-
Run HTTP Archive script
- Extract scores from tests in the LH JSON
- Aggregate
- Plot on a per test basis
- Known-issues: CPU performance caluculation was change a while ago so those tests won't be comparable before and after a certain time.
-
We will never know if a person sees a test failure in LH and then fixes it. There's no connection between the test and the work.
- One has to assume that people are doing it because of LH because the tests are rather specific.
- If you fix one test and then another... would that be an indication.
1 min read
Are developers fixing the issue identified by lighthouse
Paul Kinlan
Lead of Chrome DevRel
Disclaimer. This article is a draft.
This article might look complete but it is completely un-developed and is a snapshot of my thoughts at the time (which might also be a long time ago). It will be refined over time with research, education and conversations with people in the industry.
There are many errors including and not limited to: spelling, grammatical, lack of ecosystem awareness understanding and plain ignorance.
I am making it public to get feedback. You might completely disagree, you might agree violently. Either way, teach me. I want to hear from you: (Email, @paul_kinlan.)
See Also
Stay in the loop.
I'm trialing a newsletter. Join for monthly insights into web dev, Chrome, and the open web.