Quote from: georgem on March 12, 2016, 08:14:26 PM

Quote from: rhinomonkey on March 12, 2016, 06:39:08 PM

Georgem,

Upon a node being kicked from the network due to not meeting the criteria of a productive node, will the network automatically recognize the next node "in line" with the next highest collateral? Meaning, will there be a queue of nodes bidding to get on the SN network (if for instance someone couldn't or didn't want to buy more SPR to get on the network - they merely keep their node with less than sufficient collateral running on a vps waiting for a spot to open)?

The way I focus on implementing it is not by kicking nodes from the network.
What will happen is that your "bad" node will be left out of the payment-cycle and receive no reward, but your node doesn't stop doing it's (although crappy) job.
Ofcourse it is also possible that your node is misbehaving so badly that other servicenodes have put it on a temporary blacklist, then your node is like a mental patient talking to itself.
But there is no "shutting down" mechanism involved at all.

Furthermore, the network knows at all times the amounts of collateral that any servicenode is currently willing to invest.
Those are locked SPR amounts, visible in the blockchain.

Quote from: rhinomonkey on March 12, 2016, 06:39:08 PM

i.e. Say a node gets kicked for being a crappy node, and I have a node on a VPS with the collateral of say 995 (the highest of any other sitting on a VPS), would my node be immediately inaugurated into the service network, or would I have to do something to make it a SN after recognizing there was a spot open?

Forgive me if you have addressed this somewhere. I didn't see it in your original update - only that one could *take* a seat. Not *receive* a seat after waiting.

Yes, that's what I'm trying to achieve, a more or less automatic re-entrance.
If your node was already running, but couldn't occupy a seat because all seats were taken, then the moment the situation changes, you'll immediately re-enter the game.

In this regard... seats don't really exist, you never occupy anything. it's rather a continuum that is constantly recalculated to find out if you are eligible for a reward.
And this continuum (the spread) can be visualized in the form of seats. This visualization is for us humans, not for the network.

Well there is more to the story, because the scoring mechanism has some say too, and it is quite possible that you..

a) have a running servicenode
b) are eligible to receive a reward
but
c) your scoring is so bad that your reward is nullified.

I am doing experiments with letting your servicenode be prepared so that it can automatically react to a constantly changing situation.

If you look at this screenshot under "Auto collateral tasks"...

... this will be a mechanism you can setup to let your servicenode automatically do all kinds of things so it can quickly react when the "collateral market" changes.

For example you can say here:

If your current COLLATERAL falls lower than the MEAN of all other collaterals in the network, then go ahead and ADD enough SPR so that your COLLATERAL is again atleast as large as the MEAN.

There will be more such trigger points you can define in an IFTTT (if this then that) fashion.

Would you be interested in a testnet version where several of us could help in beta testing?