. I am strongly influenced by those practical considerations from time I spent in a not-for-profit organization, prior to working for Microsoft, trying to find technology solutions for thousands of language workers working in divers language communities throughout the world, with sizeable investments made in solutions that were *not* widely adopted in industry. Such solutions did not scale and fostered dependence on a small support infrastructure rather than giving long-term empowerment.

In fact, the opinions I was expressing were in line with an opinion I formed some years before I started to work at Microsoft: that sub-language variations are not appropriate objects for a comprehensive encoding because there’s no clear way to define dialects. I first discussed this in 2000, but the first time I published may have been in a paper I presented to a Unicode conference in 2002. The full paper is available here:

http://www.sil.org/silewp/2002/SILEWP2002-003.pdf

See section 4.6, p. 29f:

“It was
claimed earlier that it is possible to enumerate a comprehensive list of individual languages based
on a particular operational definition—to “tile the plane” of languages. In contrast, this is, in
principle, impossible for dialects. The reason has to do with operational definitions….”

I simply hold the same views that I had then.

I’m completely in favour of doing what is good for goodness sake. That is what has motivated my work over the past many years in character encoding and to develop ISO 639-3 and get it incorporated into BCP 47 – I think those have been steps toward goodness from where we were before. (Gerard may think ISO 639 is a mess, but IMO it’s less of a mess and a more well-defined one than it used to be.)

I remain unconvinced that ISO 639-6 will, on the balance, provide overall goodness to BCP 47. Even so, I haven’t done anything to stand in the way of ISO 639-6 as an ISO standard, and I wouldn’t necessarily oppose a revision to BCP 47 to allow ISO 639-6 subtags – depending on how it is handled. All that I said in this thread I would not support is putting these in the reserved alpha-4 subtags. I have various reasons for that, but I’ll refrain from going further in respect of Randy’s request.


Peter


From: Gerard Meijssen [mailto:[e-mail]]
Sent: Tuesday, March 17, 2009 1:17 AM
To: Peter Constable
Cc: LTRU Working Group
Subject: Re: [Ltru] ISO 639-6

Hoi,
I found the reply that Peter made in his reply on the thread "Re: [Ltru] ISO 639-6 (was: Geocoordinates)" problematic. In it he said: "But I can tell you up front that I will most likely push back on a revision to make use of the reserved alpha-4 subtags not only because it's inconvenient for my derivative application but also because I remain very skeptical that there will be benefits (let alone ROI, considering the costs).". I find this problematic because it puts commercial interests clearly ahead of a structured approach that comes with the ISO-639-6.

Now that Peter states that he is expressing his own opinions and not the opinion of Microsoft, I find the statement quoted above unacceptable. It is unacceptable because he is spouting FUD on a standard that intends to reconcile the different standards in one body, a standard that intends to finally give a place to the different orthographies. When Peter speaks for Microsoft about ROI and costs, I find it problematic because I wonder how he reconciles this with being a member of the community, when Peter "only" expresses his own opinion it becomes unacceptable to me because he clearly presents himself as a Microsoft employee. ROI and costs are business related  considerations and making these "personal opinion" is disingenuous probably dishonest and in my opinion unacceptable.

There are clear benefits to a standard that reconciles the current mess of ISO-639 standards that are problematic to reconcile and have made the current RFC process a waiting game. At this time there is no software that provides the "most basic" language support. The most basic language support is recognising that people want to write in their own language and enabling them to do so. Enabling this is nothing more then setting the appropriate meta data and allowing people to express themselves. The best argument against providing a "most basic" language support has been the standards do not support this and with "personal opinions" expressed in this way it is less likely when we will see the day that people can express themselves in their language.
Thanks,
        Gerard
2009/3/17 Peter Constable <[e-mail]<mailto:[e-mail]>>
Due to off-line mail I received in relation to some of my comments in this thread, I feel I need to make one additional comment:

The opinions I have expressed are my own and not those of my employer or any other party with whom I may be affiliated.


Peter

-----Original Message-----
From: [e-mail]<mailto:[e-mail]> [mailto:[e-mail]<mailto:[e-mail]>] On Behalf Of Randy Presuhn
Sent: Friday, March 13, 2009 11:39 PM
To: LTRU Working Group
Subject: Re: [Ltru] ISO 639-6

Hi -

As co-chair...

I don't see much progress in this discussion.  I suggest that those
wishing to see work in ltru on the use of ISO 639-6 as a source of
subtags should submit a specific charter update proposal to this
mailing list for discussion.  Please refrain from further comment
on this topic until someone provides a specific (i.e. containing dates
and delivarables) charter update proposal.

Randy

_______________________________________________
Ltru mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/ltru

_______________________________________________
Ltru mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/ltru