THINGS THAT GET SOFTWARE DEVELOPERS FIRED FASTER THAN BAD CODE:
1. Breaking production and hiding it.
2. Saying "it works on my machine" like that ends the discussion.
3. Missing deadlines again and again without warning anyone.
4. Arguing in code reviews instead of understanding the feedback.
5. Building fancy abstractions when the task needed simple working code.
6. Ignoring tests, monitoring, and rollback plans.
7. Pushing risky changes without telling the team.
8. Blaming QA, DevOps, product, or juniors for your own misses.
9. Staying silent during incidents when communication matters most.
10. Writing code nobody can maintain and calling it smart engineering.
11. Not documenting critical flows because "I will remember it."
12. Overpromising in planning meetings and underdelivering in sprint end.
13. Being hard to work with even if you are technically good.
Bad code can be fixed.
Bad judgement, poor ownership, weak communication, and ego damage teams much faster.
A developer usually does not get fired because of one bug.
They get fired because nobody trusts them with important systems anymore.
Wrote my own DNS server in Node.js from scratch
Iโve been wanting to build a DNS server for a long time finally sat down and did it
Spent time into how DNS actually works things like RFC 1035, UDP (dgram), packet structure, parsing, all of it
Also wrote a detailed article breaking everything down step by step
If youโve ever been curious what really happens behind a domain lookup, this might be worth your time
Give it a read ๐
https://t.co/mYnHsbFd71
@OneTab
What an upgrade guys,
Literally from the folder view to task, star, remove duplicates, and all. Literally amazed guys, just wanted to appreciate you guys first.
Well done, who needs browser tabs now, maybe add onetab as a permanent sidebar in each of the browsers?? ๐
This is my assignment for a SF startup as a backend and infra engineer
the main goal was to build a workflow engine that executes directed acyclic graphs of tasks for this assessment the task would arbitrary HTTP endpoints.
But I replaced it with my own usage of news aggregator from hackernews, github (top repos for VAE, and libs related to Pytorch) and threads on reddit related to topic ML
This I can use in many paid gigs too as a CD deployements like
test > build > deploy staging > tests > deploy prod > notify
but for now it handles simple tasks and its built on go, redis, grpc with htmx and tailwindcss
This Reddit user shared his Ultimate No BS $0 Techstack
Bookmark to ship your next Big Product !
1. IDE - Google's AntiGravity (100% free + higher access if you use student ID)
2. Al Documentation - SuperDocs (100% free & open source)
3. Database - Supabase (Nano plan free, enough for basic needs)
4. Authentication - Stack Auth ( Free upto 10K users)
5. LLM (AI Model) - OpenRouter or Gemini via Al Studio for testing and a custom tuned model by Unsloth Al for production. ( You can fine-tune models using Unsloth literal in a Google Colab Notebook )
6. Version Maintenance/Distribution - Github/Gitlab ( both totally free and open source )
7. Faster Deployment - Vercel ( Free Tier Enough for Hobbyists )
8. Analytics PostHog, Microsoft Clarity & Google Analytics ( All 3 are free and independent for different tracking )
@arpit_bhayani You can have a million ๐ณ๐ฆ๐ฒ๐ถ๐ฆ๐ด๐ต๐ด(๐ฐ๐ข๐ฌ๐ก๐๐ฌ) but you need a ๐ญ๐ฐ๐ข๐ฅ ๐ฃ๐ข๐ญ๐ข๐ฏ๐ค๐ฆ๐ณ ๐ธ๐ช๐ต๐ฉ ๐ณ๐ข๐ต๐ฆ-๐ญ๐ช๐ฎ๐ช๐ต๐ช๐ฏ๐จ(๐ฐ๐ข๐๐) to make the ๐ด๐บ๐ด๐ต๐ฆ๐ฎ(๐ฅ๐ข๐๐) ๐ด๐ค๐ข๐ญ๐ข๐ฃ๐ญ๐ฆ(๐ก๐๐ฉ๐ฉ๐ฒ).
Thanks.
๐งต You can call OpenAI APIs. Good.
Now let's talk about what actual AI engineering looks like.
1/15 Prompting is programming with constraints.
You should think about:
- system vs user vs developer messages
- invariants
- few vs zero-shot tradeoffs
- adversarial inputs