[Webkit-unassigned] [Bug 153230] Network cache: old pages returned by disk cache on history navigation after session is restored
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Fri Feb 26 09:56:10 PST 2016
https://bugs.webkit.org/show_bug.cgi?id=153230
--- Comment #47 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to comment #46)
> Comment on attachment 272315 [details]
> Go back to initial approach
>
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=272315&action=review
>
> r=me with changes.
Thank you very much!
> > Source/WebCore/history/HistoryItem.h:262
> > + // Item was created from a session restore.
>
> I don't think this comment is useful.
Ok.
> > Source/WebCore/history/HistoryItem.h:263
> > + bool m_restoredFromSession { false };
>
> we need a prefix for boolean members, I suggest m_wasRestoredFromSession.
Ok.
> Also, could you please move this new member next to the other boolean
> members above?
Sure.
> > Source/WebCore/loader/FrameLoader.cpp:3340
> > + if (!item.wasRestoredFromSession())
>
> I would do the #ifdef iOS logic here.
Ok.
> > Source/WebKit2/WebProcess/WebPage/WebPage.cpp:2234
> > +void WebPage::restoreSessionInternal(const Vector<BackForwardListItemState>& itemStates, bool restoredByAPIRequest)
>
> Do we really need a new method? Maybe we could just add the extra parameter
> to restoreSession() ?
Yes, because it's a message handler called from generated code.
> > Source/WebKit2/WebProcess/WebPage/WebPage.cpp:2238
> > +#if !PLATFORM(IOS)
>
> I would prefer if we moved the IOS ifdefing to
> FrameLoader::loadDifferentDocumentItem() and set the wasRestoredFromSession
> flag here no matter the platform. Otherwise, we end up with a boolean that
> has an inaccurate value on iOS.
Makes sense.
> > Source/WebKit2/WebProcess/WebPage/WebPage.cpp:2239
> > + // iOS always wants stale data after session restore.
>
> s/wants/allows
Ok.
> > Source/WebKit2/WebProcess/WebPage/WebPage.cpp:2248
> > + restoreSessionInternal(itemStates, true);
>
> Could you please make the boolean argument an enum for readability? e.g.
> enum class WasRestoredByAPIRequest { No, Yes };
Sure, my original patch did that indeed, but I was not happy with the names I used.
--
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20160226/b5b5b3a7/attachment.html>
More information about the webkit-unassigned
mailing list