- From: John Arwe <johnarwe@us.ibm.com>
- Date: Fri, 10 Oct 2014 11:28:39 -0400
- To: James Leigh <james@3roundstones.com>, "public-ldp-wg@w3.org Group" <public-ldp-wg@w3.org>
- Message-ID: <OF4C63149B.A709ED0B-ON85257D6D.0053C068-85257D6D.005505AE@us.ibm.com>
Attempting to move this off of the comments list and onto the WG mailing list, given that the former has extra tracking hence is not intended for "long" discussions. public-comments is more for comments from those outside the WG, and where we try to get down to problem-solution-ok triples. We discuss proposed responses along with internal WG stuff (still public, just less tracked by automation for document transitions) on public-ldp (no -comments) and in meetings. Once we have a consensus proposal we can use public-comments. I don't think your proposal to list the prohibitions is a good idea, for several reasons: 1: It was never the intent IMO to *prohibit* all LDPC behaviors, it was the intent to say that the client guarantee is that it's an LDPR, and therefore (like any LDPR) it might have additional non-standard behaviors (including some LDPC behaviors). If we prohibit, we can't un-prohibit later and remain compatible. 2: Listing the prohibitions means recapping all LDPC MUSTs in a single place, inverting them, and keeping that up to date as the rest of the spec changes. People are horrible at making that model work over time. If a middle ground works for you, like listing all the current LDPC-Musts that client is not guaranteed to get with an LDPR interaction model *informatively* along with a statement saying that list might be incomplete, I might go for that - I do have sympathy for spec readers in addition to editors. Best Regards, John Voice US 845-435-9470 BluePages Cloud and Smarter Infrastructure OSLC Lead
Received on Friday, 10 October 2014 15:29:24 UTC