Re: [quicwg/base-drafts] HTTP SETTINGS: define setting content encoding (#1556)

[Lucas Pardue <[e-mail]>]

I think there is merit in making SETTINGS simple numerical values, arising from lack of compelling argument otherwise. 

Are use cases that would like more forms of expression really going to be satisfied with an unmodifiable SETTINGS value?

> Here's an example that might make sense, though I doubt that I would implement it... a setting that included a base URL so that request targets could be identified relative to that URL if desired. That might be justified on the basis that we can't compress parts of header fields (for good reasons), but that particular tweak is probably safe.

This is curious example but I'll bite. You could achieve the same by defining an extension, enabled by a new numerical SETTINGS value, that articulates the base URL in a new extension frame. 

To take this to an extreme, we could imagine a request group concept *cough* where a set of requests are linked to a particular base URL via stapling to this extension frame. The base URL can be modified through the lifetime of the QUIC connection if needs be. 

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/1556#issuecomment-411422898