I now have 2 Windows machines running clients for the Testnet.

Machine A is mining with the new client and Machine B is running the old client. I have successfully sent coins from the new client to the old client, but have not yet been able to send coins from the old client to the new client.

Our latest fixes for the clients (Windows):
1.  old-fix missing library file for miner to work
2.  new-fix several things.  (i) allow free transactions to be relayed.  (ii) allow free transactions to be mined.  (iii) allow old client transactions to be received and mined.

There are two versions to use in the testnet.

• The Old Client - 0.7.5
• The New Client - 0.14.2

Install the software
The install packages are zip files.  They contain the programs and folders that will be used for the blockchain.  It doesn't interfere with other wallets on your system.  Unzip these folders where you want and run the "1. start-client" command.  Other commands such as mining may be present in the packages.   If you delete the package after unzipping and running the client, you will loose your wallet.

Windows
old - https://github.com/Stouse49/goldcoin/releases/download/vv0.7.5.0-testnet/Windows.GoldCoin.0.7.5.TestNet.zip
new -

Mac
old - https://github.com/Stouse49/goldcoin/releases/download/vv0.7.5.0-testnet/Mac.GoldCoin.0.7.5.TestNet.zip (may not connect)
new -

Linux

Mining
Information to be added later.

Super Users (Exchanges, Mining Pools, etc)

Source Code:
old:  https://github.com/Stouse49/goldcoin [branch = testnet]
Build instructions: https://www.goldcointalk.org/index.php?topic=3764.msg18102#msg18102

new:  https://github.com/Stouse49/Goldcoin-GLD [branch = testnet]
Build instructions:  https://github.com/goldcoin/Goldcoin-GLD/tree/goldcoin-master/doc

Mining
old - solo - https://github.com/pooler/cpuminer, $ minerd -o 127.0.0.1:<rpc port, default=18121> -u <user> -p <password> -t <threads, such as 2>new - Solo - https://github.com/mikeghen/litecoin-solo-mining-tutorial

server.js - example
var myCoin = {
    "name": "Goldcoin",
    "symbol": "GLD",
    "algorithm": "scrypt",
     "peerMagic": "fdc2b4dd",
     "peerMagicTestnet": "fcc5b4dd"
};

var Stratum = require('stratum-pool');

var pool = Stratum.createPool({

    "coin": myCoin,

    "auxes": [],

    "address": "n33ZW9SXvtCXRvsP7gjQqYFaEwxwpXoWpC", //Address to where block rewards are given

    /* Block rewards go to the configured pool wallet address to later be paid out to miners,
       except for a percentage that can go to, for examples, pool operator(s) as pool fees or
       or to donations address. Addresses or hashed public keys can be used. Here is an example
       of rewards going to the main pool op, a pool co-owner, and NOMP donation. */
    // "rewardRecipients": {
    //     "n37vuNFkXfk15uFnGoVyHZ6PYQxppD3QqK": 1.5, //1.5% goes to pool op
    //     "mirj3LtZxbSTharhtXvotqtJXUY7ki5qfx": 0.5, //0.5% goes to a pool co-owner
    // },

    "blockRefreshInterval": 1000, //How often to poll RPC daemons for new blocks, in milliseconds


    /* Some miner apps will consider the pool dead/offline if it doesn't receive anything new jobs
       for around a minute, so every time we broadcast jobs, set a timeout to rebroadcast
       in this many seconds unless we find a new job. Set to zero or remove to disable this. */
    "jobRebroadcastTimeout": 55,

    /* Some attackers will create thousands of workers that use up all available socket connections,
       usually the workers are zombies and don't submit shares after connecting. This features
       detects those and disconnects them. */
    "connectionTimeout": 1200, //Remove workers that haven't been in contact for this many seconds

    /* Sometimes you want the block hashes even for shares that aren't block candidates. */
    "emitInvalidBlockHashes": false,

    /* Enable for client IP addresses to be detected when using a load balancer with TCP proxy
       protocol enabled, such as HAProxy with 'send-proxy' param:
       http://haproxy.1wt.eu/download/1.5/doc/configuration.txt */
    "tcpProxyProtocol": false,

    /* If a worker is submitting a high threshold of invalid shares we can temporarily ban their IP
       to reduce system/network load. Also useful to fight against flooding attacks. If running
       behind something like HAProxy be sure to enable 'tcpProxyProtocol', otherwise you'll end up
       banning your own IP address (and therefore all workers). */
    "banning": {
        "enabled": true,
        "time": 600, //How many seconds to ban worker for
        "invalidPercent": 50, //What percent of invalid shares triggers ban
        "checkThreshold": 500, //Check invalid percent when this many shares have been submitted
        "purgeInterval": 300 //Every this many seconds clear out the list of old bans
    },

    /* Each pool can have as many ports for your miners to connect to as you wish. Each port can
       be configured to use its own pool difficulty and variable difficulty settings. varDiff is
       optional and will only be used for the ports you configure it for. */
    "ports": {
        "3333": { //A port for your miners to connect to
            "diff": 32, //the pool difficulty for this port

