Nexus - Pure SHA3 + CPU/GPU + nPoS + 15 Active Innovations + More to Come

[Videlicet]

go6ooo1212,

I begin work on an open source GPU miner for the CPU channel

GPU Channel Activates at Unix Timestamp 1413914400. GPU Blocks are Rejected until then.
Viz.

[1715015220]

1715015220

[Supercomputing]

Work on a GPU miner for the GPU channel would seem more fitting as logic goes [GPU Channel Launch in ~2 weeks]. I'm sure the little CPU miners with few cores such as myself would appreciate a reasonable difficulty level for their amount of computing power.

Not all of us have Supercomputers
Viz.

@Videlicet

Do you have an existing code base (OPENCL/CUDA) for the GPU channel that I can review? Is the GPU channel activated on testnet?

[Videlicet]

Supercomputing,

Yes sir I do have a code base [OPENCL] that Bitslapper has been working on. And yes GPU channel is Active on Testnet. You'll need to use the command SetChannel(2) in your mining LLP to tell Coinshield what type of block you will be working on.

Thank You,
Viz.

[phzi]

Work on a GPU miner for the GPU channel would seem more fitting as logic goes [GPU Channel Launch in ~2 weeks]. I'm sure the little CPU miners with few cores such as myself would appreciate a reasonable difficulty level for their amount of computing power.

Not all of us have Supercomputers
Viz.

This is just silly.  If an algorithm is more efficient on a GPU, then of course someone will be mining on their GPUs.  Better that the miner is publicly released instead of only privately available.

CPU mining is kinda silly anyway, and mainly works to encourage botnets.  

---

As for people commenting on insufficient marketing and a lack of exchanges, I think this is to be fully expected.  Nobody in their right-mind should trust closed source cryptography.  Until the CoinShield wallet source code is released, this project is simply a bunch of promises with no proof, and no good exchange is going to adopt it.  Providing source specifically to exchanges is equally stupid - why would you trust them but not the masses, and how would they know the source code is what client's are running unless there is a reproducible binary build process (which I am sure Viz has not created).

Viz's continued claim that the source code is private to keep people from cloning his work is an extremely poor excuse.  One of the base ideas of this project was to absorb the economies of blatant coin clones.  If he had any faith in the idea, I don't see why he would be even slightly threatened by people using his work.  I'm sure he's using other people's GPLv3 licensed work - CoinShield certainly wasn't built from scratch.

[KryptoKash]

Shkembe,

As I've mentioned, the best source of information is coinshieldtalk.org. You will find fresh updates on the chatbox, I am rarely here on Bitcointalk anymore. Pool is almost done, in final testing phases. Keep in mind I cannot take any existing pool software to build a Coinshield pool, so I am coding it from scratch. This doesn't happen overnight.

Exchanges will be given source code upon their interest to add us, nothing more.
Viz.

At coinshieldtalk.org you have 92 posts vs 1270 here, I think you can take us more serious. As I see you have no intention to promote your coin and perhaps you expect help from the comunity to focus exchange's interest to it. If yes just tell us. Personally I am thinking to abandon this coin.

We are currently building all our infrastructure. Unlike other coins (which some people have become accustom to), quality is not sacrificed for quantity (If this was the case, we could have released 11 coins each with 1 new feature). We have reached out to a few exchanges which were recommended to us on our chatbox, however please reach out to anyone about us (exchange or not) help spread the word about stopping scam/clone coins, protecting good coins, and about the most secure coin with 11 new features that is ONLY 15 days old.

This is what I outlined previously as re-quoted by viz:

@Viz, whats your marketing plan? We need to build a strong community if you want this project succedeed

In the next few hours the forum (coinshieldtalk.org) will be going offline for some updates.

Unlike other coins, CoinShield's value directly stems from our community. The more people vote, comment, create petitions, and exchange shitcoins for CSD the more CSD's value will rise.

CoinShield is the first ALT. coin made for miners and investors who are tried of owning non-Verified coins, supporting non-Verified coins, being scammed by non-Verified coins, and losing money on non-Verified coins, Our internet marketing won't focus on trying to persuade the crypto community that we are the best non-centralized currency in the world (we all know that is bitcoin). Instead it will be focusing on introducing CoinShield to:

• the entire mining community
• all the pool operators
• all the ALT. coin investors

Once CoinShield is fully operational (all channels are launched, voting and exchange are ready) we will be buying adspace on various sites (bitcointalk, coinmarketcap, etc.) to help inform the community. We may incentivise users to add CoinShield on their signature here on bitcointalk. Currently we are building our followers on twitter to let the community know what coins are being killed in real time (follow us now if you currently do not: https://twitter.com/CoinShieldTalk).

