Celestia Forums

For those curious about progress on the 1.5.0 release of Celestia, I'm posting the current ChangeLog. This is not final; there are more improvements coming for 1.5.0.

--Chris

---------

* New translations: Ukrainian, Russian, Arabic, Dutch, Chinese
* Updated translations
* Separate i18n domain for constelation names
* Windows: Added support for i18n
* Fixed bugs with UTF-8 in object name completion
* Enabled new star rendering path: uses point sprite extension, new Gaussian glare textures
* Added dynamic star labeling
* Added support for non-power-of-two textures (newer graphics cards only; on older cards the textures will be automatically rescaled to the nearest power of two.)
* Mac: i18n improvements and bug fixes
* Mac: enabled support for hardware antialiasing
* Fixed crashes that occurred when Lua script errors were reported
* Made overlay textures work in GLSL path
* KDE: accept relative dir/files in --conf, --dir, and --extrasdir command line options
* KDE: fixed by ing deep sky objects info url where the default url was always used.
* Implemented event handlers for celx scripts
* Localized planet labels
* New extrasolar planets:
- XO-1
- HD 69830 system
- HD 164922 b, HD 66428 b, HD 99109 b, HD 107148 b, HIP 14810 b
- HIP 14810 c, HD 185269 b, Gliese 849 b
- Mu Ara e
* Revised orbits of many extrasolar planets to reflect new data
* Added stars for new extrasolar planets: GSC 92941-01657
* Mac: Show menu bar when moving mouse to top of screen in full screen
* Mac: added Japanese translation
* Fixed bug in cmod loader that caused binary cmod files to be read incorrectly on 64-bit machines.
* KDE: removed a dependency on STL support in Qt
* Mac: added missing 'Hide Others' menu option
* Mac: cursor now changes on draggable view boundaries
* Added Plutonian satellite names: S/2005 P1 -> Hydra, S/2005 P2 -> Nix
* Added 9 new numbered Saturnian satellites
* Fixed bug that caused time scale in celestiacore and simulation to get out of sync.
* Improved rendering of meshes in OpenGL 2.0 render path:
- Switched to local viewer model for better quality specular highlights
- Added support for specular and normal maps for meshes
- Enable rendering of eclipse shadows cast onto meshes
* Added a ReverseMouseWheel option for configuration file
* Enabled per-pixel specular lighting in OpenGL 2.0 path
* Made specular lighting work properly with normal maps
* Enabled separate specular color when rendering models in pre OpenGL 2.0 render paths
* Fixed gcc 4.1.1 build errors caused by bad constructor declarations in Star and CommandGotoLongLat classes
* Fixed bad calculation of specular exponent when converting materials from 3DS files.
* Implemented cloud shadows
* Changed eclipse shadow shaders to use a single interpolant for all shadows instead of one per shadow
* Made relative InfoURL relative to current add-on directory instead of main data directory.
* KDE and GTK: added ogg/theora video capture
* Fixed bug in eclipse shadow test; objects close to the surface of the lit side of a planet were incorrectly rendered with shadows
* Made OpenGL 2.0 render path respect the ring shadow render flag setting
* Replaced Hubble Space Telescope made with Cham's much improved version
* New RECONS data added to nearstars.stc
* Windows: generate an accelerate command only when the 'a' key is pressed without any modifiers.
* Improved atmosphere rendering in OpenGL 2.0 path; uses new ssc parameters to specify scattering and absorption coefficients
* Added Lunar-Lambert photometric model (linear combination of Lambert and Lommel-Seeliger) for much more realistic appearing dusty objects (e.g. the Moon)
* Modified eclipse finder dialog for i18n
* Display a '?' when glyph is not found in a font texture
* Fixed a bug in the octree culling code that was causing star culling errors
* Fixed ambient light in OpenGL 2.0 render path to match other render paths
* Updates to parser:
- Allow leading + (no effect, just symmetry with - )
- Allow numbers to begin with a decimal point (e.g. accept .12 as well as 0.12)
* KDE: improvements to popup menu--orbiting bodies are now sorted by type
* Added cmodsphere, a utility for producing cmod meshes from height samples regularly spaced in longitude and latitude.
* COPYING, controls.txt, start script, and guide can all be localized
* cel and celx scripting:
- added openclusters and cloudshadows render flags
- added location, nebulae, openclusers, and i18nconstellations label flags
* Established Barycentric Dynamical Time (TDB) as the time scale used internally by Celestia
- Convert to UTC when displaying the time
- Modified Set Time dialog to convert from UTC to TDB
* Added normal maps for cloud layers
* Allow smaller objects to cast eclipse shadows (e.g. can now see eclipse shadows from Miranda on Uranus)
* Added support for compressed normal maps (DXT5nm)
* Added capability to use different reference planes for orbits. Previously orbits were always referred to the equatorial plane of their parent body. Now, the J2000 ecliptic plane is also an option.
* Implemented a new depth sorting algorithm that fixes many depth sorting bugs.
* Rewrote orbit path rendering; orbits are now properly depth sorted
* Mac: middle mouse button implemented (option-click on one button mice)
* Mac: show nearest and brightest stars properly sorted in star browser
* Added spectrum2rgb tool for generating RGB asteroid colors from SMASS spectra
* Made star orbit paths visible

All sounds very good indeed

* Established Barycentric Dynamical Time (TDB) as the time scale used internally by Celestia

What's this all about?

I know the list isn't complete, but what about the new galaxies rendering, templates, etc ?

