HNHacker News
TopNewBestAskShowJobs

azuanrb

398 karma · joined October 4, 2022

I'm a software engineer based in London, originally from Malaysia (we have great food!)

https://azuanz.com

submissionscomments
azuanrb··on The last six months in LLMs in five minutes
It depends on what you’re comparing it against. For $20, OpenAI is still probably the best value for SOTA models. In terms of limits, you can use GPT-5.4 instead of 5.5. The intelligence feels similar, but it’s cheaper. You can also experiment with other harnesses like pi. It’s lightweight but capable enough, and its token usage is definitely much more efficient.
azuanrb··on No More JetBrains Products for Me
> Granted, I tend to run older hardware, but it seems that most other programs on my machine run fast and happy

Curious what hardware you’re on. I’m in the same camp with JetBrains products, performance has always been my biggest complaint. Apple M chips made a huge difference though. It’s still not my preference, but at least it’s a lot more usable now. Most of my colleagues run multiple instances daily without issues.

azuanrb··on I am worried about Bun
If I had to guess, it comes down to speed of iteration. Claude Code is built on JavaScript, so Bun aligns well with their current stack.

Switching to Go or Rust would only make sense if performance were the main priority, which doesn’t seem to be the case. Their current setup lets them ship quickly. A rewrite in Go would likely slow that down.

Codex moved to Rust, and you can see the trade-off. Performance improved, but release velocity dropped. They’re also still catching up to Claude Code, so they don’t face the same pressure to ship as fast.

azuanrb··on Claude Code refuses requests or charges extra if your commits mention "OpenClaw"
It's probably their system prompt. Unlike Claude Code, they don't ban you for using different harness with their subscription (for now). If you use pi, their "safety" is off. Works great for me.
azuanrb··on DeepSeek v4
I have $20 ChatGPT subscription. Stopped Anthropic $20 subscription since the limit ran out too fast. That's my frontier model(s).

For OSS model, I have z.ai yearly subscription during the promo. But it's a lot more expensive now. The model is good imo, and just need to find the right providers. There are a lot of alternatives now. Like I saw some good reviews regarding ollama cloud.

azuanrb··on Claude Code to be removed from Anthropic's Pro plan?
In the end, companies are made of individuals. In my previous company, I'm one of the individual who advocate for Claude Code adoption. Amongst my circle too, most who have the authority to make an impact on the AI direction. Safe to say I no longer suggest Anthropic to anyone anymore.
azuanrb··on Anthropic says OpenClaw-style Claude CLI usage is allowed again
FAANG already did this all the time isn't it? Regardless of their policy. US is no better than China from my point of view. In this case, I see no difference between sending my prompts to US or China companies. At least China models are open source.
azuanrb··on Not buying another Kindle
One of the last physical books I owned was The Three-Body Problem trilogy in a single volume. It was massive, heavy and honestly beautiful. I loved it.

Then I switched to the Kobo Libra Colour. The weight and portability make a huge difference. Having my entire library with me means I am no longer stuck with whatever I decided to bring before leaving home.

The color display is not amazing, but it is good enough for comics. I have been reading things like Attack on Titan and Spider-Man on it. Reading tech books has been great too, especially those with graphs and images.

If I had to sum it up in one word, versatile.

azuanrb··on Anthropic downgraded cache TTL on March 6th
As a Pro user, even though these issues and bugs are “new,” the downgrade has been noticeable since January. I’ve unsubscribed because the Pro plan is no longer usable for me.

It’s only making the news now because it’s affecting Max users as well ($100/$200 plans). I understand the need for change, but having zero communication about it is just wrong.

azuanrb··on ChatGPT Pro now starts at $100/month
Same for me. I'm on $20 plan for both and I use them both interchangeably. Similar "intelligence" imo. Just different way of doing things, that's all. But Claude is getting worse in terms of token usage so I've cancelled my plan last month.
azuanrb··on GLM-5.1: Towards Long-Horizon Tasks
Have you compared it with using Claude Code as the harness? It performs much better than opencode for me.
azuanrb··on Tailscale's new macOS home
I’ve been using Bartender (paid) and Thaw (free) to manage my menu bar. Recently, both apps have become quite buggy. I’m not sure whether this is due to macOS or if there are better alternatives I’m not aware of.
azuanrb··on GPT-5.4
5.2 to 5.3 is the big leap for coding agents, so I'd say you're already missing out quite a bit.
azuanrb··on Setting up phones is a nightmare
I’m not sure about the formal evidence, but to me it’s quite straightforward. I installed the official banking apps on my parents’ phones and told them to use only those apps for banking. Nothing else.

If they use a website, they might mistype the URL or click on a fake link. They don’t really use bookmarks either. Even if they manage to reach the correct site or add it to their home screen, they still have to log in again each time they use it. The app removes all of that—they just tap one icon. Passkeys may improve web security, but they still have a long way to go for non-technical users.

