RE: Update on the Constraints changes draft page

Looking at https://www.w3.org/2016/poe/wiki/Constraints I see a question about “industry”:

“Michael: It is not clear to me how a use related to an industry could be measured. Example: if the use “print” is constrained by “must be equal to the bank business sector” what does that mean in practice? Printed only in products published by a company of the bank business sector? Or printed only on a paper/poster shown in a bank?”

This is a legitimate question. But it also applies to many other concepts in ODRL, such as “purpose” and, I would argue, “media”, “spatial”, “language”, “event” and several others.

Essentially, someone has to decide whether a given entity should be categorized by the corresponding property value. For example:

Is a given video in a particular language? Sounds straightforward, but what happens if the audio track has a mix of say, English and French spoken? Or maybe you’re talking about the subtitles on the video? Or is it the language used to describe the metadata?

How do you determine whether a particular asset (say a photograph) is being used for an educational purpose?

How do you determine whether a use is within a particular geospatial area? If I’m syndicating news articles to a website but that are not to be distributed within the UK, then does that distribution restrict the physical location of the webserver (does that make sense in the era of the “Cloud”?) or the legal owner of the domain? Or is it the location of the viewer of the web page?

For “media”, we list examples of electronic, print, advertising or marketing. (Which is a pretty heterogeneous list!) But how do we determine whether something is advertising or marketing? Advertorial and paid content deliberately blur those distinctions.

As I said on a recent POE phone call, the way that we at the AP today is that we categorize an assignee by country, purpose, industry and so on. I’m sure that other organizations do something similar.

Regards,

Stuart



From: Renato Iannella [mailto:renato.iannella@monegraph.com]
Sent: Monday, November 28, 2016 8:03 PM
To: W3C POE WG
Subject: Re: Update on the Constraints changes draft page


On 29 Nov. 2016, at 03:06, Michael Steidl (IPTC) <mdirector@iptc.org<mailto:mdirector@iptc.org>> wrote:

•         I’ve modified some proposed changes this way: if they only provided a more precise definition but did not change the semantics substantially the constraint term got a modified label but the term (= id) was left untouched.
•         … and I withdrew proposals to deprecate terms with the “old” term id and to create terms with a new id as the modified label shows the refined definition. (I hope users of ODRL will get that.)
•         As suggested at the call I added “NT (= Narrower Term) of” relationships to new terms which should be at a hierarchical level below another term.

That's looking good ;-)

The “eventEndPeriod” is still not clear to me.
Is this a new constraint to address our previous (classic) example of “30 minutes after the end of the Football match”?

For your comment on “industry” use - the example I have seen is “adult-entertainment” (so, I have been told ;-)


•         The “Proposal regarding the Constraint class” at the top of the page got two examples added.


Are you proposing that we add 9 new properties (called leftOpResource1..9) to the vocab?

In UC#1 - does this mean you would have something like the following in the ODRL expression:

     odrl:constraint [
      a odrl:Constraint ;
        odrl:count 10 ;
        odrl:operator odrl:eq ;
        odrl:leftOpResource1 <https://my.system/my-api/?get-current-count-value>
        ] ;

In UC#2 - Another way to tackle this is to use the “rightOperand is a reference to a value” [1].

This would mean that you don’t know the datetime (until you deref the URI).
This would mean the constraint looks like a “normal” datetime constraint:

     odrl:constraint [
      a odrl:Constraint ;
        odrl:dateTime <http://football.uk/round5/manuVmanc/<http://football.uk/round5/manuVmanc/>>;
        odrl:operator odrl:gteq ;
        odrl:dataType <odrl:resourceURL>
        ] ;

Deference-ing the URI <http://football.uk/round5/manuVmanc/<http://football.uk/round5/manuVmanc/>> will return an actual datetime (plus any embargo time they like to add)


Renato Iannella, Monegraph
Co-Chair, W3C Permissions & Obligations Expression (POE) Working Group

[1] https://github.com/w3c/poe/issues/56<https://github.com/w3c/poe/issues/56>