foxPEP and TenFourFoxPEP - For A Faster Firefox / TenFourFox

[z970]

Made a new "Tips for X" section exclusive to this forum to make life better.

I love making life better.

 

[z970]

OK folks, foxPEP 2.1 Humility is out. Release notes can be found in its entry in the Releases page @ GitHub ...

 

[dextructor]

Hi @z970mp

I'm finally testing foxPEP on my A1139 with VoidLinux (5.7.15_1 glibc) and so far both Arcticfox(27.10.1) and SpiderWeb(2.2.2) presented some errors.

With SpiderWeb if I use ("layers.acceleration.force-enabled", true) the colors will be a mess (see the screenshot)

In Arcticfox if I use ("gfx.content.azure.backends", "direct2d1.1,skia,cg,cairo") the program crashes without starting (see crash below) but if I use on this specific preference the value "cg,cairo" works fine.

[32660] ###!!! ABORT: Could not initialize mScreenReferenceDrawTarget: file /home/matt/Development/Arctic-Fox-27.10.1/gfx/thebes/gfxPlatform.cpp, line 521 [32660] ###!!! ABORT: Could not initialize mScreenReferenceDrawTarget: file /home/matt/Development/Arctic-Fox-27.10.1/gfx/thebes/gfxPlatform.cpp, line 521 Segmentation fault

Since I aways copy the prefs.js to the destination folder and both of the browsers had some issues I tested every single entry on the about:config to isolate the problems but some lines wasn't present until I manually inserted on the prefs.js file. Below are the specific lines there wasn't present in each browser.

Arcticfox

gl.multithreaded, true layers.acceleration.enabled, true gfx.webrender.all, true layout.frame_rate.precise, true datareporting.policy.dataSubmissionEnabled, false media.navigator.enabled, false media.peerconnection.enabled, false network.http.referer.XOriginTrimmingPolicy, 1 privacy.firstparty.isolate, true) privacy.trackingprotection.cryptomining.enabled, true privacy.trackingprotection.fingerprinting.enabled, true network.dns.localDomains

SpiderWeb

gl.multithreaded, true image.mem.max_ms_before_yield, 0 image.onload.decode.limit, 0 layers.acceleration.enabled, true nglayout.initialpaint.delay, 0 gfx.webrender.all, true browser.tabs.animate, false browser.urlbar.delay, 0 datareporting.policy.dataSubmissionEnabled, false experiments.enabled, false media.navigator.enabled, false media.peerconnection.enabled, false privacy.firstparty.isolate, true privacy.trackingprotection.cryptomining.enabled, true privacy.trackingprotection.fingerprinting.enabled, true social.remote-install.enabled, false social.toast-notifications.enabled, false

There is any problem with manually inset any if the aren't present in the original? So far the performance and fluidity on the SpiderWeb exceed my expectations. I hope that some information could be useful to the future of the project.

 

[z970]

@dextructor Although neither Arctic Fox nor SpiderWeb are officially supported browsers for foxPEP (for the reasons you discovered), I applaud your diligence and experimentation to attempt a working state in addition to compiling this information.

Arctic Fox does not launch with 'Skia' clarified as a content / canvas backend because the browser has been compiled without any support for Skia. And as for cg, it sees it as an unknown and instead falls back onto Cairo. The backend "cg" stands for Core Graphics, a native rendering library exclusive to OS X (which is faster than Cairo, but slower than Skia). If you are going to make a custom rendition for the aforementioned browsers, then it serves no place there.

I believe SpiderWeb does not support hardware acceleration due to an endianness anomaly, but I'm not totally 100% on that. In that case, that is a library / compilation issue.

Yes, looking4awayout and I encountered this problem in PowerUOC's (foxPEP's) infancy. He chose to try fighting an uphill battle with The UOC Patch and came up with the Enforcer idea to counteract this 'stubborn preference' issue, while since I was committed to the all-in-one installation model, there was nothing I could do, so I had to end up trusting the browser that it was overriding the patch for a reason. Turns out, I was right. Later down the line, I realized that the browser does not make certain changes as according to the patch most of the time because A), it is compiled without support for that preference due to a version misalignment or operating environment dependency, or B), the given altered preference has already been set that way as the browser default.

