- From: Wayne Carr <wayne.carr@linux.intel.com>
- Date: Thu, 17 Sep 2015 11:22:59 -0700
- To: W3C Process Community Group <public-w3process@w3.org>
- Message-ID: <55FB0503.9000401@linux.intel.com>
There always are discussions about how specific scope or deliverables should be. Some of the requirements met by the combination of the scope and deliverables in proposed Charters are: 1. W3C Membership (through their AC reps) and the Director can evaluate whether the WG would be useful and whether it is focused enough to achieve objectives, etc.. 2. Possible participants can evaluate what types of patents may be required to be licensed (patent policy is licensing all essential claims in all RECs, not just one's own contributions like in a CG - so have to know what you're getting into and also to be able to estimate whether an RF spec would be practical -- so not just anything related to sensors or security or payments or video) 3. Community can provide feedback on what the WG should be chartered to do to meet community needs when Charters are public as they are being developed. There are two approaches to describing scope and deliverables that provide those: 1. Scope is very clear and narrow so that it is clear exactly what types of specs could be written and what types of technologies would be needed. Deliverables can then be vague or not exhaustive (it can be examples). OR 2. Scope can be broad and vague, but deliverables list is exhaustive and carefully described so can tell what types of technologies are needed. Deliverables generally aren't treated as exactly what spec is produced. Names of specs are changed. Specs are modularized or combined. Instead of a "deliverable" that could equivalently be described as a carefully defined item in the scope. A possible model approach for defining charter scope and deliverables: 1. Broad WG Charter Scope describing the broad bounds of what the WG is created for, beyond what it will actually do under the charter. (WGs often lay out their wider claim to what they eventually will be about - which makes scope sections not as useful for defining what they actually will do and puts more emphasis on deliverables.). 2. Restricted Scope Section. Section that says the overall scope will be restricted in the Charter period to a carefully described and more limited set of restricted scope items. Links to incubator specs, use cases, or achitecture documents would all be good to include in the description of a restricted scope item. If a scope restriction item won't have all deliverables listed below, it should say so. In that case, the description of the restricted scope item should be so clear that the description of deliverables in the Charter isn't needed. e.g. that the restricted scope item description is as detailed and clear as a deliverable entry would have been without having to pretend its known exactly what specs are needed (while knowing enough to scope out the likely area for patents). 3. Deliverables. Planned specs. Each deliverable spec should explicitly link to a restricted scope item. As above, restricted scope items can indicate they can add deliverables outside the Charter that fall within that restricted scope. Starting work on a spec. 1. It should be possible to appeal to the Director if a W3C member thinks a spec is out of scope. 2. A Charter could require incubator specs from CGs or other external sources before starting work. The WG does not need to use that input spec, but it should be clear what the spec is trying to accomplish is well enough understood to write a spec. Alternately, a Charter could require an architecture document that describes how major challenges in the spec can be handled. 3. WGs can explore work outside their Charter by suggesting that a CG take it up. (WGs could also have their own CG where they control the charter of that CG and that CG could do exploratory work beyond the Charter).
Received on Thursday, 17 September 2015 18:23:29 UTC