[vcf-midatlantic] Old-school computing: when your lab PC is ancient

[Herb Johnson]

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