Open Source Airplanes

Jes(se)(ka), wasn't sure where this thread should go.

Hey all, now that vehicles work a bit better in 1.7, I thought it'd be fun to release a free (as in beer and in libre) airplane with more features than Andrew Linden's basic flight script. When I joined SL, tearing apart the free plane and modifying it were part of what got me hooked on SL.

What kinds of features do you like to see in an airplane?

Here's my list so far. Many of these features we've come to expect. Would really love suggestions and additions!

What's already incorporated or within easy reach:
-Flight and hover states
-Variable geometry wings, dynamic canopy, retractable gear
-Blinking particle running lights
-Exhaust particles scaling with throttle
-Variable throttle
-Extra passengers
-Cloaking
-Phantom in flight, nonphantom on landing
-Autolanding
-Touch dialog with:
-Custom trim and main surface coloring
-9 preset texture options
-Vehicle beacon system (based off Cubey's excellent and open one)
-Autopilot (based off Jeffrey Gomez's pathfinding algorithm
-Particle weapons
-Prim guns, rockets, intrasim guided missiles, airmines, dumb bombs
-Client SetText HUD:
-Coordinates
-Airspeed
-Throttle
-Sim warnings (time dilation, altitude, FPS)

What's I've not even attempted yet:
-Ground effects as inspired by Jillian Callahan's excellent planes
-Augmented physics as by Huns Valen's excellent planes
-Handoff control to co-pilot
-Mouselook flight
-Selective ejection of passengers
-False horizon

What am I missing?

Cute copilot?

How do you figure that vehicles work better in 1.7?

What I've noticed is that most things are pretty much the same as it ever was, but with the new script scheduler, vehicle behaviour is more sensitive to sim load.

Is saying "Jes(se)(ka)" just being cute or are these two aspects of one deity or two avatars run by one employee?

I don't think you should put everything in one free package. As you said, it's fun tearing these things apart and adding features... plus a more modular approach will avoid the 'everyone uses x-flight' problem and keep things varied and interesting.

What I'd like to see is a basic plane script that handles landing and takeoff cleanly, automatically switching from a vehicle mode to a flight mode when you hit take-off speed with the controls (flaps, boosters, whatever the user's simulating) right, and vice-versa on landing... and handles sim transitions well.

Any voice commands on channels other than 0, and let the user easily select the channel.

A simple hud (say, an artificial horizon/altimeter you can click on to drop a dialog box) would make the need for "gotta hit b return can't make it /3b that's too slow" go away.

Any non-HUD pilot responses via llOwnerSay() or llInstantMessage() rather than llWhisper(), and let a guest pilot click for a free copy of the default HUD.

Documented and consistent link_message forwarding of commands and flight states for secondary scripts. Provide sample scripts for jets and lights and canopies and stuff, but make it simple enough that people can do their own.

Panic voice command (eg, "/1sos", for when sim crossing recovery fails, that stops the plane and makes it go non-physical. I just had to derez a plane and plummet to the ground because it went into a screwy configuration and lost inputs on sim crossing.

Simple flight and sensitive flight modes.

The plane itself? Encourage people to make their own. The sample plane should be clearly a sample. If you do a fancy plane, make enough of it non-mod to encourage people to do their own thing.

Look at the Flying Tako. There's some great scripting there, but if you want to customise it you start over with LibrarySailboat.

I'm just going on my own impressions; making it farther across the continent before getting dumped out of my plane at a corner crossing, more responsive handling, less of a jolt at the average sim crossing. Maybe I've just had on rose-colored glasses for the past week. In any case, now that 1.7 is stabilizing a bit, this seems like a fun project.

Kudos on the Dominus, btw.

I thought cute significant others were mutually exclusive with scripting?

Both forum editors, both start with "Jes", both ar... wait... "JESUS" starts with "Jes" too! Jeska is the Second Coming!

OMG. Literally.

Back on topic, okay, I'll script in a deity. =P

How is the 'everyone uses x-flight' issue a problem? If 'everyone' uses it, then it clearly it fills a need. Granted, x-flight may not be the most elegant solution, (hey, we've got ROAM now, right?) but its popularity has got to stand for something. I'm truly not trying to be contrarian; I think you've raised an excellent point.



Vehicle mode to flight mode. Great idea. I'm not sure how to approach handling sim transitions well except to minimize the number of scripts and sheer size of the physical model.



Great call here. We need less chat spam, as in zero.




Yes, and being simple and consistent in my coding will make MY life easier too.



Hmm. Hadn't thought about that. I wonder if the script realizes it has lost control input.



