Learning AD2K

[aleatorylamp]

Good morning, Smilo!
How strange - mine´s transparent grey seen from the front or back, with or without enabling hardware accelleration.
Then, the blades or the disk bleed through the fuselage seen from the rear.
Could this be a graphic card issue with one failing on my machine and the other on yours?

How strange... I´ll try it on my laptop - it has AD2k on it too.
I´ll also try it on the old 2.6 Ghz Pentium 4 tower and see what it does there.

Update:
I´ve just tried it on the P4 and the prop isn´t black seen from behind, and the bleeds are the same.

Then, I´ve just finished the tail section of the Tutorial 98. There are a few mistakes in the tutorial text, and some things taken for granted and badly explained, but I suppose nothing is perfect.
The translation of the rudder to the centre for the animation vector worked, and the rudder is in position, but the one for the elevators didn´t (not explained well, taken for granted after the rudder explanation) and elevators are in the centre. I´ll manage to get it sorted out, though.
But... what a pain, the jump plane sequence. Not to write, but to make. I couldn´t create one like that - at least not yet! ...and that only for the tail-fuselage and rudder empenage. Anyway...

2nd Update:
Got it! Ha! Wow! Phew!
The Copy/Translation of the elevators towards the centre of the plane hadn´t worked, together with the vector hinge, which was positive instead of negative. Now the tail empenage of the 98 tutorial works perfectly, textured, animated and bleedfree!
I suppose once one gets used to the totally manual way of putting together the plane - the only automated part is the semi-automatic way of making the bulkheads, and the automatic way they are covered! But OK... I see the point.

Now for the forward and mid sections! On we go!!

Cheers,
Aleatorylamp

 

[smilo]

BINGO!!!

open the cubefly 3dm
go into model editor
scroll down to and open
the PROP sub assembly
scroll down to SPINNER at the bottom
see the last entry,
TPOLY 16;821, 822, 823...and so on.
this bugger is causing a dark disk
shown when looking forward.
click on TPOLY and select HIDDEN
and it disappears when you compile.
finally, viewing the compiled
prop/spinner and forward fuselage cube model
i see no bleeding and the prop animation works.
i think i'll go back to bed for a few hours.


i wrote this around 04:00
but, couldn't post it.
the soh server was too busy.

i just woke up.
...need coffee

 

[smilo]

...now that you've experienced the tedium
of manually setting up a jump plane call,
try this...

within the cubefly 3DM you sent,
open model editor, after MASTER,
clicking, JUMP PLANE BODYPROP
should highlight it in black.
okay, here's the fun part,
press, Shift+Ins

tell me what happens.
(this also works in sub assemblies)

 

[aleatorylamp]

Hello Smilo,
You´re a wizard! Thanks for the fixes. Let´s see if I can do the same as you say in your 2 posts tomorrow.
Now it´s past midnight again... we had a small dinner party with one of my daughter´s friends.

I´ve been progressing with Tutorial 98 until the afternoon and I´m just infront of the windshield, with the hand-made Z-Buffer finished for Tail and Mid-section.

There are one or two glitches in two templates regarding ties early in the tutorial that cause incongruencies in the cover template instructions later on. It prevents covering some templates and requires minor manual adjustment of panels, but nothing serious.

Did you know that the TAC scanners in hospital also use a very effective Z Buffer to get the images of the human body organs - seems logical... and we have to do it manually!

The next step is figuring out how it works...

In the sub-assemblies the sequencing works by putting labels for different groups of parts that make up sections within the assembly - interesting.

Thanks again! Cheers,
Aleatorylamp

 

[smilo]

i kinda figured you had a family gathering.
it's the weekend, after all.

z buffering in hospital scanners isn't surprising.
remember, it was fs2000 that started using it.
that's almost 20 years ago.
most of the bugs should have been worked out by now.
heck, i wouldn't be surprised if something
completely different is being used now.

as i was compiling this morning,
it dawned on me that i used
the compile for fs2000/cfs2 mode.
as i recall, i used it back with the 196.
i don't think it makes a difference.
see what you think.

sleep well.

 

[aleatorylamp]

Hello Smilo,
I was wondering why occasionally I got one full JUMP PLANE sequence (incl. PLANE vector, 4 GOSUBs and 2 RETURNs) when I didn´t want one, and usually only one GOSUB... I´d pressed SHIFT+insert instead of CTRL + Insert.
So that´s how you get it, thanks! Fine!

