Conspicuous by Their Absence

[aleatorylamp]

Hi Ivan,

OK, good tip.
I just renamed the .xaf textures to .bmp directly for a first try, and I checked them out in the dark: Ha ha! They do glow... it´s a Halloween plane...

However, I can´t really use the PCX files and rename those, because I needed .bmp textures in order to darken them uniformly, which the PCX ones wouldn´t allow, as they would alter the colours completely in order to fit the changes to their palette.

Wouldn´t that glowing also happen if I used the PCX files to change those to .bmp´s and then darkened them?

Cheers,
Aleatorylamp

[Ivan]

Hello Aleatorylamp,

You will actually have the same problem with your BMP as with the PCX files.
They all need to run from a fixed palette although you can alter the colours within the palette.

I believe Hubbabubba uses Saint Paint to alter his palettes. I use GIMP.

Regarding naming: I don't think that other than being 8.3 it matters what the file names are as long as they are R8 or BMP format.
Once I had a working solution, I didn't bother experimenting much.

- Ivan.

[Ivan]

The service ceiling seems to have gotten considerably higher at 36,490 feet.
Some of the behaviour didn't seem quite right on the climb though; Airspeed seemed too high.

- Ivan.

[aleatorylamp]

Hi Ivan,
Renaming went OK, so what names to give the bitmaps is no issue - it´s just the glow.

A sinister ghost Giant! - it was a really funny sight when I parked it on the Heliport in N.Y. at night to test for night textures! (I thought of testing the FS98 Sceneries in CFS1 as well - I was surprised they all work!!)

Well, the limited fixed pallette was actually why I thought of using another texture format. Perhaps I´ll have to find another work-around. Does the glow happen with all textures with ".bmp" extensions? Perhaps it would need bitmaps with more than 256 colours, one step up from that?

As regards performance issues on the second Giant, I was also getting ceiling issues apart from excessive climbing power after altering the torque and friction graphs, so I´ll conduct some more tests with the old friction parameters and not alter the torque so much, and fiddle around some more with the propeller tables before bothering you again, although I will accept your kind offer for testing if I arrive at another dead-end. Thanks!

Cheers,
Aleatorylamp

[aleatorylamp]

Hi Ivan Hi all!
After lowering the exorbitant friction and torque parameters again, everything has started falling into place, and the Maybach engines are performing very close to specification. Compared to the Mercedes ones, as expected, they give +3 mph, +30 RoC and +2000 ft ceiling, and also have increased range (2 extra inboard tanks).

I have also been able to correct the textures glowing in the dark - copying the darkened BMP´s back onto the PCX´s, this time, the limited palette has accepted the darker colours with hardly any changes, so the manouever has been very successful indeed!
After some minor adjustments on the engines, I´ll be uploading the machine.

Thanks again to all for such fine support and help!
Cheers,
Aleatorylamp

[aleatorylamp]

Hi all,
I´ve uploaded replacement model files to fix the failed animated propeller blur on the Porokhovschikov Twin-tail. The blur appeared when the engine was not running. The model file contains a re-SCASMed virtual cockpit.

This is the link: http://www.sim-outhouse.com/sohforum...id=19&id=19572

Sorry about the inconvenience!
Cheers,
Aleatorylamp

[Ivan]

In trying to refine a flight model to my own satisfaction and edit the visual models because of very slight inaccuracies, one has to wonder which is better?

To build a very very small number (perhaps only one) really superb aeroplane or
To build a few very good aeroplanes or
To build many more pretty good aeroplanes that are each much better than what is currently available?
Most of the audience cannot tell the difference in any case.

Opinions?

- Ivan.

[aleatorylamp]

Hi Ivan,
As many good and interesting aeroplanes as possible!! Ha ha!

The best of both worlds! Well, seriously now, as the question is about an opinion, and I don´t want to impose my opinion on anyone, here goes my opinion:

I suppose the crux of the matter is really one´s own enjoyment while working on or re-working a given model to get it right, and then to share the model. Often that depends on the model itself, how accurately one can get it within the limitations of display conditioins and/or Flight Dynamics. In my case, I always include as many animated parts as I can, which may cause more imperfections than without them, and with experience and suggestions from others, this continually gets better. There is a good saying in Spain: "He who works, has the right to make mistakes.". Maybe I should learn FSDS, Gmax, or AD2k2... but I like AF99... but that could be an interesting discussion topic for another day!

