Learning AD2K

[smilo]

have slept for a few hours
and am now on my late night jaunt.
my big monitor is turned on,
but, will take a bit to get warmed up
enough to see the fine details of the model.
the left side is clear,
but, the right is hazy,
fading into darkness.

so, what do i want to accomplish?
there is a strut, built at 0, 0, 0,
(0, 0, 0 is it's pivot point)
i want it to stand to the right,
protrude through an off center wall,
(at a pivot point defined by a vector)
then, on command, lay on it's side,
and be hidden behind said wall.
all this accomplished, using seals,
the vector and jump planes.
the tricky part, is keeping the strut
visible or invisible on command
from all viewable angles.

shoot...that doesn't sound like such a big deal.
the screen is clearing up,
except for the top quarter
and that's improving.
i'd best get crackin'.
i'm wasting wake time.

looking at the strut alignment.
something has been bugging me.
the strut is built on the negative side
which puts the pivot point
on the wrong side of the strut.
the wheel is also on the wrong side of the strut.

i think i'll change that first.
wish me luck...i'm goin' in.

 

[smilo]

been up most of the night.
none of the usual,
go back to bed
after being up for a few hours.

got down to the nitty gritty,
created new S assemblies.
one for the strut,
which i moved to 0,0,0
created another SA called WINGD
which contained one large polygon
the size and shape of the lower wing.
it's also, in the same location and angle.
i moved the vector pivot point
to well above the wingd polygon
and finally, made that polygon a seal.
i tried several combinations
of jump plane entries
and came up with the same results.

i have not tried making another seal
and placing it between the vector pivot point
and the lower wing polygon....maybe?
i also noticed something else of interest.
by adjusting the Jump Gear RANGE
minimum to 200, the strut disappears
just before full extension, but,
is visible when retracted.
the lower i make the minimum,
the sooner it vanishes when extended.
unfortunately, this is the opposite effect
i am looking for...go figure.

next, i want to add JUMP DOOR.
i have a suspicion that it will cover
the retracted gear, but,
i'm not going to get my hopes up too high.

i need a break....want coffee.

 

[aleatorylamp]

Hello Smilo,
We´re just back from the farmer´s market, with a snack lunch on the way.
Well, you certainly have been doing a lot of research experiments.
It´s a difficult one, isn´t it?
The problem of having a short computer session in the middle of the night at 2 or 3 a.m. is that it often wakes me up too much and I´m not sleepy enough to get back to sleep. If this happens 3 nights running, like last week, the accummulated lack of sleep makes giddy during the day... But one gets new ideas and wants to try them out before forgetting them.

Well... Re. the External-Gear-Retraction System:
There´s also a Yak 18 with rear-folding gear where the strut and over half the tyre stays outside the well.
This one can land without extending the gear in an emergency, but it ruins the prop.

In my case I couldn´t move the existing vector down by 0.9 ft, and a newly-made vector would automatically disappear.
It appears that a vector needs a point at the end or at a corner of the part to be rotated, so fixing one at -0.9 ft from the top of the strut is not possible. So what I did was move the strut even further out and adjust the hinge vector to the inboard top corner, successfully this time, so that after pivoting, the wheel would end up flush against the fuselage wall.

This time it worked, without bleeds. It´simple and doesn´t look bad and I may leave it without the skirt flange. If needed, a flange would require a seal between it and the rest of the skirt, as well as the corresponding Sequencing within the SA, similar to the Spinner/Propblades/Propdisk sequencing. Anyway, here´s a couple of screenshots. The belly textures still have to be polished.

Thinking of your results that were only good for either a retracted or an extended gear, but not both:
It´s a pity that a Jump Plane Sequence that can make a conditional call to a chain of SA´s in two different orders, depending on which side of the seal the viewinpoint is at, but not depending on whether your gear is being retracted or extended.

Good luck for your side of research!!
Cheers, nice Sunday!
Aleatorylamp

 

[smilo]

we are a couple gluttons for punishment, aren't we?

i am curious about your vector difficulties,
but, frankly, don't know enough
to offer any help with them.
the best i can for for now,
is set one up at 0,0,0 origin
and then, mouse click to the approximate area
i want the thing to terminate.
(I don't know why i can't just enter the origin
and destination coordinates at the beginning)
anyway, then, i edit by manually inputing the coordinates.
at least, i think that's what i do.
i still have a lot to learn.

i guess the reason i'm stuck on this,
is because, most aircraft with retracting gear,
have the retracted gear stored (hidden)
in the wing, fuselage or nacelles.
(captain obvious here)
therefore, it seems that doing so
shouldn't be a major production
and a four day battle to figure out.
...or, am i asking too much?

i have a couple more things to try
before i go to the jump door.
we shall see how it goes.

