Scala-native support #568
Scala-native support #568
Conversation
|
Nice! Is there a possibility of a snapshot, or should I bootleg this? Looking to unblock the fs2-data test suite. |
|
@armanbilge if you want it today you're gonna need to do a bootleg, but we'll set up snapshot publishing before the end of the week for the series/0.8 branch. |
| private[weaver] object Timestamp { | ||
|
|
||
| def format(epochSecond: Long): String = Zone { implicit zone => | ||
| val out = alloc[time.tm]() |
There was a problem hiding this comment.
I think this can be stackalloc, no?
We're not return the pointer to the caller, so we don't need it on the heap, and the return value is a Scala String which will be on the heap anyways.
There was a problem hiding this comment.
And this makes Zone unnecessary as well
There was a problem hiding this comment.
yeah fair enough, I copied it verbatim from 0.4.0 code I had laying around
|
|
||
| def localTime(hours: Int, minutes: Int, seconds: Int): Long = Zone { | ||
| implicit zone => | ||
| val out = alloc[time.tm]() |
There was a problem hiding this comment.
same as above - the return is Long, by value, so the memory can be allocated by stackalloc - this would make it much, much faster as well
| continuation: Array[Task] => Unit): Unit = { | ||
| val _ = executeFuture(eventHandler, loggers).map(_ => | ||
| continuation(Array.empty[Task]))( | ||
| scala.scalajs.concurrent.JSExecutionContext.queue) |
There was a problem hiding this comment.
Hmm, is this what we had before?
IIRC we are bringing the macrotask - do we want to use it instead? #476
There was a problem hiding this comment.
In this case it really shouldn't matter
There was a problem hiding this comment.
Hmm, is this what we had before?
Nah before we were using a callback based async method. We're unifying between JS and Native by relying on Future instead, and we're only needing an EC to plug the "end of the world" callback to the future.
| @@ -28,8 +28,8 @@ abstract class DogFood[F[_]]( | |||
| // for some time before getting the logs back. On JVM platform | |||
| // we do not need to wait, since the suite will run synchronously | |||
| private val patience: Option[FiniteDuration] = PlatformCompat.platform match { | |||
| case JS => 2.seconds.some | |||
| case JVM => none | |||
| case JS | Native => 2.seconds.some | |||
There was a problem hiding this comment.
For my own understanding - the comment above talks about SJS being asynchronous, hence the need for waiting.
Do we still need it for Native, which will also execute synchronously? Or is it something else?
There was a problem hiding this comment.
Addressed aa7296e
|
I'm just gonna go ahead and merge this, we can PR the fs2 version later on against the 0.8 branch |
Adds Scala-Native support
Waiting for this to be merged and for the release train to start (we need CE + fs2 non-RC versions) : typelevel/cats-effect#3138