Lightnews — Scholar-powered news

Barry Pollard

@tunetheweb.com

2.6K followers 360 following 830 posts

Web Performance Developer Advocate at Google Chrome helping to make the web go faster! All opinions my own.

Posts Media Videos Starter Packs

Barry Pollard @tunetheweb.com · 1d

The docs also link to the source code for each Insight for those really wanting to get into the detail. This is especially important IMHO since they don’t live in the Lighthouse code anymore so aren’t as easy to find.

Enjoy!

Barry Pollard @tunetheweb.com · 1d

We’ve just released some shiny new docs for the Insights so you’ll know exactly what they measure and how to pass them. Check them out:

developer.chrome.com/docs/perform...

Performance | Chrome for Developers

developer.chrome.com

Barry Pollard @tunetheweb.com · 1d

I mentioned before that we’re switching Lighthouse Performance audits to the new and improved Insights. The same Insights that you can see in DevTools Performance panel traces.

That’s happening in Lighthouse 13 due to be released very soon. Maybe even this week!

Barry Pollard @tunetheweb.com · 4d

To me identifying it is rarely an issue. It’s the fact I want the account in question to be stopped. Clicking on them I often see them raising same spam on other repos. So dealing with this on a per-repo basis when it’s obvious they should just be suspended is frustrating.

Barry Pollard @tunetheweb.com · 4d

But it does risk the same issue being raised again, makes it more difficult to upvote or reference, and means others won’t fix even if you can’t (though honestly that’s unrealistic). So I’d prefer to be left open but flagged as “unlikely to fix”. I like Chrome DevTools new “icebox label” for this.

Barry Pollard @tunetheweb.com · 4d

Saying that I can see the value in triaging issues and flagging them as “unlikely to be worked on”. That’s honest to the reporter, gives them a change to provide more evidence why that might be the wrong call, and allows maintainers to see the wood for the trees.

Barry Pollard @tunetheweb.com · 4d

Stalebot in particular annoys me. “This issue hasn’t been worked on in X days so we’re gonna assume it’s no longer an issue. Please go to the effort of filing a new issue if you wanna, but honestly we’ll probably ignore that too.” It’s just riles me up!

Barry Pollard @tunetheweb.com · 4d

Yeah I thought it was more that. Which is why I wish GitHub allowed reporting and banning/deleting of accounts.

It really can’t be that difficult to automate if ratio of work to spam complaints is higher than X then automatically suspend, subject to appeal and auto delete of no appeal in Y days.

Barry Pollard @tunetheweb.com · 4d

What’s the benefit of that over just closing them?

Reposted by Barry Pollard

Rick Byers @rbyers.net · 5d

I believe it's fundamentally impossible to ship Chrome on iOS meeting the same high quality bar we have on other platforms when we are forced to rely on WKWebView and very slow and unreliable issue handling.

Reposted by Barry Pollard

Barry Schwartz @rustybrick.com · 6d

Google's @tunetheweb.com on why Google Search Console LCP metric is not wrong www.seroundtable.com/google-searc...

#corewebvitals #googlesearchconsole #lcp

Reposted by Barry Pollard

Dave Smart @tamethebots.com · 7d

MCP Server fever! Following on from the robots.txt MCP server release of yesterday, today I've launched one for you web perf fans!

Query the CrUX API in your favorite AI tool

Example: youtu.be/et7Og7_Ii6E?...

Get the details here: tamethebots.com/tools/cwv-hi...

#WebPerf #CWV #CoreWebVitals

Using the Tame The Bots CrUX API MCP Server

YouTube video by Dave Smart

Reposted by Barry Pollard

Nordic.js @nordicjs.com · 6d

Talk about making an entrance!! 🍰✨ @una.im

Reposted by Barry Pollard

Bramus @bram.us · 6d

iOS apps on iPadOS 26 are now actually nice because you can run them windowed.

