Nxt Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Latest Nxt Client 1.11.10 - NEW RELEASE: Ardor 2.0.5e TestNet - The Ignis ICO is over!! Ardor genesis snapshots will happen at Nxt block 1,630,000 (expected for 25th December)

Pages: [1]

Author Topic: funding request to implement mthcl's forging - 1 million NXT  (Read 1875 times)

jl777

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6176
    • View Profile
  • Karma: +718/-123

http://www.docdroid.net/cckd/forging0-4-3.pdf.html describes some minor changes to the current NXT forging
ChuckOne says what is needed to fully implement this is mostly infrastructure

I do not fully understand all that is needed, but assuming that we can get the paper peer reviewed and it passes and we can clearly define the infrastructure needed to properly implement this, it will solve one of NXT's last barrier's to mass adoption. Especially by large investor types.

Maybe the 1 million NXT is way too much, but BCNext increased the unclaimed funds from 9 million to 10 million NXT and said the community should take care of infrastructure, so to me that means we should allocate 1 million NXT for this and use it for the long term to make sure NXT always has the infrastructure needed

I dont have time to deal with application forms and figuring out all the details, but I want to find someone that believe this is important and has the time to manage it.

James

P.S. Once we get this in place I believe we are one step closer to having infrastructure that can support TF.
There are over 1000 people in SuperNET slack! http://slackinvite.supernet.org/ automatically sends you an invite

I am just a simple C programmer

TeamWealth

  • Full Member
  • ***
  • Offline Offline
  • Posts: 217
    • View Profile
  • Karma: +11/-1

Skimming over it, where did the 1 million number come from? I don't see any mention of what it would take to implement it so is the 1 million a guess based on the development time required for it theoretically?
NXT: NXT-V93N-SYX2-2CNW-5TF9Y

jl777

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 6176
    • View Profile
  • Karma: +718/-123

Skimming over it, where did the 1 million number come from? I don't see any mention of what it would take to implement it so is the 1 million a guess based on the development time required for it theoretically?
9million to 10 million = 1 million
The unclaimed funds used to be 9 million, then BCnext said that community was supposed to take care of infrastructure costs and increased the 9 million to 10 million

just following instructions from above
There are over 1000 people in SuperNET slack! http://slackinvite.supernet.org/ automatically sends you an invite

I am just a simple C programmer

EvilDave

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1791
    • View Profile
    • NXT Foundation
  • Karma: +341/-40

Hmmm...i got to page 3 of the .pdf and my brain shut down.

Can we have a tl:dr outline for the non-mathematical among us ?

Nulli Dei, nulli Reges, solum NXT
NXT Donations: NXT-BNZB-9V8M-XRPW-3S3WD
We will ride eternal, shiny and chrome!

notsoshifty

  • Newbie
  • *
  • Offline Offline
  • Posts: 3
    • View Profile
  • Karma: +0/-0

This is going back a bit, sorry. But it's still interesting.

The paper describes several recommendations. For which one(s) specifically are you proposing a bounty?

1. The main one is a change from U-algorithm to Exp-algorithm, which would mean the probability of forging a block is truly proportional to the effective balance (as opposed to roughly proportional, as at present).
2. It also suggests a minimum forging balance (e.g. below 100 NXT cannot forge)
3. ... and a maximum effective forging balance (e.g. anything over 1 million NXT stakes as if it were 1 million NXT). This allows you to keep the simplicity of the U-algorithm but still have probability of forging being very close to being proportional to (capped) effective balance.
4. Lastly, it suggests a couple of possible mechanisms for introducing real randomness into the forging process. This would seem to limit the possibility of hijacking the blockchain to allow someone to forge every single block, which it notes to be possible for anyone controlling 50% of all stake.

I'm not sure what any of that has to do with infrastructure, though. Isn't it all core code?

mthcl

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 561
    • View Profile
  • Karma: +95/-8

Kind of overlooked this topic at that time. Well, as you know, I don't code, so I can only write theoretical stuff and hope the developers would be interested to implement it. Also, in any case, it would be very interesting for me to have more feedback on the TF scheme proposed there, even from a purely academical point of view. Here is the link to the last version, by the way: http://www.docdroid.net/ecmz/forging0-5-2.pdf.html

On a side note: if the TF won't be implemented within reasonable time frame, I think it would still be a good idea to modify the current Blind Shooter algorithm (https://nxtforum.org/proof-of-stake-algorithm/basetarget-adjustment-algorithm), to get rid of those occasional long block times. It is very annoying when it happens when you're buying/selling on the Asset Exchange...

abctc

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1353
    • View Profile
  • Karma: +147/-13

.. if the TF won't be implemented within reasonable time frame, I think it would still be a good idea to modify the current Blind Shooter algorithm (https://nxtforum.org/proof-of-stake-algorithm/basetarget-adjustment-algorithm), to get rid of those occasional long block times. It is very annoying when it happens when you're buying/selling on the Asset Exchange...
- agree 100%

But here is the Jean-Luc's answer:
Jean-Luc, please consider implement this:
Quote
with this adjustment algorithm it is much more difficult to get these annoying long blocktimes.
So, if there is any hardfork coming before the implementation of TF, I would strongly recommend to switch to the algorithm with γ=1/2.  And there is still some possibility we'd need this algorithm for TF as well.

Absolutely no time for that before DGS.

What are the actual changes to the java code needed? If somebody puts the algorithm in code, we can at least try it on testnet after DGS, but I don't have time to work on this myself.
Welcome to the Nxt generation of crypto!   Magis quam Moneta (More than a Coin)
"Do not worry, it is an attack" (c) Jean-Luc

mthcl

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 561
    • View Profile
  • Karma: +95/-8

As far as I understand, the changes in the code are really minimal. It's just changing γ=1 to γ=1/2, as described in the 1st post in that topic. And also put an upper limit on the difficulty, since there is no point at all to allow it to increase arbitrarily (unlike the situation with BTC mining, we cannot have more than 1000000000NXT forging).
Pages: [1]