Sergei Kotov

When I learned OOP, every book used Dog, Cat, and Animal classes. I thought: "Classes must model real-world objects." Later, I saw code with classes like: › AuthenticationHandler › DataSerializer › CacheManager Eureka! Classes aren't just mirrors of reality—they're abstractions for organizing logic. Real code rarely deals with Dogs and Cats unless you're building vet clinic software.

@EOEboh Both simultaneously and swap the database entirely if needed. That's not a problem for me. The problem for me is that the startup isn't growing 😢

Your startup is growing fast... Do you scale your database vertically or horizontally first?

@petarivanovv9 Does the bun deserve to be on the list?

If you're picking a frontend stack for 2026, learn these 15 tools (and the 6 I dropped): 👇🧵

When I can't get myself to work, I change locations. Office → cafe, home → public library, desk → couch. New place = mental reset. Sleep on the couch = physical reset.

@_jaydeepkarale Exactly this! It feels like AI usage reveals a level better than any interview ever did.

For the last 2 years, a lot of people predicted: “AI will replace software engineers.” What actually seems to be happening is more nuanced. The engineers getting the most leverage from AI are usually the ones with: - strong fundamentals - good system thinking - practical experience - and the ability to validate output critically Because AI accelerates implementation.But engineering was never just implementation. It is - Trade-offs. - Debugging. - Understanding production behavior. - Knowing when not to trust the generated answer. Ironically, AI is making strong engineers even more effective. And exposing weak fundamentals faster than ever. Good news for engineers who keep learning. Bad news for engineers who stopped thinking.

@driscollis I really don't know how I handled projects without the uv before.

What are the most popular Python build tools in 2026? 🐍 The landscape has changed quite a bit: • uv - The new standard. 10-100x faster than pip, replaces pip, virtualenv, pyenv, and pip-tools in one tool • Poetry - Still solid for library publishing and teams already using it • pip - The classic default that ships with Python • Conda - Best for data science and scientific computing • setuptools - The traditional build backend • hatch - Modern build tool with good plugin ecosystem • PDM - Standards-compliant with PEP 582 support uv has become the go-to for new projects because it handles everything: package installation, virtual environments, Python version management, and lockfiles. What build tool are you using?

Programmers build systems that run for years, but they can't maintain a sleep schedule for two weeks. The skill is there. The application is missing. You already think in systems. You just forgot you're one too.

@_jaydeepkarale Walking to the gym right now 💪

Did you workout today ???

Answer: B) [[5], [5], [5]] The trap: [[]] * 3 isn't 3 separated lists! › * 3 copies of the reference to the object (empty list), not the object itself › list[0], list[1], list[2] all point to the same inner list › append(5) on any one -- all three show [5] For beginners: * never deep-copies — it just multiplies references.

What is the output? Write your answer in the comments and share it with your friends✅✅

Answer: b) {'mom': 1, 'dad': 2, 'son': 3} The 95% beginners trap: d inside the function isn't the same d. › d = {...} inside a function creates a local variable › the original d outside is never touched › return d is also ignored (result isn't saved) For beginners: reassigning a parameter doesn't change the original.

Python Question What is the output of this code and why?

@nazanin_ashrafi Welcome back! Glad to see you again.

Okay I am Officially baaaaaaaack yall 😁 A lot has happened... stuff that I had no control over. But now let's get back to work 💪

@driscollis Indeed, it's possible to keep learning Python forever. Thank you!

Need to shorten a string while adding an ellipsis to indicate it's been shortened? You can use #Python's handy `textwrap` library for that:

@petarivanovv9 That's the real full-stack engineer list—not just backend and frontend.

8 habits that get you past senior, and skill is none of them: 1️⃣ Don't just learn your tech stack ↳ Learn how the business makes money, where it leaks, and what actually matters. 2️⃣ Don't speak in technical terms to non-technical people ↳ Translate complexity into impact, cost, and risk. 3️⃣ Don't blindly build what you're asked for ↳ Ask questions until the real problem is obvious. 4️⃣ Don't think only in terms of your team ↳ Think about customers, revenue, and long-term consequences. 5️⃣ Don't do work just because it landed in your backlog ↳ Figure out what the highest-priority work actually is. 6️⃣ Don't refuse to estimate because it's uncomfortable ↳ Offer ranges, trade-offs, and risks instead. 7️⃣ Don't assume deadlines are fixed ↳ Negotiate scope, risk, and budget like an adult. 8️⃣ Don't let missed deadlines surprise anyone ↳ Flag issues early and bring options, not excuses. Early in your career, technical ability is enough. Later on, it quietly stops being the deciding factor. The engineers who become genuinely valuable don't just ship features. They reduce risk. They clarify trade-offs. They help the business make better decisions. What one more thing would you add? Credits: Developing Skills by @johncrickett. —— 💾 Save for future reference. ♻ Repost to help others enhance their careers. 🔔 Follow me ( @petarivanovv9 ) to improve your engineering skills.

@PythonDvz Answer: NameError -- x isn't defined! No one assigned x! › Python sees x > 10 and looks up x › x doesn't exist, so it raises NameError: name 'x' is not defined › None of A, B, or C ever runs. For beginners: always assign variables before using them.

Comment your answer!

The gap isn't between Claude and GPT. It's between the person who can direct them clearly and the one who can't. Do you really understand the problem you're solving? The tool will follow.

@driscollis Any AI-agent writes awesome if someone awesome is conducting it.

Which AI agent writes the best #Python code?

@PythonPr Answer: 18 The 90% beginners trap: x += x + 4 isn't x += 9! › x = 7 (after +2) › x += x + 4 means x = x + (x + 4) › x = 7 + 7 + 4 = 18 For beginners: += expands first, then evaluates both sides with the current x.

Python Quiz: What is the Output?

@PythonDvz Answer: hello How it works: › "Good " + "Morning" = "Good Morning" (space is in the first string) › So a == "Good Morning" is True › print("hello") runs For beginners: + with strings joins them exactly as-is. You get exactly what's inside the quotes.

95% coders will fail this Python quiz 😱🐍 Do you know why it prints hello? 🤔 Comment your answer 👇

@PythonPr Answer: 44444 int(b) converts b, but a stays a string. And str * int doesn't multiply. It repeats. '4' * 5 = '44444' For beginners: want math? Convert both. int(a) * int(b) = 20

Python Quiz: Can You Predict This Python Output? 🤯

"It works" is the most expensive lie when you are learning programming. You feel like you're moving forward. But real progress isn't the result. It's the understanding underneath the result. Do you actually know what's happening here? Do you know why it works? Depth compounds. Skipping compounds too — just in the wrong direction.

@PythonPr Answer: 28.2 Two traps in one! › 2_4 is not a typo. An underscore = readability separator › 2_4 is 24 (same as 1_000_000 is 1000000) › x is a tuple (commas): (1, 24, 3.2) › sum(1, 24, 3.2) = 28.2 For beginners: _ in numbers is just visual spacing for readability.

Python Quz: Can you answer the below question?