Yacoin GPU Mining
-
[quote name=“justabitoftime” post=“13420” timestamp=“1371006809”]
Looks like Yacoin now has GPU mining.
[/quote]It seems they are using a modified scrypt hashing algorithm.
Maybe we should consider doing a similar thing?
Advantage: Hashing power can not be redirected so easily -> attacks from rouge LTC pools are impossible
Disadvantage: Hashing power can not be redirected so easily -> harder for legit users to switch to us. -
It uses a predictable N. That’s asking for being pwned.
-
[quote name=“zerodrama” post=“13599” timestamp=“1371033047”]
It uses a predictable N. That’s asking for being pwned.
[/quote]Can you please explain this (pretend your talking to your non-geek friend so we reach a wider audience here)?
-
If you know N+1 is coming at a given block and you have the hashrate you can premine it like it’s a new chain and then inject right as everybody else switches over. You end up having the longest blocks every time the thing switches.
-
[quote name=“zerodrama” post=“13657” timestamp=“1371044661”]
If you know N+1 is coming at a given block and you have the hashrate you can premine it like it’s a new chain and then inject right as everybody else switches over. You end up having the longest blocks every time the thing switches.
[/quote]What would you have done differently? I’m not talking about the 2 chain ideas you’ve presented in the feature development lead thread, but more like something minor that could have offered their solution a smaller attack surface?
-
[quote name=“justabitoftime” post=“13658” timestamp=“1371044960”]
[quote author=zerodrama link=topic=1657.msg13657#msg13657 date=1371044661]
If you know N+1 is coming at a given block and you have the hashrate you can premine it like it’s a new chain and then inject right as everybody else switches over. You end up having the longest blocks every time the thing switches.
[/quote]What would you have done differently? I’m not talking about the 2 chain ideas you’ve presented in the feature development lead thread, but more like something minor that could have offered their solution a smaller attack surface?
[/quote]“Randomizing shield frequencies, Captain. Attacks have ceased to penetrate shields. Oh and number 1 I order you to do a number 2”
Still maybe fixing YAC’s approach might give us a subchannel for microtransactions. Gah. I need like 3d notepad to keep track of all these ideas.
Longer version:
Randomize N within a given range, use a difficulty and reward window (related to N) to allow the blocks to be accepted. -
[quote name=“zerodrama” post=“13666” timestamp=“1371045649”]
[quote author=justabitoftime link=topic=1657.msg13658#msg13658 date=1371044960]
[quote author=zerodrama link=topic=1657.msg13657#msg13657 date=1371044661]
If you know N+1 is coming at a given block and you have the hashrate you can premine it like it’s a new chain and then inject right as everybody else switches over. You end up having the longest blocks every time the thing switches.
[/quote]What would you have done differently? I’m not talking about the 2 chain ideas you’ve presented in the feature development lead thread, but more like something minor that could have offered their solution a smaller attack surface?
[/quote]“Randomizing shield frequencies, Captain. Attacks have ceased to penetrate shields. Oh and number 1 I order you to do a number 2”
Still maybe fixing YAC’s approach might give us a subchannel for microtransactions. Gah. I need like 3d notepad to keep track of all these ideas.
Longer version:
Randomize N within a given range, use a difficulty and reward window (related to N) to allow the blocks to be accepted.
[/quote]How wide would that range need to be to really minimize the attack impact? If N is known within a given range, does it really offer that much more protection unless it’s fairly large?
-
Could be a short range with alternating windows. Or we could make the cost of a miss be large.
Literally, I could come up with a wide combination of ideas.
This is why I want to put out floppy fork (soon soft fork) capability. Our options become huge.
-
[quote name=“zerodrama” post=“13677” timestamp=“1371046990”]
Could be a short range with alternating windows. Or we could make the cost of a miss be large.Literally, I could come up with a wide combination of ideas.
This is why I want to put out floppy fork (soon soft fork) capability. Our options become huge.
[/quote]I thought it was clear… this is feathercoin. We don’t fork. We
[img]http://www.avenue80.com/media/2011/11/titanium-spork.jpeg[/img]
-
;D
