[ANN] SpreadCoin | True Decentralization (No Pools) | Testing New Masternodes

[defunctec]

So, will masternode payments still depend on a score, because if so it will defeat the purpose of them being dynamic.

I must admit that was something I didn't care for with the original SPR masternode testing.  I didn't understand the benefits of the highest scoring masternodes getting paid.  Am I missing a really obvious benefit?  Surely the network can only be fair if all masternodes get paid roughly the same amount over a period of time?

Why define a max number of masternodes?  Surely the ROI, coin supply and value of SPR will define this organically?  People won't run masternodes at a loss for long...

I also remain to be convinced of the dynamic masternode model other than its brilliant ease of setup.  Once again...am I missing a fundamental advantage?

The score was designed to sort the men from the boys.

Servicenode owners will try to load their nodes with as little SPR as possable (so to have move nodes/profit).
This create's instability in the network, as people are getting kicked off the list because of the small amount in the node.

So to create some stability, long term nodes that are well funded get a higher rating and probably more SPR block rewards.

That's my understanding anyway

[e1ghtSpace]

So, will masternode payments still depend on a score, because if so it will defeat the purpose of them being dynamic.

I must admit that was something I didn't care for with the original SPR masternode testing.  I didn't understand the benefits of the highest scoring masternodes getting paid.  Am I missing a really obvious benefit?  Surely the network can only be fair if all masternodes get paid roughly the same amount over a period of time?

Why define a max number of masternodes?  Surely the ROI, coin supply and value of SPR will define this organically?  People won't run masternodes at a loss for long...

I also remain to be convinced of the dynamic masternode model other than its brilliant ease of setup.  Once again...am I missing a fundamental advantage?

The score was designed to sort the men from the boys.

Servicenode owners will try to load their nodes with as little SPR as possable (so to have more nodes/profit).
This create's instability in the network, as people are getting kicked off the list because of the small amount in the node.

So to create some stability, long term nodes that are well funded get a higher rating and probably more SPR block rewards.

That's my understanding anyway

But then it would remove the competition if they were paid more, and hence defeating the purpose of a limit.

Edit: And what do you mean instability? How will it be unstable?

[Starin]

I wish there was a way to limit each user with 1 masternode, that would be perfect.

[e1ghtSpace]

I wish there was a way to limit each user with 1 masternode, that would be perfect.

If there was a way (probably only possible with a centralised coin) I'm sure it would be done.

[defunctec]

So, will masternode payments still depend on a score, because if so it will defeat the purpose of them being dynamic.

I must admit that was something I didn't care for with the original SPR masternode testing.  I didn't understand the benefits of the highest scoring masternodes getting paid.  Am I missing a really obvious benefit?  Surely the network can only be fair if all masternodes get paid roughly the same amount over a period of time?

Why define a max number of masternodes?  Surely the ROI, coin supply and value of SPR will define this organically?  People won't run masternodes at a loss for long...

I also remain to be convinced of the dynamic masternode model other than its brilliant ease of setup.  Once again...am I missing a fundamental advantage?

The score was designed to sort the men from the boys.

Servicenode owners will try to load their nodes with as little SPR as possable (so to have move nodes/profit).
This create's instability in the network, as people are getting kicked off the list because of the small amount in the node.

So to create some stability, long term nodes that are well funded get a higher rating and probably more SPR block rewards.

That's my understanding anyway

But then it would remove the competition if they were paid more, and hence defeating the purpose of a limit.

Edit: And what do you mean instability? How will it be unstable?

Their not paid more because that's in the code. There paid more because they're more consistent, handle more transactions and respond 100% of the time.

Also the larger your servicenode balance is, the larger transactions your node can handle. And we need the ability to handle large transactions so we need large wallets.

Am i right in saying Dash nodes can only handle 1000 dash at a time?

Edit: If everyone put a small amount of SPR into their SN wallets, the network would be kicking and re installing new nodes allllll the time.
If we have large wallets that dont get kicked, we always have a solid node to rely on.

