The best Hacker News stories from Show from the past day

from

All Show

from the past*

day week month

with these search terms:

Create follow Go back

Latest posts:

Show HN: Oasis – An open-source, 3D-printed smart terrarium

See the website for a demo video: <a href="https://oasis-terrarium.com" rel="nofollow">https://oasis-terrarium.com</a><p>This project is an enclosure for growing plants. Features include:<p><pre><code> - high-power LED lighting - a built-in water tank and mister - fans for airflow - temperature and humidity sensing - wi-fi connectivity and a web-based interface for control and monitoring (see https://oasis-terrarium.com/docs/usage_guide/#web-interface) </code></pre> The entire project is open-source:<p><pre><code> - electronics: designed in KiCad - 3D models: designed in CadQuery - software: written in Rust </code></pre> I initially started this project during COVID and built a working prototype using a Raspberry Pi and off-the-shelf parts. The first prototype worked, but was pretty rough. Several months ago, I picked up the project again and began a complete re-design. After two rounds of circuit board design and countless iterations of the 3D-printed parts, I'm finally happy with the results. Along the way I learned a lot about electronics design and fabrication, 3D modeling, and Rust.<p>Happy to answer any questions - thanks for checking it out!<p>Here are some direct links to parts of the project you may find interesting:<p><pre><code> - demo video: https://oasis-terrarium.com - detailed build guide with pictures: https://oasis-terrarium.com/docs/build_guide/ - interactive 3D model: https://oasis-terrarium.com/docs/3dmodel/ - detailed info on electronics design: https://oasis-terrarium.com/docs/electronics/ - 3D-printable models: https://www.printables.com/model/1315117-oasis-smart-terrarium</code></pre>

Show HN: Autumn – Open-source infra over Stripe

Hey HN, I’m Ayush from Autumn (<a href="https://useautumn.com/">https://useautumn.com/</a>). Autumn is an open source layer over Stripe that decouples pricing and billing logic from your application. We let you efficiently manage pricing plans, feature permissions, and payments, regardless of the pricing model being used. It’s a bit like if Supabase and Stripe had a baby.<p>Typically, you have to write code to handle checkouts, upgrades/downgrades, failed payments, then receive webhooks to provision features, reset usage limits etc. We abstract this into one function call for all payments flows (checkouts, upgrades, downgrades etc), one function to record usage (so we can track usage limits), and a customer state React hook you can access from your frontend (to handle paywalls, display usage data etc).<p>Here’s a demo: <a href="https://www.youtube.com/watch?v=SFARthC7JXc" rel="nofollow">https://www.youtube.com/watch?v=SFARthC7JXc</a><p>Stripe’s great! But there are 2 main reasons people use Autumn over a direct Stripe setup:<p>(1) Billing infra can get complex. After payments, there’s still handling webhooks, permission management, metering, usage resets, and connecting them all to upgrade, downgrade, cancellation and failed payments states.<p>(2) Growing companies iterate on pricing often: raising prices, experimenting with credits or charging for new features, etc. We save you from having to handle usage-based limits (super common in pricing today), rebuilding in-app flows, DB migrations, internal dashboards for custom pricing, and grandfathering users on different pricing.<p>Ripping out billing flows etc, really sucks. With Autumn, you just make pricing changes in our UI and it all auto-updates. We have a shadcn/ui component library that helps with this.<p>Because we support a lot of different pricing models (subscriptions, usage, credits, seat based etc), we have to handle a lot of different scenarios and cases under the hood. We try to keep setup simple while maintaining flexibility of a native integration. Here’s a little snippet of the architecture of our main endpoint: <a href="https://useautumn.com/blog/attach">https://useautumn.com/blog/attach</a><p>Currently, the users who get the most value out of us are founders that need to move fast and keep things flexible, but also new/non-technical devs that are more AI native.<p>You can clone the project and explore the repo, or try it out at <a href="https://useautumn.com/">https://useautumn.com/</a>, where it’s free for builders. Our repo is <a href="https://github.com/useautumn/autumn">https://github.com/useautumn/autumn</a>, docs are at <a href="https://docs.useautumn.com/">https://docs.useautumn.com/</a> and demo at <a href="https://www.youtube.com/watch?v=SFARthC7JXc" rel="nofollow">https://www.youtube.com/watch?v=SFARthC7JXc</a><p>We’d love to hear your feedback and how we could make it better!