I feel our job is to inform the community and keep them up to date with what we are killing and what is being voted on. Internet Marketing that focuses on anything else is just to create hype and make a bubble. We want CoinShield to be a stable economy where wealth can be built while clones/scams can be killed.

With this said, If you have any ideas on what else you think we should do please PM me or post here.

With that being said, Viz does post on our chatbox on a regular basis for those who are following it.

My thanks go out to everyone who is keeping up with us as we grow and build Coinshield.

Cheers,
KryptoKash

[vedran82]

Viz, so as we know, GPU miner already exists for CPU channel. At least one dev (Christian) already has it. Hats off to them, knowledge is power, if I was good enough, I would make one but I'm still far from it.
But why don't you make one and make it public then? I mean, it's more fair than just let few people have it.
Way better than "we won't work on GPU miner for CPU channel".
Just my personal opinion.

[phzi]

Viz, so as we know, GPU miner already exists for CPU channel. At least one dev (Christian) already has it. Hats off to them, knowledge is power, if I was good enough, I would make one but I'm still far from it.
But why don't you make one and make it public then? I mean, it's more fair than just let few people have it.
Way better than "we won't work on GPU miner for CPU channel".
Just my personal opinion.

The multi-channel approach and trying to offer a CPU channel is failing pretty badly so far.  The original promise was to "make the distribution as fair as possible".  If that was really going to be achieved, CPU mining is a very poor choice, GPU miner should have been released BEFORE launch (I talked about this before the coin was launched) to allow time for optimizations, and Viz would be taking a smaller dev fee.

Instead, the coin launched with a severely un-optimized CPU miner, and a PoW algorithm that is actually fairly efficient to run on GPUs.  But, only a handful of people know how, so the distribution is almost certainly skewed to those select few.

[Videlicet]

Phzi,

You are correct - nothing can be proved until source code is released. We will all be the judges then [until this time both sides are only opinion working their ways to veracity]. And yes, it was created from Peercoin Core, which is partially [Satoshi's Vision for Another] why I will continue this trend.

Everything comes in time my friend  
Viz.

edit: added more detail

[CryptoHobo]

Thanks for the info. Another method is to add the code below to the header file and then replace all instances of unsigned long with word_t in the source files.

#ifdef _WIN64
  typedef unsigned long long int word_t;
#else
  typedef unsigned long int word_t;
#endif

I added this to core.h (presuming that was the correct header), then did a search and replace through the other source files. Seems to be running better on my windows 8 system but crashes on the windows 7 system this could be because since my last build ive removed and reinstalled cygwin then built gmp from scratch or because i've bodjed the code lol

[Videlicet]

CryptoHobo,

Add this in your util.cpp
uint64 mpz2uint64(mpz_t n)
{
    unsigned int lo, hi;
    mpz_t tmp;

    mpz_init( tmp );
    mpz_mod_2exp( tmp, n, 64 );

    lo = mpz_get_ui( tmp );
    mpz_div_2exp( tmp, tmp, 32 );
    hi = mpz_get_ui( tmp );

    mpz_clear( tmp );

    return (((uint64)hi) << 32) + lo;
}

Add this to your util.h
uint64 mpz2uint64(mpz_t n);


PrimeSieve in prime.cpp:
delete nonce = mpz_get_ui(zTempVar);
add nonce = mpz2uint64(zTempVar);


And at top of PrimeSieve in prime.cpp:
delete unsigned long nonce = 0;
add uint64 nonce = 0;

Worked for me
Viz.

[CryptoHobo]

sweet thx viz i'll give it a go now

[Videlicet]

CryptoHobo,

Not a problem, let me know if you have any more issues.

Take Care,
Viz.

[CryptoHobo]

getting a build error ...

Code:

In file included from util.cpp:1:0:
util.h: At global scope:
util.h:36:19: error: ‘mpz_t’ was not declared in this scope
 uint64 mpz2uint64(mpz_t n);
                   ^
util.cpp:13:19: error: redefinition of ‘uint64 mpz2uint64’
 uint64 mpz2uint64(mpz_t n)
                   ^
In file included from util.cpp:1:0:
util.h:36:8: error: ‘uint64 mpz2uint64’ previously defined here
 uint64 mpz2uint64(mpz_t n);
        ^
util.cpp:13:19: error: ‘mpz_t’ was not declared in this scope
 uint64 mpz2uint64(mpz_t n)
                   ^
makefile.unix2:44: recipe for target 'build/util.o' failed
make: *** [build/util.o] Error 1

PrimeSieve in prime.cpp:
delete nonce = mpz_get_ui(zTempVar);
add nonce = mpz2uint64(zTempVar);


And at top of PrimeSieve in prime.cpp:
delete unsigned long nonce = 0;
add uint64 nonce = 0;

couldn't find these assumed typo for nNonce?

edit:
got it added #include <gmp.h> to util.h , succesful compilation.
crashes at same point, just after generating primes problem must lie with the build environment
i'm out for the night  thanks for the help

[cbuchner1]

Supercomputing,

I found out why it is missing shares. You are updating the nNonce with mpz_get_ui [unsigned long] when the nNonce is [unsigned long long]. Linux interprets this as 8 bytes, but not Windows 7, this give false primes.

