frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

My 600 Hours with AI Coding Assistants: A Practical Comparison

2•bv_dev•7mo 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•7mo 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•7mo 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•7mo 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.

A pointless-looking URL shortener I use as a signal probe

https://short.uare.gay/
1•thecocaineninja•2m ago•0 comments

Hotswapping Haskell at Runtime (2017)

https://simonmar.github.io/posts/2017-10-17-hotswapping-haskell.html
1•sanufar•2m ago•0 comments

Admit It, You're in a Relationship with AI

https://www.bloomberg.com/opinion/articles/2025-12-30/ai-relationships-chatbots-have-a-social-cost
1•petethomas•3m ago•0 comments

Obsidian Audio Inbox

https://alexanderweichart.de/5_Archive/1.-Projects/audio-inbox/Obsidian-Audio-Inbox
1•surrTurr•3m ago•0 comments

Reverse-engineered a Sextortion Bot: Llama-7B instance with 2048 token window

https://old.reddit.com/r/LocalLLaMA/comments/1pzwlie/in_the_wild_reverseengineered_a_snapchat/
1•perelin•5m ago•0 comments

Honey's Dieselgate: Detecting and Tricking Testers

https://www.benedelman.org/honey-detecting-testers/
1•Leftium•8m ago•1 comments

Random Quality of Life Improvements That Will Change Your Life

https://twitter.com/thebeautyofsaas/status/2006104228381721052
1•gmays•10m ago•0 comments

A Year with Graphics

https://mropert.github.io/2025/12/30/a_year_with_graphics/
1•ibobev•11m ago•0 comments

Memory Safety Is

https://matklad.github.io/2025/12/30/memory-safety-is.html
1•ibobev•11m ago•0 comments

Updated LLM Benchmark (Gemini 3 Flash)

https://entropicthoughts.com/updated-llm-benchmark
2•ibobev•12m ago•1 comments

We have all we need to make mass surveillance a reality

https://idiallo.com/blog/we-have-all-we-need-for-mass-surveillance
2•firefoxd•14m ago•1 comments

Wildfire Smoke May Be More Dangerous Than Scientists Thought

https://scitechdaily.com/wildfire-smoke-may-be-far-more-dangerous-than-scientists-thought/
1•gmays•14m ago•0 comments

California DROP: Delete the information data brokers have about you coming 2026

https://consumer.drop.privacy.ca.gov/coming-soon.html
1•irsagent•17m ago•1 comments

Turning Dafny Sets into Sequences [video]

https://www.youtube.com/watch?v=-zAhtW8YFKM
2•larrytheliquid•22m ago•1 comments

Petition: Restore Free and Open Access to the ACM Digital Library

https://www.ipetitions.com/petition/restore-fully-free-and-open-access
1•underscoreF•24m ago•0 comments

X Users Have the Power to Edit Any Image Without Permission

https://petapixel.com/2025/12/29/x-users-have-the-power-to-edit-any-image-without-permission/
2•bookofjoe•31m ago•0 comments

We still don't know what Elon Musk's DOGE did

https://www.theguardian.com/technology/2025/dec/30/elon-musk-doge-impact-us-government
3•doener•33m ago•0 comments

Slaughterbots (2019 short film) [video]

https://www.youtube.com/watch?v=9fa9lVwHHqg
1•christianqchung•34m ago•1 comments

Show HN: Vektor – A native PHP vector database using HNSW

https://github.com/centamiv/vektor
1•centamiv•35m ago•0 comments

Waymo sues Santa Monica, and the city sues right back: Court fight ahead

https://www.latimes.com/california/story/2025-12-22/fight-between-waymo-santa-monica-goes-to-court
3•lokar•36m ago•2 comments

Attention Is Bayesian Inference

https://medium.com/@vishalmisra/attention-is-bayesian-inference-578c25db4501
2•samwillis•38m ago•0 comments

Unproven air taxi company is spending $126M to take over an L.A. airport

https://www.latimes.com/business/story/2025-11-24/california-air-taxi
1•PaulHoule•39m ago•0 comments

AI and politics and stagflation = workplace fatigue

https://www.glassdoor.com/blog/glassdoor-worker-fatigue-ai-politics/
3•andrewstetsenko•40m ago•0 comments

The Complete Sega Mark III (Retail) Collection

https://nintendosegajapan.com/2025/12/29/the-complete-sega-mark-iii-retail-collection/
1•msephton•40m ago•0 comments

Project ideas to appreciate the art of programming

https://codecrafters.io/blog/programming-project-ideas
13•vitaelabitur•42m ago•1 comments

Leadership Lab: The Craft of Writing Effectively (2014) [video]

https://www.youtube.com/watch?v=vtIzMaLkCaM
1•rognjen•43m ago•0 comments

Penn and Teller Help Rob Pike and Dennis Ritchie Play a Prank on Arno Penzias [video]

https://www.youtube.com/watch?v=fxMKuv0A6z4
2•susam•46m ago•0 comments

No Longer Burying the Lead: A New Media Culture for the Metacrisis

https://www.whatisemerging.com/opinions/no-longer-burying-the-lead
1•rendx•49m ago•2 comments

Alias Method

https://en.wikipedia.org/wiki/Alias_method
1•usgroup•53m ago•0 comments

I exposed my Homelab through Cloudflare Tunnels

http://ebourgess.dev/posts/exposing-homelab-through-cloudflare-tunnel/
3•ebourgess•54m ago•2 comments