3 ms·
I'm so happy he realized this behaviour is not good. I understand that svelte templates are not HTML (but a superset as he says), so I'd like to assume that wha
by 7bit 3y ago
I'm so happy he realized this behaviour is not good. I understand that svelte templates are not HTML (but a superset as he says), so I'd like to assume that whatever is wrong HTML should also be wrong Svelte-template-language.
I really don't understand how people can defend using the self-closing div's etc, when this just contributes to the same set of problems they are criticising in the same sentence :smh:
Glad Rich is seeing it the same way and moves forward with fixing this!
- apetresc 3y agoIf Svelte templates really are a strict superset of HTML, then doesn't that mean by definition that there is "wrong HTML" that is valid Svelte? That would only not be true if Svelte was a sub-set.