[ANN] Kryptohash | Brand new PoW algo | 320bit hash | ed25519 | PID algo for dif

[UnicodeEncoding]

Anything happened? The net hashrate is pretty low,only 3.2MH/s,diff is only 30.

[wr104]

Both hashrate and difficulty are now steadily increasing.  

I'm checking the logs and I don't see indications of a fork or other network related issues.  I'm wondering if there was an issue with the pool.

[WORE]

Both hashrate and difficulty are now steadily increasing.  

I'm checking the logs and I don't see indications of a fork or other network related issues.  I'm wondering if there was an issue with the pool.

They probably upgraded.

[berbip]

Both hashrate and difficulty are now steadily increasing.  

I'm checking the logs and I don't see indications of a fork or other network related issues.  I'm wondering if there was an issue with the pool.

I've been mining on the pool for the last couple of days and had no issues. Just enjoying the low diff.

[BoscoMurray]

Does anyone know of a CPU miner being developed for this krypto? I'd like to mine but have no GPU rigs just now.

[wr104]

Does anyone know of a CPU miner being developed for this krypto? I'd like to mine but have no GPU rigs just now.

Unfortunately, CPU mining is practically impossible with this algorithm, even at current low diff levels.

You could still try to mine using CPUs but, you'll probably have more chances of getting hit by a meteor while reading this posting than resolving a block using CPUs.

About GPUs, I'd go with AMD for now. Their GPUs are loaded with ALUs which is great for PoW algorithms that calculate crypto-hashes like this one.  

[Undead_Phenix]

The block explorer stuck weeks ago.

[wr104]

The block explorer stuck weeks ago.

Which explorer? It is probably still running the obsolete version 0.4.x

[Undead_Phenix]

The block explorer stuck weeks ago.

Which explorer? It is probably still running the obsolete version 0.4.x

Yes,you are right.
http://bitexp.net/explorer.php?coin=Kryptohash

[berbip]

what is the correct syntax for two different cards ? I would like to add a 270 to my 280x

[wr104]

what is the correct syntax for two different cards ? I would like to add a 270 to my 280x

Assuming the 270 is seen by cgminer as the second GPU on the system:

  --shaders 2048,1280 --shaders-mul 8,4

Start with '--shaders-mul' equals to 4 and then increase this value until you find the best performance.  Keep an eye on the temperature.

Also, beware of the temperature mapping issue with cgminer. Unfortunately, there is no way to automatically correlate in code what OpenCL thinks it is GPU0 with what ADL reports as GPU0. The way to tell for sure is by disabling one GPU in cgminer (using --d 0 for example) and then watch the temperatures in the status bar.  

If you got an incorrect mapping, you need to use the '--gpu-map' option.  Since you have only 2 cards, it is straightforward: --gpu-map 0:1,1:0

[wand_reader]

what's the status of android wallet development?

[wr104]

what's the status of android wallet development?

It is finally coming along.  

Porting the existing BTC wallets to support all the KHC features (320bit hashes, ed25519 signatures, etc) has been harder than what I originally expected.  

[wand_reader]

what's the status of android wallet development?

It is finally coming along.  

Porting the existing BTC wallets to support all the KHC features (320bit hashes, ed25519 signatures, etc) has been harder than what I originally expected.  

Great !Can't wait to use the android wallet.

[WORE]

what's the status of android wallet development?

It is finally coming along.  

Porting the existing BTC wallets to support all the KHC features (320bit hashes, ed25519 signatures, etc) has been harder than what I originally expected.  

Hey man!  I hope things are going well and not to poke a stick or anything but KHC is trading at 60-80 sats and doing so under the new more power hungry algo, that would be <100.  I happen to be single for a while now and biting my fingernails is not exactly an attractive habit, dude please, for the love of... err... self... if you have a wallet on the mobile side close please beta test it.  There ain't shit out there that can compete with the algo, gotta strut some of ur stuff man.

[wr104]

what's the status of android wallet development?

It is finally coming along.  

Porting the existing BTC wallets to support all the KHC features (320bit hashes, ed25519 signatures, etc) has been harder than what I originally expected.  

Hey man!  I hope things are going well and not to poke a stick or anything but KHC is trading at 60-80 sats and doing so under the new more power hungry algo, that would be <100.  I happen to be single for a while now and biting my fingernails is not exactly an attractive habit, dude please, for the love of... err... self... if you have a wallet on the mobile side close please beta test it.  There ain't shit out there that can compete with the algo, gotta strut some of ur stuff man.

I hear ya!

Sometimes, I wish I didn't go this far with all the enhancements and features that I added to KHC, features that only a handful of people would appreciate anyway.  
With pretty much any other alt-coin over here, you can just grab someone else's mobile wallet source code, change the name, the logo, do few tweaks here and there and voilà! You have a brand new Android Wallet.  This isn't the case with KHC.

To make things worse, I was originally porting the well-known Mycelium BTC wallet to KHC. Once I was almost done making changes and I began testing it, I realized that by design it won't connect to  the P2P network like a normal wallet does. Instead, it connects to special backend servers running some WAPI services for the transactions.  This is a big "NO" for KHC since, I want all my Wallets to use ONLY the P2P network.

I'm now working on one based on Schildbach which doesn't require a centralized backend. Hopefully, I can have a Beta .apk in the next two weeks.

[wand_reader]

The diff is only 28 now

[Videlicet]

What's your reasoning dev for using 320 bit hashing rather than 256 bit?

Do elaborate, I love details 

[wr104]

What's your reasoning dev for using 320 bit hashing rather than 256 bit?

Do elaborate, I love details  

I would like to elaborate more but, here are the 2 main reasons.

1) To optimize the KSHAKE PoW algorithm in OpenCL (*).
2) Reduce the chances of a collisions even more (**).


(*) Earlier during development, I was using 256 bits for KSHAKE. To optimize the PoW algo in OpenCL, the block header size needed to be 136 bytes long (Keccak Rate = 1088, divided by 8 ). This meant wasting lots of space in the blockchain.  
On the other hand, using 320 bit requires Keccak Rate = 960 which translates to a block header size of 120 bytes.  This allowed me to add few extra fields to the block header without wasting space since, there is a couple fields that could allow for future enhancements.

Code:

        struct _block
        {
            int nVersion;            //  4 bytes
            int nRegion;             //  4 bytes
            uint320  hashPrevBlock;  // 40 bytes
            uint320  hashMerkleRoot; // 40 bytes
            int64_t  nTxTime;        //  8 bytes
            uint64_t nHashCoin;      //  8 bytes (currently unused)
            uint32_t sigchecksum;    //  4 bytes (currently unused)
            uint32_t nBits;          //  4 bytes
            uint32_t nTime;          //  4 bytes
            uint32_t nNonce;         //  4 bytes
        } block;

 


(**) My motto is: Higher security is always better.
People in crypto believe that a 256bit hash is good enough for today but, how about 25 years from today?
If the difficultly increases too much, there could be enough leading zeroes in the 256bit hash that can cause collisions to appear. What do you think would happen to your coin once you get hashes pointing to more than 1 transaction?  
Also, most people just see that 320 > 256. At the end of the day this is what counts.

[Xevious75]

This coin really has a lot of potential! I'm putting all my gpu firepower into it!