• Resolved Ambyomoron

    (@josiah-s-carberry)


    I am testing the impact of the various Autoptimize options using GTMetrix and Webpagetest. Averaging results of 5 trials each, I have compared Autooptimize enabled but no options selected at all (i.e., it should do nothing) to Autooptimize with only optimization of html, js and css. Most of the metrics are better with Autoptimize doing nothing at all. For example, average first bye with=2.863; without=2.289. Avg Start Render with=4.2; without=3.6. Avg. First contentful paint with=4.465; without=4.035. On the other hand, avg. Document Complete time with=7.695; without=8.291.

    There are no other plugins activated that specifically address any performance issues. Notably, no caching plugins are enabled.

    As you see, overall performance is quite bad, so I really need to determine how to improve things. So, getting longer times with Autoptimize is a surprise that I need to understand and am asking for help. It sure does look like the performance penalty of using Autoptimize is greater than the savings in optimized code.

    Many thanks in advance

Viewing 7 replies - 1 through 7 (of 7 total)
  • Plugin Author Optimizing Matters

    (@optimizingmatters)

    When using Autoptimize without page caching, your first time byte (even 2.2 indeed is pretty bad) will be higher as AO does quite a lot of heavy lifting to optimize HTML, CSS, JS & others, which in turn impacts most other KPI’s.

    Conclusion if going for AO then do install a page caching plugin and your results will be _a lot_ better 🙂

    hope this clarifies,
    frank

    Thread Starter Ambyomoron

    (@josiah-s-carberry)

    So what you’re saying is that the optimized HTML, CSS and JS is always created on the fly, rather than being created once and then cached. Is that correct?

    If it is correct, I understand the value of a caching plugin, but the heavy lifting to which you refer would have to be done whether the server is sending a cached page or not, no?

    If it is not correct (i.e., Autoptimize is caching the optimized code), then does it make any difference in which order Autoptimize and the caching plugin create their caches?

    Plugin Author Optimizing Matters

    (@optimizingmatters)

    So what you’re saying is that the optimized HTML, CSS and JS is always created on the fly, rather than being created once and then cached. Is that correct?

    the optimized HTML is never cached (that requires page caching), but the optimized JS & CSS is cached and re-used if possible.

    If it is correct, I understand the value of a caching plugin, but the heavy lifting to which you refer would have to be done whether the server is sending a cached page or not, no?

    Not really; if a page is served from page cache, autoptimize doesn’t even have to lift a finger, as the cache HTML will contain references to the cache autoptimized CSS/ JS 🙂

    If it is not correct (i.e., Autoptimize is caching the optimized code), then does it make any difference in which order Autoptimize and the caching plugin create their caches?

    typically WordPress generates the HTML, Autoptimize optimizes it and finally a page cache caches the Autoptimized HTML 🙂

    Thread Starter Ambyomoron

    (@josiah-s-carberry)

    I just to need to better understand the order of things when the HTML changes. What, if anything, must I do to ensure that Autoptimize is optimizing the changed HTML before the caching plugin caches it? Similarly, when the caching plugin creates a cached version of a page, how do I ensure that the JS and CSS cached by Autoptimize is referenced by the caching plugin, rather than the unoptimized, uncached versions? Does Autoptimize ensure that it hooks the page save function in such a way to ensure that it always does its work before the caching plugin does its part?

    Plugin Author Optimizing Matters

    (@optimizingmatters)

    Autoptimize always acts on HTML before the page caching plugin thanks to e.g. output buffer magic, tested & confirmed working in a lot of page caching solutions 🙂

    1. WordPress creates the HTML with the unoptimized CSS/ JS
    2. Autoptimize optimizes it, removing CSS/ JS references if/ where needed, creating the optimized CSS/ JS if not in cache yet and injecting links to the autoptimized cached resources in the HTML
    3. The page cache receives the autoptimized HTML and caches that

    If your page changes the page cache will know and the request will hit WordPress, starting the process over again.

    Give it a try, it works great like that and is likely more fun then having to read this explanation 😉

    Thread Starter Ambyomoron

    (@josiah-s-carberry)

    Thanks for the clarification

    Plugin Author Optimizing Matters

    (@optimizingmatters)

    you’re welcome, feel free to leave a review of the plugin and support here! 🙂

Viewing 7 replies - 1 through 7 (of 7 total)

The topic ‘Why is performance better without Autoptimize’ is closed to new replies.