response cache and matching lives in Fetch

Preload response's are one of several responses that interact with the
(yet to be formally defined) response cache. As such, the logic for both
populating and querying said cache should live in Fetch.

- Updated issue text as a warning, with some context and pointer to the
  discussion in Fetch.
- Removed "match a preloaded response" definition: this should be
  handled transparently by Fetch, as one of the first steps in its main
  fetch algorithm.

Fetch issue where we should continue this discussion:
whatwg/fetch#354

Closes #30.

index.html

@@ -229,33 +229,25 @@ <h2>Processing</h2>
  <li>If the load was successful, [queue a task] to [fire a simple event]
  named `load` at the [link] element. Otherwise, [queue a task] to [fire
  a simple event] named `error` at the [link] element.</li>
- <li>Add request to preload response cache (TODO).
- <p class="issue">TODO: define preload "response cache". Conceptually
- this should probably reuse [SW's cache
- logic](https://slightlyoff.github.io/ServiceWorker/spec/service_worker/#cache-objects)?
- In effect, the preload cache is a UA maintained response cache with a
- few interesting lifetime + eviction quirks.</p>
+ <li>Add request to fetch group's response cache.
+ <div class="advisement">
+ <p>In addition to the HTTP cache, all browser implementations
+ provide one or more levels of additional caches, which sometimes
+ live before the HTTP cache (e.g. HTTP/2 server push responses are
+ typically not commited to HTTP cache until a client request is
+ made), and after the HTTP cache (e.g. in-process memory caches).
+ These caches are not defined today and need to be defined in Fetch
+ API— see [related
+ discussion](https://github.com/whatwg/fetch/issues/354).</p>
+ <p>Conceptually, a preloaded response should be committed to the
+ HTTP cache, as it is initiated by the client, and should also be
+ available in the memory cache and be re-usable at least once within
+ the lifetime of a fetch group.</p>
+ </div>
  </li>
  </ol>
  <p>The user agent MUST NOT automatically execute or apply the resource
- against the current page context, and the user agent MUST retain the
- fetched response until it is <a data-lt=
- "match a preloaded response">matched with another request</a>, or is no
- longer valid.</p>
- <p>To <dfn>match a preloaded response</dfn>, the user agent must run the
- following steps:</p>
- <ul>
- <li>...</li>
- <li style="list-style: none; display: inline">
- <p class="issue">TODO: define matching, see <a href=
- "https://github.com/igrigorik/resource-hints/issues/5">this bug</a>
- and <a href=
- "https://groups.google.com/a/chromium.org/forum/#!msg/net-dev/cFhaIoJCRFg/3JX9lPWnDPIJ">
- chromium discussion</a>. This should probably live in Fetch? I.e.
- fetch requests should check this cache before checking browsers HTTP
- cache and going to network.</p>
- </li>
- </ul>
+ against the current page context.</p>
  <p class="note">For example, if a JavaScript resource is fetched via a
  <a>preload link</a> and the response contains a `no-cache` directive, the
  fetched response is retained by the user agent and is made immediately
@@ -356,11 +348,10 @@ <h2>Server Push (HTTP/2)</h2>
  <p>HTTP/2 ([[!RFC7540]]) allows a server to pre-emptively send ("push")
  responses to the client. A pushed response is semantically equivalent to
  a server responding to a request and, similar to a preloaded response, is
- retained by the user agent and executed by the application when
- <a data-lt="match a preloaded response">matched with another request</a>
- initiated by the application. As such, from an application perspective,
- there is no difference between consuming a preload or a server push
- response.</p>
+ retained by the user agent and executed by the application when matched
+ with a request initiated by the application. As such, from an application
+ perspective, there is no difference between consuming a preload or a
+ server push response.</p>
  <p>The server MAY initiate server push for <a>preload link</a> resources
  defined by the application for which it is [authoritative][]. Initiating
  server push eliminates the request roundtrip between client and server
@@ -411,7 +402,8 @@ <h2>Early fetch of critical resources</h2>
  penalty. To address this, the application can use a <a>preload link</a>
  to declaratively specify which resources the user agent must fetch early
  to improve page performance:</p>
- <pre class="example nolinks html" title="Early fetch of critical resources">
+ <pre class="example nolinks html" title=
+ "Early fetch of critical resources">
 &lt;link rel="preload" href="/assets/font.woff2" as="font" type="font/woff2"&gt;
 &lt;link rel="preload" href="/style/other.css" as="style"&gt;
 &lt;link rel="preload" href="//example.com/resource"&gt;