IntelliJ Plugin is missing

[gmitch215]

Woke up to find that the IntelliJ plugin no longer exists.

Can't search for it, and the direct link reports a 404.

[soywiz]

Korge plugin was removed from the marketplace as it was outdated.

There are three options now:

• Use https://forge.korge.org/
• Use Intellij without the Korge plugin. It is not required to use Korge with gradle
• If you are using an old intellij version compatible with the plugin: https://github.com/korlibs/korge-intellij-plugin/releases/tag/v4.0.4 here is a zip file you can install in IntelliJ

[wbehrens-on-gh]

As others have expressed here I think this is a step backwards. What originally drew me to korge was how simple it was to setup with just an existing Intellij installation. I personally hope this decision can get reversed in the future.

[soywiz]

If a team is formed, and they have bandwidth to support it and keeping up to date with deprecations and changes and decide it is worth to do it, of course. I'm not against it.

Also have in mind:

• Forge is installed with one command and two clicks.
• Korge can be used in intellij normally, as it uses gradle, and it is a plain kotlin library. Gradle tasks are there.

[gmitch215]

If the plugin wasn't necessary for development, what was its original purpose? Is this purpose still relevant?

[FyiurAmron]

also, FWIW @soywiz : https://korge.org/features/quicksetup is very outdated, some links are dead and things mentioned there don't really reflect current reality AFAIK.

[gmitch215]

@FyiurAmron I don't blame him - the entire ecosystem is essentially only managed and maintained by him. Given the size of this project, it's amazing it's even up-to-date.

He said he'd return from a hiatus around this time, though I haven't heard from him in a while.
I think he just did what he thought was best for his project.

[FyiurAmron]

@gmitch215 I'm not blaming him, because this is not about blame. And yes, I also do consider this choice (switch from plugin to IDEA fork) an attempt to improve, even if arguably failed one. However, it's still about taking a step back and looking at the engineering problems & solutions and whether they work or not. I present the case why they don't currently. IMVHO, Forge just doesn't make sense in its current form, precisely for the reason I stated - maintaining it as an IDEA platform fork is an order of magnitude harder both for devs and users than providing a plugin. The argument "but at least it can stay that way indefinitely instead of a plugin which will get incompatible" doesn't hold, because the IDEA fork has to be updated anyway to work (code rot happens anyway).

Also, it's a ~2 GiB and completely unneeded installation, not to mention the added moving parts here and the failures caused by them (see the screenshot - this is from my actual attempt to use Forge, dunno if it's a "hard" bug [likely] or a quirk of my system [unlikely, but possible] - still, trying to use it just for an example project just fails due to internal IntelliJ exceptions being thrown around).

And yes, I say that precisely because I'm doing exactly this kind of stuff in my day job, and has been doing it for a while. Choices like this backfire, usually quite quickly (months or years, not decades). Hell, I wouldn't mind updating that plugin if I'd to stick with KorGE (at least I'll up my game with IntelliJ plugins again, haven't worked with them in some years) - no way I'm touching a platform fork (learning how to do it is useless for me as it makes no sense for most cases, precisely for the reasons described above - for Java/Kotlin there already is a verified, stable and robust alternative, and doing so is by itself a tedious coding exercise, but nothing more).

A side note: the fact that this is a single-person project is not a viable "line of defense" for deficiencies in neither FOSS nor commercial closed-source in XXI century. If he wanted or needed more developers, he could ask around, and given enough effort they would come. There is precisely 0 need for the project to be 1-person, unless the project is obscure and mostly useless (which I really don't believe is the case here - it's reasonably useful and there is a community of actual coders around KorGE). Keeping something mostly 1-person is usually a decision of the core dev based on his personality or project management views and approach to team processes, not dire necessity. Also, it's unsustainable in the long run for incubating projects (which KorGE, arguably, is).

[gmitch215]

@FyiurAmron Everything you said is correct. I'm sure he had a reason that may not be obvious, or he just did it to gain experience for himself.

There are other members in the korlibs organization, so he's trying to pull help - most of them are inactive or don't work on the main features. There are a million explanations that are reasonable, and a million that are not; one or multiple could be true.

All I'm saying is that he'll do what he thinks is best for his game engine. One good reason for switching to an IDE is implementing a visual editor (which is probably the most likely answer long-term) along with providing more functionality that JetBrains plugins don't have the possibility of doing.

[FyiurAmron]

All I'm saying is that he'll do what he thinks is best for his game engine.

This is what I fear the most. It's literally how great projects with narrow leadership channel fail.