Yes because they have no choice: the technology has plateaued and is beyond the point of diminishing returns.

This isn’t about making things safe; it’s about getting “ahead” of the stories of them failing to meet the crazy expectations from the market bubble. They are managing expectations as the bubble is bursting and this is a stratagem to try and talk them out of this mess. But they can’t change reality by just saying stuff, as much as them seem to believe. Reality is catching up with them.

No it’s simple and not confusing at all: The models are well beyond the point of diminishing returns for investment and the AI bosses know they cannot possibly meet the crazy expectations of the stock markets. So instead of them failing to deliver, they are using this excuse to seem like they’re choosing not to deliver for everyone’s good.

It won’t work - the bubble is going to burst regardless.

You download language dictionaries, so it depends on the dictionary quality for the language

This is also a thing in the UK. It’s very irritating as I have an account that goes back to the early days of Steam, a huge library and purchase via paypal. But I had to add a credit card to my account just to be acknowledged as an adult because of the UKs “Online Safety Act”.

These laws are bullshit, infringing on the rights of and inconveniencing millions of adults, to supposedly “protect children” but there remains no meaningful regulation of children’s access to the internet, parental responsibility or proper regulation of the social medial giants that are the real problem.

I extensively use a VPN now to bypass so many restrictions, including multiple sites that just choose to block UK IP addresses rather than comply. Like many bad ideas, they start in one country and spread. Australia brought us a social media ban for children which is also likely coming to the UK - again, bad legislation that doesn’t fix the problem.

For me the solution is quite simple: ban children from owning smart phones (or even develop a “children’s phone” that can only access approved services and sites), and create a regulated standardised allow-list of internet addresses and services for internet service providers to provide access to homes with children to use. Families can then opt in or out of that, and the government can focus its regulation on those services that want to provide services to children. It would also allow separation of adults from children in common spaces as services would essentially have a service for children and a service for adults (e.g. even a game like Fortnite could have separate servers for inside the childrens firewall, and others for adults outside).

Don’t force this bullshit on the entire population; restrict the actual access of children to a curated version of the internet. If you restrict access at the entrance, you don’t have to worry where the children go inside. It’s perfectly easy to restrict children’s access - it’s just shitty parenting to not even bother. The lazy excuse that “my children know better than me how this all works” is bullshit, but can be got around by making it simpler - no physical access to the open internet for children at all.

There probably was money involved - between Meta and X/Elon Musk.

Yeah, although the big benefit of a mouse is it’s super portable, and it needs one output to the screen and that’s it. I suppose the current limitation is that the screen needs USB-C output for power?

But it’s a great idea tbh! I’d prefer a linux device rather than android but other than that, great.

Yeah, it’s an official release from rasberry pi: Pi400, Pi500 and Pi500+ are all keyboards with pi inside as an all-in-one. Bit like an Amiga 500 or Amiga 1200. Comes with a nice manual too for to help people learn a little programming.

Hey, the official documentation is here: https://github.com/keepassxreboot/keepassxc/blob/develop/docs/topics/AutoType.adoc

Essentially, KeepassXC Autotype is fully supported on KDE Plasma 6, and all configuration is done within the app itself. The KDE settings you may be thinking of was in 5, but that is no longer the case when running on 6.0+.

It uses XDG Desktop Portals when run under Wayland; it should all be set up already if Auto-Type is working but essentially you need XDG Desktop Portal and the xdg-desktop-portal-kde package to allow it to integrate with the desktop.

The limitations on wayland are that it can’t match to a specific window title and can’t track when the keyboard focus changes between windows (due to how wayland works). The overview does note that certain other input methods can conflict with Auto-Type. But other than that it should be working fluidly without issue.

The only other consideration I can think of is are you running KeyPassXC natively or as a flatpak? Flatpak can sometimes be more locked down than needed for some apps. You could install Flatseal and ensure KeePassXC has the right permissions to access services outside of it’s sandbox.

The concept of a driverless car is good, the reality is not. The technology isn’t there yet, despite all the money and resources being pumped in to it.

Driverless cars were THE big thing a few years ago, and all the big companies were putting money into it and hyping it. Tesla’s value was predicated on it being big in driverless cars. Now a few years down the line, everyone has gone quiet on driverless cars and AI is THE big thing. It all just seems like examples of tech stock market hype. We don’t have driverless cars, and it remains to be seen when they will be fit for purpose. We don’t have AI, and it remains to be seen when that will be fir for purpose.

