[ANN][RIC] Riecoin: constellations POW *CPU* HARD FORK successful, world record

[primer10]

Further updated the RIC explorer to leverage notsoshifty's code and display the prime.

To avoid cluttering things, only the base prime (n+0) of the constellation is shown.

I'll make all the primes available as a simple zipped text file later on (ie. when the db is ready).

Ahh I see. Well since your block explorer seems to have everything for Riecoin addresses you get the bounty. Please post your address on here and it will be sent in 1-2 days.

Thanks   RL9Ldk7EKHBanArEmYznLdGjTZHvazUXjb

Great job!! Thanks!

[fairglu]

For the curious and those willing to validate, here is a text file "database" of the first 20,000 constellations (n+0 primes):

Riecoin_Primes_20000.csv

(the rest is still being computed, currently around 24000)

[istudy92]

https://wiki.ubuntu.com/TrustyTahr/ReleaseNotes

Upgrading ubunto version would not affect miner would it from 13.10--> 14.04 LTS?

[GreekBitcoin]

the qt takes about 1 minute to start and downloading blockchain is rather slow. Any fixes for those two things?

btw what addnodes do you use?

[dga]

https://wiki.ubuntu.com/TrustyTahr/ReleaseNotes

Upgrading ubunto version would not affect miner would it from 13.10--> 14.04 LTS?

I haven't tested, but it should be safe.  If it breaks any of the binaries (unlikely!), you can recompile from source.  Let me know if anything's weird and I'll fix it fast.

[mumus]

Sentiment poll -

I have the linux version of the next iteration of my riecoin miner.  It'll substantially reduce memory use on many platforms, and is a bit faster, and a better long-term building block.

It doesn't work on windows and I'm having a heck of a time debugging it.

Is there any sentiment towards releasing it where it only works on Linux and letting more people stare at what's wrong with windows vs. kicking at it longer and trying to fix it to have both win/linux ready at the same time?

I feel home in Windows and I love Linux and it's challenges. My vote is to release it and I'm sure the community will find the problem with the windows fast. I personally would also try to fix it.

[dga]

Hey, Gatra -

Would it be insane to think about adding a "fast revalidate" mode to the client for startup, in which it randomly picks $k$ of the 6 primes to test for primality, and assumes that as long as nothing in the blockchain fails the check on any of those $k$, that it's probably OK to accept them and keep running?

I'm thinking just for client initial startup, obviously.  Doing it randomly prevents an attacker from getting away with fake blocks for any significant length chain, and it could reduce the number of needed Fermat tests down to 2 or 3 instead of 6.

[trny]

Ok, the blockchain explorer is now fully synchronized!

• The nonstandard transaction was an isolated case (search for "nonstandard" in the explorer to see it or check if other appear in the future)
• The multisig address that was mentioned previously is now present, it was in the blocks after the nonstandard one.
• I've added the two active pools I could find (YPool & UpCPU), though extraction stats are not very interesting right now as YPool just dominates with 95%.

Feel free to test and check things to see if everything else is okay!

Very nice implementation fairglu! Now we can finally see complete address contents. Quick question: does your explorer use the rpc interface or parse from block files?

[gatra]

I'm currently testing the new client. It comes with some speed improvements, but it's a scary thing since a bug there can cause the coin to halt. I'm being extra careful.

*gulp* can you explain a bit more? Could this affect us now, with the new client not yet launched?

no, no, there are no bugs that I know of!
 I should have used the conditional: I meant that if I introduced a bug in the new code it would be bad so I want to test the new code carefully. Current versions work fine!

[gatra]

Hey, Gatra -

Would it be insane to think about adding a "fast revalidate" mode to the client for startup, in which it randomly picks $k$ of the 6 primes to test for primality, and assumes that as long as nothing in the blockchain fails the check on any of those $k$, that it's probably OK to accept them and keep running?

I'm thinking just for client initial startup, obviously.  Doing it randomly prevents an attacker from getting away with fake blocks for any significant length chain, and it could reduce the number of needed Fermat tests down to 2 or 3 instead of 6.

