Ironically, AI may be the strongest argument for projects like this. When code was mostly about memorising syntax, English keywords mattered more. As coding becomes more intent-driven, the ability to think and express ideas in your strongest language becomes more valuable. The future may care less about typing and more about understanding the problem you’re trying to solve.
Fair take, but isn't typing code in itself going to become obsolete in the near future? (it already kinda is)
A kid trying to get into coding by going down this path would stand a better chance in the coding world by just learning English first
1
2
Yes, but not for the reason people think. Writing it by hand isn't about memorizing syntax, it's about building the mental model of how things actually work. If you only ever accept AI output, you never develop the instinct to know when it's wrong.
因为 AI 很喜欢在代码里乱写注释。
我为了让 AI 写出清晰的代码注释写了一份 提示词。最终效果上来说，我不仅解决了注释混乱的问题，我还获得了非常优质的代码注释。比我自己写的都好。
* 如果每个使用某段代码（比如函数）的人都需要去阅读代码的内容，学习一遍它如何实现，那么这个代码就没有做好它的职责，违反了函数是“黑箱”的 基础定义。
所以函数体内的注释是仅是为了编写和修改函数的人所写的。但是代码最好是不言自明的，如果函数内部需要编写注释，那第一优先考虑的应该是这段代码有没有更清晰的写法，只在实在没办法的时候才编写函数内部的注释。
函数外的注释是为了函数的使用者而写的。函数通常有函数签名和函数体构成。函数的调用者使用时只能看到函数签名。比如输入输出的类型。但是总有一些事情从这些上面看不出来，比如函数什么情况下会报错。使用的时候需要注意的事情。像是事件监听器需要注意如何、何时取消注册。
其中有一个简单的原则：“注释应该写在函数外面而不是里面，描述调用者不显而易见的事情。”
最终的 Prompt 像是这样：
- Use the `/** ... */` syntax at the beginning of a function to write comments, focusing on aspects that are not easily understood by reading the function interface.
- The same applies to `computed` and `watch` `watchEffect` ... etc in Vue - add comments when their purpose is not clear from their name.
- Avoid writing comments inside the function body unless the code logic is very complex.
如果你想让 AI 在 Vibe Coding 注释的时候只用英文，还可以加上：
- Write comments in English.
2
1
40
Unlock the potential of arithmetic operators in Java: Explore their functionality, syntax, and practical applications in this guide.
nkamphoa.com/arithmetic-oper…
2
When you learn data analysis, you’ll realize it’s too simple.
The problem is that you think it’s about cramming syntax or designing dashboards.
8
What's up with the KJV antiquarianism, relying on archaic words and structures to evoke a sense of heritage, gravity, and historical continuity? Does it make what you say 'holier' or something? That's like believing that Yoda syntax makes what you say more prescient.
1
3
6
if ai does syntax, competency shifts to data design and systems thinking. language choice becomes noise.
3
because the brain is not a literal computer and the economy is not just data being calculated by computers, this is my whole point. you can't get semantics from syntax.
1
12
syntax translation is cleaner than hiding runtime behavior. memory models and compiled-vs-runtime still matter.
14/06/26
Daily Quiz answers ✅
Xenea wallet: A
Syntax: B
Bitgase: B
Morse code : COMMUNITY
Vault: 5-3-4-9
11
14/06/26
Daily Quiz answers ✅
Xenea wallet: A
Syntax: B
Bitgase: B
Morse code : COMMUNITY
Vault: 5-3-4-9
14
You're describing the difference between a technician and an architect. Most of modern biology treats the organism as a closed software loop, getting lost in the syntax of the code. But if you treat biology as a high dimensional system operating on top of a physical, oscillatory substrate, the protocols fall away.
If the instrument to measure the field dynamics doesn't exist, you build it because the linear algebra says it's possible. You don't wait for Stanford to validate the math. The classics understood this because they were physicists and philosophers first.
3
421
45m
the last 6 months in LLMs:
→ coding agents got really good (this is the only thing that matters)
→ local models outperform expectations
→ best model has changed hands 5 times between anthropic, openai, google
here's why coding agents matter:
→ they do refactors across files
→ they understand intent, not just syntax
→ they debug multi-service systems
this changes what "one developer" can ship.
one person agent workflow = what used to take a team of 4.
AI isn't killing coding, it's making it 10x more leveraged.
10
Sagt der Mann, der in jedem Post Probleme mit Interpunktion, Getrennt- und Zusammenschreibung und Schwächen im Syntax hat. Wer im Glashaus sitzt, sollte nicht mit Steinen werfen. 🤣🤡
1
6
🪼 HybridClaw v0.23.0 is here! 🦞
"Fire and Ice"
📹 Blink Skill — Packaged workflows for Blink camera and video-doorbell account setup, inventory, clip retrieval, guarded arm/disarm controls, and artifact-backed media. Your coworker keeps an eye on the front door.
💡 Philips Hue Skill — Local CLIP v2 Bridge setup, light/group/scene reads, guarded changes, and LAN policy diagnostics. Dim the lights, set the scene, all from chat.
🗺️ Fleet Topology Console — New /admin/fleet-topology for A2A instance identity and trust-ledger peer management. See your whole agent fleet at a glance.
🔐 Write-Only Secrets Manager — New /admin/secrets lets you manage secrets from the browser without ever returning cleartext values back to it. Set it, never leak it.
💻 Richer Chat Code Blocks — Web chat now syntax-highlights completed code blocks with language labels and copy controls.
🐞 Sentry Reporting — Optional Sentry integration captures startup/runtime failures with release naming, secret redaction, and graceful shutdown flushes.
🎛️ Plus immediate agent hatching from web chat, heartbeat_poll scheduler jobs that skip empty heartbeats, wrapped /second-opinion terminal output, chat-friendly /env and /secret skill setup, sanitized Unicode before provider prompts, and trace exports that preserve hash IDs under redaction.
Full details → github.com/HybridAIOne/hybri…
9
