I would be more than happy to have modern versions that were so great to type on. While I really enjoy my iMac I cannot stand Apple keyboards or mice. The mush boards at work provided by Dell aren't much better.
Compaq keyboards weren't all that great, when compared to the IBM boards of the same era. The Compaq keyboards were quite mushy.
Where Compaq had it all over IBM was its video boards. IBM gave you a choice of two video boards: CGA and MDA. CGA gave you color and graphics, but at the expense of high quality text. With MDA, you sacrificed the graphics but the characters in text mode were rendered with a 9x14 bitmap rather than an 8x8 bitmap. This resulted in a huge quality improvement. What Compaq did was develop a video board/monitor combination that supported the higher resolution rendering in text mode, but still supported CGA quality graphics. It was a nice compromise for the time.
(The other compromise board was the 'Hercules' board. These boards connected to an IBM MDA monitor, but had a special high resolution graphics mode. They had high quality text, and better high-resolution graphics than CGA, but they weren't as widely supported.)
I enjoy referring to DEC and CP/M machines as turn of the century computers :-). But the recent turn has left us with a linguistic equivalent to the Y2K problem, which in itself is amusing in its own sort of way. :-)
In my childhood, I used to go and watch the four bridges [1] opening and closing in sync. All done by chaps in caps pressing buttons. I imagine the reason for electronic control is to remove the need for chaps. Still causes travel problems [2] even in the reduced state of the docks today.
[2] http://www.wirralglobe.co.uk/news/10988354.Wirral_bridge_to_... To enjoy the full flavour of our local politics, you need to read the comment on this article. The one written in block capitals (purple font colour isn't possible on the Wallasey Globe commenting system)
Yeah, this particular canal has two bridges that have to open in fairly short sequence (the University and the Montlake), and I'd always assumed that it was done by a guys pressing buttons in a booth that triggered some sort of relay system.
Given the presence of the guys and the booths, the age of the bridges and the generally conservative nature of civil engineering, I'm pretty surprised that there's any computer involved in the process at all, let alone one from the early days of personal computers.
It is a guy in a booth. The computer is used to time the whole process and make sure that the right events occur in the right sequence, but it's still human-supervised. Before the 1980s, it was done manually with switches from the control tower.
Much of our infrastructure runs on old computers. That's not necessarily bad. Assuming there's no network connection, configs are left unchanged, and updates are not installed, neither bitrot nor vulnerabilities will bring down the system. A hardware failure could, if compatible replacements aren't available. The latter does occur.
Part of the reason for keeping these old systems is the risk in upgrading. The software has been in service for years so it's relatively battle-tested. New software comes with the risk of bugs and deployment problems. Bugs and deployment problems can mean lengthy shutdowns of critical infrastructure (if it proves impossible to bring the old system back into service as an emergency measure).
So there is some wisdom in letting the old systems clunk along. Up to a point, that is.
I was waiting for the U-Bridge once at about 1am, and I had another ~14 miles to ride on my bike before I got home.
For about 15 minutes, the bridge would open, almost close but then stop with one side about 2 feet up.
I was NOT looking forward to riding around to another bridge.
Another cyclist rode up, watched one of these cycles, and then he looked over at me and shrugged, rode up and did the ~2ft drop w/o incident.
I gave it about 30 seconds, looked around for cops and then went over myself. It was one of the scariest things I have done, and I was just hyper-aware for the rest of the ride home and woke up my housemates to tell them this story.
The Ballard Bridge has been getting stuck more and more in recent months, it seems.
And the University Bridge (or maybe it's Montlake) needs to get hosed down with water every hour on hot days, or otherwise it will expand and be unable to lift.
In the early 90s I saw a control system for a destructive ammunition test center being run by a TRS-80 Color Computer I. They wanted to know if I could update the software and handed me a thick greenbar printout of assembly code. I didn't take the gig.
> "If it ain't broke you still have to upgrade, because delay in doing so will bring compatibly nightmares upon you" - Ivanca
It's a fucking bridge. The only thing the system operating it is connected to is the. Fucking. Bridge. Software works? Check. Software doesn't have to be compatible with anything? Check.
These are systems where feature creep is actually actively avoided. The amount of testing and deployment effort that goes into them is so large that changes are made to them only if there really is a pressing, existing need.
