W3C home > Mailing lists > Public > public-web-perf@w3.org > December 2014

[resource-hints] splitting mandatory semantics into "preload" spec?

From: Ilya Grigorik <igrigorik@google.com>
Date: Mon, 8 Dec 2014 14:04:38 -0800
Message-ID: <CADXXVKowWewkDEFcGH7-_6wDpaz+O=G0bcfLeLcKkWGAnY4mQw@mail.gmail.com>
To: public-web-perf <public-web-perf@w3.org>
One of the themes at TPAC was that the current RH spec strays from the
typical meaning of "hint" -- case in point, the whole notion of a
"mandatory hint" is a misnomer. Also, there was some confusion about the
use cases for "mandatory" vs. "optional" semantics enabled by the
primitives.

In an attempt to separate the concerns and clarify the use cases, I'd like
to propose that we split RH into two pieces:

(1) "Preload" spec that defines rel=preload with mandatory semantics
(2) "Resource Hints" spec that defines rel=preconnect and extends
rel=preload with optional semantics

To that end, I've taken a first run at separating the two:
- Preload:
https://cdn.rawgit.com/w3c/resource-hints/d5027bffbf1eaa114c1d7906256339d118ea794a/index.html
- Resource Hints:
https://cdn.rawgit.com/w3c/resource-hints/e19f621dad9856a92bf43a06a62489c0058d19de/index.html

Note that functionally there is *nothing new* in these drafts as compared
to the previous RH draft. Instead, this is strictly an editorial update
that aims to clarify the semantics and the use cases of the underlying
primitives.

Thoughts?

ig
Received on Monday, 8 December 2014 22:05:50 UTC

This archive was generated by hypermail 2.3.1 : Monday, 8 December 2014 22:05:50 UTC