Trig star

[gkellogg]

Adds a section on TriG-star similar to N-Quads-star.

It seems that most of the work is already done by Turtle-star, but the discussion could go into more detail about processing, if anything thinks it would be useful.

---

Preview | Diff

[afs]

I think the Turtle-star section should be first because it has discussion, with examples, and is the format used elsewhere for examples. While NT->TTL makes sense as a progress, there is value in having the "primary" format first.

The original Other Concrete Syntaxes was last in section. I prefer that rather than the intro header because (1) it can be marked "non-normative" (2) it is not output of this community work (3) room to add links (if JOSN-LD-star is ready in time) that are less appropriate in the 3. section intro.

[gkellogg]

I think the Turtle-star section should be first because it has discussion, with examples, and is the format used elsewhere for examples. While NT->TTL makes sense as a progress, there is value in having the "primary" format first.

I can see an argument for either, but it sounds like we need some consensus.

The original Other Concrete Syntaxes was last in section. I prefer that rather than the intro header because (1) it can be marked "non-normative" (2) it is not output of this community work (3) room to add links (if JOSN-LD-star is ready in time) that are less appropriate in the 3. section intro.

Again, I can go either way, but it seems that we need some introduction to the concrete syntax syntax section, and the introduction makes no normative statements, so it shouldn't be necessary to mark it as informative.

[afs]

Suggestion:

Add a section 3.5, add it to the list, and move the text "Other specifications may define RDF-star concrete syntaxes." to that section.

Then in 3/Intro with trailer text to SPARQL query and results.

"""This section covers changes to concrete syntaxes. <List of 5> including "other section"

Changes for SPARQL-star are given in [SPARQL-star grammar changes](#sparql-star-grammar) and the changes for the result set formats in the section [Query Result Formats](#query-result-formats). """

Aside: I have an extended TriX, reusing <triple> as an embedded triple term. Because RDF-star is about triple details, it might useful to have in an explicit triple-centric syntax.

[gkellogg]

Suggestion:

Add a section 3.5, add it to the list, and move the text "Other specifications may define RDF-star concrete syntaxes." to that section.

Then in 3/Intro with trailer text to SPARQL query and results.

"""This section covers changes to concrete syntaxes. <List of 5> including "other section"

Changes for SPARQL-star are given in [SPARQL-star grammar changes](#sparql-star-grammar) and the changes for the result set formats in the section [Query Result Formats](#query-result-formats). """

Done in 6f7c3df3.

Aside: I have an extended TriX, reusing <triple> as an embedded triple term. Because RDF-star is about triple details, it might useful to have in an explicit triple-centric syntax.

That makes sense, but what is the status of TriX? The link from https://www.w3.org/2004/03/trix/ to a Nokia page (http://sw.nokia.com/trix/TriX.html) is broken, ut HPL has the report here. Perhaps it should be re-published as a Note with RDF-star extensions?

[afs]

TriX digression:

I keep it around - a sense of keeping the system honest by having various community formats (I think it has one user).

I'm going from a later tech report https://www.hpl.hp.com/techreports/2004/HPL-2004-56.html without processing extensions.

The W3C DTD https://www.w3.org/2004/03/trix/trix-1/trix-1.0.dtd is slightly different with (<trix> not <TriX>) -- the DTD is not linked from https://www.w3.org/2004/03/trix/trix-1/ - link typo.

Trix-star is add 'triple' to subject and object positions of "ELEMENT triple".

Something for the record would be good but not as encouragement to use. It's rather verbose.

[TallTed]

Reviving a TriX (of whatever capitalization) Note/report that's been languishing in W3 space is probably doable (depending on some aspects of where/how it's been languishing), and should just need a quick review by W3 officialdom.

Taking (and updating) a later version from HPL is more questionable; IP agreements would need to be very carefully vetted, and probably require some attention from/by W3 Legal.

[afs]

@TallTed - "later" refers to a later tech report. Gregg referenced: HPL-2003-268 and later means HPL-2004-56.

The material on the W3C website corresponds to the 2004 version. This is mentioned on the W3C site. It has DTD and XML schema on the W3C site. It is not a team submission; it is community work of the Semantic Web Interest Group.

[TallTed]

@afs I cannot quickly locate anything regarding IP license. HPL-2004-56 says Copyright Hewlett-Packard Company 2004. SWIG participation would have required some IPR commitments, so the content of that w3 subsite should be fair game for incorporation into new work -- but HPL-2004-56 is not content of that w3 subsite, and is not labeled as output of SWIG, so HPL-2004-56 and related docs may not be fair game for incorporation into new work.

The "TriX homepage" to which the w3 subsite refers is no longer there, but the Internet Archive obligingly provides a capture from 2011 which shows it as "Copyright (C) 2003 Nokia".

We don't appear to have an official W3 staff contact, but perhaps our resident W3 Fellow could bring this up with the Team?

[afs]

The DTD and XML schema are on w3.org.

[gkellogg]

If you'd like to merge this, then let's fork off the section 3 intro and presentation part to a side issue and merge as-is. Approved.

I agree that discussion's gone on long enough, and there don't seem to be any concerns about the basics. I think the remaining issue is ordering of the serialization sections, but that can always be re-addressed. Besides, there is other work awaiting this PR. So, I'm going to go ahead and merge, and ask any that continue to have concerns to create new issues and/or PRs to address.