[XPM] [ANN] Primecoin Release - First Scientific Computing Cryptocurrency

[oroqen]

Not sure if this has been asked, but is there inherent value in the primes themselves? Can we extract them and sell them?

The primes...no.

The Cunningham chains, well there is certainly some notoriety for finding new ones but I don't think anyone would pay for them?

Since there in a public domain ala through the block chain you'd have a hard time proving you had the right too sell it

[salfter]

I have a Gentoo ebuild for primecoind in my overlay:

https://github.com/salfter/portage/

I built it first on my VPS, which has 8 Xeon E5-2670 cores available to it, but immediately upon starting, it bombs out with an "Illegal instruction" error.  I then built it for my home workstation, which runs a Core 2 Quad Q6600. As I write this, it's syncing the blockchain.

The VPS is an x86 Gentoo build, as it used to be memory-constrained (with 1 GB, maybe it still is somewhat).  The home workstation is an AMD64 Gentoo build.

Does primecoind not run on 32-bit systems, or is there something my ebuild is doing wrong?

(Hmm...just tried my bitcoind ebuild on the VPS, and it's also crashing right away with an "Illegal instruction" error. Must be something it's telling the compiler to do.)

I tried bringing up primecoind under gdb, but gdb crashes with the same error.   I think this might explain what's happening:

http://www.gossamer-threads.com/lists/gentoo/user/270976?do=post_view_threaded

Set your CFLAGS on your prod server to that of your dev server, if your dev server is known to work. You're using -march=native on your prod server, which depends on gcc correctly detecting CPU features from the host. There was a thread on this list just a few days ago about how that can fail in virtualized environments. (You can enable/disable exposed features piecemeal, which could well confuse the heck out of gcc's detection heuristics...)

I had -march=native set in /etc/make.conf.  Setting it to -march=i686 still caused an illegal-instruction error.  So did removing -march=* altogether.

More googling turned up this:

http://bugs.funtoo.org/browse/FL-494

Chroot into a stage3 build for x86_64 with glibc-2.15-r3 does not work on a xen based linode VPS with a Xeon E5-2630L due to linode disabling AVX in xen.

Chroot fails with the following error message: "Illegal instruction".

Linode probably is not going to enable avx, as "debian/ubuntu have gone the libc patch route ... https://launchpadlibrarian.net/101306417/eglibc_lucid_fix979003.debdiff" (quoting someon from the linode channel)

My Linode VPS was migrated to new hardware not long ago, and it looks like that may be causing problems that only started to manifest themselves when I tried bringing up primecoind on it.  I spooled up a 32-bit Ubuntu 12.04 LTS instance on the VMware box at work and successfully built primecoind on it, so it looks like an issue with my VPS.  I'm now rebuilding the whole system with -mtune=generic to try to clean up this problem.

[lucasjkr]

lucasjkr, unfortunately I can't answer every point, but here's a rebuttal of the major one on the evils of finite supply and lack of incentives for mining:

Transactions fees. Transactions fees will be the bread and butter of future miners when block rewards become negligible.
Yes, transaction costs will go up due to crowding, so there will be less transactions. The cheaper transactions will live in another blockchain - Litecoin or perhaps Primecoin.

Well, you're more open minded than many who argue that bitcoin should be the only block chain. I think there's room for more than one. And, yes, I think that the combination of hoarding and the fact that transaction fees could be much more than the trivial value they are today will severely hamper bitcoin one day. A transfer that's pennies today could easily surpass credit card fees in a few years and even rival wire transfer fees. I'm happy that a developer is providing us a chain that won't have to deal with that so much. Though we have to see if, as the reward dwindles, people stick with it. That's why I'd have liked to see a currency where the reward actually increases somewhat to keep pace with the issuance.

[canoe]

I've added primecoin to my charts website:

http://cryptometer.org/primecoin_96_hour_charts.html
http://cryptometer.org/primecoin_90_day_charts.html

I left the "Network Hashrate" chart in even though it's a bit weird for this coin.  I haven't done the work yet to make it "Network prime-chains per second" as it should be.  I'd need to know the chances of a number being the origin for a Cunningham chain of length D (difficulty).  I'll need to do more research before working on that.

So right now, network hashrate is just showing how many hashes/sec would be necessary to obtain that many blocks at that difficulty *for bitcoin*.  It's not useless.  The primecoin paper indicates that the primecoin difficulty linearly affects prime-chain finding probability just as bitcoin's difficulty linearly affects it's block finding probability.  So my chart is simply off by some constant factor -- the proportions are accurate but don't pay attention to the actual numbers.