Definitely. I'm using a plane model that is suitable to my tastes, but vehicles seem like such personal expressions in SL that I'm pretty sure I'll release the scripted plane in the plywood texture. This is doubly a good idea because I stink at texturing.

Hey, thanks for the thoughtful input. Much appreciated!

*bump* I know there have got to be more ideas out there.

Well, I will look at it from a pro/con standpoint...

Pro:

1. Make it easier for people who are good at geometry but not scripting to put together a good product relatively quickly.
2. Save people from having to do their own R&D on plane logic.
3. Possibly encourage more people to sell airplanes.
4. Raise the bar for closed-source developers.

Con:

1. Remove some of the incentive vehicle builders have to learn how to put together a good set of scripts, leading to more "generic" aircraft.
2. Releasing a fully-featured script chips away at the competitive advantage of those who have taken a lot of time and effort to develop their own. Imagine if someone decided to design giant mech avatars similar to your own, and then gave them away with full permissions. It wouldn't be very good for your sales. In real life this is considered a predatory business practice, and corporations such as Microsoft have got in trouble for it. (I'm not trying to compare you to Microsoft or anything, especially since your intentions seem good.)
3. There is already a glut of aircraft, so inducing even more people to sell them is likely to drive the prices down even further, and they are already ridiculously low for the amount of effort that goes into them. (At least from my perspective.)

That second con is mediated by the fact that vehicle makers are free to continue to innovate in their own scripts. If their scripts are significantly better than the open-source scripts, people will still buy the (closed-source) aircraft.

The third con is also mediated by the ability of scripters to continue to innovate. However, if you remove some of the profit, you remove some of the motivation to continue developing.

If you really want to help, I would suggest posting technical discussions of how to achieve these things. If you just give people a 100% finished product, and it's open-source, there isn't as much motivation for them to actually learn. You don't learn the hard stuff without making an effort to assemble it yourself, so if it's already assembled, there isn't much point.

At a more basic level, doing this could increase the "selectional pressure" (viz. natural selection) in the aircraft industry. I'm confident I can rise to that challenge, as I have in the past. However, I don't know if others in the industry would necessarily appreciate it.

There's about 40 only slightly different versions of X-Flight, all of which are kind of annoying to use if you're just trying to get past the "ground effect" zone, and at least half of the popular ones still talk or whisper and listen on zero. So if you're at a telehub you hear a constant chatter of "Carbon Rod: All go", "Purple wings: All go", "X this", "X that", ... imagine what the Welcome Area would be like if it wasn't no-script!

I know that part of this is the results of the time that it was written in, there were fewer llFunctions available, but that's the point. If this plane is a one-size-fits-all rather than a kit it'll end up having a bunch of only slightly different versions that don't keep up with the times. If there's a variety of planes built from the same family of components, there will be more ongoing revisions giving people a reason to upgrade.

I know that you can't prevent bit rot, but if you encourage people to do their own thing there will be more of a software ecosystem and less of an accidental monopoly.
Basically, having an easy way for the pilot to recover and force a clean reset if the scripts miss identifying a transition is the main thing I'd like... as for my problem, it did at least refrain from trying to use permissions it may have lost, because I didn't get any dialog spam...

In the past, the concept behind free scripts was that they're a starting point from which to learn and innovate. I agree that releasing a full-featured, free script will put several aircraft makers out of business, and bring in a host of aircraft makers who really don't understand the craft. As Huns suggested, there will be a glut of cheap aircraft all made with the same high-quality script. The business of designing aircraft would be, in effect, shut down.

So... if it's so easy to undercut a market, should there even *be* a market? If anyone can step into a given market -- be it homes, skins, clothing, gadgets, pose-balls, etc. -- and give away a free product that's as good as or better than the leading product that costs hundreds, why should anything cost money? Should content makers be allowed to charge money when someone else gives an equivalent item for free?

My feeling is that if at least one person is willing to give away excellent work for free, then charging any amount for an equivalent item is probably unethical -- it's gouging. It's like taking freebies and reselling them -- you'd be tricking unwitting customers into paying for something that's free elsewhere.

Thoughts?

