Celestia Forums

Are the SVN versions no longer available to the end users?I used to test this versions,however in Subversion,there is no file with the .exe extension.

--- edit ---

I think I will have to wait...only cartrite was implementing the changes in for example,rendering of the stars and he lost everything.The latest build from another compilers was 5023,that have NOTHING new.

Daniel, I still build Celestia regularly. There have been few changes since 5023, that's why I haven't updated it. I will upload the most recent tomorrow (but there have been changes at my lab, so the new version will be public only on Tuesday). It will have the fading orbits. I don't know about the new star rendering, I'll check.

Celestia SVN for Windows: here

Guillermo

Hello,

Have a look in the first post of the development forum
viewtopic.php?f=4&t=13449

In the download section of the first post you have
celestia built 5043 for version 1.6.1
celestia built 5044 for version 1.7

abramson wrote:I don't know about the new star rendering, I'll check.

Done. With Chris' patch, I built also with the new star rendering. It's also there:

Celestia SVN for Windows: http://fisica.cab.cnea.gov.ar/estadistica/abramson/celestia/svnexec/

(Note that the direct links to cabfst28.cnea.gov.ar are no longer accessible from the outside world. Only this new address works, and its updated at midnight.)

Guillermo

jogad wrote:Hello,

Have a look in the first post of the development forum
viewtopic.php?f=4&t=13449

In the download section of the first post you have
celestia built 5043 for version 1.6.1
celestia built 5044 for version 1.7

I noticed the 1.7.0 shows as 1.6.0 when loading up.

Reiko wrote:I noticed the 1.7.0 shows as 1.6.0 when loading up.

My builds show up as 1.6.0, but that's because I build from the trunk. Just a technicism. The new code is there, just not the tag. I presume Jogad's is the same case.

Guillermo

--- edit ---

Reiko wrote:I noticed the 1.7.0 shows as 1.6.0 when loading up.

Yup. Chris normally does not change the version number until he starts making "release candidate" versions available for download. It can be confusing at times.

The stars didn?t change at all.I can?t see any change.The rendering is the same as usual.What?s wrong?I only unzip the exe file.

danielj wrote:The stars didn?t change at all.I can?t see any change.The rendering is the same as usual.What?s wrong?I only unzip the exe file.

You need to change the star style to 'point stars' !!
Also OGL2.0 render path, of course.

Fridger

danielj wrote:The stars didn?t change at all.I can?t see any change.The rendering is the same as usual.What?s wrong?I only unzip the exe file.

You must choose Star style "points". Fuzzy points and scaled siks are unchanged. (Watch out, if you downloaded from mine builds, they will not be available until midnight (03:00 UT). I build two versions of the last SVN5063: fading orbits, and fading orbits + new stars.)

Cheers,

t00fri wrote:

danielj wrote:The stars didn?t change at all.I can?t see any change.The rendering is the same as usual.What?s wrong?I only unzip the exe file.

You need to change the star style to 'point stars' !!
Also OGL2.0 render path, of course.

Fridger

I did this and no change.

Reiko wrote:

t00fri wrote:

danielj wrote:The stars didn?t change at all.I can?t see any change.The rendering is the same as usual.What?s wrong?I only unzip the exe file.

You need to change the star style to 'point stars' !!
Also OGL2.0 render path, of course.

Fridger

I did this and no change.

I have the latest patch of Chris implemented here and for sure it works great. However, Chris sent me TWO patches! Perhaps the first one never really worked, yet has been (accidentally) implemented in the present public build?? Anyway I am convinced Guillermo tried things out before making his build public.

Here is how the alpha CRUx region should look like with limiting magnitude set equal to 12


Fridger

Reiko wrote:I did this and no change.

t00fri wrote:Here is how the alpha CRUx region should look like with limited magnitude set equal to 12

I just tested the Crux to be sure. Here are the snapshots, at 12 mag limit, no automag, using the 2M stars database, star rendering "points".

Reiko: the star rendering is VERY different, as you can see. Fridger: The result is not EXACTLY the same! (Very realistic, though.) Gacrux is more orange in yours, more redder in mine. And some differences in the dynamic range (Or you used automag and it's a crop of a wider view; if I put automag 12 at this angle bright stars com very over-extra-exposed). The render.cpp I have is dated 28 september, and render.h 5 august. I didn't tweak any parameters in the renderer.

Guillermo,

the picture I displayed above was copied from Chris's post. The essential feature of the new stars is that there is a big size/brightness enhancement of bright "foreground" stars compared to the huge number of dim "background" stars. The amount indeed depends on the limiting magnitude setting and whether automag =on | off.

Fridger

t00fri wrote:the picture I displayed above was copied from Chris's post. The essential feature of the new stars is that there is a big size/brightness enhancement of bright "foreground" stars compared to the huge number of dim "background" stars. The amount indeed depends on the limiting magnitude setting and whether automag =on | off.

Ah, OK, I hadn't realized it was from his post. Indeed. I tried to recreate it.

The rendering is much more realistic than the previous ones, as anybody that has looked at the stars with their naked eyes or wide field scopes should realize. Use it!

abramson wrote:

Reiko wrote:I did this and no change.

t00fri wrote:Here is how the alpha CRUx region should look like with limited magnitude set equal to 12

I just tested the Crux to be sure. Here are the snapshots, at 12 mag limit, no automag, using the 2M stars database, star rendering "points".

Reiko: the star rendering is VERY different, as you can see. Fridger: The result is not EXACTLY the same! (Very realistic, though.) Gacrux is more orange in yours, more redder in mine. And some differences in the dynamic range (Or you used automag and it's a crop of a wider view; if I put automag 12 at this angle bright stars com very over-extra-exposed). The render.cpp I have is dated 28 september, and render.h 5 august. I didn't tweak any parameters in the renderer.

crux-newstar-points.jpg

crux-oldstar-points.jpg

What I get is the 2nd pic. I never get anything like the first one and I do have it openGL 2.0.

Reiko wrote:What I get is the 2nd pic. I never get anything like the first one and I do have it openGL 2.0.

Reiko, have you used the celestia.exe conained in this zipfile? If not, try it. Drop the celestia.exe of the package into your celestia folder (you can rename
it or the original, so that you do not lose your original celestia). Then run the new one.