Hacker News Reader

Comment by jowiar | Hacker News Reader

My biggest challenge with (learning) Scala and the Scala community is that it feels like nobody is actually shipping software, or advocating for making shipping software easier.

And that's a shame, because Scala is language I love for writing a few dozen lines of code, and expressing a domain concept. It's a fantastic language for solving a homework problem set (and it's origins may be part of the reason why). But every library feels like it was concocted as a search for purity with zero attention given to the user experience.

The community is badly missing a DHH or Kenneth Reitz -- someone saying: "Ok, it's cool that we have this academic purity stuff and all that, now let's get serious about shipping software, and making tools that help people ship software." It's not that this view needs to dominate a community, but right now it's thoroughly unrepresented in Scala land.

View on HN

I feel like my experience is quite a bit different. I chose play framework for my web frontend and REST API, and it's quite easy to get started and ship something quickly.

Play framework itself is pretty straightforward with not much in the way of advanced Scala concepts (implicits are probably the most you have to understand). Testing is baked right in and they even handle setting up the application for you in your fixtures.

Coming from the .NET world, it was a huge relief to see that I could just use activator/sbt to compile and publish my artifacts on a build server, just as it works on my local machine. Setting up continuous integration was super easy in comparison to what Microsoft has historically offered (but looks like they're improving with .NET Core).

As for other frameworks and libraries, you may be right on a number of them, but I wouldn't say that all or even most Scala libs are designed to be pure. I have a number of dependencies that are very easy to reason about even if you aren't aware of advanced features.

Twitter? Foursquare? Linkedin?

(Also my old company, PagerDuty. Wee bit lower profile than the other three, admittedly.)

None of whom are producing well-supported public code / documentation / conference talks.

Edit: And to clarify. I think part of this is because there's a serious case of imposter syndrome among Scala developers. I feel like much Scala code I've written would make legitimately useful open source projects, but I'm not really happy with the interface, and I feel the innards could be a bit cleaner, so I don't release it, because it feels just hackish enough that I trust it myself, but don't want the outside world depending on it. I'm probably not the only one here.

> nobody is actually shipping software

What about Spark, Kafka and Akka? I actually think the Scala community is working very hard to improve user experience, it's just that they don't do in the expense of soundness.

As a ruby -> Scala developer, I don't miss DHH.

Spark is interesting, and is probably the most pragmatic focus in Scala land.

On the Akka side: Try picking up Akka-HTTP and use it to write an API client (which seems like something one would do with an HTTP library). There are lifecycle issues that are undocumented, and it really feels like this wasn't a considered use case throughout the entire development process.

I don't necessarily think DHH is right on everything, or even most things. But I do think it's valuable having a voice saying: "Does this help us ship" or "We're adding this feature because it's solving a real problem we encountered multiple times during the development of real software, and it would be nice to have a standardized solution".

Reply on news.ycombinator.com