So I have kept up with the messages in this thread via email, however I haven't taken the time to go here and reply until now. Since that was a while ago, I have a lot to reply to. Mostly on account of me either being busy at the time, not feeling like it, or forgetting that this thread exists here and that I needed to write a reply like this here.
To be completely honest here, I had said there was going to be no r3dfox 133 in the past, and well there was no "r3dfox" 133. It also coincided with moving to a new codebase for the UI, so it also made sense to do it for that reason as well. I also wanted like a third branded browser separate from r3dfox and ESR, and I actually still use the Plasmafox build to this day as that third browser.
I'm not sure why I didn't say this. Normally I check the commit history, and use that to make the release notes. Maybe I just didn't remember making any major changes and just skipped that process idk.
I know the Firefox password system is weird, but that's not something I'd have touched here. I assume it was a browser bug, and being so long ago and no other report, I assume it was either a one-off thing or got resolved in upstream. I will say though in all my browsers that I use, I've never had this happen, Except in Chromium where it nukes your profile if you look at it wrong, but that's an entirely different engine.
Same here. My PC is pretty good and modern because of pretty much just compiling. I got a 5950X when AM5 was out, and it was on discount for being last gen and old socket. GPU is nice to have for gaming, but like the realistic cap for me for a while will be a 3080 Ti/3090, and even then I had to downgrade to a 1080 from my 2080 Ti and that does fine for the games I want to play in the here and now. Outside of that it's just faster CPU for faster compile times and more RAM for more stuff open.
My main phone is going to be 9 years old this year, and my main laptop is going to be 13 years old this year. Around 8 or so months ago I was using my laptop as my main machine and remoting into build machine to make r3dfox, and albeit a little slow, it served me completely fine for that time. I only went back to main so I could play BG3 at a smooth framerate with decent graphics. Maybe with the exception of maybe soon upgrading to a phone that is a few years newer (6 years old this year), I don't see myself replacing either with something new before I upgrade my PC again. Likely won't be until AM6 comes out and I upgrade to whatever the last AM5 16 core part is that year, and probably 96GB or 192GB RAM, and repeat for next socket. (Assuming I'm alive then and I have the money, both of which I'm not sure of how likely they are.)
This is how the Eclipse website is designed, it looks nice in modern browsers, and it displays fine in even ancient browsers. Like the only major issues are old browsers don't grab the custom font, and there's not really a good fallback for the textbox background, so it's just overall lighter shaded. The only other downgrade I can name off the top of my head is the logo image requires a pretty recent browser to change on hover, not even Firefox 115 can do it. It uses the :has selector, and I don't know of any other way to implement it without using JS.
I do want to port the website to an even older layout that will display the same from IE3 up to latest browsers, and I have the base framework done for it, but I just haven't ported all of the text into it. I might even make it smart and serve the legacy one for http and the modern one for https.
The forum is similar, but the user interface might actually be better for older browsers, however it requires the user to manually switch to the legacy UI, I don't think automatic switching is possible.
I really hate this method of checking the browser. Like all they give is a non-descript website is not going to work on this browser f*** you, and there's no obvious way to force it to load. I had to block the discourse script and do other workarounds for other sites when I was using Mypal68 as my main.
For me, I'm not the most gentle, but I try not to be careless. I've had multiple parts break over multiple phones, and unless it was such an old phone where getting another used was more economical than getting the required parts, I've repaired my phones every time. I assume the problem here is phones have gotten harder to repair and parts are more expensive. I also have no idea where I would even go if I wanted someone other than me to repair a phone. Back to parts, I've never been able to take an OLED screen off of a phone without obliterating the screen, while my main LCD phone I went at with 0 heat and a razor blade and it was completely fine. On top of that, OLED screens are much more expensive, like 5x or more the price of older LCD phones, and it gets to the point where it's almost more worth it to get another or a newer phone even if used it's like $200+ because the screen costs that much.
I would say the same for laptops, but that's a bit different and I've had like a bunch of sidegrades for xyz reasons. I had a MSI GS60-2QE with a 4820HQ and 970m, but I got tired of dealing with Optimus and two RAM slots, so I got HP 8770W with a 3630QM and upgraded it with 32GB and 980m, until the 980m died and that was a $250 part and I was tired of lugging the 8770W around and on top of that MSI decided f*** you you only get 4GB of available RAM, so I got a T430 and put the 3630QM in it, but then I got tired of Intel integrated graphics being barely able to run games, so I got a M4800 with a 4810MQ, and upgraded the GPU to a M2200 and display to a 1080p 120hz panel, and here I am now. I almost want to get a 8570W and transfer over the M2200 due to weird (I assume) BIOS related issues with the M4800, but the 8770W also had weird issues that I assume the 8570W would also have, and screen upgrade is rare to get the required parts or expensive, so eh. I'll just use this until the higher end P5x/P15 Gx Thinkpads become cheap and grab one of them. (P15G2 with A5000 is still $2k wtf)
Meanwhile here, if I had the money, I literally don't know of any "new" phone or laptop that I would want to get. I'd just get what I've researched as myself wanting either because it's a better experience or just faster. The newest phone I'd want is 3 years old this year, and even then idk if I'd want to use it as main, and the P15G2 is 5 years old this year.
