DietPi - the lightweight Debian based Linux distribution.Our major changes:DietPi-Wireguard: New tool to manage WireGuard servers and clients
DietPi-Software: Software packages are setup via names instead of IDs
HAProxy: Optimized installation (apt) without recompile needed
DietPi-Config: Smaller improvements
Quartz64: Support for ATAPI&#x2F;SCSI CD&#x2F;DVD&#x2F;Blu-ray drives
UTM VMs: Now with ARM and UEFI support
DietPi-Network: Several improvements to our new network configuration tool
Fixes for DietPi-Drive_Manager, DietPi-Cloudshell, UrBackup, Immich, Immich Machine Learning, Tor Relay, Kubo, NextcloudOf course, there are further changes and fixes. See also: https:&#x2F;&#x2F;dietpi.com&#x2F;docs&#x2F;releases&#x2F;v10_8How to update? Just run dietpi-update.In addition, it would be nice, if you would contribute to our project or support us with a small donation: https:&#x2F;&#x2F;dietpi.com&#x2F;contribute.html.

DietPi v10.8 has been released

I delete the emails, refresh the screen and the emails are still not deleted.Has anyone else had this issue? It&#x27;s happening on two difference computers, one with Ad-Block and one without any extensions.For me, it has been an ongoing for several weeks now.

Expericencing an issue in Gmail where emails do not delete in Firefox

this community is by far the one i respect the most. you are well educated, well informed builders. you know best.so i ask, what will happen to humans with this ai revolution? im in a pretty depressed mood lately, not because i think it will affect me badly but because i think humanity will have to face extreme hunger, poverty and war.but what do you think? what will happen to us?

Hey, Y Combinator - What will happen to humans?

Is anyone else experiencing slow connections with sites hosted behind cloudflare? my website and server hosted behind cloudflared are experiencing slow loading times for some reason.

Slow Cloudflare Connections

Ask HN: T'was the night of Claude Reset and quota left: Funniest app you built?

How about we stop giving your skills to these guys who are putting us out of work?

Do you want to keep your job? Stop feeding AI models

Now that we have coding agents writing all our code for us, are you still needing to reach out and read books such as clean code and other software architecture books?I think they are now more important than ever as there is more code being pushed out than one can read to see if it is robust as the agents claim.

Ask HN: Do you still read clean code and software architecture books?

There have been, increasingly, a large number of posts [1] on HN in the past few weeks, where developers are slowly coming to terms with the breakneck speed with which Agentic Coding has advanced in the past few months.We all agree that an average software developer today is closely resembling the horse-carriage just when automobiles were becoming mainstream around 120 years ago.And coming to HN daily and finding at least one such lament from a software engineer is becoming quite tiresome.We are all, mostly, rational thinkers and worship logic and reasoning even more than our respective gods.With the context built-up enough, I would like to challenge the community to come up with ways to counter this death march that we find ourselves on. We are now, with every passing day, becoming more and more dependent on Agentic Coding. This is bringing the inevitable end to our profession closer at an exponential rate.It seems we are like a drug-addict that knows that soon, one day, the drug would be a likely end of him, yet carries on consuming it.How can we control this?How do we slow down the pace?How do we bring it to a halt?Is there a cold-turkey way out of this?[1] https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=50230965https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=50188865PS: This a resubmission of my post [https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=50232407]. It got flagged for some reason. I sent a mail to HN, but didnt get any reply.

Ask HN: Can we do something about this impending AI dystopia

I&#x27;ve been trying to enroll for Apple Developer for over a month and nothing works.The Apple Developer app won&#x27;t accept my Canadian passport for identity verification. The enrollment is for a Delaware LLC, so my Apple account region is set to US. I can&#x27;t switch the account region to Canada either, because I haven&#x27;t lived there in years and don&#x27;t have a Canadian bank card.If I enroll through the website, the enrollment gets approved, I accept the term and, when it&#x27;s time to pay, all my cards get declined. I tried ~7 bank cards (debit, credit, personal, business), all with matching billing addresses. All US cards. All the banks tell me they didn&#x27;t decline transactions from Apple. I tried no VPN, VPN and residential VPN. I tried from my MacBook, iPhone, Safari, Chrome, incognito, non-incognito.Phone support is unavailable every time I check. I tried talking to regular Apple Store customer support but after repeating myself over 5 times with different agents, the last person I got transferred to told me they couldn&#x27;t help with Developer stuff.If I email Developer support, they reply ~5 days later with unhelpful canned responses like &quot;have you tried Safari incognito?&quot;, &quot;you should try the Developer app instead&quot; or &quot;make sure your billing address matches your bank information&quot;. I&#x27;m at support email #9 and enrollment attempt #3.Has anyone had similar issues and been able to fix them?

