[Elfwreck]

Originally Posted by ahi

A format that is incapable of professionally typesetting (a process that includes layout work, since professional typesetting is a bit of a dirty phrase around here)

I'm not sure about this. I suppose CSS could include kerning & leading options--the hard part would be getting readers to recognize the code. There's no reason an HTML file (and therefore, an ePub file) *couldn't* have typography; it's just currently difficult, and some features aren't supported on all viewers. (Especially ebook-only viewers.)

Even reflow-enthusiasts rarely argue that reflow is viable for typographically complex books, and such books do make up a considerable percentage (if not downright the majority) of existent books.

When you make claims like this, you need to support them.
Are standard novels "typographically complex?" If not (and I'd say not), you'd be fighting *hard* to claim that complex books are the majority of existing books.

Many textbooks are typographically complex... but they're also designed to fit printing requirements. They have multiple columns because it's cheaper to print less pages, not because it's easier to understand data that way. (However, the multiple heading levels and variable chart/photo sizes, caption options, and callout text boxes keep them complex even if the two- or three-column standard went away.)

And while those *are* a significant portion of books published, they're a niche market; people not attending school rarely deal with them. It's possible that, as ebook tech gets better, they'll mostly be replaced with ebooks with entirely different layouts.

New functionality like variable font sizes, in-book links to facilitate sensible jumping about, et cetera are great... but they do not make the product viable if it fails to meet the bare minimum quality standards set by paper books for the last several centuries.

Doctorow dealt with this topic in his essay "Ebooks: Neither E, Nor Books." (Emphasis added)

there was a time when books were hand-printed on rare leather by monks. The only people who could read them were priests, who got a regular eyeful of the really cool cartoons the monks drew in the margins. The priests read the books aloud, in Latin (to a predominantly non-Latin-speaking audience) in cathedrals, wreathed in pricey incense that rose from censers swung by altar boys.

Then Johannes Gutenberg invented the printing press. Martin Luther turned that press into a revolution. He printed Bibles in languages that non-priests could read, and distributed them to normal people who got to read the word of God all on their own.

* Luther Bibles lacked the manufacturing quality of the
illuminated Bibles. They were comparatively cheap and lacked the
typographical expressiveness that a really talented monk could
bring to bear when writing out the word of God

* Luther Bibles were utterly unsuited to the traditional use-case
for Bibles. A good Bible was supposed to reinforce the authority
of the man at the pulpit. It needed heft, it needed
impressiveness, and most of all, it needed rarity.

* The user-experience of Luther Bibles sucked. There was no
incense, no altar boys, and who (apart from the priesthood) knew
that reading was so friggin' hard on the eyes?

* Luther Bibles were a lot less trustworthy than the illuminated
numbers. Anyone with a press could run one off, subbing in any
apocryphal text he wanted -- and who knew how accurate that
translation was?

And yet... the printed versions took off, despite the lack of pretty artwork and possibility of poor content quality.

Ebooks don't need to be "better than pbooks in every way." They need to fill a niche that pbooks can't fill--and they're doing that.