Compatibility with what? It's not being upgraded because it works. There's a reason that most POS systems still look like something out of Tron. They work, they're battle hardened, and they're fast. Upgrade to some new system and only if you're a glutton for pain (unless you have to absolutely change systems for some reason, then you're just forced to be in pain for a while.)
POS systems "battle hardened"? We're all paying through the nose right now because credit card companies refuse to update their POS POS magnetic stripe readers and pendants.
If it breaks, you surely need to fix it (and having plans to fix it or upgrading it due to lack of replacement parts etc is of course neccessary). However, there is also inherent risks in upgrading such appliances, so it should not be done simply because the hardware is old.
> The old-school electric resistors and other parts are wearing out, so that replacements must be scavenged from city shops or eBay, he says. Finnick recalls a 12-hour stoppage, when the city used a backup diesel motor to lift the bridge for boats. Other times, the city dispatched a second bridge tender to help open the bridge manually.
It does break and replacement parts are hard to find.
More so considering the city might want to close down the bridge (eh... open it? lift it?!) remotely in case of an emergency.
They might also want to throw in some traffic condensation sensors and cameras for security and/or tickets and so on...
It wouldn't surprise me if the new system turns out to be less reliable and not last as long, due to all these added things. This isn't aerospace-level, but it's still an application where bugs can be truly dangerous (e.g. bridge attempts to open while you're driving across?)
Why would I want to have to pay mobile when I can carry a much more durable and less valuable piece of plastic that performs just as well and never runs out of battery?
Related question: If I have some old 5.25" disks I'd like to see if I can get information from, what's the way to go? If I buy an old 5.25" drive from ebay or something, is it going to connect to a new PC running Win7 or Linux? I'm guessing no, whether hardware or software issue. So is my only option to find some more old PC hardware, or what?
Now, gaius, why do you need to use a 5.25 inch floppy disc drive instead of a more recent storage medium?
If the answer is something along the lines of 'because that is all the old computer can use' can you clarify if this is a retro hobby or an actual use case?
I like to see how much real work I can do on 8-bit hardware, e.g. BBC micros, VIC 20, so a bit of both. For example I plan technical dives using Buhlmann in 6502 assembler, printed on a dot matrix.
Come the infocalypse, everyone will work this way.
The PC seems alright to me (after all, how much CPU do you need to open a bridge...), but the floppies seem a bit fragile and unreliable, they lose data pretty fast.
That's what I'm thinking too. The fact that it's old and running MS-DOS doesn't seem like much of a big deal to me considering it's not networked and only has one job to do that it's already capable of doing. They should probably just go get a cheep IDE HDD, copy the floppy onto it and boot from there. But other then that, I don't see tons of issues.
This does however seem like a fairly obvious use-case for an embedded solution. No floppies, no moving parts, etc.. It would last much longer then this computer will. It's not like the computer was doing much anyway (I mean, besides signaling motors, I'm not really sure what the computer was even doing).
> The PC seems alright to me (after all, how much CPU do you need to open a bridge...), but the floppies seem a bit fragile and unreliable, they lose data pretty fast.
Floppy disks are fragile, but consider the application.
They're not trying to archive data for 100 years. They just need to be able to load the bridge control program when the computer boots up. If a floppy disk fails, just take it out and insert a duplicate.
There are textile mills that are still running their machines off 3.5" floppies. When they heard that Sony ended production of 3.5" floppies, they simply went out and bought the last production run. Put them in climate-controlled storage, and they had enough to use for the remaining life of their machines.
I got a tour of the Burnside Bridge in Portland twice when I was in probably middle school, a couple years apart. The first time, the operator had a big panel of giant switches and buttons with a relay interlock system to make sure things proceeded in the right order. The second time, the control system had been replaced by a single color LCD touchscreen with what I can only describe as the bridge opening wizard... complete with a "Next >" button.
The long term sustainment / archival aspects of this are really interesting. With the new replacement system, is the company just rebuilding it from scratch? Are there design documents or source code they have access to from the old system? Is the new company going to be required to escrow similar design elements?
I remember an old IEEE article on the NYC subway system. There were portions of the system that was built in the 1900's that still controls the tracks. The MTA was afraid of removing the old system because they had no idea how it worked.
Better to replace it with an embedded controller with no moving parts. Buy a couple of these and flash them with the same firmware and replace the whole thing if it ever dies.
It also includes a picture of the computer: http://i.imgur.com/oK0eknm.jpg