Ask HN: Is it not possible to enroll for Apple Developer anymore?

&quot;At the end  of September, Google Maps updated its satellite images of Gaza from the set taken in the summer of 2023. Entire cities have disappeared, wiped from the surface of the Ea&quot;
 https:&#x2F;&#x2F;www.lrb.co.uk&#x2F;the-paper&#x2F;v48&#x2F;n20&#x2F;eyal-weizman&#x2F;necro-economics

Necro-Economics Eyal Weizman

I&#x27;ve been a Max user of Claude all year and it&#x27;s crucial to my day to day work. Last night I had a new experience which makes me feel like a junior dev who blew thousands of dollars on a misconfigured AWS CDK file.To avoid agents doing online research from leaking private info I created a persona which can do online research as a subagent from a fresh context, I asked it to fact check claims as well. Little did I know subagents don&#x27;t report their token usage and my subagents ended up recursively checking claims. I burned through 80% of my Fable usage for the week in 10 minutes.I&#x27;ve never come close to using my full token allowance before and now I&#x27;m going to have to spend money on expensive usage tokens or upgrade to 20x (or change to a different provider) to complete this weeks work.Anthropic&#x27;s online support agent was terse, &quot;empathizing&quot; but ultimately said it could not do anything.Wouldn&#x27;t the right thing from a Customer Support perspective to give a one-time usage reset in a case like this? If they had a human instead of an agent doing it, would that have changed the outcome?I&#x27;ve implemented bounding now in my persona&#x27;s but it just seems like this is a preventable problem that they&#x27;re (1) not solving and (2) not willing to remedy when it did happen.Full disclosure, I love what Anthropic has built in Claude and use the product daily.

Ask HN: Why Is Anthropic's Customer Support Bad?

What to do if you discover AI models trained using your copywrighted book?

I expect more and more professions getting affected in the next 3 to 5 years, not all on the positive side. Software dev is already changing.What are your plans? Are you planning to quit software altogether? Or stop coding and adapt, accept AI but stay in software field? Or something else? If you’re not in software, what plans then?It would be helpful if you can include what career level are you at (or age range, if you prefer)

Ask HN: What are your career plans in the age of AI?

The &quot;product graveyard&quot; of SaaSHub has been tracking the status (and lifespan) of hundreds of thousands of products. It&#x27;s been hovering around 4y for quite some time. However, there&#x27;s a sharp decline in 2026 leading to about 1.5y these days. <a href="https:&#x2F;&#x2F;www.saashub.com&#x2F;product-graveyard" rel="nofollow">https:&#x2F;&#x2F;www.saashub.com&#x2F;product-graveyard</a> That&#x27;s a huge drop. And, it matches my &quot;human&quot; observations. A lot of products are being vibe coded, released, and soon after that abandoned. It&#x27;s become too cheap to produce software. So, a lot of people have been experimenting for &#x27;free&#x27; with their ideas - only to find out that they weren&#x27;t viable... or that they couldn&#x27;t bring customers in. Distribution and marketing have become much more important than before - within the context that everyone can build &quot;features&quot; for free.Anyhow, I expect this trend of shortening product lifespan to continue even further down. Or maybe, at some point people will realised that there are no users for their products as every user can &quot;build&quot; their own and custom product. Leaving us with a dozen huge platforms that host and help you build those apps. Until then - it&#x27;s a crazy gold rush and a lot of makers are abandoning their empty buckets.

Median product "lifespan" has fallen from ~4y down to ~1.5y

Ask HN: Will There Be Another iPhone Mini?

Hey there, founder of a startup here which main product is an oss project.We&#x27;ve been trying to grow the project for months, and it&#x27;s lowkey growing in stars and adoption. But it&#x27;s linear.We see other projects do big jumps and hit thousands of stars in days. Going viral in HN, X, etc.We believe that real adoption might not come from one day to another, and the teams using us are happy and want to use us more.Now, I hate to spam in places like hn or other forums, but what strategies have you found working? Specially in the realm of AI.Obviously if you ask ai, it&#x27;s always: launch in HN, PH, etc, etc.We post a lot on x, linkedin and such, but would love to push the adoption curve up.Any help welcome!Cheers,Won&#x27;t spam my project here, if you want to check it it&#x27;s on gh -&gt; desplega-ai&#x2F;agent-swarm

Ask HN: How do you grow OSS projects?

