These printer companies could also open source their drivers after a number of years. The amount of landfill printers due to driver support is kinda sad. Just tossed my old cannon that had a perfectly good scanner due to driver support and the “need” for ink for the scanner to work. Booo!
100%. Require that software be released at the end of the support lifecycle for all licensed customers to continue their use-case, and watch support lifecycles magically telescope outwards.
It would require modification to copyright law in certain aspects (requiring “pass-through licensing” so to speak) and there’s utterly no will to tackle this in the US.
But those sorts of issues are already problematic for, eg, music licensing for games/shows/etc. Some of these types of hyper-limited, time-gated, non-product-ownership-following licensing agreements need to just be outlawed as unconscionable when they’re gating hardware that ends up in landfills or cutting off the public’s access to cultural touchstones. Or just shorten copyright significantly in general.
Canon doesn't develop MacOS. They advertise support for the version of MAcOS that existed when they launched the printer. If MacOS decides to break compatibility over time, it's not their fault. If they're not selling the printer today, they can't also provide up to date support for the OS of today. It's up to the OS developer to ensure a stable API for printers.
“Break compatibility” can cover a wide range of things, however, and that’s not always so clear cut. For example, if they were relying on lax permissions or a private interface then it’s a question of how much effort the OS vendor should spend on something they never promised would work.
“What can you possibly be complaining about!? It was secure when we released it! Maybe you should just buy a new one with less vulnerabilities!”
EU isn’t wrong that people have an intuitive sense that appliances like printers should have a worthwhile lifecycle and for some classes of devices this lifecycle should be quite long. 10 years really isn’t unreasonable.
Also, just like everyone has to support usb-c charging regardless of whatever other proprietary alternative they design… vendors should have to support a couple generic standards (postscript/Ghostscript and CUPS) that relieve a lot of the ongoing maintenance. There are very few/no valid reasons you couldn’t implement cups/postscript if you really want.