W3C home > Mailing lists > Public > public-webplatform@w3.org > February 2013

RE: How should we represent the status of a spec?

From: Michael Champion (MS OPEN TECH) <Michael.Champion@microsoft.com>
Date: Fri, 15 Feb 2013 17:18:29 +0000
To: Tobie Langel <tobie@fb.com>, Robin Berjon <robin@w3.org>
CC: Julee <julee@adobe.com>, "public-webplatform@w3.org" <public-webplatform@w3.org>
Message-ID: <b5a534f88da04022a275c262ae63a265@SN2PR03MB096.namprd03.prod.outlook.com>

+1 to Tobie's points, with some caveats:

There are at least two types of status different audiences need to track 
a) Status of the spec as it moves toward standardization, and how well tests against the *spec* indicate that it is clear enough to be independently implemented

b) Status of the implementations of the spec -- what percentage of deployed platforms claim to implement it, and how interoperable those implementations are in practice.

To put it another way:  there are multiple ways to get interoperability -- write a spec that is so clear that everyone can implement it without looking at other implementations, or to start from the code (or reverse engineer the bits on the wire) of some de facto reference implementation.  W3C's testing  focuses on the first, but the real-world interoperability of technologies that aren't yet standardized often comes from a combination of specs, shared code, and reverse engineering.  End users who just want to know "can I use this feature with confidence" don't particularly care how the feature got to be interoperable, and they don't particularly care if a spec is good enough if it's not widely implemented in the real world.

The infrastructure for storing/publishing the two types of data can almost certainly be shared, but there may be significant differences in how the data are collected. In either case, we should TRY to minimize the number of judgment calls / political decisions between the raw facts and the presented data. 

From: Tobie Langel
Sent: Friday, February 15, 2013 5:01 AM
To: Robin Berjon
Cc: Julee; public-webplatform@w3.org
Subject: Re: How should we represent the status of a spec?

On Friday, February 15, 2013 at 1:54 PM, Robin Berjon wrote:
> On 15/02/2013 10:00 , Tobie Langel wrote:
> > Maybe we should start having a cross-project conversation on the subject.
> I'm all for that, but I want to make sure that we don't force ourselves
> to shoehorn all the various ways of representing spec status into a
> single unified one if it turns out not to fit.

Absolutely. There are different audiences are interested in different types of status (e.g. publication status, test status, implementation status, etc).

Of course there is some overlap.

However--and that's the part I'd like us to focus on--storing, accessing, updating this information requires the same infrastructure, and it would be a shame to not consider centralization, here.

Received on Friday, 15 February 2013 17:21:44 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:20:47 UTC