Finishing the MVP

I really enjoy building, getting stuck, and then unstuck. That last part is the best. It's enjoyable, challenging and absolutely exhilarating to see the goal you were trying to achieve. It's crazy to think that two weeks ago, this was just a curiosity. I was looking at an analytics dashboard and I was like, "I wonder how it all works?." Then I was like, "Can I build it?", I think I've answered my questions, yes I can. It is not as easy as I thought, and the parts that I thought would be the hardest, turned out to be not that hard.

The hard disguised as easy

This one I was immediately sure, on how to do it. Send a request when a user enters the site, send one when they exit the page. Easy, right? It wasn't, what I forgot to take into account was that browser cancels all outstanding requests when the code is unmounted, ex: user closes the tab. Maybe this was a skill issue, but I've never had to worry about a request failing post code being unmounted.

Getting Unstuck

It's hard when you don't have the right vocabulary to explain the problem to a search engine or AI. Took me a couple of days, as I benched this issue to the side to tackle it later. That was the smartest move on my part. This I've done before, I've often pivoted to a task that I know I can do, when I'm stuck on something I have no clue about.

When I did return to this task, my mindset was different. I was no longer using the same words to describe my problem. My search provided better results and eventually gipty brought to my attention about an API that does exactly this, it was called navigator.sendBeacon() (opens in a new tab). Although, I did not end up using this API, as I read the documentation I figured out how I could do this by updating my request with one simple parameter.

This parameter was called keepalive, which then would tell the browser that this request should be kept alive, when page looses visibility. As I'm writing this, I've realized I've not handled the scenario where the page is visible again, i.e. uses switches tabs and returns. One more item for the shipping tasks (opens in a new tab).

What's next?

The goal was to build something that would satisfy my curiosity and be useful to at least one more person. You see, programmers are great at building things, not so great at selling them. That's the other thing I'm trying to learn. To sell my software. To be sustainable and be able to build the things you want, when you want, as you want, you need money. You can either get this money from the outside or you can learn to sell. I like the freedom of building things as I want, dictating how it should work, what it does and what it will never do. Like violate privacy of it's users in the case of ioIgnition.

The goal when building ioigntion was to build a privacy first, minimal analytics tool that does not slow down your webpage. Currently the script is at 3.8kb, without any optimizations.

As for what is next, I'll be working on the finishing touches and completing the items on the list above and then talking about this to people and see what they think about it.