Win32 Executable from SVN

[cartrite]

Hi Cartrite,

I just tried your latest build (4493) on Vista x64, it shows the startup logo, but then.. nothing.
Build 4407 works just fine (I just changed the executable back to this version).

It worked on XP.
Dumb question here. Did you read the ATTENTION file? The one titled ATTENTION-READ_THIS_BEFORE_INSTALLING.txt? Did you change celestia.cfg?
cartrite

[CAP-Team]

Yes I did read (and applied) the changes in the config.
But I made an error in adjusting it , it's working now.

[Derek]

Hi Cartrite,

Are there still problems with your website as I've tried a number of times during this last weekend. I get to the site click to download the svn but nothing happens??

[cartrite]

Hi Cartrite,

Are there still problems with your website as I've tried a number of times during this last weekend. I get to the site click to download the svn but nothing happens??

I just checked both sites and they both worked. The new site had a very slow connection though. It was slower then dialup. The website icon here at shatters points to my old site and that had a good connection. Over 300 kbs. Apparently, the new site still has problems. I did try it a second time and the download speed was better. But still about half the download speed from my old site.
cartrite

[Reiko]

I apologize if this isn't the right thread for this bug report.

I'm using SVN 5412 and the asteroid.cms and roughsphere.cms models render extremely tiny to the point where you can barely see them. The strange thing is it only does this if you place them in another solar system.

The .cmod and .3ds models seem to render fine, it's just the .cms models that I've noticed this on.

[cartrite]

This sounds like you may need to add this line to your scc file.

Code: Select all

NormalizeMesh true

I'm not sure. It is supposed to be "true" by default but .cms may have been overlooked.
cartrite

[Reiko]

This sounds like you may need to add this line to your scc file.

Code: Select all

NormalizeMesh true

I'm not sure. It is supposed to be "true" by default but .cms may have been overlooked.
cartrite

Do I need to add that line to every scc file that uses .cms models? Also, do I need to add that line before each .cms reference in a scc file?
I ask because I have asteroid belts full of them.

[cartrite]

I'm sorry but I don't know how to answer that. It all depends on what your scc files look like. I suggest that you try to add it to one asteroid and see if it works. Then I guess, for now, you may need to add the line to every asteroid. But if adding the line to one asteroid gets that asteroid to work, then I suggest that you start a post in the bugs forum to catch Chris's attention. Maybe this can be fixed in the code and you don't have to change anything after the fix. That is if this is a bug.
cartrite

[Reiko]

I'm sorry but I don't know how to answer that. It all depends on what your scc files look like. I suggest that you try to add it to one asteroid and see if it works. Then I guess, for now, you may need to add the line to every asteroid. But if adding the line to one asteroid gets that asteroid to work, then I suggest that you start a post in the bugs forum to catch Chris's attention. Maybe this can be fixed in the code and you don't have to change anything after the fix. That is if this is a bug.
cartrite

I tried to add it to one asteroid and it didn't work. The .cms models look fine if I load them with a svn 4367 or older version of celestia.

[cartrite]

Post a portion of one of the scc files and I'll take a look.

[cartrite]

Reiko,
I did a quick ssc file that added an asteroid to Wolf 359. It used the roughsphere.cms as the mesh and all looked fine. I was using r4510 on Linux KDE though. I also did not have to add the "NormalizeMesh true" line.

Code: Select all

"Ida" "Wolf 359"
{
   # Data for Dactyl taken from:
   # Petit et al.: The Long-Term Dynamics of Dactyl's Orbit
   # (Icarus 1997: 130; 177-197) - linked in the InfoURL

   InfoURL "http://www.lpl.arizona.edu/~hurfordt/research/papers/Icarus130.pdf"

   Class "asteroid"
   Texture "asteroid.jpg"
   # Use the same color as for Ida--uncertain if this is correct
        Color [ 1.000 0.901 0.815 ]
   BlendTexture true
   Mesh "roughsphere.cms"
   Radius   0.7

