Conspicuous by Their Absence

[aleatorylamp]

Hello Ivan,
Yes, my e-mail was a bit long I´m afraid. I´d found what I thought was an answer to your
comment: "For some reason, it was an amazingly effective fighter in Soviet service though
I still can't quite understand why.", as well as on the manoueverability, so I thought it would
be useful. Anyway, if I find any data regarding empty weight CoG position, I´ll let you know!
The best of luck for your Hardware Project!
Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

Actually I already have found some data on the CoG location in terms of MAC at various load conditions.
The problem is putting all that data together with the station diagram and Paul Matt's drawings to figure out the exact location on the 3D model. The CoG problem was apparently recognized fairly early on and later models had slightly rearranged internal equipment to address the issue, so it is also necessary to make sure that all the data is consistent and not describing different models or if they are different models the difference must be accounted for.

The central engine location might improve responsiveness, but would not really do much for turning circles. Perhaps it made for a better angles fighter and suited the Soviet style of air combat?

Regarding Soviet service: The Airacobra was amazingly effective in terms of successes for the number of aircraft involved.
I do not believe the Soviets actually bought ANY Airacobras or King Cobras at all. I believe they were supplied under Lend-Lease which did not require actual purchase.
What is interesting is that by the numbers, the Airacobra was at best only equal to the Soviet Yakovlev and Lavochkin fighters but did extremely well in that environment. Perhaps the difference was better and more consistent manufacturing quality?

- Ivan.

 

[aleatorylamp]

Hello Ivan,
Sorry to answer a bit late! I´ve been a bit busy because the phone company has duplicated internet speed free of charge, and I´ve been improving the domestic network cables and getting advice for better router settings so as to benefit from the upgrade.

Anyway, back to the Airacobra´s success with the Russians:
Its high responsiveness would make for a fast change in rotation, increasing manoueverability in that moment, but as you say, I would expect that once the aircraft was already turning, other factors would come into action. The tighter a turn, the worse any imbalance due bad weight distribution would be. How to put that into an .air file is another matter!

I´d agree that manufacturing quality must also have been important, compensating any design trade-offs with more reliability, perhaps also allowing more prolongued, more extreme situations. I don´t know... Russian pilots perhaps enjoyed the more reliable technology, but I suppose one can only speculate.

Have a good weekend!
Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

I have also been a bit busy trying to get my Game Computer running again after a HDD failure.
The computer was running Windows 2000 Professional (Service Pack unknown) without any problems (software wise) until the drive failed completely. My attempt at a backup before the failure did not work and it does not appear to spin up an more.
I now have Windows 2000 Professional (SP4) installed but am unable to get the OS to install the MS Sidewinder software or drivers.

W2000 complains about a hardware incompatibility. This is VERY strange because I installed W98 SE on a 850 MB drive just to test and be cable to pull some pieces off for my Development Computer. With some effort at finding drivers (which all seem to behave differently when loaded from fies versus loaded from a CD, I managed to get Combat Flight Simulator up and running with sound through the sound board AND the MS Sidewinder Joystick working.

From this, I know there is no hardware problem with this combination because all I am swapping out is the Hard Disk Drive with the new Operating System.

I did find the other Airacobra I had seen before on CFS and a screenshot is attached. It is a straight side view because I don't have a joystick that I can use to alter the view.

- Ivan.

 

[Ivan]

Jpystick Working Again

As can be seen from the screenshot, the Joystick is working again.

As it turns out, the issue was not Operating System or Hardware related; It was Driver related.
I needed a later version of the MS SideWinder Drivers. There is a bit more to this story, but not worth the time to type.

Regarding this Airacobra:
The Flight Model is FS98 as can be seen from the Engine Starting sequence.
The shapes do not seem to be quite as good as the one by Eric Johnson and detail is a bit lacking.

- Ivan.

 

[aleatorylamp]

HDDs nightmare and Alpha Transparency 179 Cockpit

Hello Ivan,
Oh what fun! Hardware Problems!

I´ve run into serious problems too, with the new duplicated Internet speed, but have managed to avoid WinXP HDD reformatting. WinXP apparently caters for only half the newly available bandwidth, but not knowing this, I bought a separate, new fast Ethernet Card, which has exactly the same performance!