Something appears to have gone wrong with the CubeFly. The file I sent you was not the last version with the rear-view bleed problem. I do apologize for all this mess. It had the polygon that should have been hidden, but only the front of the plane appears, so something´s wrong. But: Despite the missing tail, the prop doesn´t bleed through the main Body anymore! The funny thing is that the file has a time just minutes prior to the file I have, long after I´d already put the disk in the Desk...

I reverted to the latest backup, where that polygon is hidden or in the Desk, and I´m still getting the glitch.
I have to see what´s going on and I´ll post the correct file. How confusing. I´m so sorry!

Then, I was thinking about our manual Z-Buffers, which actually the hospital scanners still use, but with more than 32867 bytes... and also, about the way that you have to make templates, tie some of them to others, cover them, and Bingo!, your bodyworkers have panelled the aeroplane´s skeleton!!

Ivan:
The impression I get about how your mind works, I expect both these aspects of the workings of AD2k would appeal to you, and that after getting the grips of the programme a bit, you would like it!

Cheers,
Aleatorylamp

 

[smilo]

at one point, i had a sneaking suspicion
we were looking at different cubefly 3DMs,
but, i figured, what the heck, and pressed on.
i was having fun trying to isolate the problem
and when i found it...boy, i thought i was hot stuff.
thanks for the moment in the sun.

please, don't be too sorry for the snafu.
there is a lot to keep track of and things happen.
it's not a big deal, honestly.

i'm guessing the forward view bleed issue
is caused by either, a misplaced seal
in the prop disk location, or,
an incorrect labeling sequence
in the second jump plane gosubs.
for example,
the first and second gosub gosub calls are the same
instead of reversed.

now, i'm curious...please let me know
when you figure it out.

 

[aleatorylamp]

Propeller-body bleed fixed

Hello Smilo,
Well, I´m definitely glad you are enjoying it! I was fearing I was pushing you into something you really didn´t like.

Anyway, like you say, all is not lost: Although it appears that the 3DM file showing the complete plane with the corrected sequencing seen from the front but not from behind is indeed lost.

However, I have a backup without the sequencing in the main part, but as soon I apply the sequencing, the same happens as in the file I sent to you: the tail disappears , so it´s the same as the file I sent you except for the not-hidden propdisk.

Looking at the sequence, after it, successive calls that should have been erased, duplicate the parts of the plane, overriding the sequencing. Stupid mistake.
Also, I forgot a RETURN after INSERT TAIL, so everything after that reappears too! Even more stupid.
Before, reversing the GOSUBs like you said, had no effect, so it wasn´t working anyway!! , and there´s a bug in the works.

All in all, it is all quite a mess, but the good thing is that it should be sort-out-able by just using the latest file, which is the one I sent you, but hiding the propdisk! So I´ll take it from there.

I keep wondering what Ivan is thinking about all this...

Update: Success!
Backtracking, I re-made the file that was working with the correct sequence within the PROP Sub-assembly, which has the spinner displaying infront of the blades when seen from the front, and the blades before the spinner when seen from behind. (The propdisk is hidden).
The sequencing I had put into the main text at the beginning was bad, so I removed it, and the Tail displays again.
Then, without a JUMP PLANE instruction at the beginning of the main text (!!), but with GOSUB PROP1 in first place, it works!! Too simple, probably... but the body blocks out the prop when seen from behind, which was the aim for this step of the tutorial.

Attached herewith is the CubeFly 3DM file, V1. An attached screenshot shows the propeller infront of the nose, seen from behind. Restly plane bleeds all over the place, but that´s for later.

Once I finish Tutorial 98 with the Ugly100 I called "Pony" and coloured blue, what I learn there will serve to continue with the manual Z-Buffer for the CubeFly. The Pony is only visible as a Tail, because wings aren´t built yet, so the Fuselage at the cabin, which is done, won´t display because the sequence doesn´t include wings. The texture on the fin is wrong as yet.

Cheers,
Aleatorylamp

 

[smilo]

well, it looks like you're making progress
and that's a very positive note.

my apologies in advance,
i'm more than a little ticked off today.
NO!!! it's not something you've done.

i'm getting damned tired of writing replies,
only to have a "the server is busy, try again later"
message pop up and then, loosing the message.
to say it's aggravating, is a gross under statement.
it would appear, i need to copy my replies
and paste them to a desktop notepad,
in case the flippin' server is too flippin' busy.
WTF?

there....i feel better now.

backup, backup, backup.

 

[aleatorylamp]

Hello Smilo,
You have mny sympathy. Yes, there are times here at weekends starting on Friday night when that happens. Saturday midday here yesterday had a server-busy time.

