2,200 forgotten vintage computers from a barn in Massachusetts
vice.com
vice.com
Proprietary no-source OS's were still common then, so binary patching the kernel to put in a different hard disk parameter table (to use a luxuriously large 20MB drive in place of the ST412 the machine came with - precious! Must not mess with the irreplaceable original OS image) was undaunting, especially with a .h file handy that gave the structure. Compiling "elvis" to get vi in the absolutely stripped down Minix mode, that used 63Kbytes of the maximum 64K of code space that executables could use... fun times. Of course back then you still had a hope of compiling current C with ancient pre-ANSI K&R C compiler. Most stuff that I ran on the machine didn't need much porting.
A crawler across everything tech which takes a comment such as yours, and then parses out all the systems/people/code/languages/companies/timeframe and builds a really good history of computing.
That would be absolutely beautiful.
AI Keeping its own evolutionary tree documented... and turned into a teaching platform.
The paragraph you've provided is rich with technology details, which offer a good opportunity to delve into the history of computing. Here's a summary of the key technologies, people, and companies:
1. *NABU 1200:* This is a somewhat obscure early PC built around the Intel 8086 processor. The NABU Network was a Canadian computer network that existed from 1983-1985. NABU PCs were designed for home use and offered unique connectivity options for the time.
2. *8086 Unix machine:* Intel's 8086 was a 16-bit microprocessor designed in the late 1970s, which was the basis for the x86 architecture. UNIX, a powerful multi-user, multi-tasking operating system developed in the late 1960s and early 1970s at Bell Labs, was often used in early computers with this processor.
3. *Microsoft Xenix 1.0:* This was Microsoft's version of the Unix operating system, licensed from AT&T in the late 1970s and released in 1980. Xenix was a direct port of V7 Unix from Bell Labs. The "1.0" suggests this is the earliest version of Xenix.
4. *Bell Labs:* The historic research and development company responsible for a host of important technologies and programming languages, including Unix, the C programming language, and others.
5. *ST412 hard drive:* This was an early model of hard disk drive produced by Seagate Technology in the early 1980s. It was one of the first 5.25-inch hard drives and had a storage capacity of 10 megabytes.
6. *K&R C Compiler:* This refers to the original version of the C programming language as defined by Brian Kernighan and Dennis Ritchie (hence "K&R") in their classic book "The C Programming Language" (1st edition published in 1978). It predates the ANSI C standard.
7. *Elvis and vi:* Elvis is a vi clone, vi itself being a visual text editor for Unix. It was written by Bill Joy in 1976 at Berkeley. The reference to "compiling 'elvis' to get vi in the absolutely stripped down Minix mode" refers to the process of building the Elvis software from its source code to run on the Minix operating system, a Unix-like system designed for teaching purposes.
8. *Minix:* A Unix-like operating system created by Andrew S. Tanenbaum for educational purposes. The first version of Minix was released in 1987.
Overall, this paragraph presents a snapshot of a transitional time in computing, when PCs were still relatively new, Unix was becoming more common on these machines, and software had to be patched and compiled from source code to run on specific systems. The timeframe seems to be early to mid-1990s, considering the user bought the NABU 1200 at a garage sale around 1993.
for me, i rather we go back to having time and energy to write decently about our history so others can learn in the future.
I wouldn't trust an AI to not hallucinate entire operating systems.
So get a listing of languages, companies, technologies, industries, resources, supply chains - that can be directly tied to the evolution of computing - and let it web them together?
Are there many of these still existing? If I'm not set on having something Canadian, which would be a good model to look into? Would it be a good way to introduce a kid to computers / Unix / programming?
http://www.nicolasbize.com/blog/30-years-later-qbasic-is-sti...
Hacker News discussion: https://news.ycombinator.com/item?id=11636383
https://www.computerhistory.org/revolution/supercomputers/10...
First comment: cool. I had experience with these.
Second comment: Where can I get one?
Third: I used this machine which has nothing to do with this conversation.
Fourth: Why would you use that machine when you can get a reproduction?
Fifth: You can't reproduce supercomputers.
I'm fairly certain at this point that the internet is just a bunch of people talking past each other.
If we were to gift this in the most optimum way for a person who would actually put this machine into service...who would that be? What criteria make this the correct solution?
I still miss weird stuff in sunnyvale, but this will have to do.
Reminds me a bit of the community tool libraries. And maybe could serve a similar purpose.
Like if you find some old media that you'd like to read. Or want to use an old device with a rare connector.
A few years back, I saw that a lot of my retro stuff was beginning to fail simply due to aging components. I donated what I could to those that could do something with it but it was kind of sad to see that non of this is going to be with us for a long time.
It is crazy in video games seeing stuff like PlayStation 1/2, Gamecube and Xbox DVD's are now starting to suffer from disc rot. It is looking like emulation or FPGA recreations by be the long path ahead if you want to experience as it was.
Like "gift it to Africa" where we send them a crate of useless computers with a note "you guys will love these! it's what I used as a kid (40 years ago)."
https://www.old-computers.com/corporate/donate.htm
> In our effort to preserve old-computers, and everything related to them, we are always looking for new systems ... Every minute an old computer is thrown away somewhere in the world ... Some are real museum pieces, and most deserve to at least be saved from destruction.
> Thus, we are looking for people willing to give us old computers, video game systems, books, or anything related to computing in general from the 70's to the mid 90's. Of course, we will pay the shipping costs !
They don't take everything, but it felt better to give this to them than to have folks who didn't know these systems' import taking them to the dump.
Also highly recommend attending Vintage Computer Festival events if you get the chance.
