A great post by Pablo Defendini sparked some thoughts on the production in publishing and the changes it will (well, should) implement.

First things first: the reader is your master.
There is no way in which the reader will buy for let’s say 15$ a poorly formatted e-book that he has troubles reading.
And a poorly formatted e-book is, for example, the same file you are printing, with the TOC, the title page, the blank pages, and so on. That’s great for a paper book. That’s awful for an electronic device.
Look at Liza’s slides. Memorize them. Sleep with them.
It’s easy to learn. It’s not easy doing it right.
But it’s essential, because the reader may buy something now, for the excitement of this-new-thing-that-is-digital-book, but will not be fooled for long. If what you’re selling them is badly designed and clumsy they will not pay you 15 bucks. They will not pay you even 5 bucks, actually. And that does not mean they will go and buy the trade hardcover or paperback.

Sidenote. Learn it as a mantra: a reader that wants to buy the e-book *will* have the e-book, or nothing. He will not go and buy a paperback instead. Yes, it may happen sometimes, but rarely. It's far more probable that he will download it from P2P, or give up.

Not only: apart from the layout, e-books are not static by nature. They may grow fatter, or thinner, they may change in time. Heck, they may also communicate with the external world, provided some kind of connection! So why would it be enough to put the book on a Kindle to make it magically an e-book? Sure, there will be room for this kind of static e-books, but big is the room for titles especially suited to the various devices.

For now anything good-enough or not-too-bad may be ok, but in the long run (long? let’s say medium) e-books will have to fit like a glove to the pros and cons of every device. And in order to win in the long run, it’s better to have an early start. And the early start requires some changes in production.

So what are books now? Well, almost what they have always been, a mix of text and paratext. The fact is that the paratext cannot be the same for a digital edition and for a printed one – and cannot be the same for all the digital editions.

Enter XML.
XML is a standard markup language aimed toward simplicity and interoperability. An XML file is a text file that does not contain any information on the layout of the text, but plenty of content-rich information.
An XML file may tell who is the author, may clearly identify that some word is a <place> instead of a <person>, may provide information that will help the device reading the file interact with other devices, and transmit or receive information (“hey, you are in Milan and two blocks from where you are right now, in an hour, the author will hold a conference!”).

Stripping away layout instructions and providing a standard language, XML helps a lot in tailoring the content for each device. Moreover, it permits to create a single file that *is* the book, put it in a central repository in the publishing house Content Management System (CMS), and send the content directly to all the possible devices already formatted to adapt to the potentialities of each single device – be it a book, a Kindle, the web, everything.

This is not immediate, of course. On the contrary, it requires a great amount of time and money.
Money, because there’s a system to build to take care of all the aspects of production inside the house. The best choice is with any probability to build the system in-house, so to tailor it to your specific needs and spare money in mainteinance and in buying a mammoth CMS that you then discover is almost useless for you. The relatively brief history of CMSs is full of failed experiments, of firms that bought huge systems that didn’t fit at all their needs. On the contrary, there are plenty of examples of successful smaller CMSs built from scratches.

The time expense is humongous as well, as requires to change perspective and habits to most players, and to learn new skills – but these changes are already covered well in Pablo’s post.
I do not really believe that you can get authors to provide clean, well formed XML files (but read the comments on Pablo’s post for nice insights), but on the other hand it has to be used in the workflow of the publishing house. There is no reason not to use it, unless you want to lose ground to all the other smaller or bigger publishers that will start, or already are, using it.
This is the way to go. This would have been the way to go years ago, actually.

A final thought, I will maybe build on that in a future post: the fact that there’s a design to be made for the new formats and devices open up spaces for great experiments. There’s a lot of work to do: there may be a whole new typography to be invented here.