D_Air1

Yeah, I'm using qwen 31b a3b on an amd 9070xt requires a bit of cpu offloading, but still plenty fast. Using it wall llama.cpp. Combine that with some mcp's such as ddg-search to make it truly useful by actually being able to search online.

I mostly use it for small tedious tasks with well defined inputs and outputs. For example when hyprland recently changed from their own configuration language to lua. At first I started going line by line translating my config to the new lua language until I realized oh wait this is exactly the type of thing that ML is useful for. Going from the well defined hyprland configuration language to their also well defined lua syntax. It banged it out in less than a minute with only a single mistake which I easily fixed. The mistake it made was that it forgot to translate the comments to lua. It did it in less than a minute and worked first try. Where as I had made several typos and gotten a few lines wrong when I was doing it by hand.

Not to say that I couldn't do it. I would have gotten it done in about half an hour, but less than a minute is a lot faster.

I also used it to transform a bunch of unstructured data into json data, so that I could then use purpose built tools like jq to parse that. If I'm having trouble finding certain information. I'll ask it to find me some resources to look at.

Basically small well defined tasks and parsing data is what I use it for and it seems to be pretty good at that.

What I don't like is the way companies try to market it to people. I don't believe people should be trying to summarize emails or messages from loved ones, writing essays or any other creative tasks for the most part. Translating is okay. I don't expect a machine to be able to decide things for me or to be some filter between me and others.

Agreed, I feel like people are lacking a bit of self reflection in regards to this issue. The reason why people use the AUR is because it gives access to software outside of the official repos. No distro packages every piece of software out there. Therefore there is always a need for third party repos and that is why every distro has its own AUR equivalent. Thus leading to the same problem. Blindly installing software will never be a safe thing to do.

The command I used before changing the defaults was this sync && watch -d grep -e Dirty: -e Writeback: /proc/meminfo

That should show you the progress. It should be going down as it is being written to disk.

Less, I also have this but with different numbers. The simplified answer is that the transfer speed that you are seeing isn't the actual speed that it is transferring to the disk in the first place. Instead it is the speed of it being cached into ram, before tapering off to actual write speeds. The defaults work on most machine, but there isn't a one size fits all value. You may find that playing around with those numbers causes more stable write speeds because it will force the system to actually write the data to the usb stick instead of letting large amounts of data cache into ram before actually writing it out to disk. This is often the reason why when it says it is done writing it isn't actually done writing and you have to wait some minutes on some systems to be able to eject. In fact, a lot of times on machines that experience this problem. Hitting the safely eject button is what forces it to immediately write all data to disk.

I believe there was an article about it that I can't seem to recall the name of, but I believe it was something along the lines of "your operating system is lying to you" or something like that.

Any of the adblockers can do this. Pihole, adguard home, technitium. Ublock origin to. You can probably do it pretty painstakingly through the hosts file as well.

Why Linux ended up being the big thing is pretty well answered in the historical sense if you want to go looking for it. As for its low modern adoption. No one can really answer that for certain. I'll give you my two cents on the matter, but that's all anyone can do. All of this is based on research done on and off over the years in regards to this very topic as well as personal anecdote and hearsay. I will point out specific examples that I am familiar with, but don't fault me for missing anything.

Everyone is already on Linux. Both companies and individual people. While the BSD's work just fine for some people, it is largely hardware dependent. I have heard many people liken it to where Linux was 10 - 15 years ago in terms of hardware support. That alone means that most people can't use it. Less people = less developers making things better = less people trying it. We've all seen that song and dance before. Good ole chicken and egg problem.

Furthermore, while BSD certainly has its strengths. Being technically better has never been enough with anything. There are lots of equivalents to BSD features that are good enough eg: cgroups and others for jails. More importantly with a lot more big players using and contributing to Linux. Those things also see a faster rate of development and more quickly meet the needs of companies.

There is of course the license debate. While not as important now as it was before at least to a lot of individuals, I have personally been trying to answer this question for years doing my own research. The only reason I bring this up is that companies often upstream there work. Netflix famously chose freebsd over linux for their simpler and faster networking stack. They have contributed many improvements to that upstream and there are examples floating around as to how those improvements helped to improve freebsd networking for others. Although according to many Linux has largely caught up in that regard if not surpassed it. There are after all many tech giants that use linux and also need to serve similar amounts of traffic if not more than Netflix. However, regardless of if its is better or worse. The point is I feel like examples like this are far and few between. Because companies can simply take the bsd code and choose not to give anything back. It certainly feels like they do so more often than not. I based that on my ability to find useful examples in the first place. Which is of course admittedly flawed.

You will notice a lot of the use of the word "feels" in that last paragraph because I don't have any concrete proof. It is hard to measure how much a company has contributed to freebsd. It is less talked about and even combing through commits you would need to know who is behind those aliases. There are concrete examples of things that were contributed, but in my opinion a lot of the contributions are even more company specific than those on linux.

For example when it comes to changes that matter to a desktop user. Sony contributed drivers for their ps5 controller on linux. Here is a random article for that here: https://androidexperto.com/sony-releases-official-ps5-controller-driver-for-linux/ I found many articles of bsd people digging into linux code to get the ps5 controller working on bsd as recently as 2024. Here is just one of those https://forums.freebsd.org/threads/playstation-5-dualsense-controller-pairing.80786/. In my opinion however, it is kind of strange that they would have to do any work to get it working considering that the ps5 and ps4 if I remember correctly were based on freebsd in the first place. Why did Sony not contribute drivers upstream for bsd? They must have them because the console itself needs them. This harpens back to me saying that it feels like companies more often than not choose not to contribute back when they don't have to.

It has been hard for me to find equivalent examples on the bsd side. Little things like hardware or software support for user facing things that have been contributed to the bsd's by the big names, but not to Linux

Anyways, that's the short version of a random miss mash of things I could think of.

I've used rustdesk off and on. It is pretty good. I use it to help my aunt with "computer stuff".

Had to scroll down this far just to get to markdown files. Although I write with a bit of a delay. Once I get something working. Then I document what worked and what didn't. Alternative methods and issues I had with the alternatives.

So something disabled your monitors and you were able to get them back by quickly enabling them during the brief period of time that they were visible. The monitors should still be listed in the display configuration even if disabled. May be worth filing a bug with kde. Seems like a massive oversight if that isn't the intended behavior.

My brother used to use enterprise windows 10 until he started having that vary issue with trying to play COD.

It certainly isn't a permission issue like others are claiming at least for the window rules. It could be a bug, but as far as I know the steam deck does not use the latest version of kde, so it may already be fixed and isn't worth reporting.

As for onlyoffice, if I had to guess. It is forcing its own window decorations and therefore you don't get all of the same options as with windows that use the breeze decorations from kde. However you can still create window rules for these window. If I recall correctly. The default shortcut is alt + F3 which should work regardless of which decoration the window is using.

What does the kwin rule for evolution look like?

Lastly, the password thing is the only one that could be related to a permission issue. I however don't believe you should start mixing keyrings. kwallet is probably already installed and running. Like I replied to one other person already. Both gnome-keyring and kwallet use the secret service standard. The only reason it could not work is if evolution hardcoded gnome keyring instead of relying more generally on secret service. However, I doubt that. You could even use keepassxc as a secret service provider if you wanted to. Although it is a bit more annoying in practice.

We're basically already there. KDE has already moved to the secret service standard that is used by gnome keyring. The real issue in my experience has been applications hard coding a dependency on gnome-keyring and not secret service. Some applications such as mailpring have fixed that hard coded issue, but it is still an issue in other applications.