• Suggestions about Markdown.

From what I can tell, the Markdown parser that's used here to preprocess comments is not actually suited for that task.

  1. It doesn't highlight links automatically. Must I really use these parentheses? It'd be nice to be able to simply write "http://herp/derp" and get a clickable link as a result.
  2. It replaces underscores inthemiddleofa_word with tags. Not a very nice thing to have on a website about a programming language in which that naming style is reserved for variables.
  3. Apparently, there's no way to inline some code in a paragraph. Normally, Markdown allows one to do that by using backticks (`some code` translates to <code>some code</code>), but that feature seems to be missing.
  4. Although it does not ignore single line breaks, HTML renders those efforts futile. I'd suggest adding "white-space: pre-line;" to <p> tags somewhere in the CSS.
  5. Even though it's only used to parse comments and forum posts, it's full of redundant stuff. I mean, paragraphs in list items? Headers? What's that for?


If you're interested, I have an extremely lightweight regex-based Markdown subset parser made specifically for short comments rather than long blog posts. It doesn't support some of the fanciest markup stuff, but it does solve the first three problems. (Also, it can make use of Pygments for highlighting code wrapped in triple backticks, similar to GitHub Flavored Markdown. You will probably want to disable YouTube embedding by removing lines 70 and 85, though.) Oh, and it's probably slightly faster than whatever that is you're using (python-markdown2, I guess?), if that's of any benefit to you.

End of thinly disguised shameless self-promotion. :-)