home.social

#webperf — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #webperf, aggregated by home.social.

  1. There’s a special kind of satisfaction in updating servers and sites to the latest PHP 8.5 without a single hiccup. 🚀

    That’s when you know you’re doing things the right way—the Webinoly way.

    Stability and performance, no compromises.

  2. Now, this is just the home page, I still have some work left on other parts of the site. And the performance overall has already been pretty decent.

    The two things I mainly focused on were:

    1. Color contrast. I've been a bit lazy about this, I will admit, relying on Bootstrap's own color theme. Finally cleaned this up.

    2. Custom web fonts. And I think this is a big one. The only font I really wanted to keep is the Atkinson Hyperlegible Font from Braille Institute as the main font. Huge fan of this font.

    brailleinstitute.org/freefont/

    For the rest, using system fonts is very much "good enough" and I highly recommend checking out modernfontstacks.com for this.

    #WebDev #WebPerf #WebPerformance #PageSpeed

  3. Hey fediverse, happy Friday!

    A small personal victory, but after a few days of tinkering with my website stefanbohacek.com, I finally achieved the mythical perfect PageSpeed score.

    #WebDev #WebPerf #WebPerformance #PageSpeed

  4. Speeding up oxlint and oxfmt

    Another excellent blog post in the "Speeding up the JavaScript ecosystem" series by Marvin
    Hagemeister @marvinh

    marvinh.dev/blog/speeding-up-j

  5. The April web performance data for Shopify themes is now live - check out the newsletter with all updates and the site for the full data at:
    themevitals.com/newsletter/202

    #Shopify #webperf

  6. The April web performance data for Shopify themes is now live - check out the newsletter with all updates and the site for the full data at:
    themevitals.com/newsletter/202

    #Shopify #webperf

  7. The April web performance data for Shopify themes is now live - check out the newsletter with all updates and the site for the full data at:
    themevitals.com/newsletter/202

    #Shopify #webperf

  8. The April web performance data for Shopify themes is now live - check out the newsletter with all updates and the site for the full data at:
    themevitals.com/newsletter/202

    #Shopify #webperf

  9. The April web performance data for Shopify themes is now live - check out the newsletter with all updates and the site for the full data at:
    themevitals.com/newsletter/202

    #Shopify #webperf

  10. RE: mamot.fr/@ParisWeb/11655567226

    #parisweb, ma conférence de cœur, a accepté que je vienne vous présenter un atelier pour se mettre à jour sur les outils de la #webperf

  11. A bit inside baseball -- how I made my blog faster: cut CI per push from ~6 min to ~2 min, and kept every home/archive page O(1) in corpus size. Principles: compute once at input-change time, put the result under a stable URL, let everything downstream pull it from cache.

    benjaminhan.net/posts/20260511

    #Synesis #WebPerf #Quarto #StaticSite

  12. A bit inside baseball -- how I made my blog faster: cut CI per push from ~6 min to ~2 min, and kept every home/archive page O(1) in corpus size. Principles: compute once at input-change time, put the result under a stable URL, let everything downstream pull it from cache.

    benjaminhan.net/posts/20260511

    #Synesis #WebPerf #Quarto #StaticSite

  13. A bit inside baseball -- how I made my blog faster: cut CI per push from ~6 min to ~2 min, and kept every home/archive page O(1) in corpus size. Principles: compute once at input-change time, put the result under a stable URL, let everything downstream pull it from cache.

    benjaminhan.net/posts/20260511

    #Synesis #WebPerf #Quarto #StaticSite

  14. A bit inside baseball -- how I made my blog faster: cut CI per push from ~6 min to ~2 min, and kept every home/archive page O(1) in corpus size. Principles: compute once at input-change time, put the result under a stable URL, let everything downstream pull it from cache.

    benjaminhan.net/posts/20260511

    #Synesis #WebPerf #Quarto #StaticSite

  15. A bit inside baseball -- how I made my blog faster: cut CI per push from ~6 min to ~2 min, and kept every home/archive page O(1) in corpus size. Principles: compute once at input-change time, put the result under a stable URL, let everything downstream pull it from cache.

    benjaminhan.net/posts/20260511

    #Synesis #WebPerf #Quarto #StaticSite

  16. RE: fosstodon.org/@webinoly/116489

    Still tracking the Ubuntu 26.04 rollout.

