Re: header syntax.

* Dave Pawson <dave.pawson@gmail.com> [2012-11-22 03:28 -0500]:
> 
> Just to confirm the 3 you mention.
> 
> == underline (h1 only)
> --  underline (h2 only)
> #{1..6}  All values.
> 
> What a mess! We do need to address the corner cases though and tidy
> them up?  E.g. terminating #'s, matching number of hashes, text after
> hashes etc.

Regarding matching number of hashes, do we want to require that?  Or
just a single terminating hash?  I think a single terminating hash would
be sufficient, as the only purpose of the terminator is to indicate the
end of the header.

> Alternative view. For a core profile. Support only #'s, with
> terminators.  That allows conformance across implementations.
> 
> Then the variations would fall into either intermediate or
> app-specific profiles?
> 
> Thoughts?

I like the idea of only having one method of identifying the headers and
levels, and using the most flexible method (hashes with terminating
hashes) as the method in the spec.  I also like that this prevents
confusion between horizontal rules and headers.

-- 
dave [ please don't CC me ]