Observations on explaining "Serverless"

As the SAFE Network gets closer and closer to an initial launch, I still find it difficult conveying the structure to people.

Getting people to grasp the concept that there is no server is a bit frustrating and a bit entertaining at the same time.

The first question people generally ask me is “Well, how do I deploy a server for my data on a peer to peer network?”

I then explain (again) the concept of the data being distributed and held on drives all over the world and I usually get a response like “Well, can I set up my server on a single node?” or “Well, why would you want to do that when you can just use a server…”

It is at this point that I proceed to give myself a hearty facepalm and reflect on the ingrained model of client/server being the de facto standard…and also at just how visionary and vital the SAFE structure is. (or perhaps how poorly I am explaining this new structure myself! :grin:)

I look forward to the launch when people will be able to see it in action and (hopefully) have that “AH-HA!..Oh I get it now!” moment.

Until then, keep up the good work everyone!!!


Tell them that everyone who participates in the SAFE network becomes a small piece of what will be a world wide distributed server. When they choose to deploy their site, SAFE distributes the responsibility of storing and serving the site to a bunch of random people.

There is no one point of failure because the site or their personal data is split into multiple pieces and spread over the network. To avoid data loss every piece will have at the very least 4 copies.

If they go on to ask how will they be able to manage their site if it is distributed throughout the network, inform them that when they log onto SAFE, it will be like working on their site remotely. The difference being that they wont be working on one or a group machines they have to ultimately maintain. Instead they will be working with data that is routed and stored by the network itself.

In a sense their site or data is just on a massive hard drive that everyone has a piece of. Anyone who doesn’t have their credentials will not be able to read or manipulate the site owners data. Even if they delete everything off of their personal hard drive, there will still be many copies of the site or data stored on the network. Simple…I hope:relaxed:

EDIT: Remember to inform them that if their site becomes popular, the network will automatically increase the number of copies of the site throughout the network temporarily. This compensates for increased demand by creating a large swarm of people serving the popular file. So instead of slowing down like traditional servers, file access/downloads gets faster.

Ask them, “Wouldn’t it be cool if you could place the responsibility of managing your server in the hands of a bunch of people who cannot alter or delete your data? Wouldn’t it be cheaper if you didn’t have to worry about server costs, relative electrical costs, bandwidth costs, and housing a server cluster? Wouldn’t you like the piece of mind that comes from never having to worry about your site going down? If yes then a little reading is a small price to pay.”



Asked and answered. Fabulous!


Great points! Fantastic run down… I usually liken the network to a giant shared drive/hard drive, but still find that developers minds don’t quite grasp the whys and hows…lol.

One thing that I have noticed while working on apps, sites, etc. for the network is that the current protocols are all geared toward the client/server model, for example dealing with local files/cross domains with HTML, etc. This can make it tough to use the existing model for sites/apps on SAFE… cant wait til the API gets solidified and the crew comes out with some dummy proof guides and docs! :slightly_smiling:

1 Like