            /* Variable difficulty is a feature that will automatically adjust difficulty for
               individual miners based on their hashrate in order to lower networking overhead */
            "varDiff": {
                "minDiff": 8, //Minimum difficulty
                "maxDiff": 512, //Network difficulty will be used if it is lower than this
                "targetTime": 15, //Try to get 1 share per this many seconds
                "retargetTime": 90, //Check to see if we should retarget every this many seconds
                "variancePercent": 30 //Allow time to very this % from target without retargeting
            }
        },
        "3256": { //Another port for your miners to connect to, this port does not use varDiff
            "diff": 256 //The pool difficulty
        }
    },

    /* Recommended to have at least two daemon instances running in case one drops out-of-sync
       or offline. For redundancy, all instances will be polled for block/transaction updates
       and be used for submitting blocks. Creating a backup daemon involves spawning a daemon
       using the "-datadir=/backup" argument which creates a new daemon instance with it's own
       RPC config. For more info on this see:
          - https://en.bitcoin.it/wiki/Data_directory
          - https://en.bitcoin.it/wiki/Running_bitcoind */
    "daemons": [
        {   //Main daemon instance
            "host": "127.0.0.1",
            "port": 19332,
            "user": "Stouse49",
            "password": "x"
        }
        // {   //Backup daemon instance
        //     "host": "127.0.0.1",
        //     "port": 19344,
        //     "user": "litecoinrpc",
        //     "password": "testnet"
        // }
    ],


    /* This allows the pool to connect to the daemon as a node peer to receive block updates.
       It may be the most efficient way to get block updates (faster than polling, less
       intensive than blocknotify script). It requires the additional field "peerMagic" in
       the coin config. */
    "p2p": {
        "enabled": false,

        /* Host for daemon */
        "host": "127.0.0.1",

        /* Port configured for daemon (this is the actual peer port not RPC port) */
        "port": 18121,

        /* If your coin daemon is new enough (i.e. not a [beep]coin) then it will support a p2p
           feature that prevents the daemon from spamming our peer node with unnecessary
           transaction data. Assume its supported but if you have problems try disabling it. */
        "disableTransactions": true

    }

}, function(ip, port , workerName, password, callback){ //stratum authorization function
    console.log("Authorize " + workerName + ":" + password + "@" + ip);
    callback({
        error: null,
        authorized: true,
        disconnect: false
    });
});

pool.on('share', function(isValidShare, isValidBlock, data){

    if (isValidBlock)
        console.log('Block found');
    else if (isValidShare)
        console.log('Valid share submitted');
    else if (data.blockHash)
        console.log('We thought a block was found but it was rejected by the daemon');
    else
        console.log('Invalid share submitted')

    console.log('share data: ' + JSON.stringify(data));
});

pool.on('log', function(severity, logKey, logText){
    console.log(severity + ': ' + '[' + logKey + '] ' + logText);
});

console.log("Starting Pool")
pool.start();
console.log("Pool started", pool)

With Atomica, there appears to be one working who has 248Kh/s, which is too slow to get any coins.

With a total network of 7.372 GH/s, your miner (300Mh/s would not get many blocks, perhaps 1 of 25 if mining solo or on Atomica where you would be the only one mining.  That would be about 1 block per hour, assuming everything is working right.  How long did you try to mine?  Having shares accepted only results in payout if the mining pool is successful in mining a block.

For hash-to-coins, I have no experience, so cannot comment on why your miners do not connect with cgminer.

What is your experience with respect to Bitcoin and Blockchains and P2P?

I posted this on the bitcointalk.org forum as well:

This error indicates that the blockchain database files are corrupted.  You will need to delete them (~/Library/Application Support/Goldcoin (GLD)/blkindex.dat and blk*.dat) but DO NOT delete wallet.dat and the client can download the blockchain again once you started it up again.

Let me know if this works or doesn't make any sense at all.

oblems in the GoldenRiver implementation.

We were stuck on the GoldenRiver implementation translating from C++ to Objective C, but we were successful.

Yes, it will be very hard or impossible to get your wallet back.  If it is not backed up, then there is nothing anyone can do.  Especially if the computer with the original wallet is one.

I can't provide an explanation of why it seemed as if the coins were sent to you new wallet, but now they only appear in your old address for which there is no backup.

Unfortunately I don't see anything like Previous Version after right clicking on the file wallet.dat and entering the Properties.

Could I try installing older verison of GLD wallet on the other camputer and trying to copy the wallet.dat file?

Most of the previous versions of the GoldCoin clients won't work (or sync, rather).  All versions use the same wallet.dat format, so you could install the previous version and try that.  We don't currently know of any version related issues with wallet files and not reading them correctly. 

The other thing you can try is starting the goldcoin-qt.exe file with the -rescan parameter.  Meaning if you open a command line window in the folder that contains goldcoin-qt.exe, then enter the command:
goldcoin-qt -rescan
This is a tough one.