lotdrops
@lotdrops.bsky.social
Android dev, in love with Kotlin and KMP. I love UX and dev X
While it's true that cost is driving this decision, the ratio is at best 3 to 4 engineers, not 1 to 2.
At a significant cost in user satisfaction and engagement, it may be not as luxurious as it seems
At a significant cost in user satisfaction and engagement, it may be not as luxurious as it seems
November 4, 2025 at 7:45 AM
While it's true that cost is driving this decision, the ratio is at best 3 to 4 engineers, not 1 to 2.
At a significant cost in user satisfaction and engagement, it may be not as luxurious as it seems
At a significant cost in user satisfaction and engagement, it may be not as luxurious as it seems
I introduced the culture of pairing to my team and I think it's helped a lot in learning and in solving complex problems.
But it hasn't changed the mindset of management that pairing=1/2 productivity.
So there are occasional comments about too much pairing, and we're still hesitant about doing it.
But it hasn't changed the mindset of management that pairing=1/2 productivity.
So there are occasional comments about too much pairing, and we're still hesitant about doing it.
November 3, 2025 at 11:06 AM
I introduced the culture of pairing to my team and I think it's helped a lot in learning and in solving complex problems.
But it hasn't changed the mindset of management that pairing=1/2 productivity.
So there are occasional comments about too much pairing, and we're still hesitant about doing it.
But it hasn't changed the mindset of management that pairing=1/2 productivity.
So there are occasional comments about too much pairing, and we're still hesitant about doing it.
Thanks to you samples, I've written several (medium size) kmp apps with Koin over the past few years, and I agree with this statement.
With all the hype around metro I've been thinking whether to try it.
Have you tried it?
With all the hype around metro I've been thinking whether to try it.
Have you tried it?
October 19, 2025 at 9:24 AM
Thanks to you samples, I've written several (medium size) kmp apps with Koin over the past few years, and I agree with this statement.
With all the hype around metro I've been thinking whether to try it.
Have you tried it?
With all the hype around metro I've been thinking whether to try it.
Have you tried it?
KMP permite hacer apps nativas en Android y iOS, tanto en rendimiento como en ux.
Con menos bugs y coste.
Con menos bugs y coste.
October 17, 2025 at 5:40 PM
KMP permite hacer apps nativas en Android y iOS, tanto en rendimiento como en ux.
Con menos bugs y coste.
Con menos bugs y coste.
Thank you, you've kept me from trying to delete and reconfigure everything!
It's been months of this happening to me, on Mac too. I thought it was only me as it does not happen to my colleagues, plus it was not being fixed for so long
It's been months of this happening to me, on Mac too. I thought it was only me as it does not happen to my colleagues, plus it was not being fixed for so long
September 11, 2025 at 8:54 PM
Thank you, you've kept me from trying to delete and reconfigure everything!
It's been months of this happening to me, on Mac too. I thought it was only me as it does not happen to my colleagues, plus it was not being fixed for so long
It's been months of this happening to me, on Mac too. I thought it was only me as it does not happen to my colleagues, plus it was not being fixed for so long
Great write up, as always.
If this is implemented, I see it as one big step towards the web killing mobile development.
I already saw it as a possible future, but with big barriers like this it becomes quite likely, I think.
If this is implemented, I see it as one big step towards the web killing mobile development.
I already saw it as a possible future, but with big barriers like this it becomes quite likely, I think.
August 27, 2025 at 8:58 AM
Great write up, as always.
If this is implemented, I see it as one big step towards the web killing mobile development.
I already saw it as a possible future, but with big barriers like this it becomes quite likely, I think.
If this is implemented, I see it as one big step towards the web killing mobile development.
I already saw it as a possible future, but with big barriers like this it becomes quite likely, I think.
With a lot of people saying it is super productive, I keep trying it, trying to understand if I'm doing something wrong or it is like you described.
I think this recent research may explain why many find it so productive bsky.app/profile/metr...
I think this recent research may explain why many find it so productive bsky.app/profile/metr...
We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers.
The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.
The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.
July 21, 2025 at 6:36 AM
With a lot of people saying it is super productive, I keep trying it, trying to understand if I'm doing something wrong or it is like you described.
I think this recent research may explain why many find it so productive bsky.app/profile/metr...
I think this recent research may explain why many find it so productive bsky.app/profile/metr...
How does being partially covered or visible but not "resumed" (e.g. a pop-up) affect this API?
June 24, 2025 at 1:10 PM
How does being partially covered or visible but not "resumed" (e.g. a pop-up) affect this API?
Extremely annoyed by something similar in Google maps: it auto-translates all reviews... Which costs money to Google, and is much worse UX if you happen to speak the original language.
June 21, 2025 at 1:39 PM
Extremely annoyed by something similar in Google maps: it auto-translates all reviews... Which costs money to Google, and is much worse UX if you happen to speak the original language.
