Hacker News new | ask | show | jobs
What the F-86 can teach us about software development (managingmetrics.com)
21 points by orlick 5745 days ago
8 comments

What a disappointing conclusion to an interesting historial blurb. American shot down lots of Russians because we had fly by wire planes. So Pair programming and stand up meetings are the obvious conclusion.

What a joke.

Nitpick, the controls were "mechanically assisted" (i.e. "power steering"), not fly by wire. The first FBW fighter was the F-16. http://en.wikipedia.org/wiki/General_Dynamics_F-16_Fighting_...
I had the same reaction. It felt like a sneaky publicity for Agile. On the other hand, the writer found one hell of a nice hook to get people to read the article.
It was a good story, but I stopped when it got to: "Speed through the loop beats quality and this is something we can apply to all aspects of the software development process." The jump from piloting a plane to "its therefore obviously true for software" without a shred of evidence was a little too much for me.
In addition to the tenuous link between the two activities, the 10 to 1 ratio is almost certainly untrue.

Details here (in the main text and in cited materials):

http://en.wikipedia.org/wiki/F-86_Sabre#Korean_War http://en.wikipedia.org/wiki/MiG-15#Operational_history

Isolating success in dogfights to flight controls seems like a gross oversimplification. I would like to see the original source for this (attributed to Boyd in the article).

I love how you're being downvoted for pointing to more realistic numbers.
Slight nit. >but because their arms didn’t tire as quickly from operating the flight controls

While tiring would be a factor, even if completely fresh, you can simply move a lesser weight (force) faster than a larger one.

Second problem is that there were multiple differences between MiGs and Sabres - the one I always heard was most significant, back when I was a teen, is that Sabres were better constructed, MiGs were more likely to fall apart from relatively light damage. So don't take any single article or viewpoint as the truth - reality is always more complex than will fit into a article or book.

Another important difference that went along with Boyd's thesis is that the Sabres had less obstructed canopies, letting the pilot see around themselves more easily.
See "Boyd: The Fighter Pilot Who Changed the Art of War" for a more complete description of his life. The article doesn't do it justice.

http://www.amazon.com/Boyd-Fighter-Pilot-Who-Changed/dp/0316...

Or the book "Certain to Win" which was written by one of his "disciples" with Boyd's input on how to apply all of Boyd's thoughts (which are far more complex than what the blog post shows) to business situations.

http://www.amazon.com/Certain-Win-Strategy-Applied-Business/...

Not only did Boyd make important contributions to the theory of air combat (and contests under time pressure in general), he helped develop ways to quantify the manoeuvrability of aircraft leading to the F-15 and F-16.
This reminds me of a programming assignment from college to write a chess bot.

My partner and I tweaked it's understanding of how to evaluate board positions to eek out better playing ability, but the major improvements came from implementing a more efficient specialized game tree that let us calculate another move ahead every decision.

Having one more "loop" made a huge difference against our opponents; being simply smarter or better at any of the "steps in the loop" didn't help nearly as much.

>>Boyd realized the key advantage of the F-86 was that it’s flight controls were mechanically assisted.

I am guesssing what that means is it's the tools and not the process. The process is subconscious. The tools that we use, determine how fast we can iterate. Take RoR for example, or Visual Basic(F5 anyone?), developers could iterate rapidly and after a while the process faded away, but the tools remained. Maybe, developers need to play with new toys instead of following new processes.

Pity. I thought the article would be something like this http://etutorials.org/Programming/Software+architecture+in+p...
Totally tangential but when I was a kid (in the 80s) there was a school/park somewhere in So Cal (I think) that had an actual F-86 on its playground (gutted of course). Whenever we road tripped on the weekend my parents would climb stop there and I'd climb up on it.