Hacker News Reader

Comment by 1970-01-01 | Hacker News Reader

1970-01-01

So why is it a bad idea? There's a difference between destroying information and destroying the original copy of it.

For the reasons already articulated in the article we're discussing; It's a waste of people's time to reiterate them.

I would add though that one of the top stories yesterday was about how the British Library website has been down for 10 days due to a ransomware attack. Digital information retrieval is only as reliable as your backup strategy.

Let's not forget that even that is prone to problems. I had some archives (30-35 years old) that no longer can be unpacked. The application that made them is no longer working. Source code isn't available. I also had a few "rotting" CDs (not CD-R, but factory pressed CDs). You need a very good strategy to make digital information survive the next 100 years, and keep it up to date, at least yearly.

And, given the performance of government IT projects in the past in so many countries, this will probably turn out to be more expensive than they hope.

I had some Bitcoin on a USB along with the operating system and program used to read the wallet. After 8 years or so, I couldn't open my wallet despite using the same OS and same program but different hardware. (I did eventually recover the coin but it was much harder than I anticipated. and no, the USB wasn't my only copy, I wasn't quite that foolish)

So the only legitimate reason is "Hardware goes out of date" That is a real but solvable problem. Everything else was pure speculation.

nanofiche is probably the best middle ground.

paper itself isn't durable enough but I'd still be wary of purely digital encapsulation.

I disagree with you that “Hardware goes out of date” is, in fact, a solvable problem. The truth is, macroscopic writing, in the sense of scratching or applying pigment / dye to a surface, is a very old technology, and we know very well what the revealed points of weakness are, and in many cases, how to avoid these points of weakness. Perhaps most importantly, a minimally-trained individual can recognize the content and potentially the importance of a physical document. This is not the case with modern technologies that require a separate apparatus to be decoded and displayed. In many cases, this apparatus is not available, and it is possible that the layperson will not recognize the reel, cartridge, cassette, slide, disc, drive, wax cylinder, etc, for what it is and simply discard it. This is by far the greatest risk for data preservation.

If we reach a point where people working in the archive don’t know what a computer is, we’re at the point where we don’t need to know what Dave’s will had in it.

Think about what you are protecting here. Dave’s will. Ol’ davo from down pub. Literally meaningless beyond the potential for mild curiosity in a hundred years.

If the worst happened and we lose them, it’s insignificant. Planning for the preservation of davo’s will in the event of an apocalypse is a waste of our time.

What if that is not Dave’s will, but Dave’s first novel? Or the technical manual to Dave’s computer system, where the onboard help files are just a URL that points to a dead server? Knowledge preservation in the face of disaster is not a pointless effort.

Further, there is a difference between “knowing what a computer is” and “recognizing a specific manifestation of computing technology”. For example, this picture [0]. The URL is somewhat descriptive, but say you were presented with that piece of technology. Would you recognize it? Would you think that it could be important? Would you be able to get it running? For me, personally, it would be a significant effort in terms of time spent… and I certainly wouldn’t be able to recognize it just from that picture.

0. https://upload.wikimedia.org/wikipedia/commons/7/72/ZIP_Driv...

> "Hardware goes out of date" [is a] solvable problem

Yes, theoretically, it should be practical /s

Destroying the original copy would be destroying information. A picture of an artifact is not the same as the artifact itself.

True, however these are legal documents and not art for a museum.

Even ignoring the historical value, it's easier to hack into a server and edit a file than it is to forge a legal document and break into an archive. Suddenly, you're the legitimate heir to a castle.

If the point is continued legal meaning then it would be better to digitally publish in any way that allows distributed verification of the record as what was published and not challenged.

Probably thousands of people have some means to tamper with that archive, but courts don't really care to revisit hundred year old wills in the ways soap operas imply.

Very debatable, if both are done right.

In practice these are more of historical than legal interest.

That must be why that Bitcoin thing didn't pan out.

Huh? It's very strong at the moment

This argument makes it seem like forged documents were never a thing.

It’s more that the risks and expenditure of effort are so different between changing files on a server vs breaking and entering that it becomes a separate category.

True but scanning them makes it a lot harder to determine if the original was a forgery.

True. You're at the mercy of whatever information was captured at the time using means available at the time. One reason we (hopefully) don't get rid of physical evidence from criminal cases.

Should we then destroy all legal documents, for example The Magna Carta?

Museums aren't just a store of "art", museums are a way to preserve our heritage, preferably in its original form.

That makes the originals even more important.

Not art, historical artifacts. There’s tons of info that a picture wouldn’t capture. Type of paper and ink used, isotopic analysis, etc. Who knows what questions future historians or archeologists will want to ask?

But you could say that about all trash too.

Just hope they aren't using Xerox to digitize those documents. That glyph substitution bug from a few years ago made quite a few would be digital archivists rather unhappy.

Speculating here, but carbon dating is potentially one useful vector for protecting against forgery/manipulation. Conversely, once something is digitized, you cannot incontrovertibly authenticate the origin time of the relevant bit flips and you also introduce the (now seemingly just on the horizon) possibility of AI-optimized digital forgeries that are indistinguishable from originals. That's off the top of my head, but the OP may have further insight.

If only there was a technology for storing unforgeable timestamps in a way that didn't depend on the honesty and competence of any one person or organization

It would have to have some intrinsic mechanism to encourage people to store and interact with the system though..maybe people could prove they worked on it and then trade it for other things of value?

You can trivially implement this using a timestamp authority (TSA). This can be done without the blockchain, and there are several free publicly available services that do this.

https://en.wikipedia.org/wiki/Trusted_timestamping

Seems like there's no issue with digitized wills then

You know what's easy to read 150, 500, and even 1000 years later? Physical writing on paper. No matter what changes, these documents can and will continue to be legible as long as they're properly stored.

The same can be said for microfiche.

Destroying the original destroys some information that could have legal relevance.

Things could be written on the back, or the type of paper could be an important point.

It will likely inadvertently make them susceptible to remote cyber attacks.

How much do you trust the government to keep these digital wills secure? I'm fairly confident they can keep physical wills secure. A country's army provides a good deterrent to vandalizing physical objects. In the digital world, those threats are muted, especially if the hacker is in a jurisdiction that doesn't care.

The seminal answer: an archivist explains. https://www.youtube.com/watch?v=yKNbhMCB_3g

Lots of reasons but I'll list two: 1) These documents are historical artifacts like the declaration of independence or the magna carta. They may not seem important now but in 300 years people will care about them A LOT. 2) Redundancy. If the digital records are destroyed or say encrypted in a massive ransomware attack then the UK government can simply reconstruct those digital records using the paper records without having to deal with terrorists and that's a good thing.

Ask a historian! There are several quoted in the linked article.

The digitised version isn’t a faithful copy; see the mention of the indent of the pen in the article. Information is destroyed in doing this.

Reply on news.ycombinator.com