NXT :: descendant of Bitcoin - Updated Information

[ChuckOne]

Something, I would like to add to the discussion of finding consensus.

What makes the difference between a malicious 90% forger and the rest of the world is that these forger simply CAN exclude or include transactions at will as could do any other node.

Problem with analyzing included transactions is: we cannot find consensus by looking at them because we have yet to agree on them.

The only feasible way (if there is one) is analyzing the block graph.

I hope that make the problem more clear.

[opticalcarrier]

or could we revisit the idea about the possibiliy of a TF algorithm that would support 2 or 3 different accounts that could forge the next block?  and possibly make it such that if an account forged a block then the algorithm would reduce its potential for the next block?  just throwing out ideas. 

[ChuckOne]

I was think about the 1-block-per-minute regime:

The network has 1 minute to agree on a block.

Let the n first accounts in the forging queue create a block on their own and distribute it to the network.

Each node will decide independently which is the best block for it (depending on the network latency). That is, the block with the first forger in the queue that created a block is accepted.

After that minute, a new set of n blocks are created as it was for the first minute.

If now the malicious forger sends his block from the first minute to the network, it's over for him. Nobody will accept it anymore as the second minute has already started.


*EDIT*: forging queue: for each block there is a queue of ALL accounts. The first one is the best one, the second one is the second-best and so on. If a node receives a block, it decided whether that block is within that minute AND whether it's better than nothing or the other blocks it has already received. This queue changes with every block pseudo-randomly.

[marcus03]

Can a Mac dev help me out?

How would I enable the user to start my client with a command line parameter on OSX?

On Windows, if my client is started with "NxtSolarisFM.exe testnet" it will connect the NXT testnet and use different files.

For OSX I distribute a zip file with a "NXTSolaris.app" folder in it. Could I add a script or something that would make it easy (like clicking on it...) for the Mac user to run it with the "testnet" parameter?

[Come-from-Beyond]

Hmm... I am not quite clear on what you are referring to as being the "latest target" - if this is something to do with "difficulty" then my understanding from discussions with CfB about TF is that it won't even be necessary to have that (I am not even quite sure why "difficulty" was even used at all - perhaps someone would care to enlighten me?).

Difficulty defines bias toward bigger accounts. Set it lower and 10M account will forge 1.1% of all blocks. Set it higher and the account will forge only 1.05%.

[CIYAM]

Difficulty defines bias toward bigger accounts. Set it lower and 10M account will forge 1.1% of all blocks. Set it higher and the account will forge only 1.05%.

So if we limited the forging power of a single account to x then presumably the "difficulty" would become practically irrelevant.

[CIYAM]

or could we revisit the idea about the possibiliy of a TF algorithm that would support 2 or 3 different accounts that could forge the next block?  and possibly make it such that if an account forged a block then the algorithm would reduce its potential for the next block?  just throwing out ideas.  

Effectively that is the original TF idea from BCNext (AFAIK) - 2 or 3 (or 4?) accounts would get *equal* ability to forge the next block and those that *fail* are *penalised* for not doing so (am not quite sure if the penalty is intended just for those that fail to broadcast or what and maybe this is why I am not convinced about the *penalty* thing so far).

[Uniqueorn]

Difficulty defines bias toward bigger accounts. Set it lower and 10M account will forge 1.1% of all blocks. Set it higher and the account will forge only 1.05%.

So if we limited the forging power of a single account to x then presumably the "difficulty" would become practically irrelevant.

couldn't the account just spread the NXT over more accounts and open more clients to forge?

[CIYAM]

couldn't the account just spread the NXT over more accounts and open more clients to forge?

Am only referring to the "difficulty" question (not overall power to forge - they can divide up their NXT into multiple accounts right now).

[Come-from-Beyond]

So if we limited the forging power of a single account to x then presumably the "difficulty" would become practically irrelevant.

BCNext intentionally used an algo that creates a small bias, otherwise he would use plain lottery for forging...

[EmoneyRu]

BCNext I intentionally used an algo that creates a small bias, otherwise he I would use plain lottery for forging...

Fixed

[antanst]

oh, please:


import numpy as np

import sys
import string
 

def main():
    selChars = string.ascii_letters+ string.digits
    print( selChars)
    numPws=50
    pwLen=80
    
    for pwNum in range(numPws):
        pwDigs=np.random.randint(0,62,pwLen)
        pw=''
        for dig in pwDigs:
            pw += selChars[dig]
        print(pw)

if __name__=='__main__':
    main()

Is there anything special about numpy's random generation? Why not just

Code:

password=''.join([chr(random.choice(range(33,127))) for i in range(64)])

?

Edit: You can also use os.urandom which is supposedly better since it uses the OS' native generator (/dev/urandom on Linux)

[bidji29]

Jesse James Aka DoctorEvil on the nextcoin forum (the one who cracked the genesis account passphrase) agreed to do some audit on the Crypto/Curve25519 code.

https://bitcointalk.org/index.php?topic=345619.msg5119097#msg5119097

He said it mi-february. Any news?

[marcus03]

Here is version 2.2 of my Windows/Mac OSX NXT client "NXT Solaris".


