pjcscoder @pjcscoder
Joined July 2018-
Tweets3K
-
Followers39
-
Following862
-
Likes145
Classic Operating Systems: From Batch Processing to Distributed Systems The quote from the book's preface will lay the groundwork for what to expect in this book - a good reminder of how we have arrived here today, standing on the shoulders of the legends! Highly recommended to keep it as a reference! "In operating systems courses, you learn useful principles, but very little about the difficult art of software design. So I decided to go back to the masters who made the major discoveries in operating systems and put some of their best papers together in a book, and call it Classic Operating Systems: From Batch Processing to Distributed Systems. Well, here is that book. The papers illustrate the breakthroughs in operating system technology from the 1950s to the 1990s. All of them are written by the pioneers who designed these systems." - Per Brinch Hansen.
这又是一篇值得认真读的文章:《How modern browsers work》。 读完后我有两个强烈感受: 1. 浏览器其实是现代最被低估的“操作系统”。 2. 如果前端/Agent 开发者还把浏览器当黑盒,只会越来越吃亏。 从搞产品和做 Agent 的角度,我提炼了 5 个比较有收获的点: 1. 浏览器不是在“渲染页面”,而是在跑一个小型分布式系统 一页网页背后其实在跑整条流水线:网络栈 → 解析(HTML/CSS/JS)→ 样式计算 & 布局 → 绘制 & 合成 → JS 引擎(V8 多层 JIT)→ 多进程隔离。 你在 DevTools 里看到的 Performance trace,本质上就是这套复杂系统的压缩观测日志。 2. 所有前端性能优化,本质上都是在帮这条 pipeline 减压 控制首屏体积、async/defer、只用 transform/opacity 做动画、拆分长任务……这些最佳实践,拆开来看其实都是在给网络、解析、主线程、合成线程等不同阶段减负。 以后再说页面卡,可以更精确地问:卡在 network → parse → layout → paint/composite → JS/GC 的哪一段? 3. JS 引擎早就不是简单的解释器,而是一台会自我进化的编译管线 V8 会先解释执行,再根据运行 profile 逐步把热点路径编译成机器码(Ignition → Sparkplug → Maglev → TurboFan),类型不稳定就会 deopt。 这意味着:写出“类型稳定、路径稳定”的代码,本质上是在配合编译器走高速路,而不是在给它制造 deopt。 4. 浏览器多进程架构是 Web 世界的成熟“沙箱 + 隔离”范例 Browser Process + Renderer Process + Site Isolation + Sandboxing,这套设计已经稳定运行了十几年。 对做 Agent 的人来说,这其实是一个现成的参考:高风险操作要不要拆进程?Agent 之间用什么做 IPC?出问题时能不能做到局部失效不拖垮全局? 5. 这篇其实是给 Agent / Loop Engineering 的一份「样板间」 浏览器本质上就是一个已经在全球几十亿设备上稳定运行了十几年的大型 Loop + 多 Agent 系统。 它的事件循环、任务调度、隔离机制、性能优化思路,对我们现在在做的 Loop、Skill、Harness 都有很强的参考价值。 推荐对浏览器 internals 和 Agent 架构都感兴趣的朋友读一读。 你从浏览器底层学到过什么对做 Agent 有帮助的思路?
