My Browser Builds (Part 5)

[NotHereToPlayGames]

I personally would have preferred a WARNING of that web site's CONTENT.  I just leave it at "ouch"...

[Rod Steel]

On 12/29/2024 at 4:06 AM, Leokids123 said:

Modern processors these days support AVX,so that'll not be a problem. The problem is when you have a Celeron,Pentium and even Atom,that typically never supported AVX except in latest gens.

First Generation Core i5 and i7 processors are fast and does not have AVX;

Phenom II X4 and X6 are pretty fast too. You can play even Metro: Exodus on them. Not supporting this CPU's is a bad move.

I am so grateful to roytam1 for New Moon that spare us from that bs!

Edited December 30, 2024 by Rod Steel

[D.Draker]

18 hours ago, UCyborg said:

Holy s***, this site actually wants SSE 4.1.

Most likely, for gaming, or if it's a test site, to test the gaming capabilities.

 SSE4.1 was already included in quads like Q9550 (Launch Date Q1'08). Seventeen years ago!

https://www.intel.com/content/www/us/en/products/sku/33924/intel-core2-quad-processor-q9550-12m-cache-2-83-ghz-1333-mhz-fsb/specifications.html

Non-groovy/earlier/lower/cheaper/weaker models like Q8xxx had it, too.

Who would intentionally buy Pentium, unless it's 1994?

[D.Draker]

Oh, just found, even the ancient Duo from the 7 series had it, too.

Intel Core2 Duo E7500

https://www.cpu-monkey.com/en/compare_cpu-intel_core2_duo_e7500-vs-intel_core_2_quad_q6700

[D.Draker]

On 12/29/2024 at 10:43 AM, j7n said:

Is there any appreciable benefit from AVX? People say AVX was created a long time ago now. But back then there was no gain from paying for an "i3" when a "Pentium" was exactly the same without AVX and hyperthreading that I always saw a bit of a gimmick.

I've never encountered the need for it, neither in browsers, nor in games, that said, I play games from 2020. MK11, for example. Stunning graphics on Vista,

[modnar]

7 hours ago, D.Draker said:

Oh, just found, even the ancient Duo from the 7 series had it, too.

Intel Core2 Duo E7500

https://www.cpu-monkey.com/en/compare_cpu-intel_core2_duo_e7500-vs-intel_core_2_quad_q6700

Whichever chip it is that has the Penryn core i.e. 45nm, has SSE4.1.

[Leokids123]

On Scratch,the debug help popup is broken,it's not like modern firefox or Chromium. The tutorials now crash.

[Mathwiz]

On 12/29/2024 at 12:35 PM, Leokids123 said:

They insist modern browsers are like fully fledged 3d games.... that's why they got avx....

Yeah but PM isn't a "modern" browser, so it doesn't work like a 3D game engine and doesn't really need AVX....

On 12/29/2024 at 2:27 PM, UCyborg said:

WebRender was actually documented to work like 3D game engine.

I had a hard time taking that article seriously after I saw the diagram of a jet engine, with its components labeled "Quantum" this and "Quantum" that.... Talk about hype - and that was from 2017! Is that how we're supposed to "understand" modern Web browsers now? What's next - an "explanation" of Chromium illustrated by a diagram of a rocket?

On 12/29/2024 at 3:24 PM, UCyborg said:

Holy s***, this site actually wants SSE 4.1. https://donoharm.report/

OK, the Web has officially jumped the shark. Web sites are dictating processor architecture now? How the heck is that supposed to work if you have an ARM processor, or one of Apple's new processors, or really, anything but Intel / AMD?

[NotHereToPlayGames]

5 hours ago, Mathwiz said:

Yeah but PM isn't a "modern" browser

While *I* agree with this, I suspect that *most* users will tend not to agree and that in their mind they have "modern capabilities" (a step or two behind but not a thousand steps behind).

One only needs to look at bug reports to see this.  A "modern web site does not work" so-called bug is reported.  It gets "fixed" with the next release or gets re-reported as still being a bug.

Has the developers EVER replied with, "We are not a modern browser, we will not implement the functions required on that modern web site."   ???   ???   ???

 

edit - I ask rhetorically as I do not follow PM bug reports.  Some do get cited here at MSFN and from those, no, I don't ever recall reading the developers reply with "We are not a modern browser, we will not implement the functions required to resolve that issue."

Edited January 1 by NotHereToPlayGames

[UCyborg]

6 hours ago, Mathwiz said:

How the heck is that supposed to work if you have an ARM processor, or one of Apple's new processors, or really, anything but Intel / AMD?

I'm no friend of programming, so I don't know how, but it does. SSE 4.1 requirement of WASM SIMD is specific to x86/x64. I wonder if full blown WASM requirement is default for web based projects on Microsoft's front these days or can it be avoided by some reconfiguration.

