>>>IMPORTANT: These binaries are provided by anyone who are willing to build and submit them. Because these binaries are not necessarily reproducible, authenticity cannot be guaranteed. For your consideration, each download page lists the GitHub user that submitted those binaries.
My current position is to step wide and rather carefully around anything by/from Google.
Rather than take such statements at face value, here's an example of one such instance that probably hurts its credibility: https://calpaterson.com/mozilla.html
Also, they're not exactly a world apart from Google in some respects either, for example: https://killedbymozilla.com/ vs https://killedbygoogle.com/ (though i guess the amount of projects also illustrates the difference in their sizes)
However, that's not to say that we even have many viable alternatives to Firefox either: https://batsov.com/articles/2021/11/28/firefox-is-the-only-a...
Personally, i think that we're past the peak of Firefox and are currently in the midst of it fading away, which i expect will come to its conclusion in the following decade. Apart from that, we basically have just a bunch of browsers that are all similar to Chromium.
Essentially, Chrome will have probably lived long enough to kill IE by becoming the next IE.
Who knows where the web standards will go when there will only be Google backed browsers: https://mozilla.github.io/standards-positions/
One day they added a small widget to buy crypto in the new tab page. That was the breaking point for me. Like it happened with Firefox ads, I started to understand that Brave was working against me too.
It's also like a few clicks to turn off brave settings. It's less setup for brave than I go through on any new video game settings lol
But all we got is Linux on the desktop...
The only reason Linux is popular on servers is because it has huge corporate sponsorship. Big players figured it it's more beneficial for their bottom line to commoditize the backend infrastructure and invest in OSS. Same is happening with developer tools.
It's not happening with anything consumer facing because everyone wants to take their angle at capitalising on the market. I can't think of quality user facing OSS that isn't aimed at developers or isn't trying to capitalise on the users somehow. Scratch that - blender is the only one I can think of - these guys are an amazing exception.
The only reason they aren't worse than Google is that they've been so shit at capturing the market share they don't get to make the same kind of moves, but Bing/Edge are not good alternatives for Google/Chrome.
I'm not one of those people that screams "M$ evil boooo", I actually use .NET day-to-day on a current project. But they have a really shit track record on Edge, it's the first browser I would avoid.
Microsoft is making great strides in open source and Linux. A decade ago, you wouldn't have believed MS would publish SQL Server for Linux, release something as great as VS Code, create Windows Subsystem for Linux (and Android soon), acquire npm and github and make them better rather than destroying them, etc.
[1] https://news.ycombinator.com/item?id=28975856
I gave up on Edge because it didn't support uBlock origin on Android, I view bundled add blockers as just another way for the browser to block competition while still serving you their own "respectful" ads.
[1] https://www.xda-developers.com/microsoft-edge-buy-now-pay-la...
Okay.
Unlike this project, which is very much from "internet randos"
such a well-known governance that keeps putting back telemetry and ads when users dont want them.
As for ads - where? I haven't seen any and I've been using Firefox on multiple platforms for many years.
Even if it did, we're again back at the point of "X is a spectrum". "evil" is a spectrum which, according to you, includes anonymous crash reports and then goes on to include serial killers and war criminals. Even then, anonymised usage data used to guide development is far less evil than collecting highly personal and identifying data and using it for targeted psychological manipulation to get you to buy things.
I get it, Mozilla isn't the poster child FLOSS org that we'd like it to be, but it's still orders of magnitude better than the competition.
No, crash reports are primarily data about the program, with very little if anything about you.
"Reflections on Trusting Trust": https://www.cs.cmu.edu/~rdriley/487/papers/Thompson_1984_Ref...
And even if you go Full OpenBSD, you're still relying on some chip microcode and BIOS that you bought sight-unseen.
So, unless the Free Software Foundation staffs its own fab staffed by angels, the best we can do is minimize risk.
A more practical depth would be to bootstrap with GNU Mes, which is a source based bootstrapping path, that begins with a tiny scheme interpreter (~5000 LOC of simple C) and a C compiler written in scheme that are mutually self-hosting.
