Re: An HTML language specification

On Sun, Nov 23, 2008 at 1:30 AM, Ian Hickson <ian@hixie.ch> wrote:
> On Sun, 23 Nov 2008, Jim Jewett wrote:
>> >
>> > Are you proposing splitting just the DOM, or the DOM and the
>> > implementation requirements?

>> I am not certain that I understand what you mean by "implementation
>> requirements."

> The rules that apply to implementations, like how to parse an HTML file,
> how to submit a form, how CSS and HTML interact, how the DOM should be
> implemented, etc.

>> Assuming that I do correctly understand what you mean by "implementation
>> requirements", then those should also be split from the vocabulary --
>> but it might well make sense to keep them with the DOM.

> Why would authors who write scripts be ok with reading a spec that
> included implementation details when authors who don't write scripts
> wouldn't be ok with reading a spec that included scripting details?

They might not be.

I am confident that both scripting and implementation details should
be separated from Markup vocabulary.

I am not confident about whether they should be separated from each other.

Nor am I as confident that the separation could be made cleanly,
because the effects of scripting are a major cause of the complexity
in implementation, but implementation details (and browser
differences) are a major cause of complexity in deployed scripts.

-jJ

Received on Sunday, 23 November 2008 06:45:06 UTC