I am curious who buys generative AI services? The consumers seem to be people making memes or questionable porn with free services. It can’t prepare food, unblock drains or tile a bathroom. You can’t use it for anything like medicine, law or engineering where you would be professionally liable if it fucks up. How is it sustainable?

Evolution currently. Previously Thunderbird. I wouldn’t mind a newer client but I am only interested in native apps talking to my email server over open standards.

Adding rust to a massive mature C project that targets lots of architectures and has many contributors is a difficult process. If it succeeds it is going to take a lot more time and patience.

True that copyright always existed to protect publishers and not creators. But in pre-digital times there were considerable barriers to publishing and distributing creative works at scale so while publishers in all media have often abused creators they were a necessary evil if you wanted to make a living.

The worst trick greedy capitalists have pulled recently it to bypass copyright and steal the entire digital record of human creative labor to incorporate into proprietary models and services for their own enrichment. I have no idea how society and our political representation has slept through that. The second worse is insanely destroying their own industry by fucking over both consumers and creatives with increasingly unsustainable greedy and dumb bullshit.

Access to education and other equitable causes really should be fair use. If everyone pirated, and the way things are going it will be the only sane way to get content, then new content is going to dry up unless people are happy with AI slop. We will still see indie self-published works but necessarily the creators won’t have access to the same resources we saw when they were part of an exploitative but productive industry. That sucks. A lot of people are happy to pay for convenient and affordable access to content under reasonable conditions and piracy is something they only resort to when that is denied.

You want the most common things available in a Settings app(s) as they generally are on Gnome, KDE, Windows and Mac. If we cram too much stuff in there regular people struggle. Finding a good balance is a dilemma for most platforms. You want the less obvious stuff to be available in additional specialist “tweak” apps for more experienced users as they often are on all these platforms but sometimes less so on Linux. Then the really esoteric stuff you have to edit registry settings, conf files and plists as you do on all of them. Linux tends to provide more power and flexibility but requires reading documentation due to the diversity of config methods and locations.

A Mac user very sensibly contacted me worried about pasting a command to edit a plist into the terminal from a website they found trying to fix an issue. Nobody should be pasting commands they don’t understand into terminals. A quick search and I found the GUI toggle to do the same thing. It isn’t exclusively a Linux issue. Windows and Mac have complex operating systems underneath and equivalently powerful command line tools.

GUI config isn’t practical for hardcore linux users. It isn’t scriptable, we can’t store it in version control, it is harder to document, it is harder to use remotely. We have to appreciate that we have a growing number of users where it is worth taking a bit more time and sharing an alternative if one exists. However nobody wants to configure services in a GUI as we want to version, document and distribute this stuff and managing services in a GUI is unprofessional because you lose these things.