Re: [link-relations] app data wrt HTML usage, was: NEW APP DATA

[Julian Reschke <[e-mail]>]

On 11.08.2010 11:24, Ian Hickson wrote:
>
> This thread is going off the topic of the field registrations, so I've
> only replied to the relevant parts here.

Indeed.

> ..
>> Incorrect; it's also the default for new values.
>
> Surely for new values we would want the people submitting registrations to
> consider how their relations will work with existing applications and thus
> they will explicitly say what the values should be?

Yes, but the thing we do not agree on is the level of detail required.

The application data registry was created to deal with the use case of 
distinguishing between links and external resources, for instance, so a 
client knows what to archive or to download for offline access.

I think this is a good use case, and it covers *most* of what you're 
looking for.

> I would not at all feel comfortable defaulting all future values to either
> "hyperlink", "external resource link", or "annotation link". Which one
> applies should be carefully considered in each case. If no consideration
> is made, the only safe answer IMHO is "not allowed". This argues strongly,
> again IMHO, that we absolutely must have the latter as one of the possible
> options, if we are to have this field at all.

See above: by defining the flag in general terms instead of 
HTML-specific terms, the question simply goes away.

> At this point my assumption is that the field registration is being
> rejected and we will have to deal with it outside the registry.

Again, please consider simplifying the application data, as originally 
discussed.

Best regards, Julian