Re: New Version Notification for draft-snell-search-method-02.txt

[Julian Reschke <[e-mail]>]

Am 08.09.2020 um 15:23 schrieb Philippe Mougin:
> As discussed a while back, I see a few problems in the draft.

FWIW, the dicussion we should have *right now* is whether the WG wants
to work on this general topic, not the details of the current proposal.

If this draft is adopted as work item, the WG will continue to modifiy
the proposal until we have (potentially rough) consensus on it.

That said...

> It states "The path identifies the resource processing the query (in this case 'http://example.org/feed') while the query identifies the specific parameters of the search operation".
>
> This recasts the Web model into an RPC-like system where the http://example.org/feed resource is some software we send parameters to in order for it to perform a search.

You may call that "recasting as RPC", but that's *exactly* what people
are asking for. Keep in mind that in practice the only difference to GET
with query parameters is that the query is in the payload, not in the
URI. We can sure word smith this some more, but this really is what this
topic is about.

> But actually there is no such resource involved. The resource involved is http://example.org/feed?q=foo&limit=10&sort=-published; it will typically be defined as being the result of a specific search computation.

In theory, yes. In practice of course *most* of the time there *is* a
resource identified by "http://example.org/feed", and it does indeed
process the query parameters.

 > ...

Best regards, Julian