<html>
<head>
<base href="https://bugs.webkit.org/" />
</head>
<body>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW - [GTK] No way to safely use webkit_web_view_get_snapshot() as pages are not rendered when LOAD_FINISHED is called"
href="https://bugs.webkit.org/show_bug.cgi?id=164180#c6">Comment # 6</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW - [GTK] No way to safely use webkit_web_view_get_snapshot() as pages are not rendered when LOAD_FINISHED is called"
href="https://bugs.webkit.org/show_bug.cgi?id=164180">bug 164180</a>
from <span class="vcard"><a class="email" href="mailto:cgarcia@igalia.com" title="Carlos Garcia Campos <cgarcia@igalia.com>"> <span class="fn">Carlos Garcia Campos</span></a>
</span></b>
<pre>(In reply to <a href="show_bug.cgi?id=164180#c4">comment #4</a>)
<span class="quote">> (In reply to <a href="show_bug.cgi?id=164180#c3">comment #3</a>)
> > (In reply to <a href="show_bug.cgi?id=164180#c2">comment #2</a>)
> > > (In reply to <a href="show_bug.cgi?id=164180#c1">comment #1</a>)
> > > > I don't think we should change the semantics of LOAD_FINISHED, we should
> > > > probably expose APILoaderClient::didReachLayoutMilestone, and the
> > > > applications could wait for DidFirstVisuallyNonEmptyLayout or
> > > > DidHitRelevantRepaintedObjectsAreaThreshold, I don't know the details of the
> > > > milestones.
> > >
> > > Why you think we should not change when LOAD_FINISHED triggers?
> > > What is the use case of having a LOAD_FINISHED signal that doesn't trigger
> > > always when the page is fully rendered??
> >
> > Because load finished is not about rendering but about networking, the page
> > finished loading, but not necessarily rendering. That's why there are other
> > callbacks more related to rendering like the ones I mentioned before.
>
> I think that this is an internal implementation detail that don't
> necessarily has to be exposed in our API
>
> Quoting our docs:
>
> " WEBKIT_LOAD_FINISHED This state means that everything that was required to
> display the page has been loaded. "</span >
Everything required to display doesn't mean it has displayed eferything, only that the resources are there, but you have to parse, layout, paint, etc.
<span class="quote">> <a href="https://webkitgtk.org/reference/webkitgtk/stable/WebKitWebFrame">https://webkitgtk.org/reference/webkitgtk/stable/WebKitWebFrame</a>.
> html#WebKitLoadStatus
>
> And after reading that twice, I would expect that when LOAD_FINISHED
> triggers the page is always fully rendered.
>
> Adding a new state is an option, but I think is a bad idea to do it, if we
> only add it for the sake of exposing the internal implementation.
>
> So, what is the use case for an application to have both
> WEBKIT_LOAD_FINISHED and WEBKIT_LOAD_FINISHED_AND_RENDERED_COMPLETED ?</span >
I don't think it's possible to know when a page has finished rendering, you can know if something has rendered, or if there was a layout, etc. But, if page has dynamic contents or animations. Load events has to do with the load, what happens in WebCore/loader, not with the layout or render. If we want to expose notifications about those I would not add new load events, but a different signal or whatever</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the assignee for the bug.</li>
</ul>
</body>
</html>