When comparing Gosu vs GameMaker Studio 2, the Slant community recommends Gosu for most people. In the question“What are the best 2D game engines?” Gosu is ranked 50th while GameMaker Studio 2 is ranked 67th. The most important reason people chose Gosu is:
Gosu is not a game development framework, only a media library that happens to be suited to game development. (Kind of like SDL in the C world.) That means the interface is relatively small.
Ranked in these QuestionsQuestion Ranking
Pros
Pro Lightweight
Gosu is not a game development framework, only a media library that happens to be suited to game development. (Kind of like SDL in the C world.) That means the interface is relatively small.
Pro Mature API, actively maintained and developed
Gosu has been under development since 2001. It is mature and has several toolkits built on top of it to provide additional functionality.
Pro Cross-platform, even mobile, using Ruby
Pro Quick prototyping
Pro Good user interface
Pro Well-optimized engine
Pro Has a trial version (but limited functions, can't export)
Pro Many unofficial tutorials
Most GMS1 tutorials are fine for GMS2
Pro Highly customizable IDE
Although users must work within the IDE and editor, GMS2 has many options to customize the look and feel
Pro Good documentation
Pro Huge, generous community
Cons
Con Deploying Ruby apps is a mess
Games built with the Ruby to .exe "compiler" do nothing more than extract your source code and Ruby.exe to %TEMP%, then run it. The code is not really compiled at all. The process for wrapping games as Mac apps is a bit nicer, but you'll need a paid Apple Developer subscription to code sign the app, or users will see a warning/error when running your game.
The only way to really compile Ruby is to use RubyMotion, which does not work on Windows and requires a paid subscription on top of the Apple Developer one.
(This Con is not specific to Gosu. Deploying Ruby code has never been fun.)
Con Not the best scripting language out there
GML is just weird; if you want to learn programming, it is not the best because it teaches bad habits and has many odd shortcuts and shortcomings that won't transfer to a real language
Con HTML5 export is buggy, doesn't "just work"
Con Quite expensive
Windows ($100) + HTML5 ($140) + Mobile ($400) + UWP ($400) is $1,050, plus $800 anually for each console export separately. But doesn't do anything any of the free engines can't do, and the stability and tech support aren't great.
Con Unstable
Users frequently report crashes and hangs, particularly when working with assets, and the software uses a complicated underlying meta-file structure that may become corrupted and cannot be rebuilt
Con Limited support for OOP
Con Small development team
The core programming team is only 5-10 people, with about 30 employees total, so bug fixes can take a long time to be addressed, and there aren't many official tutorials