Doug Ewell dewell at
Fri Apr 21 06:18:51 CEST 2006

Michael Everson <everson at evertype dot com> wrote:

> I would like to inform you all that I will be away in the US and 
> Switzerland 21 April to 11 May and if there is nothing urgent and this 
> list were silent that would suit me pretty well.

Apologies in advance.  I'll try to provide a summary to brief Michael 
upon his return.

Kent Karlsson <kentk at cs dot chalmers dot se> wrote:

> I was kind of hoping this suggestion would just go away...

I hope that no serious suggestion on this list will ever "just go away." 
Every suggestion that is made in good faith and with an eye toward 
improving the usability of language tags should always be considered and 
discussed publicly.  We owe it not only to the proponent, but also to 
the users who may be affected.

> Otherwise we have exactly the same issue with AX and FI. Note or not, 
> FI presumably used to include Åland, now it presumably does not 
> (though I haven't followed the issue closely). So sv-FI used to cover 
> both Swedish as used in Finland (excl. Åland) and as used in Åland, 
> while now they have separate codes, sv-FI and sv-AX. As it happens 
> they are not identical (as languages). However, sv-SE (official, 
> disregarding dialects) and sv-AX are identical as languages...

John and Debbie both noted that the underlying ISO standard never stated 
that FI encompassed AX, although de facto it probably did.  Debbie's 
proposal reflects a remark actually made, and subsequently withdrawn, 
within ISO 3166-1.

> Now, are we to note that in the LSTR?

Maybe.  You've made a reasonable case for it.

> And how about 'nn' and 'nb' vs. 'no'? There is no "guiding comment" 
> about those. In practice, I'd say 'no' is equvalent to 'nb', even 
> though 'no' should cover both 'nb' and 'nn'. This one really needs 
> some kind of guidance.

I agree completely.  The situation with these three subtags is 
particularly confusing, and will be even more so when ISO 639-3 is 
published, since that standard treats "Norwegian" as a macrolanguage 
covering Bokmål and Nynorsk, and the Registry does not.

> I'm sure there are other cases.

There may be, but you've only raised two so far.  I don't see any 
evidence yet of a great flood of superfluous annotations.

Look, folks, these are just COMMENTS.  They are not normative.  They are 
not permanent.  They do not affect processing.  We can add them, change 
them. delete them, and add them back virtually at will.  They do not 
represent a great, earth-shaking decision that will haunt us the rest of 
our lives.

Debbie Garside <md at> wrote:

> I think the GB issue has somewhat set a precedent.  That said, I think 
> each case should be dealt with on its own merits.  Whether people here 
> feel that this precedent should now be extended to include historical 
> entries is a matter for the list (I say that at risk of incurring the 
> wrath of Doug).

I think "wrath" might be a wee overstatement of my initial objection.  I 
was not originally enthusiastic about this proposal, but came to realize 
that we are talking about adding *two lines* to a Registry that already 
contains over 5,000 lines.  That kind of bloat does not scare me.  And 
if it really does help users to select the right region subtag... well, 
isn't that what comments are for?

Doug Ewell
Fullerton, California, USA

More information about the Ietf-languages mailing list