Passionate Software Engineer ๐ | Specializing in Mobile & Fullstack Development | Committed to constant growth with #100DaysOfCode | Opinions are my own. ๐โจ
The problem with slop is you canโt see it if thatโs all youโve ever seen. Once you see clean code and clean architecture and appreciate the benefits, you canโt deal with slop anymore.
Claude Code has been super slow today and has been generating some of the worst code I've seen.
Good luck to those who think code reviews aren't necessary. You'll learn the hard way: one day the pile of AI slop will be so big that even the simplest changes take forever and break unrelated features along the way.
@OmbeniSheshe You bought an MF battery when what your vehicle needs is an EFB or AGM battery due to the i-Stop system. The price should be around 20k though.
CEO of Coinbase bragging that non-technical people are now using AI to ship production code. When a CEO says this, one of two things is true: either they don't know what production code is, or they're selling you a story.
This kind of hype is how businesses end up in real trouble. The people shipping the code and the CEOs cheering them on won't be the ones held responsible when it breaks. They collect their paycheck either way. Users will find out the hard way, once unreviewed code has touched their money, medical records, identities, private messages, kids' data, legal documents, critical infrastructure. By then it's too late.
Production code isn't a vibe.
๐๐ฎ๐ฟ๐ฎ๐๐ฒ๐น ๐ฟ๐๐น๐ฒ๐ ๐๐ง๐ข๐ ๐บ๐๐๐ ๐ฒ๐ป๐ณ๐ผ๐ฟ๐ฐ๐ฒ ๐ฏ๐ ๐ฑ๐ฒ๐ณ๐ฎ๐๐น๐
Laravel moves fast.
Thatโs exactly why it needs rules.
If you donโt set these early, youโll enforce them later โ under pressure.
Here are the Laravel rules every CTO should make non-negotiable ๐
๐ญ. ๐ก๐ผ ๐๐ป๐ฐ๐ผ๐ป๐๐ฟ๐ผ๐น๐น๐ฒ๐ฑ ๐ฑ๐ฎ๐๐ฎ ๐ณ๐ฒ๐๐ฐ๐ต๐ถ๐ป๐ด
If a query returns โeverything,โ itโs a bug. Large datasets must always be constrained, paginated, or streamed.
๐ฎ. ๐ฅ๐ฒ๐น๐ฎ๐๐ถ๐ผ๐ป๐๐ต๐ถ๐ฝ๐ ๐บ๐๐๐ ๐ฏ๐ฒ ๐ถ๐ป๐๐ฒ๐ป๐๐ถ๐ผ๐ป๐ฎ๐น
Lazy loading is never accidental at scale. If relationships are accessed, they must be planned and reviewed.
๐ฏ. ๐๐ป๐ฑ๐ฒ๐ ๐ถ๐ป๐ด ๐ถ๐ ๐ฝ๐ฎ๐ฟ๐ ๐ผ๐ณ ๐ณ๐ฒ๐ฎ๐๐๐ฟ๐ฒ ๐ฑ๐ฒ๐น๐ถ๐๐ฒ๐ฟ๐
Any new filter, lookup, or foreign reference ships with an index โ not as a follow-up task.
๐ฐ. ๐ฅ๐ฒ๐พ๐๐ฒ๐๐ ๐ฐ๐๐ฐ๐น๐ฒ๐ ๐๐๐ฎ๐ ๐น๐ถ๐ด๐ต๐๐๐ฒ๐ถ๐ด๐ต๐
Anything slow, repeatable, or failure-prone runs outside the user request. Queues arenโt optional once traffic exists.
๐ฑ. ๐๐ผ๐น๐น๐ฒ๐ฐ๐๐ถ๐ผ๐ป๐ ๐ฎ๐ฟ๐ฒ ๐ป๐ผ๐ ๐ฎ ๐ฑ๐ฎ๐๐ฎ ๐ฒ๐ป๐ด๐ถ๐ป๐ฒ
If logic can be pushed to the database layer, it should be. Memory-heavy collection work is a red flag.
