User-friendly free storage service


#1

Storage is cheap, and text doesn’t take a lot of it. If I provided only 100GB of my 30€/month dedicated server’s 1TB storage space to the network, I could store one hundred billion characters of text.
(In reality, it’s of course going to be less because of multibyte encoding, metadata, encryption and the like, but the ballpark remains the same).

If I do not use the full 100GB, I get Safecoins, which I can also use to store data.

Meanwhile, for the average person, setting up a SAFE node just to post text messages to a message board hosted on SAFE is going to be a hassle. Also, they would have to dedicate system resources, no matter how little, to the network, which some may not be able to do, or buy Safecoins from somewhere else, which is also at least inconvenient - and it would be difficult to purchase an amount so small that it’s only useful for text.

So why not put them and me together?
What I’m thinking of is either a user-friendly Safecoin faucet (software) that pays out small amounts of Safecoin, but still enough for the average person’s message posting habits, in a manner that does not require fiddling around with addresses and other technical stuff (perhaps through integration with client apps through an API), or preferably a service (software) that allows people to post small amounts of data to the network without the user having to run a full node themselves, also interfaced with SAFE-based applications.

I’m aware the latter option could be a privacy risk for unencrypted content, though that could be mitigated by providing such a service as a Tor hidden service.

Both options could be protected from abuse by a proof-of-work scheme similar to Hashcash. Hashcash’s idea was to add proof-of-work to email, requiring people to spend CPU time for sending a message, making spam unprofitable because the average amount of money spent on an email would exceed the average amount of profit made with it, but not inconveniencing human users because of how small the amount actually is. My idea is to require proof of work to make running an actual SAFE node and either spending hardware resources or money cheaper and easier for larger amounts of data.

Any thoughts on this?


#2

No thoughts on this.


#3

SAFE will be used by privacy conscious people and there is no way that any marketing strategy will stop people from using GOOGLE and other extremely useful and convenient services. The reality is that we are in the minority and most people don’t care about privacy. Lets just cater our efforts to the privacy minded and make it as easy to use as possible, then maybe SAFE can become the linux of operating systems if you know what i mean.

But if the safe network allows people to host large scale websites cheaply without needing to worry about server memory/hdd space/bandwidth then maybe SAFE will take over the mainstream client/server web hosting model, people will be using SAFE without even knowing it, that i think is possible.

Maybe firefox or chrome will integrate safe support straight out of the box.

What if you could earn 1 dollar a day just by setting up a safe node? Wouldn’t everyone do it? Then maybe we could see things going mainstream.

jumbled thoughts but watever.


#4

The use case I have in my mind is an uncensorable message board. SAFE seems to be a great fit for that; traditional message boards are hosted on a server that is controlled by one entity and between the owner, moderators instated by the owner, their ISP and governments, there are a lot of people with the ability to censor. SAFE doesn’t have that problem - unless you have the ability to locate every node containing a particular chunk of data and remove it simultaneously on all those nodes, there is no way you can remove content that has been put on it.

I have not seen one good alternative to SAFE for this. There’s Freenet, but it’s slow, unstable and a pain to set up. There’s also the Bitcoin blockchain, but trying to embed longer messages than fit in one transaction on it, let alone build an entire message board upon it, would be highly inconvenient and a waste of resources.

So I feel this is a compelling use case not only for privacy-minded people, but everyone who values freedom of speech.

Besides, making the network invisible to the user doesn’t have to come at a cost of privacy. As long as the Tor network can be regarded as “as private as SAFE” (or better), using SAFE via a hidden service as an interface does not compromise your privacy. Neither does an easy way of running a SAFE node and receiving Safecoin that hides the inner workings, the other option I outlined in my OP.


#5

I think it’s great you are thinking of friction free ways of people joining SAFE. i think that is the single most important thing after SAFE’s core.

Just look at how hard it is to join diaspora, and see how different it could have been if they had been able to address that!

I don’t fully understand your scheme, but it made me interested, and I wonder if it the WebRTC / in browser vault approach might work (recently discussed) - @Erick posted about a WebRTC / Javascript SAFE client/vault

It creates a very easy way for anyone to join the network, just by visiting a URL/web page, and to be earning Safecoin anytime they are running their web browser. Since Safecoin is ideal for micropayments on the network, this could simplify your idea - by making even those casual message board users into real SAFE account holders.


#6

Actually the more I think about it, the more I think that if we can get the privacy right, a web-based node would make a tremendous difference in the speed at which the network will be adopted, even if it is less efficient that a native version.

In addition, the node ecosystem fosters libraries reuse and fine granularity in libraries. For example, the bittorrent-dht library, which implements the Kademlia Distributed Hash Table, has many dependencies such as k-bucket, that can be reused in implementing the SAFE DHT. It might actually be a reasonable amount of work to port the current C++ implementation to Node. I’ll try a few things in the next weeks and report back. I’ll also continue following webtorrent, since its planned architecture is really adapted to the SAFE context.


#7

@erick its worth following routing_v2 (in common and routing for now) branches as these should introduce some more efficiency and simpler code (I hope). It’s a re-implementation of the core routing algorithm and more importantly serialisation which should tidy an awful lot of the more complex code in that library.


#8

Interesting. Of course, a full node implemented in client-side JavaScript would be ideal.

Admittedly I haven’t worked with it, so please correct me if I’m wrong, but as I understand it, WebRTC involves a set of higher-level protocols. How do you expect to speak SAFE using it?


#9

Using nodes that would speak both protocols to bridge the two networks. See the links mentioned above.