yeah, i got that. but didn’t want to make too long of a post.
this would work if the issue was only about sourcing raspberry pis and not sourcing nearly everything.
but you’d end up wanting to make that box a class compliant USB audio interface + GPIOs (to control the screen, read buttons & encoders). so you’d end up needing a microcontroller of sorts. but lots of those microcontrollers are nearly as unobtainium as raspis…
it’s a good idea, but i’m not sure if the amount of effort (r&d time & cost) would be worth it if securing said components is uncertain.
It sounds like essentially making a Norns flavored Linux distro.
From what little I know of such things, its an entirely possible thing that would also be a huge undertaking to do with any kind of quality. At that point its almost more about having the money to fund the work hours of development than any particularly tricky feat of technical wizardry.
Which leads me back to the broader topic of the thread. This scarcity and precarity is a huge problem for small creators, so I can see how software might be a bit of a hedge. Noise Enginnering comes to mind here, having recently put out some VST plugin versions of their hardware.
I’d love to throw Norns on an old laptop and have it “just work”. But how much would I be willing to pay? When I think about it, if there ever was an audience willing to pay for a bespoke audio linux distro, itd be lines. However… the pricetag would determine the success or failure. Would you pay $25? $50? $100?
Id like to imagine a software offering helping the hardware creators I appreciate the most. But the actual logistics of doing so might not be worth it (we havent even gone into tech support for the paid software). Probably makes more sense for tehn to make t-shirts, stickers and other branded merch.
hey! yupp 666MR (~noizhardware/666cpu: vm/666MR/README.md - sourcehut git) is basically a stripped-down version of uxn without video support.
The instruction set is focused on using two different memories: progmem and actual ram, that’s because most MCUs have a small amount of ram, but “plenty” of program memory.
It runs pretty smooth on Linux and Windows, it also runs on Arduino Micro, but here I’m hitting the limit pretty soon.
The current implementation pushes a single byte of audio on each sound vector call, the absence of a buffer might be the reason for the poor performance.
I wanted to try a different implementation as a summer project, but I haven’t had time for that :((
Another interesting approach would be to emulate some simple APU and see how performance is impacted.
Sorry, Tehn, for the stress and anxiety this many-year-long situation inflicts.
Thinking out loud, it seems like it would be a good move for the government to subsidize the formation of a domestic parts manufacturer that would produce a line of all the most necessary parts needed both by large and small industry. I think the current administration is advocating to do something with stimulating the domestic semiconductor industry, but it seems like it would be most valuable if they specifically funded a plant that would make a stable line of the most common and necessary parts. Sort of a co-op where the purchasers, including the small ones, are involved in the product line. A boardgame publisher, mostly of military strategy games, GMT, changed their sales model years ago to a preorder system where the post games, or reprints of titles, and customers place preorders knowing that printing will only move forward if they hit X number of orders. This only tangentially similar to what I’m thinking of, in the way that customers determine what is needed. It would be such a good investment in the economy. I’m probably drastically uninformed and naive bu it seems worth proposing to a senator.