Does the non-power of two update also apply to VTs? Will it be possible to get rid of some of the resizing steps currently needed when breaking up the Blue Marble or other non-power of 2 textures?

Cham wrote:I know the list isn't complete, but what about the new galaxies rendering, templates, etc ?

I was just listing the items that have been checked into CVS already. Fridger and Christophe are still working on the galaxies.

--Chris

Malenfant wrote:

* Established Barycentric Dynamical Time (TDB) as the time scale used internally by Celestia

What's this all about?

Specifying that the internal time scale used by Celestia is the same as the one used by SPICE and Horizons. And using the appropriate conversions elsewhere, for example converting to UTC to display the time in the upper right corner.

--Chris

Talking about time in the top-right corner, is it possible to show a clearer format ? It's a bit cryptic.

Any plans to include Vincent's infotext patch (for ultraverbose onscreen descriptions?). I've found that to be really useful.

http://www.celestiaproject.net/forum/viewtopic.php?t=9123

If nothing else, is that still going to work with 1.5.0?

chris wrote:* Improved atmosphere rendering in OpenGL 2.0 path; uses new ssc parameters to specify scattering and absorption coefficients

What are the new ssc parameters? Will the old code still work, or will I have to manually alter all 20,342,340,987 of my addon planets?

PlutonianEmpire wrote:

chris wrote:* Improved atmosphere rendering in OpenGL 2.0 path; uses new ssc parameters to specify scattering and absorption coefficients

What are the new ssc parameters? Will the old code still work, or will I have to manually alter all 20,342,340,987 of my addon planets?

Of course, the old codes will still be working !

However, you may have to edit them, if you want to take advantage of the new capabilities.

Lots of improvements indeed!
The new depth-sorting algorithm is a breakthrough: way faster and better rendering throughout.

chris wrote:For those curious about progress on the 1.5.0 release of Celestia, I'm posting the current ChangeLog. This is not final; there are more improvements coming for 1.5.0.
--Chris
---------
*Rewrote orbit path rendering; orbits are now properly depth sorted
.....

Wow, Chris, a lot of improvements and more to come!
Regarding the above change, this means that I can see now all the orbits I switch ON, up to at a chosen distance from the central object (barycenter, star, planet, moon), as asked in this thread
http://www.celestiaproject.net/forum/viewtopic ... =ptolemaic

So the disappearing of orbits with distance will be avoided, choosing a bigger distance limit?
I truly hope so.
Thanks a lot anyhow for the enormous improvements.
Bye

Andrea

Cham wrote:I know the list isn't complete, but what about the new galaxies rendering, templates, etc ?

Since my essential new galaxy code is ready since a long time, I'll commit the stuff today to CVS, including the respective Changelog update. ElChristou and I have meanwhile a quite decent new set of PNG templates and a nice new MilkyWay custom template for that code to work with. Further little tunings of the code and the templates can always be done during the pre-1.5.0 testing phase.

Furthermore I'll also add some code for the display of labels involving MANY objects: i.e. deep-sky (stars, galaxies) and asteroids, implementing distance dependent transparency as proposed by ElChristou. This gives an additional qualitative feeling for the objects' distance, some 3d effect and sobers the screen display a bit.

Good that it's weekend...

Bye Fridger

thanks for all those good news !

I'm impatient to test new depth sorting algorithm with nebula meshes

JLL

Unfortunately, the depth sorting improvements are only for SSC objects.

The depth sorting procedure is unchanged for DSC Nebula models.

About the new depth sorting code with SSC objetcs, what do we have exactly ? For example, on my "Magnetic-Earth" addon, will I be able to see the particles paths correctly when I get close to the "Magnetic Earth" ? Currently with version 1.4.1, they tend to disappear or to show through the Earth like if they were on front.

chris wrote:For those curious about progress on the 1.5.0 release of Celestia, I'm posting the current ChangeLog.

Looks to be a great upgrade! And it seems i??ll finally get what i wanted most (...) - the local infoUrls

One small suggestion in that regard, though:
The InfoUrl attribute works with all the mimetypes the computer is set up to recognize, not just with http. So "Show Web Info" is actually a bit misleading: a better name for the Web Info menu would just be "Activate Link".

- rthorvald

rthorvald wrote:The InfoUrl attribute works with all the mimetypes the computer is set up to recognize, not just with http. So "Show Web Info" is actually a bit misleading: a better name for the Web Info menu would just be "Activate Link"

I don't know, "Activate Link" sounds like it's going to switch on a bomb or something. How about "Show More Info"?

dirkpitt wrote:"Activate Link" sounds like it's going to switch on a bomb or something. How about "Show More Info"?

Well, that only works if there are info to be had after clicking

- rthorvald

rthorvald wrote:

dirkpitt wrote:"Activate Link" sounds like it's going to switch on a bomb or something. How about "Show More Info"?

Well, that only works if there are info to be had after clicking

- rthorvald

Well, that's a problem neither for bodies (planets, astero??ds,...) nor for stars since the following default webpages are opened if no InfoURL is available in the .ssc definition :
- bodies : "http://www.nineplanets.org/[objectname].html"
- stars : "http://simbad.u-strasbg.fr/sim-id.pl?protocol=html&Ident=[objectname]"
I'm not taking into account the case of fictional addons here.

As for Deepsky object, since no default website is called in the Windows version for now, a solution could be to hide the "Info" (or" More Info"...) line in the Popup Menu if no InfoURL is available in the .ssc definition. That could be done by simply adding a line in the code :
- .../src/celestia/winmain.cpp (~ line 1614):