frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Harness-of-Harness: Multi-Day Autonomous Software Development

https://www.alphaxiv.org/abs/2609.01481
1•ijidak•47s ago•0 comments

Neural networks reveal how experience shapes learning in both brains n machines

https://medicalxpress.com/news/2026-09-neural-networks-reveal-brains-machines.html
1•mdp2021•1m ago•0 comments

The Endless Temptation of Claude

https://discardpile.pika.page/posts/the-endless-temptation-of-claude
1•birdculture•4m ago•0 comments

Shutting down our public encrypted DNS servers and sponsoring Quad9 instead

https://mullvad.net/en/blog/shutting-down-our-public-encrypted-dns-servers-and-sponsoring-quad9-i...
1•Cider9986•8m ago•0 comments

Investt in AI

https://gmdmarkets.com/how-to-invest-in-ai
1•shlomysh•10m ago•0 comments

Pipe Mania

https://en.wikipedia.org/wiki/Pipe_Mania
1•tosh•10m ago•0 comments

Wearable machines that assist just one part of the body are proving useful

https://www.economist.com/science-and-technology/2026/09/02/humble-exoskeletons-work-better-than-...
2•scythe•13m ago•2 comments

Show HN: My wife is mobile Safari Wikipedia editor with a capital issue

2•cs1996•14m ago•0 comments

Adaptive Intelligence

https://blog.cloudflare.com/introducing-adaptive-intelligence/
2•torutofu•14m ago•0 comments

Observability Turned Uncertainty into a Business Model

https://humainary.io/notes/wide-events-how-observability-turned-uncertainty-into-a-business-model/
1•backlit4034•15m ago•0 comments

Learn CSS Layout

https://learnlayout.com/
1•chistev•17m ago•1 comments

On-Prem LLM Inference

https://wirt.ee/logbook/vllm-serving/
1•portmortem•18m ago•0 comments

Why Cheap Link Building Is the Most Expensive Marketing You Can Buy

https://www.entrepreneur.com/building-a-business/business-operations-logistics/why-cheap-link-bui...
1•santiago-pl•18m ago•0 comments

Show HN: Kestero, website widgets that record what they do (no cookies)

https://kestero.com
2•oscarMeldgaard•18m ago•0 comments

"Enshittification" Isn't Real

https://www.a16z.news/p/enshittification-isnt-real
5•qikouki•19m ago•2 comments

Fingerprinting and proof of work: how CrowdSec 1.8 tells browsers from bots

https://www.crowdsec.net/blog/crowdsec-1-8-waf-bot-detection-kubernetes
3•mazzma•19m ago•0 comments

Show HN: LibreDB Studio, a self-hosted MIT browser IDE for 10 database engines

https://github.com/libredb/libredb-studio
1•devgundogdu•19m ago•0 comments

Show HN: Hot. Dog. Bench. Mark. The AI benchmark we deserve

https://hotdogbenchmark.lol/
1•naniel•20m ago•1 comments

Perverts could push Norway to become the first country to ban smart glasses

https://www.neowin.net/news/perverts-could-push-norway-to-become-the-first-country-to-ban-smart-g...
2•bundie•21m ago•0 comments

Show HN: AI Usage Ball – Claude/Codex/Antigravity Quota as Mac Liquid Orbs

https://github.com/aiusageball/ai-usage-ball
1•xbyxy2000•23m ago•0 comments

Show HN: I built a better right-click menu for macOS

https://github.com/ganeshmshetty/openclip
1•ganeshmshetty•23m ago•0 comments

Show HN: Flipdot display with an AI chat that codes sandboxed animations for it

https://github.com/mdbug/flipdot
1•mdbug•24m ago•0 comments

When 'if' slows you down, avoid it

https://easylang.online/blog/branchless
1•theanonymousone•25m ago•0 comments

Skills vs. MCP tools for AI agents: when to use which

https://blog.logrocket.com/skills-vs-mcp-tools-agent-guide/
2•deepmem•25m ago•0 comments

Training a coding model to paint watercolours

https://huggingface.co/blog/train-to-paint-with-code
2•makaimc•29m ago•1 comments

