TextRender, Origin and GLEventListener Model Questions

[macdonag]

Hi, I’ve a games developer looking to use jogl (or something similar) for prototyping of personal projects. I’ve got a couple of questions:

• The origin in screen coords is the bottom left. I prefer to use the top left. Is there a simple way of changing this? If not, I feel this would be a pretty useful addition.

• I’ve used the TextRenderer class to render some stats to my hud. For layout it’s pretty fundamental to know the bounds of the string you’re going to display, but to do this, I apparently have to call getBounds separately with the string, and later pass the same string to the textrenderer. Is this optimal? Other text code I’ve used separate the setting of text (which computes the layout) from the rendering. This allows you to then use the bounds information for layout without incurring any extra cost.

• Looking at the the model of GLEventListener, it has an init and display callback, but no update. I would ordinarily break a frame into repeated update, display calls in the main thread for anything renderable, so for now I’m simply calling updates at the start of my display call for my main Display class. How does this compare to the intended model?

Thanks!
g

[Orangy_Tang]

@macdonag

• The origin in screen coords is the bottom left. I prefer to use the top left. Is there a simple way of changing this? If not, I feel this would be a pretty useful addition.

Just change the params you pass to glOrtho accordingly.

[kbr]

@macdonag

• I’ve used the TextRenderer class to render some stats to my hud. For layout it’s pretty fundamental to know the bounds of the string you’re going to display, but to do this, I apparently have to call getBounds separately with the string, and later pass the same string to the textrenderer. Is this optimal? Other text code I’ve used separate the setting of text (which computes the layout) from the rendering. This allows you to then use the bounds information for layout without incurring any extra cost.

Please look at the text-related demos in demos.j2d.* in the jogl-demos workspace such as TextFlow for ideas on how you can separate layout and rendering with the TextRenderer.

@macdonag

• Looking at the the model of GLEventListener, it has an init and display callback, but no update. I would ordinarily break a frame into repeated update, display calls in the main thread for anything renderable, so for now I’m simply calling updates at the start of my display call for my main Display class. How does this compare to the intended model?

That’s a valid approach.

[macdonag]

Thanks. I haven’t had a chance to play much with it lately, so I’ll provide feedback when I can