Corona Labs is shutting down, will open-source everything under MIT | Hacker News Reader

Corona Labs is shutting down, will open-source everything under MIT

coronalabs.com

144 pointsminxomat25 comments

This post makes me sad.

I was friends with Carlos Icaza, one of the co-founders.

Carlos was a diabetic who died too young.

Every time I walk by the Stacks in Burlingame I think of him. We used to have breakfasts there every week or two including our last breakfast before he moved to Miami.

Carlos was a hustlers, sometimes too intense, but he had a great heart.

He used to constantly tell me: don’t be an engineer, be a fucking entrepreneur.

It took me a while to understand what he meant.

RIP, Carlos!

I hope Corona or parts of it be around as an open-source project for a long time.

>don’t be an engineer, be a fucking entrepreneur.

What did he mean?

I didn't know Carlos, but when I've heard this sentiment before I've boiled it down to mean that you shouldn't solve _a_ problem, but instead solve _someone's_ problem.

In our context, we used to talk about all the new tech innovations, languages, frameworks or new products.

I always want to build a better "mousetrap", I'm an engineer/product guy first, and he was a hustler first. If my approach was to build a better product, his approach was to be a better hustler promoting a product and getting it adopted.

I wasn't aware Carlos Icaza passed away. I was briefly involved in a collaboration between Corona Labs (then Ansca Mobile) and a iPhone game company in the early days. Never got to finish anything in Corona but I liked it a lot.

Here's a archived version of the article whilst it is getting hugged to death by us: https://web.archive.org/web/20200214140041/https://coronalab...

But perhaps you never heard of them, but now you have seen their 'name', I know what you are probably thinking these days: Nope not even close, but actually it's perfect timing.

The SEO on this right now doesn't look good right now and it would't be great in the long term. But even then, the product wasn't really generating enough revenue for them to stay afloat and now continuing this under their current name with users using search engines would have been the final nail anyway.

Open-sourcing sounds like the right way to go for a proper send off for 'Corona' the product. Probably needs a renaming too.

I remember testing Corona back in the day as it was quite novel at the time--code your game in Lua and compile to multiple platforms. I think it's safe to assume that the increased competition from Unity, Godot and the plethora of other cross-platform game engines over the years led to a major reduction in paying CoronaSDK customers.

As a web developer myself, I found Phaser a more natural fit, since I was more comfortable with Javascript and was only making games as a hobby. Many independent professional game developers tend to stick with Unity or Unreal due to their large customer base and more feature-filled editors.

I'd say the direct competitor for Corona is Defold. Also mentioned in the forums discussion.

Corona has an edge on many other smaller engines because of the broad device support. E.g. it's one of the few engines able to ship to tvOS.

Does Corona work in Firefox? Nothing done in Dapper appears to work at all/properly in Firefox (even their sample demos in the docs)

And Love2D, which is free and open-source, also using Lua.

I've never heard of Corona before, but this sounds like an impressively clean shut down. They seem to be going to great lengths to ensure Corona has a decent chance of surviving, even though the company behind it is closing. I wish more products would handle things this way.

Even if a company itself isn't going out of business, just abandoning one particular product, taking a few steps to ensure customers aren't left hanging would be much appreciated. (Imagine if Google had open sourced Google Reader, for example, instead of just killing it off.)

Pure conjecture: product wasn’t doing great and they got an amazing offer for their domain name. They figured it would never be worth more and shut it down so they can sell the domain.

Archive for anyone with connection issues: http://web.archive.org/web/20200214140041/https://coronalabs...

This brings back some memories. I used the Corona SDK to build my first iOS app and Corona and Lua were a joy to use. The app is still around but I've since moved to a different framework. But I'll always appreciate Corona for helping me get started with that particular project.

I maintained a page called learningcorona.com for several years that collected the limited number of tutorials and articles on Corona. Carlos, the co-founder, was very kind and emailed me several times to include the site in the official newsletter and to ensure that I didn't run afoul of any trademark problems (in a gentle way). I appreciated how well he interacted with and supported the small Corona community.

This is kinda sad to see. It's been almost a decade since I last used Corona (then, it was called Ansca Corona), and I found it fun and exciting to use. My employer at the time wanted to use it to build a game app, and I spent a while messing around with it, learning Lua on the fly, and in general having fun making a "game".

I found the system to be easy to use, and relatively high performance. The only thing that was an issue - at the time - was something very strange, which kinda impacted another app we were building:

Corona had a map component, for google maps - but for some reason, it wouldn't work for Android - only for iOS! So - you could create an app with that component, but if you compiled it to install on Android, it would fail, but you could compile it for iOS and load it on an iphone and it would work fine.

At the time, we had an app idea from a client that we wanted to use Corona for, but because it needed mapping, and that didn't work, we were forced to go with another option (PhoneGap and Bootstrap Mobile), that wasn't nearly as performant - but we could embed a google map easily.

Contact with Corona revealed they were working on a fix to get the map component to work with Android, but it came too late for our purposes (and actually didn't happen until about a year or so after I had left that position and had moved on).

Anyhow - I'm glad that the system will live on, although today there are a ton of other options available for easy cross-platform mobile development. Still, I haven't found anything that worked quite as well as Corona.

I maintain Planimeter's Grid Engine https://www.planimeter.org/grid-sdk/, which I also advertise as a 2D game engine. It also exists in the Lua ecosystem.

I'm sad to see Corona Labs shutting down. I'm happy it's engine will be open sourced.

There is some really great 2D game software in the Lua ecosystem, but hardly any of these products are actual engines.

Defold and LÖVE aren't game engines, they're frameworks giving you load, update, and draw callbacks and some bindings to underlying libraries. None of these solutions will load levels for you, or provide entities or actors.

Polycode is more of an engine than the former two, or even Corona.

I'm sad to see another piece of software slated for end of maintenance in this space, but I'm hoping that it will inspire the development of more actual game engines that do more than provide a collection of bindings.

I've wanted a Lua game engine that loaded levels, provided entities, had multiplayer out of the box, and everything else you'd expect from a game engine and not just some Lua C API bindings to OpenGL, OpenAL, and Box2D.

It turned out I had to build it myself.

Not that I've heard of this before, but they are handling the sunsetting and open sourcing in a great way, at least if they stick to the plan.

I remember evaluating it around 2011 when they were first picking up speed. What put us off was that, although the simulator worked offline (and writing Lua for it was a lot of fun and 10x easier than getting into native development), the app build process happened on their servers — so it was essentially a service, not a tool. All code would have instantly become useless if they would have just pulled the plug one day.

Not sure what happened afterwards, but it’s nice to see they were able to open source what they had, and to keep it possible to build old projects locally.

*edit: removed uncertainty about building old projects locally

According to the blog post, one of their tasks is to release a version of the build tool that will work locally.

I used Corona around that time - and from what I recall, projects did build locally; the only exception being that if you wanted to build an iOS version, you had to do so on a Mac with XCode installed. I also recall having to set up the Android dev environment in order to build for Android. I know that later they had where you could build on their servers, but what I recall (and I admit, it's been a while), local building was there back then.

Back in 2013 I used corona and lua to help me make a game for android, which I needed to make in order to pass a class in college. Have to say that it was extremely easy to do, way better and easier than what my teacher taught me to do in Android.

Hope that it continues living and improving.

Reply on news.ycombinator.com