The Webby Future of Structured Markup: Not your father’s XML
Canadian Centre for Studies in Publishing
Simon Fraser University
Oct 25, 2012
(Use arrow keys to navigate)
This is not a technical treatise,
but rather a manifesto for how
we think about markup.
How do we think about markup?
SGML was designed for industry
XML was designed with the Web in mind, but it still inherited SGML’s ROI model.
The case for SGML/XML:
- Content management at scale
- Efficient content re-use
- Write once, publish many times
- Output independence, especially going forward
- Open file format, software development standards
So, do publishers need XML-first workflows?
Conventional wisdom says, “Yes.”
XML is an Industrial-strength technology... with Industrial-sized cost and complexity
...because it was designed for expensive and complex problems
If your project has a million-dollar content development budget, a $200,000 savings on content re-use makes good business sense.
How many publishers live in
that world anymore?
Or ever did?
Increasingly, publishing is not Industrial anymore.
Dare I say...
“Publishing is not an Industry anymore”?
Industrial markup is not the way to the future.
It’s a solution to the wrong set of problems.
Let’s revisit the rationale
for structured markup...
Rationale for GML (circa 1976)
- vendor-neutral text-preparation format
- device-independent production format
Rationale for SGML (circa 1986):
- descriptive markup (as opposed to presentational markup)
- standardized method for indirection
- standardize parsing strategy
- document validity
- tree-shaped document structure
Rationale for XML (circa 1997):
- ubiquitous exchange format for the World-Wide Web
- single simple architecture for a large variety of applications
None of those reasons is obsolete!
How do we get back to first principles?
Industrial XML is utterly and hopelessly divorced from the world of writing.
Converting Word to XML has
been the holy grail since long before XML.
Can we have it both ways? A writing environment & a serious XML environment?
The real solution requires us
to unask the question.
Markup in actual practice. Incidental markup.
We can have the benefits of XML without the heavy industry
Because industrial efficiency isn’t what markup is about anymore.
It’s about interoperability & fluidity on the Web.
What if we treated nasty old HTML as though it were real markup?
HTML’s success is in its ubiquity.
“Just take a minute and consider how many person-years and dollars it’s taken to shake HTML down to the point where it generally just kind of interoperates and there are good authoring environments and so on.”
– Tim Bray
Encouraging things abound.
Changes in widespread practice:
- new publishing frameworks under the banner of HTML5
- beyond-blog publishing in WordPress
- writers working in Markdown
- tools that make HTML consistent, even valid
We have an opportunity to fix the disconnect between ‘Publishing’ & ‘Web’ mindflows.
The integral tension between tradition & innovation
As publishing moves forward, the internal tension between tradition and innovation threatens to tear it asunder.
How do we pull it together?
Simon Fraser University