Joint Strike Fighter Air Vehicle C++ Coding Standards [pdf]

Definitely worth noting the year, but I’d be fairly surprised if the standards for real-time avionics software changed much since then. I’d guess they still stick to a small subset of the Goliath that is C++

Re: Joint Strike Fighter Air Vehicle C++ Coding Standards [pdf]

#5

Post by ncmncm » Mon, Mar 08, 2021, 5:44 AM UTC

austinprete wrote: post #4

ncmncm wrote: post #2

(2005) It is a very, very different language, today, than when this was composed.

Definitely worth noting the year, but I’d be fairly surprised if the standards for real-time avionics software changed much since then. I’d guess they still stick to a small subset of the Goliath that is C++

Most of the Goliath results in faster, safer, more reliable code. By design.

Avoiding new features means slower, unsafe, less reliable code. At the extremum, in this direction, is the C subset that (e.g.) Linux and BSD kernels wallow in.

Re: Joint Strike Fighter Air Vehicle C++ Coding Standards [pdf]

#6

Post by 6pac3rings » Mon, Mar 08, 2021, 12:52 PM UTC

ncmncm wrote: post #5

austinprete wrote: post #4

Earlier quoted context omitted.

Definitely worth noting the year, but I’d be fairly surprised if the standards for real-time avionics software changed much since then. I’d guess they still stick to a small subset of the Goliath that is C++

Most of the Goliath results in faster, safer, more reliable code. By design. Avoiding new features means slower, unsafe, less reliable code. At the extremum, in this direction, is the C subset that (e.g.) Linux and BSD kernels wallow in.

The LM Goliath needs to put up actual performance figures under scientificly rigorous scrutiny for comparison to substantiate those claims against enthusiast driven technology (not marketing, PR) or shutup. And, factor in dollars a bang for every block of technical excellence performing in the source code.

Re: Joint Strike Fighter Air Vehicle C++ Coding Standards [pdf]

#7

Post by austinprete » Mon, Mar 08, 2021, 4:30 PM UTC

Is anyone aware of an open-sourced codebase that uses a highly restrictive standard like this? Probably more like likely just plain MISRA, since anything developed by Lockheed using this doc would be classified I'm sure.

I've seen standards like this shared a few times, but never seen the code they actually result in.