惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Jina AI
Jina AI
NISL@THU
NISL@THU
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
GbyAI
GbyAI
SecWiki News
SecWiki News
Microsoft Azure Blog
Microsoft Azure Blog
J
Java Code Geeks
B
Blog RSS Feed
Blog — PlanetScale
Blog — PlanetScale
Schneier on Security
Schneier on Security
V
Vulnerabilities – Threatpost
C
CXSECURITY Database RSS Feed - CXSecurity.com
V
Visual Studio Blog
宝玉的分享
宝玉的分享
Recent Announcements
Recent Announcements
T
True Tiger Recordings
F
Full Disclosure
Martin Fowler
Martin Fowler
D
Docker
Stack Overflow Blog
Stack Overflow Blog
Security Latest
Security Latest
A
About on SuperTechFans
雷峰网
雷峰网
Know Your Adversary
Know Your Adversary
Application and Cybersecurity Blog
Application and Cybersecurity Blog
Hacker News: Ask HN
Hacker News: Ask HN
B
Blog
V
V2EX - 技术
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Google DeepMind News
Google DeepMind News
S
Security Archives - TechRepublic
Google DeepMind News
Google DeepMind News
人人都是产品经理
人人都是产品经理
Malwarebytes
Malwarebytes
C
Check Point Blog
美团技术团队
P
Privacy International News Feed
Recorded Future
Recorded Future
博客园 - 司徒正美
T
The Blog of Author Tim Ferriss
L
LangChain Blog
Project Zero
Project Zero
P
Proofpoint News Feed
有赞技术团队
有赞技术团队
P
Proofpoint News Feed
Scott Helme
Scott Helme
C
CERT Recently Published Vulnerability Notes
云风的 BLOG
云风的 BLOG
T
ThreatConnect
F
Fox-IT International blog

