Here's my stack:
Hosted webspace (yes, like it's the 1990s!) Apache/Unix, simple PHP templates for headers and footers, basic navigation, etc w/o JS, for the blog a custom (server-side) script generating virtual article locations, navigation, meta data, basic chrome, and a RRS feed from a simple data-file and reading static HTML for the content from static files. BBEdit for editing. No analytics. (The webspace comes with Webalizer, but this stopped working some 10 years ago. I don't care.)
And yes, there are special pages with full-fledged JS apps, as well, or blog posts featuring complex things, so nothing is chiseled in stone. If there's a purpose to it. However, a visitor should be able to get an idea what a page is about without JS.