Fastlane paths usually break on folder naming/order drift; zero-pad 01..05 to keep uploads stable. Are you targeting iOS deliver, Android supply, or both?
Upvest GmbH | Software Engineers (mid and senior) | SRE Lead | Berlin, Tallinn, London, or REMOTE in Germany, Europe | Full time
At Upvest, we are on a mission to make investing as easy as spending money. Upvest empowers businesses to offer a wide range of investment products and the best experience in the field of capital market investment and retirement planning. Upvest’s Investment API is easy to integrate so that fintechs and financial institutions can save resources and fully focus on their core business.
Per-iteration times:
ROLV : 0.000476 s
Dense PyTorch (MKL): 0.023720 s → Speedup vs Dense: 49.85× Energy saved: 98.0%
CSR Sparse (MKL) : 0.053517 s → Speedup vs Best Sparse: 112.48× Energy saved: 99.1%
Hey HN, I built GridSnap because I was tired of digging through notes and docs to copy-paste things I use repeatedly — server addresses, SQL queries, config values, template messages.
It's a spreadsheet-like desktop app that lives in your system tray. Ctrl+Shift+Space pulls it up, you click a cell, it's copied. That's the whole UX.
Technical decisions:
- Tauri v2 over Electron: ~3MB bundle vs ~85MB, ~30MB RAM vs 200MB+
- Custom virtualized grid: 1000×26 cells, only visible cells rendered
- Rust backend: AES-256-GCM + Argon2id for encryption, all crypto/IO in Rust
- Sparse storage: empty cells don't allocate memory
- CSS custom properties theming, zero inline styles
Still early — the dev plan covers 7 phases and I'm working through them. Feedback on architecture decisions welcome.
Thank you for trying the app and for your feedback. In the upcoming version, I will make auto-save an option that can be disabled in the settings. The layout shift issue will also continue to be improved to provide a better user experience.
I'm not making decisions for anybody. You can stay at home and watch your garden grow. Fine with me. I described what type of people I like. And that those are rather the pyramid climbers.
Doconomy, a market leader in impact technology, dedicated to driving global climate action by equipping banks with innovative financial tools that educate and foster positive change.
Our ecosystem of tools combines expertise in behavioral science to design financial services that make it simple and fun to save and invest money, while reducing environmental impact. By integrating these solutions into existing banking apps, we help people achieve their financial goals and contribute to a sustainable future.
You had commented before I created a new account, so I will just add that I have never seen Wirth claim that he or his colleagues at ETH were the originators of the idea. Moreover, this idea was already used in Wirth’s “first” language, namely Euler, which he published in 1965 while at Stanford. An even earlier example may be Schorre’s META‑II from 1964 (UCLA).
Regarding the popular opinion here that Pascal was created for educational purposes (only) - implicitly meaning not for building real, large-scale software - Wirth considered these two characteristics equivalent, and in his view one was a necessary condition for the other.
From the summary/abstract of his original publication about Pascal (The Programming Language Pascal) written in 1970 and published in Acta Informatica in 1971:
In view of its intended usage both as a convenient basis to teach programming and as an efficient tool to write large programs, emphasis was placed on keeping the number of fundamental concepts reasonably small, on a simple and systematic language structure, and on efficient implementability.
I hope he comes to India and starts a movement, ideally leading to a public holiday tradition where everyone is supposed to pick up 10 pieces of litter.
I run 3-5 Cursor cloud agents at a time. Before this I was alt-tabbing between browser tabs to check if they finished, crashed, or opened a PR.
Built a retro TUI that connects to the Cursor Agents API - shows all agents on one screen, color-coded by status. You can launch new agents, send follow-up prompts, and stop runaway ones without leaving the terminal.
npx agents-control-tower to try it. MIT, TypeScript, zero config.
Ferriss misplaces the cause. Self-help doesn't train you to find ways you're broken — it selects for people who already carry a felt deficiency. Who sleeps well doesn't Google sleep optimisation.
The more interesting question is what that deficiency actually is. I think attachment theory gives a more precise answer: the underlying sense of insufficiency is mostly relational in origin. So his pivot to relationships has real intuition behind it — but it still mistakes the symptom for the cure.
The actual trap isn't self-help as a genre. It's using any action — including optimising your relationships — to externalise rather than confront what's underneath. The distinguishing feature is direction: are you doing this to avoid discomfort or to change your relationship to it? Rumination and productivity hacks fail by the same measure for the same reason.
