Hacker News new | past | comments | ask | show | jobs | submit login

Started doing the Jetpack Compose tutorial(s) for like the 3rd time and thought: "Everything changes, but nothing changes. I am paid to stay atop this schizophrenic elephant of an industry." Then took another gulp of ItsGonnaBeBetterThisTime KoolAid.



We've implemented Compose in production for a few months now. It's actually been my favorite new thing.

Most of the new stuff increases the time it takes to get work done, but Compose appears to reduce it by half on simple tasks, and a lot more on complex tasks. There's also some stuff on View tech debt which isn't less apparent, but I think Compose is here to stay.

The official tutorials are quite bad though. Probably better to try building something, paying for a book, or just reverse engineering something off GitHub.


In some ways I like it already as well. Doing things in code instead of XML is nice. My real struggle is knowing what different citizens of the Android lineage are are "current".

Views and ViewGroups are kinda out. But ViewModels are still Good. I have no idea about Fragments and Activities and Intents. It's like every class in the system needs one of 3 badges: "Still a Club Member", "Stinky Poo Poo, DONT use this anymore", "Use if you have to, but we're hoping to make this go away some day, so keep your eyes out here."


I hear you. The badges actually exist, but they're not very useful.

Material Design is way up in the club member territory, but my gut feeling is they're not very committed to. On the other hand, they have @ExperimentalApi slapped all over things like Compose and Flow but those are probably here to stay.

Flutter seems like it's dangling close to the chopping board. I feel sorry for all the people who think that it's going to stay because Google.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: