Hacker News: jiehong

New comment by jiehong in "Local-First and Portable CI"

jiehong — Thu, 28 May 2026 21:27:21 +0000

Dagger seems more fit, but more complex for sure.

New comment by jiehong in "No more JetBrains products for me"

jiehong — Tue, 19 May 2026 09:23:20 +0000

It's not that easy. LSP does not have a standard way to expose things not specified. So, at best you need an LSP server exposing more + the LSP client to also understand it [2].

The best that comes to mind is nvim-jdtls [0].

Maybe at some point the LSP spec could be extended with more refactoring, or maybe there could be a open rewrite LSP server [1].

I wouldn't expect intellij to release a LSP server with all the refactoring tools, though.

[0]: https://codeberg.org/mfussenegger/nvim-jdtls

[1]: https://docs.openrewrite.org/

[2]: https://microsoft.github.io/language-server-protocol/specifi...

New comment by jiehong in "No more JetBrains products for me"

jiehong — Mon, 18 May 2026 21:18:03 +0000

Jetbrains’ good deterministic refactoring tools are what I like from them (and debugging).

Other than that, I must agree with this article.

New comment by jiehong in "Where to buy a non-Apple, non-Google smartphone"

jiehong — Sat, 16 May 2026 12:13:16 +0000

HarmonyOS from Huawei is no longer based on Android, but it’s not an open OS.

New comment by jiehong in "Reimagining the mouse pointer for the AI era"

jiehong — Wed, 13 May 2026 10:47:53 +0000

Now, with vim/helix key bindings:

1. select text

2. dictate action

Feels very similar to Helix's select text and act on it.

I think text selection could also be voice controlled (with a modal voice input), so one could say: "select sentence, action mode, copy and paste it in my list and remove duplicates"

New comment by jiehong in "Reimagining the mouse pointer for the AI era"

jiehong — Wed, 13 May 2026 10:44:48 +0000

transcription gets post-processed by a LLM (with different styles, like based on prompts, so that it removes fillers, homophones, change the style, etc.

I recommend the youtube channel @afadingthought to see what people come up with (like v=283-z29TXeM).

New comment by jiehong in "SecurityBaseline.eu"

jiehong — Wed, 13 May 2026 10:28:21 +0000

It's a good way to ensure that people outside of Germany pentest German sites instead :D

New comment by jiehong in "DeepSeek 4 Flash local inference engine for Metal"

jiehong — Sun, 10 May 2026 19:16:27 +0000

Thanks for taking the time to provide this info. I appreciate it

New comment by jiehong in "Distributing Mac software is increasing my cortisol levels"

jiehong — Sun, 10 May 2026 09:15:08 +0000

This made me think of a potential workaround that would have a better UI: create a directory (maybe named 'Trust Zone'), and have it watched by something like automator. As soon as you move a file in there, it reacts and remove the quarantine bit on that file.

The enrollment app on Mac OS feels like a second citizen. I think Apple is slowly giving up on ensuring the experience on its own Mac apps is working as well as on iOS. I hope to be corrected now that the Macbook Neo is more popular, but I won't be holding my breath.

New comment by jiehong in "The surprisingly complex journey to text-selectable client-side generated PDFs"

jiehong — Fri, 08 May 2026 17:57:12 +0000

This page made me discover the tool, and I find sdocs.dev pretty cool!

New comment by jiehong in "Agents need control flow, not more prompts"

jiehong — Fri, 08 May 2026 06:27:41 +0000

This might be inherent to how the models are benchmarked.

Aren’t some benchmarks giving the model multiple shots at a problem and only keep the successful result if it appeared, ignoring the failure rate?

New comment by jiehong in "Principles for agent-native CLIs"

jiehong — Thu, 07 May 2026 20:28:11 +0000

This reminds me that agents sometimes really like heredoc in shells, and waste tokens retrying with a file.

New comment by jiehong in "DS4, a specialized inference engine for DeepSeek v4 Flash"

jiehong — Thu, 07 May 2026 18:51:21 +0000

Using it in Kagi Assistant is stupidly slow. I get like 10 t/s.

While it’s pretty fast in the official app for example.

Kagi Assistant is also kind of broken when using Qwen 3.6 Plus.

So, beware of using them in Kagi at the moment.

New comment by jiehong in "Making LLM Training Faster with Unsloth and NVIDIA"

jiehong — Thu, 07 May 2026 18:35:18 +0000

I think nowadays a lot of models are trained more at doing this than at knowing things, while being smaller. So I’d say yes!

At least that’s my impression.

New comment by jiehong in "Advancing voice intelligence with new models in the API"

jiehong — Thu, 07 May 2026 18:31:42 +0000

Looks like the GPT‑Realtime‑Whisper model isn’t open weight like the old whisper model. Too bad!

However, OpenAI had and still has a true lead on voice model interactions. That’s where Chinese AI companies don’t do as well: deepseek doesn’t have anything or like Kimi that can speak out in any language except English or Chinese.

New comment by jiehong in "RSS feeds send me more traffic than Google"

jiehong — Thu, 07 May 2026 18:17:43 +0000

CDNs exist, as well as http cache headers, though.

New comment by jiehong in "Hand Drawn QR Codes (2025)"

jiehong — Tue, 05 May 2026 10:18:37 +0000

this makes me wonder if there is an `infinite` version of QR Code that can be streamed from a printer (well, I guess things can be chunked).

New comment by jiehong in "Hand Drawn QR Codes (2025)"

jiehong — Tue, 05 May 2026 10:15:09 +0000

I once sent a letter to a friend with a hand-written QR Code whose content was the typed letter.

It's cool for the receiver, but tedious for the sender (but it's a good way to help with difficult to read cursive).

New comment by jiehong in "Clojurists Together – Q2 2026 Open Source Funding Announcement"

jiehong — Sun, 03 May 2026 06:57:08 +0000

I can see why people try this way: graalvm is limited outside the entreprise version, and it’s quite slow to compile with.

I think the aot compilation story on the JVM lacks fast tooling with good UX compared to go.

New comment by jiehong in "This Month in Ladybird – April 2026"

jiehong — Sun, 03 May 2026 06:44:16 +0000

Congratulations!

However, the screenshots for "List markers in RTL text" are the same it seems. The list markers are on the left in both cases.