I’m not saying that this feature is useless and should never be developed
I say that we need to have priority to develop any feature of any software, we need to develop first something simple that people use before developing the other additional features.
In this case, the plan is simple:
Develop a minimal object with just the feature needed to be used basically.
Publish it.
Get feedback.
Choose the next most important feature to develop (here, collisions masks).
Go back to 1 and enjoy having a finished product that is not complete but without bugs and that is available for everyone really early!
You can’t say that if you’re not developing the feature
What I want to say is that if Victor do not develop it for the first version, you’ll be able to play with the object maybe at the end of this week! Then, anybody will be able to use it and say to Victor “Hey, collisions masks would be nice!”… or maybe you’ll use it and say to victor : “In fact, this bug or this missing feature is more important”.
We can’t do everything in a day, we need to develop it in a lean way (en.wikipedia.org/wiki/Lean_software_development).
And do not worry, you’ll get your custom collisions mask, but after the first version is released! It won’t take more time for it to be developed! It will just allow you to have the object faster, and this is a clear advantage!
I finally agree with 4ian, it will help to have some feedback on the GUI for example. Afterwards, I will add more complex features (custom collision masks, actions to change tiles in-game).
Before the first release, I will add predefined collision masks (triangles, rectangle, smaller rectangle) which will be adapted automatically to the tile size and add documentation to the object (for the user and also in the source code).
Next step, add some expressions (tile size, …) and collision conditions (on one tile and on the whole TileMap).
Then, I’ll release an “unofficial” version to get some feedback.
sorry for this out of topic question, but is there any hope you’d merge some of your other excellent extensions to git? I really wish they were a part of gamedevelop for good.
I do not know if you and 4ian realize this, but with this extension gamedevelop becomes much more than just a game engine. It acquires the ability to author animations, much like brashmonkey.com/spriter.htm
and esotericsoftware.com/
More importantly, this type of animation data can play very nicely with proceduraly controlled motion, caused by game logic.
It is opening a lot of doors, a lot of possibilities.
Armor swapping will become so much easier. Reusing animation and so on. Not to mention the giant gains in memory.
Btw I really really hope gamedevelop gets a spriter game object and support, but it is not even on their implementations wishlist atm.
Here is a first beta version for the TileMapObject (both Windows and Ubuntu) : TileMapObject-beta1.zip (1.11 MB)
(extract the content of this file into the GD folder)
There are some features missing (single tile collision detection, action to change the tileset on runtime)
Note : be sure to put the object on integer coordinates : it will create blank line while editing but it will look good with preview or playing the game.
Is it Windows only?
On Ubuntu I have extracted everything into the opt/game-develop folder and merged, replaced everything inside the folder but the TileMap Object doesn’t shown as an available extensions
I have the example inside examples folder, when I want to open it an error message pops up and say that extension is missing.
I found it really easy to use, congratulations! Collisions work well, it’s simple to set up
For me, there are two points to enhance for now:
In the tile map editor, maybe you should gray out the part that are out of the map (with a large window and a small map, the bottom right part is out of the map and there are lines of the grid that are drawn).
When configuring the tile set, it would be interesting to display a preview of the first tile Would be nice to make sure that the configuration is correct without having to close and reopen the configuration.
Good job in any case, it’s impressive!
I’d take a look at the code and merge it in the official GD repo when it will be considered as finished.
But the first tile is not a good indication because it is not influenced by the tile spacing. But I can display the whole tileset (using the same panel as in the main editor, so no more code).
I had a quick test and everything works pretty good. Though I was just looking at if I could apply collision between tilemap object and a spirte object without platfromer automatis and the pixel perfect collision check and keep separate two object action just didn’t work between tilemap object and a sprite object. I don’t know if it a bug or missing feature but I hope that, tilemap object can be used as any sprite object in the future.
What I was trying to do is to make a wall using tilemap object and try to apply (pixel perfect) collision between a player object and the tilemap object.
This object can be very useful to create complex parts of scenes using a few tilemap object instead of hundreds individual objects not only for platfromers but for any genre.
An other thing I have noticed, time-time just a blank white line appeard between tiles for a sec in preview, managed to make a screenshot of it:
Are you sure the tilemap object is at integer coordinates ?
If that is the case, it’s not caused by the object but by the camera taking non integer coordinates.
Im running the example included with the automatism and as I’m walking with the player, these lines just appear time to time for a sec, I was just lucky to be able to make a screenshot but I’m experiencing the same in Stencyl on Ubuntu when usign tiles. If nobody else experiencing the problem, Maybe my graphics driver then?