I'm struggling to even understand myself, so don't think its you

[e1ghtSpace]

So, will masternode payments still depend on a score, because if so it will defeat the purpose of them being dynamic.

I must admit that was something I didn't care for with the original SPR masternode testing.  I didn't understand the benefits of the highest scoring masternodes getting paid.  Am I missing a really obvious benefit?  Surely the network can only be fair if all masternodes get paid roughly the same amount over a period of time?

Why define a max number of masternodes?  Surely the ROI, coin supply and value of SPR will define this organically?  People won't run masternodes at a loss for long...

I also remain to be convinced of the dynamic masternode model other than its brilliant ease of setup.  Once again...am I missing a fundamental advantage?

The score was designed to sort the men from the boys.

Servicenode owners will try to load their nodes with as little SPR as possable (so to have move nodes/profit).
This create's instability in the network, as people are getting kicked off the list because of the small amount in the node.

So to create some stability, long term nodes that are well funded get a higher rating and probably more SPR block rewards.

That's my understanding anyway

But then it would remove the competition if they were paid more, and hence defeating the purpose of a limit.

Their not paid more because that's in the code. There paid more because they're more consistent, handle more transactions and respond 100% of the time.

Also the larger your servicenode balance is, the larger transactions your node can handle. And we need the ability to handle large transactions so we need large wallets.

Am i right in saying Dash nodes can only handle 1000 dash at a time?

No, you are wrong. It doesn't mix any dash with its own, the dash always stays in the masternode so it makes sense.
So if 1 Dash masternodes were possible, they would be able to handle 20 999 999 Dash.

[stonehedge]

So, will masternode payments still depend on a score, because if so it will defeat the purpose of them being dynamic.

I must admit that was something I didn't care for with the original SPR masternode testing.  I didn't understand the benefits of the highest scoring masternodes getting paid.  Am I missing a really obvious benefit?  Surely the network can only be fair if all masternodes get paid roughly the same amount over a period of time?

Why define a max number of masternodes?  Surely the ROI, coin supply and value of SPR will define this organically?  People won't run masternodes at a loss for long...

I also remain to be convinced of the dynamic masternode model other than its brilliant ease of setup.  Once again...am I missing a fundamental advantage?

The score was designed to sort the men from the boys.

Servicenode owners will try to load their nodes with as little SPR as possable (so to have move nodes/profit).
This create's instability in the network, as people are getting kicked off the list because of the small amount in the node.

So to create some stability, long term nodes that are well funded get a higher rating and probably more SPR block rewards.

That's my understanding anyway

But then it would remove the competition if they were paid more, and hence defeating the purpose of a limit.

Their not paid more because that's in the code. There paid more because they're more consistent, handle more transactions and respond 100% of the time.

Also the larger your servicenode balance is, the larger transactions your node can handle. And we need the ability to handle large transactions so we need large wallets.

Am i right in saying Dash nodes can only handle 1000 dash at a time?

Ok, so in this hypothesis the people with the bigger SPR wallets with VPS hosted masternodes would receive more payments than somebody with one masternode hosted on their macbook.  Doesn't seem right somehow.  Anyway...new ANN today.  Plenty of time to thrash out semantics and test various hypotheses!

[defunctec]

So, will masternode payments still depend on a score, because if so it will defeat the purpose of them being dynamic.

I must admit that was something I didn't care for with the original SPR masternode testing.  I didn't understand the benefits of the highest scoring masternodes getting paid.  Am I missing a really obvious benefit?  Surely the network can only be fair if all masternodes get paid roughly the same amount over a period of time?

Why define a max number of masternodes?  Surely the ROI, coin supply and value of SPR will define this organically?  People won't run masternodes at a loss for long...

I also remain to be convinced of the dynamic masternode model other than its brilliant ease of setup.  Once again...am I missing a fundamental advantage?

The score was designed to sort the men from the boys.

