It's possible "CPU idle" means something like "no work for 1 second". If the first ~20s are full of small occasional bits of work, it can always respond promptly to input, but doesn't count as falling idle.
-
-
-
This reminded me of Time to Interactive, which has this problem: https://webplatform.news/issues/2019-03-22#a-web-page-may-be-interactive-sooner-than-the-time-to-interactive-metric-suggests ….
- Još 1 odgovor
Novi razgovor -
-
-
For Lighthouse lab tests "All tests are run using a simulated mobile device, throttled to a fast 3G network & 4x CPU slowdown." (https://web.dev/measure/ ). So I'd guess most of the visitors are on better networks and CPUs than Lighthouse uses?
-
The field data represents anonymized data from users in the real-world on a variety of devices and network conditions. The lab data is based on a simulated load of a page on a single device and fixed set of conditions. As a result, the values may differ.https://developers.google.com/speed/docs/insights/v5/about#faq …
- Još 3 druga odgovora
Novi razgovor -
-
-
What version of Chrome is that? I only get lab data in mine
Novi razgovor -
-
-
Interesting ! Have you tried comparing results with other instances of Lighthouse? I sometimes get very different scores between http://web.dev/measure , Chrome's lighthouse, Lighthouse cli / PageSpeed insights API, with exactly the same settings.
-
It’s all simulated throttling, so it depends on the hardware that runs the test. I prefer PageSpeed Insights over the others because it includes field data.
Kraj razgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.
I create news content for web developers on