 

[aleatorylamp]

Hello Smilo,

Ha! I´d prefer to call it preserverance... but maybe you´re right!

OK on the vector tip. I find it mystifying too.
I found out something new: Apart from the "c" button to elect a part within a SA, and the "n" button to directly go for one vertex, when you are viewing the vectors, pressing the "v" you can select the Vector to edit, and if looking at seals, it´s the "s", and texture supports "t"... it´s easy when you know...

Also, sometimes, interaction with the on-screen mouseclick comes in handy, if you know... Some things went haywire when I accidentally clicked on the screen and I didn´t know what was happening! Once I chose the wrong vector, looking from the front only, and bent the whole wing down! Slowly the mechanics of managing the programme are getting clearer.

I agree, the standard for retracted retractable gears is to be stored, so it should work. I´ll store my present external storing version because it is a clean build, and investigate further.

One quick try perhaps, could be to give Ugly100 a Strut Skirt and see how that behaves on the existing sample aeroplane .3DM, because from what I can see, wing, wheel and strut have no bleeds.

I haven´t done much on the plane today, lazing around "watching" TV I had a long nap.

Chairs,
Aleatorylamp

 

[smilo]

i finally got around to trying
the Jump Door Range.
damn...i wish i'd tried it two days ago.
what does that tell you?

to be honest, i am disappointed
in the reverse functioning
of Jump Gear Range.
that, or there is severe operator error.
why would i want the strut to disappear
when the gear is fully extended?
maybe it's a glitch.
after hours of dinking with it,
i found it to be worthless.

okay, so, the Jump Door Range
is inserted in the model editor,
via the Jump Gear Range.
basically, key Ins, J, E, Enter
while JUMP GEAR is highlighted,
in the model editor instruction list,
key D to cycle JUMP DOOR.
click on the ? and select a SA
that has a polygon large enough
to cover the retracted gear.
it should be in the same location
as the "panel" the gear is bleeding through.
make sure the "door" is facing out
when the gear is retracted.
if it isn't, select it in the graphic editor
and flip it...that's in the draw menu.
i don't know if it's required,
but, i put the "door" in it's own SA.

as for the RANGE settings,
i need to play with it a bit,
but, as far as i can tell,
228 is default and 255 min
is as high as you can go.
the higher the number,
the sooner the door becomes invisible
when the gear is extending from the wing.

now, for the bad news...
there's more work to be done.
when retracted, the gear still bleeds
through the leading and trailing edges of the wing.
to me, this indicates the need
for at least one more seal and jump plane sequence,
but, i won't be surprised if it's two.

 

[aleatorylamp]

Hello Smilo,
Thanks for your detailed instructions to build the wheel door.
Could it be that your gear is still bleeding through the wing because it hasn´t disappeared?
I am also mystified by your report on the strut disappearing when fully extended - meaning that it is still there when up.

If you have RANGE 228,0 the gear or strut should disappear quite a distance before reaching the undersurface of the wing, and it it has disappeared, then it shouldn´t bleed through.

Maybe the RANGE parameter values are thr other way round?
Just in case, the Ugly100 Right Wing Code reads: (KARMAN is the call to the wing-root fairing)

RITEWING	ASSEMBLY
	JUMP PLANE	W10
	PLANE	-3.200;-1.000,0.000,0.000
	GOSUB	W01
	GOSUB	KARMAN
	RETURN
W10	GOSUB	KARMAN
	GOSUB	W01
	RETURN
W01	JUMP GEAR	WING
	RANGE	228,0
	JUMP PLANE	W11
	PLANE	-1.153;-0.072,0.000,0.997
	GOSUB	W02
	GOSUB	WING
	RETURN
W11	GOSUB	WING
	GOSUB	W02
	RETURN
W02	GEAR RBANK	GEAR
	VECTOR	6.300,0.250,-0.700,0.00,0.00,0.00
	RETURN
WING	SUBASSY,T	click to open sub-assy
KARMAN	SUBASSY,T	click to open sub-assy
GEAR	...

Can this help?
Good luck!
Chairs,
Aleatorylamp
P.S. I had awful problems getting the text into the message, and this is the best I could do, otherwise the indents disappeared.

