Post Bugs Here

[ElChristou]

I also used to get either a giant star glare like ElChristou's, or a square glare (with Chris' experimental patch), but the more recent CVS commit seems to have mostly fixed this. I say "mostly", because one serious problem still remains: the sun disappears between about 0.05 ly to 1.0 ly. Toggle off planet orbits, and the sun (the whole shebang, not just the glare) disappears even earlier, at around 10 au. This is with all render paths, even Basic (on my Mac).

The square has been fixed by Chris second patch.

[Cham]

I confirm Dirkpitt's issue : the whole star is disappearing between two limiting distances (I didn't checked if the limiting distances are actually constant or not).

At very long range, I can see the star (a dot ?). Closing by, the star disappear. At closer range, the star and its glare reappear on screen.

I think I noticed this effect before the color patch was applied, and it appeared at the same time the new size (in km) in the upper-left corner was added.

[chris]

I confirm Dirkpitt's issue : the whole star is disappearing between two limiting distances (I didn't checked if the limiting distances are actually constant or not).

At very long range, I can see the star (a dot ?). Closing by, the star disappear. At closer range, the star and its glare reappear on screen.

I think I noticed this effect before the color patch was applied, and it appeared at the same time the new size (in km) in the upper-left corner was added.

Here's a fix to test:

http://www.celestiaproject.net/~claurel/celest ... render.cpp

The technical explanation: Celestia was setting up a perspective projection with a near plane distance of zero when rendering objects as points. That worked fine on my new NVIDIA cards, but not on the ATI Radeon 9700 in my Powerbook. It's just a small change to avoid using such questionable projections.

--Chris

[ElChristou]

Here's a fix to test:

http://www.celestiaproject.net/~claurel/celest ... render.cpp

The technical explanation: Celestia was setting up a perspective projection with a near plane distance of zero when rendering objects as points. That worked fine on my new NVIDIA cards, but not on the ATI Radeon 9700 in my Powerbook. It's just a small change to avoid using such questionable projections.

I get 3 fatal errors:

render.cpp:1425: error: no matching function for call to 'Renderer::renderCometTail(Body&, const Point3f&, double, std::vector<Renderer::LightSource, std::allocator<Renderer::LightSource> >&, const float&)'

render.cpp:6032: error: prototype for 'void Renderer::renderCometTail(const Body&, Point3f, double, const std::vector<Renderer::LightSource, std::allocator<Renderer::LightSource> >&, float)' does not match any in class 'Renderer'

render.h:421: error: candidate is: void Renderer::renderCometTail(const Body&, Point3f, float, float, double, Quatf, const std::vector<Renderer::LightSource, std::allocator<Renderer::LightSource> >&, float, float)

[dirkpitt]

Christophe, have you tried updating to latest CVS, and then applying Chris' render.cpp?

[ElChristou]

Ah, yes, must be that! Go again for a test! Tx.

[ElChristou]

Compil ok. The planets dots are fine now, the Sun itself is ok but still have a giant glare appearing at ~7.05 au...

[BobHegwood]

I can't seem to run Celestia without the little red triangle appearing on
EVERY object I select to view. If I type 3 for Earth and then either center
it or GOTO it, the little red triangle stays on the planet. This happens with
ANY object I select for viewing and it's aggravating.

At any rate, just FYI. See system in signature below.

Thanks, Bob

Edit: I have also removed ALL check marks from the location options and
I'm running Celestia as an XP-compatible program. Should I not?

Didn't see any other posts regarding this minor bug, so sorry if I'm
griping about a previously reported bug.

Thanks again.

[selden]

Bob,

Typing a Ctrl-K is supposed to toggle those pesky markers off and on. Does it do anything for you?

[BobHegwood]

Bob,

Typing a Ctrl-K is supposed to toggle those pesky markers off and on. Does it do anything for you?

Well I'll be damned... Selden? How looong have I been using this program?
In all honesty I did NOT know about CTRL-K. It worked just fine.

Sorry I'm such a boob, but my brain doesn't work like it used to. Not that
it EVER worked very well anyway.

Thank you sir... I'll shut up now.

[cartrite]

I just fixed a visibility issue with my model katrina. I tested the fix with cvs version 11/16/2007 and noticed the model is "blinking". With version cvs 10/25/2007 there is no "blinking". Links to a cel url on a html page along with the updated model is available on the home page of my site. Left side.
It is hard to show this with images but here is a shot from cvs 10/25/2007. EDIT: Both of the cvs versions above were using the kde gui.