操作系统底层原理向来是计算机领域的硬骨头。几十万行内核源码摊在眼前,翻上几页便头晕目眩,既抓不住设计主线,更谈不上通篇精读。 egos-2000 这个教学项目,只用 2000 行代码,就把操作系统的核心组件完整呈现了出来。 这套系统的三层架构尤其清晰: 1️⃣ 底层直接管控硬件交互 2️⃣ 中间层处理进程调度与系统调用 3️⃣ 上层实现文件系统和命令解释器 它还能跑在 QEMU 模拟器和真实 RISC-V 开发板上,官方配套教材直接给出 9 个课程项目,方便边学边练。 操作系统从加电到用户界面究竟如何运转?这份源码就是最好的解剖样本,值得静下心一行一行啃透。 🔗 github.com/yhzhang0128/eg…
If you got excited about the paper, you will certainly enjoy this talk by @lemire youtu.be/wlvKAT7SZIQ
Parsing Gigabytes of JSON per Second by @geofflangdale and @lemire. One of the best paper for the weekend and also understand SIMD with the quoted article. arxiv.org/pdf/1902.08318
It's customary to share another Ulrich paper - What Every Programmer Should Know About Memory. people.freebsd.org/~lstewart/arti…
How to Write Shared Libraries - Ulrich Drepper cs.dartmouth.edu/~sergey/cs258/…
有信息焦虑的朋友们,必须给你们安利一个宝藏开源项目——Horizon,它就像一个专门替你盯着海外科技圈的 AI 情报雷达。 它自动抓 Hacker News、Twitter、Reddit、GitHub 的内容,AI 筛选去重再总结,直接给你出日报。 几个我觉得真有用的点: 1️⃣ AI 自动打分,垃圾内容直接踢掉 2️⃣ 把 Reddit 和 HN 里的高质量讨论单独提出来 3️⃣ 遇到没听过的公司和技术,自动补背景 4️⃣ 同一条新闻全网只出现一次,不重复轰炸 5️⃣ 中英双语简报,推送飞书/邮箱/微信都行 每天刷海外社区刷到头大的,这东西能帮你省一半时间。 🔗 github.com/Thysrael/Horiz…
Typora 团队新作 Otty:我终于找到心仪的轻量终端了! 我之前一直在找一个带简单文件编辑功能的终端工具。像 Warp、Superset、Orca 这类偏 Agentic IDE 的工具虽然功能很强,但用起来总觉得太重,启动和占用资源都不太理想。 前段时间用了下 Muxy,它足够轻巧,功能也基本能满足我的要求,但界面还是差点意思。作为一个颜值党,我一直没完全满意。 直到前天我看到了 Otty,说实话,第一眼就被它的颜值吸引了。默认的主题虽然是当下泛滥的类纸配色,但看惯了黑乎乎的终端界面,这种浅色设计还挺让人眼前一亮。整体界面干净整洁、细节打磨不错。 后来才知道,Otty 的开发团队 @appmakes 就是知名 Markdown 编辑器 @Typora 背后的团队。之前用 Typora 的体验不错,只是后面收费了就没在用了😅。 当然,Otty 并不只是给传统终端换了一层漂亮外壳: - 核心采用 Rust,macOS 版本通过原生 Metal 进行 GPU 加速,滚动和大量文本输出都比较流畅。 - 支持标签页和灵活的分屏布局,标签页位置可以自己配置。 - 右侧的 Details Panel,可以查看当前目录、命令历史、Git 状态以及项目文件树。 - 内置了轻量级文件编辑器,能够快速预览 Markdown、查看代码或临时修改配置文件。 - 支持会话恢复,重新打开应用后可以还原之前的窗口、分屏、工作目录和滚动记录。 - 与 Claude Code、Codex 和 OpenCode 深度集成,在标签页上显示 Agent 的运行、完成和等待输入状态。 - 还可以把标签页、分屏、目录和启动命令保存成 Recipe,方便一键恢复常用工作环境。 配置项也比较丰富,外观配置非常细致,可以自己修改边距、圆角和阴影,对于喜欢折腾终端外观的人来说,可玩性并不低。 Otty 的定位正好介于传统终端和 Agentic IDE 之间:比普通终端多一些现代工作流,又不像完整 IDE 那么重。 目前 Otty 只支持 macOS 14 及以上版本,Windows 和 Linux 版本还在开发中;现在免费使用,看上去后面要做云端同步等收费功能。 官网地址:otty.sh
Un desarrollador chino llamado tw93 se hartó de que sus aplicaciones de escritorio le devoraran la RAM y el disco. Abría Slack y desaparecían cientos de megabytes. Abría Discord, Notion o cualquier otra app y pasaba lo mismo. ¿La razón? Casi todas son lo mismo por dentro: un sitio web empaquetado con una copia completa del motor de Chrome (Electron). Decidió que tenía que haber una forma mejor. En 2022 empezó a construir Pake. Usó Rust + Tauri, que en vez de incluir un navegador completo, aprovecha el WebView nativo del sistema operativo. El resultado fue brutal: - Slack con Pake → 8 MB (en vez de 524 MB) - Discord con Pake → 9 MB (en vez de 265 MB) - ChatGPT con Pake → 9 MB (en vez de 260 MB) Cuatro años después, su repositorio tiene más de 51.000 estrellas en GitHub. Tiene builds listos para Grok, ChatGPT, Gemini, Discord, YouTube, Twitter y muchos más. Todo bajo los 10 MB, ligero, rápido y gratis. Y lo mejor: con un solo comando puedes convertir cualquier página web en una aplicación de escritorio nativa. No fundó una startup. No levantó inversión. Solo resolvió un problema que molestaba a millones de personas. A veces el cambio real lo hace una sola persona que se cansa de las cosas como están. Esta brutal, repo en los comentarios 👇
The previous paper and talk focused on using SIMD (Single Instruction Multiple Data) to efficiently parse massive JSON files. SIMD is a form of vectorization utilizing built-in capabilities of hardware for faster execution. cvw.cac.cornell.edu/vector Since both the posts resonated quite well, here is a solid high-level overview and a roadmap to understand the vectorization.
