RE: Comment tracking for navigation-timing CR [Was: Re: publishing new WD of URL spec]

> -----Original Message-----
> From: Robin Berjon [mailto:robin@w3.org]
> Sent: Friday, September 12, 2014 2:39 PM
> To: Boris Zbarsky; Jeff Jaffe; Stephen Zilles; David (Standards) Singer
> Cc: Arthur Barstow; Philippe Le Hegaret; public-w3process@w3.org
> Subject: Re: Comment tracking for navigation-timing CR [Was: Re: publishing
> new WD of URL spec]
> 
> On 12/09/2014 15:39 , Boris Zbarsky wrote:
> > On 9/12/14, 8:48 AM, Jeff Jaffe wrote:
> >> Thanks Boris.  So the process issue that we should work on is:
> >>
> >> How to motivate the W3C Community and WGs to expeditiously and
> >> consistently issue errata?  Correct?
> >
> > That would be a good issue to work on, yes.
> 
> I heartily agree with Boris but I would like to strengthen the idea further.
> 
> It is understandable that the first version of a given specification may require
> time because it often requires banging heads together a fair bit and trying out
> ideas. However, once there is a sufficiently stable release to start from, it ought
> to be possible to produce the following iterations in relatively short order,
> including both errata and progressive addition of features.
> 
> For sufficiently well established parts of the platform, we ought to be able to
> issue yearly (or even less) recommendations that simply include the parts that
> the WG has been working on that happen to be interoperably implemented at
> that time. It would be better than having to wait n years, having to maintain
> errata in addition to the improved document, and it would help avoid second-
> system syndrome.
[SZ] Iterating on specifications, actually separate modules, is what the CSS group has been doing since the publication of CSS 2.1. The process is still not moving with yearly updates for, IMO, three reasons. Getting implementations is not always easy, when implementations are being created there is the discovery of edge cases and ambiguities that drive further spec work and, finally, getting a test suite that shows interoperability seems to take forever. That said, I think your statement of the goals is correct. CSS adjusts is module versions to reflect what has been interoperably implemented and pushes that to REC.

Steve Z
> 
> --
> Robin Berjon - http://berjon.com/ - @robinberjon

Received on Friday, 12 September 2014 23:10:22 UTC