Thread by @punk6529

1/ On Preserving A Digital Image for 100 Years

Have you ever thought about this in the non-NFT world?

It is basically an unsolved problem, as far as I can tell.

Let's work through an example:

2/ Let's lay out a scenario to make it real:

a) You are, say, 30 years old and will pass away at 80, halfway through this exercise
b) You have a 1 year old.
c) Your 1 year old will have a kid at 30 and pass away at 80.
d) Your grandkid will be 70 years old in 100 years

3/ What are we trying to preserve?

Let's keep it simple - the few tens of thousands or hundreds of thousands of digital photographs you have taken over the years.

Some family, some artistic

You are not world historical; you are one of us

4/ Let's dispense with the hopeless solutions (and their statistical longevity)

a) On your phone (1-3 years)
b) On your computer (1-5 years)
c) On your computer with an external hard drive backup (3 - 7 years)

5/ A bare minimum system I would consider to be reasonably likely to "not lose your photographs in the next ten years" is probably something like:

a) photographs on your computer plus
b) a redundant RAID array in your house plus
c) backed up to S3 on Amazon Web Services

6/ The setup above gives you 2 to 3 onsite backups, (so you are generally safe from a local drive failure), with an additional several backups at AWS (so you covered from an onsite disaster - fire, power surge, virus on your network)

7/ This setup covers you if you are:
a) alive and technically competent
b) diligent about synchronizing all these things
c) willing to spend $ x,xxx upfront and $ xx to $ xxx / month

It is what I do; it is a nuisance and you don't get much social utility out of it

8/ I consider the above setup to be the minimum serious setup for anyone who is serious about their digital imagery.

Anything else has non-trivial chances of being lost, even while you are in your prime.

9/ Notice that I specifically said AWS (Amazon Web Services).

There was a time when Google Photos offered great almost unlimited backups. I used that too then, but never stopped using AWS S3 because there is no free lunch in life

Inevitably, Google switched business models

10/ There is a second aspect re AWS.
I note that it is more expensive than other solutions (and others may need to go for less expensive options), but if you can afford it, well, the largest cloud provider in the world by a factor of [a lot] is the most likely to have longevity

11/ So you are technically adept, you replace your RAID drives when they fail, you synchronize your phone, your computer, your RAID and your S3 buckets.

(this all sucks, is annoying, feels like enterprise IT work, not fun and social)

12/ This can last, if you are super motivated, for 10-20 years, upgrading and updating your hardware and software along the way, dealing with new syncing and permissions issue as they emerge.

Fine, let's assume this is OK. Now comes the hard part.

13/ What is the hard part?
a) you get old and less technically adept
b) you get old and less adept in general
c) you die
d) you die unexpectedly

What happens next?

14/ Well, you need to assume that your spouse / partner / kid(s) want to now take on the mantle of doing all this PLUS doing it for their family, their spouse and their kids.

If "digital archiving" is a skill set in your family, sure, maybe this happens, but usually not.

15/ And that is just the immediate years after your death.

You are counting on your kids to do this for 20 years and then hand over the baton to their kids (your grandkids) to do it for 30 more years.

So that grandkids and great grandkids can see IRL photos of "Grandpa 6529"

16/ Are there ways around this? Not really.

You can't really prepay AWS, your credit card will bounce after your death, who is going to have your credentials?

Also, will AWS and S3 be around and/or work in the same way in 10, 20, 100 years? (I doubt it).

17/ What if you pay someone to do this for you?

Then you need to think about succession planning for them and/or their business.

Most businesses will NOT survive 100 years

Your local library/university/museum has the exact same problems to deal with; does not want ur stuff too

18/ Some semi-serious, semi-witty people say "print them out; paper has less obsolescence than electronics."

True, but a) unreasonable in era of 10,000s to 100,000s of images, b) paper over 100 year fades and c) easy to lose boxes across moves, generations, floods, fires

19/ It is a tough problem - even though we have the greatest era of digital image capture in history - most of it will be lost.

I am in the 0.1% most diligent on this topic and I am 50:50 if I will manage to hand over the family photos down a couple of generations

20/ It is in this light that I think about the the longevity of NFTs, which are as follows:
a) centralized server (someone has to keep running, poor)
b) IPFS/Arweave (protocol has to exist, someone has to pin)
c) on-chain (ETH has to exist)
d) above + social consensus

21/ My sense is that IPFS / Arweave (or some easier to use future version) plus individual pinning, plus paying others to pin, is going to be broadly equivalent to my current setup soonish

In time, we ought to be able to develop better distributed systems than our current system

22/ Specifically, what I would love to see in distributed storage systems with/without NFTs are:

a) better private/public models
b) more explicit approaches for longevity and redundancy.
c) better display options

So you can tier your approach

23/ Remember that IPFS and AWS/cloud are just words for "other peoples' computers"

AWS at least can give you clarity about which regions you are using, how much redundancy you will have and so on.

24/ I would like to be able to buy the following service:

For these 100 photos, I will prepay [x] for storage and public display for at least 25 years, with at least 5 pinners in each of US, EU, UK, Singapore, Switzerland, Japan, Brazil, Dubai and Iceland (total 45)

25/ Why the various countries?

Geographic redundancy obviously and, also, regulatory redundancy.

You never know who goes crazy and tries to "ban" blockchains one day.

26/ But the pinners should be tiered also.

I would pay 1 price for "Mike Smith to pin with his home computer", another price for "Amazon" and another price for "The British Library" to pin my photographs.

Longevity-as-a-service is a very interesting and unexplored area.

27/ In conclusion:
a) Saving our digital heritage is harder than it looks
b) Off-chain digital longevity sucks
c) On-chain digital longevity has promise
d) lots of work needed to operationalize, improve and consumerize

I am net optimistic, should be doable to improve on present

28/ If you just got here through this thread, we are trying to ensure the metaverse remains open for our children and their childrend

39 threads and 1 game plan here

https://twitter.com/punk6529/status/1448399827054833668