Re: [Algorithms] Depth-Sorting Axis-Aligned Boxes

[metanet s. <met...@ya...>]

> If what's stopping you from using a ZBuffer is memory or
> performance,you can try a span-list based system

Unfortunately the problem is the platform: Flash. This means I don't have access to the pixels directly if I want to use the flashplayer's rendering system.

In theory I could write a z-buffer or span-list system in Actionscript, however this would require circumventing Flash's built-in renderer and writing my own (on top of the basic get/set pixel commands). I suspect this would be hellishly slow, but if that's the only solution I guess it's worth trying.


> How many different camera angles will you be viewing from?  Also, by
> sprites, do you mean quads in 3-space, or something more like point
> sprites

The camera angle will be fixed (pointing along z) but will translate in x and y. By "sprites" I mean something close to GBA: 2D images you can translate/rotate/scale, but not anything as general as a textured quad; if I had access to textured polygons, I could conceivably build a BSP each frame and split each quad into pieces.


This is looking more and more grim! :(

thanks again,
raigan



--- On Mon, 8/3/09, Bert Peers <be...@bp...> wrote:

> From: Bert Peers <be...@bp...>
> Subject: Re: [Algorithms] Depth-Sorting Axis-Aligned Boxes
> To: ra...@ha..., "Game Development Algorithms" <gda...@li...>
> Received: Monday, August 3, 2009, 9:56 AM
> metanet software schreef:
> >> Unfortunately, this is the classic polygon sorting
> problem.
> > 
> > This is what I was afraid of; aside from the cost of
> building a BSP every frame, the more fundamental problem is
> that the graphics are sprites and not really amenable to
> arbitrary splitting. I was hoping in the worst case that I
> could split them up to have one sprite per face (top, front,
> side, etc), but it would appear that even that is too
> coarse.
> > 
> > So, I'm screwed?
> 
> If what's stopping you from using a ZBuffer is memory or
> performance,
> you can try a span-list based system; at every scanline,
> you keep track
> of which poly's overlap it, then you analytically compute
> the spans of
> pixels that are visible due to every polygon edge or
> intersection.
> See quake1.  If the sprites in turn have an alpha
> mask, you can still
> make it work by converting the sprites into several spans
> (ie. RLE
> compression; see, uh, 'Divinity').
> 
> 
> bert
> 
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal
> Reports 2008 30-Day 
> trial. Simplify your report design, integration and
> deployment - and focus on 
> what you do best, core application coding. Discover what's
> new with 
> Crystal Reports now.  http://p.sf.net/sfu/bobj-july
> _______________________________________________
> GDAlgorithms-list mailing list
> GDA...@li...
> https://lists.sourceforge.net/lists/listinfo/gdalgorithms-list
> Archives:
> http://sourceforge.net/mailarchive/forum.php?forum_name=gdalgorithms-list
> 


      __________________________________________________________________
Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your favourite sites. Download it now
http://ca.toolbar.yahoo.com.