Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Print it out on acid-free paper with a stable, acid-free ink. Have it bound as a hardback, and seal it into a waterproof container. Entrust it to one of your children, tell them to keep it in a safe-deposit box and take it out once a year to share with their children, and to pass it on to their children with the same instructions.

If you have it electronically, the absolute best case in 500 years is that it will be a relatively easy job for software archaeologists or historians to decode, assuming it's been periodically backed up to new media for all those years. The most likely case, though, is that in 150 years, the servers it was stored on, which have not been running for 80 years, will be picked over and/or melted down for precious metal contents by a tinker who wanders between mud-hut villages repairing their ancient metal pots in exchange for dried fish.



> “What was there to say? Civilization was like a mad dash that lasted five thousand years. Progress begot more progress; countless miracles gave birth to more miracles; humankind seemed to possess the power of gods; but in the end, the real power was wielded by time. Leaving behind a mark was tougher than creating a world. At the end of civilization, all they could do was the same thing they had done in the distant past, when humanity was but a babe: Carving words into stone.”

Death's End -Liu Cixin - The third novel in the trilogy staring with The Three-Body Problem


I recall some scifi where some Diety/SuperAI left 'commandments' for humanity carved into giant monuments made of diamond. I presume that was considered the only thing that would survive deep time.


There was an interesting article online a few years back, can't find it now. It claimed that humans can't make anything that will last more that 16 million years. This includes any kind of nuclear pollution. Sure we might get lucky like Jurassic fossils, but not intentionally.


While we're on the topic of nuclear pollution, "nuclear semiotics" is an interdisciplinary field of research focused on creating a "warning message intended to deter human intrusion at nuclear waste repositories in the far future, within or above the order of magnitude of 10,000 years."

While 10K is a few orders of magnitude greater than 500, I imagine the problems may be similar... if not more extreme.

Source: https://en.wikipedia.org/wiki/Long-time_nuclear_waste_warnin...


I haven't seen the article, but I don't buy it, there's no likely events that will harm Voyager 1 and 2 in that time period, for instance. They could get really unluckly and hit a star/planet, but it's not at all likely. Along a similar vein many of the satellites that we put in graveyard orbits around earth at the end of their useful life will also plausibly last that long, though there is a lot more debris for them to collide with.

We've also certainly... redistributed... many metals and other things around earth. Perhaps "large concentration of iron over what use to be new york" doesn't count, but arguably it should.


> I haven't seen the article, but I don't buy it, there's no likely events that will harm Voyager 1 and 2 in that time period, for instance.

Sure, but nothing will encounter or observe them ever again.

Things we put into heliocentric orbits are likely to be forever, too.

> Along a similar vein many of the satellites that we put in graveyard orbits around earth at the end of their useful life will also plausibly last that long, though there is a lot more debris for them to collide with.

While they won't decay from drag in a few million years, tidal forces and photon pressure become significant over time.


> Sure, but nothing will encounter or observe them ever again.

like trees falling silently in desolate forests?


More than that. There are plenty of atoms and photons and other particles observing those trees, much less for Voyagers.


My favorite weird theory is the "Siluran hypothesis," which states that our industrial revolution was not the first. There are some events in sediment that look like what our industrial revolution will look like. (many millions of years ago, so yes, reptiles)


Yes.

Trying to grasp a time span of one million years feels impossible, in the context of development of culture and technology.

It seems quite unlikely to me that we are the first such development on this rock.


Seems likely to me that we are, given the age of the oldest fossils.


Eschaton in Singularitys Sky?

"You shall not meddle with the lightcone of the recursive time-travelling super-intelligence you created?"


"Thou shalt not violate causality in my historic light cone," I believe.


Ah yes, that was it.



The things listed there are completely worthless though. In case of system collapse I'd want to know practical things like agriculture and simple power plants.


Isn't it actually the case that diamond isn't the strongest material? Just scratch-resistant if memory serves. Not that it's weak either, of course, but I'm curious if it's the best choice.


The claim is usually that diamond is the hardest material, which for the most part it is.

Strength is usually defined by the application and usually focuses on tensile / compressive strength. This is also why rebar is used in concrete, concrete has excellent compressive strength, but poor tensile strength.

