Jump to content

  • Log In with Google      Sign In   
  • Create Account

Banner advertising on our site currently available from just $5!


1. Learn about the promo. 2. Sign up for GDNet+. 3. Set up your advert!


GameGeezer

Member Since 07 Feb 2010
Offline Last Active Yesterday, 06:50 PM

Posts I've Made

In Topic: FBO only renders to the first target

26 January 2015 - 08:51 PM

Oh I just assumed since incrementing from ATTACHMENT0 was the root of my error that they couldn't be. Must have been another small bug that accidently got tossed out along with that code.


In Topic: FBO only renders to the first target

26 January 2015 - 07:02 PM

Figured it out, In a tutorial I used the author treated the attachment Ids like they were relative: COLOR_ATTACHMENT1 = COLOR_ATTACHMENT0 + 1. They aren't relative to each other for some reason in lwjgl so I just had to change the Id's.


In Topic: Slow performance on A*

15 October 2014 - 10:41 AM

Nesting loops will cause an exponential loss in performance and your algorithm is nested three loops deep.

 

Your worst case scenario is: 100(tries) * weightsgrid.size * getadjacent(coords) = ? some number that is probably pretty big

 

NO NESTED LOOPS! (unless they’re actually needed)


In Topic: Help with matrix rotation

15 May 2014 - 10:36 PM

So when I'm rotating I can simply multiply the old_rotation by the new_rotation? The only difference between world and local being I translate local back to the origin before rotating?

 

I'm a bit confused as to how you tell the difference between a world and local rotation if the translation is in a separate matrix. As  of now I'm storing my world and local rotation separately and multiplying world * translation* local, which seems to work fine but do you know how it's usually done?

 

4. That's an interesting idea, as of now the child is not in respect to the parent but your method looks very elegant.

 

Thank you very much that helped a lot.


In Topic: Dynamic Slime Blob Generation

03 May 2014 - 06:18 PM

I could be leading you astray but could you have a multitextured surface where the geometry covered by the "infected" texture is subdivided in the geometry shader than displayed in the vertex shader?


PARTNERS