Re: creating a formal definitions portion for the SHACL WD

Hash: SHA256

On 05/28/2015 08:53 PM, Holger Knublauch wrote:
> (Moved to WG mailing list, from public list)
(Oops, my bad.)

> The current organization is the result of discussions with WG members
> who suggested a separation into Core and Advanced parts. This has lead to
> a separation of aspects that - from a logical point of view - would
> belong together. Technically it would be possible to start the document
> with Templates, then formalize the algorithms and then simply list the
> various built-in templates at the end.

This would certainly make for a better definition document.

> At the same time, an average reader would likely prefer the current 
> separation. I believe some people feel very strongly about the
> separation between core and advanced parts, so it may be difficult to
> reopen that can of worms.

The current document may be better for the average reader, but I believe
that the working group needs a document that defines SHACL, and the current
document is not very good at doing that.

> Having said this, I welcome other suggestions for the structure, e.g. as
> a branch of the document on github. Like with many other topics, we now
> have a specific starting point, and if anyone has better suggestions then
> they should IMHO make similarly specific suggestions.

The issue is how much work needs to be done before something can be put
before the working group.  I am unwilling to fork the document without some
indication that that direction I take has some decent chance of being accepted

That said, I'll put together a more detailed proposal.

> Holger


> On 5/29/15 8:29 AM, Peter F. Patel-Schneider wrote: I'm finding it rather
> hard to tease out the formal aspects of SHACL from 
> I propose that the formal aspects of SHACL be consolidated into one or
> two sections of the document to make it clearer what constructs are
> allowed and what they mean.
> peter
Version: GnuPG v2


Received on Friday, 29 May 2015 13:26:00 UTC