The next step...

[ElChristou]

Hello everybody,

Well 1.4 is almost here with some really fantastic improvements since the 1.3 serie, so I was wondering if there is some kind of "work line" for the next steps until reaching the 1.5...

What will be the priority of next version?
What kind of improvements are in the target?
General debbugging, graphical improvements, UI improvement, else?

[Cham]

The depth sorting bugs occuring with models having transparent parts MUST be corrected! In my opinion, this should be a priority.

There are TONS of special effects (on black holes, pulsars, nebulae, spacecrafts...) which cannot be done just because of this bug.


Also, visible orbits with all binaries (double star systems) should be implemented.

[jll]

I fully agree with Cham, ...

but I am undoubtedly influenced by my nebula passion

Jll

[Malenfant]

I'd say the things I'd like to see most are:

1) realistic photometric functions for planets
2) correct multiple lighting illumination on planets
3) visible orbits for objects orbiting barycentres (including stars)

[Unfound]

Forgive my undoubtedly asinine query, but is it confirmed that 1.4 will be realeased? If so, I think I might wet my pants.

[ElChristou]

Cham, guys, I agree, those points are important, but in fact I just wanted to know if there is yet (or not) any kind of plan for the near futur...
One can create a special feature request thread to select the most important things from the users point of view, but here I am interested in the dev team point of view...

I don't know exactly how is born the FT version, but I can suppose that Fridger has been very interested in the Toti proposal of galaxy rendering, and we all know the result of this cooperation.
Now in fact I'm a bit worry because IF we have to wait some talented Toti like user/coder to create such synergy, I'm afraid to see the dev falling again in an unpleasant stagnation...

I don't know why, I have the feeling that the graphical improvement is not the priority of Chris (I can imagine it's a very big work), so I was wondering what would be the possibilities for next step...

[chris]

To answer a bunch of the items brought up here:

Forgive my undoubtedly asinine query, but is it confirmed that 1.4 will be realeased?

Yes, I'm building a Windows package right now and will have it ready within an hour or two. It will be 1.4.0pre8; if there are no serious regressions from 1.4.0pre7 or FT-1.2, it will become 1.4.0 final.

I don't know why, I have the feeling that the graphical improvement is not the priority of Chris (I can imagine it's a very big work), so I was wondering what would be the possibilities for next step...

That's not true at all. Some of the biggest items that I want to include in the next version fall into the category of graphical enhancements:

- Much more realistic atmosphere rendering that simulates Rayleigh and Mie scattering.
- Improved star rendering, with better correlation between the brightnesses of stars and galaxies.
- Point sprites in cmod files; useful for some rough volumetric effects.

Other enhancements:
- Display of more star labels; switch from using a fixed list of star labels to showing labels for all stars above some threshold apparent magnitude
- Render orbits of stars in multiple star systems
- Better handling of barycenters for planet-moon systems

Those are my highest priority items right now.

--Chris

[chris]

I'd say the things I'd like to see most are:

1) realistic photometric functions for planets

It's not my highest priority, but I want to make some progress on this in the next version. It may only get to the 'experimental feature' stage though. We'll see.

2) correct multiple lighting illumination on planets

You and I disagree on what 'correct' is. I do want to address the problem, but in a more fundamental way than your proposal to make light source brightness phase dependent.

3) visible orbits for objects orbiting barycentres (including stars)

Yes. This is on my list as well.

--Chris

[Cham]

Chris,

please, solve that nasty depth sorting problem with models having transparent parts. it prevents us to make any special effects.

[chris]

The depth sorting bugs occuring with models having transparent parts MUST be corrected! In my opinion, this should be a priority.

Care to propose a solution?

Proper rendering of general translucent meshes is a huge challenge on today's graphics hardware. There are a few ways to deal with the problem, and none are completely satisfactory. You can depth sort the triangles, but this is slow and doesn't work for intersecting triangles. One some hardware, you can use 'depth peeling' for true order independent transparency. The technique requires complex shaders and multiple render passes, and is too slow except on high end hardware. I implemented this for a 4D renderer project that I was working on a few months ago.

There are TONS of special effects (on black holes, pulsars, nebulae, spacecrafts...) which cannot be done just because of this bug.

My hope is that point sprites in CMOD files will be useful for some of these things. To give some some idea of what I'm talking about, think of Fridger's and Toti's galaxy templates--these are just collections of several thousand point sprites.

--Chris

[Cham]

Then there is this "clipping" or "flashing" problem of transparents textures defined on planes, rotating on the surface of a star or planet. The textures are "blinking" constantly, while the observer change its relative position. This is an annoying problem.

Also, models placed on the surface of planets are experiencing some "jittering" problems. They are constantly jumping and moving above the planet's plane. Is there a solution for that?

[jestr]

Chris,it sounds like a lot of increased workload to sort the transparancy problem fully,but as a first step in helping with galaxy and nebulae models,would it not be possible to depth-sort the individual parts of a mesh properly,rather than just alphabetically(or however the constituent parts are listed in the file).With careful planning at the moment it is possible to get it looking OK from one side,but when you rotate around it,it is all inside out.Cheers,Jestr
PS.I like the sound of the sprites.

[BlindedByTheLight]

Chris,

I'm pretty sure you're probably already got the OS X build covered but, if not - to speed things up for the OS X users, I'd be happy to build it off CVS in the meantime and make it available.

Take care,
Steven

[t00fri]

Chris', new Avatar looks like

"on the way up to Mt. Rainier" Right?

Bye Fridger

[ajtribick]

Some simulation of dust discs would be nice, e.g. Zodiacal Light, Vega-type systems. I guess the point sprites/volumetrics thing you mentioned would be the way to go about doing this?

[dirkpitt]

I'm pretty sure you're probably already got the OS X build covered but, if not - to speed things up for the OS X users, I'd be happy to build it off CVS in the meantime and make it available.

I've updated the version number for the OS X Celestia to 1.4.0pre8, so the OS X version should be good to go straight from CVS. Compiling takes ages on my G3 iMac - give me back my PowerBook!

[Cham]

Dirkpitt, will you compile a version with the same splash screen as the one from ElChristou ? And a French interface, maybe ? Actually, I don't remember who made the version I'm using right now. I think it was you.

[chris]

Chris', new Avatar looks like

"on the way up to Mt. Rainier" Right?

Bye Fridger

Yes, that's the one! Around 4000 meters above sea level on the Ingraham/Emmons glaciers.

--Chris

[Malenfant]

One more thing I just remembered: a more efficient way to render asteroids. It'd be nice to be able to show all 200,000 or so asteroids that are known but right now doing that using a huge ssc file grinds Celestia to a halt.

If Fridger can import thousands of galaxies and binaries in to Celestia there must surely be a way to import the asteroids too, maybe as a binary-format file?

[Cham]

Oh, and if only there was bumpmap on asteroids too ! Is this too hard to implement ?