His buried insight — "you cannot improve suffering away" — is the most important line in the piece, treated as a footnote. That's where the real work starts: developing the capacity to sit with what you've been avoiding rather than finding a better-feeling target for the same restlessness.
It is a bit ironic: the article is monetised self-help advice warning you about self-help, while introducing fresh deficiencies along the way — everything you learned was wrong — and staying carefully at the level of framework. That's precisely the move he's critiquing: redecorating the avoidance rather than confronting it. The most useful version of this piece would be considerably less optimised and considerably more vulnerable.
okay this was very interesting "To continually improve yourself, you must continually locate the ways you are broken" haven't thought of it this way. Maybe I need to look into why i keep finding new books.
What’s interesting about full-duplex speech systems isn’t just the model itself, but the pipeline latency.
Even if each component is fast individually, the chain of audio capture → feature extraction → inference → decoding → synthesis can quickly add noticeable delay.
Getting that entire loop under ~200–300ms is usually what makes the interaction start to feel conversational instead of “assistant-like”.
“Discover, install, and manage a curated collection of 1,000+ open‑source AI Skills and MCP Servers. Fast, free, community‑driven – instantly supercharge any AI‑enabled editor.”
Why It Stands Out?
100 % Open‑Source & Free – No hidden fees, no usage limits.
One‑Stop Marketplace – Browse, preview, and install AI Skills with a single click.
Universal Compatibility – Works seamlessly with Claude, Cursor, Windsurf, and any future AI‑augmented IDE.
Community‑Curated Quality – Every skill is reviewed, version‑controlled, and documented by contributors worldwide.
Rapid Integration – JSON‑based manifest makes onboarding a matter of seconds.
Scalable Architecture – Powered by Astro + Cloudflare Workers/D1, ensuring lightning‑fast search and zero‑downtime updates.
Core Features
Skill Explorer: Interactive UI with rich previews, tags, and live demos.
One‑Click Install: Auto‑generate the required MCP Server config.
Version History & Rollback: Safe upgrades with full changelog.
Developer Dashboard: Submit, test, and publish your own skills instantly.
Call‑to‑Action:
Join thousands of developers who are already extending their AI tools for free. Add Killer‑Skills to your workflow today and watch productivity soar!
This is actually a common statement people make with whom I feel bored. I call it the "evasive defense".
Me: "Let's fly to Paris tomorrow!"
People: "Nah, I'm fine just doing what I did the last 3650 days. I wonder how I deal with this issue I have with my boss at work. That is enough adventure for me."
Me: "Trash the job! Let's start a startup!"
People: "Nah, that is not for me. The benefit-to-work ratio at my current job is just too good."
This is a great idea — debugging why Cursor silently ignores rules is one of the most frustrating parts of the workflow. YAML frontmatter issues and glob pattern mismatches are the usual culprits in my experience.
One complementary approach I've found helpful: starting from well-tested rule templates rather than writing from scratch. I maintain a collection of .cursorrules files for 16+ frameworks (React, Go, Rust, Python, FastAPI, etc.) and built a free interactive generator at https://survivorforge.surge.sh/cursorrules-generator.html that lets you pick your stack and outputs a tested .cursorrules file.
Would be interesting to see if cursor-doctor could also suggest fixes for the issues it detects, not just flag them. That would make it a complete solution.
This is solving a real pain point. I manage .cursorrules files across 16+ framework configurations and keeping them consistent is a headache — especially when different tools want slightly different formats.
How does LynxPrompt handle the format translation between tools? For example, .cursorrules is free-form markdown but Claude's CLAUDE.md has its own conventions.
Yeah, those are the ones that keep us up at night. Deterministic checks catch the obvious regressions. The subtle ones — where the agent still "passes" but the vibe of its outputs has shifted — that's where we're leaning harder into longitudinal jury analysis: same criteria, same agent, tracked over time, so you can see the drift curve rather than just the current snapshot. Early days but it's the right shape of solution.
me and the team built chatgoat.ai
we integrated multiple ai chat models so you can switch anytime
gpt, claude, gemini, grok, deepseek, qwen etc
lots of ai roles/personas like writing assistant
image gen with banana and others
our app has more: video generation, create your own roles/characters
phone app and website sync fully
subscription cheaper than official sites + free trial to try it
no group chat
image gen separate from main chat
video only in app for now
link: https://www.chatgoat.ai
feedback welcome if anyone tries it
thanks
