RE: Proposed standards evolutionary lifecycle

The people who have the best ideas are not especially the best ones to write a spec. I’m not sure there exists an ‘ideal’ way to ceate specs: the web needs everybody and what he can do. The important thing is that there should be a place where the community can bring new & innovative ideas and get help in standardizing them.

 

However, I agree that the ideal standardization process involves both a working draft and a forward polyfill. Yet, if someone can bring only one or the other, this is already a good step towards standardization.

 

 

 

 


De : Brian Kardell
Envoyé : ‎15‎ ‎décembre‎ ‎2012 ‎05‎:‎54
À : public-nextweb@w3.org
Objet : Proposed standards evolutionary lifecycle


How things go from idea to standard/rec in my ideal world.. Clint and I were discussing this yesterday and it has apparently never been articulated by me, so I'd like to share it here...



* someone has an idea




* they create a proposal and send it to the list for early comment




* if they are still interested, they turn that into an unofficial draft (there are templates to unofficial drafts that look like this http://fremycompany.com/TR/2012/ED-css-custom/) and send it to the list - ideally along with a draft, but I suppose that some may post drafts hoping someone else will implement.




* Once we have a versioned implementation and draft - that's a prollyfill. It gets an initial review, gets added to the list..




*  It may change, spawn forks and variants, etc.  It might be helpful to encourage that these be referenced right off the site during development or something (we can place a CDN like cloudflare in front of it) to allow us to get some idea about how any people are using..




*  It gets tested heavily - algorithms worked out, etc with test suites.




* If it gets mature and seems to have a lot of support/traction we submit it for consideration and lobby for it.




* It gets a legitimate draft in the appropriate body (most often that would be W3C I think, but ECMA might be another)





* At this point, if we have been able to keep W3C members reasonably up to date, the normal process to real standard and implementation should be simple, relatively uneventful and fast.



-- 
Brian Kardell :: @briankardell :: hitchjs.com