Feature Requests.

[Anon]

It wouldn't be possible as in "we would need to rewrite the shaderGenerator"?

[rdb]

No, it conceptually can't work with the concept of a ShaderGenerator as we have it now.

[Anon]

No, it conceptually can't work with the concept of a ShaderGenerator as we have it now.

Well yeah, what do you think I mean by a rewrite?
Anyway, if I were to make a replacement for a shader generator for my project, how (diffently) would you suggest to make it then?

[rdb]

A system that divides up all the required behaviour into individual components, and has a system to hook up all the different components together, and a way to specify how the results of different components blend together.
Then you can let each component generate their, and generate the glue code for the output of those components.

Perhaps an illustration will help explain the concept:
http://www.blender.org/typo3temp/pics/923052fa19.png

This is also the approach we might take with Panda 2 - ability to override behaviour by simply overriding the shader code for a certain component, or adding another component and specifying how it blends in with the existing default components.

So simply rewriting ShaderGenerator won't do the trick - it'd be turned into a whole conceptually different system.

[Craig]

No, it conceptually can't work with the concept of a ShaderGenerator as we have it now.

Well yeah, what do you think I mean by a rewrite?
Anyway, if I were to make a replacement for a shader generator for my project, how (diffently) would you suggest to make it then?

Like this: https://blueprints.launchpad.net/panda3 ... -generator
https://github.com/Craig-Macomber/Panda ... -Generator

It does already work, but its not feature complete.

Note: the thing RDB linked is a fine approach for single output fragment shaders, and possibly as part of a more complete system, but I don't really see that approach as being sufficient by itself.

[Anon]

I really dont like how the bloom filter works now. It affects everything by default and everything is just too bright. You can have a black glow texture on every single model you dont want to be affected by it, but usually their amount is greater than the models that are glowing, why not have it the opposite? Not have a model glow unless it has a glow texture? Or at least allow to set it to behave like that? The way you'll do it now is have a blank (black) texture on every single model that shouldn't glow. Thats not very nice.

[wezu]

You can do it with a material no need for a texture... and maybe with some options of the bloom fillter...the docs on that are a bit poor.

I'd have a request on pview to make it support common filters (or just bloom for glow).

[Anon]

You can do it with a material no need for a texture...

um, what?
Gloss or glow? Glow doesnt seem to depend on the Emissive value of the material.

and maybe with some options of the bloom fillter...

'maybe'?
I didn't find anything: http://www.panda3d.org/reference/python ... monFilters

[Prometheus]

I now have a standing offer: anybody who wants to add a small feature to panda can ask me for help, and I'll walk you through the process. I'll help you find the classes you need in the source tree, I'll explain how they work, I'll explain how to extend them - I'll even go so far as to work on the code side-by-side with you through VNC. My hope is that you'll get the hang of working on panda, and that in time, you'll become a panda developer yourself.

If you want a new feature but aren't willing to implement it, you can post it here, but be aware - Panda3D only has a few full-time developers, and they are very overworked. We generally don't have time to do requests.

Is this offer still standing?

I've been following the Panda project for some time, and I am thinking of implementing some sort of editor tool over the summer. I guess that doesn't really count as an extension... but I'm definitely interested in finding out how P3D can be improved.

[drwr]

Josh is no longer involved with the Panda team, so no, at least not from Josh's offer. But there are others of us who are also interested in helping you to become a Panda developer. We may not have as much time to offer as Josh was offering, but we'd still like to do what we can to help you get started.

Note that there are multiple competing projects to produce a general editor of some sort, and people seem to have different ideas as to precisely what that means. My personal favorite editor project can be found in direct/src/leveleditor.

David

[Prometheus]

That's too bad! He's still with CMU, right?

And yeah, I saw some of the ideas; I was thinking more of an... object or asset editor.

I guess this is already largely covered by tools like Maya and so on, but I had some ideas on how to make it more like a 'RAD' tool rather than the current slower process.

But anyways, that was just one idea.

I mean, some other ideas that come to mind -- silly or otherwise -- are multithreading/voxel graphics, gravity (e.g., how can zero-g be handled), and so on.

[powerpup118]

... are multithreading

Do keep in mind that threading is supported through an interface much like Python's threading module (found here)

(that is of course, if that's what you meant by multithreading)

Hope this helps,
~powerpup118

[flavio]

Is it possible to add to pdeploy the option to make a version of a Linux application which includes the needed libraries (i.e. not a deb file, but a compressed file with the necessary stuff - with some exceptions, like glibc)?

I was reading this post (but the entire thread is interesting, because it covers the Linux deployment), and this post answers to some doubts about the previous one; this is an article which covers the same subject, and the part 2 covers the specific part of executables.

