Keep things as simple and single functionality as possible. If you can't explain your startup in one sentence, it's doing too much
I had a client around 2008-2009 (he was around 40 at the time and successfully launched several companies) who wanted a website/web app which basically was doing crowdfunding and videoconferencing (plus shared board and documents)
Yeah, I said basically and here we are getting a lot of functionality. I spare you the rest.
Needless to say after 2 years and a half of discussion and meetings and « what about we add that functionality ? », money ran dry and the whole thing never launched.
These days I might have an idea for a feature in my head and pitch it to people as if it exists to gauge interest. If it seems like it's a clear winner I build it into my product. If not, I'm not writing any code.
I don’t discuss which DB to use. I straight away use SQL.
That is a language though, not a DB.TiDB (distributed)
CockroachDB (distributed)
Crate (uses SQL but isn't even relational)
They all use SQL.
Picking between these and more established SQL databases is worthy of discussion.
Heck, based on what I see on Hacker News, picking between the established ones generates a lot of discussion, usually with people preferring Postgres over MySQL, etc.
Then there's SQLite, which is completely different again.
OP probably has a favourite database as a go-to regardless of the application, but for some projects SQLite might be a better fit than the others, for example. That would warrant discussion.
In conclusion, I believe you were too harsh to TekMol.
Edit: Fix autocorrect fails and add clarity.