Just remember to copy font to your own repo in case that Github user finds it funny it replace font with something else 🙂

so it’s actually harder to read. Anyone else seeing this?

kotlin uses both syntaxes

operator fun <T, U> T.invoke(x: U) = ":)"

intellij with a java file open that has python code in it

still playing with it but they will start with kotlin support tonight 😄

Re learning programming with Kotlin, I was teaching my son the basics with Kotlin, it was really smooth. You can start with simple procedural stuff with

println

, no complex graphics, no objects, no functions. First teach decomposition with functions, then introduce a little bit of data classes without any methods, etc. The most complex thing to teach was actually not something specific to Kotlin, it was “memory model”, i.e. how data is represented inside a computer memory, the concept of a “reference”, the concept of a type of a memory data (even if as simple as String or data class), etc. YMMV, but I believe Kotlin is really good for teaching.

Having more programmers shouldn't hurt, and it's not because you tried that it'll become your day job

I wonder if someone can approach http://www.android-ide.com/ and suggest adding Kotlin

@johnl Don't you think lowering the entering barrier to programming is a good idea?

@johnl ~Exactly, it is the wrong Spark. Why use the same name as such a widely popular computational framework. ~ I am not sure what logging dependencies you have had a problem with. Spring Boot on its own really is very tiny and simple but that is kind of the point. The real power comes when you start adding `starter`s from the Spring Framework. For example for making oAuth work or SSO or Redis. It is really a matter of seconds.

@trevjones Yes I know they are using mockito. I did not knew that it was possible without Roboelectric though

out of curiosity, has anyone seen a discussion about why the JVM needs to keep all `Thread`'s stack memory size the same? I feel like an awful lot of performance problems could be solved if the JVM was allowed to operate with differently-sized thread stacks. For threads that simply call blocking API's and dump the result into a

Channel

(or other coroutine/threading device), I would really appreciate the ability to tell the JVM that this thing just needs maybe 8Kb of stack.

An Rx question: I"ve just migrated to Rx 2 and only just realised I can no longer have

null

elements in streams. In cases where

null

is notionally valid; such as a stream to fetch some cached user data, where there may be none... is it advisable to model these as exceptions

NoUserDataException

?

Sigh Java 9 is becoming Duke nukem forever

wow such canine 🐕

Immutablility is a property of an object. Writablility is a property of property.

okay_face.jpg

you’re right

Long story short,

val

implies referential immutability, meaning it can't be reassigned. It has nothing to do with the mutability of object itself. That's up to the object (more specifically it's class definition) to provide. (This definition is only true for properties with backing fields)

yep, if Scala API is “scala-only”, you will get something like this: https://github.com/orangy/http-benchmarks/blob/master/src/org/kotlin/community/http/benchmarks/colossus/ColossusBenchmark.kt

Kotlin just deployed to production for the first time 🎊

We have 99.44% java and 0.56% kotlin 😂

I implemented something that follows quite closely the spirit of dagger2, but that could well be even better for simple use cases. Often the Testcomponent needs to provide more types than Component, that's when the dagger2 pattern is more powerful

I would love www.codingame.com to support Kotlin. Codingame let's you program AI bots and practice machine learning. Way more fun than keep practicing graph theory and dynamic programming.

Disappointed by the type system of your favorite language? Things could be worse:

What's the simplest way to turn a consecutive

Sequence<T>

into a

Sequence<List<T>>

, where each

List<T>

is a consecutive batch of 10

T

items?

At the fear of turning this into stackoverflow... Anyone know why the call mapType.keyType() does not compile here? https://git.io/v9VFY

you put milk in your coffee?

And you're about to have two.