[Papervision3D] Quaternions vs matrix

Thu Jan 18 05:52:56 EST 2007

Hi Richard,

I am cool. It's been a while. Wanted to come to Brighton, but as
always, no time. Probably be joining one of the coming LFUG somewhere
in the coming months.

I've been following WPF/e a bit. Haven't really tried it yet... but I
have been following some of the M$ oriented blogs, and cloud 9. As
said, currently there is no 3d indeed, but some of the blogs have
indicated it's on the roadmap for it. WPF on itself looks like it the
competition for Apollo, so if the big A would be kind enough to put
some more HW support in there, that would strengthen that position a
little...but that of course wouldn't make any sense.

As far as Director 3d vs. Flash 3d goes...Director clearly is a winner
in terms of performance, but the ubiquity and other features of Flash
/ Flex clearly puts it way over Director. With PV3D in it's current
state I have the feeling we are almost, but not totally there yet. At
least it's a good step in the right direction.

As a future vision :
I am looking forward to that keynote in a couple of years, where Lynch
will display "Flash 3D !" and shows the performance difference against
the then released PV3D version 2 ;-)
I feel it's not if, but when that happens.

Ralph.

On 1/18/07, Richard Leggett <rleggett at gmail.com> wrote:
> Hey Ralph how goes it? Just wanted to chime in about WPF/E there as I have
> spent quite a bit of time using it. AFAIK there is no indication that it
> will include any hardware acceleration or 3d engine, do you have a source,
> or could you mean WPF (which is Windows only?).
>
> Someone today saw a demo of Papervision3D and said "that should be the end
> of Director then". This is a bit of a leap but if I were a company listening
> to my users, and watching what they obviously desire, this could mean an
> expedited opengl for Flash (just dreaming). I think that with what we have
> here, we aren't quite so limited in the ways that opengl and others are. For
> example we can apply a halo style glow around 3D elements with no fancy
> render pipeline that you'd have to implement in a "traditional" 3D engine,
> usually relying on hardware. That has come about through limitations, but
> it's also showing that there are some hidden strengths to this approach.
>
> Very cool!
>
> Rich
>
>
>
> -----Original Message-----
> From: Papervision3D-bounces at osflash.org
> [mailto:Papervision3D-bounces at osflash.org] On Behalf Of Ralph Hauwert
> Sent: 17 January 2007 17:35
> To: Papervision3D at osflash.org
> Subject: Re: [Papervision3D] Quaternions vs matrix
>
> With the wpf/e player having hardware accelaration in their planning, I
> wonder when we get some official signals on adobe's pov on hardware support
> in the player. I'm also wondering where apollo will go with this. As far as
> HW3D vs ubiquity > the current mac os x tiger implements some usage of
> OpenGL already. So if the level of code portability is on that level, how
> far can some more support be :D :D :D ? (I know, I know, that's really total
> bullshit, don't flame me for it).
>
> At least I really wouldn't like to see some kind of Flash Player engine
> implemented software 3d engine, since that would probably limit us in
> extending it. With BitmapData being fine as is, I wonder how they'll aproach
> the more complex 3D support towards developers, when the time is there.
>
> Ralph.
>
> On 1/17/07, John Grden <neoriley at gmail.com> wrote:
> > Yeah, i've been pushing Carlos in front of adobe people in hopes that
> > these efforts in 3D will get them to consider adding API's to the
> > player that could really help the speed of this type of engine.
> >
> > Its certainly in their interests to support and do something like this
> > since they're stated position on 3D is that they don't want to get
> > into hardware acceleration and platform issues nor do they want to
> > increase the size of the actual player for 3D support.  Not until it
> > makes too much sense to do so.
> >
> > Adding some API's that are specific to 3D could really enable adobe to
> > get alot more life out of the player "as-is" and still remain very
> > cross platform.
> >
> > I know they're at least looking at PV3D and that the SW Demo has been
> > passed around internally.  So, they're definitely listening ;)
> >
> >
> > On 1/17/07, Carlos Lunetta <carlos.lunetta at gmail.com> wrote:
> > >
> > > Jokes apart, from a spectator point of view (not good enough to
> > > become a