If "how V8 handles parsing" is on your mind, I would start with these - great resources. Part 1 - v8.dev/blog/scanner Part 2 - v8.dev/blog/preparser
Let's Code a TCP/IP Stack in C part 1 starts with raw Ethernet frames and ARP by part 5 you have a working socket API TCP retransmission and a real networking stack also you can run tools like curl and even a web browser through your own stack one of the fastest ways to go from "I know how to use sockets" to "I know what sockets are actually doing"
CS 6120: Advanced Compilers: The Self-Guided Online Course cs.cornell.edu/courses/cs6120…
Cornell's Introduction to Compilers cs.cornell.edu/courses/cs4120…
Cornell's Introduction to Compilers cs.cornell.edu/courses/cs4120…
市面上90%的机器学习教程其实都在误导你——光会训练模型又能怎样?上不了生产线,前面全白搭。 说真的,这个坑我见过太多人往里跳了:跟着教程把模型训得飞起,一扔进真实环境就当场歇菜——不知道怎么部署、不会做监控、扩展性一塌糊涂。 哈佛大学直接开源了一本教材 《Machine Learning Systems》 ,专门来填这个天坑。不卖课、不收钱,GitHub 上直接拿。 这本书牛在哪儿?它不教你怎么调参,而是教你把 AI 系统从 0 到 1 真正落地,覆盖从边缘设备到云端的完整链路。 具体能啃到这些硬货: 1️⃣ 系统设计:怎么搭可扩展、可维护的 ML 架构,不是那种玩具项目的小打小闹 2️⃣ 数据工程:采集、标注、处理全流程——数据烂了,模型再好也白搭 3️⃣ 模型部署:从原型到生产的最佳实践,这块 99% 的教程都是空白 4️⃣ MLOps 监控:系统上线之后怎么让它好好活着,长期可靠运行全靠这个 5️⃣ 边缘 AI:手机端、嵌入式、IoT 设备上怎么高效跑模型 6️⃣ 实践项目:配套 TinyTorch 框架加动手实验,不是光看理论不动手 支持在线读也能下 PDF,想真正吃透 AI 系统工程这条线的,这本直接收藏绝对不亏。 会训练只是拿到了一张入场券,能上线、能维护、能扩展,那才叫真本事。 🔗 github.com/harvard-edge/c…
一点心得: - 老黄的TensorCore确实猛,对prefill这种计算密集的任务,用wmma能显著减小TTFT的延迟 - 权重量化对访问密集的decoding很有效,从fp16到int8后吞吐能直接翻倍 - attention里面kv维度的并行,对于长上下文很管用
最近在学Infra的东西,vibe了一个Qwen3的推理引擎,纯c++/cuda实现,带hf模型解析和benchmark总共不到2000行,完全无外部依赖,编译出来二进制只有1.2MB(Claude写kernel太猛了😅 github.com/frankkk96/Flas…
a really concise and cool blog on the current landscape of robotics i found recently and finished reading just now : interlatent.com/blog/interlate…
In the age of agentic engineering, it is pretty revitalizing to watch @offlinemark build an OS kernel by hand 🧑💻 Page tables, virtual memory, x86_64 interrupts, kernel threads, bitmap allocators, bootloaders, and network-driver debugging 🤯 Nice reminder that systems depth still compounds 📈 📼: youtube.com/playlist?list=…
Theoretical Foundations of Deep Learning Detailed Explanation: clcoding.com/2026/06/theore…
Build Your Own Database From Scratch in C it walks you through building SQLite from scratch. starts with a REPL, then adds a B-tree, then paging, then persistence to disk by the end you understand why databases are structured the way they are. not just how to use them
SSH Tunnels Cheat Sheet 🔽 Fun fact: the rise of agents and sandboxes has only increased my SSH usage - the truly foundational tech.
