The graveyard of good ideas
Solutions that were tried, and what stopped them.
Patents that were rejected or abandoned, products that shut down, projects that went quiet. Borrow the idea, avoid the dead end.
496 solutions
- startupShut down2026
Show HN: Demon – open-source real-time music diffusion engine, 25Hz local GPU
YO, I’m Ryan, lead author. I’ve been contributing open source generative audio stuff for a while now, audio reactive Comfy nodes, extended ACEstep support in Comfy, etc.. I just opened-sourced a new audio project that I've been working on for a few months and I want to tell y'all about it. WHAT IS IS DEMON: Diffusion Engine for Musical Orchestrated Noise This is StreamDiffusion but with audio instead of images, and ACEStep 1.5 instead of Stable Diffusion. It’s responsive enough that you can play it like an instrument, and remix in near real-time. I also distilled the ACEStep VAE: it’s faster at the expense of some quality. I also trained something like 200 lora/dora for ACEStep 1.5 and 1.5XL: I will release these in batches of 5 or 10 or something WHY IT IS Two reasons: 1) Making music is an inherently real-time activity 2) Why not bro SOME RUNTIME CAPABILITIES -Real-time
Why it failed: YO, I’m Ryan, lead author. I’ve been contributing open source generative audio stuff for a while now, audio reactive Comfy nodes, extended ACEstep support in Comfy, etc.. I just opened-sourced a new audio project that I've been working on for a few months and I want to tell
ryanontheinside
- startupShut down2025
Show HN: Nstart, a Nostr's Onbarding Tool
Nstart aims to guide new users to Nostr offering a easy and no-nonsense onboarding wizard, with useful hints about the protocol and some really exclusive features: - Easy local backup of your nsec or ncryptsec - Email yourself your ncryptsec, as an additional backup location - Create a multi-signer bunker URL for Nostr Connect (more info below) - Auto follow the contacts list of some old and trusted Nostr users - Customize of contact suggestions, useful for onboarding friends & family Try Nstart live at https://start.njump.me or watch this video to understand how it works: https://chronicle.dtonon.com/77b9a397cbbcd63f3bfc78ff2b2c060... For devs: it can also be used by any Nostr application, web or mobile, to offer an easy onboarding flow! You can integrate it effortlessly via a simple redirect, modal or popup: in the end, the user is automagically logged into
Why it failed: Nstart aims to guide new users to Nostr offering a easy and no-nonsense onboarding wizard, with useful hints about the protocol and some really exclusive features: - Easy local backup of your nsec or ncryptsec - Email yourself your ncryptsec, as an additional backup location - Cr
dtonon
- startupShut down2025
I am concerned about the future. Are you?
My children are in college pursuing STEM degrees (CS and Robotics), with one also pursuing an art double major (animation). While we are in the US, we also have family in Europe and Latin America. So, what's the concern? Well, there are two that I think should be obvious. The first is how AI will alter the landscape. It is easy to see that companies might be able to develop products with somewhere between 1/5 to 1/10 the engineers in the not-too-distant future. Let's define that as the time required for a fresh high school graduate to finish a Masters degree. How can you possibly advise and guide a young person given the huge variability of potential outcomes 6 to 10 years out? The second is the massive deindustrialization of the West (again, this isn't US-centric as Europe, Latin America and others are deeply affected) as China solidifies into a dominant positio
Why it failed: My children are in college pursuing STEM degrees (CS and Robotics), with one also pursuing an art double major (animation). While we are in the US, we also have family in Europe and Latin America. So, what's the concern? Well, there are two that I think should be obvious. Th
robomartin
- startupShut down2025
Show HN: A library to convert+deploy existing agent projects as MCP servers
Most of the MCP servers that I’ve seen are tools implemented in standalone projects. To onboard more tools (especially agents and multi-agent workflows) to MCP, I’ve been thinking it’s important to allow AI engineers to continue to prototype in their existing agent frameworks and deploy with minimal conversion when ready. We created the automcp library, which you can add as a dependency to existing projects (CrewAI, LangGraph, Llama Index, OpenAI Agents SDK, Pydantic AI, mcp-agent currently supported but more coming soon). You just need to run a CLI command to create a run_mcp.py file, make some edits and run it to start the server locally. You can think of run_mcp.py like Heroku’s Procfile, Codespaces configs, Pulumi/AWS CDK style IaC. We also created a demo of a deployment platform where you can enter the GitHub URL of your project, deploy with one click, and get a URL for the hos
