Revolutionary new "sci-fi" data storage technology?

[Farang]

Hi,

 

I recently watched part (the end) of a technology programme on CNN or the BBC.

 

It was about a new fascinating, revolutionary "sci-fi" technology of "etching" an unbelievable amount of data

onto an "indestructable glass(?) carrier.

 

The segment was too short, so I can't be more specific.

 

Has anyone heard of it and knows the precise term for this new technology?

Have not been able to find anything on Google.

 

Thanks!

 

'Mike the Fritz'

 

[EdEarl]

Perhaps this 5D storage on glass disk

[fredreload]

I was wondering why the data cannot be stored as analogue

For instance, if you store 2^10 bits that's 1024 for base 10 and I stored it on a piece of paper. Now when I need that data I just do OCR on the paper and I translate 1024 back into 2^10 bits

In other words, you can use another computer as a storage device, but then you need to translate the base 10 back into base 2, that would take some time and searching and decoding

[Strange]

I was wondering why the data cannot be stored as analogue

 

Well, of course, it can. But there are some technical problems. For example, lets say you want to store a value in analogue form. You could do this by filling a glass with that volume of water, or marking a stick at the position corresponding to that number. That might be OK if you are storing numbers between 0 and 10, or maybe even 1000. But how accurately can you fill the glass or mark the stick? And then how accurately can you measure it. Imagine trying to represent the contents of Wikipedia as a single number and store that as a mark on a stick ...

 

Some flash memory devices use three or four level storage instead of simple 0 or 1 (as a way of increasing storage density) but that is approaching the limit of "analogue" storage, in terms of the accuracy with which the voltage can be sensed reliably (if you end up having to add more error detection and correction then you lose the advantage).

 

 

 

For instance, if you store 2^10 bits that's 1024 for base 10 and I stored it on a piece of paper.

 

How do you store it on paper?

I was reminded of the Long Now Foundation's Rosetta Project

http://rosettaproject.org

[fredreload]

 

Well, of course, it can. But there are some technical problems. For example, lets say you want to store a value in analogue form. You could do this by filling a glass with that volume of water, or marking a stick at the position corresponding to that number. That might be OK if you are storing numbers between 0 and 10, or maybe even 1000. But how accurately can you fill the glass or mark the stick? And then how accurately can you measure it. Imagine trying to represent the contents of Wikipedia as a single number and store that as a mark on a stick ...

 

Some flash memory devices use three or four level storage instead of simple 0 or 1 (as a way of increasing storage density) but that is approaching the limit of "analogue" storage, in terms of the accuracy with which the voltage can be sensed reliably (if you end up having to add more error detection and correction then you lose the advantage).

 

 

How do you store it on paper?

I was reminded of the Long Now Foundation's Rosetta Project

http://rosettaproject.org

Well hmm, you physically write the number onto a paper like 1024, or you print it out from a computer. Then when the computer needs the data you just OCR it back into the computer and decrypt it

 

 

P.S. You just need a computer to understand what it is, make no sense lol

P.S. You'll need a converter program, and have another program run on that converter

P.S. Maybe load from database

Edited November 22, 2016 by fredreload

[Strange]

Well hmm, you physically write the number onto a paper like 1024, or you print it out from a computer.

 

 

It wouldn't be terribly space-efficient. Bearing in mind that we can store the contents of thousands (millions?) of bits of paper on a DVD.

[fiveworlds]

I was wondering why the data cannot be stored as analogue

 

For instance, if you store 2^10 bits that's 1024 for base 10 and I stored it on a piece of paper. Now when I need that data I just do OCR on the paper and I translate 1024 back into 2^10 bits

 

 

Light is a wave and waves can be subdivided into an unknown number of divisions. If I do an OCR of the paper I get a subdivision of the lightwave to a certain precision depending on how good the camera actually is. I can't store the whole lightwave unless the lightwave was actually composed of a fixed number of incredibly tiny and unknown particles making the number of waveform subdivisions non-infinite.

