Erik,
Your key claim, that RichText doesn't work, simply isn't correct. There
already is software implementing it and doing even useful things on
dumb glass TTY devices.
Is it ugly?
Could it be cleaner?
Are there gobs of spiffy additional features that could/should/will be added?
Is it inadequately close to SGML? Inadequately far from SGML?
...
These are useful questions, I think, but not if anyone uses the
discussions or the answers to alter the current processing of
Mime and RichText. Why? Because none of the above are legitimate
showstoppers? Why? Because the base technical capabilities of RichText
work and the policy issues (e.g., SGML conformance) are relatively
minor and, more importantly, could have been/were discussed in the group
earlier. The rule of group decision making is that one-writ, the hand
moves on. You only get to re-open a basic policy decision if there is
a serious, show-stopper potential that suddenly appears. I do not see
any such claim, in this case.
And the part of me that once thought of becoming a lawyer will observe that
there is an escape hatch:
Mime allows an infinite set of body part types and an infinite set of
Text content sub-types.
If one or more of you think that RichText is a major loser, then PLEASE DO
specify an alternative and recruit your community consituency. If you
accomplish this, there is a very real liklihood that you can have it declared
a standard.
I have often tried to find a simple framework for viewing the Internet
standards process and finally have settled on The Free Market System.
Competition is a good thing and we support it gladly. But the wrong way
to conduct a free market economy is to have one constituency stop
another's efforts (unless serious envrionmental impact can be claimed
and proved.)
Dave