Space Shuttle - Development

[Thorsten]

Merry Christmas from the far North!

[GinGin]

Thanks , Merry Christmas Guys

Looks like a screen from FG

[Catalanoic]

Merry christmas from sunny Barcelona!!

[wlbragg]

"Merry Christmas Houston. Oh, and did I mention we have a problem!"

[legoboyvdlp]

Oh er. Quite a situation there.

I wonder could a human survive jumping out and floating down with an EVA suit?

I suppose they are doing mach 28, so... No

[GinGin]

Very nice Wayne
I alreday have some idea of scenarios

@Legoboyvdlp: As they are at the same speed, they could EVA and will orbit at Mach 28 around the Earth waiting for help
But EVA preparation takes time to lower the N2 concentration into the blood ( pure O2 in the EVA suits). So even if they rush the preparation ( risk of N2 bubbles in the blood like for diving), it would be quite long to jump out in case of disaster in the Shuttle I guess.

[amalahama]

I was actually wondering whether you might be able to go see it (and take pictures and such...)

Oh yesss!! New hi-res textures will improve hugely the visual appaerance.

@Thorsten, any news regarding the utilization of the 3D external mesh from SSU? I find it extremely good looking, and you might make use of payload's models as well, since they would be compatible.

Best regards, and merry Christmas!

[Thorsten]

@Thorsten, any news regarding the utilization of the 3D external mesh from SSU?

No - from my side this is more an idea than an actual plan - I'm not a 3d modeler myself, so I couldn't do anything with it if I got it in nice form tomorrow.

[Hooray]

Referring to (and since this seems to be a recurring question, even among long-term contributors):
https://sourceforge.net/p/flightgear/ma ... /36504422/

I've never tried firing up the shuttle even once. [...]
Does the shuttle model have any kind of demo mode or mostly hands off flight mode?

Would it be feasible to discuss if implementing some kind of simple mechanism (UI/CLI) to provide pre-canned flights using any combination of existing features or shuttle infrastructure/code (e.g. instant replay, flight recorder, pre-recorded flights, startup scenarios etc) ?

The point being that this would also help "non-astronauts" to test fgfs-spaceflight without having to spend weeks reading a manual, thinking specifically about troubleshooting, benchmarking/profiling ?

FWIW, for 2 decades now, flying the ufo at high ground speeds has been considered the recommended way to "stress-test" certain FG components like the tile manager (and as many of the contributors know, it's a versatile and handy tool) - and since we keep talking about the shuttle's fgfs performance, it would obviously be interesting for "non-pilots" to still be able to "run" a shuttle "flight" without having to do anything, but still exercise all sorts of code paths that may otherwise not receive much testing at all.

On the other hand, if we can work out a scheme using existing infrastructure, that might also translate into regression testing the shuttle itself (think having different pre-recorded flights and using them to see if the shuttle works properly, e.g. during release preparations).

Again, this is not a feature request, but just to encourage having the discussion whether this is feasible or not (long-term) - since you (Thorsten) mentioned originally that you were not interested in supporting/using the "flight recorder" feature, whereas wlbragg was considering to give it a try IIRC, which ended up overlapping with the development of "startup situations" and "overlay profiles", and you - rightly- pointed out that it would be difficult/impossible for the shuttles state vector to be mapped to the relatively simple "flight recorder" structure.

[Thorsten]

Would it be feasible to discuss if implementing some kind of simple mechanism (UI/CLI) to provide pre-canned flights using any combination of existing features or shuttle infrastructure/code (e.g. instant replay, flight recorder, pre-recorded flights, startup scenarios etc) ?

Would it be asked too much to simply start e.g. --SpaceShuttle-ISS even once and observe what happens before starting to request features?

Spoiler below:


















(It will put you into a stable orbit with all systems configured properly, and you can remain there without any action required probably till the fuel cells need purging - which should give you >48 hours of spaceflight (!) at least).

[wlbragg]

I also feel the need to respond and say this is already somewhat accomplished. To launch and achieve orbit you only need to throw approximately a dozen switches which are detailed in the Shuttle help and also the basic Shuttle document. Really any of the startup states (variants) will get you up and running for at least a while without having to do anything. There is even a "Free Flight" which gets you a one way glider (down). The "On Approach" gets you a condition that you can play over and over with the click of a "retry" button. So anyone that uses the excuse they don't have the time to learn how to fly the Shuttle simply isn't paying attention to the features that already exist.
For my part of the development process, I strictly rely on the canned features to load and verify any of my new work, without having to do anything at all.

[eatdirt]

and I've just been contacted by the company preparing the exhibit about how the FG Shuttle can be rigged to drive the displays and external visuals.

Ah ah, so nice, congrats, that's so cool!!!

And if they need test pilots to cra..., well stress, the stuff, you know what we do right?

[legoboyvdlp]

It would be relativitely simple to program a sub-autopilot to do everything automatically... relegating the role of the pilot to a mere observer!

Obviously not a realistic feature but it might actually be good to allow the simulator to do e.g. Pilot tasks and leaving the simulator-pilots role to the things done by the commander and so on?

[Thorsten]

It would be relativitely simple to program a sub-autopilot to do everything automatically...

Assuming there are no decisions to be made... But since it's trivial to do that if you can do Nasal, I'll leave it to anyone who needs it.

[Hooray]

Some people have previously toyed with scripted "copilots", and Stuart's checklist feature makes it possible to "execute" checklists.


PS: as expected, segfault