So Neil Cherry posts about Smart Home [automation] stuff, which can be turned off from cloud access at any moment by the distributors. And Dave McGuire responds, "No offense intended, but I don't know how any intelligent person would ever buy that sort of stuff in the first place." Well, that's the discussion about operating vintage embedded computers, and decades of computing products, in a nutshell. Computing associated with control or operation of things or services, is either old and stable; or new and long-term unstable. And either one has some choices, and can take actions (repair, substitution, etc.) or one can't (and out the old stuff goes, no discussion). "Intelligence" is some other matter about skills and is just argumentative. But it references the more general point - as discussed in the article and the thread. There's some high-value services around some old computers (like operating expensive lab equipment) to compel both computer and equipment to stay in-service as original. Because, costs of replacing the lab-service or to re-engineer a new computer, are unacceptable. Neil made a point about railroad equipment, to say this is not a new thing. I'll work this a bit because the argument interests me. "Stable" means it will continue to run until hardware failure. If it was gonna fail, it already did. If replacing it made sense, it was already replaced. Software can fail but if left unchanged and un-attacked, may be "stable". So what's left as vintage and still in service defines itself. You can call that "survivor bias", and that's what is in the Nature article. "Unstable" in the modern world, means failure will occur when the providers turn off the connected services needed. Or failure of marginally designed hardware (like some modern commodity PC's). Or failure of software that accumulates issues which slow down operations (anything Windows). Or: one gets bored and moves on to the next thing. It's a race to see which occurs first. When the pace of change is faster than the breakdown of products and services, "unstable" as I described it doesn't matter. You've replaced it or moved on already. And if that's the norm, then it takes some explanation to describe an alternative world. That world, is the world of the past (in the United States). The modern bias comes from the computing world experience of the 1980's and 90's, when "smaller cheaper faster" computing due to basic physics, became some mantra. First it applied to any service that could be wrapped around information technology. Then it became about jobs and production and consumables, and other science and technology. Now we have a world where most everything seems to be about information services - with many costs at zero, the "race to the bottom". So how do you make a dollar? Charge for services; extract value from information; build cheap so items need replacement and services can be upgraded. A lot of modern problems are a consequence of that. Neil talks about "reliable" in the sense, of services and items built for durability and repair in use, which still provide a desired service, and a service of high even essential value. In the 21st century "reliable" has to be explained because of the assumption that "new" is better, and lack of (good) experience with repair. And, an effort to avoid repair or maintenance costs to encourage replacements (as a market opportunity for providers). And, a simple lack of (modern) individual experience and understanding, about the physical world and the operation of things in that world.
People at my LUG's hack night, where they try to help new people to install Linux on their systems, typically react in a bad way at my work with similar hardware. They just don't want to believe that stuff like that has been around for well over forty years and more. - Greg Levine
Some days ago, someone posted in VCFed or the IXR list, about the need to advocate "the right to repair". I said in so many words, to appreciate "repair" one has to learn about a time when repair was not only possible but considered desirable or even preferable. That's one reason I have a Web site, to preserve some of that history. The Nature article addresses historic artifacts in some of the same ways, not so much about repair as substitution. What it says and doesn't say, I find informative. Regards, Herb Johnson -- Herbert R. Johnson, New Jersey in the USA http://www.retrotechnology.com OR .net preserve, recover, restore 1970's computing email: hjohnson AT retrotechnology DOT com or try later herbjohnson AT comcast DOT net