Re: [dispatch] Building Blocks for HTTP APIs - next steps

[Lucas Pardue <[e-mail]>]

On Fri, Jul 31, 2020 at 1:54 PM Julian Reschke <[e-mail]>
wrote:

> Am 31.07.2020 um 14:49 schrieb Lucas Pardue:
> > Yes I like that distinction, it makes it clear that the group's work
> > would not have a dependency on Browsers providing explicit interfaces or
> > indirect access to these APIs. Similarly, it constrains the groups work
>
> I would find that kind of dependency rather surprising in IETF work.
>
> Are you saying that work items for HTTPBIS *do* have that dependency?
>

Forgive my overreaching language. I was speaking to the notion of running
code or intent to implement, for instance, examples I have seen are things
like cache digest. A more recent example would be WebTransport that is
splitting the work between protocol and API between IETF and W3C
respectively. I don't believe HTTPBIS has such dependencies now, but for
dispatching new work I think it is useful to know that something requiring
W3C liason is probably beyond scope for HTTPAPI.


> > to things that are not forbidden by Fetch [1] [2], providing some
>
> How so?
>

I mean if someone has an idea that works in the machine-to-machine realm
and is looking to run this on the Web, their analysis may identify that
`Sec-` prefix is needed and as such it is not a candidate for HTTPAPI.

Cheers
Lucas


>
>