When a dev dies, their apps should live on | Hacker News Reader

When a dev dies, their apps should live on

stuff.tv

108 pointsakakievich57 comments

There are huge issues with current tech and death. Basically. If you die:

1) Your family (might) need your bank info, passwords, insurance policy information, asset lists. They are all doing this, while, their favorite tech guy... is dead.

2) Where are your docs? Kids want these. Journals (evernote), writings (dropbox), stories (jira), just anything. Old iphone photos, strange 3.5 disks, etc. As time goes these are deleted, lost or most of the time: never known about.

3) What about all the things you don't want found? Old contracts, xgirlfriend notes, company transcripts, etc. Welp, I hope you remembered to encrypt your disks, well, that would mess up 1) & 2). Whatever.

> 3) What about all the things you don't want found? Old contracts, xgirlfriend notes, company transcripts, etc. Welp, I hope you remembered to encrypt your disks, well, that would mess up 1) & 2). Whatever.

Smaller encrypted single-purpose filesystems (with seamless integration, like .zip files on Windows Explorer (or Mac Finder) or encrypted folders like EncFS. Then you don't have to choose an all-or-nothing approach.

The problem is that this requires people to be constantly pro-active about where they store things, which most people aren't or don't care about right now. "Well, I can always sort that stuff out tomorrow." or even something as simple as being in the middle of an e-mail conversation and usually waiting to archive/erase that kind of thing when you finish.

I feel the worst part is making sure to have redundancy for these things. I've lost so many pictures and chat-logs through the years from disk-failure. Backing it up somewhere else which is (or is not) encrypted offsite can make this even more bothersome, but it's often neccessary.

Also, encrypted storage adds a risk that if you ever forget your password or lose the key, your data is effectively gone.

Keys in private, non-bank safety deposit boxes paid for annually with cash.

This goes back to what 'sdrothrock wrote:

> The problem is that this requires people to be constantly pro-active about where they store things, which most people aren't or don't care about right now.

Also, the solution requires you to have enough free cash every year. Data usually can last over long periods of unemployment, and there's the temptation to think that "hell, still I have those keys on me, I can live without that safety deposit box for some time, until I have spare cash"...

> non-bank

Why? I would argue that a safety deposit box in a venue entirely owned by a bank would be the safest bet.

More off-grid, less failures. Private safety deposit box companies are excellent.

Banks aren't as discreet as private vault companies and it's not at all unlikely that a government entity would subpoena for your box contents.

Certainly a private vault company could be subpoenaed, as well, but for that to happen a prosecutorial or other undesirable entity would need to actually know you have a box there.

Set the box up, making as-few-as-humanly-possible trips to it, pay for it in cash for years in advance.

One caveat: I would use a bank vault for items like passports, documents, heirlooms, etc. Things that have no potentially-incriminating value or that government would have anyway (passports, DL copies, SS cards)

When you die, your domain names expire soon or later, and your blog and personal sites disappears too. But do you want your blog to live after you passed ? and until when ? one century ? two ?

They live on in the Internet Archive. As long as people care about the Internet Archive, hopefully forever.

Unless whoever gets the domain name after you doesn't want to be on the archive. Then all the history will be deleted.

Indeed. And that totally sucks. Something needs to be done so that Internet Archive doesn't have to retroactively apply robots.txt or deletion requests (sans criminally illegal content).

If you care for something to be preserved, you preserve it and do not rely on someone else to do it for you, forever.

What if all the authors of culturally/historically significant works throughout history had thought like that?

>What if all the authors of culturally/historically significant works throughout history had thought like that?

Nothing much would have change, since it wasnt up to them whether their works surviced, it was up to their "fans".

In fact, if they had thought like that it would have helped, since they would have assisted this process (trying themselves to get their works preserved).

Besides, 99% of the internet is not culturally or historically significant except in the broadest sense (in which everything is, as an artifact of its era, etc..).

Yeah, the Internet Archive hasn't exactly been a stable repository for old stuff, really. Even without death, sites die: this is as true for archive.org as any other site on the web.

3) What about all the things you don't want found? Old contracts, xgirlfriend notes, company transcripts, etc. Welp, I hope you remembered to encrypt your disks, well, that would mess up 1) & 2). Whatever.

Well you're dead so how does that matter? :D

No need to make your family members suffer. People like kink, people might have done things they regretted. What good is it to hide it all your live for your and others' good and then have it out in the open when you are gone? It will only make the living sad/mad.