 

[smilo]

yes, you are correct,
the gear strut** is bleeding through the wing
because it has not disappeared
when in the retracted position.
if it did, i would have been done long ago.

one would think that the Jump Gear RANGE
would easily take care of that.
BUT NOOOO.
for some unknown reason,
the Jump Gear RANGE
is acting completely opposite of how it should.
as the retracted strut goes through the extension curve,
it disappears just before reaching full extension.
in a word...maddening.
i don't know if it's a glitch or operator error.
either way, i gave up on it hours ago.
i am currently constructing a "gear" box
that fits inside of the wing.
when the strut is retracted,
it can not be seen inside the box.
i had it working a while ago,
but, screwed something up
somewhere along the way
and couldn't remember how i did it.
damn the luck.
anyway, i got it straightened out
and am now fine tuning the bugger.


**i am just using the strut for simplicity.

 

[smilo]

so...what was i saying about operator error?
please, refresh me with a dose of salt.

humility is a wonderful thing...isn't it?

i just redid the model editor instructions
and, lo and behold, it works. yahooo!
the Jump Gear RANGE even works.
i couldn't/wouldn't help myself
and played with the settings a little bit.
i adjusted the min from 228,0 to 195, 0
so now, the gear disappears as it meets the wing.
of course, this will most likely change,
when i readjust the gear vector,
but, that's not a big deal.

IT WORKS!!!

that was one steep learning curve, right?

 

[aleatorylamp]

Hello, Smilo!
I´m just having my tea-at-dawn . Delicious ...and the cats are fed.

Wow! . Sense of achievement - another thing licked! And, now it´s my turn to feel useful too! Ha ha!
You are absolutely right! Humility prevents you from struggling with broken pride when you make a mistake and can´t bear recognizing it - it´s much more practical.

Anyway, now you can put a skirt on the strut! (Struts must be Landing Gear´s feminin company... how nice!)

I think I had a concept error with the adaptation for FS98/CFS1 of the RFO tutorial come-CubeFly: I thought the problem was TRANS/ROT used instead of JUMP GEAR + RANGE, but I think it isn´t!

Unless I am mistaken, the instruction JUMP GEAR + RANGE can be fitted into the sequence to also control visibility on TRANS/ROT, not only to do so on GEAR RBANK right gear, GEAR LBANK left gear, GEAR ELEV center gear and GEAR ELEVR center gear (reverse -(rear?)).

So possibly by just fitting JUMP GEAR + RANGE into the sequence before TRANS/ROT on the CubeFly it will work.
That way I can flog of my cheapo utilitarian retractable landing-gear kit on the flea market!

Interesting detail:
On the JUMP RANGE instructions for aileron movement I entered 10 for one side and -10 for the other, and automatically the minus ten value was converted to 350, so it´s like 360 minus 10.

Thanks for the 195 entry! I´ll change it to that on mine.

More, later!
Cheers,
Aleatorylamp

 

[smilo]

after i added the wheel to the strut,
i noticed a little visual over lap,
depending on the view angle.
i'm sure a jump plane will solve the problem.
while i'm at it, i might as well add the skirt.

it's nearly midnight and the early AM
wake up and no nap is catching up to me.
i'm off to bed.

 

[aleatorylamp]

Hello Smilo,
Good night!
I´m busy starting to transfer JUMP GEAR + RANGE into the CubeFly...
Get some beauty sleep!
Cheers,
Aleatorylamp

 

[aleatorylamp]

Hello, Smilo,
Er... one thing is transferring instructions into the code, and quite a different matter is getting the correct line!
I´ll get there in the end!
Cheers,
Aleatorylamp

 

[smilo]

please, explain....
what are you trying to accomplish?
post a screenie or text.
maybe, i can help.

a picture's worth a thousand words.

ps...the skirt is in place,
the tire, strut and skirt are sequenced
and the vector has been adjusted.
all in all, i think it looks pretty good.

oh yeah, the gear parts are not textured.
i find it much easier to see bleeds
by using contrasting colors for the parts.

 

[aleatorylamp]

Hello Smilo,
I put back the vector (easily this time) to the outside corner of the strut, and moved it in a bit so that now the tucked gear has the wheel flush with the fuselage again.
Then I was looking for where on the Main Code Page to sift in the listing, comparing it the way the Ugly100 has it on each wing assembly.

Ugly100 sequence:
After deciding the Wingroot priority, Jump Gear calls Wing SA and, with Range 228,0 sets Gear-movement-arc, then calls a Jump Plane, which decides whether to call Wing SA again or Gear SA.

CubeFly:
The Wings are is inside the Main Code Page, so an equivalent place (there´s no Wingroot) to insert Jump Plane + RANGE would be just before RTWNGALL, moving that label to the new instruction, and renaming the previous position that label had before to e.g. RTWNGILL, to be called by Jump Gear (this summoned RTWNGALL before).