Curiously, the software supplied didn´t work for WinXP, so I installed it on the Win7 HDD I have in dual Boot. Everything worked fine, and I got FULL Bandwidth! AND Also with the normal LAN connector!!

BUT: The new card FORCED an Ethernet Boot-Up (PXE) installation, making my DATA HDD invisible. It altered the BIOS/CMOS, the Smart Alec, and WinXP wouldn´t work anymore.
Re-setting the BIOS and then blocking the Network Boot was ignored.
Taking out the new Ethernet card was also no good: The entire computer wouldn´t boot.
It was also impossible eliminate the PXE Boot written into the WINXP HDD !!

I got round it by disconnecting the WIN7 HDD, luckily being allowed to repair the WINXP installation. Then, disconnecting the WINXP HDD, I had to install WIN7 anew - no repair allowed this time!
Now there´s no dual boot, but I can select which HDD to boot from in the BIOS/CMOS, and it remembers it until I change it again. ...and I got a refund on the new card!

OK: Airacobra:
Interesting, the CFS model. The stance is good on the on the metallized model, and the transparent cockpit is an asset, but I agree with you that EJ´s one is better, especially now.
As you have the AFX for yours, perhaps a transparent canopy could be easily and artfully done as a top-only structure with Alpha Transparency Speed Below 179 setting and a frame component. Not being pushy - just an idea!

Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

My sympathies for your hardware issues as well. Mine is still not quite resolved but resolved enough that I can continue reconstruction of the Game Machine. The problem is that it isn't just a "Game" Machine. Anna Honey also uses it for the Scanner and it also has an attached Printer and I need to make the rest of the stuff work.
In addition, CFS and Joystick work, but the utilities I used for graphics and screenshots are also gone as are the AIR file editors, DP editors, Text editors, Compilers..... The list is quite long.

It seems odd to me that a network card can cause this level of issues. I tend to stay with the big names when I can and most of the machines I have in the house are not actually networked at all except via SneakerNet.

Regarding Airacobra:
This last model is really a FS98 aeroplane as can be seen by the lack of a DP, the FS98 Flight Model and the Cessna style Panel. I just used it as a test example because it was there and I thought that more screenshots of Eric Johnson's P-39D would not be interesting. This model actually has a lot of bleeds but I just took screenshots from angles that did not show them.

Regarding EJ's P-39D:
I actually have no intention of doing any more significant edits to the model except perhaps to attempt to take out a few bleeds. Although it looks good, it really has some serious shape problems that need a full rebuild to correct.
A new build of my own is easier than doing a major rebuild of someone else's model and I would actually own the result.

The shape problems that I can see are the following:
The Fin / Rudder are too high.
This can be seen in a comparison with the Paul Matt drawings.

The Belly / Underside is too shallow.
This can be seen in comparison with Paul Matt's drawing..... AND in a quick comparison against the photograph of the USAF Museum P-39Q.
Note that the Wheels and top of the Canopy are well aligned with Paul Matt's drawing.
Note that the Main Landing Gear appears longer in the CFS screenshot than in the photograph.
The Wheels are in the same location but the underside is much too high and the difference is made up by a longer Landing Gear.

The Wing Tips are not shaped correctly when seen from the front.

There are a bunch of other issues, but the biggest issue is probably that it has an 8 sided cross section rather than the typical 12 sided cross section I try to use.

- Ivan.

 

[aleatorylamp]

Hello Ivan,
I understand your point on the Airacobra model that you reworked, and its limitations.
At any rate, it has been a fruitful excercise, so a new build will be even better!

Hardware problems:
A lot of interesting Flight Simulator equipment sadly stopped working with WinXP. Serial ports got scarce and Win 95/98 Drivers were´t updated. My FFB Sidewinder Joystick collects dust on a shelf and there´s no space here for the old Win98 Pentium 266 MMX.

As in your case, I need a 32bit OS, to run AF99 and other CFS1 development software. Funnily enough, there´s also a back-light scanner here, which won´t work on a 64bit OS. WinXP is the newest usable OS for my needs, and I bought a new motherboard that still used WinXP Drivers only just in time!