The same issues pretty much existed before $TECH.

Did they? Save for stuff in safe deposit boxes (I can't remember what happens to that when you die), everything was likely to be in your home somewhere, pretty trivially accessible unless you use a safe or secret compartments or something.

No the sheer disrespect shown by companies for their customers data and the fact that it is a lot easier to loose digital data than anything else have increased the problem.

This could end up being just another one of those dark ages in history where very little can be found since everything was digital and most of it was just deleted in an effort to reuse space or discarded by companies who went bankrupt or got bought out.

I don't buy the "dark ages" thing. Even if 99.99% of the digital data about you disappears, you'll still leave a much, much bigger footprint than even your recent ancestors did. I know a lot about my parents, a few bits and pieces about my grandparents, and nothing further back except a name and maybe a photo.

There are huge issues with death in general, if you are not prepared. Many/Most older people are prepared, with wills, directives, and instructions on file with family.

If you are younger, without many assets, without children, those things are less important. But once you start a family, preparing for its continuity upon your demise is just one of those things you should be doing.

For me, it is pretty simple. Everything is jointly owned, and I have a "In case I get hit by a bus" document, which among other things) gives instructions on how to take control of all my domains and code, which bills are on my card. And it tells my wife to make her own decisions. I will be dead. I will not give a crap if my blog stays online. She might, so it is up to her.

In any case, I update that document once a year, or before I go under general anasthesia.

death in general, if you are not prepared

Sadly, most people aren't prepared for life either. There are so many people without proper health insurance, no backup plans for anything (career, immigration, children etc). So much of people's time and energy and money goes into this (they still don't get it right or even finish it) that "when I die" stuff takes a backseat.

Coincidently, i've been thinking about creating a pet web project to solve #1 utilising PGP.

The general idea is that you (the user, manually or through a client) locally PGP encrypt the text with the public PGP keys of the friends or family whom you'd like to receive the text by e-mail. The encrypted text is then sent/uploaded to the webapp. If you don't report back after X amount of time, the web service sends the text by e-mail to the selected friends or family. The webapp would not have any private keys therefore no access to the unencrypted text.

Not sure if there's already a similar service.

One of the problems I could see here is that PGP is all but guaranteed to be irrelevant in 40+ years when most of the people who'd use such a service are currently expected to die. It'd certainly be helpful in the short term, though, which arguably has more merit.

True. Another problem is that people lose PGP keys and they also expire but i guess it's hard to find a one size fits all solution. In 40+ years, the secret info might not even be relevant anymore.

Encrypting your disks will only keep people from reading them for a century or two. Technology from the future will be able to break today's encryption in a heartbeat, whether it's simple Moore's Law, quantum computing, cryptographic loopholes in today's algorithms, or something else entirely. So if historians from the future care about you, nothing you leave behind will be truly safe.

If you really want to mess with people from the future, do what Beale did, make up a bunch of bogus stuff, and throw it in with all your encrypted documents.

> Technology from the future will be able to break today's encryption in a heartbeat

I don't believe that's necessarily true. IIRC, certain algorithms would require computations involving every particle in the universe, operating for billions of years, to brute-force, provided that they are not actually broken. Will every current algorithm eventually be broken? Probably—but if not, they should be good.

Many of our current encryption algorithms can be broken by sufficiently powerful quantum computers [1]. We're still a while off from usable quantum computers, but they are coming, eventually.

[1] https://en.wikipedia.org/wiki/Post-quantum_cryptography

Symmetric key encryption is resistant to quantum computing, you just need to double the key size

Unless something drastically changes, this does not hold true for symmatric encryption.

You're overestimating the significance of material left behind after death. What'll happen is children or anyone left behind will harvest the photos folder of your laptop and then throw the computer to trash. It's extremely unlikely that they'll find the time to go through your digital documents otherwise, let alone figure out passwords or continue your life work. Life goes on, assets will be sold, money distributed, and so on.

I think that this is a good application for secret-sharing (a way to take a secret and split it into n pieces, of which k are required to reconstruct it).

One wouldn't necessarily want to split a secret amonst one's family (because families can feud), or even put the shares into the hands of people all belonging to one legal jurisdiction (because the odds of even a good government going evil within one's lifetime are uncomfortably high). But one could, say, pick three family members, three friends and three colleagues, distributed over the world, and then split the secret amongst the subgroups {family 1, friend 1, colleague 1}, {family 2, friend 2, colleague 2} & {family 3, friend 3, colleague 3} such that two members of each subgroup must agree to recreate the subgroup secret, and two subgroups must agree to share the actual secret. Assuming a reasonable amount of independence between the individuals, this should be okay.

