Thu 17 September 2026

Since I spoke on Retrocomputing yesterday, I wanted to talk about its sister discipline/concept: Permacomputing.

Now, I can't talk about it at length as anything remotely approaching a subject matter expert, but it is something I've been fairly interested in for several years, ever since I learned about uxn and CollapseOS.

Whereas Retrocomputing looks at the past with rosy-colored glasses, Permacomputing looks to the future (or at least the present) with utterly sober, if not bitter/brutal realism.

I know more about uxn than CollapseOS, since I've enjoyed watching several of Devine's talks on youtube, and interacting with them over the years on the Fediverse. I've also compiled the uxn assembler and virtual machine, tried out several programs written for that vm, and even started learning the language (uh, several times. ADHD says hello).

To me, the central idea behind permacomputing is that our current technology standards and practices entirely untenable, to put it charitably. It's an absolute dumpster fire of stupidity, folly, short-sightedness, and rapine, to be blunt. I'm not going to beat up on young folks (because I actually regard them highly, for the most part), but the fact is that we have an entire generation of technology professionals that have grown up without a strong concept of what computing really is, what the actual, practical cost of any number of operations should be, and why things in technology were structured the way they were (both "Chesterton's Fence" and the opposite concept of "Cargo Cults" apply here).

So now, it is seen as normative to require gigabytes of RAM, gigabytes of storage, and gigahertz of compute to do the simplest tasks. The ever-lazy tech industry has brainwashed an entire generation into thinking their computers miraculously become "slow" every two to five years, when the fact of the matter is that every single computer made in the last two decades is an unthinkably capable supercomputer, if only our OSes and tools weren't absolute garbage. There are, of course, problem domains that really DO need the latest CPUs and GPUs, but they are very few. Extremely high-end gaming, of course, is one. The few cases where the use of actual, ethical neural-emulating algorithms are the actual best tools for the job is another.

But for 99% of use cases, there's simply no excuse for the fact that opening a social media website should swamp anyone's CPU and require the use of a computer no more than five or so years old. There's no excuse for the fact that a spreadsheet program can't run well in less than a gigabyte of RAM. The latest version of any spreadsheet program you could possibly name offers very few features that weren't already present in something like Excel 2003.

Ok, this is turning into a gripe session. 😅

So, to bring things back into gratitude, I'm incredibly thankful that there are a few people forward-thinking enough to create software tools and even entire ecosystems designed to run on any kind of remotely open computing device that was made in the last 35 years or so.

One of the first articles I read on the subject was cheekily entitled "These artists are making tiny ROMs that will probably outlive us all." I think that hits the ethos of permacomputing on the nose. It's technology with sustainability and survivability, not mindless short-sighted profit-seeking in mind.

Back to index

Category: Tech
Tags: ADHD   Computing   Ethics   Federated Services   FOSS (Free and Open Source Software)   Hobbies   Humor   Monthly themes   Non-religious post   Philosophy   Polemic   Productivity   Retrocomputing