一句话：梳理热补丁与热修复术语，提出融合缺陷/漏洞预测、自动测试生成与修复及运行时补丁的自动化框架愿景。
随着社会对软件的依赖持续加深，软件可靠性变得至关重要。然而，受时间限制、测试不充分以及开发者难以掌握完整技术栈等因素影响，在线系统完全避免缺陷并不现实。因此，通过热补丁与热修复在生产环境中降低风险，已成为软件开发的重要实践。本文首先梳理相关文献中用于该研究主题的术语，进而在此基础上提出一个自动化框架愿景：在运行时预测并缓解关键软件问题。该框架整合多个领域的热补丁与热修复研究，特别是软件缺陷与漏洞预测、自动测试生成与修复，以及运行时补丁。论文属于反思与展望，未报告实证评估，旨在激发不同社区之间的研究协作。
With our reliance on software continuously increasing, it is of utmost importance that it be reliable. However, complete prevention of bugs in live systems is unfortunately an impossible task due to time constraints, incomplete testing, and developers not having knowledge of the full stack. As a result, mitigating risks for systems in production through hot patching and hot fixing has become an integral part of software development. In this paper, we first give an overview of the terminology used in the literature for research on this topic. Subsequently, we build upon these findings and present our vision for an automated framework for predicting and mitigating critical software issues at runtime. Our framework combines hot patching and hot fixing research from multiple fields, in particular: software defect and vulnerability prediction, automated test generation and repair, as well as runtime patching. We hope that our vision inspires research collaboration between the different communities.