How Kuwait and Me, a small startup that I built to spread the joy of giving, died | Hacker News Reader

How Kuwait and Me, a small startup that I built to spread the joy of giving, died

steven.pw

3 pointsenamakel3 comments

>Even if you don’t create the final product with all it’s amazing features and sparkly buttons; if you have at least something out there that works, then there’s always room for improvement. Over the period of 6 months, I’ve revised the design about 5 times, changed the entire framework twice and had to completely erase and restart the database thrice. If I had continued to improve with I had in the beginning then I could have reached much further ahead. Furthermore, if I had continued with whatever I had, it would still have been better than nothing.

I'm not sure he articulated this point well. But if I understand what he is saying, he should have improved on what he had not recoded the database interface and the framework.

I would like to understand what drove him to redo so many pieces instead of incrementally improve the system.

Yes, you're right.

When I was made the first version I used PHP with codeigniter and used Backbone on the frontend. Which was okay for the beginning but I convinced myself that it wouldn't scale well in the long run. After reading about Node.js and it's advantages on scalability and performance over PHP, I switched the backend to use MongoDB and Node. And this happened over and over again, on many different things.

Basically, I was convinced that switching harder technologies and redesigning the UI should be done immediately and will help in the long run, because continuing with what I have now will prove difficult to change in the future. I didn't looked at it as a product but rather I looked at it technically (performance mainly).

But that was wrong, as I realized later on (in the hard way, when I ended up with big complicated mess).

I also want to blame my inner-desire to change everything; often I get bored with things looking the same way and hence I end up redesigning the entire UX just to keep myself content. (Which is a horrible thing to do when it comes to going to production; I wonder if anyone has been in the same boat as me?)

I don't do UI code but backend systems. Basically ETL and Middleware adapters and the like. I have the tendency to want to keep working the code to make it better. I have however found that my code usually works well long before the I get to the stop fiddling point. I don't work in a team that supports my code, it is just myself. So it is all about my own meter to gauge when to stop tweaking.

Reply on news.ycombinator.com