decoupling --plaintext and --no-headers?

[Patrice Dumas]

Hello,

I think that --plaintext should be a way to ask for a plain text output,
which would always be not split and sent to stdout, similar with
--html, --docbook, while --no-headers would mean something more restrictive, 
that is avoid doing headers, more like --footnote-style, but not necessarily 
not split nor output to stdout.

For info output --no-headers would be exactly the same than --plaintext
(mostly for backward compatibility), but for other formats it could be 
different, for example for html there could be no headers while still
output in files (especially when split at chapter, for example), or it
could be a no-op (for html when split at nodes, for docbook).

What do you think?

--
Pat