> The FAQs say that the app is a p2p based app. What is the p2p protocol being used here? * How does it compare to libp2p, SSB, and others? * What's the security model, and how would the system mitigate Sybil attacks etc?
Other questions - all good questions, but it depends on what your definition of security model is. Sybil attacks are prevented by cryptographic signatures and proof of work. Unique names are available to provide a good human-friendly polish over cryptographic signatures, as well.
> Are there any plans to have mobile apps (preferable native)? I think if people could use it on their mobile phones, it would be a game changer.
Yes, I would _love_ to have mobile apps. The system supports it, but I don't have the money to spend the time building them, or to pay someone to build them. In essence, it depends on how popular the project gets - if enough people support it financially, it would certainly be a first priority.
Can you elaborate? Googling these terms doesn't produce much other than a few papers using "MiM" to mean man-in-the-middle, as well as your comment itself!
I actually posted the protocol spec somewhere a few months ago, but it might have gotten taken down when I flipped the beta site into the production site.
Mim (comes from 'mime' from 'to replicate') is a HTTP based stateless protocol where the nodes communicate with each other via bog-standard HTTP GET and POST requests. The nodes can create caches for oft-requested data. c0 is a subprotocol of Mim that defines six graph objects, boards, threads, posts, userkeys, votes and truststates. the first four are obvious, the vote is a userkey > b|t|p relation, a trustste is a userkey > userkey relation. It creates a directed graph that is mostly acyclic.
It allows for efficient flood-network propagation of this graph based on time intervals, and it comes with certain patch semantics, so given a delta communicated from one node to another, there is always a way to apply that delta into the local node's graph in a way that does not create merge conflicts.
> I would _love_ to have mobile apps, The system supports it
If there's a library/protocol that could be embedded in a native app, I would love to play with it. I'm a mobile developer with some free time at hand. I'll drop you an email.
Not yet! The protocol docs should be out there, but I think they accidentally got taken down when I replaced the old site with the new one. It's actually a fairly simple, stateless protocol that works over HTTP, so you can very easily script it — it just outputs JSON api responses to GET and POST requests. There are JSON schemas available for it as well. You really won't need any special tools or libraries, just any bog standard HTTP library in any language should be able to handle it. If you're willing to give it a go send me an email (on profile) and I'll let you know when that is ready to give a shot.
> The FAQs say that the app is a p2p based app. What is the p2p protocol being used here? * How does it compare to libp2p, SSB, and others? * What's the security model, and how would the system mitigate Sybil attacks etc?
It's built from scratch. It runs c0 subprotocol of Mim. You can see some basic comparisons here: https://getaether.net/docs/how_is_it_different_from/
Other questions - all good questions, but it depends on what your definition of security model is. Sybil attacks are prevented by cryptographic signatures and proof of work. Unique names are available to provide a good human-friendly polish over cryptographic signatures, as well.
> Are there any plans to have mobile apps (preferable native)? I think if people could use it on their mobile phones, it would be a game changer.
Yes, I would _love_ to have mobile apps. The system supports it, but I don't have the money to spend the time building them, or to pay someone to build them. In essence, it depends on how popular the project gets - if enough people support it financially, it would certainly be a first priority.