lghassen / cosmographia Goto Github PK

Bug was filed against nVidia for hang issue, seems to work fine from source and 
build. Whats up? Contact me so I don't close it out.

1. Load the Messenger catalog file
2. Go to Messenger
3. Enable shadows

The program will crash in MeshGeometry::renderShadow when there is an unchecked 
access to the materials array with an extremely large material index (looks 
like 2^31-1). There should be a check in renderShadow, but the bad material 
index shouldn't be there to begin with.

What steps will reproduce the problem?
1. download source
2. build
3. get error

What is the expected output? What do you see instead?
build fail : file not found
data/help/sun.html, mercury, venus ... etc.
and add-on(under data/addons/...)

What version of the product are you using? On what operating system?
source / mac os x / 10.7 (lion)

Please provide any additional information below.

Due to limited accuracy in the Saturnian satellite ephemerides, Cassini can be 
sometimes observed to crash into a moon instead of performing a close flyby, 
e.g. the Aug. 11, 2008 flyby of Enceladus which had a close approach distance 
of 25km.

To do:
- Compare the TASS 1.7 theory with the latest SPICE kernels and verify that the 
problems aren't due to an implementation error in TASS17.cpp.
- If there's no problem, investigate ways to improve ephemeris accuracy: 
possibly by extracting the Chebyshev coefficients for some of the problem 
satellites from a SPICE kernel and using them instead of the TASS 1.7 theory.

Apparent orientation interpolation problem when traveling to a destination in a 
straight line. It looks like a quaternion slerp is blowing up when the start 
and end orientation are identical.

What steps will reproduce the problem?
1. select Gears icon
2. move Star Brightness slider


What is the expected output? What do you see instead?

The starry background moves.
It should not.

This is especially noticeable if there's a planetary object in the foreground: 
it rotates.

What version of the product are you using? On what operating system?

r292
Win7 Pro SP1 X64

What steps will reproduce the problem?

1. Build using qtcreator and visual C++ compiler
2. while compiling Lib3ds, output produces warnings like:   warning C4273: 
'lib3ds_file_open' : inconsistent dll linkage



To resolve add DEFINES +=   LIB3DS_EXPORTS into cosmographia.pro file in lib3ds 
section 

What steps will reproduce the problem?
1. select the (?) icon
2. shrink Cosmographia's window to be smaller than the default help overlay size
3. the help display does not change size: the lower lines disappear from view.

What is the expected output? What do you see instead?

The help and other menu overlays should shrink in area (probably not change 
text size, though) to adapt to the visible area of Cosmographia's window.

With smaller windows, one cannot reliably view the lower lines of the overlay. 
Scrolling the help text will briefly bring the lowest lines of text into view, 
but they immediately scroll back out of view when the mouse wheel is released.

This could be a more serious issue when running full-screen on a low-resolution 
screen.


What version of the product are you using? On what operating system?

r292
win7 pro sp1 x64

What steps will reproduce the problem?
1. goto Earth (or other object)
2. enable frame axes
3. rotate view so one of the frame arrows points toward the viewpoint
4. rotate view so its arrowhead is just inside the edge of the planet

What is the expected output? What do you see instead?

The labels for axes being drawn in front of an object should be drawn in front 
of the object, too.

Instead, as the viewpoint is rotated slowly, one can see the axis label 
disappear behind the object.


What version of the product are you using? On what operating system?

Cosmographia r247

Win7 Pro X64 SP1


Please provide any additional information below.

What steps will reproduce the problem?
1. define two adjacent objects with a Mesh which loads a surface texture with 
an Alpha channel, either PNG or DDS
2. view the objects from various angles


What is the expected output? What do you see instead?

The object closest to the viewpoint should always obstruct the view of the more 
distant objects.

Instead, sometimes the more distant object obstructs the view of the closer 
object.


What version of the product are you using? On what operating system?

r231
Win7 Pro SP1 x64

Please provide any additional information below.

I'll try to attach a zip file demonstrating the problem. The objects show the 
problem more obviously when shadows are enabled. 

It'd be nice if Cosmographia's included textures were located and provided 
automatically if a specified texture is not found in a catalog's own directory 
tree. While one can explicitly specify the path to a texture, one can't depend 
on either Cosmographia's or the catalog's directory names and locations being 
the same for other installations of the program.