I think it's possible, not insane...
Everytime you run the client it validates last blocks in the disk and then downloads any new blocks from the network.
By default we are currently validating the last 288 blocks from the disk: this is done just to detect curruption and will be changed down to 7 for the next version. Anyway you can configure it using the -checkblocks option, set it to 1 and it will start much faster!
For the initial download, it would be possible to randomly skip some tests. Currently we are doing too much tests: 10 rabin-miller iterations for each of the 6 candidates. I'm lowering it and maybe adding a command-line option to configure it. This will improve validation time a lot. I think it is not needed so skip tests. If it were, we could do it in the future using your $k$ idea or maybe assigning a configurable probability so each candidate of each block is tested with probability $p$ .... or some combination of that ...

[fairglu]

Very nice implementation fairglu! Now we can finally see complete address contents. Quick question: does your explorer use the rpc interface or parse from block files?

Thanks!

I'm using the RPC interface, this way I'm more independent from the implementation details, and the blockchain explorers are hosted on a different machine than the clients are, so a remote protocol would be needed anyway (I'm tunneling the RPC however, clients are firewalled and the RPC ports are closed from the outside, as I don't entirely trust the security of the RPC)

[fairglu]

On a side note, it would be nice to see crypto currency clients move away from that LevelDB slowpoke. If you ever mined a coin for tens of thousandths of blocks at a P2Pool, you'll know what I mean...

Anyway, I've got the 40,000 first constellations available in simple text file format, will update when I get back to a PC, and probably setup a small dedicated page explaining the constellations and with misc info (largest prime, smallest primes, etc.). Any ideas of useful trivia that could end up there?

Also any ideas of math/science sites where to announce that page? Could be a good opportunity to draw attention.

[primer10]

any chances of accelerating dev of GPU miner for RIC now that there is an open source GPU miner for XPM: http://www.peercointalk.org/index.php?topic=2668?

[northranger79510]

Further updated the RIC explorer to leverage notsoshifty's code and display the prime.

To avoid cluttering things, only the base prime (n+0) of the constellation is shown.

I'll make all the primes available as a simple zipped text file later on (ie. when the db is ready).

Ahh I see. Well since your block explorer seems to have everything for Riecoin addresses you get the bounty. Please post your address on here and it will be sent in 1-2 days.

Thanks   RL9Ldk7EKHBanArEmYznLdGjTZHvazUXjb

Bounty Sent Took me a while to recover them from my storage.

[vancsj]

Hmmm, blockchain explorer is currently stuck on block 26246, which has a single transaction, f16662cadd50ccf58b585938a25810913acb05f6adaf05ab9cbdc3f84c9ab688, which has the following single vout:
lol someone has got my corrupted block:)
There's a little bug in one of my pool, and it produces 5 or so nonstandard tx like this, my precious 250 RIC!

Code:

      "scriptPubKey": {
             "asm": "7692969 OP_BOOLOR [error]",
             "hex": "03a962759b44e46e064da089240f13c2c9da959816e71bba7a2a0f46ed1e8d82e9",
             "type": "nonstandard"
      }

Anyone knows what's up?

The block crawler also goes banana when looking at it.

I could just tell the explorer to ignore it and go ahead, but that sounds wrong

Nonstandard transactions are, well, non-standard, but valid. In those transactions you can't determine the receiver's address. By manual inspection I think that in this particular case the script makes no sense so those 50 RIC cannot be claimed and are lost, or burnt.
As a quick fix, you can ignore them. Later you can accumulate and report RIC on nonstandard tx

[fairglu]

Bounty Sent Took me a while to recover them from my storage.

Got it, thanks

Updated the explorer page with a downloadable text file version of the first 40000 constellations. I'll see about a more interesting trivia page once Easter is over.

[northranger79510]

Just a quick update but I am working on mini android apps for Riecoin to put more coverage for android users. I will release those next week. Android app for Riecoin is still ongoing. Ran into some problem and will try to fix.

[northranger79510]

This is the error I am getting with bitcoin android wallet when I tried to compile so if someone can help, that'd be great. I am still learning how this android app thing work.

This is the error I am getting.. Dex Loader] Unable to execute dex: Multiple dex files define Landroid/support/v4/app/_ActionBarSherlockTrojanHorse$OnCreateOptionsMenuListener;

After I fix that, I get this: Conversion to Dalvik format failed with error 1

I tried to delete android_support_v4.jar inside my project library but still doesn't work.

Help appreciated. Been looking for a solution and it is driving me crazy.

[northranger79510]

Remember to like our Riecoin Facebook site and join Reddit, Twitter, and other social media website containing Riecoin news!.

[northranger79510]

Bump.