Self Encryption + IPFS

I had a dream couple nights ago, and lightbulb went on.

What if we have a personal device, a pool that stores all of the personal chucks. The chucks cannot be open unless the owner opens it. The datamap is not stored on the harddrive. It is stored on ipfs or usb, which is effectively useless for everybody. The datamap encrypted password is broken up to three parts. One part goes to a friend, another part to another friend, and one for himself. Those friends will save the datamap on ipfs so the owner can retrieve it. He cannot open it until he gets the password from both of the friends. The owner does not know the password nor their friends. It should somehow insert the password seeing the password. So if crackers or gov want to access the laptop, they are free to do so but they will never find the information in there without the datamap. The laptop is useless to them. And the owner cannot retrieve it.

Bombproof computer. Thoughts?

So essentially a multisig wallet for your data?

I am sure this will be one of the first apps out on SAFE for sure!

It could help a lot in defeating things like key-loggers and client machine vulnerabilities.

Don’t be like EddyJohny, who’s been ranting about 2FA for years.
If you want an elegant solution, make a copy of your password, encrypt it, put it in a sealed envelope and deposit it with your lawyer. No pay, no reliable service.

If you think your “friends” will take proper care of your password chunk (while not knowing what they’ve been given), you’re wrong. You’re also wrong if you think it’s safe to give chunks to your friends and tell them what you’ve just given them.


I am pretty certain that IPFS can accomplish anything that SAFE does in more of a modular piecemeal level.

Of course you have to trust the piecemeal system more that you would trust the “whole enchilada” system of SAFE – but what can be done in IPFS can be prototyped out tomorrow, not over several years…

Probably only by recapitulating the design of SAFE in IPFS, which is like an engine that won’t run unless all the core parts are in place. Without SAFE’s XOR space being simulated on IPFS, what exactly would you be prototyping?

You probably need more than two friends, because what can you do if one goes missing? To be honest we shoulden’t be thinking about passwords anymore, especially not with software like SQRL around. Biometric is a solution, but it needs to live on the SAFE Network in the users hands.

It would be fun to see if you could have layers of biometrics before you can access your data, even more fun it would be if you use the wrong finger to access data and it would be a bogus data. Ofcourse the hardware done to do this, should also be opensource…

Sometimes I wonder why 2FA ( yeah Janitor here I go again) is not used to install software on an OS, probably because our OS has so many backdoors.

Like your lawyer can’t be forced to give that envelope, after that they’ll force you to decrypt your password.

Whereas you and your friends are ex Navvy SEALS who won’t break under torture.


1 Like

The point here is store local chucks, instead of distribute across the net. Just don’t save the datamap, and thereby making all chucks useless. Download the datamap from ipfs, or store on usb to open up the chucks.

Wouldn’t truecrypt be easier? and it can be made to appear like free space with random data rather than content stored there.

1 Like

As EJ told you you have 4 chunks of info, and if any is lost or corrupt your “backup” becomes completely worthless. Say one of your friends get hit by a bus.

Move in, this plan sucks.

Truecrypt is corrupted - designed by RSA people who collaborated with NSA. But yes, you got a point.

@janitor At least it is an idea.

GRC still has the uncorrupted version for free download.

1 Like
  • That about TrueCrypt was never proven
  • A French company reviewedd the code and released a newer version, however I didn’t see that they found anything
  • My suggestion above was to use encryption that’s considered safe (such as PGP with appropriate options)

The idea is too complex. I told you it is crazy to give a piece of data to someone, don’t tell them what it is, and expect they’ll hang on to it like a lawyer would. (You also expose yourself to accidental data loss and blackmail and so on and so forth). Move on.

SAFE is overkill for 95 percent of everything. If you have something criminal certainly the 'randomness of storage" is a feature that is vital…

Right now I can self encrypt my files and store them on IPFS nodes on various machines… If my only concern is hacker resistance, not warrant resistance, I could get a mighty secure system with IPFS… The scalability of “as complicated or as simple as fits your needs” isn’t necessarily a bad thing.

Safe will be nicer in that you don’t have to think about it most of the time. But SAFE isn’t working yet and IPFS is – Probably because they didn’t try to eat the whole elephant in one bite…