一句话：提出博士项目计划，用新度量、可视化工具和模式分类研究大型长期 FLOSS 项目中的知识集中化演化与缓解。
本文是一个博士项目计划，研究大型长期自由/开源软件（FLOSS）项目中知识集中化的演化。组织因素，如开发者负责的代码区域和代码审查职责，直接影响代码库的可持续性；维护者层级可能因责任或知识过度集中而出现瓶颈。项目计划分三阶段：(A) 提出同时考虑提交作者身份与评审职责的新度量来量化知识集中度；(B) 构建可视化工具以展示代码库中知识集中度的演化；(C) 归纳可缓解知识集中瓶颈的开发模式分类。摘要未报告已完成的实证结果，评估将围绕大型长期 FLOSS 项目展开。
The organizational factors that govern a development team, such as which parts of the code each developer works on or how code changes are reviewed, are pivotal to the survival of large software codebases. Balancing such organizational factors is an even greater challenge for Free/Libre and Open Source Software (FLOSS) projects, which must sustain hierarchies of maintainers while preventing any of them from becoming overloaded with too much responsibility or concentrating too much knowledge of the codebase. This PhD project aims to identify the effects of different development models on the evolution of knowledge concentration in large, long-lived FLOSS projects. To that end, we envision three phases for the project, each with its expected contribution: (A) A new metric to quantify knowledge concentration considering commit authorship and reviewing duties; (B) A tool to visualize the evolution of knowledge concentration in software codebases; and (C) A taxonomy of patterns of development models for mitigating knowledge-concentration-induced bottlenecks in collaborative codebases.