The house cleaning around here has been a little more literal; we’re having some HVAC work done so it was time to make the place more presentable.

I’ve added Wirebird::Handler::Profile which starts to implement some of the shadow profiling I mentioned yesterday. But that caused some side effects: checking both the Atom and RSS feeds means my Mastodon account gets two different names, and the template populator doesn’t cope well with that so it’s time to revisit that.

Now, the reason the names are different is because one is double-encoded… and the other is triple-encoded. So next week I’ll have to bite the bullet and get the Unicode stuff working consistently at all levels. (Not looking forward to that, especially dealing with older CPAN modules.)

The test suite is getting to the point where it needs to be broken up into multiple steps, which is good. The tests are still not terribly robust, but they’re at least making sure I (probably) don’t break old things while building new things.

So next week’s plan:

  • Fix unicode
  • Strip HTML for plaintext fields
  • Fix template population
  • Decide on authen/authz model

In the unlikely event I get all that done (though I do expect to have more time next week than I had this one), I’ll continue down the list of things required to reach viability.


Comment? Email it to me. (I'll assume I can publish it unless you say otherwise)


Next post: (Uni)Coding out loud: CPAN libraries and Unicode

Previous post: Rubberducking: more shadow profiling