Show HN: Autumn – Open-source infra over Stripe

Hey HN, I’m Ayush from Autumn (<a href="https://useautumn.com/">https://useautumn.com/</a>). Autumn is an open source layer over Stripe that decouples pricing and billing logic from your application. We let you efficiently manage pricing plans, feature permissions, and payments, regardless of the pricing model being used. It’s a bit like if Supabase and Stripe had a baby.<p>Typically, you have to write code to handle checkouts, upgrades/downgrades, failed payments, then receive webhooks to provision features, reset usage limits etc. We abstract this into one function call for all payments flows (checkouts, upgrades, downgrades etc), one function to record usage (so we can track usage limits), and a customer state React hook you can access from your frontend (to handle paywalls, display usage data etc).<p>Here’s a demo: <a href="https://www.youtube.com/watch?v=SFARthC7JXc" rel="nofollow">https://www.youtube.com/watch?v=SFARthC7JXc</a><p>Stripe’s great! But there are 2 main reasons people use Autumn over a direct Stripe setup:<p>(1) Billing infra can get complex. After payments, there’s still handling webhooks, permission management, metering, usage resets, and connecting them all to upgrade, downgrade, cancellation and failed payments states.<p>(2) Growing companies iterate on pricing often: raising prices, experimenting with credits or charging for new features, etc. We save you from having to handle usage-based limits (super common in pricing today), rebuilding in-app flows, DB migrations, internal dashboards for custom pricing, and grandfathering users on different pricing.<p>Ripping out billing flows etc, really sucks. With Autumn, you just make pricing changes in our UI and it all auto-updates. We have a shadcn/ui component library that helps with this.<p>Because we support a lot of different pricing models (subscriptions, usage, credits, seat based etc), we have to handle a lot of different scenarios and cases under the hood. We try to keep setup simple while maintaining flexibility of a native integration. Here’s a little snippet of the architecture of our main endpoint: <a href="https://useautumn.com/blog/attach">https://useautumn.com/blog/attach</a><p>Currently, the users who get the most value out of us are founders that need to move fast and keep things flexible, but also new/non-technical devs that are more AI native.<p>You can clone the project and explore the repo, or try it out at <a href="https://useautumn.com/">https://useautumn.com/</a>, where it’s free for builders. Our repo is <a href="https://github.com/useautumn/autumn">https://github.com/useautumn/autumn</a>, docs are at <a href="https://docs.useautumn.com/">https://docs.useautumn.com/</a> and demo at <a href="https://www.youtube.com/watch?v=SFARthC7JXc" rel="nofollow">https://www.youtube.com/watch?v=SFARthC7JXc</a><p>We’d love to hear your feedback and how we could make it better!

Show HN: Turn a paper's DOI into its full reference list (BibTeX/RIS, etc.)

Show HN: Ariadne – A Rust implementation of aperiodic cryptography

Hello HN, we're CipherNomad, the research initiative behind this project.<p>The Ariadne Protocol is our exploration of a different cryptographic model. The work began with an observation of primitives like the Lion transform, which use a static, hardcoded sequence of operations. This led us to ask: What if the cryptographic "program" wasn't a constant, but a dynamic, history-dependent variable?<p>Our first step was a "Cryptographic Virtual Machine" that took an explicit list of operations (a "Path"). This worked, but required sharing the Path object—an explicit dependency that needed to be managed.<p>The Ariadne Protocol is the maturation of that idea. It eliminates the explicit Path by making it implicit and emergent.<p>The core design is:<p>The Labyrinth: A large, deterministically-generated binary tree of cryptographic rounds.<p>The Thread: The secret path taken through the Labyrinth. This path is not stored or transmitted. It's rediscovered for each block of data by computing a keyed hash of the CVM's secret state and the public ciphertext chunk: hash(key, state, chunk).<p>This makes the cipher aperiodic: because the state ratchets forward after every block, the sequence of operations is guaranteed to never repeat. It also creates inherent tamper evidence—any modification to the ciphertext "snaps the thread" and turns subsequent output into noise.<p>This is experimental, unaudited alpha software. We are publishing it under CC0 because we believe foundational work like this should be an unrestricted public good.