What steps will reproduce the problem?
1. Create a catalog with an object which specifies "sun.jpg" as its baseMap.
2. load the catalog and view the object.

What is the expected output? 
An object with the same coloration as the Sun

What do you see instead?
A blank sphere


What version of the product are you using? 
r231
On what operating system?
Win7 Pro SP1 x64

What steps will reproduce the problem?
1. Compile the latest source of cosmographia
2. Launch the executable
3. The program dies with Segmentation Fault

What is the expected output? What do you see instead?

See above

What version of the product are you using? On what operating system?

Latest source checked out from SVN (r821)
Operating system is Ubuntu 12.10

Please provide any additional information below.

I compiled the source as debug build. I also run the strace command to trace 
the system calls and signals.

Here follow the last output of "strace" before the segmentation fault. It 
appears that the program dies while reading the file "earth.atmscat"

--------------------------------------------
open("/home/daniele/installazioni/cosmographia/cosmographia-read-only/data/earth
.atmscat", O_RDONLY|O_CLOEXEC) = 22
fcntl(22, F_SETFD, FD_CLOEXEC)          = 0
fstat(22, {st_mode=S_IFREG|0664, st_size=5767244, ...}) = 0
fstat(22, {st_mode=S_IFREG|0664, st_size=5767244, ...}) = 0
fstat(22, {st_mode=S_IFREG|0664, st_size=5767244, ...}) = 0
read(22, 
"atmscatr\1\0\0\0\0\0\0A\311\200\2736\341\34267\364?\3377\232\231\231?"..., 
5767244) = 5767244
brk(0x4216000)                          = 0x4216000
brk(0x4797000)                          = 0x4797000
brk(0x4b97000)                          = 0x4b97000
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++
Segmentation fault
---------------------------------------------------------------------------

What steps will reproduce the problem?
1. specify a surface texture image type of * either in a catalog or in a Mesh 
material definition
2. load the catalog
3. observe no surface texture image being applied to the object

What is the expected output? 
a colorful object
What do you see instead?
the object's default materials

What version of the product are you using?
r231
 On what operating system?
win7 pro sp1 x64


Please provide any additional information below.

It'd be nice, for example, to be able to replace a jpg surface texture by a dds 
texture without having to edit a catlog or regenerate a model. 

Also, some modeling packages can't use some of surface texture file types that 
Cosmographia supports. It'd be nice not to have to edit their mesh outputs to 
specify the appropriate file type.
(Anim8or can't handle png or dds textures. The only filetype supported by both 
programs is jpg.)

What steps will reproduce the problem?
1. File > Save screenshot
2. Save As dialog comes up, unpopulated, and hangs
3. Wait to see if it comes alive, then X out (OK)

What is the expected output? What do you see instead?

Normal file save selection dialog

What version of the product are you using? On what operating system?

Latest trunk on linux

Please provide any additional information below.

((Guess that a desktop thread lock is held that shouldnt be))

What steps will reproduce the problem?
1. checkout
2. build

What is the expected output? What do you see instead?

Error in "thirdparty/Eigen/src/Array/Functors.h"

What version of the product are you using? On what operating system?

trunk on linux/ubuntu (latest) with qt sdk (latest)

Please provide any additional information below.

