Midnight Pub

Building, not tinkering


How do you actually build things and finish them?

I finally got a startup idea worth considering, although I feel overwhelmed by realization of amount of work and time it would require to build an MVP. I know some very experienced business and sales people who would help me with making it public, bringing first users and monetizing it, but first I need something viable to show them. And that's the hardest part.

Thinking of complex filters that possibly would require me to use something as sophisticated as Neo4j (although thinking more deeply says that PostgreSQL would be enough) is killing me. Even SQLite would be good enough for months or even years ahead, then why bother with managing a database server? But, still, I don't even know Neo4j, so I can't be sure it'd fit there. But I don't know enough of PostgreSQL as well, so I can't be sure it'd help me.

Thinking of complex features in general is killing me. I know I should start small and then grow around it, taking one step at a time, otherwise I'm not getting anywhere.

But the reason I'm stuttering there is that I'm afraid to make a step that'd make following steps harder. Like excluding the possibility of scaling and extending just by the design decision I made in the beginning.

I've started a shitload of various projects and libraries and applications, but very, very few of them got to the point they're usable and implement the general idea. It’s not a completely bad thing though. Some of these projects, despite never being finished, taught me a thing or two. But that’s the off topic.

So the question is towards senior software engineers, that have experience with building things that grow with time: how to take a single, small step and don't be afraid of steep climbs that are ahead of me?

At this point I'm trying to take a small step, but I'm already thinking of things that will or even only *might* be ahead of me, take them into account and making this small step way too sophisticated to go anywhere.

Write a reply


~euromancer wrote:

Ironically, somewhere in the README.md of my project, I wrote:

An hour of actual coding yields more results than a week of planning!

At least it's good to know that I already have sales, business and law people in sight and reach.

Proxied content from gemini://midnight.pub/posts/597.
Get a proper gemini browser and visit!
merveilles webring (external content)

Gemini request details:

Original URL
Status code
Proxied by

Be advised that no attempt was made to verify the remote SSL certificate.