Topic: [ANN][BLC] Blakecoin Blake-256 for GPU/FPGA With Merged Mined Pools Stable Net - page 190. (Read 409641 times)

I think it depends on your network connectivity. When the 0.8.6 patch was released on Sunday I rebuilt my blockchain from scratch on a new machine but the block chain forked! I was still running 0.8.5 on my other box, and the two chains were running in parallel. It took about six hours for the "new" chain to reconnect with the "old" one and the reorg orphaned around 50 blocks (two of which were mined by me, though the diff was lower so I probably didn't lose out).

I suspect this is an issue at the moment due to the low number of miners on the network, but perhaps some additional seed IPs in the config would help (I'm not on a static IP, so I can't volunteer here).

Anyway my single Lancelot FPGA board has mined a total 11 blocks since I started mining 24/7 four days ago, so I'm not the one sitting on the big pile of loot.

no idea how many but when EC2 had an outage we had the drop in difficulty?

like kramble said it is possible that someone could have a GPU miner already and is not releasing any source code or binaries.

but atm it is just me and kramble working on open source stuff for Blakecoin and it has only been released just over 1 week 

If I had done more of a pure clone and it not changed as much of the hashing to Blake, then Blakecoin could have been more compatible with the current software for Bitcoin e.g like LTC/QRK/SRC/CPR that have only really changed the POW rest still uses sha256d see: hash.h in the wallets for these changes, so I have created more work for myself and it will take as long as it takes to get it done and I am working on it every days so far since release 

but now we have 2 types for hashing on a network level SHA-256d and my Blake-256 mod. even if they use Blake-512 or Keccak-256 or any other hashing you can name for the POW they are still the same on a network point of view. (just checked Litecoin is still uses SHA-256d for network)

so if you have to rebuild your wallet data for Blakecoin it will be super fast compared with Bitcoin due to the changes at the network chain level e.g SHA-256(SHA256(data)) vs Blake-256(data) but this also effects all the checks and those are the things I am currently rewriting for p2p

I do not have any real control over what people are mining with or how they are finding block so quick 

Please be patient I will get p2p and the GPU miner done soon as I can   

From Litecoin wallet same as Bitcoin:

template
inline uint256 Hash(const T1 pbegin, const T1 pend)
{
    static unsigned char pblank[1];
    uint256 hash1;
    SHA256((pbegin == pend ? pblank : (unsigned char*)&pbegin[0]), (pend - pbegin) * sizeof(pbegin[0]), (unsigned char*)&hash1);
    uint256 hash2;
    SHA256((unsigned char*)&hash1, sizeof(hash1), (unsigned char*)&hash2);
    return hash2;
}

From Blakecoin wallet:

template
inline uint256 Hashblake(const T1 pbegin, const T1 pend)
{
    sph_blake256_context     ctx_blake;
    static unsigned char pblank[1];
    uint256 hash1;
    sph_blake256_init(&ctx_blake);
    sph_blake256 (&ctx_blake, (pbegin == pend ? pblank : static_cast(&pbegin[0])), (pend - pbegin) * sizeof(pbegin[0]));
    sph_blake256_close(&ctx_blake, static_cast(&hash1));
    return hash1;
}

as you can see it is totally different to the others, this is creating a lot of rewrites in functions in all software not just a change to the miner!

SHA-256d is not totally gone from Blakecoin it is just most of the bulk of the hashing is done in Blake now which seemed a good idea as it is more efficient and faster, but due to so many clones of other coins about everyone is expecting me to just be able to change a few lines of code and it works, well this is just not the case with Blakecoin and I do feel like I am under quite a bit of pressure to get it done quickly

Edit:
maybe I did not sleep well last night or not had enough tea but I am a little on the grumpy side this morning

I wanted just to make an experiment, created 1 88-ECU (32-CPU) instance, ran it for 3 hours - no blocks. So there should be hundreds of such instances...

using the reference cpu miner I have mined with 8 machines ~ 24MH/s = 4 blocks since the 10th so that is <1 block a day  
on kramble's FPGA miner ~ 60MH/s = 1 block in past 2 days  

when EC2 had an outage the other day the difficulty dropped to 1/5 of the stable rate it had been, so my guess is still on a large EC2 CPU miner  

seems like an interesting coin, I was about two days behind, blakecoin wasn't posted on the 3 forums I go to so I didn't even know about it til it was posted here. still, I would have thought with no pools, p2p or otherwise, I could have at least found one block. but that's the way it goes I guess, good luck mining...

I got quite a few blocks in the early days, but since the 12th, haven't seen anything. Only got one block on the 12th.

had same issue, i7, no blocks found at all. had no issue with qrk, src, or even cpr. as much hashing power as ive used, pretty strange

Ouch Seems like someone has a GPU miner or a ridiculous amount of cloud power.