There's lots of metrics for strength, also resistance to fatigue is often an important metric.

Diamond has excellent hardness, and compressive strength (diamond anvils), it's very poor in most other metrics of strength and evaporates above 450 degrees, so it's not good for anything hot.


I don't think it's necessary to go into all this trouble. Tablets made of the much, much cheaper clay are still accessible today—4000 years after they were created.


But it wasn't just the act of creating some clay tablets - those particular clay tablets happen to have run the probability gauntlet successfully and came out the other side - just making a clay tablet and putting it in your closet is unlikely to produce the same results. The place you put the tablet is more important than the fact that it's a clay tablet in itself.


Mass production.


Posting spoilers without a warning is extremely rude.


Tbf, as someone who's never read or heard of the book, it just seemed like an interesting quote about a hypothetical future for humanity. Definitely didn't occur to me it might be a spoiler.


Ive read all 3 books, though a while ago. I have to think hard about what this could be referring to.

I remember, now, but it wouldn’t be a spoiler for me.


I haven’t read any of them, but from what I’ve heard of the plot of the first, it seems reasonable to want a spoiler tag here


without context would it's pretty meaningless imho


Worse, the past tense of "begot" is "begat."


> Past tense of beget is begot or begat

https://pasttenses.com/beget-past-tense


These were such legit books. This description, in context, being like... the 5000th reason why.


Uuugh, I just finished reading them and, while the first one was pretty good, the second one dragged a bit and the third was interminable! It felt like the author had a laundry list of ideas that he hadn't fit into the story yet, and with the third book was on a mission to cram in every last one. It's something like 2-3 times longer than the first one and by the end I was only reading out of determination to finish it.

There were also a lot of odd assumptions/conclusions about how people would behave but I put that down to a different cultural mindset (and in this sense it was pretty interesting), however some of the recurring social themes stretched credibility a bit. Also for a book that's generally held up as being hard sci-fi, the actual science aspect had some glaring flaws in pivotal events which really strained belief. (Well, either that or my understanding of astrophysics is way off...)


I think is basically the right idea, but a lot of the details are wrong.

For example, book archivists recommend against storing books in waterproof containers. https://www.sparefoot.com/self-storage/blog/3456-the-sparefo... "Be careful storing books in plastic containers. Because plastic containers form an air tight seal, any moisture residing inside your books will be trapped. If your books are not completely dry before placing them inside a plastic container for book storage, they may develop mold or mildew. If using plastic containers, make sure to insert silica gel packets to absorb moisture."

Instead, archivists recommend acid-free archival boxes. (Gaylord is a recommended brand.)

The other point is that you shouldn't just have one copy printed. Like any important data, you'll want to have backups.

At a minimum, if you have multiple children, giving one copy to each child is sensible; it would make sense for each person to have at least two copies, one to keep at home, and another to keep somewhere else that would hopefully remain safe.

If your document is suitable for public consumption, you could pay for a vanity press to make it available for publication, arranging to have copies stored in libraries. As of today, arranging to have your book archived in the Library of Congress is a reasonable approach to ensuring that some professional archivist will at least try to take care of your book.

(They'll also attempt to digitize your book, and archivists will attempt to care for the digital collection, but, as you noted, there's no way to be sure that any digital equipment will be working 500 years from now.)

But, if your thing is suitable for public consumption, consider another problem: will your great-grandchildren care to read what you wrote? Probably the only way to ensure that anyone will care to read your work is to be/become famous, and to write a successful work with millions of copies. (This also incidentally solves the archival problem: people care about protecting and preserving historically important documents.)


The United States government studied this question and came to the conclusion of... acid free paper. So it's now law that "permanent" documents in the US are to be stored on acid free paper:

https://www.loc.gov/preservation/resources/rt/perm/permpapr....

Some of us who remember the 5.25" floppy disks, the 3" "floppy" disks, the HUGE Zipdisks, the 5.25" spinning platter drives, the 2.5" spinning platter drives, then the 2.5" SSD drives, and now the M.2 SSD drives... there's clearly no hope of any digital medium lasting 500 years. It's hard enough to read data from a drive built 20 years ago!


> It's hard enough to read data from a drive built 20 years ago!

