列举在 USENIX 投稿栏目中的 Systems Benchmarking Crimes
如何理解一份 “研究工作” 的贡献?
Evaluation criteria (ICSE)
从已有的东西开始:用得多
从不存在的东西开始:前景好
如果全文的方法和实验证据能 defend 你 claim 的故事,论文基本就成功了
故事 & 问题:
Storage systems ... have a simple, basic contract: you give them data, they do not lose or corrupt it. Often they store the only copy, making its irrevocable loss almost arbitrarily bad. Unfortunately, their code is exceptionally hard to get right, ...
This paper describes EXPLODE, a system that makes it easy to systematically check real storage systems for errors.
证据:
It found serious bugs in a broad range of real storage systems (without requiring source code): ... We found bugs in every system we checked, 36 bugs in total, typically with little effort.
在某个方面改进软件的生产或使用 (usefulness)
请大家考虑一下你的研究工作:
以上是一个 “一分钟” 版本的 talk,用于向其他人推销自己 (系里偶遇的老师、开会时遇到的伙伴……) 时使用
让读者能准确、轻松地理解你想表达的内容
Important goal: Don’t make readers a hard time in reading your papers! Your technical content is already hard enough.
我们在学英语时,经常要求写一个 “不少于 XXX words 的作文”
Vigorous writing is concise. A sentence should contain no unnecessary words, a paragraph no unnecessary sentences, for the same reason that a drawing should have no unnecessary lines and a machine no unnecessary parts. This requires not that the writer make all his sentences short, or that he avoid all detail and treat his subject only in outline, but that every word tell.
—— The Elements of Style
全文的逻辑应当是一致、连贯、紧凑的
Q: 大家的 research proposal 是按什么顺序写作的?
Top-down approach
Bottom-up approach
首先,不要有拼写错误
消灭语法错误对 non-native speakers 来说其实相当困难 (对 grammar checker 也是)
(1) [Topic Sentence] Large-scale distributed systems are difficult to build and test. (2) On top of the non-determinism arising out of concurrent exchange of messages, these systems must account for partial failures, where components or communication can fail along the way and produce incomplete results. (3) Fault-tolerant components are difficult to design and reason about, ...
distributed systems → basic ideas → the hard part
每个单词都有它的 embedding
含义模糊、过强、offensive、口语化表达等都要避免
针对同一个术语,坚持使用同一个名字
(都是细节,就举几个个容易错的例子吧)
以下哪个方式是正确的?
大家可能会觉得这个论文作者是星际玩家
引号中的标点要非常小心
The one exception we agree on: If the quote is around a symbol. For instance:
这是一个积累的过程
有用的资源