Several examples:

Apparently, gl.multithreaded is an OS X-exclusive preference (I have yet to personally confirm this).

layers.acceleration.enabled is a vanilla Pale Moon-only preference, and is only included to fully turn hardware acceleration on there (it won't work with just force-enabled).

gfx.webrender.all is a feature inherited from Servo (R.I.P.), and is only relevant for Quantum-derived (FF57+) codebases.

I believe fingerprinting and cryptomining protection are features (again) native to Quantum-derived (FF57+) codebases.

However, there are still sometimes some outliers that continue to confuse me (not necessarily unborn out of my own ignorance) why the browser won't acknowledge, such as:

browser.urlbar.delay, which shouldn't typically depend on version or environment, unless it was either implemented in a newer version than what AF / SW are derived from (unlikely), or were removed outright.

nglayout.initialpaint.delay, which is a basic preference and has been around in the Mozilla codebase for a long time and should not have been unrecognized, again unless the feature was for some reason removed outright.

media.peerconnection.enabled, which controls the state of WebRTC (a security hazard), and should have been a known component of the Mozilla codebase for a similarly long time, unless the browsers were simply compiled with them disabled by default. But in this case (knowing @wicknix), the browsers were compiled with them disabled by default.

And network.dns.localDomains in AF I'm just bewildered by, as that also should have been a basic and universal component in the Mozilla codebase and forms the entirety of what is essentially foxPEP's built-in ad blocker. I have no guess as to why that might not be there, outside of the possibilities I've offered above.

However, given the above, I'll leave it up to you to decide whether unrecognized entries should still be manually added or not. And performance was better on SpiderWeb because that allowed through the Bloat Trimming Engine, unlike Arctic Fox.

-

As to the future of the project ... there is a high chance that Release 2.1 Humility will be the second to last release of foxPEP. Not only do I believe the patch has at this point in time reached its apex in effectiveness and performance, now leaving little room for improvement, I will also soon be getting less time to focus on development (contrary to my prior belief) in favor of other more promising projects, in addition to more pressing subjects (like life). Plus, no one has once sponsored (donated) the project thus far, which I'm sure you can imagine naturally decreases the incentive to keep going.

And Mozilla's current health does not help either, especially considering that 90% to 95% of the Servo team (the products of which foxPEP capitalizes on) has just been removed from the company. - And that's to say nothing of Firefox's rapidly decreasing market share, especially since this new event happened.

-

Anyway, I hope this was useful to you.

 

[dextructor]

As to the future of the project ... there is a high chance that Release 2.1 Humility will be the second to last release of foxPEP. Not only do I believe the patch has at this point in time reached its apex in effectiveness and performance, now leaving little room for improvement, I will also soon be getting less time to focus on development (contrary to my prior belief) in favor of other more promising projects, in addition to more pressing subjects (like life). Plus, no one has once sponsored (donated) the project thus far, which I'm sure you can imagine naturally decreases the incentive to keep going.

And that's to say nothing of Firefox's rapidly decreasing market share, especially since this new event happened.

It's great to have such nice folks like you around here sharing some knowledge. I know that many projects don't have enough support (either proper reports to help the development or not least important donations to keep motivating the developers involved) and that's sad to see, but thanks so far to make this little piece contributing to make the web a nicer place. I hope that your other projects (life included) could be as awesome like your contributions to this community.

The current state of the technology it's very sad and the browser applications reflect it with more bloated engines that require more ram that we could even imagine. It was nice the Mozilla Firefox early versions as an option but even today doesn't look like an real alternative because Mozilla's vision.

 

[z970]

foxPEP on Reddit: Take II

https://www.reddit.com/r/waterfox/comments/ies256
-

EDIT: Will someone please enlighten me as to why this project can't seem to get any traction with the general public because none of these people will actually bother to witness the end result for themselves because they don't think I know what I'm doing? Because I sure as hell am missing their collective point, I'll tell you that much. It most certainly does not compromise security, or stability, and neither convenience just for a negligible performance boost. When the browser leaks the client's internal IP address by default (courtesy of WebRTC / Mozilla's evidently bad vetting processes), or allows Joe Average's website to hijack the device camera and microphone without telling anyone, and foxPEP turns those exploits OFF, how in the hell is that making it less secure, or "convenient", of all things?