But, WinXP´s capacity is only 3 Gb RAM and 1/2 speed Fast LAN (the latter I just found out...), and this motherboard can handle more, so I have a second HDD on it with Win7-64bit on it, for Starcraft II, although I´m terrible at it, and for the full-speed Internet. Until yesterday the two HDD´s were on Dual Boot, not they are BIOS selectable.

I don´t have a network here either, just a router to which the family´s computers are connected to. We could configure a home network if we wanted to, but we don´t.

I thought it odd to, the problem I had. Why some Ethernet cards come with driver installers that automatically force a PXE Network Boot on you is beyond me.

It worked fine with Win7. Although there was no network, it was intelligent enough to discover that, but it completely messed up the Dual Boot for WinXP: It spent 10 minutes looking for an inexistent net-source to boot from before booting up, and on top of that, the Data HDD disk disappeared because the BIOS had replaced it with the Net-Boot!

A load of codswallop! ...and it turns out I didn´t need the card anyway. The on-board LAN is the same speed, so I returned the card to the shop. Live and learn!

Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

Perhaps I can help you out....
Maybe the details I was too lazy to type up might be useful to you after all.
With my new Windows 2000 Professional installation, I was trying to install drivers for a SideWinder Precision Pro.

Here is what I believe to be the case:
You know of course that you will need a Sound Card with a MIDI Port.
Not a Game Port, but a MIDI Port.
That is why the USB to Game Port converters will not work with the SideWinder joysticks.
The Game Port is an Analog device that can support ANALOG joysticks.
If you look at the pin assignments to the DB15 connector, you will find that there are not enough pins to support even the Hat switch on top of the DIGITAL SideWinder joystick much less the set of Base Buttons.
It certainly can't support the Shifted Base Buttons that the SideWinder Precision Pro can provide.

I can't remember what I used originally to get my SideWinder joystick working with Windows 2000, but I am guessing that it was whatever version of SideWinder Software came with the SideWinder Force Feedback joystick that I have gathering dust.
The markings on the CDs are not obviously different so I never noted it when I installed it the first time.
THIS time, I was attempting to use the SideWinder Software that came with the SideWinder Precision Pro joystick which worked with Windows 98, but not with Windows 2000. I believe it is only version 2.0.
I eventually found a version of Software distributed with the SideWinder Force Feedback 2 joystick.
It makes no mention of Windows 2000, but claims compatibility with Windows XP. This is version 4.0 of SideWinder Software.

Perhaps you should try that version as well. It is not hard to find if you are looking for it specifically.
I just did not know to look for it when I ran into problems.

Interestingly enough, I went to the local Microsoft Store to try to get advice from their technicians and to see if I could get an ISO of a Windows 2000 Professional CD that would work with the license key on the sticker on the side of my computer.
No luck there. They simply don't have it.
One of their technicians actually tried to convince me that a Hard Drive replacement was enough of a hardware change to make the old software not work! Hmmm....

I realised pretty quickly that this was not the person I should be talking to.

- Ivan.

 

[Ivan]

Lend-Lease Airacobra

A rather amusing thought occurred to me last night but I didn't have a chance to get any screenshots to post.

So far, just about every P-39 built for CFS has been done with Eric Johnson's AFX as a starting point.
Although I have also been altering the model in a few places, I have not made any maor changes to the textured pieces except to rescale everything to a slightly larger size and that was done along with a slight remapping so that the original textures could still be used. Other than moving a few stars around and messing with the nose markings, the original textures are unaltered.

What occurred to me last night was that other textures that were done for this MDL should also map properly.
Here is the result of my paint crew doing a quick copy job from another aeroplane:

- Ivan.

 

[aleatorylamp]

Hello Ivan,
Now you have to give your painters a summer bonus!

So the EJ AFX is the great grandfather of most Airacobra´s
around! Interesting. Eric Johnson´s original machine certainly
looks like a convenient developer´s advanced practice platform,
and it has whetted my appetite, to say the least.

My search for these original EJ AFX has proven futile so far, and I
was wondering if you would be willing to send them to my revived
Apprentice Training Department, for my technicians to do some
practising on.

