-
/bounty 100
-
Design weapons in more detail. List the weapons you want. How does a player gain weapons? How doe they interact with them? How do they upgrade them and switch between them? What differentiates one wea…
-
[Policy Gradient Methods for Reinforcement Learning with Function Approximation](https://papers.nips.cc/paper/1713-policy-gradient-methods-for-reinforcement-learning-with-function-approximation.pdf)
-
# Database Schema for Persistent Browser-Based Game
## Issue Description
### Overview
We need to design and implement a comprehensive database schema for our persistent browser-based game. Th…
-
Summary: Add a hitman burner phone item that creates a PVE hitman activity. Instancing an illegal short staged random encounter, that has an illegal item + cash reward at the end. Giving more heavily …
-
# ETHDenver '22 Bounty: Design an “open” peer review process
## About SCRF
The [Smart Contract Research Forum](https://smartcontractresearch.org/) (SCRF) is a premier, open discourse hub dedicat…
-
-
-
* Khi đánh thắng bot sẽ nhận được những loại phần thưởng nào
* Giá trị của từng loại phần thưởng
* Có cơ chế x2 hay x3 phần thưởng không
* Dựng UI wireframe cho phần nhận thưởng
-
# Lines of code
https://github.com/code-423n4/2024-02-uniswap-foundation/blob/main/src/UniStaker.sol#L244-L246
https://github.com/code-423n4/2024-02-uniswap-foundation/blob/main/src/UniStaker.sol#L23…