Servicenode owners will try to load their nodes with as little SPR as possable (so to have move nodes/profit).
This create's instability in the network, as people are getting kicked off the list because of the small amount in the node.

So to create some stability, long term nodes that are well funded get a higher rating and probably more SPR block rewards.

That's my understanding anyway

But then it would remove the competition if they were paid more, and hence defeating the purpose of a limit.

Their not paid more because that's in the code. There paid more because they're more consistent, handle more transactions and respond 100% of the time.

Also the larger your servicenode balance is, the larger transactions your node can handle. And we need the ability to handle large transactions so we need large wallets.

Am i right in saying Dash nodes can only handle 1000 dash at a time?

Ok, so in this hypothesis the people with the bigger SPR wallets with VPS hosted masternodes would receive more payments than somebody with one masternode hosted on their macbook.  Doesn't seem right somehow.  Anyway...new ANN today.  Plenty of time to thrash out semantics and test various hypotheses!

Not necessarily, only if you assume a macbook can't store large amounts of SPR and internet wouldn't be reliable.

If you have a good internet connection (business connection) and a few SPR, i don't see why you couldn't compete with the node you describe.

[e1ghtSpace]

Did you guys miss where I said that a smartnode can handle more spr than it holds? (same for dash's masternodes)

[defunctec]

Did you guys miss where I said that a smartnode can handle more spr than it holds? (same for dash's masternodes)

Fair enough.

So what do you suggest, get rid of "more spr = higher score"?

[chrysophylax]

Did you guys miss where I said that a smartnode can handle more spr than it holds? (same for dash's masternodes)

Fair enough.

So what do you suggest, get rid of "more spr = higher score"?

this seems to me like the existing banking situation ( yes - i called it a situation ) ...

the more you have the higher the score the better the chance of gaining more ...

the rich get richer - the poor still struggle ... but why wouldnt you reward those that put the effort AND spr in? ...

touch one that ...

#crysx

[stonehedge]

Did you guys miss where I said that a smartnode can handle more spr than it holds? (same for dash's masternodes)

Fair enough.

So what do you suggest, get rid of "more spr = higher score"?

this seems to me like the existing banking situation ( yes - i called it a situation ) ...

the more you have the higher the score the better the chance of gaining more ...

the rich get richer - the poor still struggle ... but why wouldnt you reward those that put the effort AND spr in? ...

touch one that ...

#crysx

I agree that the hypothetical macbook with 24/7 internet connection that never goes to sleep will be as "reliable" as a cloud hosted node, fair point.

I just don't like the idea that a node with a higher balance being scored higher than a node with a lower balance.  If a node has a lot of downtime, then wouldn't it probably get paid less anyway?

EDIT:  Lets take the top 10 wallets as an example...surely they have the most to gain from a balance influenced score?

[coins101]

snip-

i think that the fact that we dont even know what to call 'them' shows the state of affairs that spreadcoin is in currently ... masternodes - spreadnodes - servicesnodes ...
snip-
#crysx

There, fixed it for you.

Now you're on the same page as everyone else  

[chrysophylax]

snip-

i think that the fact that we dont even know what to call 'them' shows the state of affairs that spreadcoin is in currently ... masternodes - spreadnodes - servicesnodes ...
snip-
#crysx

There, fixed it for you.

Now you're on the same page as everyone else  

hahaha ... i even spelt that wrong ...

servicenodes ...

ok - same page now ...

#crysx

[gjhiggins]

some service nodes will generate less income for doing the same amount of work at the same expense

That's a creditable starting point. Before we kick off ...

