Microsoft Has First Major Impact on Chrome

Posted on August 17, 2019 by Paul Thurrott in Microsoft Edge, Google Chrome with 50 Comments

Thanks to the open-source nature of Chromium, Microsoft has had its first major and positive impact on Chrome, Google’s web browser. Thanks to a feature request from Microsoft, Google will issue a change to Chromium, the open-source project by which Google makes Chrome, that significantly improves battery life.

“Today, media content is cached to disk during acquisition and playback,” Microsoft’s Shawn Pickett explains in his change suggestion for Chromium. “Keeping the disk active during this process increases power consumption in general, and [it] can also prevent certain lower-power modes from being engaged in the operating system. Since media consumption is a high-usage scenario, this extra power usage has a negative impact on battery life. This change will prevent the caching of certain media content to disk for the purpose of improving device battery life for users.”

And Microsoft knows battery life. Aside from being the makers of the most popular desktop operating system on which Chrome and other Chromium-based browsers are run, it also spent several years optimizing battery life in its previous versions of Microsoft Edge. And then it would publicize the results, in which classic Edge routinely outperformed the battery life in Chrome and other browsers.

And Google’s on board. For now, the change is being tested as an experimental feature in Chrome Canary—the nightly builds of Chrome 78—which needs to be enabled by default: Just open chrome://flags and search for “Turn off caching of streaming media to disk.” (This works in Chrome for Windows, Mac, Linux, Chrome OS, and Android.)

And if all goes as well as expected, it will be implemented and enabled by default in the browser.

Tagged with

Join the discussion!


Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Comments (50)

50 responses to “Microsoft Has First Major Impact on Chrome”

  1. RonV42

    The sad thing is once it comes to Chrome the all the bloggers outside of this site will give Google credit for it.

  2. jac38

    So, where is cached media going to be recorded? This seems like an obvious change to conserve battery life. What specifically changed?

      • sevenacids

        In reply to Havoc:

        If this is the case, I wonder why they didn't do it like that in the first place. I mean, this possibly dates back when RAM was rare, but in 2019 it should be no issue for an average PC to cache media content to RAM, and only fallback to disk when the amount of memory is limited or consumption gets too high. But I don't see the innovation in here, this is nothing but a quite simple optimization.

  3. CajunMoses

    If there's anything that history tells us, it's that a good, healthy, deep-seated, intense suspicion of anything that Microsoft brings to the table is always more than warranted. Distrust aside though, this contribution isn't necessarily a trojan horse. Could I possibly be more fair-minded?

  4. Dan

    Wow, innovation here.

  5. ipaulmagu

    Oh wait!

    Has anyone determined how much additional/reduction of drainage you get ?

  6. oddgentleman

    So, does it cache to ram? Or simply no cache?

  7. dontbeevil

    LOL they needed to wait for MS help, after many years they couldn't fix their s**t

  8. skane2600

    This appears to be a feature agreement between the two companies. It doesn't really have anything to do with open source since Google is the one making the change.

    • Paul Thurrott

      In reply to skane2600:

      It has everything to do with open source. Microsoft never would have adopted Chromium or recommended this change otherwise.

      • skane2600

        In reply to paul-thurrott:

        There's really little difference between MS agreeing to make Chrome as their default browser vs making a new browser based on Chrome code. The conventional wisdom would be that Microsoft would never do either one, but here we are.

    • jrickel96

      In reply to skane2600:

      It does because Microsoft added it to Chromium. Google does not have to take all changes made to Chromium and apply them to their Chrome browser. Microsoft could execute this and add it to Chromium without Google choosing to add it to their browser.

      Ultimately it really is just Google recognizing that a change Microsoft made is a good one. Reality is they likely realized it would look bad for them if the Chromium Edge did much better for batteries than Chrome. Because of Chromium powering Edge, things like that strengthen Edge's use case in the future when it is the default browser because it will work anywhere Chrome does.

      Microsoft's move to Chromium can pay huge dividends for users whether they adopt Edge or it forces Google to pay attention to issues with Chrome that Edge fixes - such as memory usage. I've run Chrome and new Edge alongside each other and there is a noticeable difference in memory usage between the two, but not in functionality. New Edge has become my daily driver on my PCs and on my Mac. Very impressed with the rendering, memory usage, and just overall goodness of where it is right now.

      • skane2600

        In reply to jrickel96:

        Nothing in the story here indicates that MIcrosoft made the change.