[John Cuthber]

 

Well, of course, it can. But there are some technical problems. For example, lets say you want to store a value in analogue form. You could do this by filling a glass with that volume of water, or marking a stick at the position corresponding to that number. That might be OK if you are storing numbers between 0 and 10, or maybe even 1000. But how accurately can you fill the glass or mark the stick? And then how accurately can you measure it. Imagine trying to represent the contents of Wikipedia as a single number and store that as a mark on a stick ...

 

Some flash memory devices use three or four level storage instead of simple 0 or 1 (as a way of increasing storage density) but that is approaching the limit of "analogue" storage, in terms of the accuracy with which the voltage can be sensed reliably (if you end up having to add more error detection and correction then you lose the advantage).

 

 

How do you store it on paper?

I was reminded of the Long Now Foundation's Rosetta Project

http://rosettaproject.org

That's not the real problem.

In principle I can read the whole of wikipeia into an analogue tape recorder.

 

It's perfectly possible, but it takes a lot of space.

[fiveworlds]

It's perfectly possible, but it takes a lot of space.​

 

 

I was thinking that the phonograph which was an anlog sound recorder could only excite electrons to move at a fairly slow rate. Therefore knowing the theoretical maximum speed of the excited electrons it should be possible to take samples faster.

[CharonY]

That's not the real problem.

In principle I can read the whole of wikipeia into an analogue tape recorder.

 

It's perfectly possible, but it takes a lot of space.

 

There is also the issue of data corruption which is worse for analog system (due to higher ambiguity), which Strange kind of alluded to.

[StringJunky]

 

There is also the issue of data corruption which is worse for analog system (due to higher ambiguity), which Strange kind of alluded to.

Total novice here, but when you lose a bit of analogue it will still run but when you lose a bit of binary it's all buggered isn't it? I'm thinking of a bit music tape recording or a photo; basically you get drop-outs of data but the next bit will still run.

[CharonY]

The problem in terms of data corruption is when the information starts becoming ambiguous. Let's consider stored information in digital and analog format and then introduce noise to the data. In the digital format, one would still be able to distinguish between high and low values as long as the noise ceiling is significantly lower than the high signal. If your data is analog, however, the signal between would become ambiguous, even at moderate noise levels.

In your example losing a bit would not lose out all the data, same with analog data. How much you lose depends on the way the signal is converted and also whether there is error correction.

 

Edit: not saying that digital does not have issues, but I was referring specifically to that.

Edited November 23, 2016 by CharonY

[EdEarl]

Storing data depends on the surface quality of the medium used. If digital, each bit needs to be stored on almost identical areas of the medium, e.g., a glass disk, so the bits are identical. A surface that is irregular will cause errors. Paper tends to be irregular and not capable of storing high density data; whereas, a smooth glass or plastic surface can store very small regular bits/pixels or whatever.

[StringJunky]

The problem in terms of data corruption is when the information starts becoming ambiguous. Let's consider stored information in digital and analog format and then introduce noise to the data. In the digital format, one would still be able to distinguish between high and low values as long as the noise ceiling is significantly lower than the high signal. If your data is analog, however, the signal between would become ambiguous, even at moderate noise levels.

In your example losing a bit would not lose out all the data, same with analog data. How much you lose depends on the way the signal is converted and also whether there is error correction.

 

Edit: not saying that digital does not have issues, but I was referring specifically to that.

Right. I don't know if this is OT but shouldn't more attention be paid to archival durability rather than increasing data density at this point. It all seems so fragile and transient, and everything needs moving to the latest greatest storage technology every so often but none seems more durable than papyrus and burnt wood or rock and ochre paint , of which the latter which have lasted 40 000 years.... and counting.

Edited November 23, 2016 by StringJunky

[CharonY]

It really depends on the specific purpose. There is no one-size fits all solution. There are for example digital storage systems that are supposed to last a millenium (e.g. m-disk). Though it is not clear whether we will have the ability to read them in the future. And obviously more information on papyrus and tablets have been lost than recovered.

 

Assuming you can fit ~50k books onto one disk and you create a thousand of those and seal them away and try to do the same in paper form, the question would be how many of those would still be around in a few hundred years, for example.

Edited November 24, 2016 by CharonY