My Browser Builds (Part 5)

[Leokids123]

5 hours ago, basilisk-dev said:

Which operating system are you asking about? Windows or Linux?

I have not looked into this for Windows, as I am not a modern Windows user(I do use roytam1's Serpent builds on a Ryzen 3600X running Windows 2000 though).

I was already looking into doing this for Linux last week. I had to patch some code but I was able to get a successful build that works on a Pentium MMX and newer. The oldest hardware I have is a Pentium 3, it ran fine on that machine.

http://dl.basilisk-browser.org/basilisk-20260305163630.linux-i586-gtk3.tar.xz

I want to start making our x86 releases support 10.5+, right now we are stuck at 10.7. The contributor who does the Mac builds has an issue with his build environment where he was not able to do a build targeting OS X 10.5 or 10.6. It was producing SSE related errors which is strange because even the oldest CPU that can run OS X 10.5 has SSE support.

We did introduce PowerPC builds in the last release if that's something that interests you. You could run them in Rosetta, you might get Google.com to load after 1-2 years of waiting for Rosetta to run the browser.

Any but i'm thinking about Windows. Although Linux is okay.

 

Speaking about some website is broken under Pale Moon and currently Basilisk,have you implemented a polyfill to  fix CSS Nesting?

[Leokids123]

13 minutes ago, Leokids123 said:

Speaking about some website is broken under Pale Moon and currently Basilisk,have you implemented a polyfill to  fix CSS Nesting?

This is specifically about Scratch,cause for some reason they implemented CSS Nesting,which isn't implemented by Firefox before 115esr or 117 (by default) and Chrome before 109 or 112 by default.

 

They claim to support browsers of the past three years...so if that's the case they shouldn't be using that...unless you enable it yourself?

[basilisk-dev]

51 minutes ago, Leokids123 said:

Speaking about some website is broken under Pale Moon and currently Basilisk,have you implemented a polyfill to  fix CSS Nesting?

Eventually I will look into it if no one else on the UXP dev team steps up to try to look into it. I did read last week on the Pale Moon forum that someone was writing a XUL extension that un-nests nested CSS, but I can't find the post right now.

Right now my priorities are:

* Full ES2023 compliance (eventually 2024/2025 but now ES2023 is priority)
* Make Tailwind v4 playground render correctly (this will fix a lot of sites not just sites using Tailwind)
* Performance improvements
* Make WebRTC actually work with modern services
* Fix any other sites that I find that are broken

If I am being completely honest with you, I have been a bit frustrated because I have felt that the other UXP developers have prioritized things other than site compatibility, and that those other things they are prioritizing are the wrong things to prioritize.

I frequently see forum posts where users are told that the CSS or JS feature a site uses is bad and that the user should email the site asking them not to use those features because they are bad. Users don't and shouldn't have to do that. People don't care about what CSS or JS a site uses, all they care about is that the site they want to use is broken. They'll just go use another browser instead.

Due to that I've been learning C++, learning how our JS and CSS engines work, and I'm doing what I can to change that in order to keep UXP/Basilisk/Pale Moon/roytam's forks relevant.

[roytam1]

@basilisk-dev thanks for taking up #2229 for top-level await.

I wonder if you have ideas on implementing https://repo.palemoon.org/MoonchildProductions/UXP/issues/2473 ?

[VistaLover]

55 minutes ago, basilisk-dev said:

I have been a bit frustrated because I have felt that the other UXP developers have prioritized things other than site compatibility, and that those other things they are prioritizing are the wrong things to prioritize.

... Hence my previous comment you quoted  :

10 hours ago, basilisk-dev said:

On 2/22/2026 at 7:45 PM, VistaLover said:

i.e. upstream (MCP) would have to implement that support (I'm not holding my breath, though ) ...

... and I was strictly referring to "Moonchild Productions", are you officially part of that team now?

Joking here , of course, but my comment was made on Feb 22nd, your fix landed on the official UXP repo on Mar 10th (and has yet to land on a roytam1-compiled binary), so yes, I wouldn't have survived had I held my breath...

Jokes aside, my contribution was just to identify the missing JS feature; the report was made here by a UXP-fork user (on an unsupported OS), so the proper way to tackle that would've been: 

1. Get access to a machine with an official UXP sanctioned OS
2. Get hold of official UXP-based browser binary/binaries
3. Reproduce the "bug" (rather the platform's deficiency) there
4. Register an account in the official PM forum
5. Report the problem in their WebCompat subforum

(few fork users on XP/Vista can/are willing to go through the above ...)

... and then hope one member of the under-staffed MCP team takes it upon themselves to attempt a "fix"; given that the problem, as reported, affected a Polish government web site (relevant probably only to Polish citizens), MCP staff would have little incentive to spend time on fixing that, when other major/more popular sites are broken... So yes, over the years, I've become somewhat pessimistic ... 

Thanks again for implementing "object.GroupBy"  .

Regards.

[basilisk-dev]

1 hour ago, roytam1 said:

@basilisk-dev thanks for taking up #2229 for top-level await.

I wonder if you have ideas on implementing https://repo.palemoon.org/MoonchildProductions/UXP/issues/2473 ?

I haven't taken a look at how to implement this in our codebase yet, but Mozilla's implementation is written in Rust unfortunately so we will have to do a new implementation from scratch. Happy to work together with you to get an implementation figured out, you need this for your XP builds just as much as we do upstream.

43 minutes ago, VistaLover said:

... and I was strictly referring to "Moonchild Productions", are you officially part of that team now?