Well after 6 days (almost 7 days now) mining with 3 x 8 core cpu's i give up with 0,0000000000 blocks found

My i7 3770k + amd fx8320 and my E5-2690 have not found 1 accepted single block

Yes i got many unconfirmed but all of them disappeared

Its ammazing that i see almost every 10 minutes a new block being found, however non are being found by anyone other then those special people getting them all

I still have the enormous amount of 12.5 coins which i got from the give aways

I try not to think how much the price in electra it has cost me, but it will not be pretty

p2pool has build issues on window atm and on Linux have only just started testing and is not fit for release yet, will continue to work on it once I get a few nodes working and shares paid I will do a release 

on the GPU miner side of things I have only had a brief look but would like to add it to bfgminer but if I get any build issues I will just extend the CPU miner to use OpenCL and the blake256.cl kernel, I could also mod poclbm to use blake256.cl but I am not so sure about reaper does not look that good 

will get something done this week for sure 

GPU miner and p2pool will come soon?
For me it's impossible solo mining with CPU.

yes the OpenCL is on my Github but just need some time to work on the miner, but finishing p2pool first

my todo list looks like this:

1. p2pool
2. fix any small bugs and help feeleep
3. GPU miner
4. try and fix any build issues on Linux and Windows


Blakecoin is fast and efficient and more on par with Bitcoin than any other Alt coin yet

1. No current SHA-256d Asic will work with Blakecoin so it is resistant it that sense, you would need to develop a new Blake-256 ASIC and as I have said before you are best off with FPGA if you wanted power efficiency. FPGA's also have a resale value as they are re-programmable! 
2. Why would you want to be GPU resistant did it work for Litecoin?
3. The strength of the Blake-256 hashing function is far more than needed 2^192 at least, it is also more efficient than SHA-256 

I think a small amount of inflation like what Blakecoin uses will help miner's who invest by buying or using equipment to mine Blakecoin so that over a period of time you can get your ROI, I see no good coming from cutting the subsidy it only makes it harder to plan a return on your investment. Also atm the inflation is so small it might be some years before it significantly adds to the base subsidy

Note that the inflation in Blakecoin is not fixed it is directly related to block height and the current difficulty 

What's the point of this coin? It is neither asic-resistant nor gpu-resistant.
Is the strength of this coin an efficient cryptographic hashing function?
I wonder if an inflational fork will succeed in the future.

I wonder how much mtrlt would charge to do a Blakecoin version of reaper....

Are you going to release a GPU mining software soon?

0.8.6 seems to be working great.

On another note... difficulty above equivalent 120000 for pre-0.8.6 difficulty calculation. :O

Yep, I've just updated (actually rebuilt the blockchain from scratch) and I've got

pi@raspberrypi ~/.blakecoin $ ./blakecoind getinfo
{
    "version" : 80600,
    "protocolversion" : 70001,
    "walletversion" : 60000,
    "balance" : 0.00000000,
    "blocks" : 4556,
    "timeoffset" : 0,
    "connections" : 4,
    "proxy" : "",
    "difficulty" : 372.53158654,
    "testnet" : false,
    "keypoololdest" : 1381670724,
    "keypoolsize" : 101,
    "paytxfee" : 0.00000000,
    "errors" : ""
}

Seems to be stuck at block 4556, while the main chain is currently at 4585 ...

pi@tvpi ~/blakecoin $ ./blakecoind getinfo
{
    "version" : 80500,
    "protocolversion" : 70001,
    "walletversion" : 60000,
    "balance" : 425.00091893,
    "blocks" : 4585,
    "timeoffset" : 0,
    "connections" : 10,
    "proxy" : "",
    "difficulty" : 110001.58180514,
    "testnet" : false,
    "keypoololdest" : 1381152051,
    "keypoolsize" : 101,
    "paytxfee" : 0.00000000,
    "errors" : ""
}
pi@tvpi ~/blakecoin $

So I think I've just found my first orphan  ...

Share found on Sun Oct 13 14:37:12 2013 nonce 22fe248f
submitting 00000070738244283d3cc47ade94a2ea6cfaeb16e8ee1696c1ef5c4e00915c7300000000610ce62 0baf71d95151bc3f0e4c29d5cc70893831bb1de68c4b20a9140658778525aa1f51c00afeb8f24fe 2200000080000000000000000000000000000000000000000000000000000000000000000000000 0000000000080020000
hash f6822e0ca507c17bfe382db6e519a0fcd17444dd13d1e9f59be9530000000000
Upstream result: True
[1 accepted, 8 failed]

yes the difficulty was displaying incorrectly it was difficulty/256 it should be fixed now

I updated to 0.8.6 and see this:
Difficulty : 372.53158654