(Edit: ignore the "Network Hashrate" chart for now.  See below for more info.)


Also... I've been mining this for days and I've got nothing... ARy4JFrdeKSst42iS3kcxwdB81t7Vmrvc7

[eule]

Very interesting, I expected a much higher "hashrate" increase given the rate at which blocks are solved. Looks like many many solo miners.  

[lucasjkr]

This in essence is not true, so you are jumping to conclusions a bit. Already at this moment, with many transactions set to no fees and minimal Bitcoin economy, transaction fees are already 1/2 BTC per new block in average. Go check yourself if you don't believe me. I can easily imagine, with growing BTC economy and reduction from 25 to 12.5 BTC per new block in 4 years, that transaction fees in every new block in 5 years can be worth more than worth of new coins. All that without forcing larger transaction fees to anyone. In 8 years (6.25 new coins per block) I have no doubt transaction fees will be worth more than new coins.

So ending supply of new coins probably wont have any effect on mining. Satoshi Nakomoto figured that one perfectly.

OK I'm checking...

24 hours per day, 60 minutes per hour, 8 minutes per block = 180 blocks per day = 4500 coins / day generated. And blockchain.info says the total transaction fees were around 30 BTC for the day. Miners are getting 4500 BTC for mining coins vs 30 BTC transaction fees the transaction fees generated, which is righ around what they were getting a year ago. Yes, increase in bitcoin value has helped quite a bit, but at the end of the day, the incentive to validate transactions is the coins that are mined, not the fees generated, which has stayed fairly stable and FAR below actual reward per block.

I don't know if Satoshi called that one right. The blockchain, the mining, yes... the cap on coins -  I simply have to disagree on that. Again, one of the constraining factors is that the velocity of bitcoin is fairly low, because people are holding them expecting them to increase in value, and therefore not generating transaction fees.

It's okay. We're allowed to disagree. That's why there's room for more than one blockchain...

[Sunny King]

So right now, network hashrate is just showing how many hashes/sec would be necessary to obtain that many blocks at that difficulty *for bitcoin*.  It's not useless.  The primecoin paper indicates that the primecoin difficulty linearly affects prime-chain finding probability just as bitcoin's difficulty linearly affects it's block finding probability.  So my chart is simply off by some constant factor -- the proportions are accurate but don't pay attention to the actual numbers.

The linearity I am referring to is the 'difficulty model'. For the difficulty number you see in getdifficulty, it's the target length, which gets exponentially harder. With current estimate length 8 target is ~30 times harder than length 7 target.

[canoe]

So right now, network hashrate is just showing how many hashes/sec would be necessary to obtain that many blocks at that difficulty *for bitcoin*.  It's not useless.  The primecoin paper indicates that the primecoin difficulty linearly affects prime-chain finding probability just as bitcoin's difficulty linearly affects it's block finding probability.  So my chart is simply off by some constant factor -- the proportions are accurate but don't pay attention to the actual numbers.

The linearity I am referring to is the 'difficulty model'. For the difficulty number you see in getdifficulty, it's the target length, which gets exponentially harder. With current estimate length 8 target is ~30 times harder than length 7 target.

Oi.  Thanks for pointing this out.  So the network hashrate chart is nearly useless after all at the moment.  Just ignore it for now until I fix.

All the other charts are fine though (block generation, difficulty, minting, total supply, transactions).

[n4ru]

Doing 20,000 PPS at the moment and no blocks in the last hour x_X

[mr_random]

I've added primecoin to my charts website:

http://cryptometer.org/primecoin_96_hour_charts.html
http://cryptometer.org/primecoin_90_day_charts.html

I left the "Network Hashrate" chart in even though it's a bit weird for this coin.  I haven't done the work yet to make it "Network prime-chains per second" as it should be.  I'd need to know the chances of a number being the origin for a Cunningham chain of length D (difficulty).  I'll need to do more research before working on that.

So right now, network hashrate is just showing how many hashes/sec would be necessary to obtain that many blocks at that difficulty *for bitcoin*.  It's not useless.  The primecoin paper indicates that the primecoin difficulty linearly affects prime-chain finding probability just as bitcoin's difficulty linearly affects it's block finding probability.  So my chart is simply off by some constant factor -- the proportions are accurate but don't pay attention to the actual numbers.

(Edit: ignore the "Network Hashrate" chart for now.  See below for more info.)


Also... I've been mining this for days and I've got nothing... ARy4JFrdeKSst42iS3kcxwdB81t7Vmrvc7

Hmm... Interesting trend on the money supply graph so far. Looks to be exponentially increasing which surely is not desirable. I wonder if the difficulty adjustment is working as intended.