Why it failed: Most of the MCP servers that I’ve seen are tools implemented in standalone projects. To onboard more tools (especially agents and multi-agent workflows) to MCP, I’ve been thinking it’s important to allow AI engineers to continue to prototype in their existing agent frameworks and
richardblythman
- startupShut down2025
Show HN: We Built Our Own Chart Type for Live Dashboards
Most analytics tools give you 10–15 chart types and call it a day. But what happens when none of them actually represent your data well? We ran into this when building a CRM dashboard. We wanted to show how sales reps connect to open deals: not just totals, but actual relationships. A bar chart was too flat, a table too dense. We needed a network graph. The problem? Most chart libraries and BI tools don’t support that out of the box. You either say “no” to the feature request, or you hack together a standalone visualization that breaks the product’s flow (no filters, no interactivity, no theming). So we built it ourselves, but in a way that still plays nicely with the rest of the dashboard. Luzmo now lets you define your own chart type, write your own visualization code, and then drop it into the dashboard editor like any other chart. Luzmo still handles the boring stuff: querying, filte
Why it failed: Most analytics tools give you 10–15 chart types and call it a day. But what happens when none of them actually represent your data well? We ran into this when building a CRM dashboard. We wanted to show how sales reps connect to open deals: not just totals, but actual relationshi
YannickCrabbe
- startupShut down2025
Ask HN: Abandoned/dead projects you think died before their time and why?
Just curious and who knows, maybe someone will adopt it or develop something new based on its ideas.
Why it failed: Just curious and who knows, maybe someone will adopt it or develop something new based on its ideas.
ofalkaed
- startupShut down2025
Ask HN: My family business runs on a 1993-era text-based-UI (TUI). Anybody else?
Is anybody still using TUI applications for business? My family company is a wholesale distribution firm (with lightweight manufacturing) and has been using the same TUI application (on prem unix box) since 1993. We use it for customer management, ordering, invoicing, kit management/build tickets, financials - everything. We've transitioned from green screen terminals to modern emulators, but the core system remains. I spent many summers running serial and ethernet cables. I left the business years ago to become a full time software engineer, but I got my start as a script kiddie writing automations for this system with Microsoft Access, VBA, and SendKeys to automate data entry. Amazingly, they still have a Windows XP machine running many of those tasks I wrote back in 2004! It's brittle, but cumulatively has probably saved years of time. That XP machine could survive a nu
Why it failed: Is anybody still using TUI applications for business? My family company is a wholesale distribution firm (with lightweight manufacturing) and has been using the same TUI application (on prem unix box) since 1993. We use it for customer management, ordering, invoicing, kit managem
urnicus
- startupShut down2025
We got a $60k tax penalty for a zero-revenue side project
In 2020, in the middle of COVID, a few friends and I decided that obviously the world needed another presentation tool. Our pain point was simple: we hated fighting with PowerPoint/Google Slides. We just wanted to write content and get nice slides without babysitting fonts, sizes, alignment, colors, and 19 different "Title + Subtitle" layouts that all somehow looked bad. So we built a prototype: a tiny click-together slide builder. Then we decided clicking was too slow and turned it into a full markdown-to-slides thing. Write markdown in, pretty slides out. Easy. Then, like every good side project, we started bolting on features: Image uploads, PDF generation via API, AI integration and more and more stuff nobody had actually asked for.... Somewhere in there, in peak "we're definitely the next unicorn" mode, we incorporated a US company through Stripe Atlas.
