xmouse

Marking up blogs

Dave Shea mentioned moving to generic XML for blogging purposes recently. It’s an interesting idea: is it worth moving markup languages, due to the limitations of HTML? Let’s pretend this was actually possible in the first place, and look at the options.

Generic XML

This is an interesting proposition, and it could certainly work. There’s one consideration that is essential. If blogging is to move to generic XML, then everyone has to move to the same markup language; i.e. someone needs to invent a ‘BloggingML’, not just everyone start making up XML to suit their own purposes: settling on one language.

Once this language is decided upon, its semantics need to be defined. This would likely take the form of a document a bit like the HTML4 specification: listing the available elements alongside their meaning. Hell, we could even please Tim Berners-Lee and use RDF for the job. With this semantic specification comes real usefulness, as everyone begins to use entry to mean an entry, and categorylist when listing their categories. The elements take on meaning.

This is necessary, as without semantics, machines won’t be able to understand our language. This is a really bad thing as a web browser is a machine, so we’d have no way to display BloggingML! If some people are using comment for trackbacks whereas others are using trackback, things fall apart.

So, things necessary for a new XML language to work are standardisation (one language: no format wars akin to RSS vs. Atom) and a semantic specification (everyone using the same tags for the same data; little ambiguity).

Extending XHTML

The idea of XHTML is that it’s extensible. This is obvious; it’s what the ‘X’ stands for! But what if instead of moving to BloggingML, we simply wrote our own XHTML extension (with carefully designed semantics, of course) and started using that? The same procedures would have to be followed as in the generic XML section, but this couDave Shea mentioned moving to generic XML for blogging purposes recently. It’s an interesting idea: is it worth moving markup languages, due to the limitations of HTML? Let’s pretend this was actually possible in the first place, and look at the options.

Generic XML

This is an interesting proposition, and it could certainly work. There’s one consideration that is essential. If blogging is to move to generic XML, then everyone has to move to the same markup language; i.e. someone needs to invent a ‘BloggingML’, not just everyone start making up XML to suit their own purposes: settling on one language.

Once this language is decided upon, its semantics need to be defined. This would likely take the form of a document a bit like the HTML4 specification: listing the available elements alongside their meaning. Hell, we could even please Tim Berners-Lee and use RDF for the job. With this semantic specification comes real usefulness, as everyone begins to use entry to mean an entry, and categorylist when listing their categories. The elements take on meaning.

This is necessary, as without semantics, machines won’t be able to understand our language. This is a really bad thing as a web browser is a machine, so we’d have no way to display BloggingML! If some people are using comment for trackbacks whereas others are using trackback, things fall apart.

So, things necessary for a new XML language to work are standardisation (one language: no format wars akin to RSS vs. Atom) and a semantic specification (everyone using the same tags for the same data; little ambiguity).

Extending XHTML

The idea of XHTML is that it’s extensible. This is obvious; it’s what the ‘X’ stands for! But what if instead of moving to BloggingML, we simply wrote our own XHTML extension (with carefully designed semantics, of course) and started using that? The same procedures would have to be followed as in the generic XML section, but this could be a way of keeping elements like em and a which are likely to be useful no matter what format your publication is; from books to weblogs.

Conclusion

Either of these options present ideal ways toward real markup freedom: no more semantic hacks like div. Of course all of this is still a pipe dream as there’s no way we’d get browsers to understand a whole new language in the short term. I guess in that respect, extending XHTML is a prettier option than writing a whole new markup language as browsers would already understand most of what we’re writing. There’s also those that would say (X)HTML is good enough already.ld be a way of keeping elements like em and a which are likely to be useful no matter what format your publication is; from books to weblogs.