        • datameister

          In reply to skane2600: Nothing in the story here indicates that MIcrosoft made the change.

          Go gack and read the story then. You missed some critical parts. In fact the first two paragraphs should be enough to find a few places in "the story here" where it indicates Microsoft made code changes and submitted them to the open sourced Chromium project.

          • skane2600

            In reply to DataMeister:

            "Thanks to a feature request from Microsoft, Google will issue a change to Chromium"

            Please quote the part of the story that says Microsoft made code changes. Submitting code to an open source project isn't a feature request.

            • tboggs13

              In reply to skane2600:

              I do believe a feature request can be accompanied by code. From the logs on the feature request linked in the article, it does appear that Shawn is submitting code and members of the Chromium project are reviewing and accepting it.

              • Paul Thurrott

                In reply to tboggs13:

                It's almost like it's an open source project. :)

              • dmjb

                In reply to tboggs13:

                What you are describing is called a _pull_ or _change_ request. A _feature_ request is where you ask someone else to implemenent a feature for you.

                It does indeed look like they submitted a patch to the Chromium project, so the use of "feature request" in the article is misleading.

    • JoePaulson

      In reply to skane2600:

      So the Linux kernel isn't open source because the maintainers don't accept every commit from other

  9. Pbike908

    I am anxiously awaiting the global switch that stops HTML 5 video from playing. I downloaded the edge chromium the other day and if I could stop videos from autoplaying I would make it my daily driver.

  10. truerock2


    Not caching content provides Copy Protection, provides additional ways to track users, increases network bandwidth requirements and decreases playback performance!!!


    Oh - and my PC doesn't use a battery :-(

    • acemod13

      In reply to truerock2:

      Boy, you don't understand how caching works at all. As the change implies, it won't save temporally on the storage device. This won't wreck the life of the SSD and won't drain the battery. Most likely, it will use RAM for caching. RAM is heck of a lot faster than even the SSD, so playback in low-end system will actually improve. It won't affect the Tra king nor the copy protection (HDCP says hello).

    • fishtacos

      In reply to truerock2:

      "Not caching content"

      This article is about adding the ability to cache content, not the opposite.

      "provides Copy Protection"

      Neither caching nor not-caching provides any additional Copy Protection as all DRM media is already encrypted and can be accessed the same way in RAM and disk.

      "provides additional ways to track users"

      Neither caching nor not-caching provides any additional way to track users. It's simply a performance difference, which ultimately leads to greater battery life (as RAM cached data doesn't access the disk as much).

      "increases network bandwidth requirements"

      Neither caching nor not-caching increases network bandwidth usage or lowers it. It's simply a method of storage of the same downloaded data.

      "decreases playback performance!!!"

      No it doesn't. Perhaps in 1995 with 400MB 5400 RPM IDEs, but not in 2019.

      "Oh - and my PC doesn't use a battery :-("

      Cool, so this doesn't apply to you. Go away with your BS and fud.

      • truerock2

        In reply to fishtacos:

        Quoting the article:

        "This change will prevent the caching of certain media content to disk for the purpose of improving device battery life for users."

        My degree is in computer science. I have decades of experience in building and evaluating information technology. I stand by my evaluation.

    • antrenorul

      In reply to truerock2:

      If only there was a flag that you could use to toggle the feature off... Oh, wait...

  11. pepesilvia

    *in its previous versions of Microsoft Edge

  12. MacLiam

    I get the battery life advantage, which is less important to me than it probably is to others, but under this change isn't there now the disadvantage of possible stuttering or even paused playback?

    Somebody please explain if I missed something. I may be one espresso shy of getting all my neurons up and running.

    • Scott8846

      In reply to MacLiam:

      From what I understand here, there is a distinction between caching and buffering and what you are referring to is buffering, the act of keeping, say, the next minute of the video in memory to avoid problem if the connection can’t keep up for a little while.

      I think what Chrome's caching do is more akin to download the video to disk so you can, say, go back in the video without needing to get the same content again over the network.

      I don’t know if it’s the plan, but I would suggest keeping an in-memory cache of the previous minute and a buffering of the next minute to meet almost every scenarios, except of course the one where the user go back more than a minute.

      I hope I provided an appropriate perspective on what *could* be going on.

  13. shmuelie

    Sadly not in Edgium (Latest canary as I write this)