1. 9

  2. 8

    I would be happy with 20mb pages if more than 10% of those bytes actually made things better for me. As it is, my Adblock settings filter over 80% of total average http requests, and lots of the stuff I do retrieve is still junk.

    1. 6

      Funny, the entire speedcurve.com domain is blocked by my ad-blocker, so I’m going to guess they don’t see 3Mb pages as a problem at all?

      The important metric is how much of that size is real content versus advertising and tracking garbage or goofy JavaScript frameworks.

      3Mb to watch a video, a photo gallery, a high resolution image, etc. is not a problem at all. On the other hand, 3Mb for a few pages of text is ridiculous.

      1. 2
        1. 1

          I think a big part of this is how much of the 3MB is reusable after the first load. The main web app that I work on is about 2MB on initial load, but a combination of etagging, service workers, and other techniques means that subsequent loads are in the 2kb - 25kb range. Of course I always want initial page load to be better, but it’s certainly less of a concern when the cost is amortized over multiple uses of the site.

          1. 1

            The Chrome team has shown how on mobile devices the cost of parsing all of that JavaScript is non-trivial and that has to happen even on repeat visits. If mobile doesn’t matter in your use, then you’re absolutely right that it’s not a big deal!

            1. 1

              Laptops still have power constraints. A webpage should almost never cause any noticeable CPU load, but all too often they do. The author just had to add scroll animations or something.