Forum: Program and politics

When some non-technical people talk about "source code", they confuse the issue by discussing it like its a single package when clearly it isn't.
Each sub-system has its own separate processor module, running its own firmware(software) package and preloaded data.

Some areas are more sensitive than others. The radar processor, LPI Datalink & IFF are probably the most sensitive sub-system in the aircraft in terms of source code availability, and probably its signature data (target ID signature) and combat IFF codes are even more sensitive. It is likely that signature data can be reloaded without needing full source code access.

F-16.net Sponsor

I was under the impression that that is not the case for 'mission-computerised' platforms.

Any feed (input) may result in an unexpected return (output) unless the operator has full control over the software. And that full control of the software is only possible if you have the source code.

In a completely uncontrolled input method, this is possible.

But, in a properly programmed API model, input methods and output results are clearly defined. A good error handler is also required.

For example, look at an Apple computer. It's source-code is tightly controlled but still offers an API so 3rd party programs can run. They retain tight controls over who and what is allows to run on an Apple PC. It remains a very stable PC environment.

A Windows PC, OTOH, have given out a lot of its source-code and allows very deep access into the kernal by 3rd party programmers. They also allow anyone to put anything on a PC. While this has resulted in a system that is very open and easy to program for, there are many cross-compatibility issues where one program conflicts with another. There is also an issue of sloppy code causing issues.

Since we are talking about equipment that will be used for national defense, I'll stick with Apple's model of tight controls with a solid API.

The F-35 also takes advantage of "virtual machines" instead of code running directly on the hardware. This allows for fault tolerance and rapid recovery of system restarts.

SpudmanWP wrote:

In a completely uncontrolled input method, this is possible.

But, in a properly programmed API model, input methods and output results are clearly defined. A good error handler is also required.

Isn't that contradictory? I'm pretty sure the MIL-STD-1553 is designed for interoperability. Adhering to a well defined standard (API) results in things working in a controlled manner.

SpudmanWP wrote:

For example, look at an Apple computer. It's source-code is tightly controlled but still offers an API so 3rd party programs can run. They retain tight controls over who and what is allows to run on an Apple PC. It remains a very stable PC environment.

A Windows PC, OTOH, have given out a lot of its source-code and allows very deep access into the kernal by 3rd party programmers. They also allow anyone to put anything on a PC. While this has resulted in a system that is very open and easy to program for, there are many cross-compatibility issues where one program conflicts with another. There is also an issue of sloppy code causing issues.

Relatively few Windows developers have kernel source access. Most only have access to the documented API.

Using non-documented features of either platform is risky & sloppy. The relative stability of OS X is because the unix based kernel API doesn't require the kind of kernel-level hacks that make certain things in windows unstable. Device drivers and Anti-Virus software are the two worst offenders in that regard.

I occasionally write & debug C++ software, and even if the entire source code is available, I'd only need the kernel & Operating system headers to recompile a module. I spend more time writing in PHP & Python for web applications.

Israel, Turkey and other JSF partners would be more likely only have header files, and not the complete source code.

SpudmanWP wrote:

The F-35 also takes advantage of "virtual machines" instead of code running directly on the hardware. This allows for fault tolerance and rapid recovery of system restarts.

That is correct. I was making the point that the JSF software is not a monolithic piece of software running on a single computer module.

The F-35 uses LynxOS which includes Virtual Machine isolation.
http://www.lynuxworks.com/rtos/rtos-partition-178.php

All modules are interconnected, AFAIK. Any receiver module may be supplied a path to any output module.

For example;
A certain, predetermined GPS input may trigger a sub-program, that was of course NOT documented to the operator, to disable fire control module (if there is such a module or whatever that module is called).

Is that correct?

The above is the technical part for sure. I don't think that it makes much difference to have the full source code of the F-35 in case that Turkey has a conflict with the source code owner, namely the US.

Theoretically correct, although highly unlikely it would be based on GPS if any disabling occurred. I think its more about certain countries being able to determine IFF status themselves, than being able to independently integrate avionics.