Mistakes hopefully cause feedback, thus making room for improvement. Without the initial mistake, there would be no improvement, and most probably there would have been be no initial product. So, how far is fear of making mistakes or fear of inaccuracy any use? Perhaps it only acts as a deterrant to upload an otherwise nice model.

Sometimes the question arises, whether unavoidable compromises render a model unfit for uploading or not... Very subjective! It depends on one´s own appreciation of how forgiving or unforgiving the "audience" could be, and I include bugs here too. I imagine most people realize that nobody will consciously and carelessly upload a defective model, so I like to think they understand a modeller who has put his foot in it. A bug fix may easily puts it right, and if there is no fix, maybe it is better to have the model with bug than to have no model...

I am not justifying mediocrity, which a totally different thing: Mediocrity is only the bad result of an "I don´t care" attitude, cutting corners and sparing efforts.

So, the ideal situation could be: All in good fun, with a margin of tolerance for mistakes and bug fixes, and a huge number of fantastic aeroplanes!

Cheers,
Aleatorylamp

[Ivan]

Hello Aleatorylamp,

The topic I was bringing up wasn't really abut mistakes. The silly thing is that even when folks actually see a mistake, seldom does anyone comment.
In fact, the amount of feedback I have received over the years from people that I did not already know has been pretty minimal. The total number is probably around 5 or 6 and that includes you.

I was thinking more of doing a LOT of research that is not even visible in the completed model.
The Messerschmitt 109E I recently rebuilt is a pretty good example.
The overall dimensions were pretty far off in the original AFX and yet I don't believe anyone noticed.
I fixed what I thought were the biggest issues, but didn't feel like fixing everything I saw.

The P-40E rebuild is another case. The corrections I recently did were quite extensive and yet the overall appearance did not significantly change.

I rebuilt the P-40E because I wanted the most accurate model I knew how to build and am still not done yet.
My tendency is to fix everything I know how to fix regardless of how long it takes.
This experimental Ki-61 has been sitting in the workshop and paint booth since around 2006 and still hasn't seen its first release.

What do others think about "Good Enough"? I have seen others push out a new project about every two weeks.
The quality wasn't really there, but each project was an aeroplane that didn't exist before and they were generally comparable to the stock aeroplanes.

- Ivan.

[aleatorylamp]

Hi Ivan,
I understand perfectly that you weren´t referring to mistakes, it was only my own idea added to the concept of accuracy and inaccuracies, and as an answer to your question about extended research and consequent higher accuracy, I had commented "that depends on the model itself, how accurately one can get it within the limitations of display conditioins and/or Flight Dynamics.".

With regards to feedback not being very frequent, I put it down to shyness perhaps. Thank heavens there is a very small percentage of people who do, so this is what helps us develop.

In general, I love fixing and repairing things myself, as it gives physical items a longer life. With FS models, if the source files are reasonably accurate, a repair job can be quite enjoyable as it saves the initial basic work. As regards accuracy here, I go for a general degree of accuracy myself - mostly 100% is impossible because of contradictory information and diagrams on the old models I tend to go for. Here often old photos are more accurate than someone´s drawings, and it´s a nice mental 3D exercise to "get it as right as possible" from this kind of material!

When the information is available though, it is nice to be exact, for one´s own pleasure more than anything else. If it does get too tedious, however, there is a limit to my pacience/frustration, and then I go for a practical degree of accuracy - the visual model must look correct, and the flight model must behave within reasonable limits.

My concept of "good enough" probably differs from others´ "good enough". Personally, I would say that Stock Models were good enough for 1999, but not now. Upto last year, I normally used to take 3 or 4 months to produce an FS98 model, but now it takes anything from 3-4 weeks just to upgrade the visual model for CFS1 and even longer to make a CFS1 .air file. When that is done, then for me it is "good enough". I would say the minimum for me, for a decent upload is my self-criticsism criteria of 8.5 out of 10 on difficult models, that after a lot of work, I cannot improve any further.

What I hate is something that has obviously been slapped together by "pushing" parts from other models into position, and only generally fixing dimensions, and coming out with a new model, where many shapes are noticeably incorrect on the visual model. I can accept adaptations when the result looks like it´s supposed to, even though the exact measurements, if one were to investigate the source files, were not totally exact.