Incidentally, they´ve had to give up on theQbasic thing, because
when Qbasic writes a line to an output file it adds its own pair of
inverted commas at the beginning and at the end of the line, and
there´s no way of preventing this. Thus, they´ll have to learn Excell
and Batch files - if the Worker´s Union allows it, of course.

Re: Hardware: So it´s a MIDI port! Interesting, thanks!
I know the old game port is useless, and all of the old sound cards
I´ve collected over the years only have the usual 15-pin serial port.
Apart from that, the motherboard will obviously not recognize them,
for want of WinXP or newer drivers.

I´ll see if I can track down one with a MIDI port and take it from there.

At the moment I´m using a FFB Game-Pad. Spacewise, it´s very practical,
but no the old FFB Sidewinder gave totally different sensations.

Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

It appears that the Eric Johnson AFX may have to wait a while. Last evening, my Development Machine's Graphics card finally blew up. There was not just a distorted image as before, but no image at all on the screen. A full power off did not make a difference and later I noticed a faint smell of smoke in the room.
A bit after Midnight, I tried to power up the computer again. I initially got a Voodoo 3 3000 BIOS display as usual, but in about 5 seconds that changed to flickering character and multicoloured blocks on the screen before it all went black a few seconds after that.
Where are you finding your PCI graphics boards? I believe I will need to look for one to get this machine working again.

The Paint Crew does not get a bonus. It wasn't that difficult and is just part of the job.
There is plenty of room for improvement of that set of textures anyway.

Regarding QBasic:
The language can't be that primitive. There has to be more to the story.
Can QBasic handle binary files? Perhaps if you treat the output file as binary, you might have better control.
I know this is the case with handling line breaks in C programming.

Regarding Sound Cards:
The chances are that if there is a 15 pin male connector with the pins arranged in two rows (DB 15), it is really a MIDI port.
There are of course strange arrangements, but the typical RS-232 Serial Port usually appears either as a male 25 pin connector DB 25 or 9 pin male connector DB 9. I have seen other strange arrangements but not for a very long time.
I don't think I have seen a real Serial Port on a Sound Card yet though that doesn't really prove anything.

If you are having issues with XP drivers, pick one of the newer Sound Cards and send me a few photographs of the front and back and of the connectors. Perhaps I can find something.
I just finished doing the same thing with the one in my own Game Computer.

As I might have mentioned earlier, I also have a FFB SideWinder, but I don't use it much because it really isn't precise enough for testing purposes.

- Ivan.

 

[aleatorylamp]

Old Graphic Cards, and Qbasic does make good AFX!!

Hello Ivan,
Thank you for your helpful comments!

I´m very sorry to hear about your development computer,
but I believe I have some good news for you.

Please e-mail me your home address I will be only too glad to send you
a couple of cards that may be useful to you. They are looking for a new home!
I would air-mail them to you in a little package, declared for customs as used
old hardware with no monetary worth, which is the truth anyway! It would be
no trouble at all for me to do so, and I would insist on it.

I would only have to know which slot you need a card for: Legacy/ISA, PCI or AGP.

They are not super performers, but at least they will get your development machine
back into reasonable working order.

A good one I have is an 8Mb ISA slot Graphics card with elevated platform memory,
or a 4 Mb PCI Graphics card. The best one would probably be the 8Mb PCI graphics card.
This is the only one I actually bought - in 1997, and there are also a couple of
4 mb 2D/3D AGP slot graphics cards, as well as one or two 32 Mb AGP Accellerator cards.

Tomorrow I have to go to my father in law´s house, where all this old hardware is stored,
and I´ll bring all the cards here an send you pictures of them, so you can take your pick.
I also have the drivers for many of them...

All these cards were given to me over the years by friends or ex work-colleagues.

Re: MIDI port and/or Game port:
All the old sound cards I have, have 15-pin game ports, which I thought were
serial ports, where the FFB Sidewinder could be plugged into.
A newer one is a SoundBlaster with 5 jacks, not the usual 3.
I´ll bring it here tomorrow and send you a picture too.