As an American, this sounds so absurd. A lot of places here don't take $100 bills, but like that is it. A lot of places, even if they don't take $100 bills, give you a discount for paying with cash, or charge extra for using a card. It literally costs you more to use a credit card, and it costs them more to take a card payment rather than cash.
What had happened was I had gotten the feeling that the website and repo may have had the possibility of being attacked and/or spammed, and I was very tired and wanted to sleep, so I just locked everything down so I could sleep. I think I either forgot or was too busy to think about reopening them though until I was already awake for a while though, which is why they were locked for like ~20 hours if I remember correctly.
... and I'll add Vista SP2 to the OS mix above ; yes, close to "stock" Firefox but with the ability to launch on older WinOSes! This is what most site admins expect, this is what most extension authors expect and target...
I was never part of the "extreme web privacy" crowd to demand a change of route towards Librewolf (or similar forks); I understand a small portion of the LW code was needed to address a specific r3dfox technical issue, but that is different to incorporating large chunks of LW code "while we're at it" ...
As if it wasn't enough to deal with Mozilla "breaking" things (and locking down the browser) with each major version update, an "average" r3dfox user has to deal with "r3dfox-specific" changes, too (ones that not always meet with said user's "approval") ...
And my own words on DRM/EME:
I found r3dfox maintainer's "obsession" about DRM simply "blown out of proportion"; he goes to extreme lengths to disable EME at buildtime, but the browser itself provides an easy way to disable EME at runtime, if one objects to it for whatever ideological reason...
Let's face it; with Google practically owning the Web, they have leveraged the use of their own CDM (Widevine) in most media services, even the most obscure, but still free, ones... Yes, I totally understand the argument about "black-boxed code" etc., but DRM has become a necessary evil in the web era of 2025 and beyond...
A lot of focus has been put on the VMP (Verified Media Path) requirement associated with the majority of the prominent/commercial DRM'ed Video+Audio services (e.g. Netflix, Disney+, Paramount+, Spotify, Tidal, Apple Music, etc.) as a reason NOT to implement DRM on r3dfox (because VMP entails a very large sum of money, paid to Google, for certifying the browser for VMP purposes), but what about the rest of the lesser known services that don't impose VMP with DRM? Jody's example of
https://www.cp24.com/now/
is such a case, there are many others...
r3dfox originally started as mostly just Firefox but for 7/8, but then I wanted to make it so I had to change less to set the browser up for myself later, and then https://github.com/e3kskoy7wqk/Firefox-for-windows-7/ came to exist, and that fulfilled the idea better than I could have. So I started making the browser more in my image. I believe all the changes I make are beneficial for the normal user, and where realistic, I try to add options to change or revert new behavior or styling that I can see as being potentially unwanted in the settings. I would rather you use r3dfox since well it is my fork and I feel like I need a reason to exist ;-;, but like you don't have to. I've been kinda in the "extreme web privacy" crowd, but I run r3dfox announcing to every website that it is r3dfox which is a rare fork, so I definitely have a unique fingerprint, but I do all of this and run privacy extensions and tweaks idk why I'm like this.
DRM being removed is part of the same reasoning, I don't use anything that uses DRM, and although they do exist, not many services work with non-VMP or give you a good experience with non-VMP Widevine. Since I'm not using it, I don't need the DRM and I feel easier about it if my browser comes with no closed source DRM at all, so I disable it. On the Eclipse forum, they've devised methods to re-enable DRM even after my changes, and that's probably the way to go if you want DRM in r3dfox rather than me having to ship it to everyone.
So, they use different profile directories, and profiles are weird. I think just renaming the r3dfox profile directory to r3dfox ESR would be the best way to do it, but I'm not sure if that works, or if that still requires manual intervention in about:profiles to set the old profile as default. I just use portable and that accepts only one profile, and won't try to make a new profile.
Also the browser should have the code to import settings from even IE6, but due to well not having an IE6 profile or the will to make one, I have no idea if it even works. It was basically only ever in the browser because it technically broke importing from more modern IE and it is technically related to 7 support. It's still in the browser because (I assume, I don't handle the base 7 support anymore) it has practically no maintenance cost, so it's like a why not.
