New Celestia-1.4.0pre-FT1 (galaxies) for Testing

[Toti]

For your further information, I did experience some problems after running
the test version, but nothing that couldn't probably be repaired with a
better pc.

If I have the Galaxy rendering enabled, my Firefox browser simply would
NOT connect to the web, even if I shut Celestia down and ended the
program completely. If I disable galaxy rendering, then I can still click on
an "Info" button from within the test version, and be taken to the site of
interest. With galaxy rendering enabled, however, the browser simply
stops responding. I even ended the test version while Galaxy rendering
was enabled, and could still not get my browser to execute correctly
- even though the test program had ended. I had to re-boot my pc in
order to get my browser working again.

I see no problem under KDE. Any other Windows user experiencing this?

What happened to the view splitting keys?

Indeed they don't work: The menu entry in KDE also fails. Will have to see what's going on here...

[t00fri]

OK, in view of this confusion, it seems very important that we have a latest CVS compilation for Windows available that may be used to cross-check these various claims of

--vanishing render paths
--crashes when focus is on "round objects" like earth, mars etc.

It makes little sense that I try, since I cannot reproduce either of the two problems on my XP desktop machine!

@Selden: Do you have such a most recent Windows CVS compile?? You stated above already that you also get the CTRL-V crash with that version which is unrelated to our new code!

Bye Fridger

[danielj]

When I toggle Ctlr-V,Celestia crashes,too.
I don??t understand what is a "latest compilation for Windows".Where can I get this?I have the original 04/11/2004 Celestia 1.4.0 pre 6

[quote="t00fri"]OK, in view of this confusion, it seems very important that we have a latest CVS compilation for Windows available that may be used to cross-check these various claims of

[selden]

Here are wide angle (73 degrees) views of the Milky Way "problem."


Milky Way is drawn by pre6, M31 is hidden behind labels.


Milky Way is not drawn by FT1, although M 31 is.

[t00fri]

When I toggle Ctlr-V,Celestia crashes,too.
I don??t understand what is a "latest compilation for Windows".Where can I get this?I have the original 04/11/2004 Celestia 1.4.0 pre 6

OK, in view of this confusion, it seems very important that we have a latest CVS compilation for Windows available that may be used to cross-check these various claims of

Wuff...Daniel: CVS is the place where Celestia is developed.
Most people know this meanwhile.

All the latest submitted code changes are there. The latest code changes in CVS were made by Chris in July. (except for various textfile changes by Grant quite recently) .
Some of us compile this latest CVS regularly and in fact our FT1 version is based on it and NOT on Chris' latest published 1.4.0pre6 version.

Bye Fridger

[Toti]

Regarding the rendering, I've a strange result: while in the previous Celestia 1.4.0pre06 I had all the four rendering possibilities, but after installing the new celestia.exe I have only two, Basic and Multitexture.
Anyother with the same problem?

I have exactly the same problem with a NVidia GeForce MX 440 SE (with AGP 8X) and the lastest drivers installed (78.01 from the 08/02/2005).

But over all, thanks for this release !

Could you please check if the same problem happens with latest CVS? Looks like Chris changed a few things in the renderer and openGL capability querier.

[selden]

Fridger,

I have a recent complation at home, but not here, so I can't upload it anywhere for another 2 or 3 hours.

Don't forget that the CVS code does include an eariler version of the new Galaxy code. Those changes were made 4 weeks ago.

Whether the render path crashes are directly related to the galaxy code, though, I don't know.

There were changes to render.h and render.cpp 7 weeks ago for the new orbit code. The crashes might be related to that, since the Earth has an orbit associated with it and M 31 does not. The crash happens whether or not the orbit is being drawn, though.

[t00fri]

Fridger,

I have a recent complation at home, but not here, so I can't upload it anywhere for another 2 or 3 hours.

Don't forget that the CVS code does include an eariler version of the new Galaxy code. Those changes were made 4 weeks ago.

Whether the render path crashes are directly related to the galaxy code, though, I don't know.

There were changes to render.h and render.cpp 7 weeks ago for the new orbit code. The crashes might be related to that, since the Earth has an orbit associated with it and M 31 does not. The crash happens whether or not the orbit is being drawn, though.