Notification sucks, but overall I’m still in favor of it until we have something better. I’ve had to replace my credit card a few times due to fraud cases, and the same happened to my parents. Because we get instant notifications for every transaction or authentication attempt, we know immediately if something suspicious happens. That’s not really possible with web push, especially since our banks don’t support it.

Authentication is much better in the app as well. In the past, banks would ask lots of personal questions over the phone, which I’m no longer comfortable answering unless I made the call myself. With the app, the account is tied to one device, and biometric login makes it easy and secure. Any approval happens inside the app, which feels safer.

App is definitely not perfect, but for the majority of people (non-tech), it's just a lot easier and (unfortunately) safer for them.

azuanrb··on Setting up phones is a nightmare
In my country, we do. For starters, we use apps for authentication and notifications. We can debate whether web push is viable, but most banking apps simply do not rely on it. As for older people, I would rather they use an app than a website because it is far too easy to fall victim to phishing attacks, no matter how much we educate them.
azuanrb··on Google restricting Google AI Pro/Ultra subscribers for using OpenClaw
I’m aware of the tweet that says otherwise, but until they update their legal documentation, it’s still not allowed.

> OAuth authentication (used with Free, Pro, and Max plans) is intended exclusively for Claude Code and Claude.ai. Using OAuth tokens obtained through Claude Free, Pro, or Max accounts in any other product, tool, or service — including the Agent SDK — is not permitted and constitutes a violation of the Consumer Terms of Service.

https://code.claude.com/docs/en/legal-and-compliance#authent...

azuanrb··on Claws are now a new layer on top of LLM agents
When I tried it out last time, a lot of the features are macOS only. It works on other OS, but not all.
azuanrb··on Gemini 3.1 Pro
The CLI needs work, or they should officially allow third-party harnesses. Right now, the CLI experience is noticeably behind other SOTA models. It actually works much better when paired with Opencode.

But with accounts reportedly being banned over ToS issues, similar to Claude Code, it feels risky to rely on it in a serious workflow.

azuanrb··on Anthropic officially bans using subscription auth for third party use
No, it is prohibited. They're just updating the docs to be more clear about their position, which haven't changed. Their docs was unclear about it.
azuanrb··on Anthropic officially bans using subscription auth for third party use
Usually, it is already stated in their documentation (auth section). If a statement is vague, treat it as a no. It is not worth the risk when they can ban you at any time. For example, ChatGPT allows it, but Claude and Gemini do not.

https://developers.openai.com/codex/auth

azuanrb··on Anthropic bans OAuth tokens (including Agent SDK) in 3P tools
OpenAPI openly encourage users to use their subscription with their SDK and 3rd party tools like opencode, openclaw. Until they change it, they're still better than Anthropic subscription.
azuanrb··on AI adoption and Solow's productivity paradox
Curious why you like Conductor. I’m trying it out, but since I primarily live in the CLI, I might not see much value in it.
azuanrb··on AI adoption and Solow's productivity paradox
$20 is not useable, need $100 plan at least for development purposes. That is a lot of money for some countries. In my country, that can be 1/10 of their monthly salary. Hard to get approval on it. It is still too expensive right now.
azuanrb··on Show HN: I built a tool to un-dumb Claude Code's CLI output (Local Log Viewer)
Based on their Terms of Service, we are not allowed to use a Claude Code subscription outside of Claude Code itself. Although it may work with tools like pi or other harnesses, doing so puts your account at risk of being banned.

Claude Code is not developer friendly.

azuanrb··on Improving 15 LLMs at Coding in One Afternoon. Only the Harness Changed
Claude Code better than opencode for GLM models for me.
azuanrb··on GPT-5.3-Codex
Codex by a mile. Also, there's double rate limit until April. So you're paying 1 month for 2 months usage.
azuanrb··on Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT
Agree. Just that it used to be, just not anymore after December update.
azuanrb··on Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT
You have 5 hours limit (not 4), and weekly limit. But once you hit your weekly limit, you won't be able to use it anymore. That's how it works for a while now.

Not sure why you're being so hostile here. Since you mention you're on $200 plan, your rate limit is a lot higher so you probably won't notice it as much as $20 plan which to me is understandable.

I'm just sharing my experience using the same $20 plan, before and after December 2025. The difference is noticeable, that is all.

azuanrb··on Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT
Based on your other reply, seems like you're on Max5 plan ($100/m), not Pro ($20/m)
azuanrb··on Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT
Not sure what you mean by incorrect since you already validated my point about the limits. I never had these issues even with Sonnet before, but after December, the change has been obvious to me.

Also worth considering that mileage varies because we all use agents differently, and what counts as a large workload is subjective. I am simply sharing my experience from using both Claude and Codex daily. For all we know, they could be running A/B tests, and we could both be right.

← PreviousPage 3 of 5Next →