These tools can compile a slightly patched version of TinyCC that is self-hosting. Using this C compiler you can bootstrap a bunch of gnutools (glibc, binutils, gcc) and using these tools you can bootstrap a full Guix Linux distro.
stage 1: build from source using your existing compiler
stage 2: build again from source using the compiler you just built
stage 3: (just for checking): build a stage 3 compiler from the stage 2 compiler. Verify that it is byte-for-byte identical.
extra stages for the paranoid: repeat the process with a different starting compiler, as unrelated as possible. Again, verify that the output is bit-for-bit identical. You then know either that you're good, that the compiler reflects the source, or both starting compilers have the same devious hidden hack, perhaps courtesy of Ken Thompson himself.
GNU Mes achieves that paranoid step: once you verify with GNU Mes that you get the same bits, you don't really have to do this ever again, because you can certify that the compiler you get is reproducible from source and has no tricks.
How did you come to this conclusion when they stated that they'd rather use Firefox? Or are you saying Mozilla are the internet randos?
Sounds a little crazy not to trust them.
If you want chromium without Google there are several options such as Edge, Opera, Brave and probably a dozen more I've never heard of.
On the other hand, I simply don't trust internet randos, much less to install their binaries.
And it seems like a third option of Firefox w/ uBlock is less bad than either of the above.
/nerdsnipe
https://build.opensuse.org/project/show/home:ungoogled_chrom...
https://github.com/Eloston/ungoogled-chromium/discussions/14...
https://github.com/ungoogled-software/ungoogled-chromium-arc...
Worth it until the next "update", I think.
Despite that, it takes 4-8 hours to compile, due to two factors: 40,000+ separate targets, of which most are C++ files (famously slow to compile anyway).
And yes, 40,000+ separate targets is no exaggeration; ninja's first run claims about 18,000 targets, and after it finishes those, it generates a second ninja file that has to compile 23,000 targets.
I personally think it's worth it. Others may disagree, and I can't blame them. I have to run it overnight.
It is really hard to keep up-to-date with security patches if you need to compile every few days from source with that time..
Edit: Also 32GB RAM which matters
Edit: Also, 3900 is 3900X running on eco mode (same spec but half power). You might get a lot more performance by tuning BIOS settings.
I'll take a peek at the BIOS, but I'll also no good at doing such things.
I wonder how it compares to the build time of something with equivalent functionality, like firefox.
EDIT: Looks like popularity is on the rise [1]. Hopefully it will make it into the community repository.
[1] https://pkgstats.archlinux.de/packages/ungoogled-chromium
For a project with frequent point releases, building Chromium via GitHub might require a monetary commitment.
edit: "integrated email client" Ok now I see why ;)
> Note that, of the three layers above, only the UI layer is closed-source. Roughly 92% of the browser’s code is open source coming from Chromium, 3% is open source coming from us, which leaves only 5% for our UI closed-source code
That changes the equation quite a bit.
[1] https://vivaldi.com/blog/technology/why-isnt-vivaldi-browser...
I've used Vivaldi for years but now I'm getting split on it. I've got regressions that are just unacceptable (especially when it comes to video playback), but at the same time it comes with features I can't live without.
Essentials gives you the very basic stuff, classic gives you some extra features, and fully loaded gives you the built in mail client and all that other stuff.
[1] https://flathub.org/apps/details/com.github.Eloston.Ungoogle...
Here is the most recent build logs for x86_64 https://flathub.org/builds/#/builders/12/builds/8123 which I found from the GitHub green-tick status icon on the latest commit on https://github.com/flathub/com.github.Eloston.UngoogledChrom....
Personal experience, most of the source I download from GitHub either will straight refuse to compile in Visual Studio, or requires me tweaking a lot of the dependencies and stepping through the warnings and errors until it does.
"Ah, but don't use Visual Studio" you say. "Use XYZ-compiler-de-jour". Right, that's now in the 'too-hard/too-time-consuming' bucket, and I'll look for something else.
Those would be on a nix (linux/bsd/other) free source OS with distro specific binaries (or rolling their own and would be better off compiling it themselves anyways)
I don’t think this is true for macOS? the builds seem automated…