Re: shapes-ISSUE-113 (SHACL and user interfaces): [SHACL Spec]

Having seen the UX software pointed out by Bart [1] I think it is 
important that we get a grip on what part of UX support makes sense 
within SHACL, and how UX requirements will be integrated. A connection 
between UX software and SHACL is a no-brainer, but we haven't discussed 
how best such a connection is made, nor where functionality does and 
does not overlap.

I came into this group with the notion of "application profile" that has 
been bandied about in the Dublin Core arena. That was a very simple 
vision of describing a metadata vocabulary and a small set of 
constraints. A version of it has been integrated into a simple user 
interface.[2] I now envision that such a profile, useful only in very 
simple cases as currently defined, could be mid-ware between SHACL and 
the code controlling UX. Putting all of this into SHACL itself might not 
be the best solution, since it may be advantageous to build different 
user interfaces from a single set of constraints. However, I think we 
need a full discussion of this, which has not yet taken place.

Holger, as the group works, it must be possible to make changes. I would 
imagine that we could even come to conclusions that are not entirely in 
line with the charter, although it would be necessary to be very clear 
as to why such changes were necessary. I don't find it useful to assume 
that past decisions are iron-clad, especially ones that were not 
developed beyond one-sentence requirements. And note that some of the 
use case specifics, such as wanting to mark properties as read-only, 
have not been incorporated into SHACL, but without discussion. Just 
because something is in SHACL today does not mean that the group has 
agreed on it.

kc
[1] 
http://blog.resc.info/using-interface-encapsulation-to-listen-to-linked-data-predicates/
[2] https://github.com/lcnetdev/profile-edit

On 11/12/15 1:23 PM, RDF Data Shapes Working Group Issue Tracker wrote:
> shapes-ISSUE-113 (SHACL and user interfaces):  [SHACL Spec]
>
> http://www.w3.org/2014/data-shapes/track/issues/113
>
> Raised by: Peter Patel-Schneider
> On product: SHACL Spec
>
> The WG charter includes the goal of "Human and machine interpretation of shapes to [...] develop user interfaces."
>
> SHACL includes shapes and constraints.  Most constraints are expected to be property or inverse property constraints.
>
> These SHACL features provide a backbone for the development of user interfaces related to shapes.   UI tools can, for example, use property and inverse property constraints to determine which properties should be part of an input form to create data that conforms to a shape.  Because shapes and contstraints are nodes in RDF graphs they can have extra information associated with them that can be exploited by user interface tools.
>
>
> PROPOSAL:  As the RDF Data Shapes working group does not have sufficient expertise to create a good set of features for UI creation it should stop at providing this backbone and let those who build user interfaces design the information needed for connecting SHACL shapes and constraints to UI tools.   To conform with this sentiment, sh:defaultValue will be removed from the SHACL vocabulary.
>
>
>
>
>

-- 
Karen Coyle
kcoyle@kcoyle.net http://kcoyle.net
m: 1-510-435-8234
skype: kcoylenet/+1-510-984-3600

Received on Friday, 13 November 2015 20:52:54 UTC