I dunno if everyone's cued in to the fact that the discussion is actually about the management of an overlay network “An overlay network is a computer network, which is built on the top of another network” The idea's been around since forever, there's a literature corpus to be mined for algorithmic insights and a broad appreciation of a range of issues, e.g. Stanford's work on RON (Resilient Overlay Networks), now 15 years old. For a more recent overview of the domain, see “Overlay Networks: Overview, Applications and Challenges” by Galán-Jiménez and Gazo-Cervero  in IJCSNS International Journal of Computer Science and Network Security, VOL.10 No.12, December 2010, full paper (PDF): http://paper.ijcsns.org/07_book/201012/20101206.pdf

Another pertinent domain is MANET (Mobile Ad-hoc NETworks) - https://en.wikipedia.org/wiki/Mobile_ad_hoc_network for which threshold cryptography is also a key component.

Lastly, looking into the near future (at least 6 months) there's the exotica that may become mundanely relevant as the overlay network application space expands, e.g. Raftopoulou, Petrakis and Tryfonopoulos’ “Rewiring Strategies for Semantic Overlay Networks”:

Semantic overlay networks cluster peers that are semantically, thematically or socially close into groups, by means of a rewiring procedure that is periodically executed by each peer. This procedure establishes new connections to similar peers and disregards connections to peers that are dissimilar. Retrieval effectiveness is then improved by exploiting this information at query time (as queries may address clusters of similar peers).

http://users.uop.gr/~praftop/papers/pdf/dapd09-RPT.pdf <- DAPD = “Distributed and Parallel Databases”

Back to “fair” - inverting, to negate the negation and make explicit the assumption, “the same task”:

the same task, the same income,  the same amount of work, the same expense == fair

Have to rule out “the same expense” because it's incalculable in the large: it varies according local energy market conditions.

Have to rule out “the same amount of work” because it too is incalculable in the large because it varies with machine cost/performance.

So we're left with the only thing we can actually calculate: “the same task”, “the same amount of income” == fair.

That's quite egalitarian in principle and, to a degree, viable precisely because it acknowledges that variations in local conditions prevent real costs being compared, so in practice the coin is reduced to necessarily adopting a policy of simplification of modelling the cost of membership of the overlay network.

But I think that's where Mr Spread's approach has the edge. By presenting it as an auction, it enables participants' bids to reflect the locally-dependent calculations we cannot make in the large. According to our necessarily narrow definition, it cannot actually be described as less fair/egalitarian but it will certainly be more sustainable in the long term because it better reflects the actual cost of maintaining the overlay network.

Just to muddy the waters further, atm I don't see why there should always be one and only one logical overlay network. I can foresee different logical overlay networks supporting different applications offering different cost/benefit tradeoffs and levels of income, so I'd like to see the overlay network membership cost eventually pushed into the specific overlay application rather than resting on a relatively coarse and insensitive assumption that it's a single amount applicable to all overlays irrespective of differences in the complexity/magnitude of the service provided.

IIRC (I did read it very quickly) there's some reference to characterisations of differences in node capability / connectivity in the overview ref (above) which will be grist to practical discussions of these overlay network parameters.

My sense (as essentially, an outsider) is that Spreadcoin can't afford to be perceived as endorsing/enabling the kind of cynical insider profiteering that Crave seems to have attracted:

it is still good money to make watching tv and picking your nose.

That's really going to impress business clients seeking to make use of Spreadcoin's overlay network applications.

Cheers

Graham

[stonehedge]

Great post.  I'll reply in a few hours when I've had time to digest it!

[chrysophylax]

some service nodes will generate less income for doing the same amount of work at the same expense

That's a creditable starting point. Before we kick off ...

I dunno if everyone's cued in to the fact that the discussion is actually about the management of an overlay network “An overlay network is a computer network, which is built on the top of another network” The idea's been around since forever, there's a literature corpus to be mined for algorithmic insights and a broad appreciation of a range of issues, e.g. Stanford's work on RON (Resilient Overlay Networks), now 15 years old. For a more recent overview of the domain, see “Overlay Networks: Overview, Applications and Challenges” by Galán-Jiménez and Gazo-Cervero  in IJCSNS International Journal of Computer Science and Network Security, VOL.10 No.12, December 2010, full paper (PDF): http://paper.ijcsns.org/07_book/201012/20101206.pdf

Another pertinent domain is MANET (Mobile Ad-hoc NETworks) - https://en.wikipedia.org/wiki/Mobile_ad_hoc_network for which threshold cryptography is also a key component.

