elective-stereophonic
elective-stereophonic
Phase 1 results / Phase 2 start singapore
Please login or register.

Login with username, password and session length
Advanced search  

News:

Latest Nxt Client: Nxt 1.11.15

Author Topic: Phase 1 results / Phase 2 start  (Read 1923 times)

kushti

  • Board Moderator
  • Sr. Member
  • ****
  • Karma: +184/-5
  • Offline Offline
  • Posts: 384
  • Nxt Core & Apps Dev
    • View Profile
Phase 1 results / Phase 2 start
« on: December 15, 2014, 11:49:07 pm »

A lot of things made since the first fundraising letter:

- Executable forging simulation tool for a proof-of-stake with three blogposts about ( http://chepurnoy.org/ )
- Executable multibranch forging simulation tool with the paper about it

All those papers, articles, and both tools could be found on GitHub: https://github.com/ConsensusResearch

- And the final result of the phase - our paper on long-range attacks and N@S is already sent to investors & active Nxt contributors and will be publicly released this week.

So we made first milestones (https://nxtforum.org/consensus-research/first-fundraising-letter/) for the price of one! And we're starting phase2 now!


Well, after playing with multibranch forging we think it's better not to punish it(slasher etc) but use it officially. However it leads to the consensus with explicit multiple states(we're calling it "quantum consensus"), so instead statements like "i got 5K 2 blocks ago" more flexible "i got 5K recently with 99.99% probability(becoming 100% after >=n confirmations for k most powerful branches)" statement will be used (in fact we have it now implicitly with all cryptos, mostly with proof-of-stake though). And there are many concerns & problems around to solve. So we think to get deeper into this after New Year holiday.

And next two or three weeks we'll spend on experiments with blockchain quality measure. Cumulative difficulty being using currently seems to be not a good option as in multibranch environment all branches are likely to be very close so there's literally no significant difference between bad and good branches. In single-branch environment(Nxt has now) that means simplicity of blockchain to be overtaken by opponents. So better measure matters in any case. We have some proposals how to improve the measure but investigations needed.  Also we'll spend a lot of time on discussions around N@S paper probably  ;D  and I have a plan for a bunch of easy-to-read-articles about different aspects of consensus algos we're playing with.  Planned budget for phase2 is 200,000 (for two people).

Logged
for donations / messages: NXT-PKXM-WH25-UXXG-CJAVD (alias: kushti)

LiQio

  • Hero Member
  • *****
  • Karma: +50/-5
  • Offline Offline
  • Posts: 672
    • View Profile
    • NxtLoader for Windows
Re: Phase 1 results / Phase 2 start
« Reply #1 on: December 16, 2014, 05:13:57 am »

Sounds cool, thanks again :)

Daedelus

  • Hero Member
  • *****
  • Karma: +230/-12
  • Offline Offline
  • Posts: 3280
    • View Profile
Re: Phase 1 results / Phase 2 start
« Reply #2 on: December 16, 2014, 09:13:50 am »

Also we'll spend a lot of time on discussions around N@S paper probably  ;D  and I have a plan for a bunch of easy-to-read-articles about different aspects of consensus algos we're playing with. 

whoop!
Logged
NXT: NXT-4CS7-S4N5-PTH5-A8R2Q

_mr_e

  • Hero Member
  • *****
  • Karma: +88/-18
  • Offline Offline
  • Posts: 956
    • View Profile
Re: Phase 1 results / Phase 2 start
« Reply #3 on: December 16, 2014, 12:42:01 pm »

A lot of things made since the first fundraising letter:

- Executable forging simulation tool for a proof-of-stake with three blogposts about ( http://chepurnoy.org/ )
- Executable multibranch forging simulation tool with the paper about it

All those papers, articles, and both tools could be found on GitHub: https://github.com/ConsensusResearch

- And the final result of the phase - our paper on long-range attacks and N@S is already sent to investors & active Nxt contributors and will be publicly released this week.

So we made first milestones (https://nxtforum.org/consensus-research/first-fundraising-letter/) for the price of one! And we're starting phase2 now!


Well, after playing with multibranch forging we think it's better not to punish it(slasher etc) but use it officially. However it leads to the consensus with explicit multiple states(we're calling it "quantum consensus"), so instead statements like "i got 5K 2 blocks ago" more flexible "i got 5K recently with 99.99% probability(becoming 100% after >=n confirmations for k most powerful branches)" statement will be used (in fact we have it now implicitly with all cryptos, mostly with proof-of-stake though). And there are many concerns & problems around to solve. So we think to get deeper into this after New Year holiday.

And next two or three weeks we'll spend on experiments with blockchain quality measure. Cumulative difficulty being using currently seems to be not a good option as in multibranch environment all branches are likely to be very close so there's literally no significant difference between bad and good branches. In single-branch environment(Nxt has now) that means simplicity of blockchain to be overtaken by opponents. So better measure matters in any case. We have some proposals how to improve the measure but investigations needed.  Also we'll spend a lot of time on discussions around N@S paper probably  ;D  and I have a plan for a bunch of easy-to-read-articles about different aspects of consensus algos we're playing with.  Planned budget for phase2 is 200,000 (for two people).
Am I understanding right that you are saying there are many problems with Nxt's pos after all??
Logged

kushti

  • Board Moderator
  • Sr. Member
  • ****
  • Karma: +184/-5
  • Offline Offline
  • Posts: 384
  • Nxt Core & Apps Dev
    • View Profile
Re: Phase 1 results / Phase 2 start
« Reply #4 on: December 17, 2014, 08:56:53 am »

Am I understanding right that you are saying there are many problems with Nxt's pos after all??

Ah, _mr_e, you want to hear about problems with pos after all again. So, the moment is not after all, we're just at the beginning  :) And problems... the paper will be published tomorrow with tl'dr comments, so let's defer this talk :)

P.S. holding my small bunch of Nxt :)
Logged
for donations / messages: NXT-PKXM-WH25-UXXG-CJAVD (alias: kushti)

kushti

  • Board Moderator
  • Sr. Member
  • ****
  • Karma: +184/-5
  • Offline Offline
  • Posts: 384
  • Nxt Core & Apps Dev
    • View Profile
Re: Phase 1 results / Phase 2 start
« Reply #5 on: December 19, 2014, 07:26:35 am »

N@S paper was published yesterday https://nxtforum.org/consensus-research/the-paper-on-long-range-attack-nothing-at-stake/

Seems we need to update our site as we unfortunately have a lot of information dispersed through blogs and forums :(
Logged
for donations / messages: NXT-PKXM-WH25-UXXG-CJAVD (alias: kushti)
 

elective-stereophonic
elective-stereophonic
assembly
assembly