Do "time-saving" tools waste your time?
Developers are constantly told to use tool x or tool y because it automates stuff and lets you focus on domain problems. But in my experience all that happens is I end up with an over-engineered tool which means I spend most of my time reading their documentation and stackoverflow instead.
I'll name names: react.js, docker, AWS, google sign-in have all been more work than ... not using them. Is there an impedance mismatch using 'enterprise' tools for solo projects, or are other people just better yak shavers than me?
Trending on Indie Hackers
I think there’s always an added layer of complexity when trying to combine tools that weren’t specifically designed to work perfectly together.
Example: Firebase is supposed to be super easy, but if you’re using it React you probably want some kind of context/useAuth hook.. That takes time to build. Or what happens when the call to user.updateEmail fails because Firebase requires the user to have logged in recently for security sensitive actions. Now you need to build a reauth flow of some kind. The list goes on and on.
In my mind the solutions are either:
I think it just depends on your needs. Problems occur when people use tools that are overly complex for their use case.
I use and love React. It would be overkill for a very simple app. But for my medium-sized app it's indispensable and has absolutely saved me time, and I think the productivity benefits will keep growing as the app grows more complex.
I'm also using Firebase which has simplified my backend considerably. Have I had to spend time learning Firebase and reading the documentation? Yes. But I think it has saved me considerable time compared to the alternative which would look something like using Auth0 + running my own node.js server on AWS with MongoDB using S3 for file storage (I don't know if that stack even makes sense, I just have a superficial understanding of those technologies). Firebase's API is simple, and its documentation is great. I've heard lots of stories of how hard Auth0 is to implement. Firebase Auth with email took me about an hour to implement, mostly because I was learning how to do it. Now that I know, it would take me about 5-10 minutes the next time I need to do it. If I wanted to add Google Sign-In, I think it would probably take me another 30 minutes to an hour.
Yes and no...
For any tool, there's a return-on-investment to think about, and I think it boils down to that. For example, I'm just messing around right now, but one of my current projects is a desktop client for a distributed social network that I like, but wasn't really designed with the "social" part in mind. If I spent the time setting up a continuous-integration server for it, that time would be wasted, because I'm probably not going to spend more than forty hours hacking on this thing and it'd probably take me a significant percentage of that to set up the server.
However, if I was still running my automated project management system from ten years ago, I was putting in enough work on that and it was bringing in enough money that it would justify the investment in build and deployment systems. If I was still selling it as a product, it'd also be well worth the time packing it into a container to simplify installation.
Similarly, now that I've invested the time in learning React, it saves me a fair amount of time, even on smaller projects (but not the smallest projects), whereas someone who hasn't might not see any benefits for a long time.