Would it help to add some more network intelligence to NXT so it became more self-organising, so the nodes were more aware that they were part of a network and integrating into that network properly makes it stronger, rather than just being, well, nodes with some connections...

At the moment the network organisation of NXT is a mystery, I don't know if nodes look for nearest nodes, and optimum number of connections based on available bandwidth, or just grab hold of any node they happen to see.

Self optimising isn't complex - it just requires each node to be a 'good neighbour' - although in our network you need to re-define the location attributes which can be ip-addr, latency etc

Some of this I can see NXT nodes do from their behaviour, but not all of it, maybe I don't know the mechanics well enough and more is going on.

Anyway I don't know if its of interest but I thought I would add it to the mix, translating what I know of the mobile world this is what self-organising characteristics would look like;

Self Configuring means new nodes shall automatically be configured and integrated into the network. This means both connectivity establishment, and download of configuration parameters. When a new node is introduced into the network and powered on, it gets immediately recognized and registered by the network. The neighboring nodes (self optimising) would then automatically adjust their technical parameters in order to balance the required coverage and capacity, and, in the same time, avoid unnecessary conflict.

Self Optimising means each node adjusts its parameters to ensure it fits in with the its immediate network neighbourhood, based on observation of neighbouring nodes and its own traffic, for example in periods of over capacity the nodes could all configure a shared sleep time to take it in turns conserving energy when there is not enough work to justify the number that are in existence.

Self Healing means when one or more nodes in the network become inoperative, failure or ddos, self-healing mechanisms aim at reducing the impacts from the failure, for example by adjusting parameters and algorithms in adjacent nodes so that other nodes can support the users that were supported by the failing node.

The winner can be picked randomly by using all previous  blocks hashes (since the last winner)  as a seed to a pseudorandom generator. All nodes will agree as the answer should be the same

 

Would there be any proof of stake in your coin? What happens when the coin is mined to capacity? To me the whole issue with people who mine is that they only mine for the coins. What will happen when there are no coins left to mine? They will move on to the next coin. The whole thing about the mining logic is that they are not in it for the network (which I can understand). To me it seems that POW is not the best answer in the long term. A mined coin may get a lot of initial support but in the long run, POW is not going to compete with POS..... especially when the asset exchange allows people who don't know how to write program to create there own coins. I think for a coin to be long term successful POS or some form of rewarding has to be used to keep people using running the network once the coins are mined out. I like your idea though.

If the winner is chosen every hour, how many nodes will turn on and off every hour? That will take more effort then keeping it on.

How about every 30 minutes?

This could  be reduced even to every single block (1 minute). Forging nodes gets some percentage and other percentage is awarded to random node that is running.

Seems a little similar to Ripples use of the World Community Grid (leasing CPU for medical and social research) to award XRP. If something like that could be done with business analysis or similar that would be pretty cool.

Please run for infrastructure committee! These are some cool ideas. We might get emergent behavior from a self-organizing network.

can someone explain in more detail how 0.8.1e client run?
which is the file to run the client?

is necessary the start.bat file

if its the same as the 0.8.0e file

Put the file in your the root of the nxt folder, In your old start.bat file (or whatever you called it to start up NRS in the 0.7.* versions) where it had the old "java start.jar" stuffs. Type in  

java -Xmx1024M -cp nxt.jar;lib\*;conf nxt.Nxt

and replace whatever was in the .bat file previously.

Had to search through the thread on what to do...lol