Closed mnot closed 3 years ago
Just a quick question, maybe bike shedding. Why call it SEARCH
and not QUERY
given that the spec itself says:
the SEARCH method is used to ask the server to perform a query operation (described by the request payload) over some set of data scoped to the effective request URI.
I think QUERY is much more general than SEARCH, and could nicely map to complex query uses, like GraphQL. Also, in this way it would not have to deal with backwards compatibility of WebDAV, so one could request application/xml
encoding of results without having to care about WebDAV compatibility (RFC 5323).
Also:
The response to a SEARCH request is not cacheable.
Hm, why not? Shouldn't server control this using regular control cache headers? Why would we limit this in the spec?
Just a note - this repo is for managing the adoption of the draft, not technical discussion about it.
Which place is better? The issue in http-extensions was closed, and also linked draft is an expired one.
Until it's adopted by the group, we're not formally discussing it.
I read the https://github.com/httpwg/http-extensions/blob/master/CONTRIBUTING.md and as far I understood the only way to discuss is email list, right? Because I have some thoughts against the new SEARCH method (similar to what @mitar said) but not sure how to discuss it: should I find some other letter or start a new thread.
Best place is the mailing list, correct.
Discussed in the October 2020 interim meeting; candidate for Call for Adoption.
Discussed on list: https://www.w3.org/mid/F0556EC2-D5AD-47FF-A780-15949F57A911@mnot.net
Seems to be strong interest in this area, but scope needs to be open regarding things like method name, cache ability, etc. Don't detect actual disagreements yet, just areas to be discussed.
Adopting.
See draft-snell-search-method.