Show HN: I made a screenshot beautifier

I’ve always seen talented designers and developers post their work, only for it to get lost in the noise. I realized a major reason is that they often just post a simple, unstyled screenshot. As a result, nobody pays attention.<p>My solution is a tool where you can drop in a screenshot, pick from a collection of hand-picked gradients, style it, and export it—all in a few seconds. I’ve carefully designed every element of the app to be efficient, simple, and intuitive.<p>On making it "Free, Forever": I don't plan on ever charging a subscription for Moocup. I dislike the subscription model as much as you do. Instead, my plan is to include a donation option for those who wish to support the project. This allows me to keep my promise of making the app free and useful for everyone, without being burdened by hosting and maintenance costs.<p>My goal is simple: to provide a high-quality tool that is genuinely useful and doesn't waste your time. If you find it helpful, the best thing you can do is share it with others. I've included all my contact links for feedback and suggestions.<p>Happy designing :)

Show HN: Pickaxe – A TypeScript library for building AI agents

Hey HN, Gabe and Alexander here from Hatchet. Today we're releasing Pickaxe, a Typescript library to build AI agents which are scalable and fault-tolerant.<p>Here's a demo: <a href="https://github.com/user-attachments/assets/b28fc406-f501-4427-9574-e4c756b29dd4">https://github.com/user-attachments/assets/b28fc406-f501-442...</a><p>Pickaxe provides a simple set of primitives for building agents which can automatically checkpoint their state and suspend or resume processing (also known as durable execution) while waiting for external events (like a human in the loop). The library is based on common patterns we've seen when helping Hatchet users run millions of agent executions per day.<p>Unlike other tools, Pickaxe is not a framework. It does not have any opinions or abstractions for implementing agent memory, prompting, context, or calling LLMs directly. Its only focus is making AI agents more observable and reliable.<p>As agents start to scale, there are generally three big problems that emerge: 1. Agents are long-running compared to other parts of your application. Extremely long-running processes are tricky because deploying new infra or hitting request timeouts on serverless runtimes will interrupt their execution. 2. They are stateful: they generally store internal state which governs the next step in the execution path 3. They require access to lots of fresh data, which can either be queried during agent execution or needs to be continuously refreshed from a data source.<p>(These problems are more specific to agents which execute remotely -- locally running agents generally don't have these problems)<p>Pickaxe is designed to solve these issues by providing a simple API which wraps durable execution infrastructure for agents. Durable execution is a way of automatically checkpointing the state of a process, so that if the process fails, it can automatically be replayed from the checkpoint, rather than starting over from the beginning. This model is also particularly useful when your agent needs to wait for an external event or human review in order to continue execution. To support this pattern, Pickaxe uses a Hatchet feature called `waitFor` which durably registers a listener for an event, which means that even if the agent isn't actively listening for the event, it is guaranteed to be processed by Hatchet and stored in the execution history and resume processing. This infrastructure is powered by what is essentially a linear event log, which stores the entire execution history of an agent in a Postgres database managed by Hatchet.<p>Full docs are here: <a href="https://pickaxe.hatchet.run/">https://pickaxe.hatchet.run/</a><p>We'd greatly appreciate any feedback you have and hope you get the chance to try out Pickaxe.

