Getting started today

I used to build sites in PHP & HTML. I’d deploy them to a web server over FTP and that was it. But starting out today feels paradoxically both harder and easier than ever before.

On one hand, the barriers to entry have never been lower. We can spin up compute anywhere in the world in milliseconds, deploy globally with a few clicks, and scale elastically to meet demand. Gone are the days when a local power grid issue or a strict hosting quota could take down your entire site.

But at the same time, there’s so much more to consider. I remember when I could just write a page in Adobe (then Macromedia) Dreamweaver, slap together some frame sets and FTP it onto a server.

That was it. No complex build systems. No worrying about accessibility standards like WCAG, no concerns about intricate security vulnerabilities like supply chain attacks (malicious dependencies in npm packages) or side-channel exploits (like timing attacks). We didn’t have to think about GDPR, CCPA, or a dozen other compliance frameworks. There was no endless debate about frontend frameworks like React, Vue, or Angular, API rate limits, container orchestration, or zero-trust security models. Security used to be an afterthought; now we have to worry about things like CSRF, XSS, SQL injection, and protecting against DDoS attacks.

We could just build. It was a time when we could experiment without consequence, throwing together ideas just to see what worked. We didn’t need to justify every project with a business case or appease to critics on Hacker News asking how it would fit into a “long-term monetisation strategy”. We built things for the sake of building, for the joy of learning and creating. Is that still possible today? Or are we so caught in the race for efficiency and purpose that we’ve lost that freedom? Everything now needs a defined audience, a roadmap, a go-to-market strategy. But the best innovations often come from tinkering without a plan.

And beyond that, we tend to overcomplicate things now. We over-engineer solutions before we even understand the problem we’re solving. Why do we jump straight to microservice orchestration before we’ve even built a monolith? I used to love having lots of disparate microservices, each handling a specific task, but now I see it as a burden. The simplicity of a single codebase is underrated. We’re so eager to follow best practices that we sometimes forget to just make something work first. It’s like we’re optimising for scale before we’ve proven there’s anything worth scaling.

Now, the things we need to think about are often managed for us. Cloudflare, Vercel and Netlify make deploying stuff online almost effortless, but if they ever discontinue their free tier or make a major policy change, many of us would be left scrambling. Usage-based pricing makes scaling easier but sometimes feels disconnected from the value actually delivered. The convenience is incredible, but it comes at a cost—both financial and in terms of reliance on centralised services.

Of course, modern web development provides immense benefits. Tools like React and Next.js allow us to build incredibly powerful and interactive applications. Cloud platforms provide robust infrastructure and scalability, and the focus on security and accessibility has made the web a more reliable and inclusive place.

I still believe in the value of being able to run things at home. I’d love for ISPs to offer a second IP just for inbound traffic, enabling people to self-host without the convoluted workarounds of today. We should be fostering innovation, not making it harder for individuals to experiment and build new things. The more we encourage people to tinker, to create, and to push the boundaries of what’s possible, the better we all become.

We’ve come a long way from Web 1.0, but let’s make sure we’re still leaving space for the next generation of builders to start from scratch and learn by doing—just like we did. Not following a tutorial and rinsing and repeating, but truly understanding the fundamentals and building on top of them. That’s how we’ll keep the web vibrant and exciting for years to come. Perhaps, we should all try to build a small, personal project this weekend, without worrying about best practices or monetisation, just for the joy of creating.