    This release introduced several critical low-level changes and improvements that are causing a slower-than-usual turnaround for some upstream packages. While developers work through these complexities, we remain in "stability first" mode.

    We’ll pull the trigger on full support as soon as the core stack meets production standards. Thanks for your patience!

  17. RE: fosstodon.org/@webinoly/116489

    Still tracking the Ubuntu 26.04 rollout.

    This release introduced several critical low-level changes and improvements that are causing a slower-than-usual turnaround for some upstream packages. While developers work through these complexities, we remain in "stability first" mode.

    We’ll pull the trigger on full support as soon as the core stack meets production standards. Thanks for your patience!

    #Ubuntu2604 #Webinoly #SysAdmin #LTS #WebPerf

  18. Most cache misconfiguration is not carelessness, it's a missing handshake.

    Your build encodes assumptions: hashed filenames mean the URL changes whenever the content changes. The web server has to know that, or the assumption stays unused.

    If the config doesn't reflect what the build produces, the framework's work gets quietly undone at the last layer of the chain.

    #WebPerf #Caching #StaticSite #Astro #Caddy

  19. Most cache misconfiguration is not carelessness, it's a missing handshake.

    Your build encodes assumptions: hashed filenames mean the URL changes whenever the content changes. The web server has to know that, or the assumption stays unused.

    If the config doesn't reflect what the build produces, the framework's work gets quietly undone at the last layer of the chain.

    #WebPerf #Caching #StaticSite #Astro #Caddy

  20. Most cache misconfiguration is not carelessness, it's a missing handshake.

    Your build encodes assumptions: hashed filenames mean the URL changes whenever the content changes. The web server has to know that, or the assumption stays unused.

    If the config doesn't reflect what the build produces, the framework's work gets quietly undone at the last layer of the chain.

    #WebPerf #Caching #StaticSite #Astro #Caddy

  21. Your site loads. From the outside it works. But: do returning visitors re-download everything on every click? Can the connection be downgraded from HTTPS to HTTP on public WiFi? Does your homepage count as one site in Google's eyes, or two?

    For most static sites: no, yes, and yes. The web server config is the last layer most setups never touch.

    #Astro #Caddy #WebPerf #SelfHosting #StaticSite

  22. Your site loads. From the outside it works. But: do returning visitors re-download everything on every click? Can the connection be downgraded from HTTPS to HTTP on public WiFi? Does your homepage count as one site in Google's eyes, or two?

    For most static sites: no, yes, and yes. The web server config is the last layer most setups never touch.

    #Astro #Caddy #WebPerf #SelfHosting #StaticSite

  23. Your site loads. From the outside it works. But: do returning visitors re-download everything on every click? Can the connection be downgraded from HTTPS to HTTP on public WiFi? Does your homepage count as one site in Google's eyes, or two?

    For most static sites: no, yes, and yes. The web server config is the last layer most setups never touch.

    #Astro #Caddy #WebPerf #SelfHosting #StaticSite

  24. Your site loads. From the outside it works. But: do returning visitors re-download everything on every click? Can the connection be downgraded from HTTPS to HTTP on public WiFi? Does your homepage count as one site in Google's eyes, or two?

    For most static sites: no, yes, and yes. The web server config is the last layer most setups never touch.

    #Astro #Caddy #WebPerf #SelfHosting #StaticSite

  25. Single data point by the looks of it but dang, TCP RTT from North Korea to one of our CDNs is slow AF, 1.53 seconds.

    For comparison, the mean values around Europe are in the order of 25-30ms.

    #WebPerformance #WebPerf #WebDevelopment #WebDev