Continuous deployment becomes unremarkable once it becomes routine, which is exactly the point. Happy you're living the dream! 😅

I’m with you, hope our car still lasts for a very long time. Every time I’m in a new one, I’m like: what are all those screens for? Can‘t remember anyone that was excited to get a new car recently either.

It pulls your activity data from Strava and analyzes it, showing you the progress you made in the current week and how you performed in previous weeks.

I‘m also somewhat unsatisfied by that situation. Wondering if a data hub for activity data would be a cool side project?

Haha, I feel you. It worked better in Chrome for me.

However, working with AWS for the last couple of weeks, I miss GCP. And especially Logs Explorer. CloudWatch‘s log viewer is… underwhelming at best?

First impressions after two runs: very solid shoes. Feel (and sound) a bit foamy, which is different to the On experience, but that's alright.

Love this idea!

What @jamie.schembri.me said - your talk was great and very aspiring!

This looks amazing. Very well done!

But I also acknowledge the remaining 20% which one, depending on scale, needs to tune. Maybe you're right that this then shows the gap between medium and large business.

I still think that Active Record is a great abstraction, and let's you do 80% (or even more) of the things really easily. It still has sane defaults imho, which are clearly lacking in popular packages in other languages.

I guess the issue here is that Elixir docs are just _excellent_, isn't it? 😅 Totally agree with some of the newer things, had to dig into the code to figure out how to e.g. configure Kamal - shouldn't be that way (could have opened a PR, though 😬).

Wondering if this is just how things naturally play out at scale? Some languages/frameworks are clearly better suited for this, as you said. That's probably why companies like Shopify built so many tools to manage scaling, especially on the organizational side (e.g. Sorbet and Packwerk)?

Rails though - totally agree. The rough edges show up, but in my experience a lot of the pain is less about the framework and more about “soft” factors: people skipping docs, cutting corners, not caring, etc.

Curious, are you basing this more on technical limits or on organizational scaling pains (or both)?

I can’t speak from experience with Elixir/Phoenix at all (on neither scale 😅), but it seems like its handling large apps very gracefully and just as advertised.