> > developer of any) it would be wonderful do see both sandy and
> > papervision getting cozy close together.
> > >
> > > I also think it's past time Macrodobe to show more their support for
> > efforts like these - 3D is a natural next step in flash evolution. We
> > got the pixel operations the were long needed in flash 8, we got some
> > speed boost on 9... they will only profit and add a tombstone in
> > Director with a really good 3D engine for flash.
> > >
> > > Processing has some java non-openGL 3D classes that were really well
> > written (http://dev.processing.org/) - and lots of talented people
> > writing 3D code with them - check Alcys, for example
> > http://www.alcys.com/index.html,
> > >
> > >
> > > Ralph Hauwert wrote:
> > > Sandy in the Sky with Paper as opposed to Lucy in the Sky with Diamonds
> ?
> > >
> > > On 1/17/07, ricardo cabello ~ mr.doob <info at mrdoob.com> wrote:
> > >
> > >
> > > I got a better one, "Sandy's vision of a Paper, now in 3D!!!!"
> > >
> > > Wonderful name for a codenameproject!
> > >
> > > --
> > > Ricardo Cabello ~ Mr.doob
> > > Designer + Developer
> > >
> > > Web:
> > > http://mrdoob.com
> > > RSS feed: http://mrdoob.com/rss.php
> > > Skype: r08028
> > >
> > >
> > > Pete Hobson wrote:
> > >
> > >
> > > sandpaper?
> > >
> > > ;-)
> > >
> > > Good luck both of you on working out the best course of action here.
> > > Personally Im getting a lot out of following developments in these
> > > early days, as a learning experience.
> > >
> > >
> > > On 17 Jan 2007, at 09:50, kiroukou alapla wrote:
> > >
> > >
> > >
> > > It has never occurred to me that we are changing direction to
> > > follow Sandy.
> > >
> > >
> > > No, but their features will become very similar...
> > >
> > > I believe we simply want to expand Papervision3D in a number of
> > >
> > > ways, so we would be able to build what we have dreamed of.
> > >
> > > I don't think using matrices and hierarchy transformations is
> > > changing direction at all. Papervision3D will be faster and just
> > > as accurate. In fact, if we want to make a space game we will
> > >
> > > need features such as far plane clipping, which Sandy already has.
> > >
> > >
> > > I'm indeed developping such features for weeks. It is a very
> > > important algorithm in 3D development.
> > >
> > > Obviously, those features aren't unique to Sandy. I don't think
> > >
> > > it has been our reference when developing Papervision3D. We have
> > > looked at PlayStation and Quake instead.
> > >
> > >
> > > I never said that you looked at Sandy, and that those features are
> > > unique to Sandy neither ;-) I did 3 3D engines in AS1, AS2,.. before
> > >
> > > deciding to share my experience with people through an OpenSource
> > > project. My experience allowed to be orient directly my work on this
> > > kind of architecture. It wasn't my solution, but a common solution to
> > >
> > > make a GOOD 3D engine!
> > >
> > > On the other hand, I proposed to join our efforts to build a new
> > > AS3 3D engine back in October, and haven't heard from you since.
> > > In the meantime, we have developed a very fast AS3 engine that
> > >
> > > has an enormous amount of possibilities.
> > >
> > >
> > > Yes your engine has a lot of potential. I totally agree with that.
> > > But the features you are trying to add, I've done them months ago!
> > > That's why I think this is a bit a shame...
> > >
> > > I've always been agree to join our efforts on a common AS3 project.
> > > Sorry if I misunderstood something in your mail you refer to.
> > >
> > > Anyone is more than welcome to join us, and I'll be really happy
> > >
> > > to count on someone of your experience to work with.
> > >
> > > I'm sure your contribution would be invaluable.
> > >
> > >
> > > THat's something I'm allready thinking on. But as John (btw thank
> > > your for your words, and all my respect for your great work for the
> > >
> > > Flash community!) pointed out, my project as some nice features too,
> > > a documentation, some tutorials, and people are using it in
> > > commercial production... and it is not that easy to decide to stop
> > > everything.
> > >
> > >
> > > As Andre and De'Angelo are saying, your API is different and simple,
> > > so very interessting for many people. But now you are going to make
> > > it grow into a more complex engine, this good point might change, and
> > >
> > > both libraries will be quite complex to learn (in fact, 3D requires
