NXT :: descendant of Bitcoin - Updated Information

[NxtChg]

API 2 won't care about addresses. Transactions will be signed on client side.

But when you return addresses for transactions?

Like current getTransaction.

Old 64-bit unsigned number.

So we can protect from any corruption on the way from core to client by adding new addresses.

Otherwise if something goes wrong, and a wrong address gets to the client, the new addresses won't save the user.

[martismartis]

When you remember the address, could you send them to Anon for safe keeping?   

The file with the link to online wallet (blockchain.info guid) is encrypted with the BTC address...

I think guys who sold me the bitcoins should have it in their wallet history.

Edit: I'll find it in browser history, don't worry.

Maybe this: 1LLe2He9k8r8WnAnuXJbAUedX8c2HCr8bQ

[jl777]

http://www.youtube.com/watch?v=Uh6erfE9HYE

Matthew Green's half hour presentation at Real World Crypto 2014

They massively improved on the alpha version. zerocash uses SHA256, merkle trees, etc. and seems to be ideal to implement onto NXT

[jl777]

Why not zerocoin? Any unproven tech will need to be reviewed by cryptographic authority and we know that is not so easy.

I'll read Zerocoin whitepaper tomorrow and will think. Today is too late, time to drink milk and go to bed soon.

In my search for a crypto algo reviewer, I was referred to a crypto expert named Matt Green.  He did get back to me yesterday and said he was currently advising Zerocoin but might have some time to review our code.  I did mention to him that we were interested in implementing Zerocoin with Nxt POS and would like him to review our code if he could.  Here is his blog, seems like a really nice guy; http://blog.cryptographyengineering.com/

Also, I've had a hard time getting crypto people to review the code, they would rather review a protocol design document instead of code.

That's the zerocoin professor!
zerocash is scheduled to be released in May as altcoin.

[Come-from-Beyond]

But in any other case, how could Big Brother possibly add your "more info" if all transaction in the past are fulfilled?

Imagine that weight of a real (weight = 100%) link between IN and OUT is 5% (that's how Big Brother sees it). U use this OUT to buy something from a website that is later seized by BB. Now he knows one of ur real IPs, this increases that link weight from 5% to 12%.

Even worse, some other guy reveals his real transactions, this automatically increases weights of all the other links involved into the same mixing session.

And the other thing: what's the issue with the 'joincoin approach'? I.e. if n senders sends 5Nxts then n receiver receives 5Nxts. So, nobody could differentiate who exactly transferred whom 5Nxt. They send collectively n*5 Nxts and they receive collectively n*5 Nxts.

"nobody could differentiate who exactly transferred whom 5Nxt" is correct, but actually some non-zero probability is enough to use approach described above to deanonymize u.

[Come-from-Beyond]

So we can protect from any corruption on the way from core to client by adding new addresses.

Otherwise if something goes wrong, and a wrong address gets to the client, the new addresses won't save the user.

Signature would be invalidated in case of corruption.

[Come-from-Beyond]

Maybe this: 1LLe2He9k8r8WnAnuXJbAUedX8c2HCr8bQ

Yes, thank u.

Money is still there, it's good.

[NxtChg]

They massively improved on the alpha version. zerocash uses SHA256, merkle trees, etc. and seems to be ideal to implement onto NXT

Watching. This indeed looks much better.

[nxtru]

The block #53436 contains 65 transactions!

[jl777]

They massively improved on the alpha version. zerocash uses SHA256, merkle trees, etc. and seems to be ideal to implement onto NXT

Watching. This indeed looks much better.

3 totally private transactions can be embedded in a single AM
Compute time was ~one minute, but that is done by the local client computer, verification time <10 milliseconds

small enough to fit in AM, fast enough for 100 TPS

Whitepaper will be written by univ, professor
reference implementation with source code in May

If we can establish earlier workings with them, that would be much better.

James

[martismartis]

Maybe this: 1LLe2He9k8r8WnAnuXJbAUedX8c2HCr8bQ

Yes, thank u.

