On Fri, Nov 26, 2010 at 5:18 AM, Jakob Voss <[log in to unmask]> wrote:


> But if we want to find practical solutions, we definitely need a kind of
> central issue/bug-tracking system, a formal grammar, and a reference
> implementation. I started to hack something in Haskell
> (, but sure there are easier tools,
> that you are more familiar with.

An interesting aside: when James Clark documented the processing model
for Relax NG, he used Haskell:


.. I believe the Haskell fragments actually showed up in subsequent
RNG documents.

But certainly javascript would be more immediately useful. I do agree
a reference implementation is important.

> A simple wiki or collective google docs document would also help,
> but just a discussion list will not lead to a final standard.