When you pressed the power button it was attached to a steel bar that went over to the power supply. Not a wire.
No clue what my dad finally did with it. This is a guy that kept my old 1996 Gaming PC as his mainstay until 2019.
Sadly, just because something is 40 years old doesn't mean it will fetch a high price. Especially microcomputers. There's a bunch of rare and also cheap ones.
Rare because almost nobody wants them and cheap for the same reason.
For an equivalent today, think about some low end random model android phone from the early 2010s. Both cheap and rare
> Plus, they were essentially new. “It’s new old stock, but it is tested,” he says at the beginning of the clip. “I think the seller actually peeled the original tape off, tested it, and then taped it back up again.”
IBM PC complete with manuals and keyboard? Yes, expensive. Z80? Not so much
> Rare because almost nobody wants them and cheap for the same reason.
> For an equivalent today, think about some low end random model android phone from the early 2010s. Both cheap and rare
There was an episode of the Dead Zone that had a twist like this. The main character and some others were out in the woods and came across some people who took them hostage while trying to look for a plane that had crashed nearby a decade or two ago and was rumored to have really expensive cargo. In the end, they plane turned out to be full of computer chips that were high end at the time of the crash but virtually worthless in the present.
It is one seller, he has been testing them all.
Which seems an awful lot like "we're just going to, uh, you know, hang on to the ~$28,000 you just made. For a couple of months."
Was he still responsible for shipping the orders while they held onto his money? Not everyone can afford to just shovel out the shipping and labor needed to do that.
Has anyone been through this who will talk about what it involved?
Magnanimously, they allowed me to withdraw $300 per month. After 6 months, PayPal let me take out 600 per month.
In the end it took 2 years to get the money out, and I shut down the account and never touched a PayPal merchant account again.
In the meantime, I kept my business going by selling through Stripe and front loading enough sales to ship the orders for the earlier customers. Was a fun time!
So while it wasn't through eBay, like this seller, I would say he's very fortunate to reach someone in support there with a brain and common sense.
He did sell the devices on his website in the meantime.
[1] https://www.computer.org/csdl/magazine/an/2016/01/man2016010...
so the primary roadblock for development at the time was cost, not capacity.
i believe today most development happens at max capacity, because cost is no longer much of an issue.
I'm not sure if this represented the potential to sell people on the Internet, or just on online services as a whole. As I understand it, the NABU platform was still (in its original form) a walled-garden, just like CompuServe, The Source, or GEnie in the same era, or even the Minitel-style platforms in Europe. The unique proposition of connecting to another system ran by an unrelated party was still not there.
If we roll it back to "this might have kickstarted adoption of online services", that's an interesting discussion.
From my perspective, I think people were averse to early online services because most dial-up online services were extremely expensive before the 1990s. A few dollars an hour adds up fast especially at low bandwidth. I can recall my family getting its first PC (~1992 or so) and being warned never to click the icons for AOL (the early GeoWorks based version) or the terminal emulator due to cost concerns (although it would have been moot because the PC didn't have a modem!)
Part of that was that a lot of the walled-garden services used existing commercial timesharing networks for access, and those were priced for people whose business was paying for it. In the NABU design, the infrastructure was owned by the cable firm in the first place, and there's no risk you're pushing off a better-paying customer. What did the price plans look like? If they had treated it as an unlimited, flat-rate package, then adoption would have been a lot higher.
I also understand at least some forms of the NABU system were aggressively unidirectional-- a lot of the software was on an 'endless loop' where you waited for what you wanted to cycle past again to start downloading it. That probably helped keep the infrastructure costs manageable, only the much smaller upstream link needed to be managed on a per-user basis.
Cable television deployments were completely unidirectional. A cable headend typically pulled in a satellite feed and injected that into their coax. There was no upstream connection at the client side and more importantly no upstream at the headend. Cable companies didn't need so didn't have upstream links so they weren't just lying around waiting for some service to use them.
So a counterfactual NABU online service would have needed cable headends upgraded with 1) bidirectional transceivers and 2) the headend offices would need long distance backhaul. Early online services were expensive in part because that backhaul was expensive. A counterfactual NABU service would have needed that same expensive backhaul.
Note that Wikipedia attempts to revise history by claiming "LapLink cable" is only a parallel null modem cable when it meant any null modem cable.
The hope among NABU developers was that the service would become popular enough that cable services would upgrade. That ultimately did not happen, because the service itself did not become popular enough for that.
I mean, this service had Tom Wheeler, coming right off of being the head of the cable industry’s trade group, running this in the U.S., and even he could not get the industry to move faster.
Are there any specific vintage computers one should definitely buy immediately if the opportunity comes up?
There is no inherently "great" retrocomputer: they're mostly all old crap. Sure, you can make money if you buy a SGI Octane low and resell it high, but it's pretty hard to get one for a low price... and I would get emotionally attached to it anyway.
Several times I felt the impulse to buy old hardware such as Commodore C64, Commodore Amiga, Atari ST, some Z80 ZX Spectrum clones, HP Jornada, Nokia Communicator, 386 Laptops, but I am glad I didn't. I would have played one or two days with each and I would have lost interest after, so that would mean loss of money and storage space.
I'm confused.
On the user pages this doesn't happen and it shows the original time.
Oh that's interesting, thanks for the explanation! (and a bit weird too but not in a bad way)
Looks like they've raised prices a few times since. Latest listing by them is at $180:
Quantum Link and Prodigy are likely the most common comparison points, though this used a proprietary platform for much of its life, leaving out people who already owned, say, an IBM PC or C64.