GitGuild is pursuing a number of projects for us, known as “Lamassau ATM” and “Fiat Gateways”, but unfortunately many of these projects have not met all of our initial milestones. The vendor continues to work on the project, and Dash is incorporated in all the projects they are working on, including software for a number of projects for other partners like Coinapult, Bitt and CryptoCapital. You can see more information on the new GitGuild website www.gitguild.com.
When we initially contracted with GitGuild to create CTM software based on the Lamassau code, we anticipated substantial work as we were not going for the simple Dash only version replacing Bitcoin on the codebase but we felt dual functionality was required in the current environment meaning creating multicurrency support. We’d also received commitments to support the effort from Lamassu’s CEO. However, as GitGuild began their work, a number of additional obstacles came up- many of which could not have been anticipated in advance - caused significant delays.
The issues encountered were lengthy, but there were two main hurdles encountered that caused most of the delays to date, which I will attempt to explain. The resulting progress has been much slower than anticipated.
Although an initial working version of the software was created by Q2 of 2016, it was impractical for operators to deploy. First, because no Dash wallet services with APIs exist, each CTM required a full Dash node to operate. Second, the only exchange trading Dash that was included in the exchange plugin was for Poloniex, which lacked a DASH/USD trading pair. The only pricing option was therefore the illiquid USDT pair, which introduced unacceptable exchange rate pricing accuracy risk. It also required the operator to invest significantly in USDT, which meant Dash and BTC trading would draw from different resources, significantly raising operating costs. It also required a highly technically skilled operator to install the software.
To address these issues, two new components were needed. The first was a shared wallet (DeSW) available through an API to eliminate the need for a dash full node, and the shared wallet also needed a DeSW-Dash plugin. The second component was broker software (DeBroker) that could connect to additional exchanges. These additional components were not completed until recently.
The second and much larger issue was the decision of Lamassau to undertake major changes to their architecture to enable integration of Ethereum. In January 2016, when the ATM project first funded, Ethereum was not yet on Lamassau’s roadmap. By March, however, Ethereum had become extremely popular. Lamassau migrating to a multi-currency architecture might seem beneficial to Dash’s efforts on the surface, but it was actually detrimental for a few reasons. It basically invalidated much of the work performed to that point, since all of the enhancements would need to be rewritten off a fork of Lamassau’s new software, not the old one we’d been working from. Also, the software was still an “alpha” software branch. It was unclear whether the old branch would survive or whether the new branch would become the official software. So not only was the work-to-date invalidated, but which branch to develop from remained uncertain. Had the new Ethereum machines sold well, the issue might have resolved itself, but unfortunately that was not the case. Due to the uncertainty, GitGuild decided to build off of the Ethereum branch a couple of months behind Lamassu. It is still unresolved whether the Ethereum branch will become the official software.
Despite these issues, progress is continuing, albeit slowly. Ira Miller, lead developer on the project, gave us an update on the latest version of the CTM code that eliminates the need for a server and connects the machine directly to the Dash hot wallet, also developed by the GitGuild. You can see the latest code here: https://github.com/CTMGuild/ctm-machine/tree/dashctm
This is an illustration of the latest version illustrating the machine connecting directly to the hot wallet (desw):
Whereas on the previous version the setup required an extra Lamassu server using plugins as illustrated here:
Ira gave us some details on this version “This GitGuild CTM machine code is forked from the Lamassu machine alpha code which has Ethereum alpha code, but their branch is moving very slowly it has not been updated recently so we needed to move forward. So we took that and updated it, replaced some proprietary software components that were used and replaced them with open source components. Eliminated the need for a Lamassu server side installation and connected it directly to the desw (hot wallet) and the fiat gateways components that we developed for Dash and that is what we have running here. Also, the interface is updated to the latest Lamassu version, is beyond their latest stable stack, and is also beyond their ETH alpha stack while using purely open source modular components.”
Here you can see a recent video demonstration of the machine working:
On the GitGuild website www.guitguild.com you can also find links and information on the Hot Wallet and Trade Manager that are available to be used by any interested users.
