shahahmed
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @04:05PM (#49012635)
thanks for trying it - yeah i might update the distance logic to be forward facing only, good point
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @04:01PM (#49012584)
it's possible, takes a bit of finesse. i might add an auto-swing to make it easier. you can also jump to shift momentum aggressively
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @04:00PM (#49012565)
thank you for the rec, will try, and thanks for playing!
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @04:00PM (#49012558)
thanks! yes i was a big fan of the song
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @02:21PM (#49011149)
it's actually web slinging spider person
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @02:20PM (#49011131)
the blocks and building scales are accurate - based on OpenNYC, but the building styles / shapes are generic
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @02:19PM (#49011122)
uh oh, is it reversed!! i will fix
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @01:32PM (#49010374)
how so?
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @01:06PM (#49009943)
did you find your building? this was based on OpenNYC
Re:Show HN: Web swing through midtown NYC
on Wednesday July 22, 2026 @01:04PM (#49009921)
really helped to add the signifier of which building would be snapped. glad you enjoyed it!
Re:Tell HN: GitHub Having Issues
on Tuesday March 03, 2026 @02:01PM (#47237073)
Updog tracks this via Datadog logs - https://updog.ai/status/github
GitHub Is Degraded
Re:Ask HN: Could you share your personal website here?
on Wednesday January 14, 2026 @02:33PM (#46621548)
Re:It's OpenAI's world, we're just living in it
on Friday October 10, 2025 @02:06PM (#45541937)
yeah something tells me Ben does not care about being #1 on hn
Show HN: Local Email Client for AI Horseless Carriages
https://github.com/dbish/DispatchMail
The AI Horseless Carriages article spurred a lot of conversation about how we should just be giving users the system prompt box [0], and we were pretty surprised that a bunch of email clients didn’t pop up following this pattern [1].So we went ahead and created a local [2] email client that you can run that processes your inbox with your own handwritten rules. It lets you label and archive...
Show HN: Toy LLM Daydreaming
https://github.com/shahgahmed/toy-llm-daydream
After reading gwern’s post two days ago about LLM Daydreaming [0], I wanted to just see how it would behave. I grabbed the top vital article titles from wikipedia [1] and made a simple next js app to try the prompt in a loop.It was pretty interesting to see the connections made between concepts. Happy to host this if people find it useful![0] https://gwern.net/ai-daydreaming [1]...
Re:Launch HN: Leaping (YC W25) – Self-Improving Voice AI
on Thursday July 10, 2025 @02:03PM (#44523786)
congrats on launching! how are ya'll managing evals?
Re:Ask HN: Is Datadog down for anyone else?
on Wednesday July 09, 2025 @05:19PM (#44514840)
there seems to be degraded perf: https://status.datadoghq.com/
Re:Show HN: Trello Boards for Agents
on Tuesday July 08, 2025 @01:30PM (#44502169)
thanks for checking it out. i imagine 2 things, 1 - any edit actions require a "human in the loop" approval. 2 - for sensitive info, internal instances of llms or only local models accessing info
Show HN: Trello Boards for Agents
Most agent tools (Devin, n8n, Codex, etc.) create new user experiences and dashboards for managing work. It feels like we should work alongside agents and use the tools we already do.To test this experience, we built Dispatch Board. It lets you work alongside agents and humans using a kanban board to create tasks and respond to any questions that come up.Would love feedback on workflows you’d...
Re:HDR‑Infused Emoji
on Thursday April 17, 2025 @12:30PM (#43719131)
these look so tasteful and well done
Show HN: notperplexity – Emulating an AI-Search Engine over the Weekend
I challenged myself to recreate the core experience of Perplexity.ai, the $9B “Google-killer", on a Sunday, using OpenAI's latest web search capabilities and a light frontend.The result is: notperplexity.aiUnder the hood, it wraps OpenAI’s new Responses API with the web-search tool. It’s sometimes useful and mostly an excuse to see how far the primitives have come. It does beg the question of...
Re:Show HN: Dash0 – Dev-Friendly OpenTelemetry Observability with Open...
on Tuesday December 10, 2024 @02:38PM (#42380568)
gotcha, thanks for answering! i guess also disclaimer: I work at Datadog.
Re:Show HN: Beeps – the on-call platform for NextJS developers
on Tuesday December 10, 2024 @01:16PM (#42379553)
congrats on the launch! i build primarily with nextjs and vercel and will give it a shot. but we also just launched an on-call tool (i work at datadog), so i'll compare
Re:Show HN: Dash0 – Dev-Friendly OpenTelemetry Observability with Open...
on Tuesday December 10, 2024 @01:14PM (#42379518)
i like the philosophy! do ya'll provide a specific collector/agent or do users send directly from the community otel collector?
Re:David Chang on the long, hard, stupid way
on Wednesday September 11, 2024 @12:36PM (#41512975)
"The dish was meant to be a difficult pickup that required constant coordination between the front and back of house. That was what made it great."
This feels like the crux. It's not arbitrarily doing the hard, stupid thing, but the ones that are based in making the product or process better.
Re:Show HN: I am building an open-source incident management platform
on Monday July 08, 2024 @02:26PM (#40907927)
this is detailed in the Google SRE book: https://sre.google/workbook/incident-response/
Re:Show HN: I am building an open-source incident management platform
on Monday July 08, 2024 @01:15PM (#40907250)
curious, is this because you expect something other than Slack supported?
Re:Enhancing Code Completion for Rust in Cody
on Tuesday June 18, 2024 @04:13PM (#40721762)
arguably you can reduce even more latency by keeping the model on-device as well, but that would mean revealing the weights of the fine-tuned model.
If the user preferred reduced latency and had the RAM, is that an option?