greg said:
>   Not all input books or types can be reasonably
>   accurately converted to any possible format, especially
>   for the older titles with no well-formed & valid HTML version.

as i've said for years now, with a small commitment from you
to consistent formatting, i could take plain-ascii files as input,
automatically apply the typographic niceties that are expected,
and output the results to .pdf and to .html, such that the .html
can be converted to a large number of other auxiliary formats.

of course, i'm not unique.  david moynihan has done it for years.

david was willing to make a small commitment to edit the files
himself so as to obtain that consistent formatting.  i think it is
more important to teach you how to fish than to give you fish.

check with 3 tool-makers from distributed proofreaders --
thundergnat, donovan, and bill flis -- and they'll confirm that
a clear path for ascii-to-(x)html conversion is quite workable
-- due to the fact that d.p. now has the required consistency --
even with their current programs, and that if they worked on it
a bit more, they could make it into a regular part of the workflow.

there is no need for the more-complex switch to a .tei workflow.

-bowerbird

p.s.  if you only would have accepted moynihan's offer of his files
when he made it to you, you'd already _have_ a consistent library.