Call four comments 4 is out (Was: [whatwg] Web Forms 2.0 submission to W3C)

Wed Apr 13 04:58:55 PDT 2005

On Wed, 13 Apr 2005, Dean Jackson wrote:
> > 
> > That's never going to happen, just like the XHTML working group has 
> > never published a document with use cases for all their features. 
> > Ditto the SVG group,
> 
> The SVG group has published requirements documents for its features. So 
> has the CDF Group, which you participate in.

Requirements documents, sure. The WF2 spec has a whole section on 
requirements, and the WHATWG was based on a long document listing 
requirements and design criteria:

   http://www.w3.org/2004/04/webapps-cdf-ws/papers/opera.html

There's a far cry from that, and listing use cases for every feature. :-)

> One reason why I ask for this is that I've been talking to HTML 
> developers about the WHATWG work and WebForms 2.0. I ask them what they 
> think WHAT are doing. In most cases, they're not sure. Sometimes they've 
> read some press articles and think that it's two browser vendors at war 
> with W3C. If they do answer something like "WHAT are improving HTML 
> forms" I then ask them if they know what improvements are being made. 
> Honestly, I've never met anyone outside the subscribers to this list 
> that really know what you're doing. This isn't meant to be an insult. 
> It's just that I've been looking for average people so that they can 
> educate me on why they need WF2.

I'm not surprised. There's been no effort at making WHATWG a widely known 
effort -- it's only really been mentioned in the circles that people who 
care about standards development go in.

Ask random authors what the CSS, XHTML, or SVG working groups are doing 
and you'll get much the same replies: some people will say they're 
improving CSS, XHTML, or SVG (without knowing any more), others will say 
that there are turf wars being fought or that the groups are wasting their 
time. I don't think that's surprising. Few people have the time to really 
keep track of standards development.

> > Most of the features have quite obvious use cases -- for example the 
> > use case for the first feature in the Web Apps draft -- <html> -- is 
> > having a predictable root element for the document or document 
> > fragment.
> 
> OK, so why is this important? Why are the alternatives so bad?
> 
> I'm not talking about the <html> feature, I'm talking about you 
> empathising with the reader. Remember that you might be making something 
> that will last a long time. It's quite a responsibility.

If there are specific features that need better examples or explanatory 
text, I'm all for adding it. (Olav just sent such a list of WF2, and I'm 
going through it and adding text to WF2 to answer his questions.) But 
there is no chance that we're going to pre-emptively list the raison 
d'etre of every feature in the spec. That would take years. Many of the 
features date back to Tim's drafts in 1990. If we've lasted 15 years 
without use cases I'm sure we can continue a few more. :-)

> > I can maybe find the time to produce a document summarising the use 
> > cases for the less obvious features (probably by simply copying the 
> > text from e-mails in the archives of this mailing list, where the 
> > features mostly originated), but I don't want to waste time doing so 
> > for dozens of features where the use cases are obvious and nobody 
> > disagrees.
> 
> That's fair enough. It's your choice. You're a browser developer, so 
> you'll know how to implement it, and why you're doing it.
> 
> As a reader, I wanted to know.

I honestly don't believe that you want to read of the use case for the 
<html> element, the <head> element, the <title> element, and so forth, 
beyond the descriptions in the spec already.

But if you do, where is the equivalent for SVG? :-)

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'