[html5] WHATWG's lack of responsiveness (was re: Suggestion to close this list)

Domenic Denicola d at domenic.me
Wed Aug 23 13:59:04 PDT 2017

Hi Stanton,

You may be interested in the history here. W3C “HTML5” is actually forked from the WHATWG HTML Standard; your claim that the W3C version is in some way authoritative or “full”, or that the WHATWG is redefining things, doesn’t really fit with the facts. See things like https://annevankesteren.nl/2016/01/film-at-11 or https://www.reddit.com/r/javascript/comments/5swe9b/what_is_the_difference_between_the_w3c_and_the/ . In fact, if you read up, you’ll see that W3C fork is not only based on the HTML Standard, but periodically copies and pastes from it, in precisely the reverse of your suggestion.

In particular, I don’t know what this “cheatsheet” is that you reference; we certainly don’t write the specification based on it, and I don't think it's fair to call us "unresponsive" if we don't update it, given that it's an unauthoritative source that we don't control or consult. In contrast, we're pretty responsive to actual bugs reported against the content of the spec; I'd encourage you to browse through https://github.com/whatwg/html/pulse to see the work we're doing on a daily basis. I'd love to hear about any particular GitHub issues you found unaddressed with "repeated reminders", so we can fix that; we try to be as responsive as possible.

It’s true that certain W3C editors have, over the years, redefined certain elements in ways that match how they like to write their personal documents. I think it’s best to treat this just as if any other person had decided to redefine an element for their own usage, e.g. how Twitter Bootstrap redefined <i> to mean “icon”. It’s against the HTML Standard, and writing it down in a forked document and being appointed “editor” by the forking organization doesn’t really make that redefinition authoritative or “correct”. (Neither does “real-world usage”; again, if that was the criteria, we would redefine <i> to mean icon, given how many pages use it that way.)

I don’t really intend to get into a debate here, as you seem to have some strongly felt opinions, and it is somewhat off-topic, but I think it’s important for you and the others on the list to be aware of the history and facts here.

(Apologies if you receive this twice; the first version was in HTML format and was too big for the list.)


From: Help [mailto:help-bounces at lists.whatwg.org] On Behalf Of Stanton McCandlish
Sent: Wednesday, August 23, 2017 15:40
To: help at lists.whatwg.org
Subject: [html5] WHATWG's lack of responsiveness (was re: Suggestion to close this list)

> I sometimes worry that traffic has fallen away as new web developers don't really care about standards!

It may also have something to do with WHATWG's lack of responsiveness to serious problems that seem to indicate WHATWG doesn't care about standards other than their own derivative one, so people are turning back to W3C (and to less authoritative but more current sources like W3Schools and HTML5Doctor).  I reported over two years ago (and more than once, to multiple WHATWG people, and on your wiki, and on the list) that your re-definition of the cite element is demonstrably incorrect, with multiple direct contradictions of the actual W3C HTML5 specification, yet you still have not fixed it.

W3C (and real-world usage): "The cite element represents a reference to a creative work. It must include the title of the work or the name of the author (person, people or organization) or an URL reference, which may be in an abbreviated form as per the conventions used for the addition of citation metadata.", followed by examples showing these differing types of usage.

WHATWG: "The cite element represents the title of a work .... A person's name is not the title of a work ... and the element must therefore not be used to mark up people's names.", followed by examples that basically claim W3C is wrong.

As an example of what I mean by "responsiveness", I noted that the wording in the W3C Cheatsheet didn't match what the actual HTML5 spec said either, and reported this to someone at W3C.  They fixed it the very same day.

This is also not the only such error in WHATWG's version.  The problem appears to be that WHATWG's version is based on the Cheatsheet and the Cheatsheet in turn is based on the actual HTML5 spec.  This is a maintenance chain that requires attention from multiple parties.  W3C occasionally doesn't perform that maintenance until joggled by someone like me to do so, and WHATWG doesn't appear to do it at all, even when repeatedly reminded.

The sensible thing to do would be for WHATWG's version, like the Cheatsheet, to be based directly on the full W3C HTML5 spec, to cut the maintenance chain in half, and then actually do the maintenance.  HTML5 has not remained static since its initial drafting; various bits of it have changed in response to public (mostly developer) feedback, and WHATWG is not keeping up.  This should be easy; W3C's HTML5 was last changed 28 October 2014, so it's not like it's a firehose of changes.

Back to the original topic:

I tend to agree that maintaining a mailing list is important, for most of the reasons given by others. 

> +1 for a web based alternative, something I can point at my (RSS) feed reader at

That's easily provided by a list-to-Web-archive script of some sort, and various free-software ones are available.  Some are even two-way gateways, permitting posting to the list via the Web as well.

More information about the Help mailing list