Before we push for wider deployment, we decided to make an important change to the wallet.
One of the fundamental jobs of the wallet is to work cooperatively with the user to ensure that a payment is reliably sent once and not more than once (i.e., no unintended double payments).
When the send command is submitted, the user can experience three possible outcomes:
1. The wallet responds that the payment was submitted to the network and provides a TXID to track the payment.
2. The wallet responds with an error, indicating the payment was not submitted to the network.
3. The user receives no response, which can occur when the connection between the user and wallet is interrupted.
Condition #3, the no response condition, is the most complicated to handle correctly. The wallet may or may not have received the send command and submitted the payment to the network. There is no way to tell from the response. In that case, the user could query the wallet to scan for a transaction that looks similar to the one that was submitted (same address, amount and in the right block or time range). However, before making this query, the wallet would have to wait a sufficient time to allow both the network and wallet to timeout, which ensures the prior send command is not still pending or in progress. This is the method we originally envisioned to handle this situation, but it is both complex and has reliability pitfalls.
In order to address this case better, we are implementing a new send command, cc.send, which will require three parameters: a destination, an amount, and a unique reference id. This command has slightly different semantics than sendtoaddress. Unlike sendtoaddress, the user can submit the command "cc.send referenceid destination amount" multiple times, but the wallet will never send the payment more than once. This allows much easier and more reliable handling of the no response condition. If no response is received from cc.send, the user can simply submit the identical command again, as many times as required, and will eventually receive a response that contains either an error or a TXID to track the payment. The payment however will never be sent more than once, which cannot be guaranteed by the bitcoin-compatible sendtoaddress command. And as an added benefit, no queries will be required to scan for and obtain the TXID.
That is one of the changes we are working on, along with a number of other usability changes, including adjusting the timeout that caused many users to see "Insufficient Funds" when sending payments in the current release. Once all of this is implemented and thoroughly tested (which might take a few weeks), we plan on releasing an updated version of the software along with a testnet, and then we will proceed with attempting to assist exchanges in listing the CredaCash currency.
Thank you again for your interest and support, and please let us know if you have any additional questions.