SFMTA's train system in San Francisco is not only relying on humans to run it, but turns out that a floppy disk has been playing a key role for decades.
First several generations of hard drives really were awful and broke if you stared at them at them wrong. Floppies were more reliable, cheaper, and easy to get.
I’m not sure what time you talk about, but it must be before 5,25" 20MB MFM drives and 30 MB RLL. Which were way more reliable than floppy disks and diskettes. These drives were available in the mid 80’s.
Incidentally 1986 was the year I got my first hard-drive. ;)
And yes they were absolutely expensive in the mid 80’s. The first 20MB MFM i bought was almost $1000 USD. This was in Europe, prices were probably lower in USA.
But I worked as manager for a computer shop, and the 4 years I worked in that, we only had 1 defect under warranty.
I remember it clearly, because it was a woman coming in with her computer saying her hard-drive was defect, most people being somewhat ignorant of computers, often called the whole computer hard-drive, and since defects were rare, I obviously thought she meant the computer. But no she actually knew what she was talking about, and she was the unlucky one to get the only defect hard-drive we ever delivered! OK my memory may not be perfect, there may have been others, but it certainly wasn’t considered a problem in general.
But I remember I heard about defects, very old Seagate drives could get stuck, if that happened, I was told you could tap them against the table flat down, and that would often resolve the issue!!!
Apart from that, I was much more confident with drives back then, because you could actually hear if they were going bad, as the drive would make a suspicious sound in its attempt to calibrate and reread, with a surface scan you could see if they were actually going bad, or it was just some unusual file operation. Generally in time to switch to another drive before actually losing any files. There may be some truth to drives being more unreliable back then, but they were (so to speak) more unreliable in a more reliable way.
Today this functionality is hidden in the SMART system, which I find unreliable. Drives reallocate bad blocks themselves keeping the user ignorant, until suddenly they are completely dead.
First several generations of hard drives really were awful and broke if you stared at them at them wrong. Floppies were more reliable, cheaper, and easy to get.
I’m not sure what time you talk about, but it must be before 5,25" 20MB MFM drives and 30 MB RLL. Which were way more reliable than floppy disks and diskettes. These drives were available in the mid 80’s.
https://www.reddit.com/r/vintagecomputing/comments/x98scz/the_original_20mb_mfm_hdd_died_in_my_286_recently/
Maybe you are mistaking a few bad blocks that were allocated out in the allocation table, for being unreliable?
HDDs before, say, 1986, were junk. Those that came after will still very expensive until the late 90s, when prices started to drop.
Incidentally 1986 was the year I got my first hard-drive. ;)
And yes they were absolutely expensive in the mid 80’s. The first 20MB MFM i bought was almost $1000 USD. This was in Europe, prices were probably lower in USA.
But I worked as manager for a computer shop, and the 4 years I worked in that, we only had 1 defect under warranty.
I remember it clearly, because it was a woman coming in with her computer saying her hard-drive was defect, most people being somewhat ignorant of computers, often called the whole computer hard-drive, and since defects were rare, I obviously thought she meant the computer. But no she actually knew what she was talking about, and she was the unlucky one to get the only defect hard-drive we ever delivered! OK my memory may not be perfect, there may have been others, but it certainly wasn’t considered a problem in general.
But I remember I heard about defects, very old Seagate drives could get stuck, if that happened, I was told you could tap them against the table flat down, and that would often resolve the issue!!!
Apart from that, I was much more confident with drives back then, because you could actually hear if they were going bad, as the drive would make a suspicious sound in its attempt to calibrate and reread, with a surface scan you could see if they were actually going bad, or it was just some unusual file operation. Generally in time to switch to another drive before actually losing any files. There may be some truth to drives being more unreliable back then, but they were (so to speak) more unreliable in a more reliable way.
Today this functionality is hidden in the SMART system, which I find unreliable. Drives reallocate bad blocks themselves keeping the user ignorant, until suddenly they are completely dead.
I agree with it being nice to be able to hear how they were doing. But it’s nice now to manage a thousand computers.