Does the crash happen with galaxy rendering enabled AND disabled?? Ie what happens after toggling the U key?
I suspect that this crash is something that was introduced in the context of the major (orbit) patches from DrJoe where Chris was "in a hurry" sort of...

If only I could provoke it in my machine

Once again: I start celestia, place the focus on earth or mars and then cycle through the paths with CTRL-V and at some point the display collapses? Right?


Bye Fridger

[t00fri]

Here are wide angle (73 degrees) views of the Milky Way "problem."


Milky Way is drawn by pre6, M31 is hidden behind labels.

Milky Way is not drawn by FT1, although M 31 is.

Selden,

I can reproduce your problem even much easier just from the start-up configuration. BUT only under Windows and this is not understandable to us! I explicitly checked that the deepsky.dsc file used is identical for XP and KDE. You see this is why I could not understand what you were writing further up: the milky way works beautifully under KDE-Linux with FT1!

Bye Fridger

[Toti]

Selden,

A few questions:

1) Can you select the Milky Way at all? I mean typing [Enter]Milky Way[Enter]? Is the name Milky Way shown in the upper left corner as usual?
2) What happens if -with maximum light gain- you place the camera very far from Earth whilst pointing at it? Say 90000 Kpc? Can you see the Milky Way shape?

[t00fri]

...
For your further information, I did experience some problems after running
the test version, but nothing that couldn't probably be repaired with a
better pc.

If I have the Galaxy rendering enabled, my Firefox browser simply would
NOT connect to the web, even if I shut Celestia down and ended the
program completely. If I disable galaxy rendering, then I can still click on
an "Info" button from within the test version, and be taken to the site of
interest. With galaxy rendering enabled, however, the browser simply
stops responding. I even ended the test version while Galaxy rendering
was enabled, and could still not get my browser to execute correctly
- even though the test program had ended. I had to re-boot my pc in
order to get my browser working again.

All in all though, the spectacular galactic views are worth any quibbling
doubts about my brower's performance. They are simply awesome.

Again, this is probably because I'm still using my old pc - A Gateway
Intel Celeron 1400 MHz unit with only 256 MB of RAM, and NO special
graphics adapter beyond the Intel 82815 graphics chipset inherent in
the computer.

Just thought I'd let you know what it did on my PC.

Great stuff, and THANKS VERY MUCH FOR tHE EFFORT HERE...

Bob,

I now checked the situation with my Windows XP notebook and independently with my XP desktop. No doubt, when --after selection of a galaxy-- I push mouse-R and activate "Info", my Firefox browser comes up with the respective information page from Strasbourg! I suspect as I wrote earlier, that your problems arose, since you Firefox has not been set as the default browser. I wrote also how to check and correct for that etc.

Cheers,
Fridger

[buggs_moran]

First, WOW. Fridger and Toti, they're beautiful.

Second. I get crashes as well when close to Earth and ctrl-V cycling, but nowhere else when cycling. I also do have the missing galaxy syndrome in all situations once I enter the galaxy (i.e. I can't see the disk of the Milky Way from inside the galaxy.) All extragalactic objects show up fine while inside the galaxy though. The LMC and SMC are particularly gorgeous with a full starfield.

As I am not even remotely a programmer I have little idea on what might be amiss. However, I would be willing to test it on my windows machines with a newer CVS build. I tried to build it once a while ago to little avail. The instructions in the forum are a little outdated and you have to piece things together on your own, many, many errors... Anyway contact me if you need assistance in testing, I have two testers at home and 3 at work with vastly different configurations.

[selden]

Celestia crashes immediately when I type a Ctrl-V while Earth is in view.
This happens although I disabled Galaxy rendering by typing a u.

My sequence of commands was
double-click on celestia.exe
wait for Earth to come into view
[return]m 31[return]
c
u
3
c
Ctrl-V
"Celestia has encountered a problem and needs to close."

[danielj]

The same happened to me.
Only I don??t know how I will replace the regular Celestia pre version for a compiler.What does a compiler stand for?

Celestia crashes immediately when I type a Ctrl-V while Earth is in view.
This happens although I disabled Galaxy rendering by typing a u.

My sequence of commands was
double-click on celestia.exe
wait for Earth to come into view
[return]m 31[return]
c
u
3
c
Ctrl-V
"Celestia has encountered a problem and needs to close."