That particular website takes about 70 seconds to load on my smartphone that uses a CPU with ARMv7-A instruction set, specifically quad-core Cortex-A7 running at 1,2 GHz. What Intel had over 15+ years ago is kinda irrelevant in my case, all my devices have either AMD or ARM CPU. Just my oldest computer that happens to be my favorite has an AMD CPU from 2009 and AMD didn't implement those SSE 4.1/4.2 instructions until 2011, when they were implemented on Bulldozer-based CPUs along with AVX instructions.

6 hours ago, Mathwiz said:

I had a hard time taking that article seriously after I saw the diagram of a jet engine, with its components labeled "Quantum" this and "Quantum" that.... Talk about hype - and that was from 2017! Is that how we're supposed to "understand" modern Web browsers now? What's next - an "explanation" of Chromium illustrated by a diagram of a rocket?

I think it's just trying to explain it in layman's terms. Software happens to often have silly names / associated terms when you think about it. Not sure what would be "better".

Even if that article is a bit of a hype, do you prefer gaslighting on Pale Moon forum? Every single performance concern is shut down with the likes of "Just use an ad blocker", "You're comparing apples to oranges", "Go back to ChromeZilla, you normie" etc. Doesn't change the reality that UXP based browser is fully usable only if you manage to stick to few select web sites exclusively, otherwise it's a regular freeze, lag, not responding mess.

On 12/31/2024 at 12:22 AM, Leokids123 said:

On Scratch,the debug help popup is broken,it's not like modern firefox or Chromium. The tutorials now crash.

You're better of verifying the issue in the MCP's Pale Moon / Basilisk and reporting on their forum. roytam1 is just random nobody that happens to like compiling browsers for EOS Microsoft's operating systems, reporting most issues here is just useless.

In the opinion of certain individuals on the other side and reworded for my own amusement, a lazy f***er who implements hacky workarounds, doesn't support his work and shifts all the support burden to upstream.

On 12/29/2024 at 10:27 PM, NotHereToPlayGames said:

I personally would have preferred a WARNING of that web site's CONTENT.  I just leave it at "ouch"...

And I would prefer to live in a world without forced genital "surgeries". Is that too much to ask?

Edited January 1 by UCyborg

[NotHereToPlayGames]

4 hours ago, UCyborg said:

Is that too much to ask?

<OT>

My only reference here is that I once had a girlfriend that experienced both and preferred the "surgery".

To the point where she would ask any guy interested in her before even considering "dating".

The 20s (age, not era) was a very different age.  Even more so to be 20-something in today's world.

I wouldn't wish being 20-something in today's world on my worst enemy!  At least not to anybody who carries a MOBILE PHONE.  But I digress...

</OT>

[UCyborg]

@NotHereToPlayGames
Yeah, things like that are one of the reasons I rather not have anything to do with people. I feel profound disconnect even from my own family.

[NotHereToPlayGames]

Profound disconnects are almost always an "internal struggle".  I certainly wouldn't cite MSFN as the proper place to externalize an internal struggle.

Coming from a family with adopted siblings with internal struggles, with uncles and siblings both with addictions, with nephews and neices both with psychological trauma, with family with PTSD from military service, I could go on, I can only add that "I feel your pain".

But must conclude with "MSFN is not the proper place to externalize".

H#LL, as far as that goes, families of this "wide array" are just a "sign of the times" here in 21st century America.  Perhaps 21st century ALL COUNTRIES ON THIS BLUE ROCK.

But I digress...

[modnar]

22 hours ago, NotHereToPlayGames said:

..."sign of the times"... ...21st century...

Those are just phrases that mean little - any time is good time for the truth and problems can then easily be resolved.             /digressing over :-)

[Alahndro]

I've found some useful javascript tune-ups:

The Last days I was forced to use an older machine for browsing while my main rig beside it was busy or unusable due to repairs or upgrades or whatever.

Espescially I wanted to watch some youtube and found it being overloaded by the scripts.
It wasn't even able to play 360p properly. Most  of the time it ended in an endless spinning wheel until  it failed with an error.

That machine is able to play 1080 inside a native player.
- CPU is a AMD sempron3000+ @ 1,8GHz. Single Core
- Version: New Moon 28.10.7a1 (2024-06-28)

It appeared that the javascript engine went completely bonkers.

So I looked into the config to find some settings to play with and toggled the following entries:
- javascript.options.parallel_parsing = false
- javascript.options.wasm_baselinejit = true

This made a HUGE diference!
I don't know what the second option does but for me it is obvious that the parallel pre-loading and parsing and compiling overloads the CPU, precisely the caching of data and scripts.

This parallel processing is helpful if you have at least 2 cores and some L3 cache, but it utterly destroys the pipelining of single-core CPUs.

Of course this raises the latency a little bit, but it forces to process pages and scripts in a more serialized way.
Now I can even switch the playback to 720p inside youtube, though with some framedrops but it can reach this level

I also once found a site with palemoon x86 SSE2 and non-SSE2 compiled as deb-packages for linux. (Got dual-boot on this machine)
Here it is much worse, obviously these miss some single-core tune-ups that roytam includes in his builds.
Need to investigate that further