I suppose then, that this heavily "marketed" effort (or rather "little set of browser settings") will simply remain obscure and exclusive to use cases of TenFourFox and the odd insane person like myself who actually knows the end result of what it does, and thus uses it on everything else as well because it WORKS. Better yet, maybe it should have been kept as PowerUOC to fulfill our little niche and simply left at that - after all, God forbid we start allowing ISPs to inject liquid malice into adspaces because we actually dared to so much as utilize the browser's built-in hosts file to its full potential!

Nerve of some people. I'd like to see these folks package something half as well with just a third of the amount of thought that this took...

 

[CooperBox]

foxPEP on Reddit: Take II

https://www.reddit.com/r/waterfox/comments/ies256
-

EDIT: Will someone please enlighten me as to why this project can't seem to get any traction with the general public because none of these people will actually bother to witness the end result for themselves because they don't think I know what I'm doing? Because I sure as hell am missing their collective point, I'll tell you that much. It most certainly does not compromise security, or stability, and neither convenience just for a negligible performance boost. When the browser leaks the client's internal IP address by default (courtesy of WebRTC / Mozilla's evidently bad vetting processes), or allows Joe Average's website to hijack the device camera and microphone without telling anyone, and foxPEP turns those exploits OFF, how in the hell is that making it less secure, or "convenient", of all things?

I suppose then, that this heavily "marketed" effort (or rather "little set of browser settings") will simply remain obscure and exclusive to use cases of TenFourFox and the odd insane person like myself who actually knows the end result of what it does, and thus uses it on everything else as well because it WORKS. Better yet, maybe it should have been kept as PowerUOC to fulfill our little niche and simply left at that - after all, God forbid we start allowing ISPs to inject liquid malice into adspaces because we actually dared to so much as utilize the browser's built-in hosts file to its full potential!

Nerve of some people. I'd like to see these folks package something half as well with just a third of the amount of work that this took...

I'm pretty sure that for every detractor there must be 100's if not 1000's of very satisfied users (myself included) who appreciate the time and very hard work that has gone into the project. It will indeed be a sad day when it falls by the wayside unless some other bright and dedicated person can take over the reins. Again, kudos for all associated with keeping our PPC's and Intel machines in tune, and as speedy and safe as possible.

 

[z970]

@CooperBox I have not discounted this scenario of someone else taking over as project lead. In fact, it's one of the reasons why I list out rules and development guidelines in foxPEP's GitHub codebase, so that someone may in the future quickly assimilate what I have amassed in these almost 12 months of the project's lifespan.

Thank you, sir.

 

[Raging Dufus]

foxPEP on Reddit: Take II

https://www.reddit.com/r/waterfox/comments/ies256
-

EDIT: Will someone please enlighten me as to why this project can't seem to get any traction with the general public because none of these people will actually bother to witness the end result for themselves because they don't think I know what I'm doing? Because I sure as hell am missing their collective point, I'll tell you that much. It most certainly does not compromise security, or stability, and neither convenience just for a negligible performance boost. When the browser leaks the client's internal IP address by default (courtesy of WebRTC / Mozilla's evidently bad vetting processes), or allows Joe Average's website to hijack the device camera and microphone without telling anyone, and foxPEP turns those exploits OFF, how in the hell is that making it less secure, or "convenient", of all things?

I suppose then, that this heavily "marketed" effort (or rather "little set of browser settings") will simply remain obscure and exclusive to use cases of TenFourFox and the odd insane person like myself who actually knows the end result of what it does, and thus uses it on everything else as well because it WORKS. Better yet, maybe it should have been kept as PowerUOC to fulfill our little niche and simply left at that - after all, God forbid we start allowing ISPs to inject liquid malice into adspaces because we actually dared to so much as utilize the browser's built-in hosts file to its full potential!

Nerve of some people. I'd like to see these folks package something half as well with just a third of the amount of work that this took...

Not sure what to tell you dude. Your work is greatly appreciated here, as evidenced by this and another stickied thread promoting same, and the multitude of other threads on this board where your efforts have been recognized and lauded. I think it boils down to differences in the respective audiences.

