GameMakerStudio is inherently crossplatform and its games can be run on multiple platforms. the game just needs the runner (basically the universal executable) for that platform in order to load the game. Since this particular game didn't release officially with the runner for linux, It would be taking the runner from another game running on the same engine that was released on linux, and using it instead. Game assets themselves would stay the same, we would just be borrowing the executable, and then subsequently load that executable into libTAS.
So, in a hypothetical scenario:
Would a game made using GameMakerStudio that doesn't have an official linux port,
but can use the stock GMS runner for Linux without any issues be accepted to the site?
Even though its "technically" modifying the game/assets?
Or would one be better off waiting for an official linux port of that game?
Just trying to gauge where the line in the sand for acceptance would be in this scenario.
So I've been playing with this for a bit and I had some ideas to get around the no-input issue for Undertale when running under Hourglass.
1) XInput to DirectInput DLL Wrapper.
2) Decompile/Recompile the game with DirectInput Support.
3) Rebind the 'Gamepad' in the game to keyboard commands.
Each of these kinda has a different problem though:
All of the existing XInput to DirectInput wrappers I've tried don't support remapping to keyboard, only to a DirectInput controller.
De-Compiling and Re-Compiling the game with DirectInput instead of XInput may be considered too heavily modifying the game/engine for the site's liking, although this would be (in theory) the much easier option. (This is how people got the game running on Linux before the official linux port of the game)
I honestly don't know if Hourglass (Or even Resurrection) even support grabbing buttons from a gamepad. (Which considering that it doesn't support mouse input this is probably likely.) It might be possible to use Hourglass in conjunction with another program (Like Autohotkey) and record input that way to get around this though.
Then I had he T.P./Gen thing backwards. That would be less of a problem then.
My only real other question is though, Would the differences in Neu/Gen be considered enough to get their own branches, or the same one?
Just wanting to know some more of this debate, help get the ideas flowing.
Edit: Also, I just tried to run the latest automated build of Resurrection on a 32bit XP install, and its failing to run with the error: "not a valid Win32 application". Went back a few builds (79) and its throwing the same error on that version as well. Checked both on my 64bit Win10 install and it runs, and says its running in 32 bit according to task manager.
Well since the game is I guess 'open' according to Bisqwit, ill un-lurk and chime in.
Assuming the game can be ran well enough inside of Hourglass enough to sync properly, Why not just do 3 individual runs of the 3 different endings? End the first with the setup for the second, and second for the third. (Neu-->T.P.-->Gen)
Of course this would require 'save anchoring' which is normally a no no, but would be better for presentation instead of 1 longer input file for 2/3 or 3/3 endings.
(I say this as I'm doing research on the subject. Kinda looking for an excuse to mess around with Hourglass & Undertale)
Well then between the both of us we are really dropping the language ball today. If you had something else in mind I'd be more than happy to throw something together really quick.
Been kinda following this idly for a while. Do y'all still need an icon? I tried making a more modern-ish version of the existing icon. Here are a few different versions.