iiinova — where your digital self keeps evolving

Tim Kadlec @tkadlec.bsky.social · 26/03/2026

p75 and p90 JavaScript bytes on the web have exploded since May, 2025—easily the largest 10 month period based on HTTP Archive data since 2023. The long-tail is getting worse, very quickly. Gemini, what happened in May 2025?

69442

Tim Kadlec @tkadlec.bsky.social · 31/10/2025

Always, always happy to see @yoav.ws get a shout-out. So much incredible work has happened because of him. Most recent example—speculation rules in WebKit. #perfnow

1243

Tim Kadlec @tkadlec.bsky.social · 31/10/2025

I have been sooooo looking forward to this! SPA heuristics are gonna be such a huge step forward for measuring perf on the web and finally having hard data on the perf impact. @mmocny.com #perfnow

Michal Mocney presenting at perfnow about SPA heuristics, demoing being able to measure LCP and other metrics on soft navigations

0132

Tim Kadlec @tkadlec.bsky.social · 30/10/2025

@inesakrap.bsky.social starting her #perfnow talk with Lego's and Bluey has me already a fan.

082

Tim Kadlec @tkadlec.bsky.social · 15/10/2025

This is super exciting, particularly knowing Safari has it in Tech Preview so we're _this_ close to having it cross-browser. Also fun to watch the rollout through the lens of Cloudflare's Web Analytics.

061

Tim Kadlec @tkadlec.bsky.social · 02/07/2025

👀 Safari plans to ship LCP and INP in 2025?! Be still my beating heart. webkit.org/blog/16458/a...

160

Tim Kadlec @tkadlec.bsky.social · 01/07/2025

This is incredible data from @dbelson.com showing HTML requests from crawlers versus referrals from that platform. Very easy to see the cost of AI here. ClaudeBot: 70.9k requests : 1 referral (!!!) OpenAI: 1.6K : 1 Perplexity: 202.4 : 1 blog.cloudflare.com/ai-search-cr...

0105

Tim Kadlec @tkadlec.bsky.social · 22/06/2025

Lake Superior remains undefeated

1220

Tim Kadlec @tkadlec.bsky.social · 31/01/2025

Been wondering why a bunch of sites have been SUPER slow lately. Turns out it's the @1password.bsky.social chrome extension doing really bad things. :/ Gonna have to disable for awhile.

Chrome's performance timline, showing a 3.99s long task caused by a script injected by the 1Password Chrome Extension.

Screenshot showing the chunk of code (a mutation obsever) in the 1Password extension that seems to be the culprit.

4468

Tim Kadlec @tkadlec.bsky.social · 14/12/2024

Basically doing the same thing.

140

Tim Kadlec @tkadlec.bsky.social · 15/11/2024

"Always iterate". @anniesullie.bsky.social Good advice for product management, performance experimentation and so much more. I'm a huge proponent of the "what's the least amount of work we can do to see if we're on the right path" approach. #perfnow

Annie Sullivan on stage at perfnow, with a slide showing the cycle of change - measure - learn

2144

Tim Kadlec @tkadlec.bsky.social · 02/08/2023

Friendly reminder that when a third-party provider requires you to load their script in a render-blocking manner, their problems become your problems. Always build with resilience in mind.

This image shows the LCP metric from two page loads (same page, same location, same connection, same device). One is 1.96s, the other is 10.08s.

The difference is a third-party provider who was having server issues, causing their script to take 9.5s to load.

031

Tim Kadlec @tkadlec.bsky.social · 14/07/2023

Started working with these folks in mid-May and not gonna lie, pretty pleased with the progress they've made on INP in such a short span of time. From a p75 of 4.25s in April (90% of sessions having a poor experience) to a p75 of 225ms in June (9% of sessions have a poor experience).

Screenshot from Treo showing INP on mobile for the past year. There's a massive jump in INP performance in May, followed by another big one in June. INP went from 89.8% poor in April to 9% poor in June.

040

Tim Kadlec @tkadlec.bsky.social · 08/06/2023

Phew. What a difference a `:last` makes in jQuery. Third-party Shopify app with a very long list of selectors in a click handler. With the deprecated `:last` selector: ~616ms Without: ~16ms (Talked to the app devs and they're looking at it now so hopefully this gets fixed)

The heading of the image says "Same event, same selectors, with and without jQuery's :last"

Under it shows two flame charts from Chrome. One takes 616ms and is the click event with :last in the selector list.

The other takes only 16ms. It's the exact same event, but with :last removed.

240