Re: Qbasic: You are right! It does do what I´m trying.
Learning anything is not exactly a fast process - and I have just discovered
the solution to my Qbasic inverted comma predicament:

I already knew that QBasic has 2 different reading instructions:
INPUT, which uses inverted commas and commas as separators, and
LINE INPUT, which only uses commas as separatos, where Inverted Commas
are text symbols.

For writing files, the WRITE # instruction (which I was using) adds inverted
commas at the beginning and at the end, and it took me all this time to find
out that PRINT # writes to a file without adding the inverted commas!
Simply making a copy of the same file in Qbasic produces an AFX file, that
works perfectly well in AF99 !! Now I can continue working on the QBasic
versions of StretchIt and MoveIt.

Here´s 2 screenshots of partial views of the Qbasic programme result
I´m getting so far. I have it in Window mode so I can make a screenshot.
Normally I´d work in full-screen mode.
The programme identifies all part-labels and all vertices in the AFX.
The two numbers on the left are the ASCII symbols in the fourth and the last
positions in each line, to differentiate lines with vertices from lines with labels.

The duplicated column on the right is just to double-check that the resulting
isolated line is identical to the actual line in the original text file.
You can also see the structures, that have two labels - the first label contains
no vertices, but structure position and colour, unnecessary for now.

Oh dear...Another long post, but such is life when you are enjoying yourself.
Cheers,
Aleatorylamp

 

[Ivan]

Hello Aleatorylamp,

Thanks for the offer for Graphics Cards. I actually just placed an order for a replacement today from eBay.
The total cost is probably less than what you would have spent to mail something to me.
I am also thinking that I should order a Power Supply at the same time because it makes sense to take care of as many items as I can think of while having the patient in surgery.

I also found that I have an Nvidia PCI graphics card. I don't know the amount of memory on it though.
I know that a 8 MB Voodoo 3 2000 is not sufficient for some CFS models while the 16 MB Voodoo 3 3000 seems to work with everything.

By the way, check your email. I just realised that I had been keeping intermittent backups on a flash drive that I put on the back of my Development Computer a couple years ago when I started noticing odd behaviour. It doesn't have any of the current Works in Progress such as the P-47D-25 Thunderbolt or the edits I have made to EJ's Airacobra during the course of this thread, but it does have the original AFX.
This is probably the first time the Flash Drive has been removed from my Development Computer since I first started doing backups to it.

I also tried to power up the Development Computer a few minutes ago.
There is no BIOS display, no flashes, NOTHING on the screen.

- Ivan.

 

[aleatorylamp]

Hello Ivan,
I didn´t know that more complicated models required stronger graphic accellerators to be displayed. It does sound logical though! I´d thought any graphics card with 4Mb or above would do, and that only the frame rates would go down, but possibly only on FS98.

The eBay one then sounds like the best solution, and the power supply too, because you never know if the graphic card went up in smoke because of a surge from the power supply. I remember that happening to a friend´s computer some years ago during a lightning storm. Smoke started coming out the back as the graphics card smouldered, the computer froze, the screen went colourfully streaky and then black. It also damaged the motherboard and the power supply.

Re. the Airacobra AFX by Mr. Eric Johnson: Thank you very much!
That will be a great platform to try out my AFX Scaling Qbasic Program once it is finished.
Today I will be starting to write in the mathematical vertex scaling routines.

It will also do two more things for me: I get to enjoy a new model without having to build it, and get back to practising upgrades with more reliable source files than what I had given up on some time ago. My apprentices will have more fun than otherwise!

Anyway, good luck with your hardware.
These things happen just to keep us on our toes. No rest for the wicked!
Cheers,
Aleatorylamp

 

[Ivan]

Hello All,

The new Graphics Card turned out to be a dead end and the situation is still not entirely settled.
With a Development Computer that appears it will be down for a while, I thought I would test an idea I have had for a while:
Without a working copy of AF99, I am in the same situation as if I were working with other MDL file for which I do not have the source code. My Game Computer is mostly working at this point minus a few utilities that I replace when I find they are missing.
There should be sufficient tools and resources to work on the AIR file and DP file even if I cannot adjust the MDL.

