Algebraic Data Types 2019-08-19

By Max Woerner Chase

I continue to wilt in this heat. Just a month-ish more, I hope.

I just took a cursory look at Structured Data to make sure I wasn't missing anything when I just kind of ripped stuff out; I think I might need more doctests. But it looks like removing guards didn't alter the documentation, which... eh? I'll try to get back to it, but first, I read another blog, and it's clear to me that in addition to fixing things so that there are actual links to all of the pages I'm generating, I need to seriously up my aesthetic game. Ideally in some kind of opt-in way. More on Structured Data... the highest-priority bug is probably the documentation. Let's see what state that's in... Everything visible in the documentation is documented, but I should probably see if there are any return types that can show up...

I decided to add bits of documentation and annotation here and there, and ended up checking on the issues being flagged by the various code inspection services I have hooked up to the repo. And then undoing a bunch of ignores, and pushing more changes, to see what happens. The result is confusing. I'll have to check on it later to see if it manages to make sense.

I'm going to figure out what's actually needed to get this to 1.0. Previously, things were kind of stalled out by the immense complexity of a design that attempted to subsume the basic concepts of function application into a more declarative framework, which was also a bad idea for a single-person project that didn't specifically need the capabilities that would provide. Now that that plan has been replaced with "ask people to write functions specific to their use-case", I think maybe I need to redo my limited-to-myself requirements gathering and see how Dennis does what it does.

That's enough messing around for tonight. Good night.