Why it failed: In 2020, in the middle of COVID, a few friends and I decided that obviously the world needed another presentation tool. Our pain point was simple: we hated fighting with PowerPoint/Google Slides. We just wanted to write content and get nice slides without babysitting fonts,
lukaslukas
- startupShut down2025
Show HN: A word of the day that doesn't suck
I’ve long thought that the Word of the Day was a wasted genre. The goal should be to give you words you can use; to enrich your understanding of words you already know; or at least to use words to tell you something neat about the world. Instead, what you usually get is words that will never be used in conversation, held up as curios. Some examples from Dictionary.com’s daily email: thewless, balladmonger, vagility, contextomy. These words are... not useful. I’ve always thought I could do better. My friend Ben recently created a daily puzzle game, called Bracket City, launched here on HN [1], which I like because it takes about the same amount of time as Wordle but has some of the variety and artistry of a good crossword. Ben agreed to let me write a word of the day for the game’s audience. We’ve collected them all here: https://bracket.city/words . It’s such a joy to writ
Why it failed: I’ve long thought that the Word of the Day was a wasted genre. The goal should be to give you words you can use; to enrich your understanding of words you already know; or at least to use words to tell you something neat about the world. Instead, what you usually get is words tha
jsomers
- startupShut down2025
Ask HN: I don't think I want to work in tech anymore?
I’ve been working in tech as a SWE for a good while now, and I’m realising it’s making me miserable. On paper I feel like I should have a very successful career: high-paid remote work, senior/staff level, reasonably-generalist with backend specialization, generally good coworkers / managers, and so on. But in reality, I’m doing terribly! I have relatively short tenure in all my roles, because I somehow keep picking jobs where my manager/coworkers start quitting before I do; my savings were drained by health issues caused by work stresses, so I can’t afford to take an extended period between jobs; employers I end up with, regardless of contract or FTE, always seem to end up making near-impossible demands of engineering that ruin the team’s morale; the list goes on. I see a therapist, I see a number of doctors for the health issues I’ve ended up with, but I feel like no amou
Why it failed: I’ve been working in tech as a SWE for a good while now, and I’m realising it’s making me miserable. On paper I feel like I should have a very successful career: high-paid remote work, senior/staff level, reasonably-generalist with backend specialization, generally good cowo
wanttoquittech
- startupShut down2025
Show HN: Open-source Deep Research across workplace applications
I’ve been using deep research on OpenAI and Perplexity and it’s been just amazing at gathering data across a lot of related and chained searches. Just earlier today, I asked “What are some marquee tech companies / hot startups (not including the giants like FAAMG, Samsung, Nvidia etc.)”. It’s a pretty involved question and looking up “marquee tech startups” or "hot tech startups" on Google gave me nothing useful. Deep research on both ChatGPT and Perplexity gave really high quality responses with ChatGPT siding on slightly larger scaleups and Perplexity siding more on up and coming companies. Given how useful AI research agents are across the internet, we decided to build an open-source equivalent for the workplace since a ton of questions at work also cannot be easily resolved with a single search. Onyx supports deep research connected to company applications like Google
Why it failed: I’ve been using deep research on OpenAI and Perplexity and it’s been just amazing at gathering data across a lot of related and chained searches. Just earlier today, I asked “What are some marquee tech companies / hot startups (not including the giants like FAAMG, Samsung, N
yuhongsun
- startupShut down2025
FWS – pip-installable embedded process supervisor with PTY/pipe/dtach back ends
I’m releasing *Framework Shells* (`fws`): a standalone Python package for orchestrating long-running background processes (“shells”) with *PTY*, *pipes*, and *dtach* backends. This is meant for environments where you don’t want to stand up a full supervisor stack (or don’t have one): quick multi-service prototypes, dev environments, constrained userlands, etc. ### What it does * Spawn/manage shells with: * **PTY**: interactive terminal sessions (resize, input, stream) * **Pipes**: stdin/stdout/stderr streams (good for daemons/LSPs) * **dtach**: persistent sessions you can attach/detach to (survives manager restarts) * *Runtime isolation* (the big feature): shells are namespaced by `~/.cache/framework_shells/runtimes/<repo_fingerprint>/<runtime_id>/...` so two clones of the same repo can run concurrently without cross-adop
