RE: View Source

> -----Original Message-----
> From: public-html-request@w3.org [mailto:public-html-request@w3.org] On
> Behalf Of Doug Schepers
> Sent: Tuesday, March 17, 2009 1:03 AM
> To: public-html@w3.org
> Cc: www-svg@w3.org
> Subject: Re: View Source
> 
> Hi, Justin-
> 
> Justin James wrote (on 3/17/09 12:34 AM):
> >>
> >> Rick wrote (on 3/16/09 9:31 PM):
> >>>
> >>> I totally agree with the sentiment that the spec should not
> >>> dictate UI requirements.
> >>
> >> Why?
> >>
> >> I don't agree with it, as a developer, because I want to have
> >> certain constants in my application development environment that go
> >> beyond rendering the document.  This is a big advantage that
> >> environments like Air and Silverlight have.
> >
> > The HTML spec is NOT a "web browser" spec.
> 
> Sure it is.  Web browsers are one class of UAs that the HTML5 spec is
> targeting.

No, it is most definitely not. Outside of items around the DOM (for example events), the HTML spec is incredibly "hands off" in terms of defining user interaction. Please note that the HTML spec *doesn't even define default presentation*. The HTML spec is anything *but* a "Web browser" spec. It is an "HTML parser" spec, maybe.

> > browsers are free to innovate and invent above and beyond
> >  how they handle HTML
> 
> Of course.  Specifying a minimal feature set does not set an upper
> bound
> on other features, or on innovation.

Yes, and the minimal feature set that HTML requires is the ability to parse HTML. That is *it*. No where in the HTML spec does it say "the user must be allowed to access the HTML source of the document currently being view", let along if it will be the original source or a sanitized version (keeping things within the context brought up). Personally, I am all on favor of Web browsers having this feature, by the way, particularly if it is optional ("View Original" and "View Sanitized", for example). But this is completely outside of the scope of the HTML spec.

> > How do we know that the "web browser" will have
> > access to a right mouse button? Or even have an input device with an
> > analogue to a mouse? What if we discuss screen resolutions, and in 5
> > years, 24" screens are considered "small"? Etc. etc. etc. What about
> > mobile devices, which have amazing variety in input methods?
> 
> All of these things have specs.  If some of the assumptions change,
> then
> new specs are written.  This is how we adapt interoperably.

Yes, and the HTML 5 spec is currently slated to reach "recommendation" status in 2022. If many of the UI and IO assumptions that we could make today would still be valid in 2022, I would be surprised.

> We live in exciting times!

Yes, we do! Which is precisely why HTML must play to the lowest common denominator. Trying to define Web browser behavior is not part of that. Think "C and the thermostats". :)

J.Ja

Received on Tuesday, 17 March 2009 05:16:35 UTC