I don’t doubt it will come, but it feels it’s still “a few years off”. The industry likes to talk about “regulatory hurdles” but the reality is it’s too unsafe and unreliable to be deployed. So instead the tech industry is deploying something else unsafe and unreliable, but also unregulated: AI.

So the global total GDP this year is estimated to be $100trillion. Anthropic expects it’s revenue to eventually be equivalent to 30% of global GDP? US GDP is expected to be $32trillion this year. Antrhopic expects it’s revenue to be eventually be equivalent to 94% of US GDP?

And we’re supposed to believe we’re not in a bubble?

I’d say the data is pretty meaningless - at least the stuff presented in the picture and article - as it doesn’t compare X11 and Wayland running the same desktop environment. It’s X11 Xfce or Wayland KDE or Wayland Gnome etc. So who is to say whether it’s actually X11 or Wayland contributing to the differences? Comparing Xfce - a low power desktop by design - with Gnome or KDE is not a fair comparison; Gnome and KDE have way more bells and whistles. It’s interesting that X11 xfce4 under stress uses more power than Wayland KDE under stress, but it may be meaningless.

I’d rather see the figures of X11 xfce4 and Wayland xfce4 but the Wayland version is experimental at present so unlikely to be fair. X11 KDE and Wayland KDE would be a better comparison, but even then it reflects how KDE has developed their Wayland version, not necessarily just how Wayland compares to X11.

In terms of power usage, if that’s the primary concern, then go for the lightest weight desktop and run the least amount of software possible (including services). Switching from X11 to Wayland is not about power consumption, and a 1 Watt difference on a desktop PC that could run anywhere from 100 to 700 Watts isn’t meaningful. Even on a 100W low powered device, I’d question how important a 1W difference is.

If Gnome ran a restaurant, the waiter chooses the meal for you, and tut & shake their head if you try to deviate from it.
If KDE ran a restaurant, if you shake the menu it’ll grow exponentially in size until it fills the restaurant.

I found this article very interesting; the author is shorting on Oracle and makes that clear. But they give a convincing and cogent argument about the severity of the AI bubble and where it’s going, with links to interesting news stories and opinion pieces throughout.

The circular flow of cash between AI companies, and the Enron style revenue inflation is laid out in a clear way. It’s scary that such crazy behaviour is going on in the markets, and it seems the markets are catching on.

I personally prefer OpenSuSE, but Fedora is reasonably close second.

One reason OpenSuSE wins out for me is admittedly just because it’s not a US based project or related to a US based company like RedHat. Not that SuSE is “good” as a sponsor company; it’s just based in Germany and owned by a Swedish investment company, which edges out based in the US and owned by a mega corp like IBM. I know both OpenSuSE and Fedora are mostly independent, but both are under some control of their main corporate sponsors. So I prefer European.

Maybe that’s silly, but I have lost a lot of trust in the US legal system, project governance and their ability to exist without interference by the US state.

Separating them out can be a false dichotomy - some passwords are more important than others, but all of them need to be kept secure. Keeping 4 separate databases with their own passwords secure is more work than keeping one database secure. It’s generally better to be good at keeping your one database secure than be slightly less good at keeping 4 secure.

Having said I do have 2 password databases. My personal one, and a separate one I use for work. At work we have loads of different tools to use many with their own passwords, so I find it easier just to have that in it’s own password manager all in one place. I use an extension with my work web browser to access those and I keep recovery information in my personal password manager. One reason I do is it I don’t want my personal passwords exposed or available within my work computer or browser - it allows separation of my personal and work life, and means I am in no way personally reliant on my work’s IT security to keep my personal passwords safe.

Other than that, classify the types of passwords within the database; more databases increases the risk of something going wrong in my opinion.

Ok, the best way to do this is to install VM software on linux (I use KVM/QEMU) then make a new Windows guest and install from scratch using a Windows install ISO file. It’s the better method as the new Windows install will have no bloat from your original install, the boot partition will be set up properly and there won’t be redundant old drivers and configuration issues that could be very messy in the guest.

I have a KVM/QEMU windows VM with a license that runs fine; I use it rarely mainly to use work’s version of MS Office. You can migrate the license from your original install over to the VM so you don’t have to pay for a new Windows license. If you also have data to transfer into the VM from the original windows partition, you can either directly mount that partition inside the VM or mount the partition in your Linux host as a read only partition, and then share that as a folder into the VM. Then you can transfer any files you want into within your guest machine (e.g. within the windows guest copy files from the shared folder, drive D: onto the guests machines drive C:). It’s much much easier than the complexity and issues of cloning an existing partition and trouble shooting it to work in a VM.

