Author Topic: Features for a new version (round 3)  (Read 10178 times)

MyFarm

  • Sr. Member
  • ****
  • Posts: 259
    • View Profile
Features for a new version (round 3)
« on: February 20, 2015, 10:13:33 pm »
Someone correct me if I'm wrong, but based upon what I'm seeing at github, Mr. Spread left us with the sped up version of the election process.

Would anyone like to compile the builds, tweak it for round 3, and let's test it?

I have a theory on what Mr. Spread has done and if I am correct, what he left us is going to work beautifully. 

Strumpet!

  • Full Member
  • ***
  • Posts: 158
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #1 on: February 20, 2015, 10:24:24 pm »
https://github.com/spreadcoin-project/spreadcoin/blob/mn-test/src/masternodes.h   -   the current testnet paramaters are all in here - 10 votes per block, MN reward % etc.

I have no idea if just tweaking these is enough, but I suspect it wont be that easy. I don't speak C++. Maybe elbandi or gjhiggins could help out?

Basically we need an experienced dev. If I had the skillset I would take Spread and, well, royally annoy georgem for a start...  ;D

MyFarm

  • Sr. Member
  • ****
  • Posts: 259
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #2 on: February 20, 2015, 10:27:24 pm »
Actually, I was mistaken.  I was looking at elbandi's patch.  I'm an idiot.

georgem

  • Tech Admin
  • ******
  • Posts: 880
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #3 on: February 20, 2015, 10:33:21 pm »
Someone correct me if I'm wrong, but based upon what I'm seeing at github, Mr. Spread left us with the sped up version of the election process.

Would anyone like to compile the builds, tweak it for round 3, and let's test it?

I have a theory on what Mr. Spread has done and if I am correct, what he left us is going to work beautifully.

I like that.  8)

Please, if somebody thinks he's up to it, I will certainly help test it, and give my feedback.

BTW, I am working on my own fork, based on the feedback I have given in the last 2 weeks.
(slowly increasing max amount of MNs (total coins / 2880), plus a more random vote distribution that is not tied to the amount of SPR in your MN)


I would absolutely love it if we just keep on testing and tweaking different forks, this way we will all get more familiar with the internal workings of spreadcoin.

I feel we have a handful of very enthusiastic and almost ready programmers among us.




georgem

  • Tech Admin
  • ******
  • Posts: 880
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #4 on: February 20, 2015, 10:34:57 pm »
https://github.com/spreadcoin-project/spreadcoin/blob/mn-test/src/masternodes.h   -   the current testnet paramaters are all in here - 10 votes per block, MN reward % etc.

I have no idea if just tweaking these is enough, but I suspect it wont be that easy. I don't speak C++. Maybe elbandi or gjhiggins could help out?

Basically we need an experienced dev. If I had the skillset I would take Spread and, well, royally annoy georgem for a start...  ;D

Please look into it, and try to figure it out!  :)
If there ever was a chance to get into coin programming, here it is, right in front of you.

And you have a willing community ready to help you with the testing.

Maybe that's what we need... the most decentralized way of doing things.

 8)

elbandi

  • Jr. Member
  • **
  • Posts: 57
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #5 on: February 21, 2015, 12:34:50 pm »
The main goal for round3 test is to find the good value for masternodes (vote, election, etc).

For this we need to test a few times with different value to find the good values.
But while the community "build" the testnet, it is very slow. (1-1.5 week for 1440 nodes).

So we need a better solution, like: my masternode eat ~100M memory with 7 mnstart line. so if 10 mnstart line eat 100M

memory, 1440 "masternodes" running in 144 daemon requred ~14-16G total memory .

docker.io is a good technology to start separate processes in one system. so we can easily test the values on a "small"

environment. And this small environment has some advantage: easy and fast to setup the 1440 nodes, easy to reset the blockchain and test other mn values.

njs811

  • Jr. Member
  • **
  • Posts: 88
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #6 on: February 21, 2015, 05:56:29 pm »
HOLY SHIT this might be the most decentralized coin in existence. The code is all there. Perhaps we should make a closed forum specifically to teach each other what we know. Collectively we may have more than enough.

georgem

  • Tech Admin
  • ******
  • Posts: 880
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #7 on: February 21, 2015, 06:04:53 pm »
HOLY SHIT this might be the most decentralized coin in existence. The code is all there. Perhaps we should make a closed forum specifically to teach each other what we know. Collectively we may have more than enough.

Cool idea, but why closed?

I say spread the knowledge.  8)

njs811

  • Jr. Member
  • **
  • Posts: 88
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #8 on: February 21, 2015, 06:10:31 pm »
Only to discourage fud. Almost like a developers corner

Strumpet!

