Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
nafg
@nafg
Honestly sbt's implementation of ~ could use some improvements too
Greg Methvin
@gmethvin
I don’t think it helps that play has its own ways of doing things
nafg
@nafg
While I'm making a wishlist, I would also put on good Mill support. At one point I would have said Seed too...
Greg Methvin
@gmethvin
well, you can run a play server as a regular scala app without the dev mode stuff
But I know what you mean
honestly I just haven’t had much time lately to work on stuff
the real issue is that Lightbend isn’t able to devote as much developer time to it
nafg
@nafg
Yeah so I think running it as a regular scala app, under sbt-revolver, would be a good experience
Is there a straightforward recipe for an existing play app?
Greg Methvin
@gmethvin
I don’t think that’s necessarily a bad decision for them and I’m glad they contributed what they did so far, but it means the community has to pick up some of this work
It’s in the docs i know
nafg
@nafg
Meaning to convert an existing play app
is it? maybe I'll look it up sometime then
Greg Methvin
@gmethvin
oh
https://www.playframework.com/documentation/2.8.x/ScalaEmbeddingPlayAkkaHttp
nafg
@nafg
Yeah it's unfortunate lightbend priorities changed, but they need to at least do a better job of encouraging community involvement
Greg Methvin
@gmethvin
well you can still run the routes compiler and call the router there
nafg
@nafg
i don't have the head right now to start figuring out what that involves. I have an app that's built as the usual play app (with scala trait DI), would be great if there was a recipe written up...
if there isn't then it will stay for now...
Greg Methvin
@gmethvin
I thought we had a more complex example
nafg
@nafg
Same thing with Slick, they should have done a better job of encouraging the community and seamlessly transitioning to it being community run
Also I think every project needs leadership, even if that leadership is extremely hands off and just there to settle tough decisions or set a vision
I still prefer Slick over Quill for the most part. Maybe one day I'll have time to make some changes to it. But it's so poorly documented
Greg Methvin
@gmethvin
this one shows how to use the routes compiler when using play as a library https://github.com/gmethvin/play-routes-standalone-example
nafg
@nafg
Maybe dotty quill will obviate the need for slick
Greg Methvin
@gmethvin
it was just something I was testing out
yeah, the right path is to modularize stuff so it’s easy to use other libraries with it
nafg
@nafg
thanks, I starred it, but I'm not going to change my project right now if it needs me to start digging into stuff, I starred it to find it in case I feel like digging in another time
Greg Methvin
@gmethvin
I think the problem is that Play originally wanted to be a “framework” but Scala has so many different libraries for doing stuff
so it’s hard to be opinionated about things
nafg
@nafg
It's a really great example of how a framework can be very modular and not have the issues usually associated with frameworks
Greg Methvin
@gmethvin
I agree about Guice. I think we should’ve done regular Scala DI
but the fact that Play also has to support Java also made things more complicated
nafg
@nafg
I guess that makes maintenance more work because of combinatorial explosion
Yeah
I mean they could have taken the approach that Play is a Scala framework with a Java API on top to ease into it (with the idea that for more advanced stuff you can get into scala)
To maintain a first-class Java API is much more of a burden
Greg Methvin
@gmethvin
yeah, but the reality is that right now a lot of Play’s users use Java
LinkedIn for example. Last I checked they were trying to remove most of the Scala from their codebase.
I still think Play works well for Scala, especially if you can strip it down to its essential components
nafg
@nafg
Yes, it's still great
it's just disheartening that there isn't anything interesting coming down the pipe
Greg Methvin
@gmethvin
I’m trying to convince my company that we need to spend more time on developer experience. The quirks of play dev mode are a common complaint, so I might end up making some efforts there, maybe
nafg
@nafg
That would be amazing
Greg Methvin
@gmethvin
We might just decide to rip out the play plugin like I was saying
also we’re using hydra so there might be some quirks there, idk
Igmar Palsenberg
@igmar
Scala is just a pain to learn to be honest.
Matthias Kurz
@mkurz

@nafg

They are, I just get cryptic errors when something fails

Here is a fix for that: playframework/playframework#10648

nafg
@nafg
@igmar so are typescript, rust, and c#, and many other languages. The question you should ask is whether the benefits make it worthwhile
4 replies
N.S. Cutler
@godenji

In the past it's taken quite awhile for Play to catchup with latest Scala releases; is this going to be the case (and then some) with Scala 3? I understand that play has many dependencies which makes Scala 3 support challenging.

Starting to prototype an app using Scala 3 RC1, would be great to be able to integrate Play in the not too distant future. Barring that I assume we can cross compile and pull in Play 2.8.x + Scala 2.13.