But that keyboard, oh boy … #functionalhighground

Screenshot of Bsky on iPadOS with the keyboard open. The alignment of the letters is 😭

Barry Pollard @tunetheweb.com · 6d

“Added preview support for Event Timing API (Interaction to Next Paint)”

#webperf #CoreWebVitals

IntentToShip @intenttoship.dev · 6d

Webkit: Release Notes for Safari Technology Preview 229

Webkit: Release Notes for Safari Technology Preview 229

Webkit: Release Notes for Safari Technology Preview 229

Barry Pollard @tunetheweb.com · 7d

And if you work an another framework, or heck even if you just wanna use this on an app by app basis, then it's SO EASY to add these sorts of things to your performance traces too:

developer.chrome.com/docs/devtool...

Customize your performance data with extensibility API | Chrome DevTools | Chrome for Developers

Customize performance data with the extensibility API

developer.chrome.com

Barry Pollard @tunetheweb.com · 7d

React 19.2 has shipped including support for the DevTools Extensibility API so you can understand what React is doing when, and how that relates to the rest of the performance profile 🥳 🎉

react.dev/blog/2025/10...

Chrome DevTools performance trace showing some custom tracks. Scheduler Custom Track including Blocking, Transition, Suspense, and Idle sub tracks. And Components custom track.

Reposted by Barry Pollard

Chris Heilmann @codepo8.bsky.social · 8d

CFP is open for the 2026 WeAreDevelopers events in Berlin and San Jose:
Berlin: sessionize.com/wearedevelop...
San Jose: sessionize.com/wearedevelop...
Learn what we look for in a good CfP submissions:
christianheilmann.com/2025/10/01/c...

Barry Pollard @tunetheweb.com · 7d

“If you use ChatGPT to create your talk title and description, we invite ChatGPT and not you.”

Love it!!

Chris Heilmann @codepo8.bsky.social · 8d

CFP is open for the 2026 WeAreDevelopers events in Berlin and San Jose:
Berlin: sessionize.com/wearedevelop...
San Jose: sessionize.com/wearedevelop...
Learn what we look for in a good CfP submissions:
christianheilmann.com/2025/10/01/c...

Reposted by Barry Pollard

Tammy Everts 🍁 @tammyeverts.com · 7d

Every fall I hear so much FOMO from folks who missed @perfnow.nl. Don't let it happen to you!

We have amazing speakers sharing the newest best practices for the #webperf topics you care about: AI, APIs, #a11y, CSS, JS, frameworks, tools, culture, workflows & more!

Get your tickets soon! perfnow.nl

performance.now(); 30th and 31st of October, Amsterdam

On 30th and 31st of October, 2025, the sixth edition of performance.now() will take place in Amsterdam—a single track conference, covering today’s most important web performance insights.

Barry Pollard @tunetheweb.com · 8d

Anyway, hopefully that explains WHY Google Search Console can show this confusing, seemingly contradictory results, and why that doesn't mean it's incorrect.

Caching isn't the only reason of course, but it's a common one for this scenario.

Hit me up with any questions!

12/12 🏁 FIN

Barry Pollard @tunetheweb.com · 8d

Incidentally, this can also be why ad campaigns (with random UTM params and the like) can also be slower.

You can configure CDNs to ignore these and not treat them as a new pages. There's also an upcoming standard to allow a page to specify what params don't matter:
bsky.app/profile/tune...

11/12

Barry Pollard @tunetheweb.com · 12d

This is cool and been waiting for No-Vary-Search to escape Speculation Rules (where this was originally started) to the more general use case.

This allows you to say that certain client-side URL params (e.g. gclid or other analytic params) can be ignored and still use the resource from the cache.

Barry Pollard @tunetheweb.com · 8d

If it is a lot slower then you now understand the difference between your cached and uncached and can start thinking of ways to improve that.

Ideally you get it under 2.5 seconds even without cache, and your (cached) popular pages are simply even faster still!

10/12

Barry Pollard @tunetheweb.com · 8d

One way I like check this is to add a random URL param to a URL (e.g. ?test=1234) and then rerun a Lighthouse test on it changing the value each time. Usually this results in getting an uncached page back.

Compare that to a cached page run (by running the normal URL a couple of times).

9/12

Barry Pollard @tunetheweb.com · 8d

So how to fix?

There is always going to be a limit to cache sizes and priming caches for little-visited pages doesn't make much sense, so you need to reduce the load time of uncached pages.

Caching should be a "cherry on top" to boost speed, rather than the only reason you have a fast site.

8/12