Quote from: makomk on August 28, 2011, 10:59:45 AM

Quote from: Chris Acheson on August 28, 2011, 07:39:26 AM

I'm talking about the protocol. You know, that thing that's actually kind of difficult to change once a blockchain gets established? What have you done that can't just be released as a modified client for the Bitcoin blockchain?

All of the changes can't be released as a modified client for the Bitcoin blockchain; doing so would cause a fork in the blockchain which would be a lot more disruptive than simply starting a new one. In particular, the problem with forking the blockchain is that transactions can be carried from one side to the other and this can be used to double-spend bitcoins - once on each side of the fork - and obtain bitcoins that are valid on both sides, guaranteeing that exchanges and e-wallets will be screwed over.

I didn't say that all of the changes could be merged into Bitcoin. However, CoinHunter is crowing about supposed UI and API improvements in the SolidCoin client. These aren't protocol level changes, so presumably if any of them have merit they can just be copied, and aren't a good reason to switch to the SolidCoin blockchain.

The reason I'm asking about protocol changes is that those can't just be copied easily. If there were significant protocol improvements, it might make for a compelling reason to switch. As it stands, SolidCoin's protocol changes are unimpressive

Quote

Quote from: Chris Acheson on August 28, 2011, 07:39:26 AM

Difficulty algorithm aims for 3 minute blocks instead of 10 minute blocks. Achieved by changing one constant. Of questionable merit, due to scalability concerns.

Sounds like a good reason to implement it on a new, smaller network rather than Bitcoin then.

If an alternate blockchain currency is going to survive, it can't aim for being smaller than Bitcoin. As I said before, we're dealing with a natural monopoly. Some have stated that "Ix/I0/SolidCoin can be the silver to Bitcoin's gold". This is a poor analogy, as BTC is divisible enough that there's no need for a lower-value cryptocurrency to "make change" with. So really, mBTC are the silver to BTC's gold.

Quote

Quote from: Chris Acheson on August 28, 2011, 07:39:26 AM

Difficulty retargets more often. Achieved by changing one constant. Essential for an underdog blockchain to survive, but not really relevant for Bitcoin.
Difficulty increases limited to +10% per retarget. Achieved by changing one constant. Essentially an extra subsidy to miners, not relevant for Bitcoin.

Not relevant to Bitcoin now; it's an open question as to whether it will be in the future. Also, the limited difficulty increases aren't just a subsidy to miners, they're also necessary to stop the difficulty from overshooting wildly. (For some reason they're not just implemented as a simple change to one constant either; the way difficulty increases are calculated is substantially modified.)

If it becomes relevant to Bitcoin, then Bitcoin will be pretty much doomed anyway.

The difficulty retargeting algorithm doesn't really overshoot. You're not going to get a 400% difficulty increase just because blocks are being solved every 9 minutes instead of every 10 minutes. The problem occurs when the majority of your mining power leaves, not when the difficulty changes.

The limited difficulty increases aren't bad per se, it's just that in combination with the fast difficulty decrease, the average number of blocks produced ends up being far higher than the target rate, hence my subsidy comment.

Quote

(Also, am I the only person that's more worried about the security implications of their changes to difficulty targetting than anything else?)

I'm not familiar with how the network enforces time synchronization. Maybe if you find the last block in a difficulty retargeting period, you could give it a fake timestamp to mess around with the difficulty calculation. It seems like if this would be a problem for SolidCoin, it would be a problem for Bitcoin too, but with fewer opportunities to exploit it.