[SOLVED] Strange question!

[shaun9991]

Hey guys,

I've received a strange query on the itch.io page for our game Sumatra: Fate of Yandi - https://shaun9991.itch.io/sumatra-fate-of-yandi. I've pasted the query below:

"If you don't know yet, this game has a incompatibility with the newer AGS engine because the old pathfinder was inaccurate at calculating endings https://github.com/adventuregamestudio/ags/issues/784
If this could be patched, it would be nice."

"the problem manifests on users using the ags standalone executables, git versions; since the pathfinder changed in git. As mentioned later in the issue by kmar, this is *probably* because the old pathfinder was more forgiving of inaccurate non-connecting paths and could 'jump' a pixel or two to get to a target pixel even if there was no connection; though this is something only your studio can really verify. In short, it's not so much that the game won't work on the distributed version of the executable, but that people wanting to use a newer version, for features and backwards compatibility will get stuck because of this."

Why on earth would anyone want to run an AGS game like this? For me this sounds absurd and not something to issue a patch for - am I missing something? The game was compiled with AGS 3.4.1.15 and no bugs have been reported since launch - indeed the bug they refer to obviously does not exist in the released version of the game, if executing the game as normal...

In the meantime I've advised them strongly to not run AGS games in this manner... I can't see what the benefit would be? I imagine the people in the github page might be active on this forum. Does anyone know why they would want to run AGS games this way?

Thanks!
Shaun

[Dualnames]

You made this game in AGS? Well, let me run it through my custom DOS parser version of the UNITY ENGINE in combination with unreal and then report to you that it should be patched because it doesn't work. Like WTF

[shaun9991]

lol. Thanks - this confirms I am not going mad!

Literally WTF

[Privateer Puddin']

Running games with a newer version of the engine is sometimes recommended on the forums (albeit normally for much older games). I'd hazard a guess that they want to use something like OpenGL rendering in Linux added by 3.5.0, but now are faced with this 'bug' due to the updated path finder

[shaun9991]

I see...

Still. Weird.

The release is "as it says on the tin", as they say.

[Crimson Wizard]

This Linux user uses the latest version of AGS port to run older games. Often it is the preferred variant, especially when the original game was not distributed with the Linux executable.
As Privateer Puddin' sais, newest 3.5.0 version added OpenGL support on Linux, which greatly improved performance.

Linux users were running AGS games like this for many years, because roughly 99% of AGS games did not have linux support upon release.

In this specific case the person is particulary impatient and decided to mail you before I could look whether this problem may be fixed in the engine.

You made this game in AGS? Well, let me run it through my custom DOS parser version of the UNITY ENGINE in combination with unreal and then report to you that it should be patched because it doesn't work. Like WTF

He runs it with our 3.5.0 Linux port. For now we officially support backwards compatibility, meaning games made with earlier versions should continue run with new ones.
Also he did not say it should be patched, he asked "If this could be patched".

[shaun9991]

Thanks CW.

I've edited my response to him to be a bit less 'sassy'

I'm more than happy to create a Linux build for him to test out, he didn't point out to me that this was to run the game via Linux.

As a side note - is it something that might be able to be addressed via the engine? I'm unsure how I would patch this myself - as the effected code appears to something very simple that's used 100s of times in the game - literally just calling a character to walk to certain point after room fadein and some events to occur. If you'd like to take a look at the code just let me know.

[morganw]

I would imagine there is a walkable area that is very narrow in this room or a check on the players final co-ordinates that has less tolerance than in other rooms? The 'patch' might just be painting a walkable area to be 1px wider.