Finally, I think I understand a little better now. The question has been nagging at me regarding how big of a problem it would be if most Nxt users didn't bother with forging, which is what I think MOST people would not do if Nxt became a big enough thing (UNLESS it was made EXTREMELY easy to do) . As long as enough honest, large stakeholders are doing it then I guess that should not be a problem.
I hate to ask this, but have you guys examined or considered stealing some ideas from the way Ripple manages their system? Right now the "servers" are run by ripple Labs, but supposedly they will have users run them in the future. I wish I understood it better, but I know it has something to do with using a large enough and diverse enough pool of 'verifiers' that could not possibly collude with one another because people who don't know each other are incapable of coordinating in large numbers to do something dishonest, even if individually they are not trustworthy.
I wish I was more qualified to ask better questions, but all I know is that XRP is transmitted extremely quickly and there might be an idea or two to steal there. They seem to feel that a very very tiny transaction fee is enough to eliminate spamming.
As I posted a few pages back... its great NXT solves the problem of CPU for confirming transactions, but you still have the problem of network bandwidth and memory, now you can forge on your mobile but in future (when NXT is successful) unless you have the bandwidth you won't be able to keep up with the block download and it will fry your data plan and similarly you will chew through more of your home broadband plan.
I don't think bandwidth will be a problem when clients can sign transactions locally and broadcast it to a public node. Most users with less than 1000 Nxt will not have to download the entire blockchain. They will run a lightweight client. The network will only need a few hundred nodes that will maintain (upload/download) the entire blockchain.
Bandwidth will not be a problem
If 300 hub nodes are enough, then isnt that a cost of ~$5000/mo?
Infrastructure committee has 3 million NXT means worst case, 2+ years subsidy. By that time there should be plenty of businesses that are running their own servers, or NXT is worth significantly more than 5 cents, or both
On the issue of pools for forging, my understanding was that the NXT is "loaned" to the forging acct, so that even though there might end up being 10 pools, maybe it is not a problem at all. Each pool's chance is the sum of all of its participants chances and all participants share the block's fees. However, which node actually issues the block? Wouldnt it be the one with closest to the magic number? So any node that is part of the pool could end up the chosen one, just the fees end up in a pooled account. If people will get fractional NXT 144 times per day (1440/10), even though statistically it is the same reward I think they are more likely to be connected to the network.
Do we know for a fact that with pooled forging, an acct can be forging even though they are not connected to the network? Maybe all we have to do is make that a requirement?
In any case, I am designing a NXTcoin development kit that builds on top of NXT for our mining friends. They will need to run NXT in order to mine the NXTcoins. If we get just a single successful NXTcoin the network will have thousands of nodes. We could easily have a dozen successful NXTcoins.
Anybody that wants to offer a NXT service, will have to run a node. As NXT gets more and more features, it will attract more and more businesses that utilize these features.
As long as the infrastructure committee figures out what we need for varying levels of NXT network usage and a way to monitor where the current level is and is prepared to subsidize any shortfall of capacity, there are no worries.
We have 3 million NXT to figure out the requirements, monitor it and subsidize it if necessary.
In the meantime, we build tools to help people offer NXT services.
James