And a shot from cvs version 11/16/2007.



This "blinking" is also seen in Celestia 1.5.0pre4.
cartrite

[cartrite]

Sorry, I tracked this "blinking" to the solarsys.ssc file and it has nothing to do with Celestia versions. In the solarsys.ssc copy I use regularly I uncomment the line from:

Code: Select all

   # Oblateness 0.0034

to:

Code: Select all

   Oblateness 0.0034

I'll update the instructions in the model to include this.
cartrite

[selden]

cartrite,

Hopefully you'll recommend using the Modify directive in a new SSC, not editing solarsys.ssc.

e.g.

Code: Select all

Modify "Earth" "Sol" { Oblateness 0.0034 }


[cartrite]

selden,
I'm not going to recommend anything to do with the solarsys.ssc file. The reason this was happening was I used a test version to update the model and it had an ssc file that included an Oblateness 0.0034 statement. I commented that statement out in the models ssc file so there should not be any problems with the standard solarsys.ssc file staying the way it is. I can't remember why I put that statement in the model ssc file to begin with. I was probably experimenting with something. Anyhow someone sent an email to me saying that the model was not visible. It caught me by surprise in a groggy state of mind and I responded hastily.
Sorry for any problems this may have caused. There was activity at my site after the first update so this model was probably downloaded several times. If anyone downloaded the model today before I wrote this I do recommend "looking" at the cloud.ssc file in the model package and comment out the line

Code: Select all

    Oblateness 0.0054

to look like this

Code: Select all

#   Oblateness 0.0054

Anyone who downloads the model after reading this the problem should have been corrected and no action is necessary.
cartrite

[selden]

cartrite,

OK, no sweat. I understand very well the confusions caused by version skew!

[cartrite]

I actually downloaded several copies of render.cpp and render.h from sourceforge and it wasn't till I recompiled with the 10/25 version of the files and still saw the "blinking" that I even considered looking at the ssc files. SHEWWW!!!

I'll have to slow down and think things through before I act in the future.
I should have done this update a while back when I first noticed the problem. When I got the email, I had no memory that I even dealt with this before. Somehow it got placed on the back burner. The real problem was visibility.
I still need to update the model at the motherlode. SO PLEASE, Don't download the model from the motherlode and expect this visibility issue to be fixed. The visibility issue is fixed ONLY with the model at MY site. Before the fix, the model was invisible with Celestia 1.5.0prex. Because of probable code changes, and to keep confusion to a minimum, I'm going to wait until an official version of Celestia 1.5.0 is released before I attempt to update the model at the motherlode.
Referring to the last line in red, now all I need to do is remember this when the time comes.
cartrite

[Sui Ota]

Hello,
I was absent from this forum for several monthes, but I am suffering from this bugs with new iMac

Bugs with newest iMac, and newest CVS:

1.
Square star glares appear when I use OpenGL vertex program or OpenGL 2.0.
I don't encounter this behaviour with Celestia 1.4.1.
http://www.aqsp.net/square.png (1900x1200, 167kB)

2.
The Lua Edu Tools sometimes blink (continues 'enabled' and 'disabled' rapidly) when I am moving camera.

Using Intel iMac(mid 2007), Core 2 Extreme 2.8GHz, Radeon HD 2600 Pro, Mac OS X 10.5.1 with any updates for this newest iMac.
But this bug appears on newest iMac with 10.4.10.

[selden]

For item #1, which star style are you displaying? Type a Ctrl-S to step through them.

[Vincent]

Hello,
I was absent from this forum for several monthes, but I am suffering from this bugs with new iMac

Hello Sui, and welcome back !

2.
The Lua Edu Tools sometimes blink (continues 'enabled' and 'disabled' rapidly) when I am moving camera.

You may want to try the latest updated version of the Lua Tools:
http://vincent.gian.club.fr/celestia/Lua_Edu_Tools.zip

Then, if the bug still occures, please use the relevant discussion to post your bug report:
http://celestiaproject.net/forum/viewtopic.php?t=10352

That will avoid all confusion.

[Sui Ota]

For item #1, which star style are you displaying? Type a Ctrl-S to step through them.

It happenes with 'points'.
In addition, using 'fuzzy points' or 'scaled discs', any stars except near the camera don't shown, on OpenGL vertex program or OpenGL 2.0.