On May 28, 2013, at 10:11 PM, Boris Zbarsky <[email protected]> wrote:
> On 5/29/13 12:55 AM, Brady Eidson wrote: >>> This expectation was always wrong: events can be dispatched by script at >>> any time to pages in the page cache. >> >> I'm sorry, who's page cache are you talking about? > > Gecko's, which initially defined the pagehide/pageshow events. ;) Got it. Then it sounds like Gecko’s page cache and WebKit’s page cache are divergent in this regard. Which is perfectly fine, as such a browser feature isn’t spec’d anyways. But the page transition events are spec’d, and I believe we’ve been shipping ours with the use case I described since before the spec tightened down on the allowable values of “persisted,” ignoring the use case we offer. > >> Page caches are (AFAIK) not really fleshed out in any spec, and the WebKit >> page cache *intends* to represent a 100% inert page. This includes the >> inability to receive events. > > So what happens when script calls dispatchEvent on a node that's in a > document that's in your page cache? I believe the design is “nothing.” I believe the quick test case I just tried confirms that - at least in my quick test case - we meet our design. > >> Can you provide a specific link to relevant threads you're thinking of? I >> can't guess what search terms you intend for me to use. > > Searching for "unload" and "pagehide" should probably find the right threads. > At least that's where I'd start searching, if I were searching. It’s unfortunate that to follow evolution of mature loading specs originally described in WHATWG that I would’ve had to follow w3c’s web-perf group. Oh well. I’ve tried these search terms and the only obviously relevant thing I could find was http://lists.w3.org/Archives/Public/public-web-perf/2012Feb/0111.html It asks for the change, but does not describe the rationale. I’ve provided our rationale for changing this, and I’m formally asking the WHATWG community if there’s any rationale for *not* changing this, Since you clearly participated in the discussion on web-perf back in 2012, I wager you’d be more expert in digging up any thread where that actual rationale was described. ~Brady