Well, now I have to help in the kitchen...
More later,
Cheers,
Aleatorylamp

[aleatorylamp]

Hi all!
After checking everything I hope it´s all ship-shape and the upload is OK.
Here´s the upload link to the second Zeppelin Staaken R.VI Giant:

http://www.sim-outhouse.com/sohforum...id=19&id=19601

This is one of 3 night-bomber version license built by Albatros Flugzeugwerke. The darkened textures were a bit too gloomy so they are now a little lighter than I had them a few days back, but they are darker than the ones on the previous Schütte-Lanz built model.

Because of the central tailfin/rudder on this model, as well as the horn-balanced ailerons, this time there is no mechanic sitting in one of the engine nacelles, but the propeller-blurs look nicer (similar to Ivan´s!) and the machine-guns are better shaped, the front one being more accurate than the rear one. The shape was just outside the 150% parts count so I had to simplify it a tiny bit!

The model is at 149.5% parts with the virtual cockpit inserted separately using SCASM, view corrected, so basically it´s a 151.4 % parts model. The more powerful Maybach engines noticeably provide more power at altitude than the mercedes engines, although they were slightly slower lower down. They had "oversized" cylinders as well as quite a lot more compression.
Interesting, how technology was developing at that time!
The Checklists give detailed flight plans for the different bomb loads, as well as instructions for flying.

I hope you all enjoy the model. Tell me if you find any glitches of bugs.

Cheers,
Aleatorylamp

[aleatorylamp]

Hi all,
I did forget something after all. There´s always one thing, isn´t there?.. Nothing is perfect in this life, and modellers even less so!
In the Dp files, the machine-gun (and cannons here, to differentiate the forward firing one from the rear) firing speed should be at 0.08 and not 0.13 and 0.25.
Ivan had already pointed it out, but I hadn´t understood it correctly as I mistakenly took the 745 m/s bullet-speed for the firing-speed of about 700 rounds/min... I never noticed the help text at the bottom of the screen... Live and learn. Ever so sorry! This applies to the first uploaded Giant too.
Perhaps simmers will be able to correct the entry in the Dp files on their aircraft. Do forgive me!
Cheers,
Aleatorylamp

[aleatorylamp]

Hi again, everybody!

I was wondering if there´s any interest for a model of the following invention, something along the lines of A.A. Porokhovschikov´s sesquiplane, I suppose, but even more quaint, to put it mildly:

Qualifying as the smallest wartime man-carrying aircraft, we have the Pensuti Triplane with its wingspan of only 13 ft, and an all-up weight of 507 lb including pilot and 2 USG fuel for an hour´s flight. A few saw active service until 1918, in short-range reconnaisance missions for ground-troops. Powered by a 3-cyl, 35 hp Anzani radial, it flew at 59 mph and took off in only 65 ft!

It was built by Caproni in 1918 and designed by their chief test pilot Emilio Pensuti, who unfortunately died heroically before he could see the results of his invention. The purpose behind his design was to do in the air "what bicycle does for the man on the road". It was categorised post-World War I as a small sporting aeroplane, and was reputedly easy to fly.

Would anyone like a CFS1 AF99 model of it?
I could even try my hand at doing a tutorial on a separate thread if there´s any interest...

Cheers,
Aleatorylamp

[aleatorylamp]

Hi all!

I was just re-reading recent posts between myself and Ivan regarding modelling quality.

Just to make sure that I haven´t offended anyone with my comment on Stock Models having been good enough for 1999 but not for nowadays, I would like to add that my reasoning behind this comment exclusively relates to the processing speed of computers nowadays, that allows (even with AF99) more complicated models than was possible at that time.

Frame rates don´t go down anymore on a 150% parts count as happened on slower computers back then, and this often required models to be shaped with simpler cross-sections and to have less detail.

It was by no means a reflection on the quality of the modelling work, which is by any standards clean and excellent. Anyway, there are also some aspects on Stock Models like the propeller blur, that are better than what can be achieved with AF99 even on modern computers.

I thought I´d clarify this just in case.
Cheers,
Aleatorylamp

[smilo]

it's interesting that you're concerned
about derogatory comments made
about the original cfs models.
i don't think i would be overstepping
if i said, everyone agrees with you.

i have even taken to replacing the originals
with models that i think are of better quality.

