Project G42-200 StarLiner

[Moach]

quickie not-completely-sober late Saturday night progress update

well - got the engines fixed up all the way to the OMS burners, then i flew 'er off in a smooth test flight that led to a nominal ISS dock-2 rendezvous... first since engines recode :thumbup:

so we're good again - all code restructuring is in place and accounted for - we have a flying ship once again :woohoo:


i took some time and corrected the missing fire-extiguisher holes (now you can rest assured that if a fire arises behind one of those panels, it can be put out :lol

anyways, the main thing i wanna bring to attention here, is that i've come down to a general idea on how to proceed into coding the many intricate, inter-depentent systems of the big '42

i've installed a "cue system" protocol - which means, whenever a switch is thrown (by click or keyboard), instead of applying changes right away, i can "cue" the respective action, then let the OMG-so-many operating states and conditions sort themselves out in a clear, concise way...

this avoids many, many bugs - and sets out a general design-pattern for the whole project from here on (see, that's why i don't go too much for the "canned" patterns - you always find you need something else that changes everything)

anyways, for all noticeable effects, this only means i've sorted out a degree of separation between pilot input and systems response... so from now on, if a switch is thrown and there are missing systems required to support it, things just might go down in ways other than expected :hmm:

(for instance - engines cannot start without fuel pressure, which is generated by the fuel pumps, which need AC feed, which is provided by APU, which needs its own fuel and so on...)


i haven't got that far yet... electicals are still only "implied"... but i'll get to simulating those eventually - well, at least now i know just how i'm gonna do it :yes:

 

[Richy]

Sounds very promising. I'm looking forward to this all-systems-fully-implemented version! Someone has to write a huge check list for that. ^^

 

[jangofett287]

(for instance - engines cannot start without fuel pressure, which is generated by the fuel pumps, which need AC feed, which is provided by APU, which needs its own fuel and so on...)

Why do I get the feeling this is going to need batteries or a ground AC feed to get it started? And Why do I get the feeling the checklists are going to be almost as long as the space shuttle's?

 

[fhf2548]

the wip-3 have android 2.2 with market in this plane

 

[Moach]

well, then - now we have parking brakes :thumbup:

they are quite necessary, since the main engines are so large they will push the craft forwards even at zero thottle :shifty:

anyways, now you can just hit ctrl+"." (FS-pilots know why i chose this) and the brakes remain set until you press it again...


then i corrected some sound issues i was having that made the main engine shutdown sound like rocket startup :blink:
yeah, those things happen when you go ctrl+c,ctrl+v happy and make two #defines the same number... stupid mistake, took hours to find :facepalm:


i'm still not 100% happy with my sound effects... they sound good on my home rig and 4.1 surround, but then i test them in my work pc with headphones and a lot loop-hickups become most apparent... (quite odd, really)

so i have yet to fix those...


and we still need to fix the scenario read/write for the new engine features (old code is plain useless now) so one can actually save the current scenario without that meaning a certain engine failure :uhh:

then, the cockpit still needs some finer polishing (new textures look better and make old crude stuff look bad) and those RCS doors need modelling on the ext mesh....

and if i can bring myself to stop forgetting it all the time, there's a tail-strike wheel that's still not where it was suposed to be :hmm:


that, cargo, and i'll call it WIP-3

 

[RaR76]

Sounds great!

Sounds like you designed some type of "state machine" for the switches/engines, if I understand correctly.

 

[Tommy]

i'm still not 100% happy with my sound effects... they sound good on my home rig and 4.1 surround, but then i test them in my work pc with headphones and a lot loop-hickups become most apparent... (quite odd, really)

Not odd at all, actually. The nicer speakers have much heavier cones, and that increased mass tends to cancel out some transients. A nicer soundcard will do the same.

Tom Waits (musician) "tests" his mixes by recording them on a cassette, and trialing them on his el-cheapo car stereo. If they sound good there, they'll sound great on a better stereo.

 

[Moach]

Sounds great!

