Considerable movement by John Clews on -oed

[John Clews]

OK - Peter Constable, John Cowan and Michael Everson give some
very valid and reasonable points as to why it's OK.

So basically, I backtrack.

I'd be grateful if you'd take a second look at the points I mentioned
earlier, before it gets finally agreed, in case there is the odd
nugget left.

Just answering some of John Cowan's points specifically:

In message <20030530123701.GN22039 at ccil.org> John Cowan writes:
Re: Pros and cons of adding a en-GB-oxford language tag

> John Clews scripsit:
> 
> > It's proposed for dealing with written English text. Even though the
> > tag may exist, it doesn't follow that the spelling software for OED
> > English is available to a user.
> 
> True, but I'm willing to bet that if a tag were issued (and en-gb-oxford
> sounds like Oxfordshire to me; I think en-gb-oed is better), a word in
> the right quarter would get a spelling list created in short order,
> at least in the free software community.

In general, I prefer -oed to -oxford.

> > Why is the -gb part significant if "Oxford English Dictionary" if
> > OED English has such widespread international use?
> 
> It's essentially a variant of en-gb.

And I suppose it's also valid to say that -gb (or any country code
as a second-subtag) implies written language: does that assumption
definitely work for all previous xx-xx or xxx-xx examples?

> > What may be better would be to encourage developers of spellcheck
> > software and spellcheck subroutines within applications to provide
> > for more alternative dictionaries if there is a requirement.
> 
> Hence the utility of a code.  If you mark up your document with en-gb-oed,
> OED spellchecking should be applied to it automatically, and no matter
> if you are the original author of the document or not.

OK.

> > A choice of "British English" and "American English" would be
> > sufficient for most purposes.
> 
> Canadians can tell you just how annoying that particular binary choice is.

Yes, sorry - I beliieve you.

> > In each case it's obvious what is meant, if you can read English.
> 
> Well, I don't know: when I posted the "tire centre" example to another
> mailing list, a British participant said it made him think of a place
> where people go to get tired.
> 
> > And (in passing) do Canadians use "tyre" to spell the word for the
> > rubber things on wheels?
> 
> No; that was the point of the example.
> 
> > Also most _users_ would "accept language" in all cases (in terms of
> > understanding the meaning) in all the following cases:
> 
> You seem to have omitted the actual cases.

Sorry - that was just the "tire centre, tire center and tyre centre"
cases.

> > Most language tags are for use in browsers, 
> 
> That's *one* application.

Indeed - sorry: I seem to have forgotten XML developers and users,
etc, which is a rather large part of the world.

I take Peter's points about using en-gb-oxford or using en-gb-oed
in XML markup.

> > Do we want browsers somehow to change text (and how would we do
> > it?) What would happen if a browser displayed the text
> > 
> >     "American English spells words differently to British English.
> >     Tire, center, and color are examples in American English.
> >     Tyre, centre, and colour are examples in British English."
> 
> Well, my browser (if it were as smart as you are talking about) would
> change that to "American English spells words differently from British
> English."  "Different to" grates on American sensibilities as much as
> American spellings do on Brits -- it looks not so much foreign as
> plain wrong.  (I know better, obviously, but most of my compatriots
> do not -- then again, most of them write "different than", barbarously.
> Tempora mutantur, nos et *non* mutamur in illis.)
> 
> No, I don't think automatic orthography translation is either practical
> or worth doing.

Fair enough - if HTML browsers will (with few exceptions, which I
can't envisage) be dumb enough to ignore it, that's fie by me.

> >     "American English spells words differently to British English.
> >     Tyre, centre, and colour are examples in American English."
> >     Tyre, centre, and colour are examples in British English."
> 
> I don't see how en-gb-oed makes this example any worse.  All of those
> words are spelled the same in en-gb and en-gb-oed.

Alright: add in the -ise/-ation and -ize/-ation words too.

I asssume that Michael's annoyance at the -ise/-ation conventions
being used instead of -ize/-ation conventions partly led to this.

But why do "en-gb" spellcheckers generally go for the -ise/-ation
conventions instead of the -ize/-ation conventions?

Do they use - or refer to - a different authoritative source?

Does anybody know?

And I take the point (Peter's?) about the very existence of a tag
like en-gb-oxford or en-gb-oed might well stimulate somebody to
develop a spellchecker or other tools for en-gb-oxford or en-gb-oed.

> > These conventions - for allocating spelling conventions - should be
> > in the editig tools, such as spellcheckers, not as a general
> > mechanism that could be used in browsers.
> 
> RFC 3066 is about the representation of languages, not the purpose to
> which those representations are applied.

OK - I agree.

> > I would argue against adding en-gb-oxford until most of those
> > problems can be overcome.
> 
> I believe that because it meets the formal requirements it should go
> forward.

So do I now.

But thanks for letting me raise them, and you and others for
addressing them!

Best regards

John

--
John Clews,
Keytempo Limited (Information Management),
8 Avenue Rd, Harrogate, HG2 7PG
Tel:    +44 1423 888 432
mobile: +44 7766 711 395
Email:  Scripts2 at sesame.demon.co.uk
Web:    http://www.keytempo.com

Committee Member of ISO/IEC/JTC1/SC22/WG20: Internationalization;
Committee Member of ISO/TC37/SC2/WG1: Language Codes