Pedro Fayolle wrote:
Very very impressive! Once again, this is just like what I had planned to do, only better :) Congratulations on your hard work!
Just a question. I don't see what the benefit is from using a parser+presentator model when you could just convert wiki markup directly into HTML. On a more generic project it would likely fit, but since you're running your code on a browser which uses HTML anyway, what's the gain? Isn't it just slowing it down?
Just generally easier to write and debug because the HTML representation is lossy. The presentation layer takes only about a tenth the time of the parser anyway. The article never actually exists as text XML, just an in-memory DOM tree, it is faster than you'd think. It might even be faster this way because I can throw bits of the parse tree about witout worrying about presentation.
Programs are mostly easier to write with seperation of roles, there's a good article on this here:
http://www.faqs.org/docs/artu/ch01s06.html#id2877537
Can't wait to see what's next. It all looks awesome so far.
Best wishes,
Pedro