Remember the US effectively sidelined Israel during the gulf war by denying them the IFF codes. This involved F-16s with Israeli avionics.

Turkey Orders JSF Middle East Newsline June 18, 2012

http://www.menewsline.com/article-1173, ... s-JSF.aspx

"WASHINGTON [MENL] -- Turkey has quietly submitted its first order for the Joint Strike Fighter from the United States..."

One has to buy the article but I ain't buyin' it. Very Happy

The news is probably a repeat of the Reuters News. Contract for LRIP VII long lead items.
http://in.reuters.com/article/2012/06/1 ... TW20120615

Reuters doesn't mention the unit price or any conditions from Turkey. Even with the "long lead" order, changes can still be made with regard to who ends up with what aircraft. The LRIP V order hasn't been finalized, so its unlikely LRIP VII order is finalized.

OK. Thanks.

UPDATE 1-Lockheed wins $490 mln deal to buy F-35 parts Jun 16, 2012

http://in.reuters.com/article/2012/06/1 ... TW20120615

"...Lockheed Martin Corp has won a U.S contract worth $490 million to start buying parts, material and components for a seventh batch [LRIP] of 35 F-35 Joint Strike Fighter planes...

...The U.S. Department of Defense said the order would include 19 conventional takeoff and landing or "A" models for the U.S. Air Force; three F-35A models for the government of Italy; two F-35As for Turkey; six short takeoff, vertical landing or F-35B aircraft for the U.S. Marine Corps; one F-35B for Britain, and four carrier variant aircraft for the Navy.

In addition, this contract provides long lead-time efforts required for the incorporation of a drag chute in conventional takeoff air systems for the government of Norway...."

Quote:

In addition, this contract provides long lead-time efforts required for the incorporation of a drag chute in conventional takeoff air systems for the government of Norway....

Awwww, that's too bad that they went that direction.
I heard we just discovered a sure fire solution for Norway's short runway problems.
It is supposed to be completely stealthy from all aspects for the F-35!
Not only that - it should work on their remaining F-16s as well as just about any other aircraft that they use it with.

.

If life was only that simple for icy runway landings.

like they don't have ice in Russia, Canada or Alaska?

The longer runway solution keeps working year after year, day and night jets and props can use it, and doesn't require that any government officials resign in disgrace.
It also doesn't require that Norway gives a nickle to those shifty devilish Yankees and their secretive Unit Cost contracts and tight-fisted source code controls.

Last edited by spazsinbad on Jun 19, 2012 - 06:04 AM; edited 1 time in total

AFAIK the issue with iced runways (and there are many forms and thicknesses of ice to deal with as well as any crosswinds) is that the aircraft has to slow enough for the brakes to become effective without problems arising from their use - on the ice. If things do not go well it is even less far than in a straight line to GO OFF THE RUNWAY. The best runway is one that is the same length in all directions. Is that what the chap with the bulldozer is proposing?

Way back in the dream time (just after WWII) there were USN tests for circular runways. Now there is a thought. But of course there were practical difficulties (which I have forgotten). BTW I do not have experience on icy runways but enough on short wet (with inches of water over the) runways to know that slowing and braking carefully in a finite distance - IN A STRAIGHT LINE - is fraught. Thank heavens for long and short field arrest gear with dat hook. Very Happy

ADDITION: I might explain the 'inches of water'. Certain aircraft will react in different ways to different small amounts of water on a runway in a condition called 'aquaplaning' which I believe equates to an iced runway experience. Throw in any aircraft with a tendency to tip over (the A-4 Skyhawk) and much fun can be had braking under different conditions. Top that off with different tyre pressures for ashore or afloat and the recipe is really spicy. For example an A-4 had twice the main tyre pressure ashore for carrier landings. Then be catapulted with those higher pressures to land ashore on a wet runway and - brother.... Rolling Eyes

As always Spaz you make good points.
I suppose I am the problem. I just get weary of Norway waving the "I'm Special Needs!" flag.
Really Norway? Still playing that card? Is it really easier to redesign everything that flys than figure out those pesky paving machines function?

Forum: Program and politics

Turkey to buy two F-35s in 2012: minister