Show HN: Pickaxe – A TypeScript library for building AI agents

Hey HN, Gabe and Alexander here from Hatchet. Today we're releasing Pickaxe, a Typescript library to build AI agents which are scalable and fault-tolerant.<p>Here's a demo: <a href="https://github.com/user-attachments/assets/b28fc406-f501-4427-9574-e4c756b29dd4">https://github.com/user-attachments/assets/b28fc406-f501-442...</a><p>Pickaxe provides a simple set of primitives for building agents which can automatically checkpoint their state and suspend or resume processing (also known as durable execution) while waiting for external events (like a human in the loop). The library is based on common patterns we've seen when helping Hatchet users run millions of agent executions per day.<p>Unlike other tools, Pickaxe is not a framework. It does not have any opinions or abstractions for implementing agent memory, prompting, context, or calling LLMs directly. Its only focus is making AI agents more observable and reliable.<p>As agents start to scale, there are generally three big problems that emerge: 1. Agents are long-running compared to other parts of your application. Extremely long-running processes are tricky because deploying new infra or hitting request timeouts on serverless runtimes will interrupt their execution. 2. They are stateful: they generally store internal state which governs the next step in the execution path 3. They require access to lots of fresh data, which can either be queried during agent execution or needs to be continuously refreshed from a data source.<p>(These problems are more specific to agents which execute remotely -- locally running agents generally don't have these problems)<p>Pickaxe is designed to solve these issues by providing a simple API which wraps durable execution infrastructure for agents. Durable execution is a way of automatically checkpointing the state of a process, so that if the process fails, it can automatically be replayed from the checkpoint, rather than starting over from the beginning. This model is also particularly useful when your agent needs to wait for an external event or human review in order to continue execution. To support this pattern, Pickaxe uses a Hatchet feature called `waitFor` which durably registers a listener for an event, which means that even if the agent isn't actively listening for the event, it is guaranteed to be processed by Hatchet and stored in the execution history and resume processing. This infrastructure is powered by what is essentially a linear event log, which stores the entire execution history of an agent in a Postgres database managed by Hatchet.<p>Full docs are here: <a href="https://pickaxe.hatchet.run/">https://pickaxe.hatchet.run/</a><p>We'd greatly appreciate any feedback you have and hope you get the chance to try out Pickaxe.

Show HN: Pickaxe – A TypeScript library for building AI agents

Hey HN, Gabe and Alexander here from Hatchet. Today we're releasing Pickaxe, a Typescript library to build AI agents which are scalable and fault-tolerant.<p>Here's a demo: <a href="https://github.com/user-attachments/assets/b28fc406-f501-4427-9574-e4c756b29dd4">https://github.com/user-attachments/assets/b28fc406-f501-442...</a><p>Pickaxe provides a simple set of primitives for building agents which can automatically checkpoint their state and suspend or resume processing (also known as durable execution) while waiting for external events (like a human in the loop). The library is based on common patterns we've seen when helping Hatchet users run millions of agent executions per day.<p>Unlike other tools, Pickaxe is not a framework. It does not have any opinions or abstractions for implementing agent memory, prompting, context, or calling LLMs directly. Its only focus is making AI agents more observable and reliable.<p>As agents start to scale, there are generally three big problems that emerge: 1. Agents are long-running compared to other parts of your application. Extremely long-running processes are tricky because deploying new infra or hitting request timeouts on serverless runtimes will interrupt their execution. 2. They are stateful: they generally store internal state which governs the next step in the execution path 3. They require access to lots of fresh data, which can either be queried during agent execution or needs to be continuously refreshed from a data source.<p>(These problems are more specific to agents which execute remotely -- locally running agents generally don't have these problems)<p>Pickaxe is designed to solve these issues by providing a simple API which wraps durable execution infrastructure for agents. Durable execution is a way of automatically checkpointing the state of a process, so that if the process fails, it can automatically be replayed from the checkpoint, rather than starting over from the beginning. This model is also particularly useful when your agent needs to wait for an external event or human review in order to continue execution. To support this pattern, Pickaxe uses a Hatchet feature called `waitFor` which durably registers a listener for an event, which means that even if the agent isn't actively listening for the event, it is guaranteed to be processed by Hatchet and stored in the execution history and resume processing. This infrastructure is powered by what is essentially a linear event log, which stores the entire execution history of an agent in a Postgres database managed by Hatchet.<p>Full docs are here: <a href="https://pickaxe.hatchet.run/">https://pickaxe.hatchet.run/</a><p>We'd greatly appreciate any feedback you have and hope you get the chance to try out Pickaxe.