Because you don't have the hardware lying around right now. It's still quite possible to read all popular formats today and most of the ones you mentioned can even connect to the same SATA bus available on basically all normal mainboards. And those are mediums which are built for ~10 years (you'll find bit rot by then, which is what is actually preventing you from reading the data). Tapes, for example, are meant for long-term storage, are still in use and can be readily read.

To add to this, books have the same problem. Have a look at the declaration of independence: The font and language is already quite different from what we use today, and that's only from ~250 years ago. Plus the paper would probably not hold up to normal handling anymore.


Yea but there was that guy who built an arduino based contraption to read a Cray Supercomputer hard drive.


Acid free paper is a big improvement, but it's not gonna last 500 years. It's expected to last 200 years.

Archival acid-free paper -- paper with cotton added to it -- can last up to 1000 years, but be prepared to pay $2.50 for 1 page[1], so a 400 page book will cost you $1000 for one copy, just for the paper in the book. This type of archival paper might be useful for important contracts or deeds, or legal documents.

But then you have to worry about ink. Normal ink will break down as well in 1-200 years, so you need archival ink. This boils down the difference between pigment and dye based inks. Dye based inks are more expensive but more resistant to UV light.

In the end, light, heat, water, will destroy everything.

The way to make something last is social in nature -- building long lasting institutions and cultures that value your website and archive it. These must be able to preserve themselves, which means traditions that forcefully apply to successive generations.

It is not a technological problem, but a social problem. However liberalism is completely unequipped to solve this problem, because in order to create something that outlives you, you must bind future generations to some course of action they haven't agreed upon yet. So a liberal society cannot have long lasting institutions or traditions, it always eats itself -- there is another trending hackernews topic about Jefferson being cancelled. Well, of course Jefferson will be cancelled. So will Martin Luther King. So will everyone else. Absolutely nothing can last in a liberal society that believes moral progress is possible -- e.g. that children can be more moral than their grandparents. If you look at durable societies of the past, they all believed that the grandparents were wiser and more moral than they. That allowed them to preserve traditions and texts. The contingent that believes the opposite does not preserve texts, they burn them/cancel them/or otherwise try to erase them.

So once you stop thinking in terms of "what is the best way to do X" to "what is the best way to make sure my mechanisms of doing X will last", then you end up with completely different solutions for the same problem, because the social technologies of preservation are often the exact opposite of the social technologies of progress and improvement.

So no, your website is not going to last 500 years.

[1] https://www.archivalmethods.com/product/archival-paper


I believe you have misread that table. The prices (which depend on size) are all per package of 100 sheets. So you can reduce that $1000 to $10 as an order of magnitude estimate.

That's still rather expensive just for paper, but it might make a nice project, as a one-off, to print and bind a book to that sort of standard.


Microfiche can last 500 years. So can engravings in nickel, which can be read with an optical microscope:

https://en.wikipedia.org/wiki/HD-Rosetta


Cancelling is an online-driven social phenomenon that has existed for- a decade, maybe? Rather odd to view all of human civilizational progress through the lens of one overexposed, extremely online trend.


> Cancelling is an online-driven social phenomenon that has existed for- a decade, maybe?

Ignoring the tone of the parent comment, i don't believe that this subcomment is necessarily correct either.

Sure, "cancelling" is a modern term, but i'm pretty sure that exile to any degree and censorship of the exiled people (or worse) have been prevalent throughout the history of humanity, be it reporting them to the communist spies for non-communist rhetoric, claiming that they're guilty of blasphemy against the church or declaring them a witch.

Thus, regardless of where any of us stand in regards to the political or ideological climate, viewing preserving information for a long time as a social problem definitely has some merit to it. For example, just look at this from almost a century ago: https://rarehistoricalphotos.com/stalin-photo-manipulation-1...


> This boils down the difference between pigment and dye based inks. Dye based inks are more expensive but more resistant to UV light.

I think it might be the other way around: pigment-based inks are opaque particles bonded to the paper, and so even if their colour changes/fades they're likely to still be legible. Whereas dye-based inks, when they fade, can completely disappear. (One advantage of dye-based inks is that they have a larger gamut.)


