Re: An HTML language specification

On Sat, Nov 22, 2008 at 10:34 PM, Ian Hickson <ian@hixie.ch> wrote:
> On Sat, 22 Nov 2008, Jim Jewett wrote:

>> [...] if I as an author (or the writer of a simple
>> authoring tool) just want to embed a video with
>> default video look and feel, default controls, etc ...
>>  similar to what I get today with object ... I don't
>> see why I would need to know anything about
>> the HTMLMediaElement scripting API.

>> So leaving that in a separate processing spec
>> seems perfectly reasonable; the processing spec
>> would depend on the markup semantics
>> spec, but the dependency would be in only one
>> direction.

> This would be a very weird split ("vocabulary" vs
> "DOM and processing requirements"), with, to my
> knowledge, no precedent.

It isn't that far from the split betwen HTML and DOM
in previous specs.  I realize that you regard the split
as a problem there; my view is that the problem was
caused because the DOM/processing specs simply
weren't good enough.  (In fairness, the problems
there probably were in part insufficient integration;
or at least insufficient W3C documentation for the
non-document interfaces.)

> Why should authors that don't write script be
> somehow segregated from authors that do
> write script?

Because there are many authors who do not
write script, or who leave the scripting portion
to someone else.  What these authors need
is a strict subset of what the scripting authors
need, but it is already complicated enough
that there is value in keeping it as short as
possible.

> Why should implementors have to read two
> specifications to work out how to implement
> one feature?

Why should they have to read what ought to be
ten separate specifications to implement what
ought to be only one of the ten?

> What about authors who want to write documents
> with no images? Surely they don't need to know
> anything about <img>; does that mean we should
> have a spec without <img> too?

There were versions of the spec that excluded
"fancy" things like tables and forms.  There are
certainly versions that excluded frames.  You
still keep tables in a separate section, because
they are easier to understand that way.

Eliminating one or two elements won't make the
spec sufficiently shorter to be worth doing.

Eliminating all the scripting-only interfaces will.

-jJ

Received on Sunday, 23 November 2008 04:16:58 UTC