Re: Extensibility strategies

Henri Sivonen <hsivonen@iki.fi> wrote on 08/05/2008 06:39:58 AM:

> On Aug 2, 2008, at 14:44, Sam Ruby wrote:
>
> > I'm much more inclined to support the idea of evolution than the
> > notion of intelligent design.
>
> Interesting that you mention ID in this context. I've noticed a
> different parallel between this discussion and ID: It seems that
> Decentralized Extensibility is to Namespaces what Intelligent Design
> is to Creationism.

I've only argued that things should not be taken off the table until a
usable mechanism has been defined.

> Is this thread and ISSUE-41 about coming up with *some* mechanism for
> parties to extend HTML unilaterally without coordinating with anyone?

Some mechanism that people will actually use.  And has some strategy for
avoiding collisions.

> Or is this thread really about establishing a mechanism that involves
> binding a prefix to a URI using reserved attributes starting with
> "xmlns:" and then using the prefix in front of local names separated
> by a colon?

There are some advantages to such an approach, and some disadvantages.
Some of the advantages include working with the DOM API.  Some of the
disadvantages include usability issues that you have referred to.  Other
solutions which do not involve attributes starting with "xmlns:" can be
found which map onto the DOM API.  There may be ways to mitigate the
usuability issus with solutions which involve colons in element names.

> --
> Henri Sivonen
> hsivonen@iki.fi
> http://hsivonen.iki.fi/

- Sam Ruby