Re: navigationStart and the actual start of the navigation

On 7/1/14, 10:59 AM, Przemysław Pietrzkiewicz wrote:
> Every other attribute of PerformanceTiming is supposed to happen not
> earlier navigationStart as per [1]. So I'd assume that all of the other
> parameters exclude more of the events of the navigation timeline.

That's correct, but there is no attribute that corresponds to "after the 
previous page has finished doing whatever it's doing" apart from 
navigationStart.

> Because when the prompt is being displayed it's not determined if the
> navigation will happen.

So what?

If you care about the _user perceived_ latency (and it's not clear to me 
that that's the goal of this API as it stands), then that's what you 
should measure.  If the navigation never happens, there is no problem; 
nothing gets that navigationStart value from before the dialog came up.

> I think it does match the goal - which is to measure the user perceived
> time of the *browser navigation*.

Why do users care about this?

> This can't involve (possibly complex)
> script logic that causes the navigation to happen in response to an
> earlier user action.

I don't think users care about the difference, or indeed understand it, 
between a navigation taking 2s extra because the browser takes a long 
time to create a new window or because the page didn't respond to their 
click for 2s because it was mining bitcoins off the click event handler 
or because the browser spent 2s running other parts of the event loop 
before even delivering the click to the web page....

-Boris

Received on Tuesday, 1 July 2014 15:24:25 UTC