@oasisfeng 我之前也是通过软链来整合，但是现在把全部相关项目都放在一个monorepo了。然后一些需求我直接跑在kvm虚拟机里面，整个都是沙盒。
中文
TSAO
398 posts
@tsaokoming 上下文中指定并不方便（比如无法自动完成路径引用），涉及修改外部路径中的文件时被沙盒安全边界约束。
在 Codex 等工具中，我是通过在当前 repo 路径下创建「符号链接」这种 hack 方式来整合别的 repo 资料。
中文
Antigravity 2 总算有了个领先对手的设计：Project 不再捆绑单一路径 (repository)，现在可以将多个路径纳入其中。
这样我就可以把 研究资料 (Obsidian Vault)、代码工程 和 博客创作台 全部整合到一个 Project，让沟通上下文和参考信息引用在三个路径间自由逡巡。
不愧是有信息整合基因的 Google。
中文
最近公司 领导 和我聊起面 向对象（OOP）和设计模式。并说想要在公司中推广设计模式。但是我表示了 反对。
我想聊聊我为什么不喜欢设计模式，以及我信奉的软件 “设计模式”。
说实在的，我其实没看过 GoF 那本设计模式的书。以后也大概不会去读。但是设计模式的大名还是有听过
我之前在 Vue 团队写 Vapor 模式时。发现 Vue 的响应式系统确实使用了 OOP，但主要是出于性能上的考虑。当优化达到一定层级之后，哪怕是引擎层面的微小实现差异也会影响性能。所以我们选择 OOP，是因为它带来的优化便利，并不是因为我们想要套用某种具体的设计模式。
* Vue 的源代码中 ReactiveEffect、Dep、ComputedRefImpl 这些类的存在，是因为 V8 的 hidden class / inline cache 对固定形状的对象优化更好。
（如果理解有误，欢迎指出）
我在入行的初期学过 OOP，大约是 8 年前的事了，已经很久了。但是过去这几年，我逐渐觉得设计模式并不太适合 Vue 3 + TypeScript 的开发。如果不是为了性能优化，我现在也很少使用面向对象。我会这样看，一部分原因是受到了像 Anthony Fu (antfu) 这些人的影响。
* GoF 是 1994 年的 C++/Java 世界的产物。如今语言进化了，很多模式从"需要刻意设计"退化成了"语言原生特性"。 这就是为什么 antfu、Evan、Anthony Fu 那一派人不太谈 GoF。不是反对它，而是它已经融进语言里了，再单独命名反而显得笨重。
“ 模式之所以叫模式，是因为语言没原生支持它，你得手动重复实现。”
去年我读了一本关于如何更好的编程的书，我很快就爱上了它。我试着推荐给我的推友们，结果发现推友中已经有不少人读过了。而且也很喜欢！
这本书是《软件设计的哲学》(A Philosophy of Software Design)
我特别喜欢书中对复杂度的阐释方式，现在仍会时常翻阅。
我的 CLAUDE.md 提示词里也用了这本书的概念：
- 深模块（Design Deep Modules）
- 消除错误（Define Errors Out of Existence）
- 向下沉淀复杂性（Pull Complexity Downwards）
- 不同层不同抽象（Different Layer, Different Abstraction）
这边篇幅有限，就不展开这几个概念了。如果感兴趣的话，我可以之后仔细写写我自己的理解。
让我疑惑的问题：
设计模式如今还被使用，似乎更多是一种沟通时的共同语言。但是我们是否真的按照它的模式来设计代码架构呢？
它作为大家交流的语言和思想的工具如今是否还是合适的选择。是否落后，又是否有着与时俱进的新选择呢？
* 写了好多，以上都是我自己的主观想法，可以随便聊 🤗
中文
I can't help but feel personally burned by the Claude Code changes announced today.
We put so much work into wrapping the (atrocious) Claude Agent SDK in T3 Code. It was the ONLY path they supported, so we made it work. It was hell.
Now our users are getting their rate limits cut by 40x, despite us doing everything right.
I listened to the Claude Code team. I had my issues with their direction, but I trusted them and took them at their word.
I will never make that mistake again.
Until we see significant change, it is safe to assume any statement from an Anthropic employee is a lie on a timer.
The rug will be pulled, no matter how many promises are made beforehand.
English
TSAO retweetledi
We're thrilled to partner with @Google on something we've been building with them - Googlebook.
Premium, powerful devices designed for Intelligence. We can't wait to get it into your hands this fall.
Learn more at ms.spr.ly/6013vuzjt
#Googlebook #NEXT #Intel
English
TSAO retweetledi
TSAO retweetledi
@mweinbach That's fair, it's not good as gpt-5.5 on my side, overall it's a very good model.
English
Yea deepseek v4 flash/pro don't really perform that well compared to any of the major US models, even 1-2 revisions old. Looks like it's slightly behind Opus 4.5 in practice, and on par or slightly behind Kimi K2.6
Some good optimization techniques there, but overall, eh
English
TSAO retweetledi
🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length.
🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models.
🔹 DeepSeek-V4-Flash: 284B total / 13B active params. Your fast, efficient, and economical choice.
Try it now at chat.deepseek.com via Expert Mode / Instant Mode. API is updated & available today!
📄 Tech Report: huggingface.co/deepseek-ai/De…
🤗 Open Weights: huggingface.co/collections/de…
1/n
English
I’m done considering Claude. You’ve burned me too many times!
ClaudeDevs@ClaudeDevs
Over the past month, some of you reported Claude Code's quality had slipped. We investigated, and published a post-mortem on the three issues we found. All are fixed in v2.1.116+ and we’ve reset usage limits for all subscribers.
English
TSAO retweetledi
nvidia enterprise relations moves faster than any software company i've dealt with.
mar 21, mark cai(nvidia) dm'd me after i posted honest benchmark data where nemotron lost to qwen on some agentic tasks. no defensiveness, just "love to hear honest data."
apr 3, dgx spark shipped straight from santa clara to bangkok, no deck to fill out, no strings attached, no gatekeeping, just a tracking number in my inbox.
apr 9, first call with mark himself, 30 min teams. mostly listening, and the line that stuck was "we want to provide value."
every other lab i've pinged at this scale never even replies at all. the enterprise relations team that bets on honest builders wins. every time.
English