I've bought quite a few planes, and the main reason I bought the ones I did was because they were interesting for some reason or other... but that reason was rarely the scripts, and where it was I've mostly been disappointed. I've only been satisfied with one that I bought specifically because of the way it was supposed to perform... the Cygnus G-1. I bought Cubey's Ornithopter and Tiny Balloon because of their designs, just like I bought Gillian's Skitter, Chage McCoy's Mehve, and I'm sorry I don't recall who made the stealth fighter...
Well, both Apple and Microsoft have incorporated components of Open Source operating systems into their own products, without Windows NT or Mac OS turning into just another UNIX.
This is a problem, there's way too many people who just want a version of Windows that runs on Linux, and too many developers willing to accomodate them.
It can also bring people in who are accomplished modellers but currently intimidated by the difficulty of scripting aircraft. Especially if it's nicely modular. Especially if the scripts are distributed with more of a BSD style license, rather than the GPL.

All assuming that I can (a) pull off the scripting and (b) do it in a way accessible to other developers. I'm still very much learning scripting and so I have legitimate concerns about this. =)



I certainly don't want to run afoul of predatory business practices. Even though there's no governing structure in SL and therefore no laws against it, it doesn't seem nice to be predatory. SL is generally such a friendly environment inworld.

OTOH, if I do it in a clean-room approach, does that make it more acceptable? This is something I've pondered a bit because I'm aware that such decisions don't happen in a vacuum. I have great respect for you, Jillian, Cubey, Jon Marlin, and the others; from a quality control/marketing/sheer programming skill perspective I cannot match any of you. That respect, though, is in tension with a desire to open up these arcana for scripting neophytes like myself.



I agree with you-- this is a valid point. Though the script would be open source, it is clearly tweaked to one user's preferences and needs with respect to performance, interface, etc. Anyone wanting to change those significantly would need to learn at least some scripting. As far as finished product goes, the closed-source scripts *will* be better as in more polished and probably much more optimized vs. my approach. For example, to aid in usability, I broadcast linkmessages to all the children prims instead of targeting them as I normally would. The reasoning was that if the geometry changed, folks wouldn't break the scripting by making their own chassis(es?).



I'd be curious to hear the thoughts of the Seburo folks as it appears to me that gunsmithing and aeronautics engineering may have a lot in common here. There are hundreds of free guns out there but only one Seburo. Does releasing full-featured gear as open source prompt a shift toward brand management and away from script development. Something I'll have to ponder.

Seems like parallels exist here to the open source software vs proprietary software issues.

Speaking as someone who has been writing open source software since before it had that name, since before the whole question of licensing it was more than "this is public domain, have fun", I really get a kick out of seeing someone put the stuff I'm NOT good at into my software and use it in stuff they make money from. Even when that someone is Microsoft.

It really bothers me when builders in SL lock their scripts up so I can't customize them, even when the script is clearly not what makes their product valuable. It's amazing, sometimes... someone's got an object that's taken hours of detailed work to get just right, even without considering the textures and animations involved, and then they make a big deal about a 20 line script that I eventually replace with something from the scripting wiki that's more functional.

The trick is to be doing something that's beyond what the open source community can do. And there's lots of things that the open source community isn't good at, really, and you're one of the ones who's good at some of those things...

Respectfully, I guess I just don't agree with this supposition, Cubey. Sure, folks can drop in a series of scripts that add features. However, to be better than a Terra, Marlin, or VHI bird, it has got to be *awesome*, ya? Modeling, texturing, brand management, support, the whole nine yards. These scripts are only one ingredient. Not many folks can compete with y'all overall. When someone buys a Terra plane, it isn't just a plane; it is the prestige that the Aerodrome crew has built over these many months. That's not flattery, just an observation. A RL fellow researcher in my group wears $US100 jeans. I think she's crazy. But to her, it is more than jeans, it is the designer label.

Of course, the scripts in question have to not suck, also. I'm working on that, but not there yet.



/me ponders these things.

I'm reminded that I need to stop by both the Aerodrome and Kage's mecha shop. >.>

So Kage... let me just see if I follow your train of thought. You basically wish to emulate/copy whole cloth all the best and unique features of all the top air-craft's scripts from all the top makers and then give that... open source mind you, away to anyone for free?

And you don't see how that would, if not ring the death knell of the aircraft industries in SL as a whole, it would certainly cripple it until a new batch of scripts and tricks are produced (at which point one would assume you'd release all those new features for free as well)?

Right then, I guess it'd be ok if someone were to produce a steambot identical to your own current design and oh say, change the name, colour, put the rotating cog on the left shoulder rather than the right and maybe even add some better scripting, few more rotating bits, more creative use of particle effects and then give it away with full perms to any one whom wanted it, along with a note saying "Hey buddy? Want to start a great SL business? AV's is where its at! Just Plop this little puppy in a vendor of your choice and have at it Skippy! You'll be rolling in the Lindens in NO TIME!".

In fact not just with your steambot AV but your entire series... and why stop there, Marcos's bots as well, oh.. hey! Enabran's siege bot... that's getting dated, scripts wouldn't be too hard to duplicate either. Or how about a sensor based weapon system identical to frag-foo?

In short any script based object in which the script makes the object 'special' or in which said script is equally important to its success.

Brilliant.

I think Cubey is dead on, and not just about the crippling of the aircraft industry by the influx of fully featured craft and creators whom know less than beans about scripting but about any other script dependent SL industry.

Er, Mem? I think Kage is talking about independently making a script with many of these features, doing the scripting himself (or with anyone who wants to help), etc, etc.

I'm inclined to agree that a lot of what makes an aircraft sell is style, not JUST the script, along with creator support, brand name, etc, etc.

It's just most people willing to do that much work on a script aren't real keen on giving it away. But there's always the chance of someone pulling a Crystalshard Foo in the airplane biz, or any other.

And I don't think that's necessarily bad - I have the Freeview and still bought Captain Barmy's TV, simply because he went to the effort of compiling every single usable stream from archive.org. Similarly, even with a high-quality free airplane script, I imagine I'd STILL value the highly unique aircraft I've bought - which is ALL the aircraft I'm willing to buy. Hun's autogyro, Cubey's ornithoper, Ardith Mifflin's mouselook fliers... because a generic script simply could not produce those unless the creator in question spent as much work customizing it as they would building it from scratch.

I feel the parallel is a bit rough, given Kage does give away the "engine" to his mech avatars - the mech override kit, not to mention the tiny override kit. An open source script doesn't built a cool airplane by itself.

Hey Ali
Difference between the Freeview and what Kage's proposing? Freeview is akin to the LL free airplane script. Its a starting point. Thats ok... and as far as it goes so is freeview. Its a starting point, nice one with some bells and whistles but not as advanced as it could be because (and speaking for Foo here even thought its purely assumption on my part) she wanted a nice a free script out there that got the job done but it wasnt something she wanted to include every possible bell and whistle in.

Its a basic lovely script that works and works well, but it hardly has every bell and whistle that a media centre could.

Case in point and continuing with the freeview example, my friend was commissioned (and has since) a few weeks ago to create a mediacentre that could be sold. In this was scripted the abilities to sort, notecard entry, turn on or off parcel music, had web interface, its screen adjusted for aspect ratio, it had remote screens it could rez and de-rez on command -and the remotes would focus on the nearest AV, you could set it for all access, group only, owner only, by notecard, by time of day and lots lots more.

Now what Kage is proposing is giving the media centre example away, for free and open source. As opposed to the freeview, a grand script t'be sure but basic in functionality a 'starting point' much like LL's free airplane script.

Freebies are well and good. Freebies have a place and can stimulate a market, the crux of the matter is in knowing where to draw that line.

Do you give away a starting point or make an concerted effort to include every bell and whistle and innovation? Which is more likely to help or hurt? Which is more likely to cripple the learning process or encourage it?

Thanks for posing the situation in a frank way, Memir. This is a solid discussion.

Kind of, sort of; the feaurelist above is my goal. However, I can tell you right now that I cannot match the scripting skill of the top aerospace designers in SL. The entire extent of my skill is about six months worth of hammering the LSL wiki and trying to figure out how folks can make LSL do such great things. I cannot copy whole cloth (and I would not, given the option), others' work thanks to the permissions system. I will emulate to varying extents varying features.

For example, the particle-based ground effects of some helicopters is a really nifty touch. The aircraft is near the ground and its engines kick up dust. Several closed-source vehicles incorporate this effect and I think that's great. I think it'd be great as well if an open-source vehicle did too, partly so people can see one way of achieving that effect. In addition it is partly so the current influx of poor newbies on free basic accounts can learn what a fun thing that using a fully featured plane in SL can be.



No, I don't think the release of an open source plane could kill or cripple a robust industry in SL. Yes, margins are tight, and that is in my limited knowledge one hallmark of a healthy industry. But any market that supports a variety of vendors at places as large as Abbott's Aerodrome and Jillian's complex surely must have space for open source "competition." I enquote that because my implementations of features so far are (a) not replicas but analogues and (b) less elegant for the sake of robustness.



This situation is a valid comparison, Memir, and it has happened to me (albeit with another av, not the steambot). Using my DIY kit, someone faithfully replicated one of my mecha avs down to the tiniest detail, then went on to create variations and added features that I don't know how to do. It was amazing, unbelievable. Hundreds of prims and some fresh scripting. I've never since seen an effort quite like that. I don't know if the emulator went on to sell that av, as s/he certainly could have. Ultimately, I didn't pull the DIY kit because it was then and now my most popular item (next to the tinies DIY kit), even though I make essentially no profit from it.

The lesson I took from that episode is that in the end it isn't about the av (or the airplane). It is about self-realization and actualization. Insofar as I may speculate about the motivations of SL users, I think that folks want to (be able to) experiment with their own thing, in contrast to closed-source buyer lock-in. I suspect that this dynamic is part of what makes Luskwood and the Furry scene in general so dynamic and progressive. The avs are moddable (though to be fair the blink script is not afaik). Of course, I'm not a furry/vendor, so I'm admittedly not close to that situation.

For the sake of disclosure, I should note that at my old shop in Bisque, now moved elsewhere, I hosted a variety of other giant mecha-makers' vendors free of charge. I think that doing this demonstrates the difference in my perspective from yours. My take on things is that supporting fellow enthusiasts is pretty important, and at least as important as seeing to the competition of the business side of things. By supporting them I mean to say 'furnishing tools to help them realize their goals and designs,' and that included vendor space and sellable versions of the integral animations.



You are 100% correct here, Memir, but I think there's more to the issue. As I mentioned in an earlier post, there is so much more to a Terra plane's gestalt than just the scripting.

I'm getting long-winded. At the end of the day, so far, it is boiling down to this:

Position 1:
Don't release a fully featured plane because it will kill an already-glutted market.
How would you feel if someone did it to your mecha?
This will lead to a lot of cookiecutter planes.
Dude, it sucks to copy features from others.

Position 2:
Releasing a free and open plane will bring fresh minds to aerospace engineering.
Wouldn't it be nice to be able to start from the point of all these features already in place?
Don't sweat it, this plane won't be as nicely polished as yours anyway.
I am recreating features, yes, but in a clean-room way, except where open-source alternatives already are in the wild.

I see really valid points on both sides. In the shower this morning I pondered releasing these features but in a sense dumbing them down a bit. For example, take guided missiles. I have implemented them but they use llSetPos to get to the target and are clearly inelegant. Would that be a better tack, or does that inflame the "this will lead to a lot of stupid planes" postulate too much?

PS-- how could I forget to mention Konigmann Lippmann's excellent flyers?? Sheesh.

Thank you in turn for the well thought out and civil reply, mine tend towards being a bit too frank sometimes so my apologies if they were on this occasion as well.

As to the topic at hand, we're at something of a impasse as to opinion then Kage.
I think my post to Ali sums up my opinion /108/b4/72144/1.html#post753741 and I'd reiterate my questions to her to you.



Not to speak for Cubey, Huns, but I rather think that was their point as well. There are already starting points in the airplane industry in SL, several good ones that are open source, but do not do _all_ the "heavy lifting" for a potential builder LSL wise. Why then hobble them straight off by doing every neat thing they'd like to do, or might like to learn to do for them? Also keep in mind that there are a goodly number of builders and script-crafters whom would most certainly be impacted if a full featured open source were released or eventually is, no one can stop you after all if you decide to, but I'd encourage a lot more thought.




Okies, on a purely personal note, I'm going to ask you not to assume my perspective. As a member of Phase 5 we encourage and host a goodly number of new and up coming aircraft builders at our HQ in Caldbeck, providing vendor space and scripting help.

At Abbotts we've a hanger dedicated to the CCC weapons system and open to any builder whom wishes to utilize it to publicize and sell their craft.

Finally I personally spent a substantial amount of my own money promoting and setting up the largest air tournament/event to date in SL history (to my knowledge, I was told this and being relatively new I assume its correct) complete with and entire vehicle sim rental for a week and a open vendor space to any one whom wished to sell their CCC equipped vehicle and a 20k prize purse.

My point being is we both support enthusiasts, and I rather think we're both on the same page as to freebies. I think our parting comes to the degree with what should be made freely available.

In my, opinion you aren't proposing to provide a 'starting point' but rather a end point, one in which by providing all the innovations others have invented, discourages rather than encourages the learning process and in effect cripples potential builders whom would utilize said rather than to learn and do on their own.

After all why bother learning LSL, someone is surely to come along eventually and re-release another open source craft, and later another... and another... and another.

You are correct-- I did mistakenly put words in your mouth with respect to supporting other users. Looking back at my post I see that now. Honest mistake.

I'll take some time to consider the issue at hand.