   EllipticalOrbit
   {
   Epoch        2449228.2028  # 1993 Aug 28 16:52:05UT
   Period             0.96534 # stable 5:1 resonant orbit (Petit et al.)
   SemiMajorAxis     83.5     # stable 5:1 resonant orbit (Petit et al.)
   Eccentricity       0.13    # stable 5:1 resonant orbit (Petit et al.)
   LongOfPericenter 310       # stable pericenter at 90:E Ida longitude (Petit et al.)
   AscendingNode     90       # VERY approximate - chosen to place Dactyl north
                                   # of Ida's equator, as seen at time of Galileo encounter
   Inclination        8       # estimated at 7-9 degrees (Petit et al.)
   }

   UniformRotation
   {
   MeridianAngle   123       # place prime meridian facing Ida
   }

   Albedo 0.2
}


cms-test.jpg

cartrite

[Reiko]

Funny you used wolf 359 because that is what I just used to illustrate what I'm talking about.

This first one is with the 4367



This is with the 5412. Same file, same star etc.


In the first screenshot I am 1,684.4km from a 292.45km radius asteroid. That is with the 4367.

Look how close I am in the second shot. 38.9 meters from a 292.45km radius asteroid. That is with the 5412.

Same file, same cms model, different versions of Celestia.

[cartrite]

I can see the size difference. But I need to see the ssc file. Maybe I can see what you may need to change. It must have something to do with the radius you are using.

[Reiko]

ok

Code: Select all

"testroid" "wolf 359"
{
   NormalizeMesh true
   Mesh "asteroid.cms"
   Class "moon"
   Texture "gc_moons_02.jpg"
   Color [ 0.56 0.67 0.62 ]
   BlendTexture true
   Radius 292.45
   EllipticalOrbit {
      Period            18.2805
      SemiMajorAxis     7
      Eccentricity      0.0002
      Inclination       -8.0033
      MeanAnomaly       145.77
   }
   RotationPeriod 0.98
   Obliquity 1.62
   Albedo 0.54
}


[Reiko]

P.S. I tried both with and without the NormalizeMesh true line.

[cartrite]

The ssc file I did used a radius of 0.7 or 700 meters. When this is used, the asteroid is full size. When I used your radius of 292.45 km, the onscreen text confirmed that the size was 292.45 km but the model was tiny. "NormalizeMesh true" or "NormalizeMesh false" had no effect on size. But adding the line

Code: Select all

MeshScale 292.45

got the model to render the correct size. I'm not sure but this may be a bug. I'll start a thread in the bug forum.
cartrite

[Reiko]

The ssc file I did used a radius of 0.7 or 700 meters. When this is used, the asteroid is full size. When I used your radius of 292.45 km, the onscreen text confirmed that the size was 292.45 km but the model was tiny. "NormalizeMesh true" or "NormalizeMesh false" had no effect on size. But adding the line

Code: Select all

MeshScale 292.45

got the model to render the correct size. I'm not sure but this may be a bug. I'll start a thread in the bug forum.
cartrite

Thank you for looking into it. This happens with any cms model, not just that one and it doesn't seem to matter how many you have in a ssc file.

[Fenerit]

Cartrite, just FYI, no problems found until now with your new patch. I ask to me if my personal procedure of deleting al the time the shaders.log before to load any new patch is a ridiculus operation or can help in "free the road" for a stable response.

[CAP-Team]

I've been running Vista 64 bit for a couple of weeks now, I was wondering if it is possible to compile a 64-bit executable for Celestia. And ofcourse would this differ in performance?
My system has 4 GB of RAM.

[cartrite]

I've been running Vista 64 bit for a couple of weeks now, I was wondering if it is possible to compile a 64-bit executable for Celestia. And ofcourse would this differ in performance?
My system has 4 GB of RAM.

I also have a 64bit processor but I don't use the Windows64 bit OS. I'm still using Windows32 bit XP on my computer. I do have Linux x86_64 Open Suse 10.2 also installed on my computer (dual boot) and compile Celestia regularly. But I suspect it is still compiling 32 bit code. Almost all the packages installed on my computer's Linux system has 64 bit and 32 bit versions.

I don't think Celestia's code is 64 bit so you'll probably still be generating 32 bit code with Vista64. Still, I don't think I have the necessary knowledge to answer your question accurately.

My system performance is about the same on both OS systems and they use the same hardware. I think the Linux system is a little faster though but this may be because it is using a 64 bit driver for the graphics card.

cartrite