the 1.4prev9 bug report thread

[ElChristou]

GUYS, as apparently nobody is able to centralize the information, I'll do it... (I'm tired to see lost users reporting the same bugs for 1.4prev8 in various threads when 1.4prev9 is already available!!)

So First to download 1.4prev9 for windows, see:
http://www.shatters.net/forum/viewtopic.php?t=6098&start=24

Then you are welcome to post any bug for 1.4prev9 below.

Bye.

[Boux]

GUYS, as apparently nobody is able to centralize the information, I'll do it... (I'm tired to see lost users reporting the same bugs for 1.4prev8 in various threads when 1.4prev9 is already available!!)

So First to download 1.4prev9 for windows, see:
http://www.shatters.net/forum/viewtopic.php?t=6098&start=24

Then you are welcome to post any bug for 1.4prev9 below.

Bye.

Excellent initiative!
I was wondering why the devs did not start a new thread to collect impressions, bug reports, fps rendering figures, how the kitchen sink is behaving, you name it...
I am not a windows guy but anyhow.
OK, let us poor mortals do it.
There have been reports about the opengl 2 path being broken with some recent Nvidia drivers (latest beta and official).
As far as Linux is concerned, I had to regress to driver 7676 to have a functional opengl 2 path and get rid of the red Earth (and some other planets).
That is one bug solved which has nothing to do with Celestia I guess.
Otherwise, everything seems to work well here (fresh KDE build from latest CVS tree).
Ah yes, binary cmods still don't show up. But it has been implicitely suggested before that the fault is necessarily on my end (still wondering what I have done or not done properly!)

[t00fri]

Well, the "devs" didn't doit it based on the "working hypothesis" that there won't be any further bugs...

Here is another buggy that I was able to confirm. I am looking into it:

http://www.celestiaproject.net/forum/viewtopic ... highlight=

Bye Fridger

[t00fri]

Hi all,

OK I located the reason for the sudden extinction of NGC 1007. The crucial octree size parameter DSO_OCTREE_ROOT_SIZE
was just a little too small and thus the furthest galaxies in my deepsky.dsc catalog will be extinguished upon approaching them.

I will commit a fix to CVS and let you know when people may download the (corrected) code for recompiling it.

Bye Fridger

[Brendan]

My new Windows XP system has an ATI RADEON XPRESS 200 card and I got the lastest drivers from the ATI site and prerelease 9 of Celesita. In the OpenGL 2.0 render path, Earth's oceans are white where the specular map is white. For Iapetus, it is totally white. Pluto and Charon are also white where their specular maps are white.
Specular maps look fine in the OpenGL vertex program mode.

Here's my graphics info:

Code: Select all

Vendor: ATI Technologies Inc.

Renderer: RADEON XPRESS 200 Series SW TCL x86/MMX/3DNow!/SSE2

Version: 1.4.5523 WinXP Release

Max simultaneous textures: 8

Max texture size: 2048

Point size range: 1.000000 - 64.000000



Supported Extensions:

GL_ARB_multitexture

GL_EXT_texture_env_add

GL_EXT_compiled_vertex_array

GL_S3_s3tc

GL_ARB_depth_texture

GL_ARB_fragment_program

GL_ARB_fragment_program_shadow

GL_ARB_fragment_shader

GL_ARB_multisample

GL_ARB_occlusion_query

GL_ARB_point_parameters

GL_ARB_point_sprite

GL_ARB_shader_objects

GL_ARB_shading_language_100

GL_ARB_shadow

GL_ARB_shadow_ambient

GL_ARB_texture_border_clamp

GL_ARB_texture_compression

GL_ARB_texture_cube_map

GL_ARB_texture_env_add

GL_ARB_texture_env_combine

GL_ARB_texture_env_crossbar

GL_ARB_texture_env_dot3

GL_ARB_texture_mirrored_repeat

GL_ARB_transpose_matrix

GL_ARB_vertex_blend

GL_ARB_vertex_program

GL_ARB_vertex_shader

GL_ARB_window_pos

GL_ARB_draw_buffers

GL_ATI_draw_buffers

GL_ATI_envmap_bumpmap

GL_ATI_fragment_shader

GL_ATI_separate_stencil

GL_ATI_texture_env_combine3

GL_ATI_texture_float

GL_ATI_texture_mirror_once

GL_ATI_vertex_streams

GL_ATIX_texture_env_combine3

GL_ATIX_texture_env_route

GL_ATIX_vertex_shader_output_point_size

GL_EXT_abgr

GL_EXT_bgra

GL_EXT_blend_color

GL_EXT_blend_func_separate

GL_EXT_blend_minmax

GL_EXT_blend_subtract

GL_EXT_clip_volume_hint

GL_EXT_draw_range_elements

GL_EXT_fog_coord

GL_EXT_framebuffer_object

GL_EXT_multi_draw_arrays

GL_EXT_packed_pixels

GL_EXT_point_parameters

GL_EXT_rescale_normal

GL_EXT_secondary_color

GL_EXT_separate_specular_color

GL_EXT_shadow_funcs

GL_EXT_stencil_wrap

GL_EXT_texgen_reflection

GL_EXT_texture3D

GL_EXT_texture_compression_s3tc

GL_EXT_texture_cube_map

GL_EXT_texture_edge_clamp

GL_EXT_texture_env_combine

GL_EXT_texture_env_dot3

GL_EXT_texture_filter_anisotropic

GL_EXT_texture_lod_bias

GL_EXT_texture_mirror_clamp

GL_EXT_texture_object

GL_EXT_texture_rectangle

GL_EXT_vertex_array

GL_EXT_vertex_shader

GL_HP_occlusion_test

GL_NV_blend_square

GL_NV_occlusion_query

GL_NV_texgen_reflection

GL_SGI_color_matrix

GL_SGIS_generate_mipmap

GL_SGIS_multitexture

GL_SGIS_texture_border_clamp

GL_SGIS_texture_edge_clamp

GL_SGIS_texture_lod

GL_SUN_multi_draw_arrays

GL_WIN_swap_hint

WGL_EXT_extensions_string

WGL_EXT_swap_control



[fsgregs]

Chris:

I just downloaded the Windows version of 1.4.0pre9 and was excited to explore everything, when ... I tried to cycle through the rendering options with Ctrl+V.

Celestia crashed 100% of the time ... program error screen.

My system is a Pentium IV 1.6 GHz, with 1 GB RAM, 120 GB Hard Drive, Nvidia GeForce 4 TI 4200 video card with 128 MB VRAM and Nvidia driver set 7801.

Important to note that when I run Fridger's FT 1.2 Windows version from about two weeks ago, I get no such crash and it cycles through Basic to Open Gl/Nividia Combiners. The crach is, therefore, due to something that has crept into pre9. I haven't tested out pre8 on my machine, but I will if you feel it would do some good!

PLEASE FIX!!!

Frank

[fsgregs]

Found another minor bug. 1.4.0pre6 and 7 toggled labeling for both galaxies and nebulae when the "E" key was pressed. Nebulae labeling no longer appears in 1.4.0pre9. Is there a new key for it?

Frank

[fsgregs]

Sigh ... found another bug in 1.4.0pre9. The date cannot be changed manually through the time menu. Opening the time menu and trying to type a new date has no effect. The time fields do work, but the date field is frozen.

Frank

[Brendan]

To my report, I add that it is Earth's oceans that is white in the OpenGL 2.0 render path when the clouds are on. When the clouds are off, Earth is totally white. On Pluto and Charon, the areas where the specular maps are white become white on the planet, and the correspondence is rough. It looks like there is no specular texture in the default installistion for Iapetus, but a specular color and power.

[Boux]

Well, the "devs" didn't doit it based on the "working hypothesis" that there won't be any further bugs...
...
Bye Fridger

Yep, no software without bugs. Fix one of them, get a couple brand new ones for free

[selden]

This is to confirm Frank's report for Windows Celestia v1.4.pre9:
the date fields of of the time setting window cannot be selected. They work ok in pre8.

Also, the variables related to NebulaLabels are referenced only in the gtk and kde routines. They're missing from the Windows routines.

Also, pre9 still crashes for me when using Ctrl-V to step from the OpenGL 2.0 render path to the Basic rendering path.

System:
1GB 3.4GHz P4-550; WinXP Pro sp2
128MB GF 6600GT; ForceWare v81.87
Celestia v1.4.0pre9

[jgrillo2002]

I have encountered a severe problem with Celestia. Whenever I right click on the celestia program. it CTDs to the desktop. here are my current specs of the computer

Processor
Intel Pentium 4, 2800 MHz
200MHz external bus

BIOS
Intel Corp. BF86510A.15A.0045.P08.0309261344 09/26/2003

Memory
1536MB physical
90% free resources
4 memory slots, 0 free (256+512+256+512)

Video
Graphics Adapter: NVIDIA GeForce FX 5500
Screen Area/Colors: 1024x768 pixels, 16 million colors
Monitor: Default Monitor
Monitor: Default Monitor
Monitor: Default Monitor
Monitor: Gateway EV700B

Drives
C:\ (NTFS) 114470MB total, 31411MB free
D:\ CD-ROM/CD-RW or DVD Drive
E:\ CD-ROM/CD-RW or DVD Drive
F:\ (NTFS) 114470MB total, 75582MB free