I did a few calculations a couple days ago and have a pretty good guess at the proper Center of Gravity of the actual Airacobra.
It should be fairly easy to adjust the Origin of the MDL with SCASM. This was my task for this morning.
The coding in SCASM is extremely simple: a single TransformCall to the original starting point of the MDL's drawing routines with the new offsets specified would shift the entire model by that amount.

I had originally thought of this idea to adjust a rather nice looking Hawker Hurricane MDL that had a CoG that was well aft of the Wing trailing edge so that it was not possible to have a realistic AIR file that matched the MDL.

The result was interesting.
It worked to a point. The P-39D MDL has been shifted as can be seen in the simulator.
The "Problem" though is that I have no way outside the simulator of telling where MDL actuall exists.
DPED is normally quite useful, but does ot detect the TransformCall shift so we are stuck with doing everything by eyeball in the simulator.... Unless I can find another tool that will display the shifted MDL. WYSIWYG is better.

This screenshot shows the shifted MDL with an unchanged AIR file.

- Ivan.

 

[aleatorylamp]

Hello Ivan,
Interesting, all this experimenting around.

For all it´s worth, here´s an AF99 combined screenshot showing the
CoG positions on the original AiraCobra CoG, and on my stretched version.
Maybe it can be of use to you.
The propdisc on my stretched version appears to be a bit off, as yet,
and the wheels too.
Of course... they were built at origin 0,0,0, and shifted...
Oh well... they´ll have to be identified and moved then.

Let me know if I can do something else that could be useful to you.

Cheers,
Aleatorylamp

 

[Ivan]

There is MORE than you are seeing

Hello Aleatorylamp,

I don't happen to know what scaling factor you are using but there is a reason why I chose some really large (2X and 3X) scaling factors for experimentation.
With small scaling factors, something can still be wrong and you will never see it.
As an example, you only saw the Wheels and Prop Disk.

This screenshot that I originally attached back at Post 1493 will give you an idea of what else is a bit mis-located that you are not seeing at the moment.

- Ivan.

 

[aleatorylamp]

Hello Ivan,
OK, thanks!
With the small fuselage-length correction, I suppose the numbers of other discrepancies are too small. Iused the same 1.0191 multiplier as you had done, getting 30.16 ft overall length, excluding the nose-cannon, as opposed to 30.16666 ft specified length of the plane. The screenshots have the grid on to show the difference, although it is not calibrated.

Incidentally, I didn´t process any of the offsets. Probably that is the reason nothing else went astray.

Parts created at the null point won´t react to ANY scaling and stay put, and one can only really act on them through the offsets, or making new parts in the AF99 model itself.

Here, the problem is that the old parts not only have to be placed correctly in the part editor, but have to be shifted back in the building list, and it sometimes doesn´t work, so it´s better to make new ones for the original model - then it will be processed fine, including structures.

The screenshot looks like the plane has taken some severe hits - flak maybe... Things are disjointed and missing - the wheels have been shot off and are already distant!

Anyway, re. CoG:
The CoG on both the original and slightly stretched model is the same - i.e. the light blue dot is in identical position. I remembered that there are numbers attached to it, and looked them up:
3.5, -1.1, 0 ...maybe that will help you a bit more...

Cheers,
Aleatorylamp

 

[Ivan]

Npthing is Actually Missing

Hello Aleatorylamp,

Regarding the model from my last screenshot:
Nothing is actually "Missing".
Things are just moved around in such a way that they do not appear in the right place because the shifts / offsets in the Assembly are not adjusted.
If you consider it being "Shot up by FLAK", I would call it "Blown Up" instead.
The entire Aeroplane was severely overinflated.

I cannot accept your comments that the Parts have to be re-created to make things work.
It obviously is possible to adjust for the Assembly shifts because my program did just that.
I chose not to have any one piece of the program(s) do anything terribly complicated so results were easy to validate.

Here is another screenshot with the Landing Gear contact points and fuel tank locations adjusted.
The Guns are mostly adjusted, but in this case neither mattered because Ammunition was expended and Fuel Tanks were empty.
It took a bunch of tries and even then, I am not entirely convinced it could happen in real life, but it is still interesting how marginal the CoG was.

- Ivan.