The current pdeploy produces deb files, and this is the optimal solution when you can make assumptions about target systems. But, if you can't make assumptions (because you want to deploy to many different systems), you'll have some problems:

• http://www.gamedev.net/... section "The Problem"
• some distros could use different filesystem hierarchies, so you could have to provide different debs for different distros
• the user needs rights to install packages
• however you're missing non-deb distros

Previous points imply that with this kind of deployment you'll obtain:

• a simpler deployment (you've to maintain only one package (per-architecture))
• a more robust deployment (you provide the needed libraries and avoid bugs of specific versions of libraries installed by specific distros)
• a more complete deployment (at the moment you can target deb distros, but you would like to target even rpm distros, pacman distros, portage distros, ...)

I'm not against the current pdeploy method, I only think that this addition could improve very much the Linux support of pdeploy. Thank you!

[Anon]

Syntax highlight for [ code ] in the forums please

[GrizzLyCRO]

Syntax highlight for [ code ] in the forums please

You can easily use some javascript snippet in about every modern browser to achieve that. In opera its very easy, i used http://google-code-prettify.googlecode.com/svn/trunk/README.html.

And i agree that is too hacky, and i am all in for moving panda forums to some newer backend.

[Sothh]

We need a way to get the handle of the main window, so we can pass it to library's such as Tkinter.

[Sothh]

A few more request after playing around with UDK.

It would be very nice to have height fog.
http://udn.epicgames.com/Three/HeightFog.html

I really like the exponential height fog.

It would also be very nice to have a material file format. Such a format would include textures for bump, glow, diffuse, ect. and also shaders and the like. With such a format we could write a very nice material editor.

Also it would be nice to include textures in the .egg file. This could be used instead of the material format.

And, while I am at it, it would be REALLY REALLY nice to have a drag-and-drop/real-time shader creator/editor.

I would be willing to help with the coding of any editors and such.

And I know most of this will not happen...

[Anon]

I think existing fogs dont work with the shadergenerator yet, wouldnt it be better to fix existing stuff before adding new?

As for a shader editor, Craig was working on one. He was also working on his own shader generator.

[wezu]

I don't get the part about materials. In an egg file you can have materials defined just as textures. Moving that to a separate file has no sense to me. Just like putting textures in egg files - egg is a human-readable text file - a texture is an image (binary) file.

[Sothh]

I don't get the part about materials. In an egg file you can have materials defined just as textures. Moving that to a separate file has no sense to me. Just like putting textures in egg files - egg is a human-readable text file - a texture is an image (binary) file.

The idea is to include all textures, shaders, and the like in a single file.

[wezu]

I don't get the part about materials. In an egg file you can have materials defined just as textures. Moving that to a separate file has no sense to me. Just like putting textures in egg files - egg is a human-readable text file - a texture is an image (binary) file.

The idea is to include all textures, shaders, and the like in a single file.

Like a multifile? But I still don't understand what for. If I use one texture for 5 models then I'll have 5 copies of the same texture in 5 different files?

[Nemesis#13]

I don't get the part about materials. In an egg file you can have materials defined just as textures. Moving that to a separate file has no sense to me. Just like putting textures in egg files - egg is a human-readable text file - a texture is an image (binary) file.

The idea is to include all textures, shaders, and the like in a single file.

Like a multifile? But I still don't understand what for. If I use one texture for 5 models then I'll have 5 copies of the same texture in 5 different files?

No, rather one file, referenced in 5 egg files. Loading one of the eggs initiates the loading and caching of the texture, so the next time a model needs that texture, it's loaded in no time.

[lord gengoro kitsune]

1) A collision system that can work with procedurally generated meshes like GEOMIP Terrains , Craig Macomber's shader generated infinite terrains and things like procedural trees and grass. Right now it does NOT, you apparently HAVE to HAVE that whole "polyset keep descend" thing in an EGG for the geometry to even BE recognizable as an "into" object.

2) A better scheme for animated textures than EGG Texture Cards. Me I want to put animated textures onto real meshes rather than some useless square card divorced from everything. Ten years ago the folks at Flatland Online gave their users the ability to use GIF89a anims as textures complete with transparency, it wasn't a perfect implementation but it was highly useable. Heck I'd settle for being able to include the EGG Texture Card's "sequence node" in my Blender generated EGG's. Yes I could use AVI's (maybe??) but more than a few would be a performance killer.

