7 ms·

Archiving a Website for Ten Thousand Years

70 points by r721 10y ago

ekianjo 10y ago
Preserving something 10000 years is technically possible,but the real issue is that its extremely likely no one will be able to understand our present language 10 000 years from now. Cultural changes are the real deal breakers.
- mseebach 10y ago
  > its extremely likely no one will be able to understand our present language 10 000 years from now It's an academic pursuit, but the ancient greek philosophers are still readable in their original language. So are all the surviving biblical manuscripts. These are ~2000 year old text. I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000, and so chances that are present-day-English will be readable by anyone who cares enough.
  - Loughla 10y ago
    >I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000 Tell me why you believe this, please.
    - dragonwriter 10y ago
      I'm not the one who made the claim, but there is a much larger starting corpus with more copies of each individual work, and better institutions dedicated to preserving works and better preservation technology. And a lot better geographic spread of those works in modern English than for any language from 2000 years ago. So it's not all that hard to see why one might make that prediction.
      - ekianjo 10y ago
        In case of a nuclear apocalypse and very few survivors remaining to take care of whatever's left, your chances of having anything maintained over 10 000 years is close to nil. And nuclear conflict was an extremely likely scenario just a few decades ago, we have no way of telling it won't ever be again, with way more destructive power than prior civilizations ever possessed.
  - creshal 10y ago
    > It's an academic pursuit, but the ancient greek philosophers are still readable in their original language. What little of them survived, and only because for 2000 years, people made a deliberate effort of keeping the language translatable and the texts around. Etruscan or even Egyptian texts of the same period, not so much (the Rosetta stone was a chance find). Even with Ancient Greek, we're limited to a very, very small fraction of the original corpus. > I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000 Between copyright laws trying to punish people for preserving texts, DRM attempting to make it impossible; and storage moving to very ephemeral media (paper easily survives centuries, magnetic or optical storage maybe decades), I'm not that optimistic that our data will survive the next mass disruption on the scale of the collapse of the Roman Empire.
    - samastur 10y ago
      Well, sort of. We have more papyri preserved than we have people able to read them so a good chunk, if not most of them, have not been read yet. A good deal of those that also contain mundane things like shopping lists so it is unlikely they were preserved because of deliberate long term effort. Regretfully I don't have sources for this other than talking to a friend years ago who was in fact one of those people studying them.
      - creshal 10y ago
        IIRC that's more a phenomenon with clay tablets than papyri – and it's an anomaly because of a rather unique combination of extremely sturdy media (clay tablets especially) and extremely well-preserving environmental conditions (dry deserts). And without the Rosetta stone, they'd be incomprehensible still. Preserving texts and preserving knowledge of a language are two separate problems.
    - mseebach 10y ago
      I think your outlook is unnecessarily bleak. While you're right DRM certainly contributes to the unlikelihood that all, or even most, of current digital culture will survive, that's not the relevant bar to clear IMO. If even a single copy of a 20th century encyclopaedia survives (never mind a Wikipedia dump! No DRM there), the 120th century will have a vastly richer corpus of knowledge about the 20th century than we do about the 1st century (probably any pre-Gutenberg century).
  - ekianjo 10y ago
    Linear B, which is a way older language, is still a mystery to us. Greek does not prove anything. And we could not even understand old Egyptian until we found the Rosetta stone way down in History.
  - ekianjo 10y ago
    > I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000, and so chances that are present-day-English will be readable by anyone who cares enough. Note that even the Bible which is probably the most copied book in History has no remaining specimen older than 1300 years old available. And the reason why we have a great amount of Greek manuscripts is because of Roman historians as well - if we had ended up with a more destructive civilization, you can be sure we would have 100 times less Greek corpus. So your assumption that a large corpus of text will survive aeons is, in practice, very dubious.
- agumonkey 10y ago
  You watched the Onkalo documentary 'Into Eternity' ? They have similar problems, except for nuclear waste. A hard sci-fi leaking into the real world. Very odd. https://youtu.be/HeVPMzJOFrQ https://youtu.be/HeVPMzJOFrQ
  - lmm 10y ago
    You might find http://www.damninteresting.com/this-place-is-not-a-place-of-honor/ http://www.damninteresting.com/this-place-is-not-a-place-of-... interesting, on the same lines.
AndrewOMartin 10y ago
A question along the lines of indefinite preservation use to pop up on Slashdot pretty regularly. They were usually about preservation of photos, and the discussion would devolve into a comparison of types of acid-proof paper. The fact of the matter is that if you're going to put a finite amount of energy into preserving something, then its going to last for a finite amount of time. If you use a stable medium like a metal plate, you have the problem of preserving the encoding in a readable state, if you use a naive easily-interpretable encoding then you'll need more of the medium. Even if you solve these problems for your timeframe, you have the more serious problem of keeping content valuable. The only solution for this for a curator (anyone with some kind of financial, academic or sentimental incentive) to keep all the info, suitably backed up, on whatever the most cost effective consumer technology is. As this tech becomes obsolete then migrate it. If there's no one with an incentive to keep it, i.e. it's perceived as valueless, then the info will degrade over time, but that's basically a description of entropy increasing over time, welcome to our universe. (Score:1, Offtopic)
- Loughla 10y ago
  Slashdot always fascinated me when it came to this topic - the conversations always devolved into paper storage or gold-platinum-lithium eternodisks. As if there was nothing between the two.
  - kijin 10y ago
    Human-readable letters carved into granite plates with a 72-point font will probably last a million years, even a billion years if you're lucky. The real compromise, of course, is between information density and longevity. Your carvings need to be large enough to survive the blurring effect of time, but then you can't store as many bits per square inch.
    - awesomerobot 10y ago
      Until some teenager or ISIS faction comes along and decides to smash it, of course.
    - 10y ago
AlgorithmicTime 10y ago
The real solution is to keep the Internet Archive around and not have our civilization collapse into chaos in the next 10,000 years.
tnorthcutt 10y ago
Semi-related 99 Percent Invisible episode: http://99percentinvisible.org/episode/ten-thousand-years/ http://99percentinvisible.org/episode/ten-thousand-years/ This WIPP site is going to be radioactive for hundreds of thousands of years, though this panel was only responsible for keeping this place sufficiently marked for humans for the next 10,000 years—thinking beyond that timeframe was thought to be impossible.
- llamataboot 10y ago
  This place is not a place of honor. No highly esteemed deed is commemorated here. Nothing valued is here. This place is a message and part of a system of messages. Pay attention to it! Sending this message was important to us. We considered ourselves to be a powerful culture.
chippy 10y ago
I have the suspicion that what it needs is lawyers, not technology. Firms, families and corporations spread across the globe who must maintain these files via archivists. A person should be able to produce something now and pay an amount which is invested into the trust. As more people pay, the trust gets bigger and bigger and becomes wealthier via their own investments. It's aim is to keep this information forever. The cost for a user to store something should take into consideration the length of time in the future that it will be held, and risk. So in the early days, it will be cheaper to store a book than it is when the global network is established in 100 years and much cheaper in 1000 years when the planetary network is achieved. Thats the business case I'm pitching anyhow!
ezequiel-garzon 10y ago
"and a rights arrangement has been made with the Internet Archive for permanent archiving" I wonder why this is not possible in general. Say I offer IA a hash (or several) of my soon-to-appear content, and it doesn't appear in any appropriate registry. Why can't IA then accept a "non-exclusive and irrevocable license to distribute" the content such as [1]? The way things are now things can be deleted by the new domain owner, or in case the uploading user account is compromised. [1] https://arxiv.org/help/license https://arxiv.org/help/license
beamatronic 10y ago
There are not too many computer related things I can imagine will be around in 10,000 years, but ASCII seems like a good candidate.