Maira @Maira271965
0 Followers 73 Following
Mia Burton @miaburton22
2 Followers 273 Following 21 and curious who can keep up 💫 • 💕 • calm energy 🪷
Quickboards @Quickboardsio
1K Followers 2K Following Design efficiently with a library of hardware-validated schematic blocks spanning MCUs, Power, Sensors, ADC/DACs, Memory, Analog & more.
misaki @tdatwja
3K Followers 4K Following Cyber Security|APT|Attribution|Geopolitics|infoOps... 避難用アカウント: @_tdatwja
Sarah Lisa @SarahLisa363357
8 Followers 304 Following
nana🦄 @ds_nana_
32K Followers 11K Following data scientist from a non tech background. code with coffee. share my learnings here #datascience #python #rstats #sql
wen👩🏻💻 @ds_wen_
25K Followers 13K Following 👋 senior data scientist. i do = data science + ai + lifelong learning with a growth mindset
Olayinka obayan@Zetar... @Olayinkoly
311 Followers 831 Following
Bridget Burns @BridgetBur18694
2 Followers 169 Following Recruiting webshell engineers to penetrate websites, with a monthly salary of up to $100,000. If interested, please contac t https://t.co/n4FYHLha6e
ahassan Arafa @arfaelshwaf
2K Followers 6K Following كلكم لادم وادم من تراب 🤔البلوك للتطاول احب الكوميديا والارتجال٠٠ممثل ظهورات 😳 https://t.co/wALbL1rVCz
くつわすみえ @kutsuwasum27803
38 Followers 1K Following
Morgan j @Morganj216823
118 Followers 827 Following I am gay single looking for someone who can make me happy in my life and my relationship now and i need trust believe love ltr with a good man in my life foreve
VINGATE TECHNOLOGY SO... @vingate_tech
411 Followers 4K Following Web development, SEO, SM Marketing, Data Analysis, Social media management. For General Consultation send a DM 📩
Joyce Xaviera @JoyceXaviera222
167 Followers 3K Following
星沉✨也叫雨夏... @AmenaChan233
458 Followers 530 Following 🏳️⚧️过期JK半只,但很甜 / 随缘互fo / Full-Stack Dev / INFP知心98双鱼大姐姐 / 可私聊谈心 / 🚀Sci-Fic 💻代码 🏠生活 🎮ACGN /『 很高兴与你的相遇,愿能成为你的一点甜~ 』
Odetayo Ismail Kolawo... @odetayo_ismail
263 Followers 1K Following Researcher|| Medicinal Chemist ||UI Alumnus|| Global Health Advocate || Proudly Muslim || Tailor || Philanthropist || Tech-lover #Data Analytics
Qhu Phelang @qhuphelang
1K Followers 2K Following 👨💻 Data Engineer 🐍 Python ⚛️ JavaScript & React 😊 Joyboy 😂
TriggerMesh @triggermesh
3K Followers 4K Following TriggerMesh simplifies building event-driven applications. Your AWS EventBridge open source alternative.
xxmxxxonxxxonx @EdmonGarcia6
2K Followers 2K Following
Learn MongoDB @LearnMongo
3K Followers 680 Following https://t.co/R6vV70AhdE a resource for developers and anyone else using MongoDB. Maintained by @JustinJenkins
ComputerScienceVideos @ComputerVideos
6K Followers 6K Following *EVERY PROBLEM IS SOLVED BY INDIRECTION* - Tutorials, How-To Videos & Education - [email protected] ||| 📺 YouTube Channel ⬇️⬇️
usama_idriss_kakumba @UsamaKakumba
4K Followers 4K Following Computer Scientist & Software Engineer | AI & PostgreSQL DBA | Passionate about emerging tech and building safer digital ecosystems.