Then it´s different from when it normally times out and you get the black page screen with white writing (for which you have to scroll down. That took me quite a while to discover, and prevents having to backup one´s message on the notepad...
If you page back, then the message you were writing is lost.

Maybe I´m mistaken - Do I remember that with a server busy you can go one page back, recover the message you were writing, copy it onto the notepad and send it later? I think I did that yesteday. Or maybe not... my mind is not the best anymore...

Anyway, about the Ad2k tutorials:
I got the 2 tutorials from Html format into Word format, checking that all the coordinate-tables didn´t split up over 2 pages, and turned them into .pdf files: 53 and 59 pages, pics included. Then I went to a photocopying shop to have them printed out and spiral-bound, for all of 2.5 dollars each! It´s much better than reading them them on-screen!

The .pdf files are handy for printing out easily, so if anyone is interested, I could post the .pdf files here.
1- The Ad2k Tutorial 98 with the Ugly100 nice looking tutorial plane with the hand-made Z-Buffer is 745Mb.
2- The Ad2k+RFO Turorial with the Rubbish Flying Object (with no hand-made Z-Buffer) is 887 Mb.

Do let me know!
Cheers,
Aleatorylamp

 

[smilo]

yes, i was able to recover a reply early yesterday,
copy/paste it for later posting.
this morning, no...damn it.
as you are well aware,
i'm a long winded old coot,
so, loosing the reply was very aggravating.

posting the pdf sounds like a good idea.
but, what do you think about making
notated corrections first?

i'm off to outside.
i have a clear day to work on an overheating car.
i've got to grab the opportunities where i can.

 

[aleatorylamp]

AD2k Tutorial #1: RFO

Hello Smilo,
OK, before posting the .pdf of Tutorial 98 with the hand-made Z-Buffer, I´ll try to figure out how to avoid the glitch occurring in the 2 tied central fuselage templates at the beginning of the tutorial.

Then I was trying to make the wings, and right at the beginning, there is also something strange going on with those templates. They are also tied, and the wing won´t cover correctly. I have to work that one out too.

Anyway, for a possible learner, this tutorial would be STEP TWO because of the complicated manual Z-buffer emmulation.

STEP ONE would be the RFO (Rubbish Flying Object), Z-Bufferless tutorial, which is in reality for FS2000 or FS2002, but nevertheless very practical and recommendable. The building process itself is very well explained and also illustrated, even though it is a very simple model, and that´s also why it´s a good idea as STEP ONE. Moreover, there are no glitches!

As soon as I´ve sorted out the problems with the other tutorial, I´ll post the .pdf version.

Good luck with your car´s temperature!

Cheers,
Aleatorylamp

 

[aleatorylamp]

Ad2k Tutorial #2: Tutorial 98

Hello Smilo,
In the tutorial text, I inserted a note after the creation of the centre fuselage templates, mentioning the correct final dimensions and positions of the tied templates, as a double-check method to ensure that they are created properly. This will prevent problems appearing later when the cabin is built.

Then, the problem with the wing templates was that I had inadvertedly placed them in the wrong order, and it took some time to discover what had gone wrong. Further down in the text there was a comment on the importance of ordering them correctly, and that mistakes here can easily happen! So there´s no problem here.

I have just made the right wing right!
Alright! That being right too, here´s the .pdf with the second tutorial for printing out easily!
I distributed the pages so that each of the 6 sections starts on a clean page.

BTW: The already existing tail section with moving rudder and elevator has no bleeds at all!
It is striking, how effective this hand-written Z-Buffer is!
At the moment, it´s just copying from the tutorial text... To master it, is another kettle of fish altogether!

Cheers,
Aleatorylamp

 

[smilo]

thank you for taking the time to upload the tutorials.
but, i've got to say, i'm a little disappointed.
there's no spring binder included.

the overheating vehicle is an on going problem.
it keeps loosing coolant and we can't find the leak.
today, i bypassed the heater core.
we'll see how it goes in a few days.
btw, have i told you i hate auto mechanics?
okay, strongly dislike.
i used to do it professionally.
apparently, the job is not finished
until i bleed on it.

all right, about the cubefly...
i'm tempted to see if i can correct
the wing/fuselage bleed.
way to go, Stephan...thanks a bunch.

am pleased that you're beginning to see
the pluses of the ad2k sequencing process.
in my opinion, it's one of the main features of ad2k.
but, i guess you already knew that. right?

 

[aleatorylamp]

mechanics!