3) better keyboard and gamepad/joystick support ... seems ridiculous to have to drag in ALL of Pygame just to read my Xcontroller, not to mention those four totally unreadable numpad keys ( "/", "*", "-", "+" )

. This is the fourth time I've posted in a long time ... after being roundly castigated by a bunch of "Ivy League" "professional developers" over my favorite choice of text colors ( 1.0, 0.0, 1.0, 1.0 ) over black and my stunning lack of professionalism... GODALMIGHTY what would they have said if my website provider allowed me to use my favorite font too (Mistral)? What I'm trying to do isn't about money to me its about supporting creativity which moneygrubbing VR sites like Second Life care little about.

[Craig]

1) A collision system that can work with procedurally generated meshes like GEOMIP Terrains , Craig Macomber's shader generated infinite terrains and things like procedural trees and grass. Right now it does NOT, you apparently HAVE to HAVE that whole "polyset keep descend" thing in an EGG for the geometry to even BE recognizable as an "into" object.

I actually have full collision working with my terrain now (did it this week), it wasn't very hard either. Its less than ideal generation speed wise, but faster than the visible mesh generation (at least for the trees that use collision tubes). I'll push some public updates soon. I also got pre-generating and caching to bam files and async loading, so no need to use slow/big egg files.

2) A better scheme for animated textures than EGG Texture Cards. Me I want to put animated textures onto real meshes rather than some useless square card divorced from everything. Ten years ago the folks at Flatland Online gave their users the ability to use GIF89a anims as textures complete with transparency, it wasn't a perfect implementation but it was highly useable. Heck I'd settle for being able to include the EGG Texture Card's "sequence node" in my Blender generated EGG's. Yes I could use AVI's (maybe??) but more than a few would be a performance killer.

You can use 3D textures for this, or MovieTextures, or Pointer Textures. 3 ways seems like enough to me, but you can also use texture arrays (new!), sequence the shader inputs through different textures manually, or UV pan a set of frames in a texture. Personally I've only used 3D textures, but the other approaches should work fine.



There are 3 features I would really like:
- A method I can use to get collision geometry generated from visible geometry (like colliding with geoms generates. My python version is slow, and knowing there is a C++ version and no way to call it makes me sad)
- A method to get visible geometry (geoms) from collision geometry (like showing collision generates) (I want this for dumping my collision data to a mesh to generate nav meshes)
- A method of ShaderAttrib to list all shader inputs (I need to find them to serialize the parts of ShaderAttribs I need to keep, currently I have to iterate through a list that contains the names of all the shader attributes I might use and see if they are present on every geom!)

[powerpup118]

It's really a shame fog is yet to be working with the shader generator, I'd say that's top on my list of wanted things.

volumetric fog would be even more amazing, I know I'm shooting for alot here, but if someone could implement volumetric fog for the shader generator, I'd be willing to.. kill.. for that.

Kidding kidding, I seriously would love for panda3d to have some sort of fog that could be used with the shader generator though.

~powerpup118

[lord gengoro kitsune]

So it IS possible to do collisions, without resorting to using some other "physics engine" as this seemed to be my only thought, ODE and Bullet have been successfully used w/ Panda and I was sure I'd have to do something like that after I traced Panda's collision code into the C++ side of things. A place I cannot go, my skills with ANY of the C languages are S^^^fully bad, in fact C, C++, and Windows were what ENDED my carreer as a "professional" MSDOS / FORTH programmer back in the late 80's.

As to the texture animation 3d textures seemed to be the way I was going to go with it which would allow me to apply sequences of either JPG or PNG files at a much higher quality level than the old GIF 89a anim allowed for. To apply them I will have to do my first attempt at actually writing a shader from scratch, wish me luck.

Curious if you know anything about my vid chip related issues, Animate Dream's slope texturing demos ( even the early one ) do not run correctly sometimes not at all and neither dos your GEOCLIP grass. I've managed to determine my ATI Radeon HD3200 seems permanenently stuck in single precision mode. Any ideas, there was a whole thread about a year ago as I remember concerning ATI cards.

[Craig]

So it...

Nothing in this post relates to the thread. I won't answer any of it here to avoid further distracting a useful thread. Post such things elsewhere or not at all.

[redpanda]

I seriously would love for panda3d to have some sort of fog that could be used with the shader generator though.

+1

[Craig]

I seriously would love for panda3d to have some sort of fog that could be used with the shader generator though.

+1

You can do fog as a post process (aka filter) and thats compatible with the built in shader generator. Anyway, since there is interest, I made fog for my shader generator (not as a post process) as another example effect: https://github.com/Craig-Macomber/Panda ... -Generator

Discuss my shader generator here if you wish to avoid derailing this thread: http://www.panda3d.org/forums/viewtopic.php?t=10758

[Sothh]

I would like low level OpenGL access; namely the ability to create and color primitives such as triangles and points.