Mando Pi

A few days ago, on August 1st 2026, I went to a free Devin workshop/hackaton. It was delivered by Erasmo Hernandez , supported by Cognition and hosted by Xmartlabs in their Montevideo office.

First, for a couple of hours Erasmo gave us a thorough walk through Devin’s features, and he developed an app in real time as a demo.

Then came the best part: we were to develop our own idea over the course of 1h approx. I had a few ideas in the back-burner, and, on the bus ride, I had decided on the one that seemed more clear on scope. I also had a conversation with ChatGPT on whether the idea was feasible and how should it be architected.

WhatsApp went against so many startup rules.

They deliberately slowed down growth. For a while, they toggled a $1 charge on and off, not primarily to drive revenue, but to add friction and throttle signups. Their logic was that too much growth might turn their systems unstable and produce bad user experience.

While competitors hurried to add bells and whistles, in WhatsApp, most feature requests were denied and the features they actually built were tested internally and debugged for a long time, until they were polished and reliable. It took them 2 years to release video-calls, for instance.

This is if you don’t know what’s a clipboard manager.

Imagine there was a history of everything you have recently copy-pasted. Then if you needed one of those pieces of info again, you could quickly find it there. Clipboard managers make that possible.

They are available in most operating systems and their setup is usually simple. They are built in Windows 11 and Android. In Fedora I installed Gnome Clipboard History . Thanks to Alex Saveau !

I don’t know why it took me so long to realize that you can create a .gitignore with only an * in any folder, and then you don’t have to include that folder’s name in your root .gitignore. You can even rename the folder without having to touch anything else.

This is great for quick experiments or an agent scratchpad.

Example:

my-service/
└── .scratch/
    ├── .gitignore  # contains only *
    ├── prompts/
    ├── alternative-implementation/
    ├── benchmark-results/
    └── notes.md

I think you don’t see true change coming. If you can predict it, even if it’s big, it’s not a “real” change. In my experience, real change it’s very rare.

From science to software engineering

For instance, since I was a kid I thought I would be a scientist. The way I saw it, science was the noblest enterprise of humanity, the best way of giving value back to society. I went to high school and to college and then to work as a professional scientist, always maintaining that very same belief. Hence, even though my life changed a lot through all these stages, it wasn’t “real” change, because it was all predictably part of the same plan.