Network and Modem
Upgrade
(net) Intel Intel PRO/100 VE Network Connection
(net) Microsoft Direct Parallel
(net) Applied Networking Inc. Hamachi Network Interface
(modem) Intel Intel 537EP Data Fax Modem

Operating System
Upgrade
Windows XP SP2 (Windows XP 5.1.2600 Service Pack 2)
Original setup date: 1/9/2005
DirectX Version: 5.03.2600.2180 (xpsp_sp2_rtm.040803-2158)
Swap file managed by Windows
File cache managed by Windows

Browser
MSIE 6.0; 5.1; SV1; .NET CLR 1.1.4322; .NET CLR 2.0.40607
Proxy: No proxy set
HTTP level: direct 1.1, proxy 1.0
Scripting version: 5.6.0.8825

[LWK]

I have almost exactly the same problem as Brendan's:

every planet with a specular map is totally white (overbright, as if lightened by a -really- powerful light from the sun) where the specular map should be, like earth's oceans.

I have Windows XP pro with Celestia 1.4.0pre9,
1024Mb RAM, ATI Radeon X800SE 128Mb. ---> It seems related to ATI cards.

PS: I had the exact same bug also with the 1.4.0preFT1.2 installed.

Bye, LWK.

[MKruer]

High Precession Rates lead to bad orbit projection
http://www.celestiaproject.net/forum/viewtopic.php?t=6670

[jdou]

Am I lucky ?
I use "BlueMarble DDS WComer" with "Harald Schmidt 32k DDS specular maps" and I have not your problems with OpenGL paths or planets rendering.
Perhaps because I use NVIDIA drivers 78.01 (I read in this forum some bugs reported about new NV drivers releases )

Date and time can be changed with the arrows keys

[t00fri]

Am I lucky ?
I use "BlueMarble DDS WComer" with "Harald Schmidt 32k DDS specular maps" and I have not your problems with OpenGL paths or planets rendering.
Perhaps because I use NVIDIA drivers 78.01 (I read in this forum some bugs reported about new NV drivers releases )

Date and time can be changed with the arrows keys

I also don't have any significant problems under Windows with 1.4.0pre9. Also using the 78.01 NVIDIA driver for my Geforce 5900Ultra FX card. Also NO crashes with CTRL+V cycling operations. Why do people use buggy drivers??

As to setting the date, there is also no problem, except that there is a dark cursor type bar visible right over the portion of the date that may be changed. So I cannot see the previous date underneath that dark bar....

+++++++++++++++++++++
I notice however, that Chris still has my binary star data files 'visualbins.stc' and 'spectbins.stc' commented out in celestia.cfg, with the comment that there are still conflicts with nearstars.stc. Since after eliminating all known conflicts in the CVS archive, the understanding was that these binary data should be part of 1.4.0 final, I presume that Chris unintentionally used the old celestia.cfg file from 1.4.0pre7, where these conflicts indeed existed.
+++++++++++++++++++++

Bye Fridger

[cartrite]

Last night, using 1.4-pre9, celestia crashed using ctrl+v going from opengl 2.0 to basic. Next time I'll save the bug report if someone needs it. But it seems this bug is well known. I use GeForce FX 5500, Driver 78.01.

cartrite

[brokfn]

Hi,

Just downloaded 1.4.0pre9 on my main machine (XP2600, Windows XP, Radeon 9600, latest drivers - downloaded yesterday).
The earth's oceans display white with opengl 2.0.
No crash.
No galaxy either !

Bye,
Brokfn

[t00fri]

Let me just remind everyone who is observing crashes with CTRL+V cycling and|or does not see galaxies that it is paramount to download the complete 1.4.0pre9 release! If you just use the executable and stick it into an older celestia dir the consequences are UNPREDICTABLE!

Notably you will get CTRL+V crashes if the file
ell_galaxy_arb.vp

happens to be missing in the shaders directory.

Also people who don't see galaxies must

1) switch galaxy rendering ON, in the options dialog!!!
2) make sure that the galaxy templates (*.pts) properly exist in the models directory.

While in FT1.2 I had switched on galaxy rendering by default, Chris did presumably NOT do it in 1.4.0pre9...

Bye Fridger

[t00fri]

Last night, using 1.4-pre9, celestia crashed using ctrl+v going from opengl 2.0 to basic. Next time I'll save the bug report if someone needs it. But it seems this bug is well known. I use GeForce FX 5500, Driver 78.01.

cartrite

NO, NOT with FX cards using the 78.01 driver. I use it for my FX 5900Ultra (see above) without any crashes.

Bye Fridger