[Sandy] Sandy 2.0 third draft
kiroukou at gmail.com
Thu Apr 19 05:53:56 EDT 2007
Well the only thing is to simplify the primitive creation an,d having less
code at drawing time :)
But that optimize a lot, so maybe it worth the pain to implement it .
Is there still a problem with clipping ? (with the latest comit I did? )
thanks for your commend, as usual.
2007/4/19, Petit <petit at petitpub.com>:
> The primitives look great, and they now can handle textures, even if
> strange things still happen when clipping them.
> Letting the Material's render themselves in a specialized manner may be
> a good thing.
> I like generalized code, but if it is to take care of too many special
> cases, there is no advantage in a general render method.
> What do we gain by having only 'tri' mode? Other than less code of
> course ;)
> I thought the 'quad' mode was fine when you use only simple materials,
> as it leads to fewer calculations and drawing operations.
> Thomas Pfeiffer wrote:
> > I've added new primitives :)
> > If you run the test, you'll see that it is moree accurate primitives
> that we
> > have now.
> > I've also improved some things but I think we still can improve th some
> > parts.
> > I think I'll allow the Materials to do the rendering process rather
> > doing several Renderer. Why? Because the materials can replace th
> > and in the same time optimize the drawing process with some specialized
> > code.
> > I'll also remove the (at least I think ton it) the QUAD mode and only
> > the q tri one (which was the default one).
> > Any advice or help is still greatly appreciated ;)
> > Thomas
> > 2007/4/17, Petit <petit at petitpub.com>:
> >> It's starting to look great, and we should soon have some more
> >> primitives and materials to play with.
> >> The structure is better and the rendering process a bit more logic, as
> >> understand it.
> >> There are still some peculiarities in regard to missing polygons in
> >> situations.
> >> Talking of parsers, I hope for a consistent Parser interface, so that
> >> all parsers a treated in the same manner.
> >> Writing some documentation and tutorial examples will boast the 2.0
> >> version and also reveal bugs ;)
> >> Let's move on with that.
> >> sorry to say, I will be 'out of town' for a while, from Friday for
> >> ten days.
> >> My computing resources will be rudimentary, so I may not be able to
> >> But I'll join in the discussing on the list and in the forum during
> >> time.
> >> Kepp up the good work!
> >> Cheers!
> >> /Petit
> >> kiroukou wrote:
> >>> Hi all,
> >>> If you follow the SVN feed (http://cia.vc/stats/project/Sandy/) you
> >>> may have seen that a lot of work has been done ;)
> >>> You have now 2 primitives ported (Box and Line3D) and the rest is to
> >>> come.
> >>> The core seems to be stable, I'm happy about that.
> >>> I think it's time to focus on this latest version, to make it
> >>> documented (I needed to go fast, and I forgot to document some
> >>> classes), tested (You can already start to write some example in
> >>> tutorial perspective).
> >>> For sure we need to port the parsers (ASE parser will be easy, WRL, I
> >>> prefer to leave this one, COLLADA, get the PV3D one, OBJ to write
> >>> ours with P.O works). and the primitives (you have 2 examples now to
> >>> see how you can port the previous code).
> >>> I need to see how to integrate back the Sprites (2D and 3D). That's
> >>> not something very obvious, so I'll take care of this part since I'm
> >>> the most advanced I guess in the core system.
> >>> I really want to include some of the latest improvments we've done
> >>> lately (MovieSkin, that would be MovieMaterial), and lightening
> >>> (that's all?)
> >>> I'll need to focus on the AS3 version as soon as this one is ready,
> >>> so your help is very precious if we want to have this release ready
> >>> soon.
> >>> Hope you'll be available ;-)
> >>> Thomas
> Sandy mailing list
> Sandy at osflash.org
More information about the Sandy