Show HN: Lego Island Playable in the Browser

Show HN: Lego Island Playable in the Browser

Show HN: A Tool to Summarize Kenya's Parliament with Rust, Whisper, and LLMs

Bunge Bits summarizes long parliamentary sessions from the Kenyan National Assembly and Senate. Built with Rust, Whisper v3, and GPT-4o.<p>Sessions are typically 3–7 hours long, mixing English and Swahili. This tool transcribes, chunks, and summarizes them to make political content more accessible and searchable for the public.<p><a href="https://bungebits.ke/summaries" rel="nofollow">https://bungebits.ke/summaries</a>

Show HN: A Tool to Summarize Kenya's Parliament with Rust, Whisper, and LLMs

Bunge Bits summarizes long parliamentary sessions from the Kenyan National Assembly and Senate. Built with Rust, Whisper v3, and GPT-4o.<p>Sessions are typically 3–7 hours long, mixing English and Swahili. This tool transcribes, chunks, and summarizes them to make political content more accessible and searchable for the public.<p><a href="https://bungebits.ke/summaries" rel="nofollow">https://bungebits.ke/summaries</a>

Show HN: I'm building an app to replace Overleaf and Notion

Hi HN,<p>Since 2019, I’ve been working on a writing platform designed for creating complex documents (e.g., theses). I personally use it for everything as it also allows to classify documents in categories so you can organize them efficiently. As of a few months ago, the app is also available in the browser, and you can now invite coworkers to collaborate on a document in real time.<p>The app is somewhat inspired by LyX. It offers an intuitive, modern editor, but users don’t need to know any LaTeX. When it’s time to export, they can choose from a range of templates (IEEE paper, thesis, etc.).<p>A few highlights:<p>- It uses a custom-built block editor that performs well with large documents. Each block is its own contenteditable element (instead of having one massive contenteditable for the whole document)<p>- If you prefer plain text - you can insert a Markdown block and write using Markdown instead<p>- Built-in citation management<p>- Support for cross-references and footnotes<p>- Mermaid diagrams, inline LaTeX equations, and display math are all supported<p>- "To-do" sections help you stay organized while writing<p>You can try it out here: <a href="https://www.monsterwriter.com/" rel="nofollow">https://www.monsterwriter.com/</a>

Show HN: I'm building an app to replace Overleaf and Notion

Hi HN,<p>Since 2019, I’ve been working on a writing platform designed for creating complex documents (e.g., theses). I personally use it for everything as it also allows to classify documents in categories so you can organize them efficiently. As of a few months ago, the app is also available in the browser, and you can now invite coworkers to collaborate on a document in real time.<p>The app is somewhat inspired by LyX. It offers an intuitive, modern editor, but users don’t need to know any LaTeX. When it’s time to export, they can choose from a range of templates (IEEE paper, thesis, etc.).<p>A few highlights:<p>- It uses a custom-built block editor that performs well with large documents. Each block is its own contenteditable element (instead of having one massive contenteditable for the whole document)<p>- If you prefer plain text - you can insert a Markdown block and write using Markdown instead<p>- Built-in citation management<p>- Support for cross-references and footnotes<p>- Mermaid diagrams, inline LaTeX equations, and display math are all supported<p>- "To-do" sections help you stay organized while writing<p>You can try it out here: <a href="https://www.monsterwriter.com/" rel="nofollow">https://www.monsterwriter.com/</a>

Show HN: I'm building an app to replace Overleaf and Notion

Hi HN,<p>Since 2019, I’ve been working on a writing platform designed for creating complex documents (e.g., theses). I personally use it for everything as it also allows to classify documents in categories so you can organize them efficiently. As of a few months ago, the app is also available in the browser, and you can now invite coworkers to collaborate on a document in real time.<p>The app is somewhat inspired by LyX. It offers an intuitive, modern editor, but users don’t need to know any LaTeX. When it’s time to export, they can choose from a range of templates (IEEE paper, thesis, etc.).<p>A few highlights:<p>- It uses a custom-built block editor that performs well with large documents. Each block is its own contenteditable element (instead of having one massive contenteditable for the whole document)<p>- If you prefer plain text - you can insert a Markdown block and write using Markdown instead<p>- Built-in citation management<p>- Support for cross-references and footnotes<p>- Mermaid diagrams, inline LaTeX equations, and display math are all supported<p>- "To-do" sections help you stay organized while writing<p>You can try it out here: <a href="https://www.monsterwriter.com/" rel="nofollow">https://www.monsterwriter.com/</a>

