<div class="gmail_quote">On Fri, Apr 3, 2009 at 3:16 PM, Ian Hickson <span dir="ltr"><ian@hixie.ch></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div><div></div><div class="h5">On Fri, 3 Apr 2009, Jeremy Orlow wrote:<br>
><br>
> I created a page that sets a variable in sessionStorage, then I<br>
> navigated to a different domain, then I went back to the page and<br>
> checked if the variable was still set. In Safari 4 (beta) it is. In<br>
> IE8 it isn't. The spec is not terribly clear: "When a top-level browsing<br>
> context is destroyed (and therefore permanently inaccessible to the<br>
> user) the data stored in its session storage areas can be discarded with<br>
> it, as the API described in this specification provides no way for that<br>
> data to ever be subsequently retrieved."<br>
><br>
> Can anyone explain what's intended here? If not, can the spec make it<br>
> more clear?<br>
<br>
</div></div>IE8 is buggy according to the text you quoted. I don't understand why the<br>
text is unclear. Does the definition of "browsing context" not leave this<br>
unambiguous?<br></blockquote></div><font class="Apple-style-span" color="#888888"><br></font><br>Hmm...<div><br>In Chrome we also create a new browsing context when the user types a new URL into the location bar of an existing tab. This can be thought of as a shortcut for "create a new tab with the given URL and close the old tab." (We only do this when the new URL is of a different domain.) Perhaps, IE8 is doing something similar. Jeremy didn't say how he navigated. A link click or any content driven navigation cannot be treated like this of course.<br>
<br></div><div>However, I believe that we should still restore the old sessionStorage when the user navigates back to the old URL just as we do for form state, scroll position, and other attributes normally associated with session history.<br>
<br></div><div>-Darin</div>