[aleatorylamp]

Hi Smilo,
Thanks for your comment! In the early times of CFS1 hardware technology limited modelling quite a lot, it seems, so what was achieved was remarkable. Nowadays we have to deal with software limitations, and it is also quite remarkable, that CFS1 can still be used!
So we can continue enjoying ourselves!
Cheers,
Aleatorylamp

[smilo]

too true.
for me, one of the main issues with cfs models
are the excessive bleeds that "most" have,
some, much more than others.
as i see it, alleviating the bleeds is the real challenge
to building a cfs model.

i should qualify my reply. i'm no expert.
i've only built one model
and have been out of the sim for a while now.

i have the utmost respect for those who use af99.
i tried to use it, but, did not find it user friendly.
matter of fact, i think it just didn't like me.
i started playing around with aircraft designer 2000
and found we got along much better.
as much as i tried, i was unable to hit the parts limit
and was amazed at how much detail could be added.

after finishing my first project,
i started another, but, unfortunately,
got distracted by other things and there it sits.
i may pick it up someday, but for now, have no plan.

[aleatorylamp]

Hi Smilo,

I also tried using AD2k2, and was definitely impressed with its potential for quality and detail. I also liked its similitude to AF99, and of course the low level of bleedthrough problems. However, I found it more time-consuming, because of the seemingly endless room for detail, and I never knew when to stop!
Hence, I paradoxically had the sensation it was a little more difficult than AF99, which, given its limitations, requires a more "minimalistic" approach, thanks to which I find I can finish a model within a more satisfying time limit.

Nevertheless, I still have AD2k2 on my "to do" list. At the moment I´m having serious bleedthgough problems putting in the new CFS1 crew and guns on the Austro-Gotha G.IV Grossflugzeug. I have already eliminated the hollowed out gunner-wells, pilot´s cockpit and connecting corridor, and also the transparent windows, but I´m still fighting with bleedthrough, because I want to at least show the hollow parts with dark surfaces, and it´s not working yet. I even managed a reasonably good .air file for the 230 Hp Hiero-6 engine. This could well be a convenient candidate for an AD2k2 model. I wonder... Also, the pusher props grouped in the Tail left/right with ailerons and wing-trailing edges, give bleedthrough problems with the wings.

Here are some bleedthrough-free shots just for some eye-candy.

Cheers,
Aleatorylamp

[Ivan]

Hello Aleatorylamp, Smilo, et al.

I personally have been more guilty than most in my derogatory comments about the stock aircraft that came with Combat Flight Simulator.
First of all, my belief was that CFS was a hasty assembly that probably attracted much more attention than the designers ever had expected.
My guess is that this was just an introductory game, a market test, that was put together very quickly with not much investment.
It probably surprised everyone with its longevity and dedicated audience.

For me personally, it was the first game that combined everything I was looking for in a flight simulator. The graphics was "Good Enough" for a reasonable "suspension of disbelief" which I thought was a bit lacking in Flight Simulator 98.... AND it had the combat element in it!
I believe it compared well with its contemporaries.
Its basic engine apparently is sufficient for fairly detailed graphics with the models we build today AND for reasonable flight performance simulations which I am still experimenting with well over a decade after I encountered the game.

Aleatorylamp,
I can only partially agree with your assessment of the stock aircraft:
The low polygon count and poor shapes were pretty much the state of the art back in Y2K and pretty fitting with the "Minimum" recommended Pentium 166 MHz Processor, but it would have hurt nothing to have more realistically shaped aircraft.

My current Development Computer dates from that time also (1997). It started life as a 150 MHz Cyrix (Pentium 200+ Equivalent) with a blazing fast Voodoo 3 3000 Graphics Board.
This same computer still has no trouble today with the most complicated Aircraft Factory 99 project that either of us has built. It slows a bit when several aircraft are displayed concurrently and chokes pretty badly with some AD2000 projects.

For the most part, the external models of the stock aircraft don't show anything we can't do as well or better using Aircraft Factory 99 and Aircraft Animator.
Even their propellers can be reproduced in a fairly satisfactory fashion. I choose not to use that style of prop but have seen other AF99 aircraft that do.
The parts that we can NOT do with AF99 and AA are the Virtual Cockpits and especially the animated gauges in the virtual cockpit's panels.

