headers
Doug Ewell | 1 Dec 05:38

Re: Fallback vs. Macrolanguage? (was: John Cowan throws in...)

John Cowan <cowan at ccil dot org> wrote:

>> Wouldn't it be easier, more consistent, cleaner, and cherry-free...
>
> I don't think it's necessarily appropriate to fall back in all cases 
> of macrolanguages (as Peter's list shows), and it may well be 
> appropriate to fall back in some cases that aren't macrolanguages at 
> all (Breton -> French is the example we've talked about a lot).  Using 
> a different name lets us decouple from 639-3's particular choices.

We can keep the Macrolanguage name and follow the ISO 639-3 mappings 
exactly, or we can give this field a new name (such as John's 
suggestion, "Fallback") and cherry-pick the ones like Chinese and Arabic 
where one variety dominates.  I think it would be too confusing to have 
both, as I said before about having both a Macrolanguage field and 
extlangs.

However, I strongly oppose any efforts to add our own, non-ISO 639 
mappings to the Registry, even seemingly obvious ones like Breton -> 
French.  That way lies complete madness.  We will never come anywhere 
close to getting all of them, and we will be reviled for "ignoring" the 
ones we miss, and every decision to either include or not include some 
mapping will be viewed by someone as an intentional slight or attempted 
subjugation of one of the languages involved.  Please, let's leave this 
kind of mapping up to each application.

--
Doug Ewell  *  Fullerton, California, USA  *  RFC 4645  *  UTN #14
http://home.roadrunner.com/~dewell
http://www1.ietf.org/html.charters/ltru-charter.html
(Continue reading)

Permalink | Reply | 5 comments |
headers
Doug Ewell | 1 Dec 05:51

Re: Language tags in the future version of HTTP

Stephane Bortzmeyer <bortzmeyer at nic dot fr> wrote:

> That's a difficult issue. What some people want, at this time, is to 
> *include* RFC 4646 ABNF in RFC 2616bis, in order to make it 
> self-sufficient (anyone knows if there are general IETF guidance here? 
> about cross-RFC references?)

Cross-references like that are generally considered a good thing, aren't 
they?  They reduce the chances of introducing a typo that changes the 
ABNF, and serve to discourage each application from adopting their own, 
slightly different syntax.

"Self-sufficiency" should be a complete non-goal.  How many modern RFCs 
seek to define MUST and SHOULD independently of RFC 2119, or the format 
of a mail message independently of RFC (2)822?

--
Doug Ewell  *  Fullerton, California, USA  *  RFC 4645  *  UTN #14
http://home.roadrunner.com/~dewell
http://www1.ietf.org/html.charters/ltru-charter.html
http://www.alvestrand.no/mailman/listinfo/ietf-languages  ˆ
Permalink | Reply | 2 comments |
headers
Doug Ewell | 1 Dec 06:02

The true meaning of "macrolanguage" (was: Re: John Cowan...)

Peter Constable <petercon at microsoft dot com> wrote:

> If Kwanyama and Ndonga are distinct languages and if "Oshiwambo" can 
> refer to either and is in some contexts used as though it were a 
> single language rather than a collection, then it can be considered a 
> macrolanguage. But if Oshiwambo refers to a distinct variety that is 
> not the same as either Kwanyama or Ndonga but is used or coming into 
> use by speakers of both Kwanyama and Ndonga communities, then it might 
> make more sense to consider Oshiwambo an individual language.

Peter has always been very clear and consistent in his explanation of 
the term "macrolanguage," and the paragraph above is no exception. 
However, I fear there is still a great deal of misunderstanding 
surrounding this term, and I hope Peter and the ISO 639-3/RA are bold in 
explaining and promoting the correct use of this term, which is not 
synonymous with "language family," "language collection," "dominant 
language," or any other term.

--
Doug Ewell  *  Fullerton, California, USA  *  RFC 4645  *  UTN #14
http://home.roadrunner.com/~dewell
http://www1.ietf.org/html.charters/ltru-charter.html
http://www.alvestrand.no/mailman/listinfo/ietf-languages  ˆ
Permalink | Reply | 0 comments |
headers
Peter Constable | 1 Dec 08:01
Picon
Favicon

RE: Re: Fallback vs. Macrolanguage? (was: John Cowan throws in...)

> From: Doug Ewell [mailto:dewell <at> roadrunner.com]

> However, I strongly oppose any efforts to add our own, non-ISO 639
> mappings to the Registry, even seemingly obvious ones like Breton ->
> French.

I agree. Users should choose their own language fallbacks. A an appropriate fallback for Oriya might as
readily be Bengali or Hindi or Kannada, depending on the user. That in addition to the issues Doug mentions.


>  That way lies complete madness.  We will never come anywhere
> close to getting all of them, and we will be reviled for "ignoring" the
> ones we miss, and every decision to either include or not include some
> mapping will be viewed by someone as an intentional slight or attempted
> subjugation of one of the languages involved.  Please, let's leave this
> kind of mapping up to each application.



Peter
Permalink | Reply | 0 comments |
headers
Stephane Bortzmeyer | 1 Dec 14:25
Picon

Re: Fallback vs. Macrolanguage? (was: John Cowan throws in...)

