Hacker News new | past | comments | ask | show | jobs | submit login

I wrote a book in 2010. It had a references section with links to about 100 websites. When I wrote the second edition only about five years later, 50% of those links no longer worked.

What we're doing right now is borderline insane. We're putting all of this information on the web, but almost each individual bit of information is dependent on either a company or a human being keeping it online. It's inevitable that companies change their minds, and humans die, so almost all of the information that is online right now will just disappear in the next 80 years.

And we essentially only have one single entity that tries to retain that information.




> references section with links to about 100 websites.

Books deserve a github repo with PDF web archives of referenced links, the same way that Wikipedia mirrors the content of cited links.


But wouldn't that be a big waste if everyone who references the same thing is then keeping a copy of it.


> big waste

Storage cost has fallen exponentially for decades, https://ourworldindata.org/data-insights/the-price-of-comput...


Redundancy isn't really a waste.


Better many copies than none. References usually mean written text and maybe some figures, cost of storage is going down, we can afford the duplication.


> everyone who references the same thing is then keeping a copy of it.

... and it would serve as a form of redundancy, imitating the fungible nature of physical media: In order to cite the latest, copied, manuscript (for example) you needed to own a physical copy. They existence of these has enabled survival of works that would otherwise have been lost, or even reconstruction through ecdotics.-


One person’s waste is another person’s resilience.


There are all kinds of publisher and legal issues. Trust me, I did my best.


Could Wikipedia or Archive.org offer references-as-a-service to book publishers for a small fee? They already have the infrastructure and legal cover.


certainly not GitHub


What would you recommend instead?


Copyright permitting, big QR code containing the plain text.


That entity goes out of its way to hide information if you're friends with the owners.


All the more reason not to rely on it, hard to complain when no one else is willing to do what they do


> wrote a book in 2010. It had a references section

Out of curiosity, may I ask what it was about?


> so almost all of the information that is online right now will just disappear in the next 80 years.

> And we essentially only have one single entity that tries to retain that information.

Will future ages find ours a dark age, a gap in their records, a void ...

... up until the point - if ever - where a sufficiently advanced solution for permanence is found and comes online?


> ... up until the point - if ever - where a sufficiently advanced solution for permanence is found and comes online?

Like the laser printer?

The cost of permanent, physical preservation is pennies. People just don't do it for most things. And it doesn't guarantee accessibility, which has hosting costs.


> Like the laser printer?

Sure. Whatever works.-

But I meant one that is systematically and systemically and widely used.-


> Will future ages find ours a dark age, a gap in their records, a void

I think this is a very likely future, yes.


Grim, indeed.-




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: