How do 103 Early Hints responses affect measured TTFB?

[Vipul Naik]

Hello!

I'm curious how the reported value of time to first byte (TTFB) in PageSpeed Insights, Lighthouse, and the CrUX report will be affected by a site's use of 103 Early Hints, that became part of the main Chrome starting with Chrome 103?

https://blog.chromium.org/2022/05/chrome-103-beta-early-navigation-hints.html

https://developer.chrome.com/blog/early-hints/

https://chromestatus.com/feature/5207422375297024

Will the TTFB measure the time to first byte of the first 103 response (if it arrives before the 200) or will it still only measure the time to first byte of the main 200 response?

Vipul

[Michal Mocny]

Excellent question!

This list is probably not quite the right list to get this answer, but since I'm also quite curious :) I'll cc some folks who can hopefully answer.

(For some of these it may be easiest to just try and see...)

--
You received this message because you are subscribed to the Google Groups "web-vitals-feedback" group.
To unsubscribe from this group and stop receiving emails from it, send an email to web-vitals-feed...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/web-vitals-feedback/CAFGFH7d9ieuxRnTk%3D9rT_2tmp9XvZALYB2rvcqoDi_Le4Q0O%3Dg%40mail.gmail.com.

[Philip Walton]

From a Web Performance API perspective, the end time for TTFB has always been the responseStart property of the Navigation Timing API. And when I look at the HTML spec where that time is defined, it clearly includes considerations for 100-199 statuses.

So to answer your question (at least from a WebPerf API perspective) TTFB should be the first byte of the 103 response.

That being said, we should definitely double check that the timestamp in CrUX matches what is specified and what is RUM measurable.

To view this discussion on the web visit https://groups.google.com/d/msgid/web-vitals-feedback/CAEeF2TcJ8cJhwBkznayB-3S%3Ds%2Bp-ipS%3DpKPPGWRRi4HM44Otuw%40mail.gmail.com.

[Michal Mocny]

On Tue, Jul 5, 2022 at 2:04 PM 'Philip Walton' via web-vitals-feedback <web-vital...@googlegroups.com> wrote:

From a Web Performance API perspective, the end time for TTFB has always been the responseStart property of the Navigation Timing API. And when I look at the HTML spec where that time is defined, it clearly includes considerations for 100-199 statuses.

So to answer your question (at least from a WebPerf API perspective) TTFB should be the first byte of the 103 response.

And to be super crisp, just because of the way your question was worded:

- It's not the first bytes of the separate resource that the 103 early hint was hinting about...

- It's the first bytes of the 103 early hint response itself that is considered the overall TTFB.

 

That being said, we should definitely double check that the timestamp in CrUX matches what is specified and what is RUM measurable.

On Tue, Jul 5, 2022 at 8:05 AM 'Michal Mocny' via web-vitals-feedback <web-vital...@googlegroups.com> wrote:

Excellent question!

This list is probably not quite the right list to get this answer, but since I'm also quite curious :) I'll cc some folks who can hopefully answer.

(For some of these it may be easiest to just try and see...)

On Mon, Jul 4, 2022 at 3:46 PM Vipul Naik <vipul...@maven.io> wrote:

Hello!

I'm curious how the reported value of time to first byte (TTFB) in PageSpeed Insights, Lighthouse, and the CrUX report will be affected by a site's use of 103 Early Hints, that became part of the main Chrome starting with Chrome 103?

https://blog.chromium.org/2022/05/chrome-103-beta-early-navigation-hints.html

https://developer.chrome.com/blog/early-hints/

https://chromestatus.com/feature/5207422375297024

Will the TTFB measure the time to first byte of the first 103 response (if it arrives before the 200) or will it still only measure the time to first byte of the main 200 response?

Vipul

--
You received this message because you are subscribed to the Google Groups "web-vitals-feedback" group.
To unsubscribe from this group and stop receiving emails from it, send an email to web-vitals-feed...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/web-vitals-feedback/CAFGFH7d9ieuxRnTk%3D9rT_2tmp9XvZALYB2rvcqoDi_Le4Q0O%3Dg%40mail.gmail.com.

--
You received this message because you are subscribed to the Google Groups "web-vitals-feedback" group.
To unsubscribe from this group and stop receiving emails from it, send an email to web-vitals-feed...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/web-vitals-feedback/CAEeF2TcJ8cJhwBkznayB-3S%3Ds%2Bp-ipS%3DpKPPGWRRi4HM44Otuw%40mail.gmail.com.

--
You received this message because you are subscribed to the Google Groups "web-vitals-feedback" group.
To unsubscribe from this group and stop receiving emails from it, send an email to web-vitals-feed...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/web-vitals-feedback/CAChM%3Dvq0f_ZAtknCeqyA8rpGAH%2BNaMJ1fC%3DHnmCwdArYA5%3Dy%3DQ%40mail.gmail.com.