MotionMark test

[roman6626]

Why in the test MotionMark 1.1 the results of Waterfox are much less than those of Basilisk and Pale Moon? Pale Moon - 80,48 Basilisk - 66,98 Waterfox - 36,67 Very small values for: Multiply, Canvas Arcs, Canvas Lines. Repeated many times. The results are the same. It struck me. Why is that?
All the while I thought that Waterfox is a more media browser and should work faster than other legacy-browsers.

[grahamperrin]

Waterfox

Waterfox Classic 2019.10 or Waterfox Current 2019.10?

[roman6626]

It's only about Waterfox Classic 2019.10!

[grahamperrin]

MotionMark 1.1 | WebKit ▶ https://browserbench.org/MotionMark1.1/

For Waterfox Classic the shortest answers (to close this issue) are probably:

• compare with with Firefox 56.0.2
• be familiar with Mozilla's Project Quantum.

Historically: Waterfox 56.0 was based on Firefox 56.0.2.

Most recently: Waterfox Classic 2019.10 is the successor to Waterfox Classic 56.2.14, is still essentially 56-based.

Recommended reading

Waterfox 55 Release (Windows, Mac, Linux and Android) | Waterfox Web Browser

Waterfox 56 Release | Waterfox Web Browser

Quantum Up Close: What is a browser engine? - Mozilla Hacks - the Web developer blog referred from A good ELI5 guide for Quantum! … : firefox

Inside a super fast CSS engine: Quantum CSS (aka Stylo) - Mozilla Hacks - the Web developer blog referred from #332

[roman6626]

But Pale Moon and the Basilisk are based on even earlier versions of Mozilla.
I am now talking about the comparison of these browsers, and not with Quantum.
For some reason, I always thought that Waterfox should work faster in a modern web than Pale Moon and Basilisk.

[roman6626]

Is there anything that distinguishes Waterfox for the better from Basilisk?

[grahamperrin]

Is there anything that distinguishes Waterfox for the better from Basilisk?

Yes, but such things are unrelated to MotionMark.

[roman6626]

Of course not connected. But this test shows the performance of the graphics engine. And this for some reason is not for the better at Waterfox.

[grahamperrin]

For me on FreeBSD-CURRENT:

• Basilisk was never an option, and I don't expect a port
• New Moon a.k.a. Pale Moon is no longer available (port maintenance issues).

Food for thought

Those results are not intended to be accurate – I didn't bother with full screen, some memory was swapped to disk, and so on – but I imagine that they're unusual.

YMMV. Vary massively :-)

[roman6626]

I just use only Windows. :-)

[roman6626]

Why does Waterfox have 1 position in this test (Multiply) is always 1?

[MrAlex94]

Do you have e10s and hardware acceleration enabled? https://support.mozilla.org/en-US/kb/performance-settings#firefox:win10:fx60

…

On Mon, 4 Nov 2019 at 13:18, roman6626 ***@***.***> wrote: I just use only Windows. :-) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1248>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABECQWBXG4PRRUIKUULAZKLQSAOKLANCNFSM4JGUBG3A> .

[roman6626]

Hey! Tried and so and so. The result is not affected. By the way, Pale Moon and Basilisk do not have multiprocessing. But they always have this value much greater than 0.

[grahamperrin]

Why does Waterfox have 1 position in this test (Multiply) is always 1?

I don't know, but it's not always so. Please see below, where the test result was 72.95.

Pale Moon - 80,48 Basilisk - 66,98 Waterfox - 36,67

Below:

• Pale Moon 70.02
• Waterfox Classic 82.91

[roman6626]

Does it depend on the inclusion of multiprocessing in Waterfox or not? There is no multiprocessing at Pale Moon.
On different machines, the results are always the same. Is Waterfox always smaller? What could be the reason?

[roman6626]

Checked a few more times. It does not depend on multiprocessing. As before, the value of Multiply = 1! Why?

[roman6626]

I can guess, maybe the Pale Moon and the Basilisk are better optimized for older cars?
(Core i3 M370, 2,40 GHz, 8 Gb)
Could this be the reason?

[grahamperrin]

Mozilla bug 1553575 - MotionMark 1.1 - test "Multiply": scores with Firefox are inconsistent refers to:

• https://browserbench.org/MotionMark1.1/developer.html?test-interval=30&display=minimal&tiles=big&controller=ramp&frame-rate=50&kalman-process-error=1&kalman-measurement-error=4&time-measurement=performance&suite-name=MotionMark&test-name=Multiply

– 49.03 on a reportedly medium screen then a few minutes later, with the same browser session (a heavily extended profile) and the same test in a popped-out maximised window on a 1,920 x 1,080 display:

– 1.0 on a reportedly medium screen.