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

[kelsey]

It's a smart idea, but mathematically, Primecoin is on a race to the bottom, so I have to play a little Game Theory.

$40 for just typing setgenerate true 3 days ago is clearly unsustainable, sooo http://www.youtube.com/watch?v=HX0fIi3H-es

Of course. But think about this: the harder they get to generate, the more they may be worth. So price may go UP, not down.

I'm with Wolf0 on this one, anyone 'dumping' their XPM now because they think the price is going to drop is not understanding the exponential difficulty increases that lye ahead for primecoin.

or they merely understand what lack of liquidity does to a market.

[1715055719]

1715055719

[1715055719]

1715055719

[1715055719]

1715055719

[Buffer Overflow]

There is no real hashrate, because the proof-of-work is not based on hashes. The best equivalent is the "primespersec" line. ~200 seems to be a level for a desktop processor.

I'm pushing out 5500 pps, but no block for 2 hours now.

Is this on a single PC?

No.

[Buffer Overflow]

There is no real hashrate, because the proof-of-work is not based on hashes. The best equivalent is the "primespersec" line. ~200 seems to be a level for a desktop processor.

I'm pushing out 5500 pps, but no block for 2 hours now.

With the block spacing ~12-15 seconds and the default gensieveroundlimitms set to 1 second, each miner has very little time to find a chain. Consider that: some time can be lost due to peer lag in being notified of a new block; a small amount of time is lost validating the pseudoprimality of the previous block; then the remaining time is usually lost cycling through thousands of unsuccessful sieves.

All the miners have the equal disadvantage, so this is irreverent, unless someone is using a private optimized version not released.

I also run my primecoin daemons with the datadir in ram memory to reduce orphans. Every millisecond counts to get a solved block out to the network, don't want any messing about with slow hard drives.

[tonyback]

how to see my amount and address ?

how to see my speed ?

thanks

[altsay]

how to see my amount and address ?

how to see my speed ?

thanks

To see your balance:

primecoind getinfo

Adress:

primecoind getaccountaddress <your rpcuser in the primecoin.conf>

Speed:

primecoind getmininginfo

"primespersec" is your speed.

[AgentME]

It looks like the primespersec value is calculated by counting up how many primes are found in one minute (with some minor corrections if the time between primespersec calculations isn't exactly one minute), and then resetting the primes counter. It would be much more accurate if it either operated on a longer time period, or even just never reset the primes counter.

[Remember remember the 5th of November]

@Sonny King

Primecoin as it is now, does it enable custom-made miners that use getwork to submit solutions?

[vinne81]

Blocks take a LONG time to mature, "primecoind getbalance" is always lagging behind!

If your balance stays at 0, use "primecoind listtransactions" so you can see maturing blocks (if you have them)

[Stinky_Pete]

Just wondering - is anyone developing a pool?

[8bitPunk]

There is no real hashrate, because the proof-of-work is not based on hashes. The best equivalent is the "primespersec" line. ~200 seems to be a level for a desktop processor.

I'm pushing out 5500 pps, but no block for 2 hours now.

With the block spacing ~12-15 seconds and the default gensieveroundlimitms set to 1 second, each miner has very little time to find a chain. Consider that: some time can be lost due to peer lag in being notified of a new block; a small amount of time is lost validating the pseudoprimality of the previous block; then the remaining time is usually lost cycling through thousands of unsuccessful sieves.

All the miners have the equal disadvantage, so this is irreverent, unless someone is using a private optimized version not released.

I also run my primecoin daemons with the datadir in ram memory to reduce orphans. Every millisecond counts to get a solved block out to the network, don't want any messing about with slow hard drives.

I'm getting a lot of orphans (~1 per hour) so I've been comparing debug.log with blockchain records based on UTC time... In most cases I'm ~1 second late so peer lag is to blame. But a few cases have been within the second, so you're absolutely right that efficiency getting solved blocks out is key.

