Engineered complexity in our systems is what keeps us shackled to the whims of big tech. The key to devolution is simplification. Simplify, simplify, simply. Remove, reduce, distill… and then maybe we stand a chance at owning and controlling the infrastructure as individuals.
@aral Efficient Computing is not about running your energy wasting thingies on solar panels. You first want your thingies to waste less energy.
Hi Performance Computing is not about buying bigger machines, but making optimal use of what you have and use *smaller* machines if they can do the task.
Less is more! Learn to think 21st century.
@alcinnz @gert @aral Yes, kind of. The bad side effect of that paradigm is that although machines become more efficient with each advancement, chronic upgraders use that as an excuse to buy new hardware every year. I think my 12+ y.o. hardware is good for the environment because I didn't dump 12 machines/phones in that span of time.
https://storage.ning.com/topology/rest/1.0/file/get/8657487494?profile=original Page 9, from a group I'm involved in.
There can be nuance here, but agreed: Embedded energy needs to be taken into account, there is truth to the "sunk cost fallacy". Upgrading software rather than hardware *can* help with that, though the strong expectation seems to be that it doesn't.
@aral Yes I agree. Some of the complexity is necessary, because computers and networks are complex and life is complex (requisite variety, and all that), but a lot of the complexity of modern internet systems is just bloat, mostly there to facilitate surveillance.
I was recently watching this video about a search engine for retrocomputers, and it is amusing that a lot of the complexity of modern web pages can be removed and that you can have a good web surfing experience on hardware from the 1990s.
@aral there is a beauty in the selection of primitives in a CPU, how it only needs to operate on data using a finite set of instructions to solve virtually any problem. The same beauty reappears in UNIX, C and shell scripts, where data is fed through pipes, combining programs together to solve higher-level problems. but somewhere at this high-level problem sphere, we failed to define (or maybe distribute) widely useful primitives, and thus big tech throws money at catch-all solutions and wins
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!