> This method works because your mum is never going to go into conspiracy with your business partner to gain access to your information.

Unless your mom is Lucille Bluth.

I think a better way would be to have one person click 'Mark is dead' and it immediately sends an email to Mark saying 'Joe says you're dead.' If Mark doesnt' access the account within x days, then it sends further confirmation emails to Mom and business partner who also have to confirm that Mark is dead. If so, then it set 'Mark = dead' and opens the vaults as Mark configured.

What if your business partner dies? I thought everyone needs to confirm you are dead? What if they are dead?

You would have to work on the packaging for the "Mark is dead" button and the "Is Mark dead?" email you'd be sending out.

Aside from that, this seems like a really elegant solution. Occam's razor.

I had this idea too. I was going to build it until I looked for a good domain. In the process I ran into: http://www.ifidie.org/ -- which had already implemented it well enough.

Note the idea is applicable to far more than "relatives when I die".. e.g. root password for the server when you're sick in the hospital. Also, remember that many people don't want their estate documents to be viewed pre-death (e.g. don't want people to know how much they're getting), and letting people know where you put them / how to contact the lawyer in a reliable way is a hassle in and of itself.

Have a lawyer and a will.

That or just have a service setup to monitor a Twitter feed for updates. So in the case of Robert Scoble, more than 2 hours and he's probably dead!

But why would I pay $500 once to get something done right when I could pay $10 every month to get something that will almost certainly fold before I do?

(Although that does leave me with one practical issue: if I go to my trustworthy small town lawyer tomorrow and say "Hiya Taro [he's not Bob -- Bobs are quite rare in these parts], in the event that I pass away suddenly, I'd like you to take care of my final affairs as I'm about to instruct you:", and Taro gets hit by a bus, what happens to my plan? Should I be asking my lawyers "But before I give you my business can I ask you a little more about your business continuity contingency planning? How about we start with your bus number?")

Indeed. Supposing you are 30, what are the odds that the company will outlast you? Very, very poor, IMHO.

True, email polling to determine live-or-dead status is pretty dumb :)

A company exists that offers after-death services using a multiple-redundancy scheme like the one described:

https://www.mywonderfullife.com

They call the trustworthy people "angels" :)

Ideally, they would also alert you that you've been "marked as dead" and give you a week or so to claim that you're still around.

They do exactly this. There are several layers of checks because getting it wrong is kind of a deal-breaker.

A lot of things happen after confirming someone is really truly dead, and many of those things can't be easily undone :)

(Full disclosure: I work for the company that developed the site)

The biggest problem with this is what happens if one of those people dies - or simply gets a new email and doesn't tell you - before you die?

Your data is forever locked because of one person. It's even worse than you not telling the service that you changed your email. At least you should have known that the service existed.

Burdening your friends and relatives doesn't strike me as the best way t handle this issue.

The solution - if one person does not reply by clicking a button 'he is not dead', this persons opinion ceases to matter.

wha? That pretty much defeats the purpose of having that person on the list in the first place.

What if nobody responds?

1 person has responded - the person who clicked the button. If nobody else agrees, then I would say the data stays locked. Otherwise the one person could simply wait for a moment where everyone traveled and had no email access.

Which basically means to say that the entire security of the site is based on the ability of my designates to have my best interests at heart... and also not to have installed a spam filter, moved, changed email or otherwise become unreachable to a service that pretty much becomes forgettable after 6 months or so.

I'd much rather trust myself to manage things, to be honest.

I knew a few people who had a "if you read this, I am dead" cronjob setup on their machine: if they wouldn't log in by a certain time, an email would be sent saying that they're dead. Of courses there were cases where people would simply forget about these cronjobs and send email needlessly worrying the people they cared about (and possibly getting themselves 5150'd).

What about "If you are reading this, there may be a serious problem. I usually log into my system everyday and since that hasn't happened, please call or try to find me to make sure I'm OK"

That way it's only slightly alarming to your loved ones.

that's different, but do you really need to worry the people you care about like this? bad news travels fast and if you do need to call for help there are ways slightly more person than a cronjob (though this sort of a cronjob message won't get you 5150'd, which generally makes whatever problems one has worse).

What's 5150? Is it this: http://en.wikipedia.org/wiki/Code_5150 ?

Yeah. If there is a reasonable suspicion that someone has attempted/wants to attempt to take their life, the police will take you in and do a 72-hour hold.

I find it more interesting to note that not only has this sparked a conversation on HN... it has sparked a conversation about the security/efficiency/problems of the proposed solution.

I guess a lot of us are analytically minded here.

My question is: did you (and if yes, what?) do something to release your digital assets to your family?

I'd be pretty freaked out if I was 60-something and I got an email like that from one of my children, I think.

Very interesting idea. I think it would make sense to send these notifications to the party's estate lawyer, rather than loved ones, just to reduce the "morbid" factor.

Sounds like if one email ends up in a spam folder, the whole thing fails to work.