> > > some learning, that's why people are a bit frustrated with Sandy =) ).
> > >
> > > As you don't seems to be ready for a merge, I've to think about
> > >
> > > leaving my baby as John said, and that's not that easy.
> > >
> > > Cheers from France to all flash 3D lovers ;)
> > > Thomas PFEIFFER
> > >
> > > On 16/01/07, *kiroukou* <
> > > kiroukou at gmail.com
> > > <mailto:kiroukou at gmail.com>> wrote:
> > >
> > > Hi Carlos,
> > >
> > >
> > >
> > >
> > >
> > > Due to its humble beginning, a cheap and fast solution based
> > > on the cosine law was chosen. At the time, Paper was Flash 7
> > > and it seemed more than enough for the purpose.
> > > Unfortunately this approach has a lot of limitations and
> > >
> > > contrary to what I believed it actually requires more
> > > calculations.
> > >
> > > That's why Sandy already have this matrix and tree 3D
> > > structure since its first release. That's why I asked you to
> > > unify our efforts to create a great 3D library.
> > > At the beginning PV3D had its own system, very restrictive,
> > >
> > > but fast and accurate. So their goals were differents, and
> > > people jsut had to choose between our solutions depending on
> > > wht they want.
> > >
> > >
> > > But now you are thinking to develop PV3D in the same
> > >
> > > direction as Sandy, and I don't understand (once again) why
> > > to do that.
> > > Sandy has a very slow development because I'm alone to
> > > develop it, and in my rare free time (I'm an enginneer, but
> > >
> > > not in flash, in aerospace). But with the power of the
> > > community, a library with a flexible licence as Sandy has,
> > > it can grow up very rapidly.
> > >
> > >
> > > Yes the API is a bit different, but its just method or
> > >
> > > classes names. And it can change :)
> > >
> > >
> > > Anyway this was my first public request to join our efforts
> > > after our private ones. I'm doing this once again because I
> > > see that our projects are going to be almost the same, but
> > >
> > > with a different name and team.
> > >
> > >
> > > Great work anyway, and just an advice, prefer matrix
> > > operation directly rather then quaternion (at least in AS2),
> > > it is faster (because a matrix converssion of your quaternion
> > >
> > > is required during your transformation chain).
> > >
> > >
> > > ++
> > > Thomas PFEIFFER aka Kiroukou
> > >
> > > http://www.flashsandy.org <http://www.flashsandy.org/>
> > >
> > > _______________________________________________
> > > Papervision3D mailing list
> > > Papervision3D at osflash.org
> > > <mailto:Papervision3D at osflash.org>
> > >
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> > >
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Papervision3D mailing list
> > > Papervision3D at osflash.org
> > > <mailto:Papervision3D at osflash.org>
> > >
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> > >
> > <http://osflash.org/mailman/listinfo/papervision3d_osflash.org>
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Papervision3D mailing list
> > > Papervision3D at osflash.org
> > > <mailto:Papervision3D at osflash.org>
> > >
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> > >
> > >
> > ------------------------------------------------------------------------
> > >
> > > _______________________________________________
> > > Papervision3D mailing list
> > >
> > > Papervision3D at osflash.org
> > >
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> > >
> > >
> > > _______________________________________________
> > > Papervision3D mailing list
> > > Papervision3D at osflash.org
> > >
> > >
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> > >
> > >
> > >
> > >
> > >
> > > _______________________________________________
> > > Papervision3D mailing list
> > > Papervision3D at osflash.org
> > >
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> > >
> > >
> > >
> >
> >
> >
> > --
> > [  JPG  ]
> > _______________________________________________
> > Papervision3D mailing list
> > Papervision3D at osflash.org
> > http://osflash.org/mailman/listinfo/papervision3d_osflash.org
> >
> >
> >
>
>
> --
> Ralph Hauwert
> FlashCoder
>
> _______________________________________________
> Papervision3D mailing list
> Papervision3D at osflash.org
> http://osflash.org/mailman/listinfo/papervision3d_osflash.org
>
>
> _______________________________________________
> Papervision3D mailing list
> Papervision3D at osflash.org
> http://osflash.org/mailman/listinfo/papervision3d_osflash.org
>

-- 
Ralph Hauwert
FlashCoder