Replit 4+

I’ve been poking around this app despite not knowing much about coding, and found the interface to be surprisingly intuitive. Within minutes I created a nice looking spoof website from an open-source template, and for the most part it’s obvious what sort of things I can do. I would really appreciate a tutorial for beginners or at least some indication of what more technical words mean (what is “repls”? What does “shell” do?), but I understand that I’m not the target audience here and most who use the app will already know such things. Regardless, for the most part I’m finding the experience to be quite intuitive. Kudos to the developers for making a coding interface so friendly to newcomers!

I've been using Replit for launching Minimum Viable Products (MVPs), and I must say, it's fantastic for that purpose. The ease of setting up, coding, and instantly seeing your project run in the browser is unparalleled, especially for quick prototyping or when you want to share your work with others without any setup on their end.

However, as projects grow in complexity, Replit starts showing its limitations. For more sophisticated features, deeper system integration, or when you need more control over the environment (like specific versions of libraries or custom configurations), Replit might not be the best standalone solution. In these cases, combining Replit with more robust development tools becomes necessary for managing complex applications effectively.

In conclusion, Replit is an excellent tool for getting your MVP off the ground, for scaling and intricate feature development, you'll likely need to pair it with more advanced development environments. Still, for what it does, Replit is an invaluable asset in any developer's toolkit.

I’ve built two apps so far with success. Vibe coding my way through it, however, the agent makes so many mistakes through simply forgetting or making changes you don’t ask for that you end up losing so much of your allotment of credit. It’s frustrating. What I want to see is an adaptive agent that refunds you when it screws up, and is clearly not your fault. An example is Manus AI. If it screws up, it refunds you. Everything. You’re wasting credits correcting the agent. It you get the assistant involved mid build, it crashes the application. Lovable 2.0 just released an ability to select your edits and manually make changes. This is a huge need. They also gave you a separate chat option to plan. I’ve been planning with CHAT GPT then bringing over my details and it still gets it wrong. I actually prefer Replit to everything else because it builds a full stack app. But the UX needs a ton of work.