Jerzy Jelinek

@bestwildclips .i

Great site #performance = great #UserExperience 💻 Learn about #INP - the new #CoreWebVitals metrics in our latest blog post by Kacper Stodolak. Click here to read → bit.ly/4clF8MM #webperformance #FrontEnd #Tech

My very own, first official CSS trick. Enjoy ❤️

🪩 Web Performance calendar day 15 🔗 calendar.perfplanet.com/2023/who-stole… 📺 New post "Who stole my LCP?" (hum to the tune of "I want my MTV") 📈 Building on his experience at @Shopify and elsewhere, @krzksz shares some lesser-known quirks of the LCP #webperf metric

We just announced that Interaction to Next Paint (INP) will replace FID as a Core Web Vitals metric. If you want to learn more about INP, please watch the deep dive @mmocny and I gave here: io.google/2023/program/b…

Substantial #webperf change to Largest Contentful Paint metric is rolling out. Browsers (back to Chrome 109) will now ignore low entropy images 🎉 chromium.googlesource.com/chromium/src/+…

Before deploying 103 Early Hints my advise is to measure: "Time To Parsing" (TTP) Without this new TTP you otherwise have a blindspot tracking HTML Performance. As TTFB gets a different meaning... #webperf #earlyhints

Override multiple response headers in one go! Yes, you can! Modify the .headers file in the Sources panel… It’s a prototype! Test it in Chrome Canary and leave your feedback in our RFC. 👉🏼 github.com/ChromeDevTools… #ChromeDevTools #RFC

Full page prerenders are coming back to Chrome! - How does it work? - Why is this good for Core Web Vitals? - How can I use it on my site? - What do you need to do/be aware of? Lots of answers to those questions here: developer.chrome.com/blog/prerender…

@PerfNowConf @liseuek @stasyaner88 Thank you for a great time! See you next year;)

@rick_viscomi @TheGreenGreek Again, who eats pancakes with butter? ;)

Being wealthy is unusual. We need to keep this in mind when setting performance budgets. We are serving web pages to people who are not us. @slightlylate speaking truth at @PerfNowConf #perfnow #webperf

Wow! @nicj just announced Akamai’s RUM Archive (beta) open-source project 🤩 rumarchive.com #perfnow #webperf #perfmatters @PerfNowConf

Fascinating data on CLS improvements in the last 2 years from @nicj at @PerfNowConf. Different industries started with *very* different score histograms..

@liseuek Some proxies, VPNs etc. are not ready to consume EH. They fail when response has EH and users couldn't see the page at all.

Q: What are the tradeoffs of Early Hints? #perfnow #perfNowConf

This is golden! Chrome 107 (currently in beta) comes with a new `renderBlockingStatus` field on ResourceTiming entries. 👏 Wanna find and monitor all the render blocking resources in a page? Here you go!

Seems like TTFB measured by @RealWebPageTest is unaffected by this. #webperf

@colinbendell @yoavweiss @colinbendell meanwhile check DM

@jerzyjelinek Thanks! Can you also file a CR in Chrome? I suspect we also need to clarify if PerformanceTiming.responseStart should be a) the first non-informational http response b) the first byte of the http header or body frame cf: #dom-performancetiming-responsestart" target="_blank" rel="nofollow noopener">w3.org/TR/navigation-… /cc: @yoavweiss

If you use TTFB to measure your backend, be careful with #EarlyHints — they're gonna take over the responseStart and obscure the actual metric. #webperf

@colinbendell I saw that... I will provide sample repo later

@jerzyjelinek In our analysis, we found TTFB (using responseStart) actually increased, as expected, because of TCP Head of Line blocking from Early Hints. see: x.com/colinbendell/s…