W3C home > Mailing lists > Public > public-html@w3.org > July 2015

Re: Proposed split of the HTML specification

From: Robin Berjon <robin@w3.org>
Date: Mon, 06 Jul 2015 17:26:24 +0200
Message-ID: <559A9E20.30702@w3.org>
To: Adrian Bateman <adrianba@microsoft.com>
CC: "HTML WG (public-html@w3.org)" <public-html@w3.org>
On 01/07/2015 22:54 , Adrian Bateman wrote:
> The HTML 5.1 spec contains substantial new features in many sections.
> This makes it a poor choice for bug-fix maintenance to HTML 5.0. This
> Means our choice will be bug fix the monolithic spec or don't bother. With
> the interdependencies between sections we also likely won't be able
> to take the pieces to Rec if they depend on WD status parts that aren't
> aligned with HTML 5.0.
>
> If we produced a split spec from HTML 5.0 we could pretty much move
> to Rec straightaway and then never touch some finished pieces (many of the
> definitions for example). Using the 5.1 document we are going to need
> to better understand the interdependencies to have a chance at moving
> any part forward.

I can probably produce a split version of the 5.0 spec relatively 
easily, but before jumping in and doing so I would like to make sure it 
is actually useful.

In the 5.1 split we made, some of the big new features are contained in 
modules (e.g. sortable tables). Some of the modules have fixes, but not 
necessarily that big (e.g. the inert attribute). I'm not sure that for 
maintenance purposes the work is all that different.

The question is: if I produce a split 5.0 for maintenance purposes, I 
will need people to help work on the various modules to figure out which 
of those modules can just re-ship because they have non-substantial 
fixes, which have important fixes that justify shipping as a new edition 
or even the start of 5.1, etc. This is work that will require extensive 
implementer input — without that it's pretty much busy work.

Part of the problem is that the spec markup has changed *extensively* 
between 5.0 and 5.1, such that a direct HTML diff is impossible. It may, 
however, be possible to produce a text-based diff that would be 
meaningful. This could enable us to produce modules both for 5.1 and for 
5.0, and compare them. Then based on those diffs make decisions as to 
what to push forward, publish, etc.

If that sounds like a plan, and if there is genuine commitment to then 
push the work forward, I'm happy to look at generating a split version 
of 5.0 that more or less matches the 5.1 split (just with less stuff). 
But I doubt that I will have the time to go through the diffing and 
comparison process before my time with W3C is over, so we'd need someone 
to actually commit to that work.

-- 
Robin Berjon - http://berjon.com/ - @robinberjon
Received on Monday, 6 July 2015 15:25:27 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:16:44 UTC