Two answers. Yes & No.
There are two types of Mesh networks.
- Ad-hoc Mesh, in which devices come together, forms a network and traffic is routed from one device to another in p2p fashion. Users with their mobile devices are enough for this type of network.
- Infrastructure Mesh, in which we need to create a wireless back-haul mesh among wireless routers and Access Points. Users with their mobile devices + wireless mesh infrastructure is required for this type of network.
It’s easy to achieve ad-hoc mesh, since it doesn’t require anything other than users and their devices, but these networks cannot replace the mainstream Internet which is hierarchical. So, Yes, we have tried ad-hoc mesh using Project Byzantium and we are already using Rumble (I am still confused whether rumble is just ad-hoc or it’s a ad-hoc mesh).
Infrastructure mesh needs to be setup with routers and access points and that involves additional cost. And moreover not all routers can participate in the mesh network, we need to find the OpenWRT supported devices and flash with it or we could use a SOC like beaglebone / raspberry pi. Either way, without identifying proper mode of funding, we cannot proceed with this type of mesh. This would be time consuming, nevertheless this is what we need to push for. These infrastructure based mesh have the real potential to replace the existing hierarchical Internet thus eliminating the need for ISPs. So No, we haven’t yet started working on infrastructure mesh networks.
Whatever type of mesh network it may be, we should remember that we also need applications that are also p2p to deploy on these networks. Though decentralized but not p2p applications will work on these networks too, but as more people join those services, they too will become centralized. For example, Diaspora can be hosted on a mesh network, there can be many pods, but if only one pod grows in number of users and the other pods fail to attract people, that will lead to centralization again.
P.S: I bought a wireless router from TP-Link exclusively for setting up a mesh node in my house. Before attempting to flash the router with OpenWRT, the router has got a manufacturing defect and I have submitted it for repair. When I receive the router, my first step would be to flash it with OpenWRT Image and install mesh protocols like olsr / batman-adv / cjdns.
Blockstack is decentralized DNS and identity.
With the Blockstack software, a network of computers collectively maintain a global registry of names.
When you run a Blockstack node, you join this network, which is more
secure by design than traditional DNS systems and identity systems. This
is because the system’s registry and its records are secured by an
underlying blockchain, which is extremely resilient against tampering
In the registry that makes up Blockstack, each of the names has an
owner, represented by a cryptographic keypair, and is associated with
instructions for how DNS resolvers and other software should resolve the
Blockstack is already being used in production and currently more than 46,000 names have been registered using it.
There is also another chat application called the ring. Get it here. https://ring.cx/
Pirate Bay on the mesh ??
Looks so . .
I find IPFS very promising. It really seems to have the potential to redistribute the web getting rid of the large centralized datacenters.
I am now playing with IPFS on my home server (unlimited BSNL landline with FUP). My first impression is very positive. But, I notice that it consumes a lot of bandwidth (some kind of “seeding”, I guess). There’s no way I could use this on my mobile 3G connection with very expensive and limited data. We really need to eliminate all these irritating data limited plans. Or, we should prefer to use land line connections, or probably better still, run on our own mesh networks.
By the way, is there much content on the IPFS network at this point? Does anyone know any websites or other data available on the IPFS network? Is there some index or directory of this somewhere? It would be really fun to start using the IPFS network for real. So, do let me know if you know anything about this.
I found this article shared on GNU Social. I thought it might add to the discussion here.
The blockchain will be very useful for registering large corporate
capital markets and making cross-border banking transactions, but as a
system for the development of everyday applications on the Internet,
it’s a danger to the distributed structure of the network.
I shared this in FSFTN Telegram group and we had few discussions there. @venkatesh.thennarasa was asking, how come blockchain could be a threat for decentralization while the primary objective of blockchain is the same. He also asked, isn’t time vs memory trade-off is what we should expect in such systems?
The article argues that the infrastructure cost involved in setting up a blockchain social network will be so high that a few large nodes will dominate the network.
When I started reading about adoption of Blockchain into financial systems, I too was worried, puzzled and amazed, how quick the capitalist system absorbed an idea which is alien to it. I am still not clear on its implications and on what basis system ready to leave control of the centralized control.
But I am sure, system will make necessary ecosystem to sustain itself. Like cost of hardisk is historically in lowest point and further getting lower, and all the devise we have have atleast few GBs by default, so in future definitely space constrain would not be of a problem, we need to be worry about. But how the financial institutions and capitalist system confident enough to go for block chain and decentralize their control/power. Which is in my opinion will revel a greater clarity on issues with Blockchain.