Smilo,
I agree that AF99 is not a great tool, but it happens to be one that I have worked with long enough to know how to do things. Sooner or later, I hope to get a project into AD2000 to see how it can be improved.
I can see how bleeds are incredibly distracting, but from a designer's point of view, I would much rather inherit a project with good shapes and bleeds than one that had no bleeds but not a very good shape
I am sure you have seen the AD2000 projects that produce a nice bleed free aeroplane but with shapes that are not so good.

My daughter insists on a smiley for this message

- Ivan.

[aleatorylamp]

Hi Ivan, Hi Smilo,

Yes, it is interesting to see how some software products like CFS1 endure the wear of time more than others, and retain their attractiveness, and their shortcomings are more than balanced out by their virtues. Only a year after FS98, the improvements supplied are quite striking. I remember reading that the developers were testing the market for some of the .air file and scenery improvements they were planning for a later edition, FS2000 and/or 2002, and they decided to put them in for their combat-game FS version.

I suppose in 1999 most people were using Pentium 150 or 166MMX computers with 2 or 4 Mb Graphic cards, so the stock models must have been geared towards this, much to the frustration of the lucky ones with 16 or 32Mb, or even 64Mb accellerator cards. But that´s where the world opened for aircraft developers!

In my case I would say that despite its shortcomings, AF99 is a great programme after one knows these and how to squeeze the maximum out of it. What I like about it is the logical way it works for building a plane. It also allows one to concentrate on the field one chooses - either moving parts in my case or more exact detail like in Ivan´s case.

For my brain geometry, AF99 and even AD2000/2k2, seem to be more practical than other building programmes where I get lost with the endless possibilities, even though the advantages that AD2000/2k2 offers over AF99 also require more building time, that I haven´t got yet.

Must rush off to buy some biological food at the farmer´s market - I can´t eat other stuff because of migraines... Not one body is perfect!

I think I´ll build the Pensuti Triplane to cheer me up while I battle the Gotha Bleedthrough!

Edited update: I finally managed to fix the AF99 texturing problem caused by the punctuation-mark conflict between AF99 and WinXP: Other than the WinXP region/language change option with its consequent keyboard-entry confusion and the fact that it kept reverting back to default, I discovered that puntuation options can be edited within an individual language/region, so now AF99 texturing works perfectly! It´s nice to have one more thing solved.

Cheers,
Aleatorylamp

[Ivan]

Hello Aleatorylamp,

I find again that I must respectfully disagree with you on the quality of Aircraft Factory 99 as a tool.
I believe it has many limitations that are the result of a very basic fault in design.

To illustrate, I will make the assumption that much of the audience for this post has had at least a minor familiarity with AF99.
The way AF99 assembly sequences work is pretty much a straight line:
The designer first adds a piece to the assembly.
Each additional piece is then either glued to the wad of pieces already in place OR
is added without Glue which then creates a new sub assembly to which pieces may also be glued.
The BIG problem is that with all the subassemblies, There is no way to control the order or direction they will be put together.
The result may be predictable, but is hardly controllable.

If it were possible to specify how these subassemblies fit together, many problems would be solved.

The AF99 assembly sequence may be considered a computer language.
Unfortunately, this language executes in a straight line with no hierarchy.
This is the same as a computer language that has no branches or subroutine calls.
The end result is that although the language can be used for some tasks, there are many things it simply cannot do.

The recent P-38 Lightning is a great example of why this is needed:
The inner wing section needed to be glued to the nacelle in a well determined fashion.
The inner wing section itself needed its own sequence to fit together flaps, fillets and the wing section itself.
In a real language such as SCASM, this is trivial to do.
In AF99, we cannot do it without special exceptional rules (Templates) which unfortunately had a bug I could not work around. In other words, the inner wing needs to be its own subroutine.

Another obvious issue is the Component versus Structure item count.
There is almost nothing that a Structure can do that a Component cannot do better.

Even on your aeroplanes from the Great War, the biggest reason for the high resource usage is because of the need to use many Structures for lack of Components.
From a SCASM coding perspective, a Component is typically no more complicated and in many cases LESS complicated than a Component.
Two serious limitations with wing Structures are the inability to taper them or build in dihedral.

