Great dive for the advantages part but the disadvantages seems very much picked at random with no clear investigation
Participation node requirements are not that high, I run it on a normal pc and know some that use a rasberry Pi, those are the recommended future proof requirements that can handle all sorts of traffic.
About relay nodes:
1) they will become lighter to run
2) they will become optional for txn to go through since peer to peer node network will be introduced alongside dynamic round times to allow for extremely fast execution.
2) there are 120+ relay nodes geographically distributed and each txn is sent to 5+ different relays so it's not so easy to apply censorship.
3) anyone can run a relay the code for it is the same of the participation node but the permissioned part is only the default link from nodes to the relays, so you could set everything up without relying on external infrastructure network if you run both relay + part node pointing to your relay. No censorship there.
Currently you can run a participation node on lightweight hardware because no one is using the network. If Algorand was actually doing 6000 tps your machine wouldn't be able to keep up if it doesn't have the minimum requirements.
Either Algo can do 6000tps and the minimum node specs are high or Algo can't do 6000tps and the node specs are low. It can't do both.
1. Any source on Relay nodes becoming lighter to run? From the widely accessible data I don't see that anywhere. It seems unlikely unless they plan to lower throughput in the future.
2. Any source of there being plans for a peer to peer network? Again I don't see that from the widely accessible data anywhere.
3. You only need to prevent 25% of participation nodes from receiving a block to censor it as blocks need 75+% votes from a subsample to reach consensus. It's still difficult but this is much more censorship prone than other cryptos.
Also whilst there may be 120 relay nodes, they are all of a similar profile. If a government bans a smart contract, what percentage of relay nodes are you confident would ignore the rules and not censor? Given their profile of being entities like universities, it seems unlikely many would.
4. Yes I said this in the post that anyone could run one. The incentives to do so aren't there though and the access to subsidies requires permission.
If I set up my own external node and connected to a permissioned system would that mean the permissioned system couldn't censor me? No.
The network is periodically tested by the Inc and community testing showed also a throughput of 10k tps, 6k tps is a "theoretical minimum". But the same team that tested the swap also did test the throughput and reached these numbers, the network could handle it, those requirements are still not a problem.
1 and 2) https://www.youtube.com/watch?v=ghmLVxPrmSA in the video here Gary VP of Engineering talks about the future development for Algorand. So very much likely we'll see these changes coming in the future.
About relays we could debate the actual practicality of what you want to imply with that but still the system could work around it very easily since community could if needed set up it's own relay system by coordinating the nodes to point to those relays too.
There is no hidden code to set them up or anything. The default line is there to guarantee stable performance.
However considered that these infrastructure design will change in the future I don't see that as a disadvantage vs things that other blockchain are facing to fix their design.
Nice, I watched the first 20 min of the interview and he did address a lot of the problems I had and everything he said sounded great. He also mentioned better staking incentives which was another one of the negatives I had.
If they do manage to do these things it will make Algorand much stronger and will remove most of the negatives I have. I'll make some edits in the post later to at least say that they are working on the problems
Great dive for the advantages part but the disadvantages seems very much picked at random with no clear investigation
Participation node requirements are not that high, I run it on a normal pc and know some that use a rasberry Pi, those are the recommended future proof requirements that can handle all sorts of traffic.
About relay nodes:
1) they will become lighter to run
2) they will become optional for txn to go through since peer to peer node network will be introduced alongside dynamic round times to allow for extremely fast execution.
2) there are 120+ relay nodes geographically distributed and each txn is sent to 5+ different relays so it's not so easy to apply censorship.
3) anyone can run a relay the code for it is the same of the participation node but the permissioned part is only the default link from nodes to the relays, so you could set everything up without relying on external infrastructure network if you run both relay + part node pointing to your relay. No censorship there.
Currently you can run a participation node on lightweight hardware because no one is using the network. If Algorand was actually doing 6000 tps your machine wouldn't be able to keep up if it doesn't have the minimum requirements.
Either Algo can do 6000tps and the minimum node specs are high or Algo can't do 6000tps and the node specs are low. It can't do both.
1. Any source on Relay nodes becoming lighter to run? From the widely accessible data I don't see that anywhere. It seems unlikely unless they plan to lower throughput in the future.
2. Any source of there being plans for a peer to peer network? Again I don't see that from the widely accessible data anywhere.
3. You only need to prevent 25% of participation nodes from receiving a block to censor it as blocks need 75+% votes from a subsample to reach consensus. It's still difficult but this is much more censorship prone than other cryptos.
Also whilst there may be 120 relay nodes, they are all of a similar profile. If a government bans a smart contract, what percentage of relay nodes are you confident would ignore the rules and not censor? Given their profile of being entities like universities, it seems unlikely many would.
4. Yes I said this in the post that anyone could run one. The incentives to do so aren't there though and the access to subsidies requires permission.
If I set up my own external node and connected to a permissioned system would that mean the permissioned system couldn't censor me? No.
The network is periodically tested by the Inc and community testing showed also a throughput of 10k tps, 6k tps is a "theoretical minimum". But the same team that tested the swap also did test the throughput and reached these numbers, the network could handle it, those requirements are still not a problem.
1 and 2) https://www.youtube.com/watch?v=ghmLVxPrmSA in the video here Gary VP of Engineering talks about the future development for Algorand. So very much likely we'll see these changes coming in the future.
About relays we could debate the actual practicality of what you want to imply with that but still the system could work around it very easily since community could if needed set up it's own relay system by coordinating the nodes to point to those relays too.
There is no hidden code to set them up or anything. The default line is there to guarantee stable performance.
However considered that these infrastructure design will change in the future I don't see that as a disadvantage vs things that other blockchain are facing to fix their design.
Nice, I watched the first 20 min of the interview and he did address a lot of the problems I had and everything he said sounded great. He also mentioned better staking incentives which was another one of the negatives I had.
If they do manage to do these things it will make Algorand much stronger and will remove most of the negatives I have. I'll make some edits in the post later to at least say that they are working on the problems