Money is still there, it's good.

5% for storing your wallet number 

[NxtChg]

Signature would be invalidated in case of corruption.

You're not listening! Forget about transaction signing, I understood already long time ago what you're trying to say.

This is another issue:
- client calls getTransaction
- core returns sender address
- something goes wrong and client receives a corrupted address
- user decides to send money to that address and loses money

If that would be the new address, it would be protected from corruption.

We have a chance to add extra protection, why not do it?

[Come-from-Beyond]

You're not listening!

I am. I'm just sleeping. I got u. I hope.

[Std52stp]

Why is NXT falling down?
Do you think it will reach 0,00005000 btc again ?

[Anon136]

Why is NXT falling down?
Do you think it will reach 0,00005000 btc again ?

from my observation of the charts it appears to be driven from the supply side. strong demand but big seller(s) taking very large dumps.

[msin]

Why is NXT falling down?
Do you think it will reach 0,00005000 btc again ?

Maybe, we had a dip before it skyrocketed before, so don't wait to long.

[ChuckOne]

But in any other case, how could Big Brother possibly add your "more info" if all transaction in the past are fulfilled?

Imagine that weight of a real (weight = 100%) link between IN and OUT is 5% (that's how Big Brother sees it). U use this OUT to buy something from a website that is later seized by BB. Now he knows one of ur real IPs, this increases that link weight from 5% to 12%.

Even worse, some other guy reveals his real transactions, this automatically increases weights of all the other links involved into the same mixing session.

Social engineering, I see.

And the other thing: what's the issue with the 'joincoin approach'? I.e. if n senders sends 5Nxts then n receiver receives 5Nxts. So, nobody could differentiate who exactly transferred whom 5Nxt. They send collectively n*5 Nxts and they receive collectively n*5 Nxts.

"nobody could differentiate who exactly transferred whom 5Nxt" is correct, but actually some non-zero probability is enough to use approach described above to deanonymize u.

Correct me if I am wrong, but if that is so, that even mixing with sum(in)=/=sum(out) would give my links a non-zero probability.

[jkoil]

a sneak peak to my idea about a green Nxt video:
http://www.youtube.com/watch?v=ic_WLMPX888&feature=youtu.be

It's cheesy, yet funny and spot on

It made me laugh, very accurate though!

good video,
some watchers may still be interested in the numbers and ask for the proofs: could there be also a link to those calculations.
Or is it better that now is only raised some questions ...

I agree, that's why NxtMyths projects start as a paper, will continue as infographics and will be completed in the video(s). I just wanted a little more fun for 2 hours

ok
Sometimes it may be better to start with the video. But, ofc it's safer to start with facts and proofs

BTW, is there any other example of some system, which takes as much electricity as Bitcoin (17 million USD / day) ?
http://qz.com/156479/miners-spend-17-million-a-day-for-a-shot-at-4-4-million-of-bitcoin/

Just wondering, how Bitcoin can justify that kind of waste ... is it really worth of 17 M ...

[msin]

Why not zerocoin? Any unproven tech will need to be reviewed by cryptographic authority and we know that is not so easy.

I'll read Zerocoin whitepaper tomorrow and will think. Today is too late, time to drink milk and go to bed soon.

In my search for a crypto algo reviewer, I was referred to a crypto expert named Matt Green.  He did get back to me yesterday and said he was currently advising Zerocoin but might have some time to review our code.  I did mention to him that we were interested in implementing Zerocoin with Nxt POS and would like him to review our code if he could.  Here is his blog, seems like a really nice guy; http://blog.cryptographyengineering.com/

Also, I've had a hard time getting crypto people to review the code, they would rather review a protocol design document instead of code.

That's the zerocoin professor!
zerocash is scheduled to be released in May as altcoin.

I can send him a proposal with bounty offer if you would like to put some text together regarding what you want.  

[Emule]

do you all like it? not even started ... plenty more to come I hope i don't kill the coin