Performance optimizations?

[UCyborg]

Huh? My Pale Moon is at 780 MB when that page is opened along with this forum, one short thread on old Reddit, another forum and a video playing on YouTube (using PROJECT VORAPIS frontend). Maybe everything on Windows just shows as using more memory if it's available? Or do you have millions extensions in both browsers? I have 6 GB of RAM total.

Idle consumption does seems to indicate like it could matter, but I'm not 100% certain as my only other reference is work laptop with 12 GB of RAM which I haven't setup myself from scratch.

Though that Pale Moon's reading I took is from Lubuntu 20.04, I'll check Windows 10 later, though I don't expect it to be much different.

[suzyne]

I guess I can say that my observation was equally casual, which is why Windows Task Manager was good enough for the point I was making? It is a particular gripe of mine that screen captures of browser resource use without any explanation of the open tabs is like showing a chart without the unit of measurement shown on the axis.

And with today's computers is 28 processes, really a shit ton?

Is my modern CPU going to "run out" of them because Edge requires that many?

If a chromium-based browser uses a different design from Pale Moon, but loads fast and runs well, comparing process counts feels unnecessary or like meaningless point scoring. I am reacting against what I see as petty Edge/Chromium/Whatever bashing.

Huh? My Pale Moon is at 780 MB

I turned off the ad blocking on both browsers, so it is the core browser was being compared not the add-ons or extensions. Maybe that explains the difference result that you got?

[UCyborg]

Just thought that memory consumption is abnormally high. Here's mine:

I don't use other browsers much on my computer TBH. Maybe I'd be more distressed about memory usage then. I only upgraded from 4 GB to 6 GB of RAM last year.

[suzyne]

For sure, the memory usage is abnormally high on Pale Moon and Edge. I noticed the high memory consumption after spending several minutes reading the content and scrolling a fair bit as I considered the list. In other words, a somewhat typical visit to a page.

I wonder, did you visit the page, scroll to the bottom and immediately do the measurement? Maybe there is a difference in what happens as more time is spent there?

Interestingly, I also see that your CPU use is very different from mine, where Pale Moon is giving it a good workout when I am on the page. I visited again this morning and here is a more complete Task Manager list, for what it's worth, whether inaccurate or just a vaguely general ballpark figure.

Yours and my results do have curious differences, but my original point was only the importance of stating the actual open tabs in any discussion about browser memory consumption, and nothing more. It doesn't bother me that Pale Moon and this URL don't get along.

[UCyborg]

I wonder, did you visit the page, scroll to the bottom and immediately do the measurement? Maybe there is a difference in what happens as more time is spent there?

Yes to first. You also have 52 processes in Edge. I have about 20 with 10 extensions enabled. My Edge is still at version 94 though, newest Chromium browser I have is at version 117 (Thorium). The picture there is similar. I have a hunch it's something specific on your system.

I also don't run any antivirus/antimalware, haven't run one in over a decade. These probably snoop inside any process they can get.

[Eduardolucas1]

I honestly doubt people are really checking in full depth the use of memory in blink and gecko browsers compared to UXP and pale moon. Because the difference is abysmal. And the browsers that spawn processes DO NOT free memory within time, because even the process killing solution they found likely don't work anymore, as its so much javascript garbage collection failure that it starts to accumulate on the most vital processes that it usually does not kill (which may indicate modern javascript or its implementations in modern sites is very memory unsafe, because, imagine if some exploit in their electrolysis sandbox happens, which is very usual and frequent?). If it wasn't from that, the browser would just free this memory from the main processes for allocation, or, it tends to hold so much more crap that it forces to hold more memory virtual addresses to itself until its able to clean it up, by design.

Pale moon after the last update frees memory normally, although some base address memory stays allocated, but now, restarting the browsers takes DAYS instead of hours. the multiple spawning process browsers do not do that. The high memory usage is usually on sites which likely clearly ignore pale moon user agent and try to fingerprint it, throwing polyfills and backup code, such as reddit, tiktok, discord and youtube. Some others refuse altogether to run and accept to be incompatible since forever, such as whatsapp web, and leak (you can see it in about:memory) its attempt of creating a database by workers and not even the ghostbuster is able to clean its leak, or fail if you don't have a graphics driver to load (not by requiring webgl but by simply giving up due to not following the maintream rendering engines unstandard behaviours) such as telegram. But these things very clearly indicate the terrible software development and architecture standards behind these "online web service front-ends", not from pale moon itself

My performance is very ok even in SSE2, but the thinness of the OS below it helps dramatically on that. The hardware too. I use an AVX2 capable CPU. But i do not need AVX2, again due to the lightness of the OS. Maybe depending on too heavy OSes may demand a better hardware...