On Fri, Nov 30, 2007 at 08:38:58PM -0800,
 Doug Ewell <dewell <at> roadrunner.com> wrote 
 a message of 38 lines which said:

> However, I strongly oppose any efforts to add our own, non-ISO 639
> mappings to the Registry, even seemingly obvious ones like Breton ->
> French.

Like many people here, I agree. (One of the reasons being that such
mappings would not be very stable.)

> Please, let's leave this kind of mapping up to each application.

But I don't think we should add this solution. I understand the need
for user-specific mappings but I do not see why an
application-specific mapping could be useful.

An "official" "fallback mapping" somewhere (ISO 639-something?) would
be very useful. But I agree that it is not our job.
Permalink | Reply | 3 comments |
headers
Frank Ellermann | 1 Dec 14:37
Picon
Picon

Re: Language tags in the future version of HTTP

Doug Ewell wrote on the LTRU list:

> Cross-references like that are generally considered a good thing,
> aren't they?  They reduce the chances of introducing a typo that
> changes the ABNF, and serve to discourage each application from
> adopting their own, slightly different syntax.

Depends.  For 2616bis they could use a simplified syntax in the
direction of  tag = primary *( "-" auxiliary )  directly expanded
in  tag = 1*8( ALPHA / DIGIT ) *( "-" 1*8( ALPHA / DIGIT ))  with
a note in prose that tags are actually supposed to be BCP 47 tags.

> "Self-sufficiency" should be a complete non-goal.

As proposed above, where is the problem ?  I guess 2616bis needs
also ranges (with "*" stars) based on RFC 4617 somewhere, that is
also covered by BCP 47, but references in RFCs to BCPs, STDs, etc.
are typically given by their RFC number (at least for documents
created with xml2rfc).

> How many modern RFCs seek to define MUST and SHOULD independently
> of RFC 2119

RFC 2821 and I-D.klensin-2821bis come to mind.  It's a royal pain
to check that these definitions are verbatim copies not citing the
source for a very esoteric reason (arguably 2821bis doesn't follow
the one and only MUST in chapter 6 of 2119).

> or the format of a mail message independently of RFC (2)822?
(Continue reading)

Permalink | Reply | 0 comments |
headers
Frank Ellermann | 1 Dec 14:45
Picon
Picon

Re: John Cowan throws in the towel on extlangs

Addison Phillips wrote:

> if John's towel stays thrown, it wouldn't impact RFC 4646bis.

The extlang-ABNF and the "longest tag" discussion are affected,
or are you talking about 4645bis ?

 Frank
Permalink | Reply | 4 comments |
headers
Frank Ellermann | 1 Dec 15:26
Picon
Picon

Re: draft-ietf-ltru-4646bis-09.txt

Elisabeth Porteneuve wrote:

 [2. about the 2006 ed. of ISO 3166 and 50 years]
> Yes.

Thanks for info, maybe the Netnews RFC can be updated in
AUTH48 (it's actually almost irrelevant for Netnews, not
the same situation as for 4646bis).

> All change of the world in Europe in 1990's until the most recent
> split into Serbia and Montenegro are related to the second world
> war.  To some extent the second war is being finished now.

Kosovo will be the next.   

> I am optimistic, at least as far as Europe is concerned, to not
> see much changes in country codes.

I'm optimistic to see changes for regions where the population
apparently wishes to gain independence of their "encompassing"
country.  For regions like Iraq or Somalia registering existing
entities like Somaliland might be the best possible solution.

> Concerning languages, I noted the inflation related to the
> former Yugoslavia split - the Serbo-Croatian, or Yugoslavian,
> was replicated with N different names. That is of course 
> politically sad, as all people from the former Yugoslavia
> speaks Yugoslavian, Southern Slavic language (yug=south).

Right, but ISO 639 has alpha3 code elements, they're likely
(Continue reading)

Permalink | Reply | 0 comments |
headers
Frank Ellermann | 1 Dec 15:41
Picon
Picon

Re: action item: "other" collections in 639

Peter Constable wrote:

>>> I've suggested in the past that removing collection IDs from
>>> BCP47 might not be a bad idea, but on several occasions people
>>> have indicated that they wish to have the collections kept.

>> "Removing" can't fly.  Do you mean "not add new" (in the 4645bis
>> bulk update), "deprecate old", or both ?

> Deprecate, and not add new.

Okay.  Then it could make sense to deprecate some giving a new
preferred value based on your list.  Ideally in a form where
Doug can reference a better source than an archived article in
4645bis, but if all else fails a "WG decision" is good enough.

 Frank
Permalink | Reply | 1 comment |
headers
Peter Constable | 1 Dec 16:19
Picon
Favicon

RE: Re: Fallback vs. Macrolanguage? (was: John Cowan throws in...)

> From: Stephane Bortzmeyer [mailto:bortzmeyer <at> nic.fr]

> An "official" "fallback mapping" somewhere (ISO 639-something?) would
> be very useful.

I disagree. A user with a system set for Breton as the preferred UI language might want to fallback to French
or Japanese or Swahili -- it all depends on the user. Perhaps French is a reasonable default that could be
assumed in that case in the absence of other information, but that's just one case. What is the fallback for
German in Switzerland? What is the fallback for Hindi in India? What is the fallback for Inuktitut, or Filipino?

Peter
Permalink | Reply | 1 comment |

Gmane