[itod]

We are in need of a tool which will group found blocks by IP address, XPM address. That will show us how many users are there who get the most blocks.

Edit: There is a possibility to get info about IP if you are a node, is it?

[lucasjkr]

I've added primecoin to my charts website:

http://cryptometer.org/primecoin_96_hour_charts.html
http://cryptometer.org/primecoin_90_day_charts.html

I left the "Network Hashrate" chart in even though it's a bit weird for this coin.  I haven't done the work yet to make it "Network prime-chains per second" as it should be.  I'd need to know the chances of a number being the origin for a Cunningham chain of length D (difficulty).  I'll need to do more research before working on that.

So right now, network hashrate is just showing how many hashes/sec would be necessary to obtain that many blocks at that difficulty *for bitcoin*.  It's not useless.  The primecoin paper indicates that the primecoin difficulty linearly affects prime-chain finding probability just as bitcoin's difficulty linearly affects it's block finding probability.  So my chart is simply off by some constant factor -- the proportions are accurate but don't pay attention to the actual numbers.

(Edit: ignore the "Network Hashrate" chart for now.  See below for more info.)


Also... I've been mining this for days and I've got nothing... ARy4JFrdeKSst42iS3kcxwdB81t7Vmrvc7

Hmm... Interesting trend on the money supply graph so far. Looks to be exponentially increasing which surely is not desirable. I wonder if the difficulty adjustment is working as intended.

It's only day 5! Of course monetary base growth will look exponential, you're starting from zero.

Give it a few weeks or months and it likely will look a lot different,unless there is wholesale adoption of this coin immediately

[lucasjkr]

We are in need of a tool which will group found blocks by IP address, XPM address. That will show us how many users are there who get the most blocks.

Edit: There is a possibility to get info about IP if you are a node, is it?

There's what, 400,000 coins in existence? And the v1.1 binary for windows has 8000 downloads, plus linux users... If that makes it 10'000 on the network, the average person will have 40 coins/have found 2-3 shares

I've found quite a few more than that, but I started at the beginning... And even so, I've seen 24 hrs pass without a block anywhere, only to see 3 get found 3hours later.

But know, if the average is 3-4 blocks per person, and I've found more than that on most of the computers I'm running it on (6-7), then there'll be other people who have found fewer and others who have found none at all.

Remember, this isn't pooled mining. We're not compensated by out amount of work contribution alone, there's a huge amount of luck involved too...

[Sunny King]

I've added primecoin to my charts website:

http://cryptometer.org/primecoin_96_hour_charts.html
http://cryptometer.org/primecoin_90_day_charts.html

Minor correction, 'starting difficulty' is 7 for primecoin. Only genesis block has difficulty 6, which is also the minimum allowed.

[xcezzz]

CrypDough Dice Now Supports XPM!

http://XPMdice.crypdough.com/

Max Bet = 3 XPM

Max bet increasing shortly after everything works smoothly!

If you have any questions, comments, or issues please PM me! Or post into the feedback thread!

Sorry about downtime folks! Had an issue with double payouts! Back up and running. Let me know if you have any other issues/questions

[maco]

Seeking Community Feedback for the 'Primecoin Buzz ~ Mobile App' to release a new update.

Get involved at this thread: https://bitcointalk.org/index.php?topic=253060.0

[canoe]

I've added primecoin to my charts website:

http://cryptometer.org/primecoin_96_hour_charts.html
http://cryptometer.org/primecoin_90_day_charts.html

Minor correction, 'starting difficulty' is 7 for primecoin. Only genesis block has difficulty 6, which is also the minimum allowed.

I suppose there's an interesting distinction there.  I can get the software to show both a "Genesis Difficulty" and a "First Block Difficulty".

[maco]

These are nice charts.

I've added primecoin to my charts website:

http://cryptometer.org/primecoin_96_hour_charts.html
http://cryptometer.org/primecoin_90_day_charts.html

Minor correction, 'starting difficulty' is 7 for primecoin. Only genesis block has difficulty 6, which is also the minimum allowed.

I suppose there's an interesting distinction there.  I can get the software to show both a "Genesis Difficulty" and a "First Block Difficulty".

[AtomSea]

Pardon my ignorance - Is it true that Primecoin is only CPU mining?
Is there a way to GPU mine this?
I love the idea.  I have been waiting for funky coins like this to emerge - I want to mine them and help out the cause

[OnlyC]

Ok, here's the real fun story:


There had a man, who bought 10XPM at 0.1. But there have some guys, who do not want, selling at 0.003