g++ -c -m64 -pipe -O2 -Wall -W -D_REENTRANT -DGLEW_STATIC 
-DEIGEN_USE_NEW_STDVECTOR -DQJSON_EXPORT= -DQT_NO_DEBUG -DQT_DECLARATIVE_LIB 
-DQT_OPENGL_LIB -DQT_GUI_LIB -DQT_NETWORK_LIB -DQT_CORE_LIB 
-I../QtSDK/Desktop/Qt/474/gcc/mkspecs/linux-g++-64 -I../cosmographia 
-I../QtSDK/Desktop/Qt/474/gcc/include/QtCore 
-I../QtSDK/Desktop/Qt/474/gcc/include/QtNetwork 
-I../QtSDK/Desktop/Qt/474/gcc/include/QtGui 
-I../QtSDK/Desktop/Qt/474/gcc/include/QtOpenGL 
-I../QtSDK/Desktop/Qt/474/gcc/include/QtDeclarative 
-I../QtSDK/Desktop/Qt/474/gcc/include -I../cosmographia/thirdparty/glew 
-I../cosmographia/thirdparty/curveplot -I../cosmographia/thirdparty 
-I/usr/X11R6/include -I. -I../cosmographia -I. -o GlareOverlay.o 
../cosmographia/thirdparty/vesta/GlareOverlay.cpp
In file included from ../cosmographia/thirdparty/Eigen/Array:28:0,
                 from ../cosmographia/thirdparty/Eigen/Geometry:8,
                 from ../cosmographia/thirdparty/vesta/Entity.h:19,
                 from ../cosmographia/thirdparty/vesta/Material.h:14,
                 from ../cosmographia/thirdparty/vesta/RenderContext.h:16,
                 from ../cosmographia/thirdparty/vesta/GlareOverlay.h:14,
                 from ../cosmographia/thirdparty/vesta/GlareOverlay.cpp:11:
../cosmographia/thirdparty/Eigen/src/Array/Functors.h:242:8: error: partial 
specialization of ‘Eigen::ei_functor_traits<std::greater<T> >’ after 
instantiation of ‘Eigen::ei_functor_traits<std::greater<float> >’ 
[-fpermissive]
../cosmographia/thirdparty/Eigen/src/Array/Functors.h:246:8: error: partial 
specialization of ‘Eigen::ei_functor_traits<std::less<T> >’ after 
instantiation of ‘Eigen::ei_functor_traits<std::less<float> >’ 
[-fpermissive]
make: *** [GlareOverlay.o] Error 1
make: Leaving directory 
`/home/jdp/src/cosmographia-build-desktop-Desktop_Qt_4_7_4_for_GCC__Qt_SDK__Rele
ase'
The process "/usr/bin/make" exited with code 2.
Error while building project cosmographia (target: Desktop)
When executing build step 'Make'

What steps will reproduce the problem?
1. get Cosmographia sources
2. get QTFFMPEGWRAPPER sources and put them under thirdparty 
3. edit cosmographia.pro to enable FFMPEG support
4. build

What is the expected output? What do you see instead?
Expected: program compiles
What I see: compilation errors
How to fix it:
- add "using namespace ffmpeg;" to the end of ffmpeg.h (see 
http://code.google.com/p/qtffmpegwrapper/issues/detail?id=2&colspec=ID%20Type%20
Status%20Priority%20Milestone%20Owner%20Summary%20Opened%20Closed%20Modified )
- add in -DSTDC_CONSTANT_MACROS to CXXFLAGS in cosmographia.pro
- move the methods "unsigned getWidth();" and "unsigned getHeight();" in 
QVideoEncoder into the public section (they are being used in Cosmographia.cpp)


What version of the product are you using? On what operating system?
latest SVN, Ubuntu 12.04 64bits

Please provide any additional information below.
QTFFMPEGWRAPPER does not seem to be maintained anymore ( 
http://code.google.com/p/qtffmpegwrapper/issues/list ). Issue 1 and 2 seems to 
be old issues. Therefore, it would probably be best if the package was imported 
into cosmographia and fixed there (that's only 5 source files: QVideoEncoder.h, 
QVideoEncoder.cpp, QVideoDecoder.h, QVideoDecoder.cpp, ffmpeg.h. The ffmpeg 
library itself should of course not be delivered with Cosmographia).

Purpose of code changes on this branch:
We are finding that it would be extremely useful to have the ability to define 
UTC times within the json catalog file. This enhancement allows the user to add 
the "UTC" suffix to the end of a time to specify that the given time is in UTC. 
Users can also add the "TDB"suffix to the end of times if they want, but since 
TDB is the default input time, this suffix is not necessary. 

After the review, I'll commit my change to:
/trunk

What steps will reproduce the problem?
1. Select and GoTo Earth
2. view Earth very close up (e.g. examine Cape Canaveral)
3. observe that the Earth is rotating buildings out of view
4. type on space bar

What is the expected output? What do you see instead?

The Earth's rotation should stop and "(paused)" should appear in the display at 
upper right.
Neither happens.

Selecting the menu item Time --> Pause does work.