But, the model won´t do anything new, and I´m still trying to figure it out.
I didn´t get much time to do so today, as every time I sat down, I was summoned to do different other things.
I know I can do it if I have enough sit-down time!
Cheers,
Aleatorylamp

 

[smilo]

as a personal preference, i disagree with the ugly order.
ignore the wing root, it just gets in the way.
save it for after the wing stuff gets done.

first, do the gear movement...test it alone.
then, join it to the wing using the jump plane...test it.
if everything lines up properly, add the jump gear range.
test it...does it work? adjust the range.
the reason mine didn't work was,
i had a return after the jump plane
and before the gear jump range.
live and learn.

now, when all that is done,
go ahead and add the karman jump plane.

if there are still problems, either post a screenie
or the 3DM in a zip and i'll have a look.

i'm kinda busy myself, today.
i guess i need to catch up
on everything i ignored
over the past few days.

 

[aleatorylamp]

Hello Smilo,
Thanks for the indications!

OK, the Ugly100 sequence as a guideline is a bit different,
and the Wing sequences are in separate Assemblies, but I
was just looking for the principle, which should be the same.

I wasn´t writing my own new sequence for the CubeFly yet.
I´d thought of a less ambitious task to start off with:
Using your CubeFly Sequencing code, just to find the correct
point to insert the Jump Plane + Range instruction.

I´m still trying to follow and understand the chain of instructions.

OK, I´ll try my un-sequenced model (with just the Prop-sequencing
working) and start to apply a gear sequence there, step by step.

This would imply I knew how to start one though... but I´ll see.

Cheers,
Aleatorylamp

 

[smilo]

wonder of wonders...
something unusual has happened.
i, actually, slept through the night.
to bed around eleven and awake at 06:30.
i never sleep for longer than three or four hours.
never, except last night.

jump plane sequencing...where to start.
divide your model into sections.
say, tail, mid fuselage, nose/prop, wing.
keep them separate.
there are many advantages to separation,
especially, when you get into more complex models.
remember when i talked about loosing work?
it will happen...be prepared for it.
better to loose a section than the whole damn thing.
not to mention, trying to keep track of all the codes
and scrolling through a giant list can be daunting.

okay, enough of the scare tactics.
i think you get the point.

say, for example, we want to work on the wing,
but when we compile, we have the wing
and everything else we've build floating around
and blocking our view of the wing.
say, we want to work on just the gear.
same problem, only worse.

open the model editor, see the SUBASSY,T list,
click on the top entry, then the Ins key.
you'll get a Instruction Code window.
the default selection it Return.
while you're here, notice, each entry
has a letter in its title underlined.
these are the keyboard shortcuts.
also, OK is highlighted with a dotted line.
yes, you can make your selections with the mouse,
but, trust me here, when you do this hundreds of times,
tapping a couple keys is much quicker.
okay, so, we just want to insert a return.
key Ins, Enter your done.
a return is like throwing up a wall.
nothing after it on the code list will be seen.
but you want to look at the gear assembly, right?
click on that Return, just to highlight it,
key Ins, S (for Go Sub) Enter.
see GOSUB highlighted with a ?
that is a sub assembly call
click on the ? and then click on the SUBASSY
you want to see when the model is compiled.

not so fast, turbo.
did you try to compile and get an error message?
let's say you selected the gear assembly.
scroll down to the assemble you called with the GOSUB,
double click to open it...look at all those TPOLYs.
DO NOT BE TEMPTED to assign them to textures.
that's a whole other can of worms
we don't have time for...just yet. trust me.
click on the first one to highlight it,
key and hold Shift and click the last one.
all should be highlighted.
now key Ctrl and P. select shaded(S) OK(ENTER)
they are now all GPOLYs.
look at the list of parts.
some have the first number 16, some 4, some 5
this is the number of vertices in each part.
click to highlight the top 16 GPOLY,
Ins, C, Enter , click on the ? and select a color (preferably a bright one)
do the same thing for the 4 group and the 5 group,
each with a different color.
now compile
if you get another error message,
look at the left side of the main model editor page.
if you see a blue dot...there is an error.
if you see untextured TPOLYs, just change them to GOPLYs
you don't have to assign colors until you need to see them.

okay, so much for the first lesson.
apologies if it's convoluted.
there are a lot of details to cover
and i don't know how much you already know about.

i'm going up for my morning coffee.
let me know what you see in the sim.

 

[aleatorylamp]

Hello Smilo,
Thanks a lot for your detailed e-mail full of tips and tricks!