This argument may sound strange coming from me because I have spent a lot of years trying to prove that nice looking aeroplanes CAN be built with AF99.
Consider this to be a case of a man trying to demonstrate that a Hatchet can be used to turn screws, drive nails and spread plaster even if it is not the best tool for the job.

Good Evening.

- Ivan.

[aleatorylamp]

Hi Ivan,

Ha ha! I agree with you! I meant something else which I´ll explain further down...

I realize AF99 often gets very frustrating, but why are we using it then? Good question. A potentially good tool was messed up from the start (AF5) with an inconvenient programming style, containing simplifications most probably to speed things for the slow computers of the time. Later, these simplifications were not corrected when AF5 was upgraded to AF99 by the time computers had got a little faster.

Later still, unfortunately, instead of updating and correcting AF99, it was discarded altogether and FSDS1 came out for 55$, I think. That was about the same time as I bought AF99 for 25$. Do I like AF99 because I paid for it? Why do I insist on using it, if AD2k2 is better? I suppose I have got used to it and it´s difficult to tear away from a familiar tool even if it is far from perfect. A voice in my head keeps telling "use AD2k2!!, use AD2k2!!"

When I said AF99 building style was logical, I really only meant the X-Y-S viewing axes and how it is quite straight forward for the builder to place a piece in its position, as opposed to the more complicated way in which more modern 3D programmes extrude shapes from a wire mesh.

I should have clarified what I meant, but I couldn´t edit the post anymore when I realized my comment was ambiguous and could refer to how AF99 puts things together, and that is often NOT logical! Its limited grouping possibilities and incongruent display sequence DON´T logically always "show what´s furthest last".

Moreover, what you say about components and structures is too true! If the AI contained in the programme were to put together structures like we put together components, things would be much better.

Well, now what...? After this, I can either continue using AF99 or switch over to AD2k2 - it´s a free world and we are only limited by our own fears or laziness, or even stubborness I suppose. It´s easier to say..."when in doubt, don´t do anything!!", than "when in doubt, throw it out!", or even "mit dem Kopf durch die Wand" (German for "through the wall with your head").

Have a nice Sunday!
Cheers,
Aleatorylamp

[aleatorylamp]

Hi Ivan, Hi Smilo,

I have just noticed that for CFS1, it´s not AD2k2 but AD2000! When I built the Tutorial Rubbish Flying Object with AD2k2 it was so long ago, I´d forgotten that I´d built it for FS2002. I also started the Gotha G.IV with it, but chickened out when things got tedious, and went back to my usual AF99...

Well, out of curiosity, I´ve just downloaded AD2000 and I think I´ll do the RFO tutorial again, just to see how it goes - no harm in learning something new. It looks deceptively similar to AF99 but in reality it´s very different. I must get my old brain oiled again - that will be good exercise against Alzeimer.

Nevertheles, I´m also making good progress with the AF99 Gotha G.IV bleedthrough battle wherever possible, and I will definitely not waste all my work there, so hopefully I will be able to upload it soon!

Edited update: I looked into the AD2000 RFO tutorial to see how things work, and it is rather a lot more than I can chew at the moment, I´m afraid, so it will have to wait until the summer holidays when I have more time on my hands. It looks very powerful though: It it lets you attack the display priorities directly by programming routines and subroutines into the model itself. Quite amazing! Over the years I´d completely forgotten how it was.

Cheers,
Aleatorylamp

[Ivan]

Hello Aleatorylamp,

Yes, we knew about the AD2000 versus AD2002. AD2002 was free for a very long time and I tried it out but was unable to build working models for CFS because of the different way that viewing planes were handled.

Smilo for certain knows about the difference because he has already built something (Arado 196) with AD2000 and for other not so obvious reasons.

I downloaded AD2000 a while back but don't remember where it is now.

By the way, I am finding that my free time is pretty limited these days.
Anna Honey just got back from Paris almost two weeks ago and now she is on the big jet out to China....
Sheesh!

- Ivan.

[smilo]

i think it would be best for all concerned
if i withold any ad2k comments at this time.
i'm currently in the western montana woods
on an extended road trip vacation.
any comments would be from distant memory.
(a very bad idea)
it's been a long time since i've used the program
and it would not be a good thing
if i went on about an ad2k feature,
only to find out, my memory was playing tricks.