Lastly, looking into the near future (at least 6 months) there's the exotica that may become mundanely relevant as the overlay network application space expands, e.g. Raftopoulou, Petrakis and Tryfonopoulos’ “Rewiring Strategies for Semantic Overlay Networks”:

Semantic overlay networks cluster peers that are semantically, thematically or socially close into groups, by means of a rewiring procedure that is periodically executed by each peer. This procedure establishes new connections to similar peers and disregards connections to peers that are dissimilar. Retrieval effectiveness is then improved by exploiting this information at query time (as queries may address clusters of similar peers).

http://users.uop.gr/~praftop/papers/pdf/dapd09-RPT.pdf <- DAPD = “Distributed and Parallel Databases”

Back to “fair” - inverting, to negate the negation and make explicit the assumption, “the same task”:

the same task, the same income,  the same amount of work, the same expense == fair

Have to rule out “the same expense” because it's incalculable in the large: it varies according local energy market conditions.

Have to rule out “the same amount of work” because it too is incalculable in the large because it varies with machine cost/performance.

So we're left with the only thing we can actually calculate: “the same task”, “the same amount of income” == fair.

That's quite egalitarian in principle and, to a degree, viable precisely because it acknowledges that variations in local conditions prevent real costs being compared, so in practice the coin is reduced to necessarily adopting a policy of simplification of modelling the cost of membership of the overlay network.

But I think that's where Mr Spread's approach has the edge. By presenting it as an auction, it enables participants' bids to reflect the locally-dependent calculations we cannot make in the large. According to our necessarily narrow definition, it cannot actually be described as less fair/egalitarian but it will certainly be more sustainable in the long term because it better reflects the actual cost of maintaining the overlay network.

Just to muddy the waters further, atm I don't see why there should always be one and only one logical overlay network. I can foresee different logical overlay networks supporting different applications offering different cost/benefit tradeoffs and levels of income, so I'd like to see the overlay network membership cost eventually pushed into the specific overlay application rather than resting on a relatively coarse and insensitive assumption that it's a single amount applicable to all overlays irrespective of differences in the complexity/magnitude of the service provided.

IIRC (I did read it very quickly) there's some reference to characterisations of differences in node capability / connectivity in the overview ref (above) which will be grist to practical discussions of these overlay network parameters.

My sense (as essentially, an outsider) is that Spreadcoin can't afford to be perceived as endorsing/enabling the kind of cynical insider profiteering that Crave seems to have attracted:

it is still good money to make watching tv and picking your nose.

That's really going to impress business clients seeking to make use of Spreadcoin's overlay network applications.

Cheers

Graham

graham ... eloquence is obviously and evidently your strong point ...

pity it was never mine ...

#crysx

[gjhiggins]

graham ... eloquence is obviously and evidently your strong point ...

pity it was never mine ...

Thanks for the kind words. Does it help if I reveal that i) it can take me a couple of days to research and couple of hours to write a post like that and ii) I probably get the most benefit out of the exercise?

Cheers

Graham

[coins101]

Did you guys miss where I said that a smartnode can handle more spr than it holds? (same for dash's masternodes)

No.

Nor did we miss your point about hackers facing the risk that they could have their stolen coins taken back by the legitimate owners. Which is a great insight.

[stonehedge]

graham ... eloquence is obviously and evidently your strong point ...

pity it was never mine ...

Thanks for the kind words. Does it help if I reveal that i) it can take me a couple of days to research and couple of hours to write a post like that and ii) I probably get the most benefit out of the exercise?

Cheers

Graham

+1 Its all too easy to write rapid responses on internet forums without really researching and thinking about what you want to say.  And as you mention, it benefits yourself as well as the community.  I shall try to learn from your example.