I'm not sure the latter argument summarized "in liberal societies cultural values change, so no information can be preserved" makes sense. We certainly can access certain culturally preserved texts that are thousands of years old with a pretty high fidelity, even when those texts may contain positive depictions of acts that are considered morally abhorrent today (slavery, genocide, murder, rape, etc) or simply physically falsifiable worldviews.


When you are talking about millions of works, you need to approach it statistically.

That we have some works preserved does not mean that we are succeeding at preserving works in general. Nor is the fact that we have books preserved by previous generations a guarantee that the same books will be preserved by current and future generations.

So one way to think about this is that society has a big junkyard into which it throws documents. And sometimes, people do some archeological digging and reconstruct portions of some documents or bring them to light and popularize them, bringing them out of the junkyard, with a bias towards those documents that reconstruct whatever fashions are happening in that society.

That does not mean that the society as a whole is preserving documents well, even though you will always have some ancient texts available, and people are still mining the junkyard.

Moreover once we move to information stored online in the present climate of account deletion and deplatforming, we are again reconstructing the ingredients for a dark ages as mass deletion of online data is a lot easier than burning individual books. Many of the manuscripts we have were literally pulled out of ancient trashheaps or were written over by other texts. That's a lot less likely to happen with modern technologies.


You weren't originally making a statistical point, you said:

>Absolutely nothing can last in a liberal society that believes moral progress is possible

Which is clearly not true.

In any case, toward the more general claim that this post is making, which I will summarize as "Authoritarian societies are statistically better at preserving works in general" this is also demonstrably not true in the wide statistical sense. They are just as likely (if not more) to deliberately discard works that disagree with the general ethos of that society. Heresy is not really a big deal in a liberal society, but it will get some books burned and practitioners slaughtered (reducing social transmission of ideas) in a autocratic theocracy, for example.


> Which is clearly not true.

Huh? Pounding your fists on the table is not, you know, an argument. That not everything is deleted after 50 years is not proof that something can last permanently.

> In any case, toward the more general claim that this post is making, which I will summarize as "Authoritarian societies

That's a complete misunderstanding of the post, which contrasts traditional societies with liberal societies, by changing the subject to authoritarian societies. As if this was the only choice.

Authoritarian societies are not the opposite of liberal societies. In fact authoritarian societies -- e.g. communist and nazi societies or other societies in which individuals are micromanaged -- only came into existence in the 20th Century when the technology for mass micromanagement became possible. And whatever words you use to descibe authoritarian societies, "preserving tradition" is not one of them. These are big book burning, history-rewriting societies because they try to address the issue of social reproduction by the fist of centralized top-down control that monitors and micro-manages every aspect of life. Human beings are not compatible with that type of control, which is why authoritarian societies don't last very long -- the Russian czars lasted a thousand years whereas communism lasted only 70. And people were much more free under the czars than under communism, because the czars never tried to control every aspect of social life, and never needed to setup networks of gulags, or a vast secret police force, or party functionaries throwing people in prison for skipping work without a doctor's note.

There is the old saying "the right talks about authority, the left talks about control". For a society to be able to preserve knowledge, it must develop long lasting institutions and a culture that reveres the past and seeks to preserve it. Therefore while you need a cultural respect for authority, you cannot actually have a centralized system of social control. So you need basics like "honor your father and your mother" to be taught in churches and other civil institutions, but you cannot have a world in which political meetings decide which author is going to be erased from history today or whether so-and-so is allowed to attend university because their parents were class enemies.


500 years from now, at a grand reconstruction ceremony:

"And there seems to be text scrolling and blinking upon the display surface... never.... gonna... give you up? What do you think it means, Tharl?"


The second part of your answer is hilarious and depressing.


Never underestimate the power of basic human needs over layers of abstractions.

Software's intangibility will be its downfall when our modern society eventually collapses.


It always brings me joy to read doomsday prophecies from decades, centuries, and millenia past. They were legion -- so confident, yet so wrong.

I hope your predictions bring similar joy to future readers in the decades, centuries, and millenia to come.


The thing is, occasionally those predictions have been right and society reverts to a state less advanced than it was for hundreds or thousands of years. A lot of the technology from Ancient Rome, building techniques, steam engine technology, etc. were lost for thousands of years. It’s totally conceivable to me that 500 years from now scientists have access to a wealth of electronic media, but haven’t invented the computer and thus can’t read it.


