gerald said:> the idea is to have the source
> AND the rendered book in hypertext
aha! great point. i see the utility now...
> For Project Gutenberg you get the source
> (but not a rendered book)
which is where i entered this thread... :+)
as, indeed, for some time, p.g. has offered
.html versions of all its newly-done books,
and even .epub and .mobi/kindle versions...
once you understand the p.g. conventions,
it's relatively easy to leverage them and
generate .html at will. indeed, it was my
effort to do that, at scale, for the whole
library -- as it approached 10,000 books
at the time -- which led me to first make
z.m.l., during the period of 1998-2003...
_however!_ the .html versions over at p.g.
are usually _not_ auto-generated from their
plain-text file... no, instead, the process
by which the .html file is created is left
to the individual in charge of the book, so
that means each book is its own "snowflake",
similar to each other, but "unique", in ways
that are not apparent without close analysis.
so inconsistencies in the e-text versions
have been doubled in the .html versions!
as you can imagine, this makes updating the
library extremely difficult, virtually impossible.
even the generation of the e-book formats
has been fraught with problems due to that.
back in 2003, i was unsuccessful in getting
the p.g. powers-that-be to grok the power of
z.m.l. to create and maintain a cyberlibrary.
they constantly informed me a light-markup
system "won't work" and mocked me for it.
they wanted to use t.e.i. instead, while
the minions who were doing the real work
preferred to create their own snowflakes,
and rolled their eyes at complex t.e.i.
the conflict would've been comical, but
their mistakes led to the decline and fall
of the world's very first great cyberlibrary.
50,000 e-texts now, but a shadow of itself.
> and for Leanpub you get the rendered book
> but not the source.
yeah, i've often thought that that's very ironic.
you don't get the most useful version of the book.
-bowerbird