"The total amount of information stored on its 50 spinning iron-oxide-coated disks--each of them a pizza-size 24 inches--was 5 megabytes. That's not quite enough to hold two MP3 copies of Elvis Presley's "Hound Dog.""
So maybe a dumb question... what were they doing with that thing back in 1956? That's such a small amount of data in such a large space, was this actually used to do anything useful after it was sold?
Don't think of it as two MP3s. Think of it as 25-30 book-length volumes of text.
That was plenty to handle payroll, inventory, sales, etc. for a business, especially since the data model back then was to only keep current, live data on the disk, while archiving old data off on tapes.
Also, reading from the article they are mentioning that there was already tape storage (non random access). I can imagine that the 5MB random access disks were using as "working memory" and then results where still stored in tape.
Those were the 'I/O operators' where I worked... the 'system operators' (computer operators) were the guys that got the big bucks to alter jcl and drink coffee :-P
I also worked at places with dedicated 'print pool' operators, that basically fed paper into printers/bursters/collators/binders.
My first computer with a hard disk was 10MB. You could store the operating system (MSDOS 5) and several programs (I remember Professional Write and QPRO. Not only that, but you could even store games. Prince of Persia was available in 2 floppies (5 1/4") with 720KB each one. You could store the whole game in the hard disk as well... This was around 1990.
Back in the early 90's, the first computer that I owned was a Psion 3 PDA with 256kB of battery-backed RAM, which was shared automatically between "disk" and working memory. Applications were held in ROM. On one occasion I needed to prepare a church address list. I entered the contact details in to a new contacts list (actually a simple flat-file database), exported to the spreadsheet to sort them, then exported to the word processor for formatting and printing. Not only was it possible, but it was easy. So yes, 5MB was a lot.
Before RAMAC, information had to be entered by running a stack of cards through a punched card machine, and answers would arrive in hours or days. RAMAC could find data in seconds, alter it, and move on to find a completely different piece of data. It let enterprises think about data in new ways, mixing and matching it on the fly. Random access made the relational database possible.
The introduction of the term database coincided with the availability of direct-access storage (disks and drums) from the mid-1960s onwards. The term represented a contrast with the tape-based systems of the past, allowing shared interactive use rather than daily batch processing.
It's not a dumb question. Consider the value in moving calculations from paper to a spreadsheet.
That's _about_ what it was like in the early computing days in terms of overall amount of data and the types of general processing that were done with it. Incredibly useful but only applicable to a narrow range of tasks.
Hi. You seem new to HN, and have already written a few such comments. Don't do that here, thanks. Please read the "In comments" section of the guidelines, which starts with "Be kind." Thanks.
Hard drives are at this stage where no one will really miss them given their performance vs SSDs, but still common enough to not attract nostalgia. I think it is still mind blowing that such a cheap device can move a physical head over a disk within a minuscule precision, both in term of position and distance above the disk, and this many times per second. Yes it doesn’t do a million iops, it only does like 10 iops, but I am still impressed we get those 10 iops on a cheap consumer devices that stores TBs of data.
I know you're making an exaggerated comparison, but 10 IOPS would be closer to the performance of a floppy disk than even a 5400rpm consumer hard drive (depending on workload, of course). Imagine if we still had to deal with the performance of floppy drives on a daily basis!
WD says that the track width on their newest drives is about 50nm. The servo has to position the head with an accuracy considerably smaller than that, especially for writes - for reads you can just try again if you're off center. (which is why some drives have different seek time numbers for read and write - they don't need as much "settling time" for reads) For comparison, green light has a wavelength of about 550nm.
"changed the world". a couple of 5MB drives were important, but delivering 4 300MB drives to every credit union, bank, and medium sized business is what changed the world.
I remember watching one of these drives in operation back in the late '60s. It was BIG. The 360s started out with 2311s, then 3330s came with 370s. The multi platter disks were removable.
30 comments
[ 3.1 ms ] story [ 66.9 ms ] threadSo maybe a dumb question... what were they doing with that thing back in 1956? That's such a small amount of data in such a large space, was this actually used to do anything useful after it was sold?
That was plenty to handle payroll, inventory, sales, etc. for a business, especially since the data model back then was to only keep current, live data on the disk, while archiving old data off on tapes.
"Batch processing" often involved reading the data off one tape into the CPU, processing it, then writing it back out onto another tape.
"Computer operators" basically did nothing but load the proper tapes onto the drives at the proper times (plus replacing printer consumables, etc.).
I also worked at places with dedicated 'print pool' operators, that basically fed paper into printers/bursters/collators/binders.
In 1956, 5MB was plent of storage.
You aren’t running npm install on these machines, you’re mostly working with numbers.
Before RAMAC, information had to be entered by running a stack of cards through a punched card machine, and answers would arrive in hours or days. RAMAC could find data in seconds, alter it, and move on to find a completely different piece of data. It let enterprises think about data in new ways, mixing and matching it on the fly. Random access made the relational database possible.
https://en.wikipedia.org/wiki/Database#1960s,_navigational_D...
The introduction of the term database coincided with the availability of direct-access storage (disks and drums) from the mid-1960s onwards. The term represented a contrast with the tape-based systems of the past, allowing shared interactive use rather than daily batch processing.
That's _about_ what it was like in the early computing days in terms of overall amount of data and the types of general processing that were done with it. Incredibly useful but only applicable to a narrow range of tasks.
https://news.ycombinator.com/newsguidelines.html
http://archive.computerhistory.org/resources/access/text/201...
More storage than you thought possible. And it was widely used for about 10 years.
I remember when I saw my fist DEC disk pack how slim they looked in comparison :-)