YouTube journey on 18yr old computer

[NotHereToPlayGames]

So below are my five oldest computers.
The Intel Atom N450 (Acer Aspire One, 16yrs old, barely used, worst purchase ever!) is officially going to Goodwill.
The Intel Celeron N3050 (Acer Aspire X, 11yrs old, freebie) is also officially going to Goodwill.

The Intel Core2 Quad Q6700 is undergoing a small journey.
I'm unsure of the depths of that journey at the moment.

I installed Q4OS (based on Debian Linux) on the Intel Core i5-6300U.
I did Plasma instead of Trinity, but eating up 6.5GB RAM out of only 8GB was a "H#LL NO!" and it got uninstalled.
But that basically tells me that it would be a WASTE OF TIME to install Q4OS (or even Zorin) onto the Q6700 computer.

So I have narrowed my options to Win7 Enterprise and Win10 LTSB 2016.
I'm looking to compare/contrast YouTube "watchability".

 

 

[NotHereToPlayGames]

Today's "find".
Darik's Boot and Nuke 2.3.0 (DBAN) - secure 3-pass "wipe" of hard drive to ensure that no data can be recovered after donating it.

The POC Acer Aspire One showed ELEVEN HOURS for a "wipe" using Active@ Boot Disk's KillDisk (was just over 2hrs for the Acer Aspire X).
DBAN is 43min in and shows 4hrs 33min remaining - using DoD Short 3-pass option.
Largely, of course, due to the collosal POC nature of the Acer Aspire One.

https://en.wikipedia.org/wiki/Darik's_Boot_and_Nuke
https://sourceforge.net/projects/dban/

Edited 4 hours ago by NotHereToPlayGames

[anton12]

Hello NotHereToPlayGames,

MSFN certainly is no forum for Linux experts.

Nevertheless here are some charts for you.

Contrast these charts with M$ Windows11 !!

(By Gary Sims & Andy Walker)

Edited 5 hours ago by anton12

[NotHereToPlayGames]

I may revisit Q4OS.  I actually suspected from the get-go that I made a *MISTAKE* when *not* selecting the TRINITY desktop.

I did not screencap, but Q4OS ate up 6.5 GB of only 8 GB RAM.

That was on a DIFFERENT computer, but still, that's one helluva lot of RAM consumption just for the desktop!

[NotHereToPlayGames]

I opted to install 10 first.
Admittedly, I fear that to be a mistake because I will very likely end up with 10 so I should have done it last.

[NotHereToPlayGames]

In 10, my test-case browser is not going to be the same as my test-case browser when I install 7.
Mainly because 10 has MORE AVAILABLE and 7 will be "limited" to something like e3kskoy7wqk or r3dfox.
While some may consider this to not be "apples to apples", that is not my goal here, I want to know if 7 or 10 is "better" for YouTube.
And that will determine which OS *stays* on this 'junk hardware'.
Should still be "very close" (e3k-needs-easier-name has a HIGHER version of Chromium as I write this).

Edited 4 hours ago by NotHereToPlayGames

[NotHereToPlayGames]

My wi-fi is pretty fast.  Yeah, I'm only on a 500 Mbps ISP but my wi-fi connects anywhere from 300 all the way to the full 500.
Two TV's are currently streaming news so that does bring me down a tad during these screencaps.
At any rate, this is PLENTY FAST ENOUGH for YouTube "watchability".

[NotHereToPlayGames]

1 hour ago, anton12 said:

Contrast these charts with M$ Windows11 !!

Here is my Win11 --  only 0.9 GB of RAM used at startup, but yeah, it did bump up to 1.0 when I opened the "about" screen.

That seems very much in-line with the "more-capable" versions of Linux.

"Apples to apples" here is 11 with all AI removed since Linux doesn't have these AI features, so not really "fair" to compare Linux-without to Win-with.

And yeah, it's a GOOD THING that there is no AI and the end-user doesn't have to jump through hoops to remove/disable.

Edited 4 hours ago by NotHereToPlayGames

[NotHereToPlayGames]

Test video at 2160p, VP9, no extensions - https://www.youtube.com/watch?v=KLuTLF3x9sA
Laggy as H#LL (obviously), all cores pegged to 100%.

[NotHereToPlayGames]

CPU drop at change from 2160p/vp9 to 1080p/av1.
No extensions.  No lagging, but no CPU overhead for multitasking.

[NotHereToPlayGames]

1080p but avc1 instead of av1.
Some "breathing room" for cpu multitasking, but not much.

 

 

"h264-ify" userscript:

 

// ==UserScript==
// @name - YouTube 02 - Workaround and H264ify
// @version 2025-02-27
// @match https://www.youtube.com/*
// @match *://*.youtube.com/*
// @match *://*.youtube-nocookie.com/*
// @match *://*.youtu.be/*
// @run-at document-start
// @grant none
// ==/UserScript==

(function () {
    "use strict";

    const h264ifyEnable = true;
    const h264ifyBlock60fps = false;

    const originalGetContext = HTMLCanvasElement.prototype.getContext;
    HTMLCanvasElement.prototype.getContext = function (contextType) {
        if (contextType === "webgl" || contextType === "webgl2") {
            console.log("WebGL is disabled by Tampermonkey");
            return null;
        }
        return originalGetContext.apply(this, arguments);
    };

    if (!h264ifyEnable) {
        return;
    }

    function override() {
        var videoElem = document.createElement('video');
        var origCanPlayType = videoElem.canPlayType.bind(videoElem);
        videoElem.__proto__.canPlayType = makeModifiedTypeChecker(origCanPlayType);

        var mse = window.MediaSource;
        if (mse === undefined) return;
        var origIsTypeSupported = mse.isTypeSupported.bind(mse);
        mse.isTypeSupported = makeModifiedTypeChecker(origIsTypeSupported);
    }

    function makeModifiedTypeChecker(origChecker) {
        return function (type) {
            if (!type) return '';

            var disallowed_types = ['webm', 'vp8', 'vp9', 'av01'];
            for (var i = 0; i < disallowed_types.length; i++) {
                if (type.includes(disallowed_types)) return '';
            }

            if (h264ifyBlock60fps) {
                var match = /framerate=(\d+)/.exec(type);
                if (match && parseInt(match[1], 10) > 30) return '';
            }

            return origChecker(type);
        };
    }

    override();
})();

 

 

Edited 1 hour ago by NotHereToPlayGames

[NotHereToPlayGames]

CPU overhead gained by using 720p over 1080p.