First off, those Pale Moon people, despite the excellent product they put out, are a bunch of grouchy curmudgeons. The higher up you go, the grouchier they seem to get. Just read around their forums to see what I mean. They've sat on their Mac build forever, refusing to make it official despite having discussed doing so since practically the beginning of the project...and what really, really gets me about the state of their Mac build is that they blame it on past developers who've put their best efforts into the project for, as far as I can tell, nothing...except to be abused in absentia as the supposed cause of the lack of an official Mac build. One of those former developers is, like yourself, a beloved contributor here, who regularly pumps out browsers and Linux stuff for PPC, clearly has a solid work ethic, and is just as clearly devoted to his craft. I can't speak for him and won't identify him here because I have no idea how he feels about any of that; but when I saw what a particular higher-up in the Pale Moon project had to say in their Mac forum about former Mac developers, I was aghast. I don't know what their problem is over there, maybe it's just that they're Windows users which is enough to piss anyone off. But whatever it is, they're toxic IMHO.

As for Reddit, man, what a cesspool that place is. I go there too, because there are some - some - good open-minded, congenial and knowledgeable people there. But I've yet to come upon a subreddit that, no matter how narrow its focus, did not have its share of trolls, idiots, troublemakers, chronic critics and/or twelve-year-olds. You'll get it all, the entirety of humanity, on Reddit. Since people generally suck, it's not a place I'd go looking for affirmation or even constructive criticism.

Also, consider the very nature of your product. Despite how easy you've made it to use, it's still only noticeable to people who have an interest in tinkering. Nobody else will even go looking for such a thing as foxPEP; the vast masses are just going to go get newer hardware/OS'es and/or switch to Chrome when FireFox slows down. People using Waterfox, Pale Moon, etc, are the edge cases who will even bother searching such alternatives out. They're tinkerers, at least to a degree. The problem with tinkerers is that we all have different opinions as to what we think is best for our own hardware/OS/use case. So you're marketing a one-size-fits-all solution to an audience that is by its very nature individualistic. Many will appreciate what you've done, but there will always be many who do not, and many who believe they know better than you.

Your own satisfaction, and the gratitude of the people you've helped, will probably have to be enough for you in the end. Expecting anything more may lead to frustration, as you're experiencing now. So FWIW, thanks for all you've done here, and I for one hope you continue to contribute.

 

[z970]

@Raging Dufus I realize all of that. Perhaps I should have labeled my comments as a mere rant for clarity - or even so, practiced better judgement with what to make public in the first place (this frustration was not entirely sourced from Reddit). I suppose it's just another day on the Internet, right?

I didn't approach Reddit looking for affirmation or approval, and I am not here for that either (even though constructive criticism is helpful to better the end result), but to simply widen the audience of people who may put the product to good use, and nothing more. Effectively, to improve the landscape itself where I think might be applicable. And as it happens, gratitude seems to naturally follow if that is successfully carried out, but in any case, that's not a concern for me. I'd like to stress that's not why I'm here, or there.

But thank you always for your focused inputs. They are appreciated greatly.

 

[Raging Dufus]

@z970mp sorry, I didn't mean to imply that affirmation or approval were your primary goals. Of course you want people to make use of your work, nobody would put that much effort into something just to let it languish. You remind me of my dad, God rest his soul, who was a man of the land. He grew up farming, and despite becoming a top-notch aerospace technician, just couldn't stop playing in the dirt. Every spring he was out breaking ground and planting everything he could've just gone to the grocery store to get. And he always produced far more than we could consume. Just about every day, he'd park a wheelbarrow at the roadside in front of our house so people could take, for free, his watermelons, tomatoes, onions, okra, squash, you name it. For some time after he passed, people who noticed the wheelbarrow was no longer there would stop by and ask Mom about him, and tell her how much they loved and appreciated the food he grew. It wasn't out of need, none of these people were going hungry, they just appreciated the man's artistry and generosity. That's how I feel about the contributions of you and other here who don't have to do any of this, but do it because you have the skills and want to benefit this community. So again, thanks!

 

[wicknix]

This is what @Raging Dufus more than likely saw, although as he said, just browse the forums in general. You'll see the high and mighty a$$hats belittle 90% of their users every day. Anyway, the reason why i quit... (which lead to me creating my own OSX and PowerPC builds for people who actually appreciate it).

