The SAFE network: Giving a decentralised internet back to the people

Posted 10 April 2016

I have nothing to do with this article, just found it online today.


Good article - up to date, good background, describes project evolution, current status, mentions the community & collaboration. Even mentions a project Decorum (near the end @Seneca :slight_smile:).


Great catch!,

Really good article. Cheers, Harmen! You’re in the international spotlight!

1 Like

Nit picking but…

The platform addresses privacy by implementing three distinct data security stages. The first is an algorithm encrypting the data before it reaches the network. Keying material is produced using the data within each file, providing the user with secure access. This encryption is applied by the client, prior to data chunking and dispersal.

The second stage is implemented by the MaidSafe protocol. The already encrypted files are chunked and then further secured with the Encrypt Library. This is not an encryption algorithm per se, but an algorithm that utilises existing security algorithms, predominantly AES 256.

Thirdly, the protocol distributes the chunks across the decentralized network, choosing the geographic storage locations randomly. This is performed by an algorithm which ensures a minimum physical distance between each chunk.

The file is not encrypted before chunking as the article alludes to. The self-encryption handles both chunking and encryption in a single step on the client and then the pieces are distributed out.

Other than that, it’s one of the more accurate articles I’ve read. And they didn’t call it a blockchain! :thumbsup:


From the article[quote]
This is performed by an algorithm which ensures a minimum physical distance between each chunk.

This is incorrect, right? The chunk is distributed according to its hash and then as close as possible to its corresponding address in the mathematical XOR space. So the physical machine can be anywhere on the planet. Correct? I´ve had a hard time to grasp this and I don´t think it is very well explained in the Wiki. I would like a more detailed step-by-step explanation.


Fundamentally yes but I think they’re referring to the caching algorithm which will help to serve those chunks quicker (but always at least 1 hop in between to preserve anonymity). Caching is also important to help mitigate DDoS since the more a piece of content is requested, the closer to the requester a copy is stored and the less the attack is felt by the greater network.

1 Like

I thought they were trying to describe how the chunks were distributed on the network, not the caching mechanism specifically.

If they were… then perhaps they said minimum when they meant maximum? On rereading it does seem like a mistake.

Do you have a contact with them? If so they should be open to a discussion of the points you raised. If not, considering their userbase, it may be worth it to contact them and establish a working relationship with them.

Re: the wiki. I agree. I am beginning the long arduous journey to correct this.

1 Like

@nicklambert is on leave this week but maybe he can double check that part of the article and see if we’re misinterpreting something on his return next week. I believe he was in contact with the journalist.


Yes I’m in touch with the journalist who used a combination of research and some Q & A with me. I’ll ping her and ask for a couple of amends. Thanks.