They're starting to break the code on it, but there are concrete docks that Rome built 2000 years ago that still exist today - we have trouble building salt resistant concrete docks that last ~100 years.

Apparently it has to do with using a certain type of volcanic ash in the concrete...


We've known for a good while how Roman concrete works. The reason that nobody uses it is that the economic incentives aren't there. Roman concrete is more annoying to work with and takes way, way longer to usefully solidify than cheap modern concrete. Yes, it will last longer, but almost nobody these days cares about paying through the nose for a building that will last more than 100 years when you could just force future generations to pay to rebuild it when it collapses.


It's a test of intergenerational commitment, because you are asking the generation that bears the cost of construction to build something that benefits many future generations. But the future generations, even though they enjoy all those beautiful old buildings provided to them by their ancestors, are tempted to go cheap for the building they need to build themselves.

So as people become disconnected from the great chain of being that connects your ancestors to you and you to your descendants, they start to build ugly, disposable buildings.

Perhaps one approach might be to provide lower interest rates to buildings based on the expected life and maintenance cost. So a building that is expected to last 500 years would have a much lower annual interest burden than a building expected to last 70 years. That would require some type of government guarantee for the 500 year bond.


> we have trouble building salt resistant concrete docks that last ~100 years

Because of failures in technology/ability, or lack of incentive/motivation? It isn't obvious to me that anyone cares to build long-lasting structures..


I really wish the Romans had invented apocalyptic fiction, in the modern sense. Would've been interesting to read what their anxieties were in non-poetic, scrutable terms.


That's basically what the Book of Revelation is, although it's referring to Romans rather than being written by them.


Too poetic and inscrutable.


I was very young at the time, but I was absolutely convinced that the world would end at 2020. It didn’t, and that deeply ingrained in me a very substantial amount of skepticism from doomsdaysayers. The only doomsday I believe in now for absolute certainty is the heat death of the universe, but no one reading this will be around for that.


Luckily, you are much older and wiser today, a full year later! :)

I am assuming you are thinking of 2000, though, and only mistyped 2020 because... well, that one was weird, but only a year ago.


Haha! Yes, you are correct, I meant 2000. To be fair, had someone predicted the end of the world in 2020 that would have seemed extremely accurate in retrospect.


There was quite a lot of doomsday talk during the Cold War, and there has been periodic environmental doomsday predictions since the 60s, starting with overpopulation and chemicals like DDT. Interspersed with that was AI and nanotech apocalyptic concerns. Climate change is the latest. The idea that civilization will manage to avoid the worst case scenarios and find its way through is not as exciting. You probably won't sell as many books or public appearances that way. And it doesn't make for the best Black Mirror episodes. Although there are a couple exceptions.


This is a little bit like the bird who avoids the cat for 3 days in a row using that as evidence that cats can't eat birds. Be careful extrapolating possible futures only by sampling past events. That's why we have physics, because we are notoriously bad at doing that with just our intuition.

Climate change is a pretty simple proposal with pretty simple and direct evidence. Carbon traps light, which means more heat is trapped within the atmosphere. We can measure the amount of carbon in the atmosphere. That atmosphere will obey physics. More heat will be trapped.


> Climate change is a pretty simple proposal with pretty simple and direct evidence. Carbon traps light, which means more heat is trapped within the atmosphere. We can measure the amount of carbon in the atmosphere. That atmosphere will obey physics. More heat will be trapped.

Yes, but that's different from an apocalyptic scenario. Predicting the population increase in the 60s was also scientific. Predicting that we couldn't feed several billion people turned out to be wrong. Increased sea level rise and more extreme weather is one thing, predicting that human civilization ends and we all die is an entirely different matter.

Apcoalytic scenarios make worst case assumptions. That we'll fight a nuclear war which will trigger a nuclear winter, or that population won't peak and there's no green revolution, or that feedback loops will lead to a hothouse Earth scenario. But the likely projections don't show that.


It's funny that there is a set of people who think we can't survive +2 degrees of heat but can easily be an interplanetary species surviving in Mars.


