intertwingly

It’s just data

WHATWG FAQ


The WHATWG has a FAQ.  Here are some thoughts on the points made there.

But first it is worth noting that I am a member of the WHATWG.  I have been for years.  I certainly can go and update the wiki directly.  In a number of places what is written there does not reflect my thinking on this matter.

It occurs to me that it must reflect the thinking of some.  Perhaps it doesn't reflect the opinion of others in the group?  As this wiki is not operating under the burden of rules like [citation needed] it is difficult to determine exactly whose opinion this page does reflect.

How does the WHATWG work?

I question the presumption implicit in the notions of “the” editor, and “the” spec.  I reluctantly accept the notion that any individual spec development process need not employ processes requiring consensus or voting, but I reject any implication, however subtle, of inevitability or entitlement.

Simply put, there needs to be a recourse if a person or a group disagrees with a decision made by the editor of the WHATWG document.  That recourse is forking.

I realize that that is a very high bar, and will say that is intentionally so.  Simply put, specs don't write themselves... I don't care how good you think your idea is, either you need to step up and directly write the spec text yourself, or accept that you need to be persuasive.

Note: the granularity of this activity need not be at the entire spec level.  Within the AtomPub working group, we created a culture where camera ready spec text was the gold standard for proposed changes.  While there never is any guarantee that such proposals will get adopted or that they won't get significantly modified, concrete spec text tends to focus the discussion on tangible and real issues.

A corollary to this is that people who desire to contribute need not limit themselves to any notions by any other individual as to how specs should be split out.

I will acknowledge up front that such an approach will create confusion in the short term.  However, until we collectively can develop a spec development process based on establishing and continuously maintaining consensus, the only remaining alternative is to defer the evaluation of consensus until the work product is produced.  Furthermore, note that none of this prevents any particular editor from practicing consensus.  In fact, such an approach would be preferred, at least by this co-chair.

How should browser developers interact with the WHATWG?

While the WHATWG was founded by browser vendors, and browsers do have a special and prominent place in the development of the web, this position too is not an exclusive one.  There is nothing in the W3C charter that says that non-browser user agents need to be excluded.  However, this is moderated by the point made above: any progress along these lines needs to be pursued either by spec writing or by persuasion.

Related reading: application neutrality.

When will HTML5 be finished?

For starters, I want to operate under the assumption that there will be an HTML6, and hopefully even an HTML7, HTML8, and HTML9.  Hopefully they will be as unversioned as HTML5 is in that using a single HTML8 parser will be the best way to consume an HTML4 and HTML6 document alike, but that’s not something that needs to be decided at the moment.

Meanwhile, I would like to see a W3C Candidate Recommendation in 2009, even if it means significantly jetisonning function.  That can and should be followed up in 2012 with an HTML6 release with more function.

Notes: again this is predicated on there being sufficient volunteers to do the work.  With sufficient volunteers, work on 2009 and 2012 recommendations can proceed in parallel.

I also recognize that both dates may slip slightly, so 2010 and 2013 are possibilities.  I also would not be opposed to recommendations being issued in 2010 and 2012.

What about Microsoft and Internet Explorer?

While I fully understand and appreciate that no browser vendor is in a position to make absolute commitments as to when future releases will occur and the functionality provided in such releases, the rather limited amount of feedback that Microsoft has provided to the W3C drafts produced to date is of significant concern to me.  Hopefully such feedback will be forthcoming in “camera ready copy” format.

Will (X)HTML 5 finally put an end to the XHTML as text/html debate?

With the full benefit of 20-20 hindsight, I agree that the fact that XHTML 1.0 provides for the serving XHTML with the mime type of text/html is unfortunate.  Beyond that, the “debate” is largely a semantics one.  What’s more important, and largely unaddressed, is the fact that the domain of documents which can be serialized as HTML5 and as XHTML5 do not observe a proper super/subset relationship.

What will the DOCTYPE be?

I disagree with “not recommended”.  It would be unfortunate if a conformance checker were to flag such occurences.  I suggest that this be recast as

In XHTML: no DOCTYPE is required, but if present must be the same as in HTML.  Note that case is significant in XHTML.

How are pre-HTML5 documents parsed?

I strongly endorse this.

What is an HTML Serialisation?

At this point, it is inspired both by SGML, XML, and by reality.

Based on a number of comments by a number of participants, I sense an unhealthy desire on their part to distance this effort from XML.  Some appear to go further and appear to want to actively crush XML.  I'd suggest a kinder, gentler approach: share credit copiously, and win gracefully.

Should I close empty elements with /> or >?

Trailing slashes also play a role in foreign markup.  And I would again suggest dropping the recommendation.

If I’m careful with the syntax I use in my HTML document, can I process it with an XML parser?

“No” as an absolute is clearly false.  It may be difficult, and may only apply to a subset of potential documents (i.e., inline scripts beyond a certain level of complexity would effectively be precluded), but if you are careful (which, after all, is how the question is phrased), it most certainly is possible.

Will there be support for namespaces in HTML?

With the inclusion of MathML elements, this answer needs to be updated.

Which group has authority in the event of a dispute?

The previous comments on inevitability and entitlement apply here.

To be clear, the WHATWG has done the W3C, the browser vendors, and the web in general a great service.  I hope and expect that they will continue doing exactly that — not that anybody needs my encouragement one bit, as this work is clearly going to proceed with or without my endorsement.  But they/we have it anyway.