๐ฒ. ๐ฃ๐ฎ๐ด๐ถ๐ป๐ฎ๐๐ถ๐ผ๐ป ๐ถ๐ ๐บ๐ฎ๐ป๐ฑ๐ฎ๐๐ผ๐ฟ๐ ๐ณ๐ผ๐ฟ ๐น๐ถ๐๐๐
No exceptions. If a screen can grow, it must paginate from day one.
๐ณ. ๐ง๐ต๐ฒ ๐ฑ๐ฎ๐๐ฎ๐ฏ๐ฎ๐๐ฒ ๐ถ๐ ๐ป๐ฒ๐๐ฒ๐ฟ ๐ฎ ๐ฐ๐ฎ๐ฐ๐ต๐ฒ
Repeat reads require a caching strategy. Database load should represent truth, not convenience.
๐ด. ๐ฃ๐ฒ๐ฟ๐ณ๐ผ๐ฟ๐บ๐ฎ๐ป๐ฐ๐ฒ ๐ถ๐ ๐บ๐ฒ๐ฎ๐๐๐ฟ๐ฒ๐ฑ, ๐ป๐ผ๐ ๐ฎ๐๐๐๐บ๐ฒ๐ฑ
Query counts, execution time, and memory usage are reviewed regularly โ not only during incidents.
๐ต. ๐๐น๐ผ๐พ๐๐ฒ๐ป๐ ๐ฐ๐ผ๐ป๐๐ฒ๐ป๐ถ๐ฒ๐ป๐ฐ๐ฒ ๐ถ๐ ๐ฎ๐๐ฑ๐ถ๐๐ฒ๐ฑ
Accessors, casts, appended attributes, and global scopes are treated as performance-sensitive code.
๐ญ๐ฌ. ๐ฆ๐ฐ๐ฎ๐น๐ถ๐ป๐ด ๐ฟ๐๐น๐ฒ๐ ๐ฎ๐ฟ๐ฒ ๐ฑ๐ผ๐ฐ๐๐บ๐ฒ๐ป๐๐ฒ๐ฑ ๐ฒ๐ฎ๐ฟ๐น๐
Every Laravel codebase needs written performance boundaries โ before the team doubles.
Laravel doesnโt collapse because itโs slow.
It collapses because leaders let convenience replace discipline.
CTOs donโt need to block speed.
They need to protect the system from its own success.
Thatโs how Laravel products survive real growth.
@fidexcode It's just pure preference for what works for you as a developer, and the rest is nonsense. I use both light mode and dark mode depending on whether I am coding during daytime or at night.
Mark Zuckerberg scaled Facebook in 2005 - without Kubernetes, Serverless Functions, Redis, Managed Auth, Rust, or Kafka.
No fancy orchestration.
No distributed event streams.
No cloud-native anything.
Modern devs love to over-engineer.
We build as if our projects will scale to 1 million users tomorrow.
But here's the truth:
Most apps die with 100 users, not 1 million.
We predict every possible future case:
โ So we add layers, queues, caches, and microservices...
for problems that don't even exist yet.
That's not architecture.
That's trend-driven engineering.
Always architect your apps based on current needs (not possible future guesses), leaving room for extension and evolution.
Complexity kills projects - simplicity scales.
Don't design your MVP like it's Facebook in 2025.
Design it like Facebook in 2005 - simple, fast, and focused on what matters most: value.
๐ Join 16,000 software engineers who learn how to build software based on current needs and not the trends:
โณ https://t.co/dajCViEeyi
๐๐ผ๐ป๐๐: every subscriber gets a PDF with 650+ exclusive resources for mastering C#, .NET, ASP .NET Core, EF Core, and Microservices.
โโ
โป๏ธ Repost to help others learn how to build software based on current needs
โ Follow me ( @AntonMartyniuk ) to improve your .NET Skills
@rqobela Turbo Pascal. My first programming language (2006) that I used for an examination project at the secondary school level in 2007. I never used it again since it was (almost) obsolete. I tried Visual Basic in 2008 then started learning C++ in 2009 in preparation for university.
It's quite unbelievable that 14 years have already lapsed since I joined X (formerly Twitter) - from the good old days when Twitter #Bootstrap (the frontend framework) was first released to the public! #MyXAnniversary