In practice, this would probably have to involve sending each person an annual notice with his share, as well as detailed notes on how to recreate the appropriate secrets and information on the others who have received a share. And of course one would need to think hard about how difficult it would be for a second or third party to suborn the system (in the example above, all that's needed is for two family members and two friends in the same subgroups to live in the same oppressive legal jurisdiction: one subpoena later it's all over).

If I am not sharing something alive, I have no wish for that to be shared after I am dead.

> people often say similar things when it comes to art and literature, and even film and music. But those mediums have the kind of longevity that just isn’t afforded to modern digital apps.

This line stood out to me the most. Along with payment issues, one of the larger issues is bitrot. What if Windows phones gets a 99% market share next year? That's 99% of users who will never be able to play Hogarth's game because it was never ported. Even if we put aside proactive adaption, there's also an issue of the code stagnating in relation to its environment. Entire architectures might change such as Apple's move from PowerPC to Intel, or requirements might get tighter like sandboxing in the app store or the recent SPI mechanism that protects system files. Sure you might have emulators but that mostly relegates the software for later generations with more powerful hardware; to be played with nostalgia rather than in the moment.

I don't know what the solution is, but we definitely need to do something to preserve these kinds of things for the future, like how the Internet Archive is breathing new life into MS-DOS games using in-browser emulation. [0] For one thing we need to preserve the binaries of games like these in case they do go down, and related to that, the specific software versions they ran on. (I'm contradicting myself here because these are the only non-radical approaches I see right now as opposed to something like a reserve for source code to be released when it enters the public domain).

[0]: https://archive.org/details/softwarelibrary_msdos_games

We already had a lot of architectural transitions throughout the computer history (a very small part of which I witnessed) and we could emulate nearly all of the deprecated architectures quite well. I am optimistic that it can be repeated.

Not when you have companies forcing online registration, expiring software, cloud services/storage, DRM, ...

Yes, that is also a concern. One can only hope that when they turn into abandonware, those software will be "freed" ("cracked" but legally, although IANAL) by the community.

There's no legal concept of abandonware, and even redistributing it is illegal. It's just that, well, it's abandoned because nobody wants to take legal responsibility right now. That situation can still change when a copyright holder sells it to a patent/copyright troll later.

Thanks for the clarification. I falsely assumed abandonware as public domain. I also found some explanation about the public domain here: http://programmers.stackexchange.com/a/72916

There's also the small problem that not all countries have a concept of a copyright in the first place. Quite a few EU nations have laws around the concept of Authors' Rights instead. Public domain does not even exist for those (because there is always an author, and authors always have rights to their work, even if they're publishing anonymously).

At least here in Portugal, while we do have Author's Rights (which include moral rights) instead of copyright, they still expire to Public Domain, and I believe the same is true for most EU nations: https://en.wikipedia.org/wiki/List_of_countries'_copyright_l...

Authors' Rights expire, but (at least in Germany and Austria) you cannot voluntarily put a work into public domain before and give up your rights on it, like you can under the US concept of Public Domain.

As far as I know all EU countries are signatories to the Berne convention on copyright, and so all of them have the concept of copyright. It is just that copyright is not the only intellectual property rights that apply, as you pointed out.

Authors rights (or "moral rights") are not a big problem. While you usually can't sign those rights away, they don't generally prevent the author from signing away distribution in perpetuity, so while you can't get "total" public domain in those cases, you can usually sign away most of the rights people are most interested in.

not if SAAS becomes the standard way of doing things, you can't emulate what you don't have, if all we're going to end up using are thin clients to cloud-based software once it's gone it's gone forever

On a more wider subject, I always wonder what happens to the developers just simply disappeared.

Phone apps, browser add-ons, you may only notice something happens when it stopped working.

Then you go to the forum, you go to the blog, there may be some rumors. There would be some people trying to pick up and carry on. But, sometimes, no one knows what really happens.

No emails has been responded from that weird email address ever since, no activities from that account.

Finally, you may go to tweak that apps you loved so much yourself. Or when you are lazy, you will try to find an alternative and accept it. But the question is still hanging in your head.