Show HN: I'm a doctor and built a responsive breathing app for anxiety and sleep

Hey HN!<p>I’m an NHS doctor and the founder of Pia (<a href="https://www.piahealth.co" rel="nofollow">https://www.piahealth.co</a>) which developed Lungy (<a href="https://www.lungy.app" rel="nofollow">https://www.lungy.app</a>). Lungy is an iOS app that responds to breathing in real-time and was designed to make breathing exercises more engaging and beneficial to do. It’s been two years since Lungy launched (here’s the original ShowHN: <a href="https://news.ycombinator.com/item?id=34534615">https://news.ycombinator.com/item?id=34534615</a>) and it's had a huge update and complete redesign. We rebuilt the whole app, and added a real-time 3D soft body solver which gives some really cool interactions like blobs / objects that inflate as you breathe. We also made a version for Vision Pro, called 'Lungy Spaces'.<p>My background is as a surgical trainee and I started building Lungy in 2020 during the first COVID lockdown in London. During COVID, there were huge numbers of patients coming off ventilators and patients are often given breathing exercises on a worksheet and disposable plastic devices called incentive spirometers to encourage deep breathing. This is intended to prevent chest infections and strengthen breathing muscles that have weakened. I noticed often the incentive spirometer would sit by the bedside, whilst the patient would be on their phone – this was the spark that lead to Lungy!<p>Since making the first version we’ve made exercises fully customisable (you can dial in exact timings for each breath phase), added new breathing indicators, learning modules, e.g. self-care for anxiety symptoms, and lots of new visuals. The free version gives you access to a new breathing exercise each day, whilst premium unlocks the full library of exercises, exercise data and visuals..<p>The visuals are mostly built using Metal (a couple use SpriteKit) and there are lots to choose from - boids, cloth sims, fluid sims, a hacky DLA implementation, rigid body + soft body sims - each one reacts to breath and touch. The audio uses AudioKit with a polyphonic synth and a sequencer plays generated notes from a chosen scale (you can mess around with the sequencer and synth in Settings/Create Music). The nice thing about the visuals + audio being generative is that the download size is relatively small with no other downloads. We’re still working on improving the breath detection, using ML - currently, it uses microphone input, with optional camera input to guide positioning.<p>We’re also close to finishing the medical device version - <a href="http://lungy.health" rel="nofollow">http://lungy.health</a> - designed as a pulmonary rehab platform for patients with asthma, it should hopefully undergo early trials in the UK in 2026.<p>Thanks for reading - would love to hear any feedback!<p><a href="https://www.lungy.app" rel="nofollow">https://www.lungy.app</a><p>Lungy Version 2 here: <a href="https://apps.apple.com/app/apple-store/id1545223887">https://apps.apple.com/app/apple-store/id1545223887</a>