The situation remains uncertain with the CTM software as it relates to Lamassau, and we are still uncertain whether the goals of the project will be fulfilled. However, regardless of the outcome, the software is receiving interest from larger operators wishing to launch their own CTMs, and we will pursue those opportunities. Many of the components developed could be ported to other machines, even if Lamassu doesn’t resolve their architecture direction in the near future.
The non-CTM components continue to be developed and enhanced as well. These other tools can assist partners with accelerating their integration to Dash. The full list of tools is also available on the website.
Although our agreement with GitGuild called for 12 months of payments, the service period is actually 24 months, through the end of this year. Given the challenges facing the project and continued uncertainty, it has been frustrating for us to attempt to extract the intended value. Going forward, the project is paid through this year, and while we intend to allow GitGuild to continue working to produce improved solutions, we feel it will be more beneficial to focus our efforts in more promising directions, such as working with operators wishing to develop their own machines. We will continue to check on the progress being made, and report periodically on any additional progress. In the meantime, the software is freely available for download and Lamassau will flash the current version on their factory floor for new buyers upon request.
The project vendor represented by Ira Miller has started to communicate directly with the community on different channels and will continue to do so going forward when there are new updates to share.
Lessons from this agreement:
We’ve learned a number of lessons from this contract that the core team has discussed at length, many of which have led to changes in the way we operate.
First, despite the fact that funds were insufficient to pay for these services up front, our price can change quite rapidly. Consequently, we should not be signing agreements denominated in Dash lasting more than three months, and we have not done so since this proposal. Any proposal unable to be accommodated within three months is likely too large in scope to address until later.
Second, we should not submit proposals that distribute funds directly to a vendor. The core team should act as custodian of the funds until it can assess often complex situations like this one to determine the best course of action.
Third, we should prioritize working with partners that possess direct control over the outcome of the projects we request of them. In this case, being at the mercy of Lamassau’s architecture changes has significantly impacted the vendor’s ability to deliver the CTM portion of the contract. We should assign much higher execution risk to third-party projects.
Fourth, we should include exit clauses in our agreements if there are situations in which proceeding makes little sense. In this case, when Lamassau announced major changes to their architecture and the Ethereum branch development ceased in the fall, an exit clause surrounding this uncertainty would have been valuable. Any number of outcomes, from the bankruptcy of Lamassau to being bought by another company to a change in the CEO commitment to the project could be triggers for the option to discontinue. A simply 30-60 day notice clause would have also worked.
When we initially contracted with GitGuild to create CTM software based on the Lamassau code, we anticipated substantial work as we were not going for the simple Dash only version replacing Bitcoin on the codebase but we felt dual functionality was required in the current environment meaning creating multicurrency support. We’d also received commitments to support the effort from Lamassu’s CEO. However, as GitGuild began their work, a number of additional obstacles came up- many of which could not have been anticipated in advance - caused significant delays.
The issues encountered were lengthy, but there were two main hurdles encountered that caused most of the delays to date, which I will attempt to explain. The resulting progress has been much slower than anticipated.
Although an initial working version of the software was created by Q2 of 2016, it was impractical for operators to deploy. First, because no Dash wallet services with APIs exist, each CTM required a full Dash node to operate. Second, the only exchange trading Dash that was included in the exchange plugin was for Poloniex, which lacked a DASH/USD trading pair. The only pricing option was therefore the illiquid USDT pair, which introduced unacceptable exchange rate pricing accuracy risk. It also required the operator to invest significantly in USDT, which meant Dash and BTC trading would draw from different resources, significantly raising operating costs. It also required a highly technically skilled operator to install the software.
To address these issues, two new components were needed. The first was a shared wallet (DeSW) available through an API to eliminate the need for a dash full node, and the shared wallet also needed a DeSW-Dash plugin. The second component was broker software (DeBroker) that could connect to additional exchanges. These additional components were not completed until recently.
The second and much larger issue was the decision of Lamassau to undertake major changes to their architecture to enable integration of Ethereum. In January 2016, when the ATM project first funded, Ethereum was not yet on Lamassau’s roadmap. By March, however, Ethereum had become extremely popular. Lamassau migrating to a multi-currency architecture might seem beneficial to Dash’s efforts on the surface, but it was actually detrimental for a few reasons. It basically invalidated much of the work performed to that point, since all of the enhancements would need to be rewritten off a fork of Lamassau’s new software, not the old one we’d been working from. Also, the software was still an “alpha” software branch. It was unclear whether the old branch would survive or whether the new branch would become the official software. So not only was the work-to-date invalidated, but which branch to develop from remained uncertain. Had the new Ethereum machines sold well, the issue might have resolved itself, but unfortunately that was not the case. Due to the uncertainty, GitGuild decided to build off of the Ethereum branch a couple of months behind Lamassu. It is still unresolved whether the Ethereum branch will become the official software.
