For one thing, dynamic publishing is not necessarily publishing that uses dynamic content (by Ann Rockley’s definition). Users have been able to get the information they need at the time they need it by search since the days of Gopher, which predates the Web. SneakerNet (transporting data to other computers by using a floppy disk) has always had higher bandwidth than most digital networks. Even card catalogs and interlibrary loan could make the information you needed available in a timely enough manner, considering. Links to PDFs are just a carry-over of this older concept of dynamic delivery–it was dynamic in its day, I suppose. 

But what makes today’s dynamic “publishing” possible has been 1) the refactoring of metadata into finer, more semantically-linked keys for smarter search capability, 2) the refactoring of lengthy content into smaller, semantically-enriched chunks that are more appropriate to the scope of knowledge needed, and 3) the creation of web services that enable smart search to hook up to intelligent content in response to a wide number of things that take the place of the old ILL office: portlets, bookmarklets, feeds, forms, Ajax widgets, apps and applets, and more. Very few of these resemble publishing in the traditional sense, but they all accomplish getting you the information that YOU have selected, on the go, and hopefully in the timeframe you expected. 

Clearly, any definition coined decades ago would be inadequate today, and your efforts today may suffer the same obsolescence decades hence… unless you approach the definition in terms of the benefit you receive, which tends to imply all the protocols, services, and practices that make that benefit possible.

And what about SneakerNet? Copying a movie to a capacious thumb drive and walking it to the HTPC in my living room is sometimes faster than waiting for my LAN to buffer enough of it to begin showing it. Long live sturdy sneakers?

Noz Urbina (one of the best minds in the business) shared this contribution to our efforts via email. I’ll ask him to stop by the site and register and post some additional thoughts later.

“Dynamic Assembly is a process by which units of content (‘units’ often called components, topics, articles, modules) can be delivered as a logically organised group – on request.  This means that what were discrete units previous to the moment they were asked for, become a unified deliverable ready for publication or print.  A ‘request’ could take several forms with various levels of automation.  The most interactive and user-driven would be allow users to browse for and select units of content and arrange them into their chosen order and/or hierarchy and request output to be delivered to them in some way.  The most automated would be to apply metadata (tagging, labelling, semantic mark-up, even simple dates, statistics and anlytics) to the units so that computer systems can use these to assemble new publications based on algorithms or queries.  Examples might include frequently accessed or requested content, or more complex queries like recent topics surrounding a certain event, e.g. ’2011 articles on the financial crisis that were published in Japan/. ”
 
“Dynamic Publishing allows content to be assembled on request (see Dynamic Assembly) and delivered to its destination ready for consumption.  It may add value above and beyond simple assembly and rendering by automatically enriching content with additional relationships, features or even additional content during the actual publishing process.”What do you think? Does Noz have it right. What else is needed to understand dynamic publishing?You can learn more about Noz by reading his very informative blog, “Less Work, More Flow”: http://lessworkmoreflow.blogspot.com/Scott

A definition should be as short as possible but leave
nothing out. With that in mind, here is a first attempt at a Dynamic Publishing
definition that I would like to offer for discussion:

“Dynamic Publishing is a system that automatically and
logically assembles discrete components of content on demand and produces an
acceptable published product.”

It may seem shocking that this doesn’t mention XML,
Intelligent Content, single-sourcing, multi-platform, personalization, or most
of the other things that we might consider absolutely essential parts of
Dynamic Publishing. Why would we want to start with such a simple definition and
how can it be accurate when so much is left out? I think that starting with
this bare bones definition is the key to helping others understand what Dynamic
Publishing is, and more importantly, what it can be in the future.  

Let’s look at another software application for which there
is already an authoritative definition. Most dictionaries define a word
processor as “a system for manipulating, storing, and outputting text entered
from a keyboard.” This mentions nothing of the vast features of Microsoft Word,
and also qualifies Notepad as a word processor—something that sounds ridiculous
now, but actually would have been pretty credible in the 1970s.

Even though hundreds of “essential” features have been added
to word processors over the past forty years, the dictionaries still hold to
the original 1970 definition because it is the foundation for understanding the
technology. Everything else we know about word processors enhances their
usability and value based on requirements that are changing all the time.

If we can contribute an authoritative definition, I think
the next step would be to describe the range of capabilities available under
the umbrella of Dynamic Publishing today and the technologies that drive them. They
will be different from what we would have listed just three years ago and certain
to evolve and need revisiting in the future.

It’s taken me a while to weigh
in, mostly because with the calibre of comments on this post, I wanted to be
sure that my remarks made a position contribution to the discussion. If I were
to explain dynamic publishing to a non-technical professional at the director
level, I would do so as follows:
 
Dynamic publishing
doesn’t happen auto-magically; there are four specific ingredients that go
into the recipe for the dynamic publishing of content.
 
First, let’s define
content. Content is anything you can publish that is usable (readable,
viewable) by humans. In other words, weird computer code that only geeks can
decipher doesn’t count. Text, video, audio counts.
 
Next, let’s define
publishing. Publishing is the process of making content visible to your
audiences. In the world of electronic publishing, that means getting content
out of storage and put where someone can see it. The content could be stored
in a flat file, a repository, a database, or similar place. Publishing it
could mean sending the content to a web page, a CD, an e-book, or any other
place where your audience can find it.
 
Dynamic publishing,
then, is sending a different combination of content to different people,
through the use of software that automates the process.- The “automated”
part is controlled by writing business rules that determines how the
software decides to differentiate who sees what content.- The “dynamic” part
is defined by how the software carries out the business rules.
 
So dynamic
publishing is the process of showing specific sets of content to different
audiences,  carried out by software that
analyses audience member choices, and then responds based on a set of business
rules.
 
