Use browser caching to eliminate duplicate file downloads

[billyc]

Datasette does something really cool: they

• (1) append a hash to every URL, which changes if the file changes, and
• (2) if you request a file without a hash (which is the normal use case), an HTTP 302 redirect always redirects you to the URL with the latest hash
• (3) and that response has a far-future "HTTP cache expiry header"

This way the browser only needs to download a file once and it then retrieved from cache for all subsequent requests. That is, until the file gets changed! Then the hash redirect forces an update. Pretty slick!

See https://simonwillison.net/2018/Oct/4/datasette-ideas/#Farfuture_cache_expiration

Can we do this??!?