SoylentNews is people
SoylentNews
When Allan Lasser went looking for the oldest computer used by the U.S. government, he found a surprising candidate: the Voyager probes.
When I started this project, I hadn't even considered that the oldest active computer might not even be on Earth. But after my first post, I received a few tips encouraging me to look at the computers onboard Voyager.
Benjamin Levy pointed out how, "the actual computers on board are probably older than [1977] because it takes time to design and build space probes and to certify their computers for their mission," and another tipster sent me a link to a story about the Voyager team needing to hire a new programmer with experience in FORTRAN.
I'll admit I was reluctant to pursue these computers at first, but I soon realized that it was silly to disqualify a government computer from this hunt simply because it's billions of miles away. While the hardware hasn't been upgraded since it left Earth, the software has been upgraded and maintained to meet new mission requirements. We're still in touch with these probes and they're still performing science at the edge of our solar system. Most important, these are government computers and they are both old and active.
How much computer infrastructure of today will be operable, let alone reliable in 40 years?
(Score: 3, Informative) by AthanasiusKircher on Sunday January 08 2017, @06:30PM
But for everyday computers, they would cost so much more that hardly anyone would buy them, especially given the fact that they likely would be tossed out before that time anyway because they no longer fulfil the requirements of modern software, lack newer hardware interfaces, etc.
While I agree with you that 40 years is an awfully long time in electronics, I think there is also an argument to be made from the opposing side, i.e., that most consumer computers today are built on the principle that they are EXPECTED to fail or become unusable in a few years, necessitating another purchase. Also, the continuous war in increasing specs isn't exactly necessary to handle 97% of what most people do with their computers. Gamers want faster computers, sure -- along with people who do a lot of video or intense graphics work and editing, etc.
But look at what average folks do -- they don't play processor or memory intensive games: they get addicted to Candy Crush or Words With Friends or whatever the most recent BS is. Most don't do video editing on any scale (except if what you mean by "video editing" is creating a video file of their family slideshow animated with Ken Burns Effect). They check Facebook and email. They use word processors and spreadsheets at work. About the only hardware-intensive task they do on a regular basis is probably watching some videos, but a much weaker "computer" in the form of a phone or tablet is even up to those tasks today.
And keep in mind that a lot of actual computers (not "mobile devices") are sold today for work, primarily. Does office software REALLY need that much power? We've just accepted the bloat and the eye candy and whatever, but really -- has office software in terms of core functionality actually improved that much in the past 20 (maybe even 25) years? If not, why exactly does MS Office require something like 100-1000 times the system requirements of that era?
Of course the greatest change since 25 years ago in computing is probably the use of the web. But aside from when people are looking at DEDICATED video or image sites, they're often reading text primarily. Text that probably occupies less than 1% of the resources necessary to download and render the website. (A huge problem here, of course, is the deployment of ads.)
Anyhow, my point here is that the typical computer "lifespan" today isn't just driven by concerns about hardware expense or requirements of "modern software." It's also driven by the fact that computer manufacturers are happy to sell you another computer in 2-3 years, software manufacturers are happy to force you to buy an upgrade (which may not run so well on an older system, even if core functionality hasn't changed much), etc. If we didn't have those driving forces, the desirable computer lifespan for consumers might not be 40 years, but it might be well over 10 years. Unfortunately, there is little reason to force the kind of optimization that designers and programmers used to deal with decades ago. If you know that most people have computers 1000 times more powerful than what they were in 1995, why try to optimize common software like Office so it could still run on such a system?