[buggs_moran]

I just posted a small (1.6 MB) clip of what happens as you approach Milky Way (the choppyness is caused by using the scroll wheel, I get 30fps all the time with the new renderings). Contrast on all objects is pumped way up for visibility. As you can see, M31 stays visible once Milky Way pops out of existence, it does not seem based on distance, just entry into the disk. Hope this helps...

http://www.clanmoran.net/temp/capture.avi

[t00fri]

Celestia crashes immediately when I type a Ctrl-V while Earth is in view.
This happens although I disabled Galaxy rendering by typing a u.

My sequence of commands was
double-click on celestia.exe
wait for Earth to come into view
[return]m 31[return]
c
u
3
c
Ctrl-V
"Celestia has encountered a problem and needs to close."

OK Selden,

I went again to my Windows XP desktop and the notebook and typed /exactly/ what you wrote. Everything fine in both cases. NO CRASH. But remember: I have NO Celestia registry entry...


The next test was this: I noted already that indeed in Windows (only) I do not see the MilkyWay arms from inside (distance to center < Radius), although the name is completed and selected on the command line. Next I selected SMC and went there (G). Then I typed again Milky Way and centered it. From SMS I could see the MilkyWay from a distance as a nice barred spiral as it should be. So for some mysterious reason rendering has been switched off if the observer is located /within a galaxy/, here the Milky Way.

So it must be something in winmain.cpp, since in the KDE GUI program everything works fine. Wuff...I only changed 1 line in winmain.cpp rather unrelated with this


Bye Fridger

[ANDREA]

If I type "Milky Way", I see it, very nice, but if then I type "Sun" and give a look around, I can see LMC ans SMC, M 31 and other near galaxies, but not the Milky Way.
It looks like this one be considered in a differnt way from all the other galaxies.
Obviously I have "galaxies" option ON.
Moreover I checked all the Solar System objects, all show well, except the rings shadow rendering.
BTW, my nVidia driver is the last 78.01
Bye

Andrea

[t00fri]

Hi,

I now have compiled in Windows XP (Vc++.net) the code from the latest CVS without any additions, for which Chris was responsible (integrating Dr. Joe's patches etc) .

The result is as I suspected. All galaxies are rendered EXCEPT the Milky Way from inside. If Selden can confirm this within a few hours then we can be sure that this is a problem introduced by Chris' recent code integration and NOT by our new galaxy code.

Bye Fridger

[danielj]

I??m no programmer and while in cvs,can??t find where can I get a Celestia compiler.Why didn??t you tell that this test version was exclusively to programmers and not to general users?Now I have to reinstall the program.

Fridger,

I have a recent complation at home, but not here, so I can't upload it anywhere for another 2 or 3 hours.

Don't forget that the CVS code does include an eariler version of the new Galaxy code. Those changes were made 4 weeks ago.

Whether the render path crashes are directly related to the galaxy code, though, I don't know.

There were changes to render.h and render.cpp 7 weeks ago for the new orbit code. The crashes might be related to that, since the Earth has an orbit associated with it and M 31 does not. The crash happens whether or not the orbit is being drawn, though.

[t00fri]

I??m no programmer and while in cvs,can??t find where can I get a Celestia compiler.Why don??t you tell that this test version was exclusively to programmers and not to general users.Now I have to reinstall the program.

Fridger,

I have a recent complation at home, but not here, so I can't upload it anywhere for another 2 or 3 hours.

Don't forget that the CVS code does include an eariler version of the new Galaxy code. Those changes were made 4 weeks ago.

Whether the render path crashes are directly related to the galaxy code, though, I don't know.

There were changes to render.h and render.cpp 7 weeks ago for the new orbit code. The crashes might be related to that, since the Earth has an orbit associated with it and M 31 does not. The crash happens whether or not the orbit is being drawn, though.

This release is certainly for users to test.

What are you talking about? It's not our problem if you do not know what a compiler is or CVS. You are part of this community since a long time and these notions have been discussed MANY times by MANY people.

Most people had little problems with it EXCEPT you.
It seems right now that some of the problems were not due to us but due to Chris' last integration of DrJoe's patches.

You are not complaining again, are you??