It's especially interesting because this same idea is also responsible for saving Rails 3's ass from being far slower than Rails 2. Read through this discussion of how Rails 3 queries were sped up by using an AST decoupled from its final use: http://www.slideshare.net/tenderlove/zomg-why-is-this-code-s....
The original upskirt markdown library, written by Natacha Porté[1], had a specific goal of separating the parsing of markdown and the generation of html. When github forked it, they apparently (according to Natacha's comments on her site) slowly added more and more logic which mingled and blurred the lines.
Now this article seems to be coming from a standpoint that separation of the two concerns is somehow novel. A bit odd to say the least. I think this is where the o.p. was deriving sarcasm.
Perhaps the real story is the separation of concerns in the ruby module wrapper, instead of at the c library level? I don't know.
It allows real computer scientists to laugh at stupid webdevs reinventing the wheel. </sarcasm>
(Except no one was claiming that it was a novel idea, just that in the past, most Markdown processors just convert directly between Markdown and HTML, but now they are changing that. It's also funny that there seems to be some animosity in that post towards Ruby programmers for whatever reason. It's not like Markdown is an invention of Ruby, the reference implementation was written in Perl.)
