Anyone could list a comparison about these two coins? which coin is better? Strengths and weakness ??
Primecoin has 1 minute blocks, no limit on coins, continuous difficulty adjustment, and difficulty based block rewards. It also has checkpointing.
Riecoin has 2.5 minute blocks, an arbitrary limit of 84 million coins, difficulty retargeting every 12 hours, and block rewards of 50, halving every 4 years.
Riecoin also claims that it is able to be able to estimate the average time before finding a block, contrary to the misinformation on Riecoin’s website that is also possible to do with Primecoins.
Both have proof of work that look for chains of prime numbers. Riecoin looks for chains of primes that have the minimum distance possible between each other, which Primecoin looks for three different types of chains, Two forms of Cunningham chains, and bi-twin chains. As far as I know I don’t think either method provides any more scientific value than the other.
The last difference is that Riecoin uses Rabin Miller test to test for primality while Primecoin uses a combination of the Euler-Lagrange-Lifchitz and Fermat tests to prove primality.
The Rabin Miller test is more accurate, it is also slower. The tests that Primecoin uses are faster but less accurate.
In my opinion, Primecoin is superior, obviously I am biased otherwise I wouldn’t be here. But I do have opinions to support that.
I think Riecoin is just a another Alt-coin made to serve the purpose of making the dev money with no real intention of it being a competing currency. It is basically just a fork of Bitcoin with some of the basic parameters changed and then ripping off the concept of using prime chains as a proof of work to try to set itself apart from the other “scam” coins.
Riecoin claims that 2.5 minute blocks are better because “litecoin” and 1 minute blocks causes more orphans, stales, and a bloated blockchain. Most new coins now have faster block times than 2.5 minutes and don’t have any significant issues with orphans or stales. As for the a bloated block time, the difference between 1 minute and 2.5 minutes isn’t very significant. Maybe if you were talking about 10 minutes you’d have a good point.
Then there is the debate about the limits on coins. While being very inflationary isn’t good for people holding, it also isn’t good for there to be a set number of coins. Coins get lost very frequently. Everyone also seems to ignore in the inflation/deflation argument is that all coins that are mined are going to be inflationary until they reach their cap, which will be in decades. Then you are stuck with the same amount of coins forever, no matter however many are lost.
That is why Primecoin with difficulty based block rewards and limitless coins is better. You can have inflation to keep the circulated coin supply at a reasonable level while not having inflation go out of hand.
Continuous difficulty adjustment. Why wait an arbitary amount of time, like 12 hours to adjust the difficulty if you can do it continuously and have it accurately reflect the power of the network?
The main point of contention, I think is the Rabin Miller test. Again, Riecoin provides misinformation to make their coin sound better than it is. Pseudoprimes in Primecoin are not as common as Riecoin makes it out to be. Riecoin claims there are infinitely many Psuedoprimes, well there are infinitely many composite numbers that pass the Miller Rabin test as well. Riecoin also emphasises the general chance of finding a psuedoprime while skips the details when he acknowledges that with the Fermat tests that Primecoin is using the chance is lower. To give you an idea, there are only 21,853 psuedoprimes less than 1,000,000,000 using the same arguments for the Fermat test that Primecoin uses.
They are much more rare than finding actual primes, so it poses no risk to the security of the network and occurs so rarely that it isn’t significant. Even if chains were found with Composites and accepted, any researcher using the information could easily verify the chains with the Miller Rabin test before using them.
Also, as far as I know, Primecoin uses the Eular-Lagrange-Lifchitz first and then the Fermat test to for primality as the former is faster and will exclude some composite numbers before using the slower Fermat test.
The final point I have to make is, if the majority of the time the Fermat test will be accurate and the Rabin-Miller test is slower, what is to stop custom Riecoin clients from using the Fermat test and submitting the results to the network anyways? If it is valid they didn’t have to waste extra processing. If it isn’t valid the network will waste time figuring it out and it happens so rarely that it would be impossible to detect.
If there are enough individuals and pool operators that are concerned about performance and decide to implement this optimization Riecoin will probably end up forking itself anyways to use the Fermat test instead of the Rabin-Miller test.
Thanks for the elaborate reply okzx. I guess you don’t believe in the “fair launch” Riecoin claims to have had?
I have only one comment: Riecoin seems a lot simpler. Not just counting lines of code, but also counting commits. Primecoin seems like it’s in development, getting patches merged from upstream. How do I see the Primecoin specific changes? Until now I have been browsing the commit log for “Primecoin:”. With Riecoin it’s simpler, cause all changes were done in three commits :P.
I forgot the fair launch claim. Check pointing also. Although I’m too lazy to go into that now.
I don’t think either Riecoin or Primecoin’s launches were unfair. While I agree that having the first blocks give no reward gives more people time to get their clients in order, but I don’t think it makes it that much more fair. First, you can just pre-release your client but not enable it for mining, secondly, anyone who is smart would just not start mining until block rewards have started. It is just a waste of electricity. Then the difficulty wouldn’t be accurate because all of the smart miners weren’t even mining. Although, honestly, not a big issue with 12 hour re targeting times. Although it isn’t very fair for the miners wasting electricity and the miners after the difficulty adjusts.
Riecoin has a block reward of 50 every 2.5 minutes, with the intention of it halving after 4 years while primecoin started with a block reward of 20 every minute (same rate as Riecoin) but with difficulty based rewards it has already more than halved in less than a year. I don’t really have an opinion on if either if fairer than the other. Both favor early adopters. I guess Riecoin is more fair as it will allow for more early adopters to benefit due to the longer time before halving.
Primecoin’s launch was fair enough and Riecoin’s launch was something that tried to sound good but shouldn’t work well in practice. They both seemed fair enough regardless, so i’m really indifferent on that.
As for going through Github, I can’t help you with that. But I can imagine Riecoin is simplier, the developer probably just pulled Bitcoin, changed a couple values, replaced the proof of work function, and boom, another alt-coin lacking any real innovation.
Perhaps more work than that. For one thing you have to find a new way to define difficulty for a new algorithm:
Perhaps more work than that. For one thing you have to find a new way to define difficulty for a new algorithm:
That is true. It’s still just two functions instead of one. I would concede on that point if Riecoin argued that it was an innovation over Primecoin’s method, other than it being different just to be different. Looking at it, it just sounds like they took Bitcoin’s method for difficulty and applied it to prime numbers instead of a hash. Granted, It is a little more complicated than that, but not much more.
How about a comparison between the Primecoin proof-of-work and Cuckoo Cycle at https://github.com/tromp/cuckoo ?
The latter has simpler code, faster (and exact) verification, more precise memory requirements (that can be set arbitrarily), and is less compute intensive.
The major downside is that it has nothing to do with primes:-(
Still, finding long cycles in random graphs does sound vaguely scientific…
gatra (Riecoin boss) has replied to your post, okzx: http://riecointalk.com/index.php/topic,175.msg474.html#msg474
Riecoin scores one if it has an open source pool.