In my two last e-mails I made the mistake of saying I wanted to insert a
Jump Plane + Range instruction, when I really meant Jump Gear + Range instruction.

I hadn´t been able to post anything because there wasn´t much to pinpoint or say.

I tried in a few inconvenient points, just after the Jump Plane + Vector instruction
that gives rise to the calls leading to the Gear Bank + Vector instructions for each Gear SA.

Results were varied, none any good, although the gear did disappear when retracted,
but wasn´t animated. One try hung AD2000 and the whole computer because I´d called RGEAR instead of RWING from Jump Gear, and of course RGEAR from GEAR BANK straight after, causing an endless loop, probably. Another try hung the Sim, because I´d stuck in Jump Gear right after the Jump Plane + Vector further above, messing up the conditional flow.

Well, now I´ve just tried try a point before each of the the Gear Bank + Vector instructions, marked in red boxes in the code screenshot attached here. There are also 3 screenshots of the plane.
This is the best so far: No bleed when retracted, but no animation, and THREE landing gears!!!
Why is there no animation? The gears just appear and disappear?

The next step will probably be starting my own Gear sequence, for which your detailed e-mail will come in very handy, unless some able and cunning manouever can get rid of the the center gear, and animate the other two...


Thank you for your time and effort!
Cheers,
Aleatorylamp

 

[smilo]

i hate to say this, but must,
this is exactly what i was talking about
in the last post.
i see a main page, jam packed with calls.
small wonder there is confusion.

try this.. be forewarned, it's radical.
close ad2k and save when prompted.
i'm assuming your work directory
is in the main ad2k folder.
(mine is C:\AD2000.25)
i made a desktop short cut for ease of access
find your and open it and open your cubefly folder.
copy and paste the cubefly (or whatever you've named it)3DM.
restart ad2k and open the copy of cuberfly 3DM.
click on File in the upper left corner
and select Rename from the menu.
change the name to RITEWING,
make sure the master project box is checked and click OK.
(how are we doing so far? still with me)
before we forget, reduce ad2k to the task bar
and open the cfs/aircraft/cubefly folder.
locate the air file and copy and paste it.
(it's probably named cubefly.air, right?)
rename the copy, ritewing.air.
without it, you will not be able to view
the work you do in cfs.
okay, close out folders cfs folders.
restore ad2k to full size.
just to make sure we're on track,
compile the model, go to cfs freeflight
and make sure there is an entry for Ritewing.
it's important that the Ritewing model
is in the select aircraft list.
if it's not...STOP..we need to find out why.
if it is, key Alt, A, A, press the arrow up key
(the keyboard should have a small group
of 4 arrows, left, right, up and down)
to move one aircraft up the list and key Enter.
okay, minimize cfs to the task bar.
ad2k should still be maximized,
if not, make it so.
now for the really fun part...
in the graphic editor,
first, click on View and uncheck Show templates.
(at this point, they're a distraction)
select the canopy sub assembly.
click the Priority chn button.
now, click on the all button,
then, the sel button.
(this way, you'll only see what in the canopy assembly)
are you ready for this?
up on the menu bar, click on Sub Assy,
click, Delete sub-assy.
yes, you heard right.
continue this procedure until all that's left
is RWING, RAILERON (which is a useless piece of crap)
and the RGEAR sub assemblies.
now, go into the model editor (F9)
and delete everything that doesn't relate
to the right wing, right aileron or right gear.
(gosubs, jump plane, returns...everything)
now, you can see what needs to be seen
to accomplish the task at hand.

unfortunately, ad2k does not have a back button,
which would be hand if you accidentally make a mistake
and intentionally delete something you didn't want to.
there are a couple solutions, but, that's for later.
for now, just be careful.

when you've finished the hack job,
click the save icon, open the ad2k cubefly source file
and copy/paste the ritewing 3DM
right there in the file.

that should keep you busy for a bit.
i'm sorry, but i have errands to run.
if you cant stand waiting,
in model editor,
see the RGEAR2 GEAR RBANK
and the RGEAR1 TRANS/ROT instructions.
these entries are the right gear animation calls.
scrolling up, see the right wing to rgear2 jump plane.
this instruction ties the two together.
the jump gear range is added above the jump plane.
be sure there is no return between the two.
if it was me, i would insert a return at the top,
then a gosub above that.
click on the ? and select rgear2, then compile.
go to cfs and have a look at it, does it work?

the trick is not to get ahead of yourself
until you understand what you've accomplished.
one step at a time, buddy, one step at a time.

okay, i'm off.