It is my opinion, though I'm not an authority on such matters, that the
simple
approach would be that a parser should not attempt to parse the contents
of a
<param>...</param> command section. Since the simple parser doesn't
understand
the contents of the <param> section anyway, what does it gain by
pretending
that it does?
You have to parse the contents, because you need to be able to
recognize the end of the block. Parse is actually too strong a
word - its really just lexical analysis.
<param> looks like it meets the needs. However, this whole
discussion is probably not too important. The basic usefulness
of enriched, if it ever gets nailed down, is in being a simple,
clear standard for passing a bit of marked-up text. It's value is in
being a lingua franca, not in having lots of whiz-bang capabilities.
My bet, given my experience with a couple of systems that provide
more powerful editing systems, is text/enriched will simply
be an mp/alternative format that provides a little more value than
plain text. In fact, given the wider use of windowed systems and
the resultant variety in linewidths, its linewrapping semantics may
be more useful in heterogeneous exchange than anything else.
Production systems will continue to use private or de-facto
standard protocols for sending more complete information.
There's no way text/enriched can ever be complete enough
to handle that. Technically its possible, but politically it will
never happen.
Terry