Messages

@Walluce I just shared that plugin code in case it help you since there aren't many references for C++ plugins + AGS, it was not with intent to force or tell you to use it, it was more in to explain a possible way to get all things together and then sorting.

There's a plugin function IAGSEngine.GetRenderStageDesc, that was added recently, which let you get the matrixes for use with 3D renderers, I mention this in case you do not need to create a entirely new alternative renderer, but can leverage what already exists and just apply what you intend. Also, if this is for your game as you mention, about the walkbehinds, you can skip them altogether and not use them. I never use those and use objects instead, as this gives me a better control of the alpha, when using walkbehinds things are either transparent or not, and this is usually fairly limiting.

Dualnames made a FakeScreen plugin, this is done like this

https://github.com/ericoporto/agsfakescreen/blob/3ebb2236379f6268a5549850cf12fba64520d8e0/AGSFakeScreen/AGSFakeScreen.cpp#L333-L342

The way to do is to organize everything that needs to be draw and then sort everything by ZOrder/baseline. In his plugin each item is given the name ScreenItem.

(He didn't do the actual drawing in the plugin though and paired a script module for some reason)

I think the mode you mention is only supported in software renderer, and not in the hardware accelerated renderers (Direct3D and OpenGL).

As new characteristics are added to the graphical items (rotations, blend modes, ...) it may require updating the plugin. If there is some specific use, it may be easier to leverage the existing renderers somehow or add new plugin interface features.

@AGA , is the wiki username and password the same as the forums?

I don't like excessive log messages and checks that are not necessary that will only reduce performance. There's no need for it. This will only reduce log usefulness.

Also 0 was used in place of null in optional parameters - before some change in the compiler. Not sure if it was zero instead of null a long time ago somewhere.

0 is not a blue cup, it's the 0 sprite, which is ignored in some cases (GUI background and button graphics), and has a special meaning. It happens in the sierra template the sprite 0 is a blue cup.

I was going to comment here that it was working fine for me but just now I can see the march mentioned here - I don't remember seeing it as this before except once.

Spoiler

[close]

I see you are using that keyboard module that comes with the template. That control is quite unnatural when compared to more modern games. I think the controls from my Controlz module would work better for the isometric world, specially if your character is designed with 8 directions.

Ahn, the path finding algorithm has been completely rewritten since and this should not happen anymore afaik.

Bug reports without a way to actually reproduce makes them not useful. There's a template for bug reports where the information that makes them easier to act is mentioned.

For this specific case, an export of the walkable area that is problematic would be a reasonable way to reproduce, and mention of the AGS version and operating system where the issue occurs would be also useful.

UHM, it seems we are back to 2013

Can you explain a bit more of what you are trying to do, why and what you want to achieve?

Technically an AGS game can run another AGS game, just not sure if this is the question.

@Brian925 you can disable that behavior from the top bar by looking into the GUI properties, it has a specific setting there (I forgot the name) that does this behavior. You can disable that and handle it through script, so you would have full control.

image1 is the name, it could be anything like oCharmander or whatever. The number there has no meaning for the computer.

Assuming they are objects, you would write something like

Code: ags

#define OIMG_COUNT 4
Object* oImage[OIMG_COUNT];

// call this in room_Load
function initArray()
{
 oImage[0] = oImage0;
 oImage[1] = oImage1;
 oImage[2] = oImage2;
 oImage[3] = oImage3;
}

// Later in your code
  for(int i=0; i<OIMG_COUNT; i++)
  {
    oImage[i].Visible = false;
  }

I typed this on my phone, so try to run and see if it works. I renamed your objects as oImageX because I think it makes easier to pick up the AGS internal type of it.

I have a good idea for this, not sure if I will have the time to develop it. 

Thank you for this spreadsheet! It makes much easier to make a copy so you can play and make your own notes of all games.

I found the abstract of the chapter of a book online that apparently mentions the AGS interface. Leaving here in case someone wants to read it.

https://link.springer.com/chapter/10.1007/978-3-031-05214-9_11

The book is named The Authoring Problem, and the chapter in question is named Mapping the Unmappable: Reimagining Visual Representations of Interactive Narrative (doi 10.1007/978-3-031-05214-9_11)

Unfortunately there's something wrong with my institutional access, so I haven't been able to read it yet.

Edit: Being onsite my access to springer normalized. There's not much in the article, AGS is in a context of other text based narrative development software, so in this context it's interface is much more visual - as others (Twine, Inform7 and Storyspace 3) focus more on text.

This article gave me an idea for a simple tool that naively grabs all player.ChangeRoom in the room scripts, notes down the script room number and the destination room, and then produces a small graph of the rooms and their connections. While I have my guesses to do this in either python, R or JS, I am not sure if there are useful graphing libraries in C# so it could be cooked up as an Editor plugin, which is more elegant.

(of course change rooms in the global script and other non-room scripts would be missed, but I guess this would still have a little use to check your story development)

Here's a way to do isometric movement

- divide the walkable area instead of a tile, in quarters, so a corner gets only a quarter of a tile, a border half tile and inner regions get full tile.

- divide the rest of the map in two layers, background and foreground. Split wall and wall-like tiles in the middle, the upper half goes to the foreground, the bottle half goes to the background.

This way there should be no overlapping - IIRC this is how things were done when sorting was too expensive, or not possible in game consoles.

music1 is a bad name for a variable that represents the audio channel, this makes the code really confusing. 

Sorry to ask again, but, is there any other setting that resets or ignores the config in some way? Or this happens only with the setting for the touch method?

This will help me figure out what is happening - I am not sure if I am doing something wrong, but I haven't been able to reproduce this yet.

@Athanor , do you kill the app between sessions? Meaning, using the button/gesture that shows things loaded in Android, and swiping to kill the app.

Like, if you kill the app, does it fix for you?

Also, are there other configurations that "resurrect"/"reset", or only the touch method resets for you?

Ah, good question, I don't remember if the PlayingClip property is set in the same frame or only in the next one.

Any problems in opening the audio device can cause the audio channel to be null. Say you are in a desktop computer, with no Speakers or Headphones connected to it, and using Windows, Windows will tell that no audio devices are available, and then AGS will fail to open the audio device. So if you try to play something, there will be no channel to assign the clip as it fails to play.

I don't remember correctly, but I think if the game uses external vox packages but they are missing for some reason, it returns null too.