And of course they are viewing the domain largely through the
tools they use for their work; which unfortunately creates a little
world of its own that they must train into without much transferable
knowledge or skill; and which I think people at PG are completely
unfamiliar with. But that's how they view the problem domain. If
they make guiprep happy, and see a document on the only output
medium available to them for validation (i.e. the browser screen),
they have no way to appreciate or adjust to your objections. Their
world is A-OK.

On Sun, Feb 5, 2012 at 6:48 PM, don kretz <dakretz@gmail.com> wrote:
For example, PG and DP need a vocabulary for discussing
the representation of book contents that is familiar, precise
enough, and comprehensive enough to discuss how they
will represent the book and how PG will interpret their
representation.

The common language I would expect woutd include
paragraphs, headings, chapters, poetry, emphasis,
tables, illustrations, captions, viewing devices (screen
capabilities and controls), ebook identification, acquisition,
distribution, storage, maintenance, ... These are all
concepts that have roughly similar meaning to both
sides. And you need to agree on the scope and workflow
in which you mutually participate; and have a roughly
good idea what the participation feels like to each other.

When the conversation instead devolves around markup
and divs and floats and margins and RCS, you no longer
are discussing the problem domain, nor are you using
vocabulary that is equally useful and meaningful to
both of you.