I don't know what to do here, I like the idea of including uBO by default, and if possible, I would like to make uninstalling it actually uninstall it and not have it come back. However as far as I know that's not possible. So when it comes to do I make both my experience in testing better (not dealing with ads in a fresh profile or take the time to manually install uBO), and make the experience of the layman who may not even know what uBO is better, vs not including uBO. I'm going to take the choice of including uBO. I don't know what to do here. I feel like a decent ad blocker is the bare minimum for a usable web experience.
I don't know of any good way to add uBO on first run, but let the user uninstall it without editing policies.json. Unless there is some other way I just don't know what to do here.
I feel like that would be just as time consuming as manually grabbing it, because I would have to manually replace the policies.json every time, and I'd have to keep a separate version of it up to date which I know I will forget to do. Also I would go about testing r3dfox, not grab uBO, see ads, get annoyed, then get uBO. I'd need to preemptively apply the uBO installing policies.json, which I also know I will forget to do.
In that case, I'm not going to ship something I won't use. If uBO gets enshittified, I'll replace it with something else. The only way that scenario could play out is if "Poogle" or someone gave me millions to sell them my browser, something which nobody will ever pay millions for. If that somehow happened I'd see that as a no-brainer easy money, and I would still need a decent browser for myself, so I would just make a new fork, or even sit back and pay someone to develop it for me and continue there. (Or I guess it could happen if I die, r3dfox gets discontinued, and then enshittifed uBO, but in that case r3dfox will be outdated and something better will exist and you would probably just want to use that instead anyways.)
I want the AI to replace me. When it gets to the point where AI can handle the process of making the browser and do a better job, maybe this could happen. I have no idea how it could work though, or if I will see hardware capable of hosting a local AI able to do all these steps within my lifetime.
How I make r3dfox, is I import e3k's patches as commits into the next version branch, then I cherry-pick over the r3dfox changes. If a conflict is encountered, I check the commit to see what it's changing and what commit to the browser resulted in the conflict. If it's simple, I just adapt the old commit to the new code, or if it's complex I have to revert the bad commit. Which since I'm cherry-picking code over is a pain. I can't just cancel the cherry-pick there, cancelling it will revert all the applied commits, which may also include previously resolved conflicts. So I have to upload the changes, cancel the cherry-pick, which will go back to the previous branch I went to cherry pick from. (Also multiple times I've had the cherry-pick just silently fail mid-way through undoing everything I've done.) So I have to switch to the new branch, and then pull all those changes from the uploaded branch, then revert given commit because if I start by reverting the commit that will f*** up the tree and require merge, then go back to the previous branch, select the rest of the commits to cherry-pick over and continue. Then I have to go and test everything, and if I have an error, I have to dive into that and that's it's whole own thing depending on the component. Mozilla has been f****** with Brotli lately, and I was able to fix it before, but why the seemingly simple changes in 147 broke it to the point where I was unable to compile with it hurt my head so much that I just dropped the Brotli compressed omni.ja idea.
Doing nightly means doing this process more often, and I just don't have that in me. Also even for like updating to point releases or updating the ESR which normally is easy just rebase current branch or pick the changes to new branch and build, except Mozilla has made breaking changes in 140 ESR twice in a row now, I just don't feel like doing it or running that on my PC at the time, so the builds get delayed. The long gaps between versions at times are because I just don't feel like I want to make a new build and I don't feel like it's worth it to force myself to update it yet.
When was this? I did the Vista compatibility mostly on my own because I finally felt decent and I wanted to get closer to XP. Vista is an obvious target to get on the way, and it was pretty close, so I went for it.
Also I don't think you can just write your way around the compiler. You would also need to replace the current compilers, something Mypal68 did. I never figured out how to set things up for Mypal68 to compile, so I have no idea how to make a similar setup for r3dfox.
I feel like compiler would be the main starting point, and/or get one step closer to XP by attaining Vista RTM compatibility by either bringing back layers and/or removing the WebRender dependency on DWrite, and then getting GDI and/or Cairo mode to work and not be super unstable (may also be WebRender related).
Now these are outside my current skillset and are open issues. Maybe I could figure something out finally out of sheer will sorta like how I got Vista support in the first place, but currently I feel like s*** and at this rate there's a higher chance of me blowing my own head off than things getting better enough for me to feel not like s***.
So that's all the replies I have for you all. I'll try to reply to the thread more often rather than making one big post after a number of months in the future, but I have no idea how that will go.
.jpg.c64c1204584c92cda024c0e0a7b33494.jpg){kind=tracking}
.thumb.jpg.b251015fc522288a9691c6db583793fd.jpg)
I wanted to compile the driver and I decided your one might be better to do this with. However I noticed your link doesn't exist anymore and I thought I had downloaded it, but I can't find it. Can you reupload pls.
.jpg.d863fbd8b1067921ccb9866a5c2776e8.jpg)
