- From: Dominique Hazael-Massieux <dom@w3.org>
- Date: Wed, 22 Aug 2012 13:58:57 +0200
- To: Harald Alvestrand <harald@alvestrand.no>
- Cc: Jim Barnett <Jim.Barnett@genesyslab.com>, public-media-capture@w3.org
Le mardi 21 août 2012 à 17:38 +0200, Harald Alvestrand a écrit : > If people are happy with that, we could. If others who watch us from > outside have dependencies on us getting past certain stages, or if > others don't believe anything is stable until it's passed some stage, I > assume they will raise their voices and tell us. > > Dom might want to tell us whether there are IPR implications of not > going through the CR/and so on stages. The only stage that freeze the RF licensing commitments is when a spec reaches Recommendation (which to me at least is a good motivation to get small set to Rec early rather than to wait to a full spec many years later). The impact of other stages on the patent policy are as follow: * each Last Call starts a new call for exclusion, under which any feature that wasn't available in the previous last call can be excluded by a WG participant * you can only go to Proposed Recommendation if there is no open call for exclusion (which means there needs to be at least 60 days between the last last call and moving to Proposed Recommendation) * the minimal time between a first public working draft and a proposed recommendation is 150 days (due to the first window required for the first call for exclusions) My separate perspective on the "long time to go through the Recommendation track" is that the time needs grows non-linearly with the size of the spec (due to the rule that to go out of Candidate Recommendation, you need to create test cases and show interoperability for each feature — and creating test cases and showing interoperability becomes harder as you add new edge-cases features). Also, WebIDL/JavaScript are sufficiently flexible that in many cases, you can reasonably easily create modules (e.g. partial interfaces, partial dictionaries) which allow to complete a particular aspect of a spec. Dom
Received on Wednesday, 22 August 2012 11:59:14 UTC