Isn't the content model complicated enough as it is?

Authors > Hixie

It's bad for authors that you can put script/template anywhere in a <table> but not anywhere in a <ruby>.

If we're going to make it more consistent, I'm far more tempted to do so by removing the ability to put <script> in <table> than the other way around.

But I'm not convinced these need to be consistent.

Note that a complicated content model is also bad for authors.

What's the use case for ever putting <script> or <template> in <ruby> in a way that isn't already allowed? I don't understand what you would do. Note that <template> in tables makes a lot more sense, since there you might not know in advance what data you're going to be injecting.

I don't have a use case for script/template in ruby. I just expected the spec to more consistently allow script/template anywhere where the parser can put them, and thought it would be easier for authors to understand that they're allowed anywhere (modulo parser behavior).

It's definitely easier to understand "you can put stuff anywhere and it's conforming", but being easy to understand isn't the goal of content model conformance criteria. The goal is to try to catch unintended mistakes.

Yeah but <script> in <ruby> doesn't seem like an unintended mistake to me. Can you give an example of a page that has <script> somewhere that is a mistake (and isn't weirdly parsed)?

<script> is probably reasonable, I was thinking more about <template>, as a replacement for other elements. I guess if we're just talking about putting script elements in ruby then that's ok...

Let’s re-raise this in the GitHub issue tracker if necessary.