Spending 5 minutes playing with the voice-activated spectrogram

As you can see :)

With @OnlyInputTypes the message from the compiler wasn't good enough for the mistmatch plugin to work, but with @Exact it does!

Error reporting is going to get much better soon 👀

But the rollout is *so weird.* Chrome desktop wants my phone screen PIN *on the desktop*. I am not typing the password for unlocking my device in any other input form that the one on that device. These kinds of passwords shouldn't reach the net.

Build tools are redundant, ask Junie to output the JAR that corresponds to your code

Yes.

Indeed, if the clients already contained the Tailscale lib, that would be convenient.

That would require a tailscale-js lib though. I don't think the Chromecast can load native libraries.

Jellyfin works great behind a caddy reserve proxy + the tailscale plugin

Haven't figured out casting yet though

INDUSTRIAL ESPIONNAGE

I'm getting a 404

Prototyping a new design for overloads…

OpenSavvy Spine Share your Ktor endpoints between client and server

Looks like Spine will be able to upgrade to Kotlin 2.4.0…

Since 2.4.0 also brings stable context parameters, Spine will probably go 1.0 shortly after. If you have any feedback, now's the time!

spine.opensavvy.dev

Also: branch coverage is much more telling than line coverage.

It's also much harder to inflate

(nevermind that KotlinJS can't tell the difference between the different number types)

KotlinJS is more typesafe than KotlinJVM confirmed!

nice!

classifier == List::class || classifier == MutableList::class

…this is pointless, right? Since they're the same interface.

I'm hearing "code coverage doesn't tell you that you don't have bugs" more and more often nowadays.

That's true! But code coverage is a very good tool to tell you which code you forgot to test.

I'm very successfully using Kover in KtMongo to find rare edge cases.

AndroidMakers :)

Right. The tool could measure how long it takes each time, then average it to get an order score.

Does nothing to help the very first time, but it should be fairly effective when upgrading to new versions?

The Pacman package manager downloads bigger files first, and it does lead to the benefits you're mentioning

The official MongoDB driver (org.bson:bson-kotlin) doesn't support Kotlin's default arguments

It's already over??

First time at @androidmakers.bsky.social!

Schedule | KotlinConf 2026, May 20–22, Munich KotlinConf is the official Kotlin conference by JetBrains. It is a place for the community to gather and discuss all things Kotlin.

The KotlinConf schedule is stacked this year!

kotlinconf.com/schedule/

Thanks! ❤️

I'm a finalist at Golden Kodee Awards, the Kotlin Foundation community-recognition award!

Next step is KotlinConf 2026, Munich 👀

In the official Bordeaux transport network ad, the tram door is out of service

And, it's the entire point of open source that it's fine if people get tired of maintaining a project, because whoever cares can fork& continue.

On one side, if it's really a single file change, then yeah it's not great as a maintainer to break it.

On the other side, I can understand that a maintainer would be tired of constant fix requests for an older version than what they use themselves in an experimental project.

It's the default behavior for block comments, but not for line comments