If you want your stuff to live on, there is an easy fix. You can set up your will so some or all of your stuff will be uploaded to a public place, such as GitHub or whatever. Or you can ask a friend to go through your digital possessions and use their own judgement regarding what to do with it.

Simply doing nothing and expecting app stores to come up with something like memorialized accounts is not going to work. The "best" they can do is delete your apps faster when notice of your death comes in.

If you're writing a will anyway, why not just include a clause about the data on your hard drives? Arguably, including a "data will" is probably more important to your legacy than the boring minutiae of how to split up your bank account.

This becomes so much more apparent when you work for an app startup (or downloadable console game publisher) that shuts down. The games I put hundreds of hours into while working for those companies have disappeared from their corresponding downloadable stores after only two years.

I still have copies of them on my devices and on my computer, and there's still a few copies floating around in torrent-land, and there's still a few videos and a couple reviews out there, but for the most part it's already difficult to see any evidence that those games ever existed.

In contrast, the Flash games I made on my own you can still access from about 2000 websites, including my own and Newgrounds, over a full decade later. Although even that is going to be a problem soon now that Flash is on its way out.

How the hell can a game developer keep up with this? I decided I can't, and coincidentally I got into board game design anyway, so now I spend more of my time working on board game designs. If those get published, then there will be many hard copies out there, and if they're popular enough, either myself or another company will develop a video game version of them, so they can exist in both mediums.

And you don't need to keep up with the latest tech to make sure your board games survive for decades, even hundreds of years. For me, this satisfies my preservation itch a lot more than the current state of digital preservation, which seems in terrible condition for modern games.

Hopefully the retro game emulation scene can keep up with the demands of it all. I'm thankful they allow me to play SNES games that would cost me $100+ to get a physical copy today, a physical copy that is not guaranteed to work on my machine (I bought a few SNES cartridges that didn't work recently).

I remember being so sad when Dan died: http://web.archive.org/web/20050814083139/http://www.danb.di...

He wasn't a real world friend, but we'd spoken via email a few times and I was in love with the game he'd written. He had really big plans for it, but sadly it died with him, and doesn't seem to be playable now without digging out an old browser and versions of java.

This was the days before github - I can only imagine that today this project and his others could have lived on in some way.

I would love to see a similar proposition with proprietary products run by companies.

If you run an infrastructure style product, there should be a living will involve where the product will become open source upon the bankruptcy or shutdown style sale of the company.

For example, FoundationDB was bought by apple and the product was shuttered. Under a living will, the product would become open source, as it is no longer developed by the company.

I realise this may lead to a lower selling price, all things being equal, however people/companies are also more likely to trust you when you are small, as they have a greater chance of survival should you fail.

My personal feeling is that most of these products should be open source from the get go, and it will probably have to be this way going forward, however this is a reasonable middle ground.

"For example, FoundationDB was bought by apple and the product was shuttered. Under a living will, the product would become open source, as it is no longer developed by the company."

Would it? I assume the tech is still being developed. It's just that the continuing development is not offered as a product.

Further, once someone buys the company, don't they gain control over that company's assets? So if they no longer want to offer a product for sale, why should they?

In theory there should be a trade off when offering such a solution. Companies are more likely to trust you, and use your product if they can be sure there is a fallback plan (open source) upon your product being shuttered. Yes, you may lose some potential fallback plans for your company if it is being bought for its technology, but hopefully the number of additional adopters who would trust you early on in your company's life would compensate for this.

I'm not saying there is a moral obligation to do so, merely that it might be a good decision for both sides. See, for example, how open source is increasingly becoming a requirement for infrastructure products that companies come to depend on. This may be a good middle ground. Personally think open sourcing from the start is a better choice.

The idea expressed in the title is a kind of hell. Think how hard it would be for all of us if we were obligated to maintain every app past the lifetime of its creators.

Software dies. Maybe there will be some OS code that runs for a hundred years. But I think every application is going to die that doesn't have some cult like group of fans who maintain it, probably mostly for sentimental reasons.

Perhaps it might be appropriate to link to these:

Get Your Shit Together[0], and related discussion[1]

[0] http://getyourshittogether.org/ [1] https://news.ycombinator.com/item?id=9569934

well the article identified but failed to address both of the core issues: copyright and who pays for it.

but most importantly failed to mention or aknowledge this https://archive.org/details/internetarcade which is doing exactly what he vouches for, preserving memory (especially relevant since he's talking about a game)

Reply on news.ycombinator.com