Show HN: I'm a doctor and built a responsive breathing app for anxiety and sleep

Hey HN!<p>I’m an NHS doctor and the founder of Pia (<a href="https://www.piahealth.co" rel="nofollow">https://www.piahealth.co</a>) which developed Lungy (<a href="https://www.lungy.app" rel="nofollow">https://www.lungy.app</a>). Lungy is an iOS app that responds to breathing in real-time and was designed to make breathing exercises more engaging and beneficial to do. It’s been two years since Lungy launched (here’s the original ShowHN: <a href="https://news.ycombinator.com/item?id=34534615">https://news.ycombinator.com/item?id=34534615</a>) and it's had a huge update and complete redesign. We rebuilt the whole app, and added a real-time 3D soft body solver which gives some really cool interactions like blobs / objects that inflate as you breathe. We also made a version for Vision Pro, called 'Lungy Spaces'.<p>My background is as a surgical trainee and I started building Lungy in 2020 during the first COVID lockdown in London. During COVID, there were huge numbers of patients coming off ventilators and patients are often given breathing exercises on a worksheet and disposable plastic devices called incentive spirometers to encourage deep breathing. This is intended to prevent chest infections and strengthen breathing muscles that have weakened. I noticed often the incentive spirometer would sit by the bedside, whilst the patient would be on their phone – this was the spark that lead to Lungy!<p>Since making the first version we’ve made exercises fully customisable (you can dial in exact timings for each breath phase), added new breathing indicators, learning modules, e.g. self-care for anxiety symptoms, and lots of new visuals. The free version gives you access to a new breathing exercise each day, whilst premium unlocks the full library of exercises, exercise data and visuals..<p>The visuals are mostly built using Metal (a couple use SpriteKit) and there are lots to choose from - boids, cloth sims, fluid sims, a hacky DLA implementation, rigid body + soft body sims - each one reacts to breath and touch. The audio uses AudioKit with a polyphonic synth and a sequencer plays generated notes from a chosen scale (you can mess around with the sequencer and synth in Settings/Create Music). The nice thing about the visuals + audio being generative is that the download size is relatively small with no other downloads. We’re still working on improving the breath detection, using ML - currently, it uses microphone input, with optional camera input to guide positioning.<p>We’re also close to finishing the medical device version - <a href="http://lungy.health" rel="nofollow">http://lungy.health</a> - designed as a pulmonary rehab platform for patients with asthma, it should hopefully undergo early trials in the UK in 2026.<p>Thanks for reading - would love to hear any feedback!<p><a href="https://www.lungy.app" rel="nofollow">https://www.lungy.app</a><p>Lungy Version 2 here: <a href="https://apps.apple.com/app/apple-store/id1545223887">https://apps.apple.com/app/apple-store/id1545223887</a>

Show HN: Luna Rail – Treating night trains as a spatial optimization problem

Hi HN, I'm Anton, founder of Luna Rail.<p>I've always thought night trains are a fantastic, sustainable alternative to short-haul flights, but they're often held back by a lack of privacy, comfort, and poor economics due to low passenger capacity.<p>I became overly fascinated with this puzzle. I view it as a kind of night train Tetris (my wife less charitably calls it "sardinology"). I spent way too much time learning about and sketching various layouts, trying to figure out how to fit the maximum number of private cabins into a standard railcar, while making them attractive for both day and night travel.<p>This eventually led to a physical workshop (in Berlin) and a hands-on rapid prototyping process. We've built a series of full-scale mockups, starting with wood and cardboard and progressing to high-fidelity versions with 3D-printed and CNC-milled parts, with various functional elements.<p>Hundreds of people have come in to test our various iterations, because you can't test ergonomics or comfort by looking at renderings (although we did create a bunch of nice ones).<p>The link goes to our home page showing our approach and some of the thinking behind them. It’s been a lot of fun working on this puzzle, and we're excited to share what we've come up with. We hope you think it's cool too and would love to hear your thoughts.

Show HN: Luna Rail – Treating night trains as a spatial optimization problem

Hi HN, I'm Anton, founder of Luna Rail.<p>I've always thought night trains are a fantastic, sustainable alternative to short-haul flights, but they're often held back by a lack of privacy, comfort, and poor economics due to low passenger capacity.<p>I became overly fascinated with this puzzle. I view it as a kind of night train Tetris (my wife less charitably calls it "sardinology"). I spent way too much time learning about and sketching various layouts, trying to figure out how to fit the maximum number of private cabins into a standard railcar, while making them attractive for both day and night travel.<p>This eventually led to a physical workshop (in Berlin) and a hands-on rapid prototyping process. We've built a series of full-scale mockups, starting with wood and cardboard and progressing to high-fidelity versions with 3D-printed and CNC-milled parts, with various functional elements.<p>Hundreds of people have come in to test our various iterations, because you can't test ergonomics or comfort by looking at renderings (although we did create a bunch of nice ones).<p>The link goes to our home page showing our approach and some of the thinking behind them. It’s been a lot of fun working on this puzzle, and we're excited to share what we've come up with. We hope you think it's cool too and would love to hear your thoughts.