The bug seems to be specific to the Devuan GNU+Linux installer. The graphical Debian GNU/Linux installer (Debian Bookworm 12.1 Netinstall) does not have that bug.

The installer step "Select a keyboard layout" could be the easy solution, but it isn't working:

In the Devuan daedalus installer, it first asks about the keyboard and gives two selections: "PC-AT ..." and "keep settings" (or something like this). When I select the first entry, the installer step fails and I get the "Installation step failed" message. In the Devuan chimaera installer, the installation step fails instantly when I select it.

I'm planning to add Gemini support to MoeNavigator, my multi-protocol web browser written from scratch (including its engine):
https://codeberg.org/moenavigator/moena … /issues/23

It may take some time though until Gemini is supported.

Hmm... when they removed all the code comments without storing them somewhere else to look them up, they will probably suffer from code amnesia in a few months (code amnesia = when you don't remember why something is done in a particular way in the source code). That could lead to slower Audacity development, giving the Tenacity project an advantage. Documentation is important, inside and outside of the source code. If Muse Group thinks otherwise, they may be up to an interesting development adventure.

I consider tenacity (https://github.com/tenacityteam/tenacity) the real fork. First of all, the issues of the sneedacity "project" make it look like a bad parody of a software project:

https://github.com/Sneeds-Feed-and-Seed … ity/issues

Most important, there is the report of attempted murder of the main tenacity maintainer. Read more here: https://github.com/tenacityteam/tenacity/issues/99

If the people from 4chan behind the sneedacity project (see its README.md) really see murder as a legitimate way to solve problems, then the sneedacity project should be removed.

This thread is not about browser extensions or the need to update them from time to time. It's more about Firefox (and Chromium) removing features like FTP that are still useful for some people while introducing other questionable "features" like Pocket, new UI and hiding useful configuration entries.

Too bad to see that the Palemoon developers apparently act in the same manner as Mozilla does in regard to unwanted changes (like the user agent setting). And the arrogant attitude of two main developers certainly doesn't contribute to a good working environment in the Palemoon project.

I took a quick look at the current Firefox source code (see https://hg.mozilla.org/mozilla-central/file ) to get a first-hand impression of the code quality. I looked at some files in the "browser" and "netwerk" folders. The files I looked at are written in C++98 (raw pointers, const char* instead of std::string), even the newer ones like those for the HTTP3 protocol. I haven't found a header file with documentation blocks, but instead multiple classes in one header file and cpp files with almost no comments. Furthermore, I found hints in the form of include directives that the Firefox developers implemented things like mutex and weak_ptr themselves instead of using the implementations in the standard library. If the rest of the code is the same way, then I can understand why Firefox forks are so rare... Then the only option indeed would be writing own browsers, made by the community, for the community.

Protocols and malware are similar to pizza boxes with tainted pizza: It doesn't matter how the box looks like if the pizza is tainted. If you deliver a malicious web page via HTTP(S), FTP(S), Gopher or another protocol, the protocol doesn't matter.

Doesn't the FTP protocol removal follow the same pattern that can be seen around SystemD? Before it descended upon the GNU/Linux world, we had multiple init systems that other software had to interface with. Now a unification process is on its way, either by using SystemD directly or by using compatible implementations. The web also had many protocols besides HTTP(S) that were more common in the past to retrieve data: Gopher, FTP, (that one newsgroup protocol), QOTD and many more. A huge part of the Web is HTTPS now, so that we see a similar unification process.

Indeed, Firefox is bloated and there seems to be no incentive for Firefox project leaders to make the browser better, since Firefox and Chromium have almost no concurrency at the moment (Microsoft Edge is also Chromium) and Firefox can be seen as the lesser of the two evils in regard to privacy issues since you still can disable some antifeatures via about:config, which you cannot do with Chromium.

In the long term, the free software community has to build new browsers. Either by starting to code from scratch (like I did) or by taking the Firefox or Chromium source code and creating something better out of it (Pale Moon looks promising).

EFI would allow building an installer image once without having to care about the specialities of the hardware. That's what an U-Boot that is already installed on the machine (in on-board flash for example) would do. The installer image would just contain Grub and a generic kernel which gets its hardware information (FDT) from U-Boot.

A few days ago, apt finally reported some packages that can be updated. In short: Everything is fine again. Thank you all for your help :-)

@florine:

I think, we have different approaches here. I try to build my Devuan system "from scratch" using multistrap. Regarding the boot process, I'm attempting to start my ARM boards just like x86 hardware by letting U-Boot load Grub first so that no extra steps after upgrading the kernel will be required.