Re: [Justfont] Working Group Last Call: draft-ietf-justfont-toplevel-02.txt

[Martin J. Dürst <[e-mail]>]

Hello Wendy,

On 2016/05/25 04:54, Wendy Seltzer wrote:
> Thanks Chris and Martin,
>
> On 05/09/2016 01:49 PM, Chris Lilley wrote:
>>
>>
>> On 16/05/09 03:45, Martin J. Dürst wrote:
>>>
>>> - The start of section 7 introduces @font-face format identifiers. It
>>> uses the word "defined" for these, but only talks about the types
>>> defined in the document itself. This raises the following questions:
>>>   - Are registrations of new subtypes supposed to give this
>>>     information, too? If yes, who is responsible to avoid conflicts?
>>>     (we could just ask IANA to check that the identifier doesn't clash
>>>      with that of another subtype)
>>>   - Will a future version of css-fonts-3 reference the IANA registry
>>>     for these values, or how will potential discrepancies be solved?
>>>   - If the information in css-fonts-3 is supposed to be normative, and
>>>     the information in the IANA registry just informative, then
>>>     clearly say so in the draft.
>>
>> This is a very good point. I have opened an issue
>> https://github.com/svgeesus/ietf-justfont/issues/14
>>
>> This requires discussion, as there are several options and it is not
>> immediately clear which is best.
>
> Maybe you could spark that discussion with a proposed way to resolve the
> issue?

I have thought about this a bit, and have come closer to a proposal.

I trust all of the W3C, the IETF, and the IANA, so I don't have a 
preference on that front.

IANA does bookkeeping extremely well, so asking it to do it will make 
sure these identifiers don't get forgotten if there's a font/... 
registration. But that wouldn't apply to font formats registered under 
e.g. application/... (well, that shouldn't happen anymore once we have 
defined font/..., so not a big issue) and font formats that don't get 
registered. Having IANA do it might also be an additional incentive for 
registering the media type.

On the other hand, having to update the CSS spec just because of a new 
font format is work I'd want to avoid if I were in charge of that spec. 
So my proposal would indeed be that all new font/... registrations have 
to provide such a value, and that IANA checks that value and lists it 
somewhere (maybe just the registration template would be enough).

css-font-3 could then be updated to point to that information, but that 
could take its time, no need to hurry.

So that would be my proposal, but if somebody closer to the 'metal', 
e.g. Chris, has different preferences, that would be absolutely fine 
with me, too.

Regards,   Martin.