Sounds like you designed some type of "state machine" for the switches/engines, if I understand correctly.

exactly! :thumbup:

and not one - many! - each switch is a little state machine of its own, neatly wrapped into a class that allows me to define them in a list-like format (one entry = one switch)

then each of those fire up a message when clicked (that's an orbiter thing, really) - it's just a numeric value - but from that i can figure out what switch was clicked (i had their switch-ID's encoded into the orbiter hit-area indexes with binary ops) - and figuring out what was clicked, its' easy enough to have it flip, play a sound, and trigger the appropriate command


and there's the "separation" i was talking about - instead of triggering the systems response right away, i submit the input into a separate routine that checks if whatever would happen can actually do so under whatever conditions apply....

so each system is by itself, another state-machine :hmm:


this also allows any other non-input event to trigger a systems change using the same logic...
so, for instance - if an engine runs out of fuel, it's fuel-line pressure will drop - and when fuel pressure is lost, the engines quit... this is triggered by "cueing" the same process that would shut down the engines if the pilot were throwing the switches to "cut"


doing things this way is an effective way of stopping the unpredictable nature of player input from creating "special cases" which your code isn't ready to handle (thus, crash or bug-fest)...

and it keeps code readable... which is a good things, otherwise after a few thousand lines, a project starts approaching critical mass... :lol:


and good call about the sounds - i do remember in my years of rock-band'ing, i learned that the worse speakers really do make the best testing conditions
tho, these here are actually quite neat headsphones... then again, that's not my point

 

[ky]

This all sounds great!2 things:1)Will it be Ummu compatible?(I havent followed this thread in a while so I may ask dumb questions)2)Will the spacecraft break up on a bad reentry angle?

 

[Moach]

yes :thumbup:, and yes :headbang:


...eventually

 

[marcogavazzeni]

I am very impressed by you work, I really like ... if I need help with the code for the heat shield.

 

[Moach]

I am very impressed by you work, I really like ... if I need help with the code for the heat shield.

thanks! - and if you can give us a quick run on how to infer proper thermal loads using the orbiter API (getAirspeed?, getMachNumber?, getAtmDens?), i'd thank you some more for it

 

[marcogavazzeni]

thanks! - and if you can give us a quick run on how to infer proper thermal loads using the orbiter API (getAirspeed?, getMachNumber?, getAtmDens?), i'd thank you some more for it

http://www.orbiter-forum.com/showthread.php?t=19023&page=2

 

[Cras]

Why do I get the feeling this is going to need batteries or a ground AC feed to get it started? And Why do I get the feeling the checklists are going to be almost as long as the space shuttle's?

What is wrong with that?

 

[Moach]

What is wrong with that?

who knows... :lol:


but it's not gonna be as long as the shuttle... this is a near(ish)-future craft... lots of things that had to be checklisted on the shuttle are computer-controlled on the '42, so that makes things a lot easier... (for me to program too :hmm


anyways - there WILL be checklists -- more than there are already... fuel pumps must be worked, APU must run, electrics, hidraulics, bleed-air - all should be simulated at least to some degree


i'd expect a 747' sized checklist... not so much an STS' sized one

 

[cymrych]

SO glad to **finally** make it back to this project! Replaced my g-card, downloaded WIP 2, and I'm back in the air!

Can't wait to really give the 42 a thorough shaking down.

Moach!

PS I obviously made it to ISS rendezvous and docking, but it wasn't exactly pretty. As near I can tell, I have enough fuel in the tanks for about 2m/s dV, lol, maybe a bit more with linear rcs but definitely not enough to get my butt home. Gotta work on that ascent routine some more.

 

[Moach]

just to let you folks know - i got some progress done with those missing forward RCS thrusters...

due to the nature of the heatshield setup, they must extend/retract from doors located at the sides of the forward gear bay - now, i'm coding to make those (and the lower-wing roll thrusters) inoperable whenever the rcs doors are closed

it's no biggie... but it takes yet another stretch from what "conventional" vessels can do in Orbiter :lol: - that's when working with dll modules pays off


still remixing and triple-checking those sound effects... i'm not happy with it until what i hear makes me happy - and i can be tough to please sometimes :hmm:


and i think i'll add an attachment point in the bay, then we're ready to call it WIP-3 :thumbup:

 

[Arrowstar]

Looking forward to it!

 

[Evil_Onyx]

Ive just had a go with WIP-2 and got some things to report.(If they have been mentioned before i apologize)

1. It looks great and flys like i would expect it to, given its something akin to Concorde

2. Whilst taxiing the nose tends to bounce, making a 300m trip form the hanger to the runway almost imposable.

All in all it shaping up to be one of the must have addons.

 

[Moach]

the ground steering thing has been fixed :thumbup:

i have no idea why it was acting out like that... quite odd

anyways, i got it taxi-worthy by completely overriding the default setting and "manually" commanding the turns based on rudder position

wip-3 should be more responsive on ground :hmm: