Orbiter 2024 Launch readiness

jarmonik

Well-known member
Orbiter Contributor
Addon Developer
Beta Tester
Joined
Mar 28, 2008
Messages
2,666
Reaction score
795
Points
128
Lot of time has passed since previous Orbiter release so it might be good idea to push forward with a new release Orbiter 2024. I haven't had a word with Martin yet, but he said long ago to prefer a release sooner rather than later. So, let's try to get the launch somewhere in late March this year. There is a lot of work waiting for this to happen so that we can start doing API cleanup and get new features going forward.

What is the current status of launch readiness ?

Documentation ? There is no documentation in a pre-build binary package in a GitHub. Is anyone else able to build the documentation than Martin ?
Do we have crashes or failures that do not occur in Orbiter 2016 ?
What other preparations would be needed for a launch ? How does the GitHub work exactly, would there be a download page for an official release that could have links to texture packages needed to run the Orbiter ?
 

GLS

Well-known member
Orbiter Contributor
Addon Developer
Joined
Mar 22, 2008
Messages
5,917
Reaction score
2,921
Points
188
Website
github.com
I haven't had a word with Martin yet
What other preparations would be needed for a launch ?
would there be a download page for an official release that could have links to texture packages
Maybe running it by @martins (is the account gone?) first would probably make some sense, as perhaps the existing pages could be updated, and thus we would not need github for the official releases.
If we have to go with github, the release page allows for text to be added, so any links can go there. Here is an example:
1708438686830.png
(hopefully with a better description than what I have...)


What is the current status of launch readiness ?

Documentation ?
This should probably be the main target now, both updating it as well as building it.


So, let's try to get the launch somewhere in late March this year.
So, the legal days for a release in March are: 5, 12, 19, 26. 😅
 

Urwumpe

Not funny anymore
Addon Developer
Donator
Joined
Feb 6, 2008
Messages
37,617
Reaction score
2,337
Points
203
Location
Wolfsburg
Preferred Pronouns
Sire
How about testing? Can we maybe start with a short beta phase to test compatibility with at least the 2016 add-ons? Also, did all scenarios and example spacecraft get tested yet?

I'd suggest making a special "Test and Release" project in github, then anybody interested can see how much work to release is needed and if they want to contribute, help in testing.

I can try looking at the documentation, as far as I understand it, its only a minor configuration issue that has to be found. And since its minor, its hard to find.

You can make a download page, at least in the wiki. Also you can release precompiled packages in the special releases panel.

What else?
 

n72.75

Move slow and try not to break too much.
Orbiter Contributor
Addon Developer
Tutorial Publisher
Donator
Joined
Mar 21, 2008
Messages
2,696
Reaction score
1,353
Points
128
Location
Saco, ME
Website
mwhume.space
Preferred Pronouns
he/him
@CaptainSwag101 and @Max-Q have both done some extensive testing building NASSP against OO and flying missions. Their input would be good to get.

I have one tiny little feature that I would like to add....after that I promise I'll stop creating pull-requests, and start testing.
 

Max-Q

99 40
Addon Developer
Joined
Jul 5, 2021
Messages
765
Reaction score
1,183
Points
108
Location
Cislunar Space
Website
www.orbiter-forum.com
@CaptainSwag101 and @Max-Q have both done some extensive testing building NASSP against OO and flying missions. Their input would be good to get.

From what I've seen, NASSP is stable and working well in OpenOrbiter/hopefully soon Orbiter 2024.
Also, I'd like to get @Mr. Residuals input on this, he's done quite a bit of testing and full missions with NASSP in OpenOrbiter.

I might add though, I have tested almost exclusively in x64. I don't see how x86 would be any less stable than x64, given that both NASSP and Orbiter have always been x86 projects and both were ported to x64 only recently.

I assume we're only talking about an x86 release at this time, and not x64?

As far as addons, I can test my stuff against OO x86. If they don't work, a simple recompile should fix it, given the bit-ness is still the same.
 

jarmonik

Well-known member
Orbiter Contributor
Addon Developer
Beta Tester
Joined
Mar 28, 2008
Messages
2,666
Reaction score
795
Points
128
I assume we're only talking about an x86 release at this time, and not x64?

Yes, this is x86 only release. Many planetary ephemeris are in a broken state and wont work on x64.
 

misha.physics

Well-known member
Joined
Dec 22, 2021
Messages
399
Reaction score
515
Points
108
Location
Lviv
Preferred Pronouns
he/him
I have some thoughts regarding planet textures. Maybe we have to include the default planet textures (from the Orbiter 2016 official zip-archive) into the Orbiter 2024 release in order not to require Orbiter 2016 installation and write a path for the "PlanetTexDir" option in "Orbiter_NG.cfg". Just so that Orbiter users could install/unzip the Orbiter 2024 release and use it "out of the box".

Also, maybe Minor bodies could be include into the 2024 release. It's only ~261 Mb.

And also, there're some additional modifications for the highy-detailed textures for Mercury, Mars and Titan. Here and here. But maybe we can't include these modifications into the 2024 release, since they may be incompatible with the default ones. I think these aspects must be studied in order to make the 2024 installation as easy as possible, namely optimize this all for the 2024 version.
 

WingC3

Donator
Donator
Joined
Feb 10, 2022
Messages
106
Reaction score
35
Points
28
Location
UK
I have some thoughts regarding planet textures. Maybe we have to include the default planet textures (from the Orbiter 2016 official zip-archive) into the Orbiter 2024 release in order not to require Orbiter 2016 installation and write a path for the "PlanetTexDir" option in "Orbiter_NG.cfg". Just so that Orbiter users could install/unzip the Orbiter 2024 release and use it "out of the box".

Also, maybe Minor bodies could be include into the 2024 release. It's only ~261 Mb.

And also, there're some additional modifications for the highy-detailed textures for Mercury, Mars and Titan. Here and here. But maybe we can't include these modifications into the 2024 release, since they may be incompatible with the default ones. I think these aspects must be studied in order to make the 2024 installation as easy as possible, namely optimize this all for the 2024 version.
I support that sentiment.
I think it's important to keep in mind that most of the people who we might seek to inspire towards a career in aerospace or aviation through exposure to Orbiter are very unlikely to simultaneously be software engineers, pilots, and physicists. They're also likely to be younger.

Very few people in the younger generation are familiar with the level of file manipulation that Orbiter often requires. They've grown up installing apps. That's unfortunate, but it's a fact. This should not preclude them from experiencing the wonder of flight and space. So, the lower the barrier to entry, the better.

I don't think Orbiter should ever compromise its focus on realism and fidelity within the simulation itself. If that makes it difficult to master, so be it. Space is hard.
KSP, Re-Entry, and others can serve as easier on-ramps to the complexity of Orbiter if someone needs that.

However, it does sometimes feel like we lose sight of the forest through all the trees: It shouldn't be difficult to install and run.
There's an incredibly talented team of people donating their valuable time to work on APIs for cargo, EVAs, new vehicles, new systems, better performance, etc.
These are all incredibly important things, but I think it may be worth devoting a little of that time to the less glamorous task of bringing the installation and management of Orbiter's files and configurations within a reasonable distance of what one would expect from any modern application. Perhaps just a post-install tool like CKAN to automate the discovery, installation, and uninstallation of mods. But ideally an executable installer with selection options for texture packs to install automatically as well.

Right now, I think we can all agree that the installation process is nowhere close to any other modern simulator; even ones that support extensive modding. If we want this community to sustain and grow, I think that should be high on the list of priorities for an update.
 

Urwumpe

Not funny anymore
Addon Developer
Donator
Joined
Feb 6, 2008
Messages
37,617
Reaction score
2,337
Points
203
Location
Wolfsburg
Preferred Pronouns
Sire
I would really like to have a tool like rpm for maintaining orbiter installations. But it can and should be a separate project, especially since it isn't that easy without some server support AND support by the add-on developers.

AND because of the skills that we can expect, maybe it would be smarter to run as UI application, where you can just click and select the packages that you want to install or deinstall or update.
 

Gondos

Well-known member
Joined
Apr 18, 2022
Messages
231
Reaction score
268
Points
78
Location
On my chair
I would really like to have a tool like rpm for maintaining orbiter installations. But it can and should be a separate project, especially since it isn't that easy without some server support AND support by the add-on developers.

AND because of the skills that we can expect, maybe it would be smarter to run as UI application, where you can just click and select the packages that you want to install or deinstall or update.
We already had one orbiter-mods disaster, not sure we need another one...
 

misha.physics

Well-known member
Joined
Dec 22, 2021
Messages
399
Reaction score
515
Points
108
Location
Lviv
Preferred Pronouns
he/him
Personally for me an installing of addons just unzipping an arhive into Orbiter directory is pretty convenient since we see what files the addon includes. Such zip-mode of addons is better for me, than just an executable file where we can't see what files we install. Even if there will be some *.exe installers I think the *.zip addons must be preserved as an alternative for developers and everyone who prefers this way.
 

WingC3

Donator
Donator
Joined
Feb 10, 2022
Messages
106
Reaction score
35
Points
28
Location
UK
Personally for me an installing of addons just unzipping an arhive into Orbiter directory is pretty convenient since we see what files the addon includes. Such zip-mode of addons is better for me, than just an executable file where we can't see what files we install. Even if there will be some *.exe installers I think the *.zip addons must be preserved as an alternative for developers and everyone who prefers this way.
As the above commenter mentioned, I think a package manager is preferable to an opaque .exe installer for addons. It's good to know what is going where. Totally agree.
The exe installer would be for the main simulator. You can see the files it installs after.

I am completely aware that this is not a simple task, and I didn't mean to imply that it was. I only meant to say that I think it's an important task. Nor do I think it is a prerequisite for an Orbiter 2024 initial release. I would suggest that it should be prioritised for work post release.
 

misha.physics

Well-known member
Joined
Dec 22, 2021
Messages
399
Reaction score
515
Points
108
Location
Lviv
Preferred Pronouns
he/him
I'd like to share my some other thoughts as well.

Maybe it's not necessary to hurry setting some "deadline" for the Orbiter 2024 release, because haste and time limits don't always benefit projects. I think some additional testing, optimization, tuning, and issue fixing could be done to get a good stable 2024 release.

I consider the last Orbiter build on GitHub (v240201 x86) pretty nice to become the base for the 2024 release. Personally for me, it has some key advantages compared to Orbiter 2016.

Maybe it makes sense to collect noticed issues in one place, structure them by type, update their current status, and set the priority (make something for this or the next releases). Just to know what issues (and requests) we have. So, the corresponding thread could be created (maybe something like the addon status thread Addon compatibility and development status by @N_Molson).

Personally, I have some such notes and ideas, too. That's not necessary for the development, but just to familiarize the community with it, hear their opinion and their own proposals. Just not to lose/forget all this.


Also, I think a zip-installer for the next even base Orbiter versions must be preserved (along with a possible *.exe installer), since it may be useful to extract some files from the zip-archive of the base Orbiter distribution without installing all Orbiter files.

I'd also like to take this opportunity to thank all the developers and users for their contributions to improving the Orbiter :)
 

Buck Rogers

Major Spacecadet
Joined
Feb 26, 2013
Messages
359
Reaction score
279
Points
63
Just my 2 cents
Keep the download structure as is, basic core package that one can setup and build on, so keeping some of the key features of Orbiter; accessability and flexability.
Additionally add pre-setup packages for eg. med-high end computer, a 1 or 2 step download, check your gf card and away, minimum setup.
If there's absolutely no need for an installer stay with zip- one of the positive things about orbiter adding to accessability.
IMO the deltaglider IS the perfect example for orb. but maybe something even simpler for a first demo .scn like a PB with a really nice VC so in a few minutes after installation one can hop around cape can. and the moon.
When I first installed Orb2016 I was like ok looks nice, but a lot of work to do, and I'm still at it flattening, improving etc. and still haven't done cape can. The landing pads are wonky, this is an unneccesary bad impression for the first place you'll see. I'm sure others have tinkered too, if we consolidate these efforts it would not look so half done.

On the question of add-ons. I've been playing around with Orb for over ten years and still don't know half of what has been produced, and a wealth of productivity seems to be lost to obscurity. I think you could get enough modders to agree to put some packages together, like Moon Bases (was working on) etc.. But I think it's maybe more a question of organisation.

I was thinking if we Showcase addons (addon bundles): an organized download page, approved and tested collections. Even AMSO for ex., probably the next thing your going to download, complete package, easy, accessable, effective, but till you got all landing sites fixed on your setup can take a while. Have all these things together on one page.
I think if we iron out these frustrations Orb would be even more likable.

P.S. JSGME is an easy mod management tool- especially if it's pre-setup
 

jarmonik

Well-known member
Orbiter Contributor
Addon Developer
Beta Tester
Joined
Mar 28, 2008
Messages
2,666
Reaction score
795
Points
128
Maybe it's not necessary to hurry setting some "deadline" for the Orbiter 2024 release, because haste and time limits don't always benefit projects. I think some additional testing, optimization, tuning, and issue fixing could be done to get a good stable 2024 release.
I believe the original goal was to release Orbiter a year ago. (Sooner rather than later). There's been testing, there's been bug fixing and tuning over a year. It's a process that never ends, so, we have to draw a line somewhere. By my knowledge Orbiter is now in a pretty good stable condition and we still have a month to carry out tests. If there are stability related problems then let me know.
 
Top