  • Full Member
  • ***
  • Posts: 158
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #9 on: February 21, 2015, 06:29:27 pm »
The main goal for round3 test is to find the good value for masternodes (vote, election, etc).

For this we need to test a few times with different value to find the good values.
But while the community "build" the testnet, it is very slow. (1-1.5 week for 1440 nodes).

So we need a better solution, like: my masternode eat ~100M memory with 7 mnstart line. so if 10 mnstart line eat 100M

memory, 1440 "masternodes" running in 144 daemon requred ~14-16G total memory .

docker.io is a good technology to start separate processes in one system. so we can easily test the values on a "small"

environment. And this small environment has some advantage: easy and fast to setup the 1440 nodes, easy to reset the blockchain and test other mn values.

If you know how to do this then go for it! There's no real show stopper with the voting paramaters as they are, it's just slow. If more votes per block speed the process up that's all we really need at this point.

If future development by the community occurs though I'd really like to abandon the dynamic pricing. Imagine there were only 1440 mining slots, and the richest miners with the most hash could come along and kick out the smaller miners...

Dynamic pricing was a terrible idea from the start. Call it 1000 SPR and have done with it. Simpler code and the small guys don't get screwed.

georgem

  • Tech Admin
  • ******
  • Posts: 880
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #10 on: February 21, 2015, 06:38:33 pm »
The main goal for round3 test is to find the good value for masternodes (vote, election, etc).

For this we need to test a few times with different value to find the good values.
But while the community "build" the testnet, it is very slow. (1-1.5 week for 1440 nodes).

So we need a better solution, like: my masternode eat ~100M memory with 7 mnstart line. so if 10 mnstart line eat 100M

memory, 1440 "masternodes" running in 144 daemon requred ~14-16G total memory .

docker.io is a good technology to start separate processes in one system. so we can easily test the values on a "small"

environment. And this small environment has some advantage: easy and fast to setup the 1440 nodes, easy to reset the blockchain and test other mn values.

If you know how to do this then go for it! There's no real show stopper with the voting paramaters as they are, it's just slow. If more votes per block speed the process up that's all we really need at this point.

If future development by the community occurs though I'd really like to abandon the dynamic pricing. Imagine there were only 1440 mining slots, and the richest miners with the most hash could come along and kick out the smaller miners...

Dynamic pricing was a terrible idea from the start. Call it 1000 SPR and have done with it. Simpler code and the small guys don't get screwed.

I disagree but I will not stop you or elbandi from creating a fork with fix 1000 SPR requirement, and I will even help you with the testing.
In return, I ask only that you will help me test my own fork in a few days, when it's ready.

Deal?

We could even test the forks simultaneously by agreeing to use different ports and data folders, so they can run side by side.

Everybody, spreadcoin is far from dead, and I want to invite everybody here to just continue following his dream.

Is this a crisis or rather an opportunity?  8)

georgem

  • Tech Admin
  • ******
  • Posts: 880
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #11 on: February 21, 2015, 06:42:34 pm »
Also,

We all agree that the testing phase we are in now, will have to be extended considerably because we will not be ready in a few days or weeks as originally planned,
but probably in a few months.


That's why I say: Yes, go on, test your fork, adjust your parameters, and let's see what happens.
I will prepare my own fork.

We shall then compare results and let the community vote which way to continue.  8)

Strumpet!

  • Full Member
  • ***
  • Posts: 158
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #12 on: February 21, 2015, 06:43:39 pm »
georgem, have you given some honest thought to the sentence I bolded? There's no difference between letting the richest MN ops kick out the poorer ones, and letting the richest miners kick out the poorer ones. None. But if anyone suggested that only the richest 1440 miners could mine the coin there would be outrage.

What advantage to the network does rigging the setup to favour the rich and exclude the poor bring?

njs811

  • Jr. Member
  • **
  • Posts: 88
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #13 on: February 21, 2015, 06:46:15 pm »
I think voting will take to much time. This is basically a new coin with a firm foundation. Are there any decisions we could make together? Ww can debate and go forward with one mindset

georgem

  • Tech Admin
  • ******
  • Posts: 880
    • View Profile
Re: Masternodes Testing, Round 3
« Reply #14 on: February 21, 2015, 06:48:37 pm »
georgem, have you given some honest thought to the sentence I bolded? There's no difference between letting the richest MN ops kick out the poorer ones, and letting the richest miners kick out the poorer ones. None.

What advantage to the network does rigging the setup to favour the rich and exclude the poor bring?

Yes, you keep coming back with this.
I think this has been debunked 1 month ago.
The rich can't control the market if the protocol makes it easy for newcomers to enter the competition.

Mr. Spread has completely disabled the competition by giving high tSPR amount MNs voting privileges.

That is a farce. And I want to prove it to you. But not with words.
Stay tuned.