Why it failed: I’m releasing *Framework Shells* (`fws`): a standalone Python package for orchestrating long-running background processes (“shells”) with *PTY*, *pipes*, and *dtach* backends. This is meant for environments where you don’t want to stand up a full supervisor stack (or don’t have o
mrsurge
- startupShut down2025
Ask HN: Want to leave my job with nothing lined up
I’m currently working in a very toxic environment and it’s starting to affect my life outside of work in negative ways. I have money saved up, don’t live an extravagant lifestyle, and have around 7 years of experience but I do know how challenging the job market has been which worries me. I would also be leaving my current company towards the end of the year when not many people are hiring. When asking my friends and colleagues what to do, I always get the same answer of either wait until you find something else to quit or wait until they fire you, but whatever you do don’t leave with no offer in hand. I agree with their approach in general but I’m just really burned out and don’t have the mental energy to come home from a 10 hour day and start preparing for interviews. Ideally I would like time to prepare adequately for interviews and really think about what I want to do next instead of
Why it failed: I’m currently working in a very toxic environment and it’s starting to affect my life outside of work in negative ways. I have money saved up, don’t live an extravagant lifestyle, and have around 7 years of experience but I do know how challenging the job market has been which wo
bsjak
- startupShut down2025
Ask HN: As a developer, am I wrong to think monitoring alerts are mostly noise?
I'm a solo developer working on a new tool, and I need a reality check from the ops and infrastructure experts here. My background is in software development, not SRE. From my perspective, the monitoring alerts that bubble up from our infrastructure have always felt like a massive distraction. I'll get a page for "High CPU" on a service, spend an hour digging through logs and dashboards, only to find out it was just a temporary traffic spike and not a real issue. It feels like a huge waste of developer time. My hypothesis is that the tools we use are too focused on static thresholds (e.g., "CPU > 80%") and lack the context to tell us what's actually an anomaly. I've been exploring a different approach based on peer-group comparisons (e.g., is api-server-5 behaving differently from its peers api-server-1 through 4?). But I'm coming at this f
Why it failed: I'm a solo developer working on a new tool, and I need a reality check from the ops and infrastructure experts here. My background is in software development, not SRE. From my perspective, the monitoring alerts that bubble up from our infrastructure have always felt like a m
yansoki
- startupShut down2025
DrawAFish.com Postmortem
Related: Show HN: Draw a fish and watch it swim with the others - https://news.ycombinator.com/item?id=44719222 - July 2025 (233 comments)
Why it failed: Related: Show HN: Draw a fish and watch it swim with the others - https://news.ycombinator.com/item?id=44719222 - July 2025 (233 comments)
hallak
- startupShut down2025
Notion's Lies Sunsetting Skiff Mail
With the announcement of "Notion Mail", here's a reminder that after Notion acquired Skiff Mail in early 2024, their entire product line was sunset and, despite promising[1][2] email forwarding and support would remain functional until 2025 for trapped users that couldn't immediately access their accounts, both turned out to be complete lies. Their support through support@skiff.org was shut off[3] not long after the fake promise, responding only with automated acknowledgements to even refund requests. Forwarding was also effectively shut off around the same time thanks to Notion's willful negligence maintaining Skiff's certificates[4], pulling the rug out from under users needing mails forwarded from inaccessible Skiff email accounts. To add insult to injury, Notion's support through team@makenotion.com for months responded with incoherent, AI-generated
Why it failed: With the announcement of "Notion Mail", here's a reminder that after Notion acquired Skiff Mail in early 2024, their entire product line was sunset and, despite promising[1][2] email forwarding and support would remain functional until 2025 for trapped users that c
notioned
- startupShut down2025
Show HN: Woomarks, transfer your Pocket links to this app or self-host it