Changelog
-Mac OSX version added (source is included in the Windows downloads)
-GUI Performance improvements
-Changed the default testnet address to "http://tn01.nxtsolaris.info:6876/"
-The JCL is no longer needed to compile the source code
-Fixed a bug that prevented sending NXT to accounts without a public key in the BC
-Fixed Bter market data retrieval
-Fixed "secret maps to expected account number" check for secrets with characters beyond ascii
-Added checks for enough funds before broadcasting transactions


Download
64-bit Windows Version: NXTSolaris-v2.2-Win64bit.zip (36.3 MB) – Download here
SHA256 checksum for NXTSolaris-v2.2-Win64bit.zip: BCDC7C27755A138E44AE51FB3F842BB8E12FAC6AF3117F44BB99B9B01B581F95

32-bit Windows Version: NXTSolaris-v2.2-Win32bit.zip (34.3 MB) – Download here
SHA256 checksum for NXTSolaris-v2.2-Win32bit.zip: 0F6851CE16411756EB5ED9ADFBD1AF730CEC0B68D8566660D8A8257C0DF3F7B1

32-bit Mac OSX Version: NXTSolaris-v2.2-OSX32bit.zip (9.5 MB) – Download here
SHA256 checksum for NXTSolaris-v2.2-OSX32bit.zip: 76B416EB3783410E33AC14353501F9FC735863CE725205AE0DF78C1E678F76BD


For screenshots and everything else please visit: http://nxtsolaris.wordpress.com/


Installation
Simply unzip the archive and run the NXTSolarisStarter.exe file.

The current version 2.2 is compatible with the previous versions 2.1.

Roadmap
1.) Move current version of my code to XE5 and release closed source 32-bit and 64-bit versions for Windows. [DONE]
2.) Refactor code to a MVC architecture, so that I can create separate GUIs while using the same version of the business logic code. [DONE]
3.) Create new GUIs for Windows and OS X using the same business logic, but based on the cross-platform Firemonkey libraries. [DONE]
4.) Release Windows version (32 and 64 bit) based on FireMonkey. [DONE]
5.) Implement new features from the TODO list [DONE]
6.) Release source-code for FireMonkey Windows version (February 16th, 2014 latest) [DONE]
7.) Implement new features from the TODO list
8.) OSX compatibility changes
9.) Release OSX version based on FireMonkey.
10.) Release source code for FireMonkey OSX version.


I've reached the end of my roadmap! :-)


TODO list
-Arbitrary message encryption
-Offer a simple and advanced GUI, with the simple GUI targeted at new users



Donations
I hope you like my client and I’d be extremely happy to see some donations for the future of this project!

NXT: 1758531264253431177

[l8orre]

oh, please:


import numpy as np

import sys
import string
 

def main():
    selChars = string.ascii_letters+ string.digits
    print( selChars)
    numPws=50
    pwLen=80
    
    for pwNum in range(numPws):
        pwDigs=np.random.randint(0,62,pwLen)
        pw=''
        for dig in pwDigs:
            pw += selChars[dig]
        print(pw)

if __name__=='__main__':
    main()

Is there anything special about numpy's random generation? Why not just

Code:

password=''.join([chr(random.choice(range(33,127))) for i in range(64)])

?

Edit: You can also use os.urandom which is supposedly better since it uses the OS' native generator (/dev/urandom on Linux)

no, nothing special about that. just had it sitting there in the script .

[Uniqueorn]

CFB, are you the only one communicating with BCNext? Who will be in touch with him when you depart in 3 1/2 weeks?

He had left the day before Nxt source code became public. Went for good to work on other things. I'll get 3rd part of the plan in April but I can't contact him anymore.

somebody in the dgex forum said they talk to bcnext frequently, maybe they could help you out?

should we hallmark the testnet nodes or just leave them non hallmarked?

You may be referring to me saying that BCNext told us that NXT is a experiment for him and that he has abandoned the project.
I haven't seen anyone talk about talking to BCNext there.
I got to ask some question through proxy (CfB) to him and he answered, but as far as I know noone else is talking to him

[Come-from-Beyond]

somebody in the dgex forum said they talk to bcnext frequently, maybe they could help you out?

Maybe.

[Come-from-Beyond]

From http://mag.newsweek.com/2014/03/14/bitcoin-satoshi-nakamoto.html:

"Everyone who looked at his code has pretty much concluded it was a single person," says Andresen. "We have rewritten roughly 70 percent of the code since inception. It wasn't written with nice interfaces. It was like one big hairball. It was incredibly tight and well-written at the lower level but where functions came together it could be pretty messy."

BCNext followed Satoshi even in this...

[Uniqueorn]

From http://mag.newsweek.com/2014/03/14/bitcoin-satoshi-nakamoto.html:

"Everyone who looked at his code has pretty much concluded it was a single person," says Andresen. "We have rewritten roughly 70 percent of the code since inception. It wasn't written with nice interfaces. It was like one big hairball. It was incredibly tight and well-written at the lower level but where functions came together it could be pretty messy."

BCNext followed Satoshi even in this...

My theory is that Satoshi simply had this idea and then had to learn how to code, not the other way around.
I think the same goes for BCNext.

[mcjavar]

Jesse James Aka DoctorEvil on the nextcoin forum (the one who cracked the genesis account passphrase) agreed to do some audit on the Crypto/Curve25519 code.

https://bitcointalk.org/index.php?topic=345619.msg5119097#msg5119097

He said it mi-february. Any news?

Could you please follow this up?