I am reeeeally disappointed. I've used the non-Steam version for 638+ hours. You should understand that you are not buying a click-and-export solution. The mobile export is not maintained. The instructions for basic setup are outdated and don't work, and you will wrestle with googling and trying different installs of underlying libraries; it took me the same amount of struggle to get the basic export working as when I did this with libgdx (which is free).
The Google and Facebook extensions/plugins are partly out of date and have outdated or no documentation. The current version of the Google-Admob extension simply does not support rewarded ads because it is out of date; these are the most recommended/least hated way of implementing ads. Support is fast for very small questions but nonresponsive (literally, doesn't respond) for significant issues. These are the things you're supposed to be paying $400 for. If you end up manually wrestling with SDKs or buying an extension for these mobile features, you might as well be using a free engine/framework like Unity, Unreal, Godot, Phaser, love2d, Defold, libgdx, Monogame, whatever. (And even there, if it's open source, there's a great chance someone's already done it for you. Worst case, you can hire someone to fix any bugs caused by platform changes.)
GMS2 itself is janky---the project database structure becomes corrupted, creates ghosts, and produces unavoidable conflicts when more than one person uses git/svn. The IDE crashes when working with assets. It requires an internet connection to verify your license; although this is supposed to be once a month, it's not consistent and you may be locked out without internet. Whenever internet is available, it creates persistent connections and, among other things, downloads required splash-screen advertising for the developer's other products. Finally, you're not learning portable programming knowledge. GML is not an object-oriented language. It locks your hard work into a proprietary software with a company that I really don't think is going to last long in its current form. I am also uncomfortable supporting the corporate owner, Playtech, which is the world's largest gambling software company.
If you decide you want to make a desktop and a mobile game, and want to export to HTML5 (normal for game jams or demoing your game), then you're paying $100 + $400 + $150 = $650. If you want to get onto the Windows store or Xbox's "indie" channel, that's another $400 = $1,050. If you ever want to go to consoles, that's an additional $800 annually, and the Nintendo Switch is not yet supported, although support has been announced for the future.
Bottom line: use something else, probably Unity, or anyway one of the bajillion free 2D game engines out there, because you won't get what you paid for with this.
Update 1/23/19: The lack of support for mobile has not been remedied. If you go to marketplace.yoyogames.com, you can search for the YoYo Games official extensions. IAP for Android, Google Licensing, Google Analytics, MoPub, and Facebook haven't been updated since 2016; AdColony since February 2017; and Google Play Services since May 2018. There are constant reports of bugs and problems trying to make these out-of-date services work, since Google and the other companies have updated their own SDKs dozens of times in the interim. To some degree, you're worse off with GMS2 because it's a closed system and random users can't always make an extension to provide full functionality. The upshot: All the time and money you put into your game can fail at the finish line, and no matter how easy GML was when you started out, there's a lot of regret waiting on the back end.
Posted September 22, 2017. Last edited January 24.
Plays perfectly with a stylus on a Surface Pro 3. Also a really fun game, although I was disappointed that it essentially has no story. The dev's other game, Reverse Crawl, had a light-hearted but compelling story.
I'm on my 5th replay and 17/18 achievements; I don't know why Steam says I've only played 4 hours, that's not true, although most of it was offline. This is a GREAT game that is very absorbing even as it can be picked up and put down. The graphics are adorable, the interface is fluid, intuitive, and clear, and the gameplay has a good amount of depth although it's true that there are some minor balance issues between different minion groups. (I usually end up playing revenant, wolves, rats, violet's, or greybeard, or maaybe bonehead, unless I'm just futzing around.) Most of all the game is FAIR. The enemies run on the exact same rules that you do. I do not understand the "RNG" criticisms; my experience has been that all mechanics are even. Overall: a perfect game if you have a trip or just want something distracting and more casual. Worth more than its price.
As a game, this is an admirable failure. As an artistic experience, it succeeds, but not in a way that makes for a playable video game. The bottom line, however, is that you almost certainly should not play it.
The writing is excellent, and not just by video game standards. The dev could write good sci-fi fiction.
The game has permadeath. The core game loop is good, but clunky. Progress is punishing. It takes 10 real life minutes to make a bunch of arrows, which aren't particularly effective, but they're what you get. I recommend the sling. Even taking all precautions, you can die instantly because a dogman or blue frog pops up at close proximity or in your sleep and kills you in one round. You realize that there's nothing you could have done, and often there's nothing you can learn from to prevent it in the future.
Because getting a character into survival state, which allows exploring the story, takes several hours, the clunkiness becomes an insane ping-pong of clicks (left screen, right screen, ten ping-pong clicks to sleep; seven to scavenge a single building; repeat until you die or quit, then repeat some more). I view this as an artistic statement about life that, like the sudden and unpreventable deaths, cannot be understood until wearily experienced.
The problem is that the game breaks into choose-your-own-adventure encounters that are both the reason to play the game and what breaks it. A majority of the options in these encounters will instantly end your game. Single encounters will have branching choices, the majority of which will end your game. They cannot be predicted. Because of their branching nature and the immense real life time to get to the point where you see them, you cannot realistically finish the game. Further, the game hard-limits your exploration; you get X number of choices, then time's up. The internet is full of gnashing and wailing, true believers who spent over 100 hours on the game begging the dev to release the text files of the end-game, which he basically did by telling people what to search for in the XML files.
If you do buy it, some guy made a batch file that acts as a save-game manager. Get it.