You too can get rich quick in real estate --- details in this book; only $49.95.If there really was a sure fire way to get rich quick, why sell books to the public? Why not keep it to yourself and apply it to make money?The same applies to AI. If it was really all that, they wouldn&#x27;t be selling tokens to the public. They would be forming corporate partnerships, building products and conquering markets --- starting with the financial markets.Those who can --- do. Those who can&#x27;t --- sell books and &quot;tokens&quot;.

Is AI just another get rich quick scheme?

I think there is still place for this type of coding where you autocomplete predictable chunks and review it as you go.But I haven&#x27;t seen much development in this space. Are there any notable models optimized for this task or full&#x2F;semi-autonomous coding has taken over?

Ask HN: Are tab completion models dead?

With the age of agents CI costs are generally going up, how is your build setup handling things? Have you been able to optimize, and how did you? Most importantly, how fast is it?

Ask HN: How fast is your CI these days?

steamcmd is the official command line tool from Steam. Gamedevs work with it to get builds uploaded to Steamworks. And it doesn&#x27;t support Apple Silicon.To upload games to Steam, devs either go through Steamworks website, or use steamcmd. The web upload on Steamworks website mostly works for small zips, just it is unreliable. At bad times one can stare at the web upload for 10 minutes of failure, errors, &quot;did not receive the entire file&quot; after timeout.I have finally decided to migrate to upload using steamcmd today. I usually work on a Windows machine and build my game for Windows and Linux. The build and the upload worked flawlessly. It seems steamcmd compares my local build with contents in the depot and only upload some deltas. Saves time and bandwidth.Until I try to set it up on my Mac. Because of the hellish codesign and notarization, it is better (or less bad) to build everything macOS on a Mac. Then this hits me.
&gt; .&#x2F;steamcmd
zsh: bad CPU type in executable: .&#x2F;steamcmd
&gt; file steamcmd
steamcmd: Mach-O 64-bit executable x86_64
In comparison, this is Godot:
&gt; file &#x2F;Applications&#x2F;Godot.app&#x2F;Contents&#x2F;MacOS&#x2F;Godot
&#x2F;Applications&#x2F;Godot.app&#x2F;Contents&#x2F;MacOS&#x2F;Godot: Mach-O universal binary with 2 architectures: [x86_64:Mach-O 64-bit executable x86_64] [arm64]I&#x27;ve cursed Apple because Xcode, codesign, notarization are all hell of a maze to navigate, but the expectations of developing a game on a Mac has always been low (at least when you&#x27;re not developing exclusively for macOS and latest iPhone). I just never thought I&#x27;d be mad at Steam. A developer tool, at 2026, under folder `builder_osx`, doesn&#x27;t support Apple Silicon.My disappointment is immeasurable and my day is ruined.

Tell HN: Steamcmd doesn't work on Apple Silicon

Hey all,
I have a very stupid confession, back when DOGE was first introduced, I didn&#x27;t think much of it and followed a bunch of tutorials on how to mine it on free (and much less limited) Azure account.I used up all of the credits and didn&#x27;t think much of it...Now, I am infra engineer with about 5 years of experience, I have multiple Azure certs, I use it everyday at work but I still cannot get MSFT to unlock my personal account or allow me to create a new one, citing fraud. At the beginning of my career I even got to be a part of the terrible Azure technical customer service and internal requests to lift it off of me didn&#x27;t help.Please advice is there any way to get to be a legitimate user of service again. Thank you.

Ask HN: Mined crypto on Azure when I was 13, after 12y still cannot open account

So, I&#x27;ve created my own OS called LimonOS. It is currently in the early stages of development. Its uniqueness lies in the fact that it is neither Linux nor Unix—in short, the kernel is unique and 98% written in C. Yes, I used an AI assistant for help because I had a headache and encountered many errors and problems since I was working on it and thinking through it alone. Development has been ongoing for 4 days. Here is the link to the OS itself—we recently added a browser and internet connection:https:&#x2F;&#x2F;github.com&#x2F;AnonimusShamshiAlex&#x2F;LimonOSThe code and the entire OS system are there. There is a catch: if you are on Windows, you need to click on the batch file, and everything will compile using WSL Linux and Limine for implementation. Everything is on GitHub, including all instructions, and it will probably be easiest for you to check it out in a virtual machine. I&#x27;m writing this because I want your opinion and advice on how to improve this OS system so that it works and becomes as stable as other operating systems like Linux. It&#x27;s cool to build my own OS like Terry Davis, and I hope I don&#x27;t become a maniac. Here are some screenshots of the interface:Thank you everyone.I recently added AI to the terminal, which you can connect to via the command:aikey &lt;your openrouter.ai API key here&gt;then you can use it with the command ai:ai hi my name Jucyhere are the screenshots of the interface in githubhttps:&#x2F;&#x2F;github.com&#x2F;AnonimusShamshiAlex&#x2F;LimonOS

