gemini://inconsistentuniverse.space/gemlog/2021-11-21-sustainableimplement.gmi

combined with a lot of my own thoughts about salvage computing/permacomputing/heirloom computing just that general vibe a lot of us seem to be thinking about right now as we all scramble towards something different, something new, something long-lasting.

I'm in a position of being able to work on a grant proposal for sustainable computing, applying to a grant opportunity related to Clean Energy/Clean Tech. This round I'm applying for is going to involve getting funding for a year of research, community outreach, and such that will---if we get it---give us the right to apply for a larger multi-year implementation grant that we can use to do more on-the-ground work.

In my city (Portland, OR---I'm pretty open about where I live roughly) there are already some efforts that deal with refurbishing desktops and getting them into the hands of people who need computers. This is, despite some of the ways my home gets talked about in media, a historically very poor city with pockets of tech sector excess. I have rants about that that I'll just elide for now.

Despite these efforts, much of Portland has no computational materials at home beyond smartphones and even those smartphones are often wi-fi only, without reliable data plans. During the pandemic there were programs to try and put wifi hotspots in low income homes so youth could still attend school virtually. This had mixed success. I also have rants about that that I'll elide. Instead, lets talk about schools and computers.

Our city is run by Google. Chromebooks are the only computers youth touch and after watching the consequences of that continue to unfold I think my previous rant about Chromebooks wasn't nearly angry enough.

Not only do youth get locked into a platform that hides how computers work from them and makes them dependent on Google for basic tasks, but it also limits the ability to grab hold of autonomy. While, yes, you can use Linux as a subsystem on a Chromebook and that gives access to a commandline that certainly isn't how schools setup their chromebooks.

Instead, you are rendered solely as a consumer of technology via the browser: the web browser having become its own separate operating system run on top of your own, one built around invasiveness, inefficiency, and being permanently connected to the internet to accomplish the simplest thing.

This is, of course, antithetical to how permacomputing needs to work. With all our coding living in the browser, there's massive inefficiency introduced into everything we do: layers of indirection, data lag. Just try using chrome on a little rpi and you'll see what I mean: a machine that can run multiple servers at once while also playing 1080p video and sending emacs over ssh -X will grind to a halt simply trying to play a video on youtube in the browser. Yes, I know, that's anecdata but it's not really hard to see what I mean when you compare native versus web applications.

In fact, the underlying point for sustainable computing is that even old computers, "underpowered" computers, can do far more than they seem once you strip away the inefficiency of a web-first/always-online paradigm of computation.

We're walking around with all these powerful devices that can be used for art and music and all sorts of utilities to make our lives better and yet the constant bloat of parasitic software makes them feel like toys after just a year or two.

And it's not just the web with all the surveillance tech. I have a smartphone and I've been noticing that even with keeping the Android operating system but moving towards foss apps from the fDroid store that my phone is running faster and using less power than ever.

But let's go even further: I think media, especially the games industry, are pushing this rapid obsolescence. I love gaming, don't get me wrong, but I fall into the camp of "I want shorter games with worse graphics..." like the viral tweet

I've grown very *uncomfortable* to say the least with games having grown to be these massive projects that provoke an arms race in storage and processing. Everything is being made in excess as AAA games have swollen to be > 100GB in storage and are always pushing the limits of current graphics tech. For what, though?

There's something horrifying to me about the way we're using entertainment to drive further endless growth and obsolescence of our tools but both viznut's thoughts on demoscene versus maximalism as well as "shorter games with worse graphics" point towards an effective way to fight this trend.

We can simultaneously, complementarily, pursue both making art with fewer resources and promoting greater autonomy with computers. By focusing on the limitations incurred by relying on older, salvaged, hardware then we can encourage using tools that have fewer layers of indirection and dependency, that are less virtualized in viznut's parlance. This means learning more about how computers and underlying concepts of computation and hardware work rather than a tool-of-the-day like the corporate tech sector wants us to learn.

This means also means rejecting big centralized projects like the gpt-series with massive datasets and instead working on doing more with less, researching machine learning techniques that lead to something much more "compressible" inline with Greg Chaitin's sense of the power of a theory.

Now all of this so far is very abstract but, in practice, if you had a year long project of r&d and then a multi-year project of implementation what might you do to promote permacomputing?

First, I think solderpunk is right: old phones and tablets with postmarketOS is probably going to be a good start to getting general computational materials in peoples hands. What we can spend that first year on is helping to port pmOS to more phones while also testing out putting together phone conversion kits, with the peripherals to connect the phone to an external screen and or keyboard so that it's usable in many different contexts such as public computer labs. We can supplement by creating booklets that are an introduction to all the things you can do to use your pmOS phone as a computer, including introductions to the linux command line and basic programming. We can test out these instructional booklets in free workshops over the course of that year and interview participants in follow-ups in order to see how what kinds of needs this actually was able to meet.

Second, I think we need to support getting more refurbished/salvaged laptops and desktops into peoples hands. There are already efforts in this direction as mentioned earlier, but I think these can be stepped up in terms of teaching repair & maintenance and helping people become linux powerusers. The materials mentioned earlier on the linux command line will be useful but to really promote the demoscene style of computing we also need to produce guides for the creation of art/music/&c. with limited computation. Again, we could run free workshops that also highlight and display works created by participants. From the experience of these prelimary workshops we should be able to learn what the real educational needs there are and what actually makes for proper support.

In other words, something that teaches low-level programming and about how computers work in ways that allow people to have real high-level skills and autonomy.

This is an effort that will, of course, involve starting small but imagine the end goal: more people learning how to work with the devices they have, fix them, and still do far more than they can do with the locked down/intentionally inefficient systems people---especially youth---are being handed every day. This means less purchasing, less waste, and eventually less demand. The best way to recycle is to never throw away, and the best way to produce sustainably is to not produce at all.

What we'll be doing is a small first step towards heirloom computing, the use of devices for not just a few years but decades, treating computational devices not as disposable trash but as previous artifacts that can have history and meaning over the course of years. What we're proposing is essentially the bootstrapping phase of building the knowledge, interest, and culture needed to tackle heirloom computing on the community level.

Implementing Salvage Computing: Rough Thoughts