Quote from: gjhiggins on April 01, 2017, 12:15:28 PM

Quote from: avitas on March 31, 2017, 09:23:08 PM

1. what are the 11 hashing algorithms chained and what are the benefits of such chaining.
2. can we start the masternodes remotely like other masternode coins, and do we need a unique ip for each or just a different port.
3. and research or plans to expand the masternodes to do elastic computing instead of mining. "elastic chains" sounds great.
4. can the dev team introduce he/she/themselves?

1. what are the 11 hashing algorithms chained and what are the benefits of such chaining.

C11 -> https://github.com/chaincoin/chaincoin/blob/master/src/hash.h#L194

(for reference, also see Darkcoin-now-Dash):

X11 > https://github.com/dx11/darkcoin/blob/master/src/hashblock.h#L62

The benefits are at best arguable and in peer-reviewed cryptography papers the technique (of serially chaining the output of different hash algorithms) has been characterised as folklore. The characterisation is accurate, as far as I can ascertain, according to my understanding of the bitcointalk origins of the approach:

The original notion of chaining the output-to-input of different hash algos (as opposed to Satoshi's double-wrapping of SHA2 to elide a speculated-upon length extension vulnerability) was proposed by the Russian developer(s) of SIFCoin, ANN, here translated informally:

Quote

hesh function for the signature block header changed from sha-256 (sha-256 ()), on the daisy chain of the candidates / finalists and winner sha-3. Blake, BMW, Groestl, JH, Keccak, Skein. All functions of the 512-bit, but the end result is truncated to 256 bits. (Rejection of the sha-256 and scrypt - to protect against a possible startup sharp influx, and then just as sharp outflow capacity miners from other currencies, and the subsequent "paralysis." Complication chain to a length of 6 different hash functions and increased bit depth to intermediate 512 - an attempt to protect from further development of highly efficient Mh / s gpu-algorithms and theory, "simple" Gh / s devices).

I've highlighted the key sentence - the aim was anti-ASIC, they ignore any implications for the integrity of the cryptography.

The notion was picked up by the developers of QuarkCoin, who threw in another algo:

https://github.com/quark-project/quark/blob/1c0961810b65226bf74ba98e7c0a11ada8b7cd32/src/hashblock.h#L18

and then stirred 'em up a bit:

https://github.com/quark-project/quark/blob/506b294665eb61033cee7bf7404235a93db4e535/src/hashblock.h#L18

Unfortunately, it was at this point that the bogus claim for strengthened security appeared in the bitcointalk ANN for Quarkcoin:

Quote

Super secure hashing: 9 rounds of hashing from 6 hashing functions (blake, bmw, groestl, jh, keccak, skein). 3 rounds apply a random hashing function.

(my emphasis)

Although the claim of super secure hashing was later quietly dropped from the Quarkcoin rubric (presumably as indefensible), the fiction quickly spread amongst altcoin devs, starting with Darkcoin (which didn't even change the copy'n'pastad function name --- Hash9):

https://github.com/dx11/darkcoin/blob/master/src/hashblock.h#L62

and a short time later, Chaincoin, which used the same algos but two appear in a slightly different sequence to Hash9.

https://github.com/chaincoin/chaincoin/blob/893af8e4bd73344c007b61c20e9b97b2a1cf7b02/src/hashblock.h#L54

Subsequently, according to the MOAR-HASHES!!! principle - I don't have any cryptography literature to reference, only psychological literature which, if you'll forgive me, I presume is probably outside your bailiwick - the fiction was developed up to and including X19 (IIRC).

I haven't any evidence that there has been a C11 ASIC released thus far but I understand there are one or two X11 ASIC offerings around.

From a cryptography standpoint, the approach fails on first principles - not only can the technique not be proven to strengthen the cryptography, it cannot be proven not to weaken it. For a more informed discussion, see http://crypto.stackexchange.com/a/3763, one example amongst many in the literature.

2. can we start the masternodes remotely like other masternode coins, and do we need a unique ip for each or just a different port.

Chaincoin 0.9.2 is based on Dash 0.9.2, you get whatever Dash was offering at the time, so the answer is probably no-yes-no (testnet is your friend).

3. and research or plans to expand the masternodes to do elastic computing instead of mining. "elastic chains" sounds great.

i) This is bitcointalk, talk is cheap and plans even cheaper.
ii) Chaincoin is open source software
iii) Ask yourself, is there a budget?

You may conclude: if there is an unadvertised research effort, it will remain opaque to casual enquirers. OTOH, if there is no public research effort, anyone is free to initiate one.

4. can the dev team introduce he/she/themselves?

i) After you
ii) I encourage you to reflect on where you first encountered the notion of the dev [team] w.r.t to an altcoin and whether this also might be a popular fiction, given the open source context.

Cheers

Graham

Thanks for the detailed reply Graham!

Just a clarification, Chaincoin was released before Darkcoin and was originally based on Zetacoin/Bitcoin, so the 11 chained hashing algo was first done by Chaincoin. Chaincoin only later adopted the Dash code base.