Topic: Math of Nxt forging (Read 22579 times)

CIYAM · « **Reply #20 on:** March 26, 2014, 05:59:14 pm »

Guys - aren't we "on the same side" here?

The point is to improve the randomness not to have a "pissing match".

ChuckOne · « **Reply #21 on:** March 26, 2014, 06:03:02 pm »

Quote from: CIYAM on March 26, 2014, 05:59:14 pm

Guys - aren't we "on the same side" here?

True. Nevertheless, you have to understand mczarnek'. He is concerned of his intellectual contribution not being recognized sufficiently. At least to me, it is understandable.

L5Society · « **Reply #22 on:** March 26, 2014, 06:03:58 pm »

Assigning credit for the original idea matters when there is a hypothetical bounty on the line.

It can make things ugly.

ChuckOne · « **Reply #23 on:** March 26, 2014, 06:07:39 pm »

I agree with all what you said. Let us give a break here.

We need to focus on the more important issue and do justice to the topic of this thread.

mthcl · « **Reply #24 on:** March 26, 2014, 08:12:46 pm »

Well, I propose to stop all discussion about priorities, who first posted on the forum etc. I only remember this algorithm came to my head last Friday, when I was driving on a highway

When smart people think about the same problem, they often come to roughly the same solution, independently. So, we are on the same boat, and let's work all this out together!

Also, about bounties, donations, etc.: of course, it's very pleasant to receive them (and, please, do send donations to us

). But, anyhow, I guess that everybody in this thread already owns a decent amount of NXT, so the valorization is our best friend (and this partly depends on our work)

mthcl · « **Reply #25 on:** March 26, 2014, 08:37:09 pm »

Quote from: ChuckOne on March 26, 2014, 05:50:03 pm

Thank you, mthcl, for initiating that discussion.

This paper http://www.links.org/files/decentralised-currencies.pdf is a very good starting point to gain knowledge about the consensus problem and why including random numbers is not as easy as it seems.

Thanks for the reference!

It seems to me, by the way, that what we are proposing here is more or less what is written in Section 7 of that article. Don't you think so?

mthcl · « **Reply #26 on:** March 26, 2014, 08:44:56 pm »

About XORing, can it be vulnerable to the following attack: the attacker copies the hash of the good guy, and then copies the number itself. Then, he would just "erase" the number of the good guy. Does this work?

mthcl · « **Reply #27 on:** March 26, 2014, 08:51:33 pm »

Quote from: mthcl on March 26, 2014, 04:31:37 pm

Quote from: mthcl
Quote from: ChuckOne
Quote from: mthcl
OK. Not sure I understand the picture; but is it possible that an account somehow broadcasts (i.e., makes public to everyone) its (signed) hash?

Yes. Of course. But these published signed hashes are like raw transactions. As long as they are not included in the block, the network has yet to agree on them. So, they can easily be removed by anybody (and so by the next forger as well).

Then I think my technical knowledge is not enough to answer this. Maybe, we discuss all that openly at https://nxtforum.org/transparent-forging-*/ , so someone else could contribute?

So, does anyone have an idea, how the accounts can publish the hashes of their numbers in a reliable way (so that the network agrees on what is actually published)?

mczarnek · « **Reply #28 on:** March 26, 2014, 08:58:32 pm »

ChuckOne and I chatted off of the forum, it looks like he and mthcl, came up with the approximately same idea independently and upon further and finer reading, he did not disclose anything we discussed.

Sorry. For the record though, everything in my version of the paper was thought up the day before they came up with the same idea. I told him I'd solved it and I thought he had more information than he did.. a google doc link that in hindsight I don't think I gave him permission to view was involved, and long story short, ChuckOne and mthcl independently arrived at the same basic idea.

ChuckOne · « **Reply #29 on:** March 26, 2014, 10:17:43 pm »

Alright. I think this thread can be considered closed aka continues over here: https://nxtforum.org/transparent-forging-*/transparent-forging-algorithm-245/

mthcl · « **Reply #30 on:** March 26, 2014, 11:11:01 pm »

Quote from: ChuckOne on March 26, 2014, 10:17:43 pm

Alright. I think this thread can be considered closed aka continues over here: https://nxtforum.org/transparent-forging-*/transparent-forging-algorithm-245/

Closed only for now. I expect to have more versions of the math-of-forging paper...

mthcl · « **Reply #31 on:** April 23, 2014, 01:32:38 pm »

Version 0.4.2: http://www.docdroid.net/bfwb/forging0-4-2.pdf.html

But, maybe, better discuss it here: https://nxtforum.org/technical-development-applications/transparent-forging-algorithm-755/

mthcl · « **Reply #32 on:** May 18, 2014, 08:41:31 pm »

Version 0.4.3: http://www.docdroid.net/cckd/forging0-4-3.pdf.html Made it some time ago, but forgot to publish

Come-from-Beyond · « **Reply #33 on:** May 18, 2014, 08:54:42 pm »

Quote from: mthcl on May 18, 2014, 08:41:31 pm

Version 0.4.3: http://www.docdroid.net/ccka/forging0-4-3.pdf.html. Made it some time ago, but forgot to publish

When will we get 1.0.0?

mthcl · « **Reply #34 on:** May 18, 2014, 08:58:10 pm »

Quote from: Come-from-Beyond on May 18, 2014, 08:54:42 pm

Quote from: mthcl on May 18, 2014, 08:41:31 pm
Version 0.4.3: http://www.docdroid.net/cckd/forging0-4-3.pdf.html. Made it some time ago, but forgot to publish

When will we get 1.0.0?

We may call this one 1.0.0

But I'm not sure if the random number generation and the TF algorithm should be a part of this paper, maybe better leave only probabilistic things there.

mthcl · « **Reply #35 on:** June 29, 2014, 10:22:50 pm »

Version 0.5.1: http://www.docdroid.net/e29h/forging0-5-1.pdf.html

mthcl · « **Reply #36 on:** July 05, 2014, 07:52:17 pm »

Version 0.5.2: http://www.docdroid.net/ecmz/forging0-5-2.pdf.html. Very minor changes: updated one picture and a couple of other small things.

mthcl · « **Reply #37 on:** December 22, 2016, 07:50:51 pm »

http://www.ledgerjournal.org/ojs/index.php/ledger/issue/view/2/showToc

http://www.ledgerjournal.org/ojs/index.php/ledger/article/download/46/60

abctc · « **Reply #38 on:** December 22, 2016, 09:26:04 pm »

Quote from: mthcl on December 22, 2016, 07:50:51 pm

http://www.ledgerjournal.org/ojs/index.php/ledger/issue/view/2/showToc

Congratulations with the publication in the 1st editoin of Ledger Journal, Serguei!

mthcl · « **Reply #39 on:** December 22, 2016, 09:51:10 pm »

Quote from: abctc on December 22, 2016, 09:26:04 pm

Quote from: mthcl on December 22, 2016, 07:50:51 pm
http://www.ledgerjournal.org/ojs/index.php/ledger/issue/view/2/showToc
Congratulations with the publication in the 1st editoin of Ledger Journal, Serguei!

Спасибо!

News:

Author Topic: Math of Nxt forging (Read 22579 times)