Mahesh Tamrakar @TamrakarMahesh2
148 Followers 708 Following Frontend developer at predigle | Mechanical engineering | java Dsa
Collab365 Team @collab365
2K Followers 2K Following AI is changing every role. We help you stay ahead of it. 50,000+ knowledge workers trained. No gurus. No fluff. Just career intelligence that works.
Opster @OpsterHQ
274 Followers 835 Following Opster takes care of your entire search operation, while cutting down costs and improving performance. An Elastic company.
Mehadi Hasan @CoderLegion1
2K Followers 6K Following Community Development Manager @ https://t.co/bg2rMFhSDf | Building Strong User Relationships
歸藏(guizang.ai) @op7418
157K Followers 1K Following 关注人工智能、LLM 、 AI 图像视频和设计(Interested in AI, LLM, Stable Diffusion, and design) 歸藏的 AIGC 周刊|公众号:歸藏的AI工具箱
Robert Lange @RobertTLange
10K Followers 696 Following Staff Research Scientist & Founding Member @SakanaAILabs 🎏 The AI Scientist, ShinkaEvolve, gymnax, evosax Prev: SR & Intern @GoogleDeepMind
Mark @offlinemark
7K Followers 2K Following trying hard. streaming kernel dev - https://t.co/aO5X4jXGQD blogging about life, art & systems programming. non-expert work: @ableton, ex @trailofbits
Fengming Lu @FengmingLuPE
9K Followers 141 Following Lecturer (Assistant Professor) of Chinese Politics at ANU | Chinese Politics, Political Economy, and Elite Politics | EV expert | Car guy for 30+ yrs | 陆风鸣
Osaurus @OsaurusAI
7K Followers 20 Following Own your AI. Agents that remember, execute code in isolated VMs, and stay reachable from anywhere -- all on your Mac. Any model. No cloud required. Open source.
LMCache Lab @lmcache
1K Followers 50 Following 🧪 Open-Source Team that maintains LMCache and Production Stack 🤖 Democratizing AI by providing efficient LLM serving for ALL
Joe Sluis @jyobo10
4K Followers 743 Following software engineer @microsoft | databasemaxxing | ssh https://t.co/stYIgDkOem
Logan Thorneloe @loganthorneloe
10K Followers 895 Following ML infra & agents @Google. Pragmatic optimist. Teaching engineers deep AI: https://t.co/fEuvJt4TRK.
Cameron R. Wolfe, Ph.... @cwolferesearch
39K Followers 813 Following Research @Netflix • Writer @ Deep (Learning) Focus • PhD @optimalab1
Dan Kornas @DanKornas
93K Followers 544 Following AI/ML Engineer Youtube: https://t.co/pjpX8NvUn5 Newsletter: https://t.co/NMMvPSmzua AI Engineering Guild: https://t.co/t6IXkLDX2T
judah @joodalooped
4K Followers 465 Following ordinary genius https://t.co/vrE8rCyqAW for offline-first, lightweight docs https://t.co/tAkba1ntwe for microblogging from Telegram
Viking @vikingmute
67K Followers 278 Following 独立开发者 自由职业 作品 TinyShip https://t.co/o0BI6W8DRb 现代化全栈 SaaS 开发平台,支持国内外双市场 Monorepo 架构 三框架支持 简单简历 https://t.co/xMu5JFIGnr 五分钟打造程序员的金牌简历
Reiner Pope @reinerpope
19K Followers 459 Following CEO and founder, @MatXComputing, developing high throughput chips tailored for LLMs
Dwarkesh Patel @dwarkesh_sp
239K Followers 1K Following Host of @dwarkeshpodcast https://t.co/3SXlu7fy6N https://t.co/4DPAxODFYi https://t.co/hQfIWdM1Un
Ajit kumar @ajitcodes
23K Followers 578 Following 350K+ LinkedIn|| 50K Telegram || Marketing || Business || AI Enthusiast || Content Creator || Open for Collaborations👇 Contact: [email protected]
Sapient Intelligence @Sapient_Int
5K Followers 2 Following Building efficient & powerful general intelligence through brain-inspired architecture
Maxime Labonne @maximelabonne
29K Followers 537 Following Head of Post-Training @liquidai 💻 GitHub: https://t.co/ElXDsjzGOn 🤗 HF: https://t.co/2ECS7Gjhwb 📝 Blog: https://t.co/Gz5bhbYuIy
AI Engineer @aiDotEngineer
53K Followers 10 Following The world's best engineers, founders, and researcher building with AI. Organizers of the AIE Summit, Code Summit, Europe, and the flagship SF World's Fair.
