Team Members: @Permie, @MarkoPaasila, @Kimchiking, Cryptik, Taconator
DISCLAIMER: DEXBOT IS NOT A RISK-FREE ENDEAVOUR AND ANY TRADING BOT CARRIES SIGNIFICANT RISK. THE DEXBOT TEAM DO NOT GIVE TRADING ADVICE AND RELY ON USERS TO DO THEIR OWN RESEARCH AND EVALUATE THE RISKS BEFORE PARTICIPATING IN THE PROJECT OR USING ANY PRODUCT FUNDED, CREATED, OR MENTIONED BY THE DEXBOT TEAM.
DEXBot is an approved BitShares Worker Proposal receiving funding equivalent to $667 per day to build an open-source and simple to use trading bot to help provide liquidity to the DEX.
Liquidity is the life-blood of an exchange and we believe that a widely distributed trading bot will help crowd-source the provision of liquidity.
The DEXBOT team is looking for new developer talent.
It was decided that when talking about the whole software we use the word Bot. The word Worker is to be used when referring to a single account+strategy+history+configuration combo. The Bot can contain multiple Workers.
Now you can set the Center Price for the Relative Orders strategy either manually or let the bot calculate it from the closest opposite orders in the market in question. This makes the strategy work in any non-empty market, volatile or stable. However it is most suitable for at least moderately active markets. In very silent markets the strategy can be tricked to make a loss.
Decimal places was increased to 8 for price and amounts
Now an existing Worker can be edited. If the configuration is changed, the Worker replaces existing orders.
Pausing a Worker now cancels all open orders, but remembers configuration and history. When "Pause" is clicked the user likely expects no action to happen until resumed, which requires removing market orders. Deleting a Worker cancels all orders and deletes configuration.
Installing the software with PIP now automatically installs dependencies
Continuous Integration for Windows (Appveyor), Linux and Mac (Travis) was "integrated" into developer workflow (Github) to streamline development and ensure build success after new code.
Work has been done to merge the CLI and GUI codebases, so from now on all new features should appear for both user interfaces.
Lots of progress has been made on making a simple-to-install application for all platforms. However there is "still one small problem". You know what I mean. This remains a high priority but we won't let it hinder progress on the features front.
The website representing DEXBot was migrated to dexbot.info. The website has been improved both visually and content-wise, preparing for it to be a central source of information and of help to new users getting an understanding of what they are up to.
We decided together with rahemanali (from the Autonio team) to find a way for Autonio users to request liquidity for any market-pair. Traders look for liquid markets, and if a market is "empty" they will just skip it. The expectation of volume is an incentive to Market Makers to bootstrap a new pair. That way we can get over the chicken-egg problem of in-existent pairs.
Bugfixes
Lots of experiments and work has gone into some fundamental challenges like
How to find, validate and choose a node to connect to
How to build the bot so that multiple Workers, like 10 or 100, won't make it slow down. If the software waits for one action to be confirmed before doing the next one, a slow node would render it unusable.
Maybe the most important "development" has been that the process of development is beginning to work. People are finding their natural roles in the process and things are moving on all fronts. This is an entirely new project and team, and we are spread all over the globe. It would have been significantly easier to make it a One Man Project, but in the end teamwork will yield much better results. Develpment has picked more speed, and will continue to do so as we improve collaboration and experience is gained. I would like to thank all participating developers for being so co-operative.
Imminent improvements
Hopefully "binaries" for all platforms.
Relative Order Size for the Relative Orders strategy. This will help keep the strategy in balance - not to sell all of the other asset when price trends. It will also automatically reinvest profits automatically, hopefully creating a compound-effect (assuming any profits are made). The inverse is also true; the order sizes will be decreased in case of losses, so the strategy remains "healthy".
Support for multiple Workers in the GUI. The CLI already has this.
Bugfixes
After these enhancements the bot should be actually useful in most existing markets.
Here's some work done on the UX
The goal is to make DEXBot an easy-to-use and useful tool, but also appealing. We want to keep it as simple as possible, self-explanatory and "glanceable", so you can see what you need to with a glance. There should be no unnecessary stuff cluttering the view, and anything requiring manual intervention should be immediately obvious.
Possible layouts:
Possible tab layout:
Possible "menu layout":
Money Spent
The cost of LTM and WP
Programming work for Milestone 1: $11160
UX design work: $630
Project management January, 30 hours: $900
How to help
If you are somewhat technical you can give the software a try and report bugs.
Let us know any ideas in the chat or by creating a github issue. We are interested in all kinds of ideas and opinions regarding at least User Experience, code, Market Making strategies, community building, analysis, security, collaboration, project management, and what not.
The multisig participants responsible for allocating the funds are willing to pay for whatever contribution brings the most value to BitShares (of course relevant to the DEXBot project), so please join us and bring your intellect and expertise.
Other
Please expect the software to be buggy until explicitly declared bug-free.
The next Milestone is due by March 26th. We will try to eventually achieve a regular release rhythm of two weeks.