I see no reason to suspect there's much overlap there, and in any case this misrepresents both arguments. Climate change doesn't need to render the planet uninhabitable to all life in order to devastate human civilization; if we are forced to abandon the coasts, it will be the biggest refugee crisis in human history by orders of magnitude, and hundreds of millions will die in the resulting chaos and scramble for resources. Meanwhile, Mars is a barren tomb world, and, at best, living in a colony there will be a physically and psychologically exhausting prospect, and has no chance at exhibiting any real self-sufficiency for hundreds of years at best, let alone anything resembling civilization.


Nobody thinks we can't survive 2 degrees of heat. After all typical climate fluctuations of more than 100 degrees are common in human experience. What you're describing is a strawman argument.

The actual concern is 2 degrees of average heat change and how that will change the web of systems that we rely on for current human civilization. Break enough subsystems in there and the human experience becomes dramatically different.


> There was quite a lot of doomsday talk during the Cold War,

Unfortunately, all the existential risks from the Cold War are still around, plus we have new ones we didn't know about or didn't take seriously back then.


Society as we knew it definitely took a hit in 2020 though.

Collapse is not a singular event at a fixed point in time like the start of a war would be, it's a process.


Which is why some people have taken to using the term crumble rather than collapse, crumbling happens at different rates in different places and can sometimes, at least temporarily, be fixed.


Yeah, that's a good term and a good way of thinking about it. When all the crumbling overlaps for a century, you get Gibson's Jackpot.


Unfortunately, this entire thread is about how it's unlikely for future generations to get joy from our ramblings on the web today.

Such a shame!


I hope so, too, but it's not the way to bet.


Hopefully some remnant of humanity will cast out among the stars with ultra-high density storage holding archives of most of human knowledge.


The Voyager Golden Record is an attempt to do this. It's a record that contains sounds and images to portray life on earth [1].

This is a present from a small, distant world, a token of our sounds, our science, our images, our music, our thoughts and our feelings. We are attempting to survive our time so we may live into yours - President Jimmy Carter

[1] https://en.wikipedia.org/wiki/Voyager_Golden_Record


a song i composed is being sent to the moon on an SD card or flash drive of some sort, i was told. Realistically if there is some faster-than-light mechanism, any interested alien or human could just "fast forward" and "rewind" through time to listen/watch our broadcasts from the appropriate distance from the original location.

Barring that, I'm guessing the equivalent of stone tablets, or the golden record are the best bet. I wouldn't place any bets on anything that requires magnetism or electricity to survive in open space for long enough to matter. Like, platter bit-flips due to radiation from outer space are a thing, and current leakage would eventually render something like SSDs unreadable.

So hard copy for earth, radio and hard copy for the universe?


How is that going to repair anyone's kettle?


It may help if the kettle is somewhere between the Earth and Mars :)

https://en.wikipedia.org/wiki/Russell's_teapot


And some ancient civilization will entertain themselves by bidding against each other to win the rights to the contents of the archive.


When society collapses, software of any kind will probably be among the least of our concerns.


Wait so I shouldn't be spending all this time learning vi?


I guess there’s always the chance that society is collapsing because not enough people are trained with software to fix whatever the problem is.


Yes, the idiocracy scenario does appear more and more relevant when you go out and look around and realize that 99.9% of civilization treats tech just as black box appliances....


No, you should learn emacs.

M-x purify-water is gonna save a lot of people one day.


depends on how gullible you are..


> The most likely case, though, is that in 150 years, the servers it was stored on, which have not been running for 80 years, will be picked over and/or melted down for precious metal contents by a tinker who wanders between mud-hut villages repairing their ancient metal pots in exchange for dried fish.

somewhat reminds me of: https://en.wikipedia.org/wiki/A_Canticle_for_Leibowitz


That's such a good book. I first read it after reading Wikipedia's "Terminal Event Management Policy" (a humorous page) which noted:

> In the longer term, archivists are encouraged to pool the resources of the encyclopedia for the common good. A suggested model of collaboration is based on the Leibowitz-Canticle report of the 1960s, which suggests pooling of archives in a centralized location, which might serve as a hub for reconstruction.

https://en.wikipedia.org/wiki/Wikipedia:Terminal_Event_Manag...

After reading that I quickly found a copy in the library to enjoy. Not many sci-fi books from 1959 have remained as relevant and fresh 6 decades later.


