Feed aggregator

Here are hints and answers for the NYT Strands puzzle for Feb. 8, No. 707.&

My first job in management was not good. I became so stressed I got a mild case of shingles[0], and at that point decided to quit for my own health. I didn’t stay long enough to overcome the difficulties—I had to run away.

I’m wondering if this job was just especially bad, or if I can expect similar troubles in any management job.

It was consulting company work, and I enjoyed it when I was an individual contributor on the team. The company cared about our project in the beginning and things were well planned. If we fell behind, the consulting company would give us extra people to help catch up.

Then they made me team lead and all of that stopped. We were left to flounder. Or maybe I was just bad at planning and asking for help?

The team was filled with good people, but I always felt like I was leading “the B team.” In part because the team was originally assembled to do Python work, but then the client decided to start using TypeScript, but only for some things. None of us knew TypeScript.

Ultimately, I couldn’t trust that the other members of the team would get their work done well. I was expected to spend some of my time programming as the team lead, and so I felt a lot of pressure to personally fix all the code the team produced.

Was this team lead position just an especially bad one? I think I’m about to answer that question with this next observation:

Our client was a company run by Alice, and our technical manager was Bob (names changed). Bob wanted us fired so he could hire another consulting company made up of people he was friends with, but Alice was friends with executives at my consulting company. So my team and I were forced to work with Bob, even though Bob didn’t want to work with us.

It was really screwed up, and management at the consulting company told us at one point to stop talking to Bob—we literally couldn't talk to our client for a time. This seemed like a red line for me. Once everyone stops working toward solving a problem and instead company politics are the #1 consideration, it’s time to go. Or so I thought—once you’re actually in a job, leaving isn’t always as easy as you expect.

I’ve also been frustrated throughout my career because it seems like in most jobs, company politics are actually more important than solving the problem. Is this true?

Anyway, I guess this turned into a bit of a ramble.

I’m looking for new work again, and I’m wondering if I should be open to management roles again. I really didn’t like my first leadership experience. But I have a lot of experience and enjoy being included in the higher-level decisions that come with management. I do hate the politics, though.

Is it unreasonable to hope for a job where everyone works toward solving technical problems without politics?

----

[0]: I also learned that I have celiac disease (which is an autoimmune disease) during the same time I had shingles. It was a rough time. The shingles cleared on their own, but there was one evening where they were hurting worse than anything I've ever experienced. All of this made it feel like the stress was going to kill me.

Comments URL: https://news.ycombinator.com/item?id=46927652

Points: 1

# Comments: 0

I've worked on the design-end of engineering for some time, and have done my share of crimping for prototyping / development, but have never done electrical design for mass production. I'm now working with a small company who is scaling up their early prototypes, and we're going to assemble the early ones in house -- think 10s of units. This is the first time that I'm involved in this scale of manual assembly. I find I'm spending significant time _just making crimps_, and this makes me wonder if I can get better at making crimps, or make a design change to simplify the assembly. I take about 60 sec per crimp, and I don't have the exact tools in this guide, but it is basically what I do: https://www.youtube.com/watch?v=jHfYzrSF4pY

Here's the design constraints that come to mind:

1. We have about 15 connections to make

2. Most of our connections are JST-XH, some are JST-SM. We could maybe simplify to just JST-XH and then buy pre-crimped wires, but we like having the locking connections for wire-to-wire connections

3. Sometimes we need different connectors at opposing ends of wire. If we buy pre-crimped SM and SH, we could just solder the appropriate lines together.

4. we want to keep the 2.54mm pitch, to interface with simple OTS boards

My best ideas:

1. Buy pre-crimped JST-XH wires, and change one end to JST-SM as needed. This should get rid of ~80% the crimping effort.

2. Move entirely to JST-XH, buy pre-crimped wires, and accept slightly worse wire-to-wire connections.

3. Buy better crimp tools, practice like hell, and 'git gud'

Are there other ideas I should consider? Is there a secret or a better guide to crimp faster? I know this is a noob-ish question, any help is appreciated. I have googled, searched youtube, and asked the various LLMs, and the ideas listed above are the result.

Comments URL: https://news.ycombinator.com/item?id=46927634

Points: 1

# Comments: 0

Article URL: https://arxiv.org/abs/2511.01815

Comments URL: https://news.ycombinator.com/item?id=46927580

Points: 1

# Comments: 0

Article URL: https://www.nature.com/articles/s41467-025-67747-9

Comments URL: https://news.ycombinator.com/item?id=46927562

Points: 1

# Comments: 0

Article URL: https://www.wsj.com/world/india/why-big-tech-is-throwing-cash-into-india-in-quest-for-ai-supremacy-a5d02f0e

Comments URL: https://news.ycombinator.com/item?id=46927559

Points: 1

# Comments: 0

Article URL: https://github.com/aweussom/HowToShootYourselfInTheFoot

Comments URL: https://news.ycombinator.com/item?id=46927552

Points: 1

# Comments: 0

Article URL: https://crawshaw.io/blog/eight-more-months-of-agents

Comments URL: https://news.ycombinator.com/item?id=46927533

Points: 3

# Comments: 0

Article URL: https://www.psychologytoday.com/us/blog/the-digital-self/202602/from-human-thought-to-machine-coordination

Comments URL: https://news.ycombinator.com/item?id=46927529

Points: 1

# Comments: 0

Article URL: https://developer.x.com/

Comments URL: https://news.ycombinator.com/item?id=46927518

Points: 1

# Comments: 0

I’m building RMA (Rust Monorepo Analyzer) and a small dashboard for it.

