@scallisto Thanks for feedback. :) That convinces me doing it myself, but have to compile it for ARM64 myself (no build for it as far as I see). Otherwise I cannot replace XC on my phone.

@scallisto They mentioned it in a pinned issue, haven't tried, yet. ChiPass supports everything that KeepassXC could do at least one year ago. Newer features are likely not included, unless they already developed those in the last 6 months the project exists.

About Linux kernel, not all of them share the opinion that LLMs are really helpful in every way. There is a talk about how they affect the kernel work on a daily basis, the first 35 minutes is the talk, rest questions and answers: https://inv.nadeko.net/watch?v=NnV_cWeoo5Q -> I don't want to know how much energy is only wasted for all these thousands of bug reports that are duplicated, already fixed, stolen from researchers that were using LLM chats and so on. How can Linus still make such a strong claims when it is clearly not that helpful? When patches even include human fails from 20 years ago, because the data is trained by the whole history of human mistakes, too? Sad to hear about Bitwarden, but never used it. I always had the feeling this will happen. Even more sad about KeepassXC, because I used it and was happy, but never realized the vibecoding policy and I don't trust those who are not only using them on such important security applications, but especially that early. Was looking deeper into it and it seems there is an anti-AI fork called Chipass that forked the version before KeepassXC was introducing the new policy. I will look a bit deeper next days, but currently I am 80% sure to switch over to it. The browser integration for KeepassXC is not allowing LLMs and support Chi, so no issues at this front. Chi is hosted on Codeberg for those also want to give a look.

@danny_laursen The pump from $200 to $850 means you can pay a developer full time for 1,5 months each year. The funding of a single source. Most free software projects have not enough money to pay even a single developer full time, so this amount would be amazing to many projects, also many other distros.

Just saying, it is a difference if an actual user complains about UX in specific vs a vibe-coder just bashes the entire project to promote his own two prompts.

(1) No, more KLAC smashed onto Linux is not good at all. As told earlier, it will escalate and now we reached level 2 of the escalation. And since there are these huge holes which cheaters definitely will use, the escalation level 3, 4 and 5 will come for sure, probably an endless fight where all honest gamers will lose. Epic also could just check out kill/dead-ratio at server level and would already ban 95% of all cheaters they ban with KLAC malware.(2) About LLM disclosure: I would suggest a middle-way, don't call the exact model, just "written/co-authored/assisted by LLM" as long as it is not a fully FOSS LLM (from future), but naming the exact model when using FOSS LLMs. That should address the issues on both sides (no advertisement, but still being transparent enough). At this point a huge thanks for how you speak about these LLM topics in general.(3) GIMP sucks indeed ... using it for 20 years now (so I am experienced enough to make that claim), still makes a lot of things unnecessary complicated. Just try to package textures in the different RGBA channels and compare it with other programs that can do this waaay easier or try to format text and compare. It is like GIMP has never reached the UX level of software in 2010+. On the other hand it is also a great tool and can do almost everything people need, just not always very accessible. It still sucks less than proprietary software as Adobe Suite with abo, cloud and also quirks. LibreOffice on the other hand is just great. I love that I still can use the non-ribbon GUI-standards and personally don't miss any feature. With Germany and others on board, missing features will be added at some point, so I don't worry about its future.

While I am not against the LLM driven recreation of Adobe Suite (in fact, one of the best use-cases), calling something LLM created "Open Source" is at best a strong claim. It reads code from the model (in form of weights) that contains all forms of licensed code, even like GPL2 (which you can't simply put into other free software licenses) and on top you create tons of lines of code nobody actually did read. A major point of "being open source" is that at least some people read the code to verify the functionality. Let's start to call LLM generated stuff "source available" unless they used a fully free software LLM (trained on specific licenses and made sources public and how they trained it etc) and the generated code is well understood. Otherwise it is openwashing.

LibreWolf as best privacy respecting browser for daily usage. Not bloated, not supporting Google (due Chromium usage), not proprietary as Vivaldi (absolutely NoGo). TOR as second browser for more sensible stuff. Servo from time to time to track progress to hopefully replacing Firefox-based browsers one day.

@alex@eloquer.de Yes it matters for many reasons. First, neuronal networks degenerate without new input. And once they are trained on 99% LLM output, it just takes 10 generations to get garbage out on the level of a child's first words. That is especially true if future models are bound to one or two companies, because nobody else can pay the costs of development or the hardware to run on. That also causes heavily democratic issues. Furthermore the models will be manipulated more and more over time -> enshittification. How can you be sure that it creates the program you want without build-in tracking? And do we have 100% green energy in 50 years (which we require much earlier) or is LLM one of the reasons we could not reach it? In best case it is a bet that "everything will just go right", but the experience in tech with super scaled companies is more that everything will go wrong that could go wrong. What if one of the companies becomes so big, that one or two produce 95% of all software in the world via LLM? It is a highly political topic, not just for "is LLM output still FOSS compatible or not?" (which is not as far as the LLM itself is not 100% FOSS, which is far more than "open weight").

Kernel Level Anti-Cheat (KLAC) should just die with Windows. Don't welcome the abusive technology on Linux.

I checked statcounter and was surprised how useless it became. The data for the almost last two years was not reliable any longer (probably due LLM crawler and agents) and now they decided to just remove the "unknown" graph. In India over 50% of all collected data was "unknown", now everything is gone. Linux dropped from 17% to 5% in that time where unknown went up from nearly none existent to those 50+%. Now it shows 12% while I am pretty sure it is around 20% in reality. Same with the rest of the world, without knowing the unknown, data is not useful at all. And even before they never showed the total number of collected data points, which makes analyzing data very hard. Just a service I will not use any longer after removing "unknown".

Looks like Ubuntu Touch is pretty limiting. I am daily driving PureOS and use Mobian for testing, both with Phosh and it works pretty well. Not as well as a real desktop, but most apps from desktop just work. Sudo apt install, Flatpaks etc or using native Firefox/LibreWolf or the desktop file manager ... it just works. The best of all? They are no smartphones, but pocketcomputers (computers that can also do phone stuff). I also saw people using PostmarketOS, Fedora and KDE Plasma Mobile, sxmo and alike. Pocketcomputers are so cool, all the freedom of desktop Linux on your phone device.