After explaining
this, I imagine I’d get a bunch of questions which would end up having answers
such as:
The quality of
the dynamic publishing depends on the quality of the business rules, which
depends on how well the business and user requirements have been thought
through.
The audience
might make no choices (as in deciding what to show a person based on the
country they’re in, which is decided by their IP address), or they might
make a series of choices (as in filtering products until they have a
sufficiently small subset of information to make an informed decision).
Content is a
garbage-in, garbage-out situation. To be able to deliver up the content in
response to choices made (whether these are back-end or audience-end
choices), the content has to be in a technical format that allows the
appropriate software to understand where it fits in the business-rule
delivery system.
 
I know this is a
less technical view of dynamic publishing, but probably a more
business-friendly way of explaining it.
 

Wow! You folks are bringing up all sorts of issues. And, I hear you. Definitions versus implementation. Simple but encompass everything needed to understand it and make it accurate. Vague, but not useless. I think we should be able to craft something that most will agree as the basic description and then add to it from there. I am going to invite a few more thought leaders to the conversation. Let’s see where this takes us. Thanks. If you have new revelations or thoughts, share them here. — Scott

OK. I have taken another run at a “definition” and this time with the benefit of pouring over the contributions of my colleagues. Below, I have tried to map parts of my proposed definition to ideas I borrowed from each of them. (Immature poets imitate. Mature poets steal. – T.S. Eliot)

Here goes:

Dynamic publishing delivers customized information products that help users to be optimally effective. Dynamic publishing achieves its goal by leveraging automation to assemble content resources according to users’ requests and according to the applicable business rules, and then rendering, packaging and
delivering contextualized information products that they can use with whatever technology is available to them.

Some background on sources:

From Steve I took the importance of packaging and delivery – the things that make up the last mile and that sees the information products actually arrive in the hands of the users that requested them.

From Rahel, I took the emphasis being placed on business rules which, in enterprise environments, can be an almost overwhelming concern. And by the way, I don’t think that the formatting of your comment looks wonky Rahel – if E.E. Cummings were to write a definition of dynamic publishing it would look like that.

The stress that Mark places on dynamic release is a very interesting one. Also my experience confirms his assertion that the dynamic release of up-to-the-minute content can be the most profound change that a dynamic publishing capability can effect. I think that this becomes an “emergent property” of a dynamic publishing environment and so I have not found a graceful way to bring this out in the definition. Maybe someone else can. Or perhaps it is implicit in the capability laid out by the definition.

From Michael, I specifically plucked the emphasis placed on automation and on the assembly of content resources (discrete components).

I like Sarah’s definition because I spot in it some delightful details buried within its apparent simplicity. Specifically, I like the word “rendering”. In one of my more torturous decompositions of the “publishing” concept (in my past “Content Engineering” workshops) I break publishing into a whole sequence of process steps: selection, assembly, resolution, compilation, rendition and delivery. The hidden gem in Sarah’s definition (at least to my eye) is the fact that the “process of rendering” determines not just what content will be published but the presentation that will be applied. On one level this is obvious but on another it is not all that obvious. Perhaps I am reading too much into it, but to me this speaks to environments where the formatting rules are dynamically assembled and tailored in parallel with, and in exactly the same way as, the content itself. These are my kind of dynamic publishing solutions!

From Noz, I jumped on the points he makes about “enriching” information products to make them optimally useful. This can include things like contextualizing the products so that related products can be recommended to users. This way, if they have further questions, or their interest takes them in new directions, they have somewhere to start. This is how “contextualized information products” made it into my elaboration of the core definition. Perhaps “enriched” is better but after years of using that term I have found it a little too abstract (again people might find that funny coming from me).

From Don, I took further emphasis of the importance of semantically enriched (there’s that word again) chunks of content as the starting point for this next generation of dynamic publishing solutions.

I am reading David Weinberger’s Too Big to Know, in which he talks about how science has changed in the Internet age and how traditional publication was actually an important defining element of the traditional scientific method. Publishing was a major step in the evolution of a scientific idea, and involved the prolonged and careful evaluation of the results and conclusions by a network of peers. This meant that science presented itself to the world as a series of stable authorized ideas, and that it consisted largely of conclusions. Most of the data, especially the data on failed experiments, was never made public. Today, however, in the age of open notebooks and networked data, the world has access to, and can participate in, scientific work in progress. This unprecedented access to data has produced some important discoveries that otherwise might not have been possible.

Reading this made me realize that we are all (or most of us, at least) talking about publication in its old pre-Internet sense, as a monolithic act of an authority blessing information for distribution and thereby affixing to it an imprimatur. Publishing, in short, as marking the end of a lengthy process of collection, collation, evaluation, selection, analysis, and composition.

But this is not what publication is in the age of the Internet. When I press the post button when I finish typing this comment, it will be published, without oversight or debate or delay, to everyone from Toronto to Timbuktu and from Manchester to Mandalay. In being so published, it will dynamically change the article which Scot originally wrote. The experience of the reader who reads it after I press the Post button will be different from the experience of the reader who read it before it did so. The conclusions and impressions they take away from reading it may be quite different from those of the people who read it yesterday (or not).

In short, the answer to “What is dynamic publishing anyway?” is “this is”. What we are doing right here in this thread is dynamic publishing, for this topic is public and it is dynamic.

The particular form of dynamic publishing we are engaged in here might be called distributed manual dynamic publishing (a category that would include Wikipedia), but there is very clearly a great deal of value to be realized from automated and centralized models of dynamic publishing. But as to the question of what dynamic publishing is, its this, what we are doing right here, right now.