Re: [EXTERNAL] Re: Call for Consensus: Incubations adoption

On Wed, Aug 26, 2020 at 12:10 PM Benjamin De Kosnik <>

> Microsoft supports most of the proposals (event timing, element timing,
>> isInputPending, layout instability) except some concern over LCP.
>> Quoting an engineer working on large web apps:
>> “I think this metric can be misleading. Although there was a switch from
>> FMP to LCP (Largest), we should be careful assuming that the largest
>> element is the most important to the user when measuring user experience. I
>> might be wrong, but to me it feels like this metric was designed to support
>> automated tools like LightHouse, without any developers’ annotations. “
> Agreed. It seems as if LCP is being pushed as a key part of Web Vitals for
> 2020, and that rush is unfortunate. It would have been much easier for the
> rest of the browser vendors to support that effort if FCP was used instead,
> and work on LCP or something like it was refined through 2020 for later use.

Happy to discuss the tradeoffs between heuristics and developer annotations
in an issue as Yoav suggested.

As for rush, can you clarify what you were referring to? From our
perspective, LCP has been repeatedly discussed in working group meetings
for well over a year
during which we received great feedback and addressed it by presenting
analyses of the heuristic in July 2019
and December 2019
<>. We
appreciate the recent feedback we've received from Mozilla and WebKit on
the spec, and are working hard on addressing it as well.

The CFC is to continue to work on these incubations as a WG draft, and we
intend to continue addressing feedback whether they are adopted as working
drafts at this time or not.

> best,
> Benjamin

Received on Wednesday, 26 August 2020 17:24:44 UTC