Unveiling the Cognitive and Social Toll of ChatGPT Over-Reliance

https://onlinelibrary.wiley.com/doi/10.1002/acp.70205
2•geox•31m ago•0 comments

I turned the periodic table into a live market

https://elementswar.lol
1•tunarob•31m ago•0 comments

HotDog and Sausage Software: the HTML editor that barked at the world

https://karamatli.com/posts/hotdog-html-editor-sausage-software/
1•sparsesignal•32m ago•1 comments

Show HN: jeden.day – a compact habit tracker with flexible schedules

https://jeden.day/
3•mdbug•32m ago•0 comments

A Story About Joe

https://alisnic.github.io/posts/story-about-joe/
1•alisnic•35m ago•0 comments
Open in hackernews

My 600 Hours with AI Coding Assistants: A Practical Comparison

2•bv_dev•1y ago
After spending over 600 hours using various AI coding assistants over the past 3 months, I wanted to share my experience for those navigating this rapidly evolving landscape.

What I Mean by "Agentic Mode" First, to clarify: by "agentic mode," I'm referring to the assistant's ability to understand project context, reason through multi-step problems, and autonomously make coherent code changes across files without constant hand-holding. True agency means the tool can maintain context across interactions and execute on high-level directions.

The Current Landscape (May 2025) Augment Code - Current go-to tool despite higher costs

Strengths: Maintains context remarkably well across complex refactors; actually understands project structure; can implement feature requests that span multiple files Weaknesses: More expensive than alternatives ($30/month vs $20 for others); occasional hallucinations when venturing outside codebase context Best for: Complex refactoring tasks and implementing features that span multiple files

Windsurf - Slightly edges out Cursor for agentic capabilities

Strengths: Better context retention than Cursor; decent file traversal; good understanding of code relationships Weaknesses: Can get quite stuck in their full agentic mode as it starts editing things. While they have removed their flow credits part, it is still painful to watch it go completely out of context. Best for: Mid-size projects where you need moderate autonomy

Cursor - Popular but underwhelming for true agentic work

Strengths: Good IDE integration; clean interface; works reasonably well for single-file tasks. I like the ability to Cmd+K and insert a bulk of code in the middle. Also, I like the @Docs feature to bring latest documentation for popular libraries.

Weaknesses: Context falls apart in agentic mode; often loses track of previous instructions; requires excessive prompting Best for: Single-file optimizations and modifications, but not complex cross-file tasks

Claude Code - Declining quality since public beta

Strengths: Used to have superior reasoning and contextual understanding 3 months ago Weaknesses: Super expensive (like always), but recent updates have significantly degraded agentic capabilities; now requires much more hand-holding than before as it goes compleltely off base. Best for: Simple tasks that don't require deep contextual understanding Note: Most disappointing decline in quality - was previously much more capable. I spent $500 in Feb-Mar and thought it was worth.

Cline, Roo, and Aider - Conceptually interesting but practically limited

Strengths: Cline has good terminal integration; Roo offers interesting visualization; Aider has straightforward CLI Weaknesses: All three struggle with maintaining context; limited understanding of project structure; frequent need to repeat instructions Best for: Very simple, isolated coding tasks or experiments

Real-world Performance Differences The gap between these tools becomes most apparent when trying to implement complex features. For example, when asked to "add user authentication with email verification to my Express app":

Augment Code: Identified relevant files, added middleware, routes, and email service integration, then explained how the pieces fit together Windsurf/Cursor: Added authentication to single files I pointed at but needed explicit instructions for each additional component Others: Generally required file-by-file guidance with frequent context reminders

Conclusion If budget isn't a concern, Augment Code currently offers the most truly agentic experience, but still has a long way to go. For more budget-conscious developers, Windsurf slightly edges out Cursor for agentic capabilities, though both still require significant guidance for complex tasks.

Comments

SoMomentary•1y ago
I'm always surprised by people sleeping on GitHub Copilot. Is this because people truly don't find any value in it?
bv_dev•1y ago
I have used Github copilot since their beta release in 2023 and I don't find it anywhere near good these days. Automplete was good, but the industry has moved way beyond 2025. Copilot is slightly worse than Cursor which is itself a pretty average tool now. If you use truly agentic code generation, you won't be able to go back to Github Copilot.
SoMomentary•1y ago
You don't consider copilots agent mode to be agentic? I've had some pretty great results with agent mode + mcp to have it check it's own work.