Pocket is shutting down and I really, really liked it. So I built woomarks, an app that let's you save links with a similar UI. It's very minimal, but it's doing everything I liked from Pocket and you can bulk import your links and use the app or self-host. - Public app that you can test: https://woomarks.com/ - My self-hosted version, where you can see my saves: https://roberto.fyi/bookmarks/ - Repository if you want to self-host: https://github.com/earlyriser/woomarks Export links from Pocket here: https://getpocket.com/export the last day will be on October 20025. Features: - Add/Delete links - Search - Tags - Bookmarklet (useful for a 2-click-save) - Data reads from: csv file in server (these links are public) local storage in browser (these links are visible just for the user) - Local storage s
Why it failed: Pocket is shutting down and I really, really liked it. So I built woomarks, an app that let's you save links with a similar UI. It's very minimal, but it's doing everything I liked from Pocket and you can bulk import your links and use the app or self-host. - Publi
earlyriser
- startupShut down2025
Ask HN: Scheduling stateful nodes when MMAP makes memory accounting a lie
We’re hitting a classic distributed systems wall and I’m looking for war stories or "least worst" practices. The Context: We maintain a distributed stateful engine (think search/analytics). The architecture is standard: a Control Plane (Coordinator) assigns data segments to Worker Nodes. The workload involves heavy use of mmap and lazy loading for large datasets. The Incident: We had a cascading failure where the Coordinator got stuck in a loop, DDOS-ing a specific node. The Signal: Coordinator sees Node A has significantly fewer rows (logical count) than the cluster average. It flags Node A as "underutilized." The Action: Coordinator attempts to rebalance/load new segments onto Node A. The Reality: Node A is actually sitting at 197GB RAM usage (near OOM). The data on it happens to be extremely wide (fat rows, huge blobs), so its logical row count is low, bu
Why it failed: We’re hitting a classic distributed systems wall and I’m looking for war stories or "least worst" practices. The Context: We maintain a distributed stateful engine (think search/analytics). The architecture is standard: a Control Plane (Coordinator) assigns data se
leo_e
- startupShut down2025
Show HN: Klarity – Open-source tool to analyze uncertainty/entropy in LLM output
We've open-sourced Klarity - a tool for analyzing uncertainty and decision-making in LLM token generation. It provides structured insights into how models choose tokens and where they show uncertainty. What Klarity does: - Real-time analysis of model uncertainty during generation - Dual analysis combining log probabilities and semantic understanding - Structured JSON output with actionable insights - Fully self-hostable with customizable analysis models The tool works by analyzing each step of text generation and returns a structured JSON: - uncertainty_points: array of {step, entropy, options[], type} - high_confidence: array of {step, probability, token, context} - risk_areas: array of {type, steps[], motivation} - suggestions: array of {issue, improvement} Currently supports hugging face transformers (more frameworks coming), we tested extensively with Qwen2.5 (0.5B-7B) models, b
Why it failed: We've open-sourced Klarity - a tool for analyzing uncertainty and decision-making in LLM token generation. It provides structured insights into how models choose tokens and where they show uncertainty. What Klarity does: - Real-time analysis of model uncertainty during gener
mrciffa
- startupShut down2025
Show HN: Molab, a cloud-hosted Marimo notebook workspace
We launched marimo [1], an open-source reactive Python notebook, last year on HackerNews. Today, the most popular recent feature request in Google Colab’s issue tracker asks for marimo support in Colab [2]. Rather than try to convince Google to replace their notebook with marimo, we decided to just build our own cloud-hosted notebook service instead. We're calling this molab (mo for marimo), and we're launching it today. You can try it at https://molab.marimo.io Some features: - Persistent storage - Link sharing (notebooks are public but undiscoverable, like secret GitHub Gists) - Download notebooks to your machine, reuse them as Python scripts or apps - Upload local notebooks to the cloud from our CLI (coming soon) - Real-time collaboration (coming soon) - Configure computational resources to obtain more CPU or GPU (coming soon) marimo is a modern notebook for modern
Why it failed: We launched marimo [1], an open-source reactive Python notebook, last year on HackerNews. Today, the most popular recent feature request in Google Colab’s issue tracker asks for marimo support in Colab [2]. Rather than try to convince Google to replace their notebook with marimo,
akshayka
- startupShut down2025
Evaluating GPT5's reasoning ability using the Only Connect game show