[shinkicker]

[Stinky_Pete]

1 block now less than 17 XPM. I miss the days when it was over 20 <sigh>

[masterOfDisaster]

Hi,

has anybody tried to compile primecoind him/herself on linux (in my case Ubuntu 12.04)?
I tried to do so, installed the required packages and as recommended "BerkeleyDB.4.8" (I followed primecoin-0.1.0-linux/src/doc/build-unix.md)
The first attempt to run make ended with errors regarding db_cxx.h (db.h:14:20: fatal error: db_cxx.h: No such file or directory).
I set the variables regarding the BerkeleyDB that are used by makefile.unix to
BDB_LIB_PATH=/usr/local/BerkeleyDB.4.8/lib/ and
BDB_INCLUDE_PATH=/usr/local/BerkeleyDB.4.8/include/
without success.
make still ends with errors and I might be too newbish to find out what's wrong...
Can anybody help?

Kind regards
masterOfDisaster

<edit>
If someone could post IPv6 addresses of nodes, that would be great, too. Maybe the IPv6 flag has been used for compiling primecoind and I only lack of nodes...
</edit>

[Remember remember the 5th of November]

Just wondering - is anyone developing a pool?

Right now I don't believe Primecoin has any protocol defined for custom miners, let alone a pool.

[AgentME]

Blocks take a LONG time to mature, "primecoind getbalance" is always lagging behind!

If your balance stays at 0, use "primecoind listtransactions" so you can see maturing blocks (if you have them)

Use this to show your balance including immature funds:

Code:

primecoind getbalance "*"

[8bitPunk]

It looks like the primespersec value is calculated by counting up how many primes are found in one minute (with some minor corrections if the time between primespersec calculations isn't exactly one minute), and then resetting the primes counter. It would be much more accurate if it either operated on a longer time period, or even just never reset the primes counter.

In debug.log you will see a more meaningful prime/h which updates at 2 minute intervals.

edit: I was wrong primes/h is calculated as dPrimesPerMinute * 60.0

[vinne81]

Blocks take a LONG time to mature, "primecoind getbalance" is always lagging behind!

If your balance stays at 0, use "primecoind listtransactions" so you can see maturing blocks (if you have them)

Use this to show your balance including immature funds:

Code:

primecoind getbalance "*"

Thx, that works!

[maka]

I've noticed some of the new "release" versions by helpful people may have minor bugs that it won't notifying you even if you mine a block, neither does it show the hint on the UI except you may find the immature amount increased.

To make sure you know how many blocks you have mined, use "listtransactions" to see the details.  This way there is no confusing messages.  

[AgentME]

It looks like the primespersec value is calculated by counting up how many primes are found in one minute (with some minor corrections if the time between primespersec calculations isn't exactly one minute), and then resetting the primes counter. It would be much more accurate if it either operated on a longer time period, or even just never reset the primes counter.

In debug.log you will see a more meaningful prime/h which updates at 2 minute intervals.

This message?

Code:

printf("%s primemeter %9.0f prime/h %9.0f test/h\n", DateTimeStrFormat("%Y-%m-%d %H:%M:%S", nLogTime).c_str(), dPrimesPerMinute * 60.0, dTestsPerMinute * 60.0);

That's just the same counter of primes found in the last minute extrapolated to an hour.

[oroqen]

Hi,

has anybody tried to compile primecoind him/herself on linux (in my case Ubuntu 12.04)?
I tried to do so, installed the required packages and as recommended "BerkeleyDB.4.8" (I followed primecoin-0.1.0-linux/src/doc/build-unix.md)
The first attempt to run make ended with errors regarding db_cxx.h (db.h:14:20: fatal error: db_cxx.h: No such file or directory).
I set the variables regarding the BerkeleyDB that are used by makefile.unix to
BDB_LIB_PATH=/usr/local/BerkeleyDB.4.8/lib/ and
BDB_INCLUDE_PATH=/usr/local/BerkeleyDB.4.8/include/
without success.
make still ends with errors and I might be too newbish to find out what's wrong...
Can anybody help?

Kind regards
masterOfDisaster

<edit>
If someone could post IPv6 addresses of nodes, that would be great, too. Maybe the IPv6 flag has been used for compiling primecoind and I only lack of nodes...
</edit>

I use gentoo and have too add "BDB_INCLUDE_PATH=/usr/include/db4.8/ BDB_LIB_SUFFIX=-4.8" too my make command