Should Foam build it’s own blockchain?


#1

I believe that Foam should move from being a geospatial protocol on top of ethereum, to building it’s own blockchain. This is essential for adoption as we can’t rely on an ERC 20 token model to build incentives. Consensus mechanisms can be defined effectively only on a standalone blockchain.
We may have scaling solutions like Casper in the future. But the gas costs can clog the network on a particular day. Imagine time delays when dPOL goes live.
I may be wrong, but let me know if this makes any sense…


Thinking Out-Loud - How will the Future of FOAM Play Out?
#2

I like this exploration @Regan. We should always question the reliability on other solutions. I’m on board with doing that with you! Let’s not get complacent and just accept things as they are. Let’s continually assess. Always re-evaluate. :facepunch:

However, it’s my understanding that the quantity of nodes running the software is important. The fewer the nodes the easier (less expensive) an attack. That’s why, at the present time, Ethereum is good. It’s big. It helps protect the network. At present, today, Ethereum is running at least 11,800 independent nodes.

https://www.ethernodes.org/network/1

It may be more. I didn’t see a summary page at first glance but that node explorer website is very cool. I read somewhere Ethereum had 25,000 nodes on it last year.

There are only 1,004 cartographers. All of us aren’t going to be nodes. Many not interested in that. Lets say half the community run nodes. That’s 500 nodes. That makes a very small attack vector for a malicious actor. An attack, at this stage, could highly lucrative for a malicious actor financially and someone like google maps or Apple maps could shut FOAM down fast. FOAM is building its reputation for security. Ethereum still offers this.

I welcome feedback, input and corrections.


#3

Why would you even need an own blockchain lol


#4

what do you mean @foammarine ? the OP listed a whole bunch of reasons. instead, don’t you think what would be helpful is if you made a point specific credible argument for why you don’t think it’s a good idea?


#5

the attack surface/vector argument is an interesting and important one to be considered. however i believe it can be mitigated by implementing things like sidechains. this would allow the separation of data (storage) and computing (consensus). so the suggested FOAM ledger could run consensus on fewer (heavier ledger) nodes referencing from many nodes acting as (lighter ledgers) sidechains running on ethereum and/or cardano or elsewhere.

this hybrid approach would make it more affordable to run and create a lot of room to scale. having separate domains with unpredictable variability of randomised selection of nodes/ledgers - if implemented right would theoretically optimise for both speed/accuracy and security/reliability - two very integral necessities for FOAM.

these sidechains could, in the future, be hosted/running on the FOAM chain as well as on dapps utilising location data in exchange for a fraction off fees.

compromising the network would mean having thousands of nodes conspiring for an extended amount of time without detection. which would be quite difficult, not only because the cost opportunity disincentivises it, but in running some basic mental simulations - a few passes would raise an alert highlighting the nodes with anomalies in their ledgers - who would be a comparative minority. which is to say that probabilistically getting away with it let alone doing it in the first place a very very slim.

i believe if FOAM is serious about supplanting GPS it needs to invest heavily upfront - cause what it’ll cost today will be made back many many times over in the very near future. initiatives like this own blockchain and the one to have (micro) satellites - would not cost millions but thousands - and would mean serious adoption because it’ll be reliable. as things are i wouldn’t run an application on FOAM let alone want to be on a aeroplane running on it.


#6

Gold talisman.

I understand how this helps in scaling. But can you kindly please explain what you mean by affordable? Thanks. [quote=“talisman, post:5, topic:179”]
this hybrid approach would make it more affordable to run
[/quote]

Yeah. Cool. :point_down:

Brilliant, fascinating thoughts here👇

I have questions about this because I have no real life side chain experience myself. In my limited knowledge of side chains: consensus is reached on side chains and only disputes are resolved back on main chain. But sometimes a discrepancy can require the main chain, in the case of your example, FOAM, to need to soft fork to correct the discrepancy.

Perhaps the benefits of running side chains in FOAM and Dapps outway maintenance, risk or other “costs” of side chains? I would like to hear a little more detail on the case for this. At the end of the day, which chain or configuration of chains is just a cost analysis exercise (“cost” not necessarily being monetary). Do you agree?

I can see there might be value here. Just interested in knowing more.

:point_down:important to have these at the front of discussion imho. Time is of the essence. The choices made now will be amplified with time. (That’s not to say I disagree with any decisions at this point, just agreeing with you about small investments now may mean big shifts in future direction) The micro satellites are particularly appealing since we live on a blue planet… not a green planet. Would be interesting if there was a group in the community that had this as their main focus.

Thanks for your post.


#7

You can already think of FOAM Dynamic Proof of Location as being its own blockchain, but thank you Regan for bringing up and kickstarting this discussion. You are correct that there will be latency to the finality of a presence claim going from a Zone to fully verified and checked for fraud, and this is something that needs to be optimized. Additionally, different use cases can select their own levels of security in terms of what stage a Presence Claim is in, similar to how some applications of Bitcoin make you wait 6 block confirmations or in some cases many more.

That said, the architecture of Dynamic Proof of Location needs to rely on a root chain and this chain needs to have the highest level of security. This is where proofs are posted as final and the initial staking occurs. Right now the most secure root chain is Ethereum and Ethereum 2.0 is likely to be a strong candidate in the future for dPoL.

Zones run a time synchronization protocol over radio and share a state machine to store the data and Presence Claims they may have received. Zones use Tendermint consensus to update and share their state machine data log. This data is then checked by a verifier for fraud and then posted to the root chain. To become a Zone operator you need to stake tokens and enter into a service level agreement. Each zone of dPoL is its own blockchain, but a Proof of Authority one that in order to be a validator in you need to stake tokens on the root chain. Zones are side-chains in this sense and could be implemented with Plasma or other solutions.

Additionally, there are exciting alternative developments in the space for a root chain. For example Cosmos as well as the recently revealed Subtrate by Parity, which allows a custom blockchain to be spun and attach itself to Polkadot to rely on their security. See this talk and demo by Gavin Wood:

The final architecture of dPol will be a function of general developments in the blockchain ecosystem and a result of community testing and optimization.


If the foam runs on IOTA it's great. Here is an app that is using GPS.
#8

Very good observation. In light of recent attacks on minor chains (https://www.crypto51.app/) and considering FOAM’s vision to be part of technology layer that bridges cyber and physical worlds, one cannot be too careful with its security design. All the nice use cases can be compromised by fundamental vulnerability of the underlying tech.