Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[3cats]

Hey guys !

-IF- one were to create a 1st person adventure game (Rhem, Myst), with lots (100's) of locations, what do you think would be the best way to go about it ?
I was thinking : 1 world location (x,y on world map) = 1 scene ; the scene would contain different views (combination of : forward view, backward view, view to the left, view to the right, up, down) ... Suppose : 1 view = 400x300 px

1) combine the images in the scene editor : main layer contains all entities (1 entity / png view) + region entities for hotspots 'look left', 'right' , ...
   => use Scene.SkipTo(x,y) to show correct area of png on "LeftClick"
   => how expensive is empty area in main layer (size eg/ (400x4)x(300x3))  

2) combine the images in Gimp ; main layer contains 1 entity (for the 1 png) + region entities for hotspots 'look left', 'right' , ...
   => use Scene.SkipTo(x,y) to show correct area of png on "LeftClick"
   => size eg/ (400x4)x(300x3)

3) in scene editor : use 1 layer / png view + region entities for hotspots 'look left', 'right' , ...  => layerFW, layerBW, layerLeft, ...
   => use LayerToBeShownOnLeftClick.Active= true;  CurrentLayer.Active= false    : to switch between views
   => all layers are eg/ 400x300

4) no difference between the 3

Thanks for your input ! 

PS-I was going to insert some images to try and clarify, but I can't get it to work 

[Dan Peach]

For Corrosion, I used a new scene for every single view. I created separate folders for each location, and then created separate scenes for each of the 4 views on each node. So for example with a corridor, I could have 5 nodes, with 4 views each, totaling 20 scenes. Like this for example:

scenes/main_corridor/main_corridor_node_001
scenes/main_corridor/main_corridor_node_001_back
scenes/main_corridor/main_corridor_node_001_left
scenes/main_corridor/main_corridor_node_001_right

Then I'd repeat that for nodes 2 to 5.

This worked best for me because I like things to be as separated as possible. It made it easier to control all the changes in the game because the changes were confined to just one scene. If you have all your views in one scene, then you need all your hotspot regions in one scene, and they need to change when the main view changes, but change back again when it changes again. Seems like a lot to keep track of, and seems very easy to make mistakes and create bugs that will take ages to hunt down because you have to so much going on in just one scene. It was relatively easy to solve problems for me because the scene scripts were quite simple and short because they only dealt with a single viewpoint.

We're all different though.

[fireside]

I would use separate scenes, but I would create a subfolder which held the 4 view scenes so it didn't end up too messy.

[3cats]

Yeah ... I guess maybe I'm making things needlessly complicated ...
I'll definitely think about using the solutions you suggested (thank you both !). I'm also still considering my '#3', because that could also be a convenient way of organizing things. It will depend on how complicated the scripts will need to be to make it all work...
I'll try your way, and my #3, and report back here when I have something more to say.   

[metamorphium]

Keep in mind, that WME loads all visuals for the scene immediately. having four views in one scene might potentially feel that the game is stuttering because turning would be instant while stepping would be slower (of course this depends on resolution / computer). I would never do this on iOS for example.

Having scene per view could lead to much stable performance.

It also really depends on what is inside. If it's eg. just static shots, you can simply have one scene for 4 sides and dynamically load each image while having all region entities layered in SceneEdit.

It would look like this:

main layer - just the background image
front layer - all region entities for front
right layer - all region entities for right
etc...

Then with each turn you simply load correct image for background and activate corresponding control layer.

But I would not recommend this with heavily animated scenes because it could become messy unless you are good with creating dynamic environments.

Just my 2 cents.

[3cats]

Well, it's a 1st person, so there wouldn't be an actor walking around...
But you are right : I did not think about image loading. Very good point !   

[metamorphium]

just for the record - by stepping I meant location change. Not actor walking.