Re: [Ietf-languages] Fwd: I-D Action: draft-msporny-d-langtag-ext-00.txt

[Martin J. Dürst <[e-mail]>]

Hello Manu, others,

On 2019/05/27 00:22, Manu Sporny wrote:
> On 5/25/19 10:22 PM, Martin J. Dürst wrote:
>> (Manu, I suggest you subscribe to this list for the duration of this
>> discussion, to make sure you are not missing any comments.)
> 
> Subscribed, thanks Martin. :)
> 
> On 5/26/19 2:10 AM, Phillips, Addison wrote:
>> There is a discussion in the W3C space (and particularly the Linked
>> Data, I.e. the RDF/JSON-LD space) about this. I happen to think Manu
>> has jumped the gun, but it is an internet draft and thus is meant to
>> spark discussion of this sort. When I'm at a real computer in the
>> morning I'll send some links for those interested in this debate.
> 
> Yes, I jumped the gun to push the conversation along... background for
> those that need it:
> 
> The W3C Verifiable Credentials defines a data model for the expression
> of things like Professional Licenses, Driver's Licenses, Education
> Certificates, Passports, Business descriptions (summary, hours of
> operation, etc.). The data model works across a variety of syntaxes
> (JSON, JSON-LD, N-Quads, HTML, CBOR, etc.). We want first-class i18n
> support, and a common approach to base direction has proved to be
> challenging across all of the expected syntaxes.

Yes. This is nothing more than expected. Working across syntaxes (and 
data models!) IS hard. You shouldn't work under the assumption that this 
is easy, and you shouldn't try to make it seem easy when it's not.


> One approach is to
> extend the language tag to also support base direction in a way that is
> backwards compatible. The proposal can be found here:
> 
> https://github.com/w3c/rdf-dir-literal/issues/3#issuecomment-496006350
> 
> There is a massive thread w/ background here:
> 
> https://github.com/w3c/rdf-dir-literal/issues/3

As very experienced people such as Ivan Herman (@iherman) and Charles 
McCathieNevile (@chaals) have expressed in that thread, it's 
inappropriate to force something like this direction information on HTML 
(which has solved the problem since more than 20 years, see 
https://trac.tools.ietf.org/html/rfc2070). It would be very much the 
tail ("Verifiable Credentials") wagging the dog.

So at a minimum, you have to make sure that when data gets converted 
from RDF to HTML, the -D extension gets removed and the corresponding 
HTML dir attribute gets added (and vice versa for the conversion in the 
other direction). You should also check out formats such as CBOR, which 
work somewhat differently from JSON, RDF, or HTML. This is actual work, 
make sure it gets done.

Once you realize that actual work is involved, you may realize that 
using a separate attribute (or whatever the equivalent in other formats) 
may not be that bad a solution.


> ... which led to this I-D, written by someone that is definitely not an
> expert in this area (me):
> 
> https://tools.ietf.org/html/draft-msporny-d-langtag-ext-00
> 
> There is a time pressure here. Our i198n concerns have been hanging out
> there for more than 9 months and our WG charter is up in a couple of
> months. We need to wrap this up in 3 weeks. Or to put it another way, if
> we don't wrap this up in 3 weeks, we won't be addressing this issue,
> which would be a shame.

This wouldn't be the first standards effort where a group's duration is 
extended, and it wouldn't be the last. And in this case, an extension 
might actually be justified. After all, this is something that shouldn't 
just work for Verifiable Credentials, but for a lot of other RDF, 
JSON-LD, JSON,...

3 weeks or 2 months are nothing given that the first teleconference I 
have been on discussion language tags in RDF was more than 20 years ago.
(see https://www.w3.org/TR/1999/REC-rdf-syntax-19990222/ and search for 
xml:lang).


> In any case, if you're going to engage in this discussion, the issue #3
> above is probably the place to do it.

I'll try. But I know I'm better with email than with github issues, sorry.

> Martin, I'm tracking your comments here:
> 
> https://github.com/digitalbazaar/d-langtag-ext/issues/1

Thanks.

> Martin, almost all, if not all, of your input is reasonable and I'll
> make those changes if it looks like there is consensus to proceed with
> draft-msporny-d-langtag-ext-00.. that is, if Addison doesn't strangle me
> first. :P

Okay.

Regards,   Martin.