The motivation: I love tools like Semgrep/CodeQL, but on big monorepos I kept ending up in “run scan → go make tea” territory. RMA is my attempt at a single-binary, fast SAST + code intelligence scanner that still outputs something you can actually triage.

What you can try in the dashboard:

Browse scan results with dedupe (unique issues vs occurrences)

Export SARIF (GitHub Security tab friendly) and JSON

Filter security findings by subcategory so “real vulns” don’t get drowned out by audit/hardening noise

Default suppression for generated files (toggle to include)

A detail I’m pretty happy about: we now propagate rule metadata end-to-end (confidence, CWE/OWASP, references, impact/likelihood, etc.) so the findings carry enough context to be actionable.

If you try it, I’d love feedback on:

what feels noisy / what feels missing,

which repos it breaks on,

what you’d want for CI gating (vuln-only vs audit/hardening).

Dashboard: https://rma-dashboard.bukhari-kibuka7.workers.dev/

CLI/source: https://github.com/bumahkib7/rust-monorepo-analyzer

Comments URL: https://news.ycombinator.com/item?id=46927513

Points: 1

# Comments: 0

Hey everyone,

I've built a source code graphRAG for Java/Kotlin project agentic analysis and development. It is built on top of jQAssistant with a detailed knowledge graphRAG in Neo4j.

You can check out the code here: https://github.com/2015xli/jqassistant-graph-rag

What it can do:

* "What's the main purpose of the com.example.auth package?" * "Show me the call chain leading to the processPayment method." * "What services use the UserRepository class?" How it works:

1. Graph Creation: It uses jQAssistant and Java/Kotlin source file parsers to analyze your code's structure, dependencies, and relationships. It essentially builds graph overlays for a source code tree and build artifact tree. 2. GraphRAG Enrichment: It then enriches the graph and generates summaries from individual methods and classes all the way up to packages and the entire project. Embeddings are generated for the summaries to facilitate semantic query. 3. MCP server and Agent: It exposes the graphRAG capabilities through an MCP server and an example coding agent. You can use them to accomplish complex tasks. Other design features:

* Modular design that can be easily adapted to new graphRAGs for other languages. * Parallelized summarization process and summary cache management to save the cost in money and time. The project is still a work in progress, but I'd love to get your feedback. Thanks for taking a look

Btw, I also built a source code graphRAG for C/C++ development at https://github.com/2015xli/clangd-graph-rag.

Comments URL: https://news.ycombinator.com/item?id=46927460

Points: 1

# Comments: 0

Article URL: https://mccue.dev/pages/2-6-26-python-competitor

Comments URL: https://news.ycombinator.com/item?id=46927441

Points: 3

# Comments: 0

Article URL: https://www.mauriciopoppe.com/notes/tmux-to-zellij/

Comments URL: https://news.ycombinator.com/item?id=46927429

Points: 1

# Comments: 1

I'm curious how you all are going beyond AI-powered editors (Cursor, Claude Code, etc.) and vibe coding to meaningful accelerate your workflows and improve output quality using specialized agents.

There's a lot of talk about decomposing work into sub-agents with specific responsibilities, but not a lot of specifics (unless I'm looking in the wrong places). I'd love to hear concrete examples of how you are actually doing this in practice.

What specialized agents have you built and what's each one responsible for? (e.g., code review agent, test generation agent, integration validator, design critic) How are you wiring them together? Are you using frameworks, custom scripts, or API orchestration?

What's your handoff strategy between agents? How do you ensure they work cohesively instead of creating inconsistencies?

How do you maintain quality and catch issues that might slip through? (testing, integration problems, design coherence, etc.)

What unexpected benefits or challenges have you encountered with your approaches?

Comments URL: https://news.ycombinator.com/item?id=46927410

Points: 1

# Comments: 0

Article URL: https://signoz.io/blog/otel-baggage/

Comments URL: https://news.ycombinator.com/item?id=46927400

Points: 1

# Comments: 0

Article URL: https://davmail.sourceforge.net/

Comments URL: https://news.ycombinator.com/item?id=46927388

Points: 1

# Comments: 0

Article URL: https://github.com/sqlmodel/sqlmodel

Comments URL: https://news.ycombinator.com/item?id=46927362

Points: 1

# Comments: 0

OP here. We built Tharos because we were tired of "security theatre" tools that produce read-only reports.

Tharos is an open-source security engine written in Go. Key features:

Polyglot: Scans Go, Python, JS/TS (AST-based + Regex). Interactive: tharos --interactive opens a TUI to fix issues one-by-one. CI-Ready: Returns strict exit codes for blocking builds. Visual: HTML reports and rich terminal UI. Repo: https://github.com/chinonsochikelue/tharos Docs: https://tharos.vercel.app

Would love feedback on the rule engine and the TUI ux!

Comments URL: https://news.ycombinator.com/item?id=46927353

Points: 1

# Comments: 0