I’d alternatively recommend microfilm. It’s specifically engineered to last ~500 years and has much higher storage density than paper and ink. The technology needed to read it is fairly low tech and trivial and the format is quite durable.


Particularly if you store it at 0°C.


My favorite part of this excellent answer, and to me the most important part, is the yearly reading tradition.

The other idea might be to encode your website in a song that's popular today but complex enough to be worthy of study by future generations.


Have you been playing a lot of Fallout lately?


As entertaining as this is, I have a bit more faith in archive.org.


Serious question, but why?


I think it's a historically significant dataset. We've seen other datasets be preserved, such as GitHub arctic vault.

I agree that it's tenuous. I would give it 20% odds of hitting the 500 year mark at best. And I don't think all of the data will survive.

But if archive.org ever becomes unsustainable to run, the existing data will likely be preserved. Lots of companies will be incentivized to continue hosting the data, as it's excellent PR if nothing else. They don't need to continue gathering the data, just host it.

Hosting is only going to become cheaper as t -> infinity, and given the massive amount of compute I've seen Google wield, it's hard to imagine that an operation like archive.org can't find some way to be preserved.

All that said, the biggest threat is sudden data loss. This only works as long as the data doesn't get lost. Has archive.org posted their operations policies anywhere? It would be interesting reading.


Archive.org has substantial legal risks too.

Imagine a future gdpr-like policy that gives people's descendants ownership and copyright over everything they've said. Suddenly every word written into archive.org has an owner, who might come and sue archive.org or its managers. Soon every person alive has some grandparent who wrote something in the archive and some of them are wanting compensation for all the decades archive.org has been distributing grandpa's words for free.


It's less about the "getting it done" aspect. It's more about are they going to be around in 50/100/500 years. Will the tech be around that long? Will they keep up with the conversion of old tech into new tech? In my opinion, any kind of digital archive is just not a sound way to go about it. Analog all the way for long term archival.


not sure I share your sentiment about companies hosting the data, considering what happened to Geocities and others.


Mm, you're right, but Geocities might be less interesting to historians than an archive of all internet history.

Also, as someone who has trained a few large GPT models, I think ML has a chance of preserving a lot of this data. Training datasets are only growing larger and larger, and although those aren't updated (yet), there's no reason to think they won't last for a long time.

I imagine that in 500 years, imagenet2012 might still be around as a historical curiosity, at least somewhere.


Well, everyone was hyped on perl at the turn of the millennium. Yet not many people write it anymore. I keep waiting for the re-surge, but it just doesn't look like it is going to happen.

At nuclear waste sites, even the feds have come up with a few ways of saying "Don't enter. It is bad" with different languages, pictorial signs, and such.

It is really tough to figure out what the next few hundred years looks like. And to be a bit political, I don't think anyone saw the invasion of the capitol building in January.

It isn't easy to predict the future. With the original poster in mind, I think the best bet would to be with archive.org.

Maybe archive.org should provide this service. It could be a way to generate revenue - say "here is a thousand bucks, keep it for eternity."

I'm not sure I would want my thoughts to last that long though.

(And I'm still not sure that it would survive for more than a few hundred years.) Maybe the right thing to do is do something so great for society that they want to write books on you (eg: George Washington).


> Maybe the right thing to do is do something so great for society that they want to write books on you (eg: George Washington).

Rarely do the people that seek this actually achieve this. --quoth the raven


So true. Anyone that is smart enough to be president is smart enough to not be president.


I think the chance of future generations having the motivation to continue preserving OP's specific website would be quite low but there would be a much greater motivation to maintain a large organised archive.


That works until the FBI raids the entire safe deposit bank and steals everyones possessions.


Put it on Vellum!

Create a social ritual to read it so it can stay pliable. Vellum lasts forever


I see no technical reason for why we can't create some e-reader that will keep your library much safer for much longer than paper. I see no reason why we can't make some that last for a millennium, if the power supply and storage aren't included and it's kept off in some dry place, out of the Sun's light and never overheat.

But well, there is no demand for tech that will last for a millennium. In fact, people are pushing for degradable tech that won't stay as waste after it stops being useful instead.


