Celestia 1.7.0 Development Thread

The place to discuss creating, porting and modifying Celestia's source code.
Avatar
SevenSpheres
Moderator
Posts: 824
Joined: 08.10.2019
With us: 4 years 11 months

Post #761by SevenSpheres » 16.03.2020, 23:08

selden wrote:no GUI (corner) text is visible whether terse or verbose is selected.

This is because the latest builds require TTF fonts, but for some reason they have not been added to the repository.

Added after 2 minutes 25 seconds:
That said, I see another problem: the splash screen is solid black with no picture.
My Addons: viewtopic.php?f=23&t=19978 • Discord server admin
Celestia versions: 1.5.1, 1.6.1, 1.6.2, 1.7.0, and some unofficial versions like Celestia-ED

Avatar
Anthony_B_Russo10
Moderator
Posts: 672
Joined: 03.07.2018
Age: 21
With us: 6 years 2 months
Location: Tallahassee, Florida, US

Post #762by Anthony_B_Russo10 » 16.03.2020, 23:33

@onetwothree, I can not test the build you posted due non-OpenGL 2.0 capable graphics and Wine incompatibility. The only way I could test it is if it was compiled for MacOS X Snow Leopard without the need of dragging out one of my iMacs and going into firewire target disk mode.
Anthony B. Russo, I like Pluto. Mod of the Celestia subreddit: https://www.reddit.com/r/Celestiasoftware/
I have over 40 computers, trying to list them here would be a pain.
Responsible for the NEO catalog: https://celestiaproject.space/forum/viewtopic.php?f=23&t=22203
And mod of the Discord server.

Topic author
onetwothree
Site Admin
Posts: 705
Joined: 22.09.2018
With us: 5 years 11 months

Post #763by onetwothree » 17.03.2020, 09:34

selden wrote:1. no GUI (corner) text is visible whether terse or verbose is selected.

As already said, you need to update celestia.cfg and drop some TTF fonts to fonts/ subfolder.

selden wrote:2. There's a gap in the orbit next to the Earth.

Fading orbits are enabled by default.

selden wrote:3. Note that the OpenGL info window is improperly formatted, too.

I hate windows, it causes too many issues. It seems you are using win 10, is it?

Added after 1 minute 15 seconds:
SevenSpheres wrote:That said, I see another problem: the splash screen is solid black with no picture.

splash.png is beeing searched in another location - in the splash/ subfolder. This was changed to be consistent with other interfaces.

Added after 2 minutes 48 seconds:
Anthony_B_Russo10 wrote:The only way I could test it is if it was compiled for MacOS X Snow Leopard

The change was in win-only code. Windows OpenGL initialization is nightmare, now we effectively use 2 libraries - glew and own one, we want to switch to libepoxy which not only eliminates requirements in own wgl extension manager but also supports gl es (read as "ios & android").

Avatar
selden
Developer
Posts: 10190
Joined: 04.09.2002
With us: 22 years
Location: NY, USA

Post #764by selden » 17.03.2020, 11:27

1. I was expecting everything necessary to be included in the .7z archive.

2. Unfortunately, the orbit gap isn't a fade. It's an abrupt turning off, which I suspect is a bug. Also, please note that the Fading Orbits option is not checked in the View Options menu as can be seen in the screenshot that I provided.

3. Yes, I'm using Win10 now that Win7 is no longer supported. The upgrade is still free if you have a valid Win7 or Win8 license. While Win10 has its own problems, it has fixed many features which were broken or missing in Win7.
Selden

Avatar
LukeCEL
Posts: 373
Joined: 26.09.2017
With us: 6 years 11 months

Post #765by LukeCEL » 17.03.2020, 15:43

selden wrote:2. Unfortunately, the orbit gap isn't a fade. It's an abrupt turning off, which I suspect is a bug. Also, please note that the Fading Orbits option is not checked in the View Options menu as can be seen in the screenshot that I provided.

I can confirm that the same bug appears on MacOS. I have "Fading orbits" turned off.

orbitstop.png

Topic author
onetwothree
Site Admin
Posts: 705
Joined: 22.09.2018
With us: 5 years 11 months

Post #766by onetwothree » 17.03.2020, 15:53

It seems to me, that this is not a bug but made intentionally. With correct orbit theory like vsop87 you can't have closed orbits.

Avatar
LukeCEL
Posts: 373
Joined: 26.09.2017
With us: 6 years 11 months

Post #767by LukeCEL » 17.03.2020, 16:03

onetwothree wrote:It seems to me, that this is not a bug but made intentionally. With correct orbit theory like vsop87 you can't have closed orbits.

But in Celestia 1.6.1, Earth's orbits are closed.
161smooth.png

I noticed another interesting thing: in 1.6.1, the Moon's orbit line (green) has a strange bend in it. Am I the only one that sees this?

Avatar
SevenSpheres
Moderator
Posts: 824
Joined: 08.10.2019
With us: 4 years 11 months

Post #768by SevenSpheres » 17.03.2020, 17:16

onetwothree wrote:splash.png is beeing searched in another location - in the splash/ subfolder. This was changed to be consistent with other interfaces.

Then why isn't it there in the GitHub repository?

LukeCEL wrote:I noticed another interesting thing: in 1.6.1, the Moon's orbit line (green) has a strange bend in it. Am I the only one that sees this?

I've seen it too, that's one reason fading orbits were added.
My Addons: viewtopic.php?f=23&t=19978 • Discord server admin
Celestia versions: 1.5.1, 1.6.1, 1.6.2, 1.7.0, and some unofficial versions like Celestia-ED

Janus
Posts: 537
Joined: 13.08.2016
With us: 8 years 1 month

Post #769by Janus » 19.03.2020, 16:41

If anyone is interested here is commit 5772 in x86 & x64, txf font only.
I can do a ttf build if desired.
As usual for me, VS2015, static linked w/lua5.1

Sorry I have not been around, been busy rebuilding after some kinetically assisted hardware failures.
Is anyone besides me interested in a fork WITHOUT the greek letter stuff?
I do strictly ascii, and the greek lettering doesn't work for me.


Hope this build helps.

Avatar
Chuft-Captain
Posts: 1779
Joined: 18.12.2005
With us: 18 years 9 months

Post #770by Chuft-Captain » 20.03.2020, 04:12

LukeCel wrote:I noticed another interesting thing: in 1.6.1, the Moon's orbit line (green) has a strange bend in it. Am I the only one that sees this?

viewtopic.php?f=8&t=17343
"Is a planetary surface the right place for an expanding technological civilization?"
-- Gerard K. O'Neill (1969)

CATALOG SYNTAX HIGHLIGHTING TOOLS LAGRANGE POINTS

Avatar
LukeCEL
Posts: 373
Joined: 26.09.2017
With us: 6 years 11 months

Post #771by LukeCEL » 20.03.2020, 16:20

Chuft-Captain, thanks for the info.

Avatar
selden
Developer
Posts: 10190
Joined: 04.09.2002
With us: 22 years
Location: NY, USA

Post #772by selden » 20.03.2020, 16:48

A newly observed glitch (I don't use fullscreen mode often):

When running in full-screen mode at 1920x1080, my on-screen cursor is not being drawn. It appears when I move it up to the top of the screen, causing the menu bar to appear, but it goes away again as soon as I move it down so the menu bar stops being drawn. This was first experienced while using the default "crosshair" cursor. Switching to "inverting crosshair" or "arrow" makes no difference.

This problem is seen with 1.6.2 (all video modes, including OpenGL 2.0), and recent versions of 1.7, including Janus' variants dated 9-jan-2020 and celestia-7552-txf-x64 of 19-mar-2020

This problem is NOT seen with 1.6.1 (all video modes, including OpenGL 2.0) or when running 1.6.2 or 1.7 in windowed mode.

Added after 38 minutes 17 seconds:
Another 1.7 issue, which would never be encountered by developers:

The very first time v1.7 runs, when no previous "View Options" have been saved, the checkboxes "Body Stars" and "Body Planets" are not checked. As a result, the user sees a blank, black window.

I think they should be checked by default when Celestia v1.7 first runs and creates the initial on-disk copy of the "View Options".

Added after 29 minutes:
Some more info on the invisible cursor problem: the cursor is visible when Celestia is in fullscreen mode while any of Celestia's "control" windows are open. E.g. when the "View Options" window is open, the cursor is visible. As soon as the "View Options" window closes, and no others are open, the cursor disappears.
Selden

Janus
Posts: 537
Joined: 13.08.2016
With us: 8 years 1 month

Post #773by Janus » 20.03.2020, 18:12

@selden

I just checked and the cursor is visible in full screen on my system.
At the moment it is dual screen W7 64-bit using open shell & 7+ taskbar tweaker, with classic theme selected and 3d/shadows enabled.
Will only full screen on primary monitor, which is annoying, I wanted to test both.
Still getting the rest put together after some hardware issues.

Before anyone asks, which so many do for reasons I fail to comprehend, there is no chance I am allowing a W10 machine on my network.
It is insecure by design, the telemetry renders all other security measures irrelevant.


Janus.

Avatar
selden
Developer
Posts: 10190
Joined: 04.09.2002
With us: 22 years
Location: NY, USA

Post #774by selden » 20.03.2020, 18:51

Janus,

Thanks for investigating. It certainly has me puzzled. I had a problem like this under Win7 with Celestia 1.6.1, but it went away when I upgraded to Win10 with the same graphics drivers. It might be something funky in the Windows cursor configuration, but then I'd expect it to affect 1.6.1, too.

FWIW, I've been using O&O Shutup10 plus Comodo Internet Security Premium's firewall to try to minimize the outbound network traffic. Unfortunately, it isn't just Win10 itself which does a lot of telemetry. Many other products do, too, especially browsers. :( For example, if you turn on the security feature "protect against known bad web sites," they transmit all of the URLs you visit to a cloud-based scanner to check them. *sigh* Avast's "security" software was doing something similar until they got caught. See https://www.cnet.com/news/antivirus-firm-avast-is ... lling-users-web-browsing-data/
Selden

Topic author
onetwothree
Site Admin
Posts: 705
Joined: 22.09.2018
With us: 5 years 11 months

Post #775by onetwothree » 03.04.2020, 08:58

I've just merged self-shadowing branch into master, windows build are available here.

Known bugs and limitations:
* Only the 1st light source casts shadows.
* Thin objects have incorrect shadows, see ISS.

To enable them add the following option "ShadowMapSize 4096" to your celestia.cfg. You can use lower value but in that case visual artifacts will be more visible. I also suggest using anti-aliasing to decrease artifacts "AntialiasingSamples 1".

Despite 7spheres request this option will not be added to celestia.cfg even in commented form because I'm not sure if it present in future. We may want to implement another algorithm, or if we stay with this one we will move this option to the Preferences window.

Markerz
Developer
Posts: 274
Joined: 29.01.2009
Age: 29
With us: 15 years 7 months
Location: Suzhou, China

Post #776by Markerz » 03.04.2020, 11:33

macOS users can try the corresponding macOS build here, it is built with most current commit (with a few changes that landed in Android/iOS build) should have the same functionality as the one onetwothree provides

this is a native macOS application instead of the QT build in Celestia repo.

Avatar
LukeCEL
Posts: 373
Joined: 26.09.2017
With us: 6 years 11 months

Post #777by LukeCEL » 03.04.2020, 15:55

Hi Markerz, I just tried the latest build, but right after it loads all the files it crashes.

Also, when I click on it, it's reading files from where my 1.6.1 installation is (/Applications/Celestia.app/). I don't know if this is intended or not.

Here's the crash report, since it's too long for me to copy-paste:
Attachments
Celestia-2-crashreport.txt
(64.6 KiB) Downloaded 143 times

Markerz
Developer
Posts: 274
Joined: 29.01.2009
Age: 29
With us: 15 years 7 months
Location: Suzhou, China

Post #778by Markerz » 04.04.2020, 02:26

LukeCEL wrote:but right after it loads all the files it crashes.
Thanks a lot! i have identified the issue, should be able to provide another version fixing it in a few days!

LukeCEL wrote:Also, when I click on it, it's reading files from where my 1.6.1 installation is (/Applications/Celestia.app/). I don't know if this is intended or not.
What do you mean by reading files from your 1.6.1 installation? for the data it should be reading file from the content it was bundled with, unless you have specified a location to read from with 1.6.2

Avatar
LukeCEL
Posts: 373
Joined: 26.09.2017
With us: 6 years 11 months

Post #779by LukeCEL » 04.04.2020, 04:23

Markerz wrote:Thanks a lot! i have identified the issue, should be able to provide another version fixing it in a few days!

Nice!

Markerz wrote:for the data it should be reading file from the content it was bundled with, unless you have specified a location to read from with 1.6.2

That makes sense. I specified 1.6.2 to read from the same place as my 1.6.1 application.

Markerz
Developer
Posts: 274
Joined: 29.01.2009
Age: 29
With us: 15 years 7 months
Location: Suzhou, China

Post #780by Markerz » 07.04.2020, 02:31

LukeCEL wrote:That makes sense. I specified 1.6.2 to read from the same place as my 1.6.1 application.

can you try if this fixes the crash, the new download is at
App Center


Return to “Development”