DEV Community

Beginning to market Why Your Treasure Hunt Engine Kept Crashing at 1.2M Concurrent Connections Introducing Batch Processing for ZeroGPU Kiln Crisis Management: Controlling Irregular Raw Meal in CCR Using Python The Grilling Optimizing a High-Throughput Browser-Based Box Shadow Generator: Debounced State Updates and Chunked File Readers I Was Spending $3,200/Month on GPT. Then I Tried Chinese Models. Why You Must Stop Pasting Production Payloads into Web Decoders: Building a Secure Base64 Decode Strategy Message Brokers Comparison 2026 — Kafka, RabbitMQ, NATS & Redis Streams: Which One Should You Choose? Your Git Tree Looks Like a Crime Scene: How to Write Commits That Don’t Suck I tried every popular library for programmatic PDF form filling. None of them survived production The const enum that took down our payments Architecture of Chaos Part 3 — Event Sourcing Saved Our Audit Trail, Then a Fiber Cable Broke Stop Paying Per Cert. It's Crazy. Building Embeddable Browser Games for Website Engagement Build a Privacy-First Tampermonkey Script for Long ChatGPT Conversations XSS Attacks Are Everywhere: Reflected, Stored, DOM-Based — How to Actually Fix Them (2026) Stop letting LLMs hallucinate dates — a tool for AI agents The Platform Team Became a Finance Team /align v0.8 — personal evals for Claude Code, maintained by an LLM agent Copilot helped me deploy my passion project to the App Store Software Engineering: The Art of Thinking Out Loud (with AI) Leaked Kubernetes Secrets: Impact Assessment and Mitigation Strategies First 90 days as a junior engineer on an AI-heavy team: what to learn first Something Honest About Being a Developer on This Kind of Team JSON Schema Validator Advanced Techniques for Power Users I Built Hermes Immune System — A Safety Lab for AI Agents Google I/O 2026: MCP Is Now Infrastructure (Spark, Managed Agents, WebMCP & More) Probabilistic Graph Neural Inference for deep-sea exploration habitat design for extreme data sparsity scenarios QuantConnect Review: Running 2,400 Backtests Without Installing a Single Python Library The Complete Guide to Video APIs in 2026 (And Why Your Choice of Tool Actually Matters) Alpha Vantage vs Yahoo Finance API: Free Market Data for Side Projects — An Honest Comparison Day 20 of 60: I Built a Production-Grade Authentication System with JWT Tokens and API Key Managemen Nobody on the internet knows if you are a human The fastest way to optimize images for your web projects (Zero Server Roundtrips) We Got Burned by Veltrix Configuration Layer and Lived to Tell the Story Why Block Handed Goose to the Linux Foundation: Agentic AI Goes Open The Delve Scandal Proved SOC 2 Is Broken — Here's What Micro-SaaS Founders Should Do Instead OpenTelemetry: The Foundation of Modern Cloud-Native Observability — Traces, Metrics, Logs, and the Future of Observability Arc Browser Review: 18 Months With a Browser That Thinks Differently [Boost] Docker healthchecks: what they actually measure and what you shouldn't promise Docker healthchecks: qué miden de verdad y qué no deberías prometer I Built an AI That Roasts Cold Emails — Here's What 18,000 Drafts Taught Me Are You My Parent?: Scaffolding in the architecture necessary for keyboard handling between components. The AI Labs Found Product-Market Fit in April How I Stopped Fighting AI Context: JetBrains AI vs. Copilot in Rider I Accidentally force-pushed to main at 11 PM — So I Built an Interactive Git Undo Tool Perplexity Spaces vs You.com vs Phind: which AI search fits your dev research workflow I'm 14, can't code, and built a cognitive state app in one day — here's what happened Three Cloudflare Patterns Earned the Hard Way Aider Review: The Open-Source AI Pair Programmer That Works With Any LLM How to Measure and Improve Core Web Vitals in Under 30 Minutes Standardizing Feature Flags Is Easy to Agree On. Migrating Safely Is the Hard Part. What if UI tests validated user experience instead of selectors? Why I Stopped Believing 'Best Practices' and Started Trusting 'Works For Us' PrestaShop Doctrine: Automatically Manage the DB Prefix PrestaShop Enterprise vs Shopify Plus A .NET Dinosaur in Web3 — Day 15: DAO Voting Halyra IDE Wearable App Development Cost: How to Build a Quality MVP Without Overspending New in Vue - May 2026 427 Remote Companies Using TypeScript in 2026 MCP CI gates need receipts: tools/list is not enough 📖 DICTIONARIES IN PYTHON: THE SMART DATA VAULT I Generated a Tableau Dashboard Using Gemma 4 — Locally, No API Key, No Cloud The Hidden Way Electronics Can Start a Fire — Even Without an Open Flame I Built a Beginner-Friendly NGINX Automation CLI for Linux Servers Vibe Thinking - The PM Who Writes Requirements That an AI Can Actually Use A Refreshing Perspective on AI and Truth Kubelet Metrics: How cAdvisor and CRI Collect Kubernetes Stats How to Optimize MongoDB on Bare Metal Servers: SRE Playbook Why I Built Bamise Instead of Using Laravel How to Build a Clean Academic Dataset Without Losing Your Mind (or Your Weekend) Kubernetes Is Eating Your Budget: How to Fix EKS Over-Provisioning What Awnings Taught Me About Developer Experience Tree Traversal: Why the Order You Pick Is a Data Flow Decision I built my own forum using PHP- it came out great Optimizing Chunking and Data Extraction for Zero-Hallucination RAG Controlling Blender with AI — Building an MCP Server for 3D Creation 5 Smart Contract Vulnerabilities Every Developer Should Know in 2026 Cursor users who write failing tests before prompting the AI complete features in 37% fewer iterations than those who pr When AI Becomes a Danger: 370,000 Grok Conversations Exposed I Refactored 100 Functions With Claude. CI Was Green. Production Got Slower in 7 Spots. I read my own commits like a stranger Child Safety vs. Data Center Dollars The Reason Your AI Chatbot Feels Fast Has Nothing to Do With a Better Model Beyond Vibe-Coding What I learned testing AI translation tools in 2026 (DeepL is still good, but LLMs caught up) AWS ECS Fargate Cost Allocation: Why Your Per-Cluster Spend Shows as One Line How to Surface License Violations in GitHub Advanced Security with feluda We Deleted 10 Real Users with a Test-Cleanup Script — RCA The Decision Subtraction Framework: How to Evaluate Any AI Tool How I Access My Home PC From Anywhere Without Spending a Penny # agents.md: Teaching AI Agents How to Scrape (The Future of Web Automation) KAI vs Global vs Tojiro vs Miyabi: How to Actually Tell Japanese Knife Brands Apart Why We Accidentally Blocked Our Users: A Deep Dive into Idempotency in Distributed Systems I Connected Hermes Agent to a Live MCP Server with 59 Tools and Here's What It Actually Built Our first app is finally live on the Play Store after 4 months of hard work 🚀 I Built UUIDs That Look Random But Sort Like Timestamps (50% Smaller Indexes!)
I Thought Coding Was The Job
Aryan Choudh · 2026-05-28 · via DEV Community

Two years ago, when I got my first freelance client, I was still in my final semester of college.

A guy approached me on LinkedIn because he wanted an app built for some gym equipment he was selling.

To this day, he’s probably one of the most professional people I’ve worked with.

And being the naive amateur I was, I thought the hard part would be building the product.

You know… the actual coding part.

The React Native components.
The backend logic.
The deployment.
The bugs.

That was the “real work” in my head.

Everything else felt secondary.

I was wrong almost immediately.


As the project progressed though, things actually started going surprisingly smoothly.

Which was honestly lucky because life outside the project was kind of falling apart at the same time.