I'm creating my own OS, check it out, LimonOs no Unix no Linux

I paid $19 for Kimi Code to test K3.1 Pro Standard (not Highspeed) using the official CLI v2.3.1.I ran two tasks on small repositories:1. Recommend movies based on an existing movie database.
2. Analyze a website&#x27;s codebase.Neither task completed. The entire 5-hour quota was exhausted within minutes.I analyzed the actual session logs, including background agents:- Fresh input: 368,203 tokens
- Cached input: 1,183,744 tokens
- Output: 39,324 tokens
- Total: 1.59M tokensAt Kimi&#x27;s published API rates, that&#x27;s just $1.37 worth of usage before hitting the 5-hour limit which is easily less than 5% of what Claude Code offers at $20 subscription.The logs show: &quot;429 You&#x27;ve reached your 5-hour usage limit.&quot;For comparison, Claude Code on a $20 subscription processed roughly 381M tokens over 4-5 hours using Opus 5.5 xhigh, mostly cached input. That&#x27;s not a controlled benchmark, but the difference in usable capacity is staggering.Both Kimi tasks failed without delivering results. I have video evidence and detailed logs, but won&#x27;t publish private repository data.Video: https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=hSr-DLn88AMIs this a quota-accounting bug, or are Kimi&#x27;s $19 subscription limits intentionally this restrictive?Disclosure: I also develop a competing coding agent. These tests used Kimi&#x27;s official CLI.

Ask HN: Kimi Code hit its 5-hour limit after $1.37 of API usage

The fastest way to learn harness engineering? Build Claude Code from scratch.Why harness engineering? LangChain kept the same model and changed only the harness. Their coding agent went from 53% to 67% on Terminal-Bench, a top coding-agent benchmark. From outside the top 30 to the top 5. Same model. Better harness.Over the past few months, I&#x27;ve taken apart Claude Code, OpenCode, and Pi to see how they work under the hood.And you don&#x27;t need to ship a coding agent for this to pay off. The same foundations power any agent, in e-commerce, finance, or sales.On top of that, people are increasingly realizing they need to build their own harness instead of relying on Claude Code or Codex. Ramp scaling its operations with a custom harness is a perfect example.So I turned what I learned into an in-depth course on harness engineering by building Claude Code from scratch. It&#x27;s free and open source: 8 articles, 6 videos, and one codebase. It&#x27;s perfect for people who learn by building. Everything in Python.For full transparency, Modal, Opik, and Kitaru sponsor the course, which lets me keep it 100% free as an independent AI engineer.Here are the 8 steps that I think are core to building and maintaining any harness:1. Design the harness 
2. Build the bare-bones agent loop 
3. Sandbox it, locally and remotely 
4. Apply 4 essential context engineering techniques
5. Add subagents as a context engineering technique
6. Deploy it headless to the cloud: CLI, webhooks, cron
7. Build benchmarks to measure it 
8. Replay failed runs to debug agentsEverything is on GitHub: https:&#x2F;&#x2F;github.com&#x2F;decodingai-magazine&#x2F;building-a-coding-agent-from-scratch-courseCurious what others here think. Which of these steps has been the hardest to get right in your own agents?

Took apart Claude Code, then turned it into a free course on harness engineering

I&#x27;ve already cancelled my OpenAI subscription due to all their recent shenanigans. But before my month ends I thought I&#x27;d see what this Dot thing is all about.First thing I did was install Firefox to test out a project of mine (wouldn&#x27;t work in Chrome). It failed to connect:&gt; The cloud proxy uses an OpenAI certificate authority that the operating system already trusts, but the fresh Firefox profile doesn’t. The game’s connection verifies successfully against that existing authority.&gt; May I add that same cloud certificate authority to this isolated Firefox test profile? Certificate validation stays enabled, but Firefox would then trust certificates issued by the cloud proxy.Wait. What? No. Of course not.Aren&#x27;t Dots ostensibly designed for personal use? Accessing your email and the like. With this thing intercepting web traffic, I imagine almost everyone using these things are in violation of all sorts of policies and agreements.