We evaluated OpenAI GPT5 lateral reasoning abilities against other models using an approach based on the notoriously difficult and highly-challenging british game show Only Connect, which challenges contestants' pattern-matching and trivia skills. Insights: - GPT-5 does extremely well, but only marginally better than o3. - Model verbosity has little impact on accuracy and cleverness, except, interestingly, for the sequences round - "minimal" verbosity however causes accuracy to drop sharply. We'll be publishing additional results in the coming days from our extended tests. We're looking at different types of evals (how do the models fare with a single item in a sequence vs. 2, 3, 4). We would also like to look at how the models behave in a team of 3, replicating the format of the game show. We were unable to find evidence that the Only Connect games are in the tr
Why it failed: We evaluated OpenAI GPT5 lateral reasoning abilities against other models using an approach based on the notoriously difficult and highly-challenging british game show Only Connect, which challenges contestants' pattern-matching and trivia skills. Insights: - GPT-5 does extr
scrollaway
- startupShut down2025
Show HN: FlyCode – Recover Stripe payments by automatically using backup cards
We built FlyCode after seeing subscription businesses lose ~35% of recurring revenue each year to failed payments — even when customers had other valid cards on file. *The problem:* When a customer's primary card fails, Stripe retries a few times then cancels the subscription. If that customer has a backup card, it isn’t tried. At least 20% of active customers have more than one card on file, which means a lot of preventable churn. *Our solution:* FlyCode automatically identifies if a customer has other valid cards on file and retries them when a subscription payment fails. You can configure when these retries happen during the dunning period (beginning, middle, end) and define validity rules (e.g. “card was used in last 180 days”). It’s a Stripe app — no code changes needed. We've seen 18%-20% higher recovery rates from our core retry engine, plus another 5–10% from using back
Why it failed: We built FlyCode after seeing subscription businesses lose ~35% of recurring revenue each year to failed payments — even when customers had other valid cards on file. *The problem:* When a customer's primary card fails, Stripe retries a few times then cancels the subscriptio
JakeVacovec
- startupShut down2025
Mozilla Firefox Is Dying
We are undoubtedly witnessing a rapid decline of Mozilla and its products: the question we should be asking ourselves is: how long before the total collapse? And furthermore: are those Linux users (like me) who are still relying on Firefox for doing their development job ready to abandon it? Let me list at least three facts to support my thesis: 1) Over the past year, there have been various controversies announcing a radical change in the core values of the Mozilla Foundation 2) After its rewrite (Project Fenix), the Firefox Mobile browser has never regained the completeness it had before, despite the passage of years 3) One of the web extensions developed by Mozilla (and marked as “official”) was recently removed from the store because it violated Mozilla's own add-on policies: this suggests serious internal disputes. I wonder what is your opinion [1] https://www.osnews.
Why it failed: We are undoubtedly witnessing a rapid decline of Mozilla and its products: the question we should be asking ourselves is: how long before the total collapse? And furthermore: are those Linux users (like me) who are still relying on Firefox for doing their development job ready to
leonixyz
- startupShut down2025
The Death of Utilitarian Programming
Utilitarian coding is defined as follows: The code you write should be *directly* useful or serve the interest of at least one actual human being. It might appear somewhat abstract or vague, so examples might help. For example, I don't consider frameworks as utilitarian code. What you create are like the "frames" of a picture box, someone else (the user) will take it and draw the actual picture. Though you did help with part of the process, it's indirect at best. You're part of the supply chain here, not part of the team. A clever and witty bash script running on a unix server somewhere is also not utilitarian coding, no human ever directly benefited from it. Libraries can be somewhat utilitarian, at least more than frameworks. At least they provide some reusable functionality to the user out of the box like logging, scanning a barcode, fetching data from a URL,
Why it failed: Utilitarian coding is defined as follows: The code you write should be *directly* useful or serve the interest of at least one actual human being. It might appear somewhat abstract or vague, so examples might help. For example, I don't consider frameworks as utilitarian code
pyeri