Lmst

Tonight at 6pm the NYC WebPerf Meetup is running another in person "Meet for Speed" session. We'll be taking requests from attendees for sites to do a web performance analysis on. Fastly will be hosting us at their NYC office.

https://www.meetup.com/web-performance-ny/events/304709044/

This Friday afternoon I'll be at the Boston Web Performance meetup giving my talk on "Performance Mistakes". If you are in the Boston area and interested in web performance, join us!

https://www.meetup.com/boston-web-performance/events/304542099/

Last week I gave a talk at the #PerfNow conference titled “Performance Mistakes”. I started off explaining a bit about Etsy’s blameless culture, and how it’s ok to make mistakes. Then I shared 17 common #webperf mistakes using HTTP Archive data, with examples.

Slides are here - https://www.slideshare.net/slideshow/performance-now-24-performance-mistakes-final-pdf/273431101

A few attendees told me they found some of these mistakes on their websites while we were still at the conference! So now they can be considered performance opportunities!

Tomorrow at 6pm the NYC WebPerf Meetup is running another "Meet for Speed" session. We'll be taking requests from attendees for sites to do a web performance analysis on. Fastly will be hosting us at their NYC office.

RSVP here - https://www.meetup.com/web-performance-ny/events/300512872/

I wrote a new tool to help analyze potential third party performance risks on your site. It uses data from WebPageTest to group results in a sortable/filterable table with lots of metrics that might be helpful for estimating whether a third party is a performance/failure risk.

Give it a try and see if you learn anything new about your third party content! More details on how to use it here - https://paulcalvano.com/2024-09-03-discovering-third-party-performance-risks/

After last week's DigiCert certificate revocation incident, I thought it would be interesting to dig into some HTTP Archive data. I found that 1,241,943 websites would have been affected in some way (due to third parties). Imagine how disruptive that could have been!

I wrote a blog post about this here - https://paulcalvano.com/2024-08-04-third-parties-and-certificate-revocations/

Yesterday I gave a talk about font performance at the NYC WebPerf Meetup. Using data from the HTTP Archive I was able to find and share examples of sites that could benefit from performance optimizations on their font files. The talk was recorded and will be shared soon, but for now here's the slides!

https://www.slideshare.net/slideshow/font-performance-nyc-webperf-meetup-april-24-1abf/267363735

After experiencing my first earthquake, I checked the
US Geological Survey's aftershock forecast page. The loading spinner causes layout shifts and is displayed for a long time because their text based assets are not compressed. Fortunately this #webperf issue is an easy fix.

Twitter thread here in case anyone is interested. https://twitter.com/paulcalvano/status/1776302447108972663

Paul Calvano boosted:

Next in-person speaker event on April 18 @ 6:00PM EST at Fastly's new office in Union Square.
Font Performance with @paulcalvano.
RSVP: https://www.meetup.com/Web-Performance-NY/events/299395150

@simevidas @paulcalvano thanks! I’ll take a look and fix that later tonight!

Earlier this month Chrome shipped support for the zStandard compression. I wrote a blog post about determining which compression is best for your content, and what levels to use. https://paulcalvano.com/2024-03-19-choosing-between-gzip-brotli-and-zstandard-compression/

A few interesting findings:
- Brotli is now used more than gzip for some content types!
-There may be a TTFB benefit to using low level zStandard compression on dynamic payloads
- A surprising number ofsites are using gzip level 1!

I also updated my compression tool to support zStandard.