Middle-man abstractions.
Like adding use cases with no logic, just to follow the same pattern everywhere.
Like adding use cases with no logic, just to follow the same pattern everywhere.
May 25, 2025 at 7:00 AM
Middle-man abstractions.
Like adding use cases with no logic, just to follow the same pattern everywhere.
Like adding use cases with no logic, just to follow the same pattern everywhere.
It makes sense, they may feel like they lose control and also job security.
At the same time, it makes A LOT of sense from the technical and business perspectives, IMO.
I don't see a reason for doing pure native apps anymore, actually.
At the same time, it makes A LOT of sense from the technical and business perspectives, IMO.
I don't see a reason for doing pure native apps anymore, actually.
May 13, 2025 at 6:17 AM
It makes sense, they may feel like they lose control and also job security.
At the same time, it makes A LOT of sense from the technical and business perspectives, IMO.
I don't see a reason for doing pure native apps anymore, actually.
At the same time, it makes A LOT of sense from the technical and business perspectives, IMO.
I don't see a reason for doing pure native apps anymore, actually.
1, to enforce their rules, which often benefit them
2, to make it look like they provide high quality to the end user
3, to justify charging money to devs and users
2, to make it look like they provide high quality to the end user
3, to justify charging money to devs and users
May 8, 2025 at 5:15 PM
1, to enforce their rules, which often benefit them
2, to make it look like they provide high quality to the end user
3, to justify charging money to devs and users
2, to make it look like they provide high quality to the end user
3, to justify charging money to devs and users
I've seen a few argue in favor of TDD (not an arch, but related)
March 24, 2025 at 3:23 PM
I've seen a few argue in favor of TDD (not an arch, but related)
Yes, it's quite different to jetpack nav, for example. I guess I've been burned so many times by jetpack nav issues that I was OK with trying something very different. But it does take some time figuring out how it works.
February 9, 2025 at 8:45 AM
Yes, it's quite different to jetpack nav, for example. I guess I've been burned so many times by jetpack nav issues that I was OK with trying something very different. But it does take some time figuring out how it works.
I looked into the different options a couple months ago and decompose was the more complete in terms of supporting ios features (such as back gesture, animation, tab bar, etc).
I like it's state model, but it is a bit harder to get started than other options.
I like it's state model, but it is a bit harder to get started than other options.
February 8, 2025 at 3:55 PM
I looked into the different options a couple months ago and decompose was the more complete in terms of supporting ios features (such as back gesture, animation, tab bar, etc).
I like it's state model, but it is a bit harder to get started than other options.
I like it's state model, but it is a bit harder to get started than other options.
Pues te sugiero este libro si no lo conoces. El foco es en programación funcional, pero es con kotlin, con arquitectura Hexagonal y una forma de testear bastante interesante pragprog.com/titles/uboop...
From Objects to Functions
Learn Functional Programming by building a complete web application that uses Kotlin, TDD with end-to-end tests, and CQRS and Event Sourcing microservices architecture.
pragprog.com
January 15, 2025 at 10:10 AM
Pues te sugiero este libro si no lo conoces. El foco es en programación funcional, pero es con kotlin, con arquitectura Hexagonal y una forma de testear bastante interesante pragprog.com/titles/uboop...
Or tagging components for screen readers is quick, but trying it all as a user to make sure you didn't miss anything and the order and grouping are good... That's slower
January 4, 2025 at 8:03 AM
Or tagging components for screen readers is quick, but trying it all as a user to make sure you didn't miss anything and the order and grouping are good... That's slower
Thanks for the answer! This makes sense, but what about those cases where it does take significant time? E.g. Suporting larger fonts is worthit at the start, but doing it for extreme cases (very large font, small screen) where you need a specific layout for these cases seems less straightforward.
January 4, 2025 at 8:01 AM
Thanks for the answer! This makes sense, but what about those cases where it does take significant time? E.g. Suporting larger fonts is worthit at the start, but doing it for extreme cases (very large font, small screen) where you need a specific layout for these cases seems less straightforward.
What about accessibility? It seems to be a task you should wait on, thinking in terms of agile. But there are ethical reasons for supporting it early (and in some cases legal reasons too).
January 3, 2025 at 8:19 AM
What about accessibility? It seems to be a task you should wait on, thinking in terms of agile. But there are ethical reasons for supporting it early (and in some cases legal reasons too).
I remember this TIL moment too.
I couldn't understand why was I getting those crashes if I had tested config changes, until the big "oh" came.
I couldn't understand why was I getting those crashes if I had tested config changes, until the big "oh" came.
December 6, 2024 at 9:11 AM
I remember this TIL moment too.
I couldn't understand why was I getting those crashes if I had tested config changes, until the big "oh" came.
I couldn't understand why was I getting those crashes if I had tested config changes, until the big "oh" came.