When time has been paused, typing the space bar does not resume its flow, 
either.

What version of the product are you using? On what operating system?
r319
Win7 Pro SP1 x64

What steps will reproduce the problem?
1. insert an extra comma somewhere in a valid json catalog
2. load the catalog
3. observe the line number
4. go to that line in a text editor
5. observe no error at that location

A workaround is to use another json interpreter to read the erroneous catalog 
and provide an accurate line and character count.
e.g. 
python -mjson.tool < test.json > test_out.json
(which requires python v2.6 or later)

What version of the product are you using? 
r231
On what operating system?
win7 pro sp1 x64

Firefox
* Type "about:config" into the address bar
* Click through warning
* (To see others, enter "network.protocol-handler" into the filter)
* Right click to "New > String"
* Enter "network.protocol-handler.app.cel" (for example)
* Enter "/usr/bin/celestia" (example ubuntu/dpkg location)
+ For some interesting notes see 
+ http://code.google.com/p/chromium/issues/detail?id=18113

Chrome
* See http://code.google.com/p/chromium/issues/detail?id=18113#c5
* Desktop registry for Gnome or KDE
* gconftool-2 --set --type string /desktop/gnome/url-handlers/cel/command 
"celestia %s"

What steps will reproduce the problem?
1. specify something other than 1.0 as the version value in a catalog.
2. load the catalog
3. observe no error mesage

What is the expected output? 
A message complaining that this catalog version is unsupported

What do you see instead?
No error mesasge.


What version of the product are you using?
r231
 On what operating system?
Win7 Pro SP1 x64

What steps will reproduce the problem?
1. define an object with a surface texture file name which does not exist
2. load the catalog file
3. goto the object

What is the expected output? What do you see instead?

The Application Output in QtCreator dislpays a loadTexture message for the 
texture, but no message showing that it failed to load.


What version of the product are you using? On what operating system?

r231
Win7 Pro SP1 x64

What steps will reproduce the problem?
1. right click on a selected object
2. choose either frame or body axes

What is the expected output? What do you see instead?

A letter designating an axis should be shown in addition to the multi-colored 
arrows.

Instead, only colored arrows are drawn.

What version of the product are you using? On what operating system?

r231
Win7 Pro SP1 x64


Please provide any additional information below.

I have trouble remembering which color corresponds to which direction. 

(It was nice to discover that the axes point in the same as the directions 
Anim8or uses and match my long-ago choice of CMOD output orientation.)

What steps will reproduce the problem?
1. select Pluto
2. goto Pluto
3. select RMB popup menu item Plot Trajectory

What is the expected output? What do you see instead?

I expected to see either the trajectory of the barycenter or of Pluto going 
around the barycenter. Neither is shown. The same problem is seen with Charon.

What version of the product are you using? On what operating system?

r348
Win7 Pro SP1 x64

It'd be nice to get a list of keywords that aren't recognized when a catalog 
has finished loading.
It's too easy to make a spelling error that one keeps overlooking.

Certain combinations of particle system parameters will cause some particles to 
be drawn in obviously incorrect colors. These badly colored particles are 
always the oldest ones in the particle system It appears that the particle 
system's color map may not be getting properly initialized.

What steps will reproduce the problem?
1. create a catalog which defines a Mesh which is near the surface of a Globe 
(planet, moon or defined in the same catalog with the Mesh)

2. load the catalog and turn on Shadows and Eclipse Shadows

3. observe that no shadow is cast on the Globe.

What is the expected output?
A dark region on the Globe's surface

What do you see instead? 
The Globe's natural coloring, as illuminated by the Sun

What version of the product are you using? 
r231
On what operating system?
Win7 Pro SP1 x64

What steps will reproduce the problem?
1. open the on-screen help
2. not all help text is visible


What is the expected output? What do you see instead?

the on-screen help overlay has neither a slidebar nor top and bottom arrows (or 
carets) to indicate that there is help text available outside what can be seen.

It'd be nice if there were an explicit indication that more is available. Right 
now it's an "easter egg hunt" for a novice to discover that one can scroll the 
help text. It might not even be noticed that more is available if the text 
changes so that the bottom line isn't partially cut off.

What version of the product are you using? On what operating system?

r292
win7 pro sp1 x64