Despite these issues, progress is continuing, albeit slowly. Ira Miller, lead developer on the project, gave us an update on the latest version of the CTM code that eliminates the need for a server and connects the machine directly to the Dash hot wallet, also developed by the GitGuild. You can see the latest code here: https://github.com/CTMGuild/ctm-machine/tree/dashctm
This is an illustration of the latest version illustrating the machine connecting directly to the hot wallet (desw):
Whereas on the previous version the setup required an extra Lamassu server using plugins as illustrated here:
Ira gave us some details on this version “This GitGuild CTM machine code is forked from the Lamassu machine alpha code which has Ethereum alpha code, but their branch is moving very slowly it has not been updated recently so we needed to move forward. So we took that and updated it, replaced some proprietary software components that were used and replaced them with open source components. Eliminated the need for a Lamassu server side installation and connected it directly to the desw (hot wallet) and the fiat gateways components that we developed for Dash and that is what we have running here. Also, the interface is updated to the latest Lamassu version, is beyond their latest stable stack, and is also beyond their ETH alpha stack while using purely open source modular components.”
Here you can see a recent video demonstration of the machine working:
On the GitGuild website www.guitguild.com you can also find links and information on the Hot Wallet and Trade Manager that are available to be used by any interested users.
The situation remains uncertain with the CTM software as it relates to Lamassau, and we are still uncertain whether the goals of the project will be fulfilled. However, regardless of the outcome, the software is receiving interest from larger operators wishing to launch their own CTMs, and we will pursue those opportunities. Many of the components developed could be ported to other machines, even if Lamassu doesn’t resolve their architecture direction in the near future.
The non-CTM components continue to be developed and enhanced as well. These other tools can assist partners with accelerating their integration to Dash. The full list of tools is also available on the website.
Although our agreement with GitGuild called for 12 months of payments, the service period is actually 24 months, through the end of this year. Given the challenges facing the project and continued uncertainty, it has been frustrating for us to attempt to extract the intended value. Going forward, the project is paid through this year, and while we intend to allow GitGuild to continue working to produce improved solutions, we feel it will be more beneficial to focus our efforts in more promising directions, such as working with operators wishing to develop their own machines. We will continue to check on the progress being made, and report periodically on any additional progress. In the meantime, the software is freely available for download and Lamassau will flash the current version on their factory floor for new buyers upon request.
The project vendor represented by Ira Miller has started to communicate directly with the community on different channels and will continue to do so going forward when there are new updates to share.
Lessons from this agreement:
We’ve learned a number of lessons from this contract that the core team has discussed at length, many of which have led to changes in the way we operate.
First, despite the fact that funds were insufficient to pay for these services up front, our price can change quite rapidly. Consequently, we should not be signing agreements denominated in Dash lasting more than three months, and we have not done so since this proposal. Any proposal unable to be accommodated within three months is likely too large in scope to address until later.
Second, we should not submit proposals that distribute funds directly to a vendor. The core team should act as custodian of the funds until it can assess often complex situations like this one to determine the best course of action.
Third, we should prioritize working with partners that possess direct control over the outcome of the projects we request of them. In this case, being at the mercy of Lamassau’s architecture changes has significantly impacted the vendor’s ability to deliver the CTM portion of the contract. We should assign much higher execution risk to third-party projects.
Fourth, we should include exit clauses in our agreements if there are situations in which proceeding makes little sense. In this case, when Lamassau announced major changes to their architecture and the Ethereum branch development ceased in the fall, an exit clause surrounding this uncertainty would have been valuable. Any number of outcomes, from the bankruptcy of Lamassau to being bought by another company to a change in the CEO commitment to the project could be triggers for the option to discontinue. A simply 30-60 day notice clause would have also worked.