70 points by r721 10y ago

Preserving something 10000 years is technically possible,but the real issue is that its extremely likely no one will be able to understand our present language 10 000 years from now. Cultural changes are the real deal breakers.

mseebach 10y ago
> its extremely likely no one will be able to understand our present language 10 000 years from now It's an academic pursuit, but the ancient greek philosophers are still readable in their original language. So are all the surviving biblical manuscripts. These are ~2000 year old text. I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000, and so chances that are present-day-English will be readable by anyone who cares enough.
- Loughla 10y ago
  >I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000 Tell me why you believe this, please.
  - dragonwriter 10y ago
    I'm not the one who made the claim, but there is a much larger starting corpus with more copies of each individual work, and better institutions dedicated to preserving works and better preservation technology. And a lot better geographic spread of those works in modern English than for any language from 2000 years ago. So it's not all that hard to see why one might make that prediction.
    - ekianjo 10y ago
      In case of a nuclear apocalypse and very few survivors remaining to take care of whatever's left, your chances of having anything maintained over 10 000 years is close to nil. And nuclear conflict was an extremely likely scenario just a few decades ago, we have no way of telling it won't ever be again, with way more destructive power than prior civilizations ever possessed.
- creshal 10y ago
  > It's an academic pursuit, but the ancient greek philosophers are still readable in their original language. What little of them survived, and only because for 2000 years, people made a deliberate effort of keeping the language translatable and the texts around. Etruscan or even Egyptian texts of the same period, not so much (the Rosetta stone was a chance find). Even with Ancient Greek, we're limited to a very, very small fraction of the original corpus. > I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000 Between copyright laws trying to punish people for preserving texts, DRM attempting to make it impossible; and storage moving to very ephemeral media (paper easily survives centuries, magnetic or optical storage maybe decades), I'm not that optimistic that our data will survive the next mass disruption on the scale of the collapse of the Roman Empire.
  - samastur 10y ago
    Well, sort of. We have more papyri preserved than we have people able to read them so a good chunk, if not most of them, have not been read yet. A good deal of those that also contain mundane things like shopping lists so it is unlikely they were preserved because of deliberate long term effort. Regretfully I don't have sources for this other than talking to a friend years ago who was in fact one of those people studying them.
    - creshal 10y ago
      IIRC that's more a phenomenon with clay tablets than papyri – and it's an anomaly because of a rather unique combination of extremely sturdy media (clay tablets especially) and extremely well-preserving environmental conditions (dry deserts). And without the Rosetta stone, they'd be incomprehensible still. Preserving texts and preserving knowledge of a language are two separate problems.
  - mseebach 10y ago
    I think your outlook is unnecessarily bleak. While you're right DRM certainly contributes to the unlikelihood that all, or even most, of current digital culture will survive, that's not the relevant bar to clear IMO. If even a single copy of a 20th century encyclopaedia survives (never mind a Wikipedia dump! No DRM there), the 120th century will have a vastly richer corpus of knowledge about the 20th century than we do about the 1st century (probably any pre-Gutenberg century).
- ekianjo 10y ago
  Linear B, which is a way older language, is still a mystery to us. Greek does not prove anything. And we could not even understand old Egyptian until we found the Rosetta stone way down in History.
- ekianjo 10y ago
  > I find it plausible that a much, much larger corpus of text written in present-day-English will survive the next 10000 years than that which has survived the past 2000, and so chances that are present-day-English will be readable by anyone who cares enough. Note that even the Bible which is probably the most copied book in History has no remaining specimen older than 1300 years old available. And the reason why we have a great amount of Greek manuscripts is because of Roman historians as well - if we had ended up with a more destructive civilization, you can be sure we would have 100 times less Greek corpus. So your assumption that a large corpus of text will survive aeons is, in practice, very dubious.
agumonkey 10y ago
You watched the Onkalo documentary 'Into Eternity' ? They have similar problems, except for nuclear waste. A hard sci-fi leaking into the real world. Very odd. https://youtu.be/HeVPMzJOFrQ https://youtu.be/HeVPMzJOFrQ
- lmm 10y ago
  You might find http://www.damninteresting.com/this-place-is-not-a-place-of-honor/ http://www.damninteresting.com/this-place-is-not-a-place-of-... interesting, on the same lines.