When there is a space constrain app like Twister, GNU Social, or Diaspora, with necessary freedoms, a community can host a node collectively and manage it based on transparency and democratic. When there is no space constrains like chat apps, which may not need max space would be ideally using blockchain. There cant be just no one solution for a systemic change. Users rights and freedom is all matters now.
libp2p is the product of a long, and arduous quest of understanding – a
deep dive into the internet’s network stack, and plentiful peer-to-peer
protocols from the past. Building large scale peer-to-peer systems has
been complex and difficult in the last 15 years, and libp2p is a way to
fix that. It is a “network stack” – a protocol suite – that cleanly
separates concerns, and enables sophisticated applications to only use
the protocols they absolutely need, without giving up interoperability
and upgradeability. libp2p grew out of IPFS, but it is built so that
lots of people can use it, for lots of different projects.
Awesome discussion guys, i too have read and understood some of the protocols (at least papers or at functionality level) and had a similar effect after seeing a good movie that immerses us with its mood. In such analogy cinema observed for few hours that aligns with our idea is very much similar to protocols that aligns with our idea. Initially which looks right in every dimension after a lot of contemplation and analysis reveals its limitation / vulnerabilities / barriers, etc…
On trying to take out the common pattern, of all the recent protocols each propagated with different view points of the same pole, it seems that the orchestration of p2p and encryption gaurantee protocols, as @prashere mentioned above. So it evolves into a system that distributed systems are being built of top of these fundamental support structures.
We currently have these strategies of Inter-Networking:
- Totally Proprietary and Centralized systems
- Proprietary but Decentralized top-down (Geographically) with Central control (confused ???) - A real weapon
- Partially - Open source systems (Single End Open) with Central black box & control - A nightmare
- Fully Free & Transparent systems without encryption
- Fully Free & Transparent systems with encryption, & Share Tracking (Decentralized bottom-up) - similar to Torrent
- Fully Free & Transparent systems with encryption & Distributed Sharing - like Retroshare
- Fully Free & Transparent systems with encryption & Distributed Sharing + Hosting - like Zeronet, IPFS
- Fully Free & Transparent systems with encryption for Naming - like blockstack
am i right ?
Block chain being a different animal, can it scale up in present situation ?
I am interested to know, how come we can evaluate a projects outcome as a distributed or decentralized system ?. Looking at code is one thing which is solid and good. Is there any other way ?
With diversity in protocols, we also have responsibility to manage interoperation. On the whole i am happy that these projects are started without any hope for success or failures. But as a triumph. I am personally proud to have CJDChat built by @suriyadeepan, @Vanangamudi & @nnpths01 to observe sub sproutings. I believe soon many others will do similar things.
We have to remember with more diversity, systems tend to evolve and nobody know which one will adapt every changes…
One more question, what you guys feel about the difference between the words Decentralization & Distribution ? I was hearing people using these words interchangeably in discussions with their biases and so there is a loss in communication due to ambiguity. I see decentralization as a transition phenomenon between centralized systems to distributed systems. Some say it is a special case of centralized system and other say it as a special case of distributed systems. We have understood commonly that Distribution is possible with bottom-up approach. Shall i call Torrent as decentralized and ZeroNet as distributed. One more thing is, in network fundamentals at the physical layer afaik, there is distributed mesh network, there is distributed systems which is more aligned to our thought process.
People also use decentralized in political and social terms, similar to power decentralized from central gov. to state gov, and the tree structure keeps on going until it reaches the grassroot forming a hierarchical system. So if decentralized system result in a hierarchical (possible for subjugation) how come we are still using that word too in our conversations. I do not want to make it alike s a serious issue between the gnu/linux or gnu or free or open source… These are my doubts lingered for years.
For layman explanation i dont want to make it serious, but with critical thought i want to make myself clear.
This should help you
@ramaseshan, i know such visual/abstract difference… my point is need to know which words the community peers are meaning ? Particularly in presentations, deliberations, community meetups most people use the word “decentralized” while conversing about mesh/distributed networks & systems.
Also I like to know about the tech cast initiative and the resources they have to do so !
GNU Social is extremely light.
Blockchain based systems will likely be the heaviest.
Wow! Something worth trying.