Re: Eric's description of core SHACL

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

The algebraic semantics for shape expressions has, I think, a less dramatic
change between open and closed shapes, but that's not the point.

The point is that there can be a large difference, and that picking an
approach without information on this change is probably not a good idea.

peter


On 03/17/2015 09:25 AM, Eric Prud'hommeaux wrote:
> * Peter F. Patel-Schneider <pfpschneider@gmail.com> [2015-03-17
> 05:44-0700] My takeaway from Eric's description at 
> http://w3c.github.io/data-shapes/semantics/ is that 1/ informal
> descriptions can leave quite a bit open (e.g., how recursive shapes are
> handled) and 2/ going beyond the core (for example to closed shapes) can
> require significantly different machinery.
> 
>> Do you know of some specification or implementation strategy in which
>> the difference between e.g. open and closed schema or single and
>> multi-occurance is less dramatic? It's more terse in formal notation
>> <http://w3c.github.io/data-shapes/semantics/Axiomatic>, but one still
>> has to change function prototypes and impose extra logic, no?
> 
> 
> peter
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iQEcBAEBAgAGBQJVCJi0AAoJECjN6+QThfjzeHwH/3tDkQqkET2Yk+iFcgvlF87+
sVzNk5D+drcpbmZ2EgiKGtPRKMfEy8jxThhKgelukTaUhky306seDlWPZUyP2cxP
mbAdU0ObPNPDyMJMD4iOz7B78KKVDsBHYZyICgMYx4R2VtWvMK6JIurha0RWIvYg
RYlhIG8wXe9mbNxn3a860CJEtDAANGrdEnPpm35FlbJDNPQkSUmm3TqY3YMYfmI+
1pZ2TnW1J6h0wUt2vpxvIOKdSE1FckvZwAqAUXPeywFptNMOh8zGLrz5qkBI3I/l
CjDG+zNnrIp+/hXIsK4uhoWMbaDwn8n55wqTD7VmeHGlUhlLK9D1qj3Ny+76yEI=
=Ymbl
-----END PGP SIGNATURE-----

Received on Tuesday, 17 March 2015 21:12:50 UTC