AndrewOMartin 10y ago

A question along the lines of indefinite preservation use to pop up on Slashdot pretty regularly. They were usually about preservation of photos, and the discussion would devolve into a comparison of types of acid-proof paper. The fact of the matter is that if you're going to put a finite amount of energy into preserving something, then its going to last for a finite amount of time. If you use a stable medium like a metal plate, you have the problem of preserving the encoding in a readable state, if you use a naive easily-interpretable encoding then you'll need more of the medium. Even if you solve these problems for your timeframe, you have the more serious problem of keeping content valuable. The only solution for this for a curator (anyone with some kind of financial, academic or sentimental incentive) to keep all the info, suitably backed up, on whatever the most cost effective consumer technology is. As this tech becomes obsolete then migrate it. If there's no one with an incentive to keep it, i.e. it's perceived as valueless, then the info will degrade over time, but that's basically a description of entropy increasing over time, welcome to our universe. (Score:1, Offtopic)

Loughla 10y ago
Slashdot always fascinated me when it came to this topic - the conversations always devolved into paper storage or gold-platinum-lithium eternodisks. As if there was nothing between the two.
- kijin 10y ago
  Human-readable letters carved into granite plates with a 72-point font will probably last a million years, even a billion years if you're lucky. The real compromise, of course, is between information density and longevity. Your carvings need to be large enough to survive the blurring effect of time, but then you can't store as many bits per square inch.
  - awesomerobot 10y ago
    Until some teenager or ISIS faction comes along and decides to smash it, of course.
  - 10y ago

AlgorithmicTime 10y ago

The real solution is to keep the Internet Archive around and not have our civilization collapse into chaos in the next 10,000 years.

tnorthcutt 10y ago

Semi-related 99 Percent Invisible episode: http://99percentinvisible.org/episode/ten-thousand-years/ http://99percentinvisible.org/episode/ten-thousand-years/ This WIPP site is going to be radioactive for hundreds of thousands of years, though this panel was only responsible for keeping this place sufficiently marked for humans for the next 10,000 years—thinking beyond that timeframe was thought to be impossible.

llamataboot 10y ago
This place is not a place of honor. No highly esteemed deed is commemorated here. Nothing valued is here. This place is a message and part of a system of messages. Pay attention to it! Sending this message was important to us. We considered ourselves to be a powerful culture.

chippy 10y ago

I have the suspicion that what it needs is lawyers, not technology. Firms, families and corporations spread across the globe who must maintain these files via archivists. A person should be able to produce something now and pay an amount which is invested into the trust. As more people pay, the trust gets bigger and bigger and becomes wealthier via their own investments. It's aim is to keep this information forever. The cost for a user to store something should take into consideration the length of time in the future that it will be held, and risk. So in the early days, it will be cheaper to store a book than it is when the global network is established in 100 years and much cheaper in 1000 years when the planetary network is achieved. Thats the business case I'm pitching anyhow!

ezequiel-garzon 10y ago

"and a rights arrangement has been made with the Internet Archive for permanent archiving" I wonder why this is not possible in general. Say I offer IA a hash (or several) of my soon-to-appear content, and it doesn't appear in any appropriate registry. Why can't IA then accept a "non-exclusive and irrevocable license to distribute" the content such as [1]? The way things are now things can be deleted by the new domain owner, or in case the uploading user account is compromised. [1] https://arxiv.org/help/license https://arxiv.org/help/license

beamatronic 10y ago

There are not too many computer related things I can imagine will be around in 10,000 years, but ASCII seems like a good candidate.