Mark Nottingham

> The harder to deploy thing is almost certainly a feature, for sure. I'd say that once software is in place, flipping the switch should be easy, it's the process...

Wait a few weeks; the HTTP specs are about to be revised as new RFCs.

Some discussion at TPAC. There are many use cases for range requests, initial approach might be to identify bare minimum and only support those. E.g., 1. Pass through range requests...

1. You can take the approach that every 206 is combined into one cache entry as soon as it occurs, or you can alternatively store multiple 206s and combine them...

> Is something for the Cache API to decide. I mean the HTTP cache, not the SW cache. Or do you mean it's effectively an implementation decision?

FWIW, I've written some basic tests to see if the HTTP (!SW) cache that Fetch uses supports partial content (i.e., range requests). AFAICT, the only browser that does anything is...

If you can agree on a common limit between browsers, we can refine this in the next HTTP; tell us on httpwg/http11bis if so.

Doing it generically would indeed be very broken. To do this for a specific application (e.g., HTMLMediaElement), you need a _really_ explicit assertion that not only are the two resources...

7230 3.2.2 only puts the #-rule constraint on senders; recipients can combine headers without knowing their syntax. WRT structured headers, see https://github.com/httpwg/http-extensions/issues/596 and most recent text at the bottom of...

3986 [defines userinfo](https://www.rfc-editor.org/rfc/rfc3986.html#section-3.2.1) to include `sub-delims`, which includes `'`. ~~~ abnf userinfo = *( unreserved / pct-encoded / sub-delims / ":" ) unreserved = ALPHA / DIGIT / "-" /...