Electronics aren't that sturdy. Hard drives demagnetize, solid-state storage decays, and there's always the chance that a stray cosmic ray will fry something. Even if we could construct something that sturdy: After a millennium, how would anyone know how to operate it? Would we include an instruction manual? Printed on what sort of paper?

Microengravings on metal plates [1] will be more durable than electronics could ever be, and easier to read as well. No power source necessary — just a lens.

[1]: https://en.wikipedia.org/wiki/HD-Rosetta


Program once memory is quite resistant, and of course, plain old ROM is a rock, literally. Transistors do not deteriorate unless you use them, just like resistors and a lot of the available capacitors.

On how to operate it, it's an e-reader, you turn it on there's the manual there. There is the really difficult part of telling people "put 12V DC here, ground on the outside", but it's trivial compared to the problem of people understanding the books inside.


> ground on the outside

A diode bridge (https://en.wikipedia.org/wiki/Diode_bridge) can easily take care of polarity inversions. Voltage is a harder problem.

Since we're going through flights of fancy in these threads, another option would be to add a built-in solar panel to your hypothetical e-reader. Expose it to enough light (and you can probably assume the lighting won't be much brighter than a tropical midday sun) and it turns on.


Yes, but then you would both incentive people into placing your device under sunlight and need some of the shortlived kind of capacitors.

It should still be possible to make a solar powered e-reader long-lived, but a replaceable power supply is a good idea even if your device has to last for a couple of years.


But 21st-century ROM isn't human-readable without a 21-st century computer. Would that survive a millenium? It's not enough to have a storage medium that lasts 1000 years; the entire tech stack has to.


A more accessible approach than the High-Density Rosetta might be the M-DISC.[0] It's essentially a DVD/Blu-Ray which materials engineered to last significantly longer than standard ones. I attended a chemistry lecture by Prof. Matthew Linford (BYU) and from the gist of his description,[1] the materials that they were using have a good shot at lasting 500 to 1000 years if properly stored.

[0] : https://en.wikipedia.org/wiki/M-DISC

[1] : https://purl.pt/24107/1/iPres2013_PDF/Permanent%20Digital%20...


That assumes we'll have DVD drives 1000 years from now. The thing I like about HD-Rosetta is that it doesn't depend on any special technology to be readable. Magnification is the sort of technology that will always exist in any sufficiently-advanced society, while the specifics of DVD encoding could easily be lost to time.


> After a millennium, how would anyone know how to operate it

Same way they know now. Humans will still be humans. If my 3 year old son can figure out how my e-reader works from nothing, so can future humans.


Can he figure out how to build a charger for that e-reader from scratch? Wall sockets are going to be different in a millenium, if they even exist at all. With no manual, they could easily fry it with too much voltage before they actually get it working.


Somebody wrote down their thinking on exactly this subject: http://canonical.org/~kragen/eotf/


The weak link is the children - would you take out and read your great-grandfather's book every year? Now take it further yet and eventually it gets boring and not done anymore.


It's really interesting how Jewish culture has done something like that for thousands (I'm not an expert), with the annual passover family conversation.

Edit: and/or other traditional Jewish yearly family events.


I've considered using mylar paper-tape as a long term digital storage medium.

I thought about either standard sized paper tape, or six foot wide reels of mylar (in any length) which can be punched at a pretty high bit density, and read back optically.

With instructions printed on the outside (and on the first dozen layers), much like the voyager record, explaining how to play it back, and construct a playback device, and how the encoding works.


> keep it in a safe-deposit box

on a practically non-declining 2000km high orbit where USSR Uranium nuclear reactors are parked.


I think we will clean that stuff up. You have to be pretty sure that humanity is going to suffer societal collapse for hundreds of years in order to put something in an orbit that people will want clear.


The physical book have a symbols array limit.

I'm not sure if archived sites will cost less than 10MB by images and unlimitied private photo streams. Because we created the Internet to look forward, not backward.

Looks like 1+ TB for the minimal common case for the human race


I once had an email service which dismantled its servers one day after it stopped flowing email.

I lost a lot of email.

They had sent a couple of warnings, which I had cleverly filtered into a special folder. And missed.


Oh. You assume that in 500 years there will still be fish.


you just may not recognize that as fish.


You are better off with microfiche.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: