> Further, it’s extremely hard to statically analyse selectors to see whether any rules are broken (or even a typo!) this means constraints have to be enforced

> Further, it’s extremely hard to statically analyse selectors to see whether any rules are broken (or even a typo!) this means constraints have to be enforced via code reviews and education, which is a harder problem to solve and scale. We should use a computer's assistance us on this, and not rely on humans being infallible (they’re not!) Is there a more general name for the argument in defense of a technological movement on the grounds that it prevents users of said tool from making a certain mistake, and that mistake is so disastrous that we see it necessary to implement a drastic change with wide-ranging demands on an entire class of users or eliminates them entirely from the user base? I see this kind of argument prop up all the time in these types of discussions and something about it just fundamentally irks me, but I can't put my finger on it. It's kind of like Google's AMP, where instead of making incremental changes to the technologies in question, we'd rather draw an arbitrary line between "good" and "bad" uses of that technology and create a fork or subset of that tech where only the "good" uses are allowed. For a less controversial but equally relevant example, see Crockford's "Javascript: the Good Parts," which irked me at the time as well for the same reasons, even though it became universally accepted (although not-so-universally applied). I'm trying to come up with an example from my favorite language (Ruby) and I can't. Ruby's approach seems to be to discourage "bad" use cases by making them ugly (ex. instance_variable_get) rather than preventing them outright.