PS... Sorry to hijack the thread, but i felt the need to share. Maybe it'll help you not be discouraged as much. If people use it, great. If not, so what? I read crap about how old or slow AF or Remix is all the time. Do i care? No, because 98% of the time my projects are enjoyed by the majority, and that is why it's worth it.

 

[Raging Dufus]

This is what @Raging Dufus more than likely saw, although as he said, just browse the forums in general. You'll see the high and mighty a$$hats belittle 90% of their users every day. Anyway, the reason why i quit... (which lead to me creating my own OSX and PowerPC builds for people who actually appreciate it).

View attachment 947838
View attachment 947839

PS... Sorry to hijack the thread, but i felt the need to share. Maybe it'll help you not be discouraged as much. If people use it, great. If not, so what? I read crap about how old or slow AF or Remix is all the time. Do i care? No, because 98% of the time my projects are enjoyed by the majority, and that is why it's worth it.

Holy cow, that is so much worse than what I saw. Didn't know about that conversation, but they kept on belittling users and "Mac people" after you left. It's little wonder you did.

 

[wicknix]

Yeah, for them wanting to be everything that Mozilla was "supposed" to be, they sure act like the Mozilla devs that they belittle daily. I don't even know how the remaining "volunteer" devs for any platform can stand that environment and forum.

Cheers

 

[z970]

Yup... Once I refuted one of their community members' arguments that it should not be a prefs file, he then (seemingly fresh out of rebuttals) responded with that I "should have consulted Moonchild before advertising these tweaks as compatible with Pale Moon". I then realized that the developers are control-freak pieces of work that think the world revolves around them, as evidenced by the tons of projects and collaborators they've pushed away thus far.

What's more, their advertising is confusing because they've got this motto "Your browser. Your way." with a reputation that you can customize and modify it greatly, yet you are frowned upon when you do exactly that and share it with the world because "it's already been optimized". ... Then what the hell did they mean by having the user "modify" the browser?! I suppose then that it's only "your browser" in a figurative sense, but not a literal one... Even better, maybe Moonchild wrote that specifically for Tobin and forgot to set the changes as private.

Either way, when I saw for myself the reason why OpenBSD has not supported Pale Moon in recent times, I then quickly understood that this was no special case.

Pale Moon Official Branding Violation · Issue #86 · jasperla/openbsd-wip

You will revise your mozconfig located at www/palemoon/files/mozconfig to remove the following: ac_add_options --with-system-jpeg="${LOCALBASE}" ac_add_options --with-system-zlib ac_add_options --w...

github.com

foxPEP only supports Pale Moon (irregardless of what these people think) because many of its users already make use of Pale Moon, it has already been tested, it does not break anything, and it can fully leverage what the browser is capable of doing. Not because of someone's ill-tempered decree.

Anyway, nothing's been hijacked - though I suppose we're all ranting at this point. I do what I do because not only do I create for myself, it also makes wonderful folks like you all here have a better time with what you've chosen to be part of, and as far as I see it, that's just one of the thousands of ways that can give life its charm. Following this philosophy, there are many ways that this can take full shape and transform into something much larger than just an optimized patch script.

But being someone that, pretty much by profession, does not see people every day (especially in these times), I consider myself very lucky to get to have colleagues such as you all, even if only in text form. Therefore, I think it accurate that I say thank you to everyone for the lessons, support, and conversation I've gotten to take part in throughout these past three years plus. For if you know where to look, you can find yourself endlessly grateful for what God and the universe gifts you every day.

 

[alectrona6400]

i'm curious to why tenfourfox isn't supported with this anymore

 

[z970]

@antimony1095 It is unofficially supported as a special exception (hinted at in the Tips for PowerPC Users section), although this isn't stated in the system requirements because TenFourFox is not compiled to leverage GPU acceleration, and neither is OS X 10.4 (or 10.5, to what I understand). Naturally, this takes a toll on the end result that foxPEP enables, so only 10.6 and up is listed as that is the first release that is capable of offloading Web and video rendering to the GPU, lessening the load on the CPU, and thus taking full advantage of what foxPEP puts on offer.