I had just gone through a heartbreak, my final semester was ending, career uncertainty was sitting in the background constantly… and somehow in the middle of all that, I was building this app with my friend @hisukurifu. Really grateful for his support, even today.

nod gif

Looking back, I think having work to focus on genuinely helped me hold myself together during that phase.

And somehow, despite all the chaos, the project itself stayed stable.

The client paid on time.
The scope stayed reasonable.
Communication stayed respectful.

Hell, the guy even paid for renting a Mac because we needed it to build the iOS version and neither of us owned one at the time 😭

At that point, freelancing looked very simple from the outside.

A client needs something.
You build it.
You get paid.
Everybody wins.

Clean. Logical. Straightforward.

At least that’s what Instagram reels and “How I Made $10k Freelancing” YouTube thumbnails had convinced me of.

Reality felt very different.

Because suddenly there were things nobody really talks about when you’re learning development.

Things like:

  • figuring out what the client actually wants
  • discussing timelines
  • dealing with uncertainty
  • deciding pricing
  • revisions
  • feature changes halfway through
  • awkward conversations
  • scope creep
  • waiting for replies
  • wondering if the project will even go through

And honestly?

At the beginning, all of that felt more overwhelming than the code itself.

But slowly, we figured things out as we went.


One thing I still remember is that I never made a written agreement with that client.

Everything was done through pure verbal trust.

And thank GOD he didn’t suddenly increase the scope or disappear halfway through because looking back now… yeah that could’ve gone very badly 😭

Guess some small part of the world is still rainbows and sunshine lol.

But around five months after that project ended, another client opportunity came up.

This time we actually started properly planning things out beforehand.

Discussions.
Requirements.
Agreement drafting.

And then after about a week…

the whole thing just ended abruptly because the project itself was still being aligned internally on their side.

Nobody’s fault honestly.

But that experience finally made something click for me.

I remembered one piece of advice a friend had given me during my freelancing phase:

“You don’t get lucky twice with clients. Make agreements beforehand.”

At the time, that sounded overly serious to me.

Like bro… we’re just building websites 😭

But eventually I understood what they meant.

Because the more I zoomed out on freelancing as a career path, the more I realized I wasn’t just building software.

I was managing:

  • expectations
  • communication
  • trust
  • uncertainty

The code was only one part of the system.

And honestly, that realization changed how I looked at work entirely.


The strange thing is… I thought this problem only existed in freelancing.

Then I entered corporate.

And somehow the exact same realization came back wearing formal clothes.

Before starting my job, I still had this very simplified mental image of software engineering.

I thought:

“Okay, now I’ll finally work in a proper engineering environment.”

Meaning:

  • coding
  • solving technical problems
  • building systems
  • learning architecture

And yes, those things do exist.

But again, there was this whole invisible layer nobody really prepares you for.

Things like:

  • hierarchy
  • communication styles
  • meetings
  • visibility
  • asking questions correctly
  • understanding team dynamics
  • knowing when to speak
  • knowing when not to
  • learning how people actually work together

And once again, I realized:

The technical part was only one layer of the job.


Even learning Japanese started feeling the same way after I began using it professionally.

At first, language learning felt like:

vocabulary + grammar = communication

Simple.

Then workplace conversations arrived and suddenly communication became:

  • timing
  • confidence
  • hierarchy
  • context
  • listening
  • reading atmosphere
  • adapting to people

Different field.

Same realization.


Even marketing ended up teaching me this.

I originally got into it thinking:

“Okay this is probably just posting content and promoting things.”

But behind that was:

  • positioning
  • audience psychology
  • negotiations
  • understanding attention
  • figuring out why some things spread and others don’t

Again, the visible part was tiny compared to everything underneath.


And honestly, I think this pattern exists in almost every career.

From the outside, most professions look simple because we only see the visible output.

Everything is an iceberg.

We see:

  • the deployed app
  • the successful freelancer
  • the polished presentation
  • the fluent speaker
  • the viral post

But we usually don’t see:

  • uncertainty
  • communication
  • failed attempts
  • invisible coordination
  • emotional pressure
  • relationships
  • trust
  • adaptation

The deeper you go into anything, the more human it becomes.

And I think that was the biggest surprise for me.

Not that coding was hard.

But that almost every meaningful thing around coding involved people.


Looking back, I think younger me believed careers were mostly about skills.

Now I think they’re more about people skills.

Technical systems.
Social systems.
Communication systems.
Even emotional systems sometimes.

And honestly? That realization used to overwhelm me a little.

Now I weirdly find it interesting.
Because it makes every field feel deeper than it first appears.


So yeah.
Turns out coding wasn’t “the job.”
It was just the most visible part of it.
And maybe that’s true for a lot more things in life than we realize.


I’m curious now though:

Have any of you had a similar realization after entering a field professionally?

Where the thing you thought was “the main skill” ended up being only a small part of the actual job?