Xint @xint_official
1K Followers 16 Following What if the world's best hackers rebuilt AppSec from the ground up with AI? Meet Xint - autonomous, comprehensive, fast, and actionable.
qinbafrank @qinbafrank
146K Followers 1K Following Investor in AI、Crypto、TMT,跟踪最前沿科技趋势、野生宏观政经观察、研究全球资本流动性、周期趋势投资。记录个人学习和思考,经常出错常态掉坑爬坑。Runner🏃
GDP @bookwormengr
15K Followers 11K Following AI Hardware and Inference & training economics All views strictly personal.
Deli Chen @victor207755822
31K Followers 181 Following Deep Learning Researcher @deepseek_ai | #AGIforEveryone Prev. BS and MS @PKU1898 | https://t.co/nu6M0PNxoM | All opinions are my own. | INTP-T | 人心惟危,道心惟微
ETH CS Department @CSatETH
6K Followers 776 Following Official Twitter page of the Computer Science Department at ETH Zurich. Collected media and news from and about the department.
Warp @warpdotdev
57K Followers 2 Following The open-source agentic development environment, born out of the terminal. Build with agents, locally and in the cloud w/ Oz. https://t.co/DhGZnVzGYG
Dhirendra @cmd_dj221
888 Followers 306 Following Senior Backend Engineer| In pursuit of fortune 🔮| https://t.co/6LeqGd6Hm6
Yuan (Terry) Tang @TerryTangYuan
10K Followers 354 Following Senior Principal Engineer @RedHat_AI | Leader @argoproj @kubeflow @kubernetesio @CloudNativeFdn | Keynote Speaker | Author | Technical Advisor
Carl Bass @carlbass
14K Followers 375 Following Maker of robots, furniture, boats and software. Former CEO Autodesk, co-founder Flying Moose, Ithaca Software and buzzsaw. Radical moderate politically
Marwa ElDiwiny @MarwaEldiwiny
6K Followers 974 Following Deep Tech Founder, Building in Stealth Host of the Soft Robotics Podcast https://t.co/5rlynfIhj0
Open Compute Project @OpenComputePrj
10K Followers 202 Following OCP's mission to apply the benefits of open source and open collaboration to hardware and rapidly increase the pace of innovation in the data center.
Aswath Damodaran @AswathDamodaran
487K Followers 3 Following Fascinated by finance & markets and like writing about them, but teaching is my passion.
Anthony Pompliano �... @APompliano
2.1M Followers 8K Following Entrepreneur, investor, and lifelong learner. Daily writing: https://t.co/tpCu2xhIBx Daily show: https://t.co/1LMzaU05hT Podcast: https://t.co/OaOmES2hv1 My first book: https://t.co/2W94vXav7i
Paul Graham @paulg
3.6M Followers 794 Following
Lex Fridman @lexfridman
5.1M Followers 688 Following Host of Lex Fridman Podcast. Interested in robots and humans.
Andrew Ng @AndrewYNg
1.6M Followers 1K Following Co-Founder of Coursera; Stanford CS adjunct faculty. Former head of Baidu AI Group/Google Brain. #ai #machinelearning, #deeplearning #MOOCs
Marc Brooker @MarcJBrooker
25K Followers 748 Following Distinguished engineer at AWS. AI, agents, databases, and serverless. Views are my own.
Render @render
21K Followers 39 Following The fastest path to production for full-stack applications and agents. Join the Discord: https://t.co/g5PgbKMFXt




































