Quote from: Anotheranonlol on October 04, 2014, 10:10:12 PM

Could you elaborate on the con's of Wild Keccak's proof-of-work function approach implemented in Boolberry over vanilla cryptonight implemented in bytecoin, quazarcoin,aeoncoin monero etc?

1. Getting pool support, miner support, etc. was much slower and contributed to the lag of uptake of the coin. There are still many more pools for even the failed cryptonight coins than for BBR.

2. A much higher GPU performance ratio, especially earlier. Arguably a plus or minus depending on your point of view.

3. Lag on getting optimized miners, which also contributed to the lag of uptake, and also contributed (much) more to the dumping than the claims of manipulation. BTW, for the conspiracy theorist(s) who think(s) all BBR dumping was manipulative, it is well documented that XMR had a lag in getting optimized miners and this contributed to hundreds of thousands of USD of dumping. The difference is that XMR had the demand to absorb those coins, and BBR did not. Nevertheless I count this as a negative on WK because of the one-month launch lag. BBR could have piggybacked on the optimization (de-unoptimization) curve of the rest of the cryptonotes and avoided its privileged mining period almost entirely.

Quote

Seems that it satisfies the same ASIC resistant properties of cryptonight

In reality who knows.

Quote

Whilst you may not agree with the numbers there (obviously the 17x slower syncing of monero was not simply down to using cryptonight instead of wild keccack) with same transaction flow, it was demonstrated to synchronize 4x faster - of course we know that there are not 4x as many blocks in XMR than BBR, that seems to go against your claims.

When crypto_zoidberg measured it he confirmed that only a portion of the difference in sync time was due to PoW. My recollection (not 100%) is that he was slightly surprised that portion was as small as it was. This is not rocket science, you can measure the PoW time on a typical computer at about 20ms, which means 50 PoW verifies per second. Roughly an hour worth of (XMR) blocks per second at that rate. A week takes a few minutes.

BTW, I slightly questioned the 4x number when it was reported because during the period in question XMR had 2.5x the number of blocks of BBR (and therefore 2.5x the number of coinbase transactions, etc.), while on average it would have 2x. A small difference but likely significant on a few of the measurements and the 4x overall result. No question it is still faster to sync, I have acknowledged that several times.

Quote

Quote from: smooth on August 31, 2014, 06:46:23 PM

some advantage -- or disadvantage -- in the future[/b]).

Could you give more details to the disadvantage it might have in the future that you alluded to there?

Only in so far as we will have to see what the future holds.

I'm always skeptical of homebrew PoW (and crypto generally). Cryptonight probably qualifies as that too, but at least it was already in use on several coins, well on its way to being optimized by the time BBR launched, etc. If you're going replace homebrew, you should probably replace it with non-homebrew.