Re: [whatwg/fetch] Define a cache for H2 server push & friends (#354)

@annevk I'm curious what's intended by the fetch group semantics, as I don't believe that's something implemented in Chrome (nor easily implemented). The closest parallel I can think is if it's meant to be the abstraction layer detailing how our renderer process works (and the relationship to the renderer-side memory cache), but if that's the case, then it certainly is not part of our implementation or notion of Fetch, nor would I be keen (nor would it be possible) to hang more stuff off it.

I think I'd disagree with @jakearchibald on the relationship between PUSH and the cache being weird, and it may simply have to do with the choice of verb attached to it implying a semantic meaning unmet by implementations; that is, had it been called, say, OPPORTUNISTIC_RESOURCE, would that have made it clearer or not? I'm not aware of any implementation letting H/2 into the cache, and while that may be something to discuss, I absolutely want to make sure we don't entangle it with the notion of fetch groups unless/until we have a solid understanding of what they're trying to describe or prescribe. I'm not sure if this is exactly the right bug to hash it out, and I suspect I'd need to rope in a few colleagues to make sure we're accurately describing it, but for now, I'd like to separate out protocol-handling discussions (e.g. H/2) from "Fetch/Web Platform" level discussions (e.g. prefetch, preload; whether or not that's a fair characterization I'm not sure, but that's certainly reflected in how Chrome has implemented them and architected around)

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/fetch/issues/354#issuecomment-240885774