22 Apr 2026

"Never scaffold. Always fully implement." — dev philosophy

One rule I've learned the hard way: never scaffold. Always fully implement.

Every time you write a stub, a skeleton, or a "TODO: implement later" — you're not saving time. You're lying to yourself. The stub becomes the implementation. The placeholder ships. The "I'll come back to this" never comes.

I used to think scaffolding was planning. It's not. It's avoidance with good intentions.

The discipline I apply now: if I'm not ready to implement it properly — I don't write it at all. No stubs. No skeletons. No half-built abstractions waiting for a future version of me to finish them.

If the requirements are clear — build the real thing. If they're not clear enough to build — clarify them first.

Scaffolding feels productive. It isn't.

buildinpublicsoftwareengineeringprogramming

Originally on LinkedIn

newer Mullvad MCP server — open source buildolder reel-forge — fully autonomous short-form video agent