The site comment handling has now been greatly improved. If you're entering comments in "plain text" then the formatting should be preserved a lot more accurately.

Some simple transformations are applied to text, such as making hyperlinks and converting from *bold* into bold. But otherwise your input should remain unmangled.

Hyperlinks will still be allowed and because people tend to put HTML in their sigs <i>, <b>, <p>, and <a href="">....</a> should work.

Due to some minor optimizations each article read should now be significantly faster.

Essentially the whole page content will come out of the fast memory cache of content - with only two database lookups:

1. One to see if you're logged in.
2. And one to update the read count for the relevent article.
   • (Needed for the hall of fame.)

There is some invalidation going one when things are editted, or new articles added. But otherwise performance should be much improved.

More work is going on breaking up the monolithic codebase. I'm confident that this will soon be over.

I'm pleased with the progress, and the way things are shaping up. I still need to write more test cases though, to ensure I don't cause regressions.

The following perl modules are lovely:

• Test::More
• Test::Harness

I'm tempted to make a small article on their usage ..