Hello everyone, I hope you are well.
It's been a week since my last post and a lot has happened, including a huge reboot.
Indeed, as an unsatisfied insider, I found that my application was not well structured enough and I had concerns about its stability and my ability to make it more optimal and scalable with many users.
That's why I decided to do a ...
I don't know about you, but that's how I see my app.
Third reboot
Okay, so a new reboot, but for what purpose and what changes? Don't move, I'll explain everything. Be careful, it's going to sting a bit, it's a technique.
A real architecture from the start.
In my previous versions, I started by developing the application logic, then I wanted to integrate with the various services, servers, and databases. But this caused several architecture and code organization problems that were not optimized at all.
Here, we do the exact opposite. I started by focusing on the backend hosting and the fact that the service had to run 24/7, without interruption, even when no user is connected. To do this, I chose RailWay. It is efficient, a database is directly integrated, and the pricing will allow me to start small, without too many financial constraints.
Well-thought-out dependencies.
This also allowed me to completely rethink the dependencies I wanted to rely on. My professional experience taught me that there were several important points to respect. First, are they compromised? This is the basis of security, especially when we are talking about private active keys stored in the database. Even if they are hashed, there should not be a leak due to a poorly maintained dependency or with known vulnerabilities.
It will certainly not be compatible with a whole bunch of the latest versions of the dependencies I want to use, including main services like Node. As a result, we are even more targeted at pirates.
The second thing is, is it still maintained? Have there been recent releases? Is the community around it tight-knit? Is it talked about? It's really important if I want to have an application that will last over time.
The last point is not the least. I have done my utmost to limit dependencies internally. Always a better practice, even if you should avoid reinventing the wheel.
A solid architecture.
The second point is the separation of responsibilities pushed quite far.
Each service of my application will be responsible for a specific task. For example, scanning the market to search for deals, updating a price, transferring a detected and assigned deal to another table that will allow it to be processed in the user's dashboard.
It's much easier to maintain, much easier to evolve, and especially easier to debug.
Speaking of which, and I'm sorry, it's a bit technical, but the entirety of the involvement will be covered by unit tests and even integration tests, in order to secure it as much as possible and not break anything, so that you have the best user experience once the product is finished.
Knowing how to take your time to deliver quality.
At the time of the Vibe coding, it is tempting to deliver an application in a week, or even less, and iterate on it to improve it. However, this system has its limits, and the first version of the application has proven it.
Here I choose to take my time. I think there will still be a good week of development before the backend is considered finished for a first beta test. And after that, I will start introducing the frontend and so the first beta testers will arrive.
Speaking of that, if you are interested, it is clearly the time to show your support. I already have two people certified in the team who will do the first beta tests and will have the chance to give me feedback and certainly have free access to the application when it is released.
In conclusion
The application has never been as solid and well thought out. This allows me to see many cases that could have occurred and caused problems, and that would have ruined your user experience and possibly your trust. Period. Instead, we will start with something solid from the beginning, and just as I am focusing today on the backend, I plan to offer you a frontend with small onions, with a hyper-pleasing user experience.
Stay online, things are going to change soon. See you soon and thank you for following me on this adventure.