If you really want to go down the route of taking the current installed OS and move that into a VM image, then the best way is making a disk image of the whole windows drive (assuming it’s a boot partition + C: drive) or clone both the boot partition + windows partitions into their own files if it’s a shared drive with Linux. There are lots of tools that can do this; “dd” and “partclone” are 2 particularly common tools you can use from a linux terminal to create a .img file of the partitions, and then you mount those into a VM. But messing with Windows boot partitions is always messy, and you’ll likely also need a Windows Recover disc to also mount in the partition to “repair” the boot partition in the new machine. You will also have to troubleshoot hardware issues if Windows can’t cope with the sudden complete hardware change from your host PC to the new virtual hardware.

Honestly, make a fresh Windows machine in a VM and transfer your files in. It’s far easier.

EDIT:
Here is a good guide to setting up a Windows guest in KVM/QEMU including TPM 2.0 for linux.

https://computingforgeeks.com/windows-11-installation-on-kvm-with-virt-manager/

Then with once you have a working Windows VM, this is a good guide for setting up a shared folder (I’d mount your windows drive within your Linux host as RO, and then share the mount point as a folder into your guest; that reduces the risk of data loss):

https://www.debugpoint.com/kvm-share-folder-windows-guest/

I didn’t write these guides but have read them through and they’re what I’d do/did for my Windows VM. If you choose to migrate your drive instead, you’ll still need to set up the hardware of the virtual machine as in these guides, but instead of creating a new virtual drive file, use the disk image you create from your windows partitions.

Before you do anything, backup /home.

You can reinstall your system while preserving and expanding your /home partition; this is probably the cleanest and safest way to sort your disk. It allows you to move the boot and filesystem partitions to the start of the disk while keeping /home untouched, and then separately resize /home to fill the rest of the drive. You could not reinstall and manually move the partition but it’s slow and riskier when messing with a boot and main filesystem partition; much easier to start again tbh.

I’d get a USB and install Ventoy on it. Ventoy is a great bootable USB tool that lets you drop multiple different bootable ISOs on it (instead of reflashing the drive every time) & pick one at boot; great for installs and also to keep around as a recovery drive. I’d then put on it an ISO file of your preferred linux distro, and also a separate ISO file of a good live distro for recovery. GParted Live is particularly good USB live distro for this because resizing the partition is the aim, but almost any good USB Live Distro will do

I’d then boot up the USB drive and select the ISO for your Linux distro’s installer. During install, in the partition section, I’d then use the partition tool in your installer. Dlete all the windows partitions (sda1, sda2, sda3 and sda4), and then delete the exisiting boot (/boot/efi) and root file system (/) and create new ones at the beginning of the disk: 1gb /boot/efi and 85gb / system partiton as you have now, and ensure the existing /home partition is kept and mounted as /home in the new install. You’ll have loads of free unpartitioned space; leave that for now.

After the system is reinstalled, I’d boot in, check everything is ok, and then restart and boot the USB again, this time selecting GParted Live. Then with GParted Live, I’d resize the /home partition to fill all the empty space.

But as I said, before you do anything, backup /home. Also before you do anything you can use a partition tool now (like KDE Partition or Gparted) to add a label/name to your /home drive so there is no confusion when you use use the Linux installer or Gparted later. But it should be clear from the size alone.

Switchdesk is a GUI for changing the default desktop once they’re installed, not for installing desktops.

The top of the page you link explains how to install alternative desktops which is using DNF. The GUI route to do this would be to use a DNF frontend like DNFDragora and search for the relevant desktop metapackage, and then install that.

NixOS config just sets Fstab. The issue was “nofail” was not set. This config calls that out which is good, but the issue isn’t fstab, it’s wherever OP got guidance on automounting drives in fstab as that guidance clearly failed to highlight the importance of the nofail mount option.

The issue isn’t fstab. The nixos thing posted there would have the exact same issue: if nofail wasn’t included in the configuration, the system would go to emergency mode when it can’t find a drive.

Fstab is the “File System Table”, and NixOS still uses it. In your case, a non-essential auto-mounting drive was not correctly flagged as “nofail” in your Fstab; when the system couldn’t find that drive it treated it as essential and locked the system down.