Re: [sdw] ssn implementations need for rec track/standard

Kerry, 

No evidence of implementation is required for an OGC standard. We do have a standards track that includes a "standard with compliance suite" that is more mature than a "standard," but that requires an OGC maintained compliance test suite in addition to reference implementations. 

Best regards, 
Scott 


From: "Kerry Taylor" <kerry.taylor@anu.edu.au> 
To: "SDW WG Public List" <public-sdw-wg@w3.org> 
Cc: "Scott Simmons" <ssimmons@opengeospatial.org> 
Sent: Wednesday, September 28, 2016 6:40:04 AM 
Subject: [sdw] ssn implementations need for rec track/standard 



As discussed at the Lisbon F2F and again at in the ssn meeting this morning (or yesterday for many, see minutes https://www.w3.org/2016/08/23-sdwssn-minutes ), 

we require proof of implementations for W3C Rec track deliverables. This is defined in the W3C process here https://www.w3.org/2015/Process-20150901/#rec-advance . 



Phila has advised that there is no more formal definition of what an implementation means for a vocabulary deliverable. However, he points us to these two successful precedents 



(1) The implementation report for Prov is at https://www.w3.org/TR/2013/NOTE-prov-implementations-20130430/ - and is very impressive. It's more than is required but they had a whole stack to prove, we just have a vocabulary. 



(2) The implementation report for the ORG ontology might be a better example, see https://www.w3.org/2011/gld/wiki/ORG_Implementations . 



In my opinion, we can use the latter style. I know for the former this was put together through a survey call (I was one of the implementation contributors). In our case I expect we would be smarter to do an analysis ourselves of public material, starting now, and then work harder to source (through a survey or through explicit soliciting for implementations) for the rest – ie anything that we cannot find in the published material plus anything we might add to ssn here. I suggest this strategy because I believe we can get a very long way, faster, sooner, by analysing public material (and this could save us a lot of time at the CandidateRec -> Rec stage later on) . However, * if * we agree with this approach, we also need a volunteer to take on the task. This might be a particularly attractive task to someone fairly new to ssn who wants to see how it is used. 



Is this a reasonable approach? Will you volunteer? 



@Scott – by this asking kindly whether there may be some OGC principles for implementation of standards that we additionally need to satisfy. 



--Kerry