yangcheng

github failure is a live lesson of dysfunctional org. A business unit like github need a CEO.

cool! it's way faster on desktop. I also recompiled to include 3MF support. However on mobile phones compile is still slow even if I set fa to be lower.

It's looks like antigravity's agent manager or codex app. Guess we have new unified interface now , IDEs have out grown vscode UX

I have tried OpenSCAD, it seems very slow to compile to display on web. are you using the official wasm or some other ways?

Best doctors in China works at state owned hospitals. They may choose to retire or work part time at private hospitals

While the article emphasis on chaos and short time with China's doctor, they are very accessible, 2 USD you visit a fully trained doctor at top hospital, have your CT scan done the same day with 50 USD even if you pay out of packet.

As for AI, my personal experience is ChatGPT and Gemini is more effective then DeepSeek for healthcare issues. I do hope DeepSeek or Doubao can catch up.

it will get more exciting once those solar panel can charge electronic cars

The public dataset only contains 3 or 4 languages. go-280 python-266 js-165 ts-20

I hope in future the benchmark can cover other widely used languages, such as c++, java, swift, rust etc.

I wish Zed can better support claude code, like offering native IDE integration. with Claude code SDK this seems doable?

Having worked at both FAANG companies and startups, I can offer a perspective on AI's coding impact in different environments. At startups, engineers work with new tech stacks, start projects from scratch, and need to ship something quickly. LLMs can wrtie way more code. I've seen ML engineers build React frontends without any previous frontend experience, flutter developers write 100-line SQL queries for data analysis, with LLM 10x productivity for this type of work. At FAANG companies, codebases contain years of business logic, edge cases, and 'not-bugs-but-features.' Engineers know their tech stacks well, and legacy constraints make LLMs less effective, and can generate wrong code that needs to be fixed

Thanks for sharing! would be super nice if notebooklm can automatically include reference papers from a single paper.

latest llama 3.1 is in a different repo, https://github.com/meta-llama/llama-models/blob/main/models/... , but yes, the code is shared. It astonishing that in software 2.0 era, powerful applications like llama has only hundreds of lines of code, and most work hidden in training data. Source code alone is no longer that informative as Software 1.0

Can you elaborate a bit more why render is good? we are on heroku and I have evaluated alternatives every 6 months since heroku/github outage 2 years ago [1]. But I don't see how render is better. 2 years ago render postgres did not have PITR. now they have build it, but Render's postgres offering is even more expensive than heroku, and queries run a bit slower on similar spec machines based on my test. I also don't like render charges per seat in addition to infra cost.

[1] https://status.heroku.com/incidents/2413

Aside from the content itself, the "Listen to Article" button uses a robotic, outdated TTS voice. Shouldn't a company like Google use their latest technologies in public-facing content, particularly when discussing AI progress?

I'm genuinely curious about the decision-making process behind this choice.

First time saw it, would love to try, do I need to uninstall co-pilot plugin to use double?

Agree, twitter layoff set an example that it's ok to layoff. Then all FAANG followed

For my particular cases there are features like shared-mailbox and distribution list that requires additional setups. I don't see a clear advantage to move over a critical provider so I kept things the same. If from dayone I have a choice, I would go with all-in-one.

For example, when sharing a google doc in Gmail, the experience is more feature rich. And if an organization is already paying for gmail, it's hard to stop people from using gdoc and gsheet. In real work, I can't just reply a gdoc share with "use nino or I refuse to read".

Have you thought about use flutter for web as well? It seems a lot of effort to maintain 2 codebase for web and desktop

it's nice that everything is integrated into one system! I have always been searching for similar solutions. But one thing that bothers me email. If I use an email provider from Google workspace or Microsoft365, then sooner or later I was forced to use their calendar, and soon docs because the preview and integration.

The closest I found is larksuite[0] which they recently added email hosting. But I am still paying microsoft so I haven't switched email. I just use lark's IM as email client as well.

While I am trying the product, I am curious what's your thought about add email hosting.

[0] https://www.larksuite.com/

I am getting email from anthropic `Anthropic is inviting you to access the Claude API using the one-time link below:` immediately after the OpenAI outage. I hope it's a coincidence.

but for code llm to be useful , the local machine need to have some very powerful GPU

I beleive google must use domain internally too, they even have a .google

http requests, js execution and a kv store. Actually very useful. You can run many stuffs there from simple webhooks to full CMS

me too! it's unfortunate that more services are starting ban google voice numbers.

For people who use a personal notes tool like obsidian, how would you deal with notes that turned to work related later? I always find myself end up with half-finished notes in multiple different places. First versions in personal notebook and refined ones in company's google doc/etc. Thanks

Brotli is not enabled by default on Android's webview yet.

I have a wide tech foot print but boring stack.

website+e-commence: shopify liquid, later moved to typescript+hydrogen

embedded Android: Kotlin

mobile app: Flutter

Backend: go + postgres on heroku, with some lambdas for periodical work.

The thing surprised me most is the amount of downtime Github and heroku had in past year.

you saved my users! thanks so much

I feel it's impossible for all first 50 to get >1% mathematically. it's more like first 5 other than co-founders.