Crosscut

Out of my long list of problems, let's start with the code database, which I've talked about many times and at length. It's a concept with far-reaching effects on many aspects of language design. But so far, I've never implemented it. I'm just planning to. And because of that, I constantly have to consider an unknown in my design decisions.

This causes a problem with these design decision. Those I make with the code database in mind, end up being suboptimal in the current situation. Others, that don't prepare for the code database, end up being preliminary. I mean, in a sense all design decisions are. But with those specifically, it's clear from the start that they are just placeholders.

But there's another problem, and it's probably the bigger one: The code database might just turn out to be a bad idea. It might not work out for any number of reasons. And the longer I go on, making decisions with this unproven idea in mind, the bigger the impact of this scenario becomes. It's a growing risk.