Hello Smilo,
I hate all the mechanics I know except the one that I have currently, until he does something and I will have to find another one.
If they don´t rip you off by causing the next breakdown on purpose because they are dishonest, it is because they are incompetent, but the effect is the same.

I had a small leak that should have been found but wasn´t - the mechanic said it was so small that it had to be an internal leak, but it was the old radiator. The water was dripping onto something hot and evaporated before it could drip on the ground so there was never a puddle, until one day it boiled, but it was raining, so I didn´t notice, and it happened twice, almost ruining the engine.

Had it been another car and not a 1985 diesel BMW, it would have ruined it. So, the mechanic changed the radiator and everything was fine. Further incompetence later caused the dash-board instrument cluster to burn out, and they don´t work anymore, but are too expensive to replace as they don´t exist in the scrapyards or on e-bay Europe anymore.

Then one day he over-tightened the distributor belt when he changed it, and a year or more later, this had eroded the injection pump bearing, offsetting the belt a few micras, and the belt jumped 3 teeth. This caused one cylinder to hit the valves, luckily only denting the shims, but I had to have the injection pump repaired, which I had already been done 4 years before. Had it been another car with slanting valves instead of vertical ones, they would have broken and I could have thrown away the car. This time I changed the mechanic, and it´s a miracle the car still runs!

Well, let´s see if I can finish the second AD2k tutorial this week!
Hopefully the .pdf files will help anyone interested. The anti-bleed sequencing feature is, as you say, the main point,and I´d say the template style semi-automatic building facility would be the second advantage!

Update:
This morning I was half asleep when I wrote this and forgot something:
Thanks for your offer to try to make a sequence for the cubefly!
That would be great, but I don´t want to impose it on you. ONLY if you fancy doing it!

At the moment I´ve done the right Gear for the right wing of Tutorial 98, and I´m doing the sequencing. As alway, this French author´s way of expressing himself is somewhat confusing: He usually explains how to do something before telling you what to do, but in some difficult parts, he explains how to do it AFTER telling you what to do! This makes you go crazy looking for things that don´t exist yet because he hasn´t told you yet!!... But OK, nobody´s perfect...

Almost ALL of the tutorial is VERY well explained, even though some bugs crop up occasionally to keep us on our toes!! E.g. in the numbers, where a short while later in the code listing, it becomes obvious that they are wrong, and one thing like assigning the "wing" texture to the wheels instead of the "pilot" texture that contains the wheel bitmap quarters... and it´s obviously the wrong texture, and it says so a bit later in the code listing... and then, I forget where there´s a minus sign omitted for building the tailplanes, but well, these small errors are easily solved. It must be extremely hard to write such a detailed tutorial like this one...

Second update:
The last part of the section on building the Right Wing is confusing as something is omitted and requires looking into the original tutorial code, so I have included a little paragraph to clarify this. The two number and texture error mentioned above have also been corrected! Here is the newly corrected Tutorial .pdf file! Enjoy!

Anyway, it is rather addictive... so it will probably be quite interesting to use Ad2k for the next project!

Cheers,
Aleatorylamp

 

[smilo]

played with the cube a bit
created a new wing/dice seal
have tried to compile several time.
keep getting error message
"master: assy does not exist.."

am going back top bed.
will try again later.

 

[aleatorylamp]

Hello Smilo,
It´s my fault - I should have known that the new .3DM file would require the compiled tail assembly file "Tail.MD1", and possibly there is a texture missing too. Anyway, here are all the necessary files so that it will compile. At least, I think so! Let me know what happens!
Sorry!
Cheers,
Aleatorylamp

 

[smilo]

actually, i'm not sure if it is your fault.
at this point, i very much want to investigate
and find out exactly why i got the error message.
call it a brain teaser....i'll let you know what i find.

not to worry, i've downloaded the new v1.zip.

in the mean time, don't wait for me.
keep plugging along.
i have several outside chores
to take care of before i can play with ad2k.

good luck and growing skills
in all your endeavorers.

 

[aleatorylamp]

Hello Smilo,
Thanks,it´s going quite well.
Now I have the left wing ready, and also the sequencing.
It compiles but the cabin and wings didn´t display! I called the wrong section! Ha ha... It just shows how one has to pay careful attention!
Now they display, but the textures are distorted... Never a dull moment.
Well... I´ll figure it out!

No hurry for the AD2k Cubefly!
Cheers,
Aleatorylamp

 

[smilo]

am pleased to hear it's going well.
very pleased.
just as a point of clarification,
you're currently working on the pony,
correct?