SDK Most Wanted

Hi all,

Thanks for all your hard work making plugins and behaviors for Construct 2! You're helping make Construct 2 the best game editor out there.

Usually I'm too busy to spend much time on the SDK, but I thought it would be wise now that a stable build has just gone out to take some new suggestions. What are you most missing from the SDK?

As I said on fb,

• Ability to define more than one dependency.
• Use image to fill a quad in edittime.js.

I'm sure that I forgot other things, I'll add it to this list soon.

Thx for opening this thread.

More control on editor (edittime.js).

It might be useful (interesting) to create plugins for editor.

Uh. just a suggestion.

There's a huge disparity between edittime and runtime, it's kind of exhausting to write different functions for each, especially for rendering. draw() for 2D, drawGL() for WebGL and, of course, edittime - they all are fundamentally different. Tis a pain. :( So much pain I haven't been able to produce edittime rendering for SpriteFont. I also gave up on Frame plugin (just like Panel in CC), because it was rendering differently in each of drawing modes (not only on different browsers).

Would be much appreciated if this could be addressed somehow!

About the new pf_nosize flag,

could it be possible to replace it by two flags : pf_noWidthSize & pf_noHeightSize (need it to prevent height resizing...)

AS Mipey and rexrainbow suggested, the possibility to add our own window at edittime for the plugins could be great (allowing more refined control of the plugins). If it's not possible, at least some control of what we can draw on the IDE' screen (perhaps loading our own images, or allowing to paste images stored as base64 strings inside the edittime.js).

If all that isn't possible, is there a possibility to add one of our own binaries in the toolchain ? (For example, at export time, having the capacity to launch a binary coded by one of the plugin dev, to have a pass on the generated files).

Two suggestion for manipulating during edittime.

1) If it's possible it would be nice to be able to launch a small (standalone) html5 browser (window) to which we can pass something and after that get back some stuff.. eg using libraries like svg-edit during edittime.. if we have this I think only the sky is the limit for what we can do in edittime.. in theory.. ;)

I've created a very simple example for launching a standalone html5-browser + svg-edit - done with a batch-script ==> all together ~7MB, perhaps it's possible to get some stuff back to C2 via cookies or something similar - QtWeb+svg-edit

2) The possibility to execute an executable-file during editime and the C2-editor loads then a created file(XML- or JSON or Textfile or ..) .

This would also open new possibilities for manipulating stuff during edittime.

very easy example to get the idea - made with CC <img src="smileys/smiley17.gif" border="0" align="middle">

simpleTxt.exe (CC-cap)

I would like to have a "recreateBody" flag in the physics behavior.

At the moment there is a local recreateBody-flag that is only active when the physics behavior is updating, and then is directly reseted.

I would like to have a flag, that is active until the physics tick function is called at the next update and then reseted, so I have the chance to react to a physics body rebuild.

And to have this SDK bug fixed <img src="smileys/smiley1.gif" border="0" align="middle">

http://www.scirra.com/forum/r90-code-error-in-physics-behavior-js-file_topic52256_post328631.html#328631

I would like to easy add 3d (camera, terrain). Maybe with using of another js engine. Just to make easiest way to work with webgl and 3d