NOTE: This project is entirely for community consideration and discussion. There are no people from DFINITY or any other orgs working on it (as far as we know).
Summary
Discuss whether the DFINITY Foundation should expend resources to develop a low-barrier network for hobbyist node providers. Debate whether it should be a subnet, a new network, etc.
I say this should be a low priority for now. It sounds like the IC has more node operators waiting to jump in than it can handle. I would propose onboarding as many of the current node operators with the current hardware spec as possible, and then move forward with cutting the requirements. Once subnets with lowered requirements are stable, perhaps lower the requirements again and create new subnets. At each stage of this process, many technical issues will probably pop up. We can slowly address them as the network becomes more and more accessible to node operators.
I agree with lastmjs, this should be low priority. If I could I would personally put it near the bottom of all the current proposals. I think onboarding with current hardware for new node operators is a much better option for a while down the road then pursuing this path. There’s definitely a lot of interesting things that could come out of this, but from my perspective, comes across as knee jerk reaction to a lot of FUD out there when that energy could be focused on more productive proposals for the Dfinity ecosystem.
Long term? Great idea, would love to steal one of my partner’s Raspberry Pis and do this. Next 1-2 years? I don’t see what’s to be gained personally at the moment vs other proposals.
I’m with the other two that it’s a fascinating concept and i can’t wait to be involved, but there are obviously bigger fish to fry. Focusing on core power and increased adoption will lead to the need for badlands naturally.
I feel that the prioritization should be maintained and not lowered. I base this reasoning on the fact that having minimal hardware (RaspberryPis) will boost the participation rate amongst the growing community. In addition the performance amongst different hardware stacks would become quite distinguishable. This will , then, justify the difference between high end nodes and the Pis; making it easier to understand why specialized hw is required.
I disagree with the notion that this topic is low priority - the disproportion of resources between the have and have nots , embed inequality and hampers entrepreneurial development in the developing world - waiting to be invited to the table after the wealthy have developed the infrastructure will be an indictment to the vision of the Dfinity foundation - democratising opportunity.
I really don’t care If Badlands exists as a subnet or as a separate chain. Either way it should serve a purpose. I think a testnet or ‘canary network’ would be beneficial.
Just following up. This is the kind of thing that makes me believe a separate badlands chain (or subnet if it can be isolated) would be beneficial for testing new technologies.
I just updated the summary to help differentiate between the proposals which are more baked and which those are still in more conceptual phases.
Fwiw, To see an example of a project which is more baked (and a model of how these proposals should look more like) check out the timeline thread on Increased Canister Storage: Increased Canister Storage - #13 by flyq
I think that I need to understand a little better.( NOTE: This project is entirely for community consideration and discussion. There are no people from DFINITY or any other orgs working on it (as far as we know).).
Implement badlands concept as new subnet chain. If forking a new chain is the selected option, it will have to be forked many times in the future to add node machines with different capacities.
Yeah, I think this relates to part of my post under Tokenomics. The communication on Badlands (and Endorphin) lead to confusion. I hope after these proposals are thoroughly discussed the foundation can help communication a clear priority and a bigger roadmap so we can all get a clearer picture.
There are certainly a lot of conflicting pros and cons here. From my perspective, taking part in the network as an amateur node provider and potentially reaping benefits from it as well is very appealing. Otherwise, the current node machines distinguish the internet computer from other blockchains and provide enough computing power to serve the web-speed apps we are currently seeing. I’d love to have more information on the type of smart contracts for which amateur node machines would be the better option and then also how the interoperability between the blockchains (badlands and main) has to play out.
On the other hand badlands could be a big step toward a more decentralised IC and it would allow people across the world and from different income groups to participate in the network without having to buy expensive equipment first. Hence, badlands could be a big democratisation step.
What could be improved from my standpoint is the proposed way in which amateur node providers should be verified (people parties). Seems a bit too analog to me and would bring a huge coordination effort with it potentially hindering the speed at which an amateur network would establish itself.
Without some type of Badlands participation on a large scale am unsure how the goal of ‘decentralizing the IC’ can ever occur.
Power centralizes …if u distribute power by increasing the number of participants you create the only real long term bulwark against your desired enemy … the centralization of power.
Is that the enemy you wish to fight or is it the one you want to emulate? The human condition is contagious folks…
From what I see the investing ecosystem is looking for the big demographic. They want B in front of the illion when it comes to eyeballs ( Worldcoin sign me up! )… if IC doesn’t implement Badlands … someone else most definitely will using comparable tech.
I think the system can be decentralized without Raspberry Pis or other low-end hardware. Does decentralization require cheap hardware? If there are enough independent parties who can afford the high-end node hardware, then we can achieve the desired level of decentralization. If Dominic did not misspeak in his latest podcast appearance on Epicenter, there is currently a queue of ~4000 node operators waiting to get into the network. That sounds like it will be very decentralized, no Raspberry Pis or other low-end hardware needed.