W3C home > Mailing lists > Public > whatwg@whatwg.org > February 2015

Re: [whatwg] Obsolete Feature [hgroup]

From: Jukka K. Korpela <jkorpela@cs.tut.fi>
Date: Thu, 19 Feb 2015 00:07:31 +0200
Message-ID: <54E50D23.3060101@cs.tut.fi>
To: whatwg@lists.whatwg.org
2015-02-18, 22:55, Ian Hickson wrote:

> If you wish to create a valid and semantically correct document according
> to the most recently published HTML standard, then using <hgroup> is fine.

I think you meant to say that <hgroup> is in the WHATWG document at 
present (and you added that you do not plan to remove it from there).

There is strictly speaking only one HTML standard, namely ISO/IEC 15445. 
Few people know about it, still fewer care about it. (It’s just HTML 
4.01 written in more standardese.)

The W3C calls their recommendations “standards”. (I think this is a 
rather recent change.)

The IETF calls some of their documents “Internet Standards”, which is 
somewhat odd, but irrelevant to HTML, as IETF abandoned work on HTML 
after HTML 2.0.

The WHATWG calls their documents “living standards”, which is really an 

Everyone could write a specification, with or without his brother, and 
call it a “standard”. Actually, you can never know which one is the most 
recent “standard” in that sense.

So what really matters is that there are different players in the fields 
of defining HTML, and there is no authority to rule them all. Every 
author/developer/designer/whatever needs to decide how to deal with 
this. Asking one of the players about the rules will most probably yield 
just the obvious answer.

> I recommend using the validator.nu service rather than the W3C one.
> They're basically the same but the validator.nu one is closer to the
> WHATWG spec's requirements than the W3C one.

The relationship between the two is somewhat obscure, but it is in that 
direction. What the different validators check against is really their 
authors’ idea of what is correct HTML, so we have yet another 
“standard”, defined very implicitly, and mutable. But it indeed appears 
to be closer to WHATWG HTML than to W3C HTML5.

The bottom line is that validators are just useful tools, at best. All 
HTML5 validators are experimental software that checks against some 
rules that have not been disclosed in detail but are supposed to match 
some idea of “HTML5”.

Received on Wednesday, 18 February 2015 22:07:59 UTC

This archive was generated by hypermail 2.4.0 : Wednesday, 22 January 2020 17:00:27 UTC