I have no idea if Moonchild Productions is just Moonchild or if there are other UXP/Pale Moon contributors involved in Moonchild Productions as well. I've never asked. In my case, Basilisk Development Team is just me because nobody else seems to be interested in contributing to Basilisk outside of the occasional PRs from other UXP developers to ensure breakages don't occur after UXP updates.

[Guest raddy]

4 hours ago, basilisk-dev said:

Set security.tls.hello_downgrade_check in about:config to false.

This affects upstream as well.

This does not help..

ps. palemoon-28.10.7a1.win32-git-20260307-d849524bd-uxp-2a33ab0e2c-xpmod.7z

[schwups]

An useragent override should work.

try e.g.:     general.useragent.override.distrowatch.com     Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:133.0) Gecko/20100101 Firefox/133.0

[Guest raddy]

31 minutes ago, schwups said:

An useragent override should work.

distrowatch banned windows xp, you can change only OS number - Windows NT 6.1 instead 5.1.

[Leokids123]

13 hours ago, basilisk-dev said:

Eventually I will look into it if no one else on the UXP dev team steps up to try to look into it. I did read last week on the Pale Moon forum that someone was writing a XUL extension that un-nests nested CSS, but I can't find the post right now.

Right now my priorities are:

* Full ES2023 compliance (eventually 2024/2025 but now ES2023 is priority)
* Make Tailwind v4 playground render correctly (this will fix a lot of sites not just sites using Tailwind)
* Performance improvements
* Make WebRTC actually work with modern services
* Fix any other sites that I find that are broken

If I am being completely honest with you, I have been a bit frustrated because I have felt that the other UXP developers have prioritized things other than site compatibility, and that those other things they are prioritizing are the wrong things to prioritize.

I frequently see forum posts where users are told that the CSS or JS feature a site uses is bad and that the user should email the site asking them not to use those features because they are bad. Users don't and shouldn't have to do that. People don't care about what CSS or JS a site uses, all they care about is that the site they want to use is broken. They'll just go use another browser instead.

Due to that I've been learning C++, learning how our JS and CSS engines work, and I'm doing what I can to change that in order to keep UXP/Basilisk/Pale Moon/roytam's forks relevant.

Yep! Considering web compatibility is key to you,you might one day enable Widevine DRM to enable watching protected media,even if it's 480p and not HD...when Pale Moon does not...

 

While WebExtensions isn't planned,do you wanna implement that one day?

Edited March 11 by Leokids123

[basilisk-dev]

Just now, Leokids123 said:

Yep! Considering web compatibility is key to you,you might one day enable Widevine DRM to enable watching protected media,even if it's 480p and not HD...when Pale Moon does not...

That would be a very large amount of work. I understand why users want it but at the moment it isn't anywhere near the list of top priorities. Personally, I view the lack of DRM in the browser as a feature and not a bug, but I could see why other users might not feel this way.

1 minute ago, Leokids123 said:

While WebExtensions isn't planned,do you wanna implement that one day?

Is it something I want to implement? Yes. Am I going to do that? Not any time soon.. If I did do it, the implementation would be Manifest V2 only because Manifest V2 is a stable target and a lot of Firefox WebExtensions still use Manifest V2. For example, the Firefox version of UBlock Origin uses Manifest V2.

[anton12]

basilisk-dev said:
"I do use roytam1's Serpent builds on a Ryzen 3600X running Windows 2000 though"

Please,can you tell which version of KernelEx for Win2000
you use ?

[Leokids123]

6 hours ago, basilisk-dev said:

That would be a very large amount of work. I understand why users want it but at the moment it isn't anywhere near the list of top priorities. Personally, I view the lack of DRM in the browser as a feature and not a bug, but I could see why other users might not feel this way.

Is it something I want to implement? Yes. Am I going to do that? Not any time soon.. If I did do it, the implementation would be Manifest V2 only because Manifest V2 is a stable target and a lot of Firefox WebExtensions still use Manifest V2. For example, the Firefox version of UBlock Origin uses Manifest V2.

I get it. I view the lack of DRM as a bug,not a feature,but i get why you view it as a feature.

 

And WebExtensions? Sure but...

 

I do think you should focus on web compatibility rather than focusing on what Pale Moon is doing,it would be bad to prioritize AVX instead of web compatibility,as it'll break support for most CPUs and Chrome/Firefox whose latter Pale Moon was forked with supported SSE2 and SSE3,something Pale Moon stopped supporting,except for its 32 bits build...

 

 They don't care if they use a certain feature,all they care is that it's broken,CSS nesting should have been implemented regardless of if it's draft or not...

Edited March 11 by Leokids123

[basilisk-dev]

14 minutes ago, Leokids123 said:

I do think you should focus on web compatibility rather than focusing on what Pale Moon is doing,it would be bad to prioritize AVX instead of web compatibility,as it'll break support for most CPUs and Chrome/Firefox whose latter Pale Moon was forked with supported SSE2 and SSE3,something Pale Moon stopped supporting,except for its 32 bits build...

I think you misunderstood something here. Web compatibility is my number one priority right now. I already said that earlier in the thread.

Basilisk does not, and will never, require AVX for release builds.

[Leokids123]

2 hours ago, basilisk-dev said:

I think you misunderstood something here. Web compatibility is my number one priority right now. I already said that earlier in the thread.

Basilisk does not, and will never, require AVX for release builds.

Yep...

 

Do you plan on supporting CSS Nesting,something scratch somehow requires but UXP doesn't require?