The unsigned long problem would only affect 32 bit builds with gcc (any platform) and 64 bit builds with Visual C++ (Windows only).
I posted about the issue with Visual Studio/Visual C++ a few pages ago.

64 bit builds with gcc should be OK (any platform), it will always use 64 bits longs.

Avoid 32 bit builds like the plage, the lower register count will kill your performance.

Christian

[Videlicet]

cbuchner1,

Yes - Good Work. I did though notice this issue still on Windows x64 builds on Windows 7. Tested it with a couple different PC's from a couple different users.


-----------------------------------


CryptoHobo,

My apologies you need to add #include <gmpxx.h> in your core.h file.

Viz.

[KryptoKash]

Core Mandatory Update: 1.0.1 Time Dependent Core Updates go live on Unix Timestamp: 1412964000


Ubuntu
Windows



Change Log:

Updates:
A. Clear Unified Time Database necessity, so it starts with fresh seeds for clock drift
B. Verify 1.0.1 updates in Coinbase TX
C. Change Protocol.h Ports for Testnet
D. Change Activation Timestamps
E. Remove Debug Output For Mining LLP and Core LLP
F. Remove 1 connection per IP group
G. Fix Unified Time to Gather Seeds Quickly if not Unified and Fails to grab Seeds
H. Change Block Trust to 3 for New Channel break Streak, and 5 for POS
I. Update Checks when Loading Database and building Block Index. Use Version 2 Check for Coinbase Tx.

Timestamps:
1412676000   //--- Block Version 2 Testnet Activation:  10/07/2014 04:00:00 GMT - 6
1412964000   //--- Block Version 2 Activation:               10/10/2014 12:00:00 GMT - 6
1413914400   //--- GPU Channel Activation:                  10/21/2014 12:00:00 GMT - 6
1415124000   //--- POS Channel Activation:                  11/04/2014 12:00:00 GMT - 6

Block Version 2: Changes from Version 1 Go Live at Unix Timestamp 1412964000
A. Modulate Difficulty from Released Rewards by changing Block Target, this will set the minimum difficulty to 1/2 target when Released Rewards are too high.
B. Allow multiple outputs besides the 3 decay equations. Allows miner to customize their coinbase transaction.


Technical Support:
Contact me on Coin Shield Talk: Chat Box or Coin Shield Talk: Support.
Follow me on My Twitter For Updates..


Viz.

Make sure you update. 

[Supercomputing]

Something is wrong with block propagation through the network

[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14604 | Diff = 6.038846 | 29:26:34
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14604 | Diff = 6.038846 | 29:26:46
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14604 | Diff = 6.038846 | 29:26:57
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14604 | Diff = 6.038846 | 29:27:08
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14604 | Diff = 6.038846 | 29:27:19
[MASTER] Coinshield Network: New Block 14607
[MASTER] Coinshield Network: New Block 14611
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:27:32

  Prime Clusters Found: Four=59379 | Five=2117 | Six=32 | Seven=0 | Eight=0

[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:27:44
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:27:55

  Prime Clusters Found: Four=59392 | Five=2118 | Six=32 | Seven=0 | Eight=0

[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:28:06
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:28:17
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:28:28

  Prime Clusters Found: Four=59416 | Five=2119 | Six=32 | Seven=0 | Eight=0

[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14611 | Diff = 6.059813 | 29:28:39

  Prime Clusters Found: Four=59418 | Five=2120 | Six=32 | Seven=0 | Eight=0


  Prime Clusters Found: Four=59419 | Five=2121 | Six=32 | Seven=0 | Eight=0

[MASTER] Coinshield Network: New Block 14612
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14612 | Diff = 6.062807 | 29:28:51
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14612 | Diff = 6.062807 | 29:29:02
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14612 | Diff = 6.062807 | 29:29:13
[METERS] 28 Block(s) ACC=28 REJ=0| Height = 14612 | Diff = 6.062807 | 29:29:24

[Videlicet]

Supercomputing,

Fixing at the moment. Will be resolved shortly.

Viz.

[Supercomputing]

Awesome, thanks.

[kev7112001]

yeah big diff drop