When Too Many Wrongs Actually Make a Right
It’s a sunny day about 5pm. I have the twins in the buggy, I’m walking around and around and around our house. Fuelled by 3.5 hours sleep last night (that’s the norm at the moment) and what feels like a vat of coffee I finally come to my decision. It’s time to build an app.
August 2025. My twins are now 3 months old. It’s a blur of feeding, changing nappies and surviving. Keeping my clinics going and finding pockets of space at night, between feeds, to code this “project”.
The dashboard embed being blocked leaves me facing up to a hard choice.
Find a way to move what I’ve built to a service I can embed and therefore support my little beta group much more quickly…
Or
Cut my teeth on trying to build my own app using Ai. Claude code is proving extremely able at writing code so that’s the only reason I’m half considering it.
It’s a sunny day about 5pm. I have the twins in the buggy, I’m walking around and around and around our house. Fuelled by 3.5 hours sleep last night (that’s the norm at the moment) and what feels like a vat of coffee I finally come to my decision.
It’s time to build an app. It’s the longer term plan anyway. I was thinking of getting some help to do this but I’m kind of at a dead end here with my dashboard embed.
My own app it is.
A Progressive Web App. An app that will work on any device and I won’t need to deal with the hassle of getting approved in app stores. Plus I will own it all - it will be my architecture, on my servers. Sorted.
And honestly, this is what I see with people in pain. When the workarounds and the short term fixes stop helping, you’re left with committing to the real plan. You leave the roundabout and take the hard road that ultimately leads somewhere.
The reality is I’ve messed up so many simple things already - purely out of not knowing. Turns out that too many wrongs might actually lead you to what’s right after all.
I reflect on the fact that a decision I made way earlier (to build with agentic design in mind) - has saved me now, as it makes it easier to start to build.
I brainstorm with a few AI systems about the best way to do this. Turns out Google Apps Script has met the end of its road. We are moving to Node.js and React/Next.js.
I write like I understand these. At the time I pull the trigger, they mean nothing more to me than letters on a page.
We begin.
Occasional updates. No spam.