That said, it has now become clear in multiple cases that this support model has caused unnecessary confusion among its users, and in some cases may have even unintentionally decreased its user base. For the sake of clarity, TenFourFox will be re-listed as a supported browser starting with foxPEP's potential final release, Version 2.2 Gratitude, likely becoming public somewhere within one month's time.

In any case, go ahead and install it into TenFourFox regardless. I can assure you that nothing will break.

 

[z970]

Good news all, FPR 27 with its 78 ESR-enhanced libraries fixed the eBay CSS issue. AquaWeb is no longer required as a workaround.

Also, props to Dr. Kaiser for this one (whatever those are). Just as mentioned in the release notes, the newer libraries are not only more secure, but faster too. Pages definitely load and render slightly quicker, scrolling is definitely a little smoother as well, and even the launch times have been noticeably reduced!

As AquaWeb and AquaVid depend on a TenFourFox install to function, this update will benefit them too. Performance boosts ahoy!

 

[rampancy]

I can really relate to what people have said about the Pale Moon forums. I was just reading a thread posted by someone who innocently asked about the difference between Firefox Legacy and Pale Moon, and I was amazed by the toxic haughtiness in the replies. Even the current contributor of Mac builds wasn't spared. Yikes.

Anyway, I just wanted to give a quick shout out to both @z970mp and @wicknix for all of the time and effort they've put into their respective projects. As a pandemic project, I gave away a large chunk of my A1181 MacBook collection to families and parents in my city who needed computers to access online learning. One of the first things I did was pre-install FoxPEP and Firefox Legacy/Arctic Fox on them. Thanks for everything you've done for people like me who are trying to keep old Macs running and useful for others!

 

[z970]

@rampancy foxPEP is not compatible with Arctic Fox due to its non-standard technology implementations, and will cause issues and instabilities if installed there. Therefore, I do hope it was only installed into Firefox Legacy.

Otherwise, you're welcome.

I will take this opportunity to issue a reminder to everyone that as v2.2 will be the presumptive final iteration, now would be a very good time for both those even slightly knowledgeable of Firefox's workings and users noticing nuances or anomalies during browsing to make suggestions on what be included, changed, or removed once more.

 

[z970]

OK everyone ... The final release of foxPEP, iteration 2.2 Gratitude, has been made available for download in the usual place @ GitHub.

The modifications that it applies to the browser should stay relevant to future releases for at least several years, given how current ones to this day contain non-deprecated switches and technologies dating back to Firefox's inception, almost two decades ago (which also explains why foxPEP has been shown to improve the performance of Classilla and Camino regardless, even though they reside far out of curated support).

And when given Mozilla's current internal situation and future roadmap ... well, I'm fairly confident that foxPEP in its present state will remain an effective performance booster for the rest of Firefox, Waterfox, and Pale Moon / UXP's development life, thus invalidating the need for further updates altogether.

-

Now with that said, even though I'm calling it a day, that does not necessarily mean that the product must end here. As previously discussed, I continue to remain open to others taking over project development in the event that they encounter room for potential improvement.

-

Failing that however, I finally offer my sincerest Gratitude to you all for the opportunity alone of not just getting to operate these projects within such a tightly-knit group as this one (itself a rarity in the forums of today), but for making something that has had a pronounced impact for all of us in navigating around the infamously slow modern Web.

Thank you!

 

[MacFoxG4]

Thank you for making FoxPEP!

Tried out the latest version on TFF Intel and it works great. Firefox Legacy, on the other hand, doesn't seem to play nice with it. It worked with FoxPEP 2.1, but 2.2 caused it to randomly quit. I reinstalled 2.1, but the same thing happened. Ended up having to delete FoxPEP entirely and let Firefox Legacy generate its own prefs.js file again.

 

[z970]

@MacFoxG4 Firefox Legacy 68.12 on Mac OS X 10.7.5 works reliably well with 2.2 on my end. What FF and OS version are you using, and for how long did you use it before it crashed?

 

[MacFoxG4]

@MacFoxG4 Firefox Legacy 68.12 on Mac OS X 10.7.5 works reliably well with 2.2 on my end. What FF and OS version are you using, and for how long did you use it before it crashed?

I used the same version and OS. It crashed in less than a minute for me.

 

[z970]

@MacFoxG4 On what website?