Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The F-35's $400k helmet

washingtonpost.com

Post reply on HN
Print view View original

111–120 of 136 posts

Re: The F-35's $400k helmet

#111

Post by tormeh »

revelation wrote: post #42
pi-err wrote: post #17

From the article: > a stealthy fighter jet years that is often called a flying computer because of its more than 8 million lines of code For all the critics, all the hate vs drones and leaner fighters and bombers, the F-35 _is_ an impressive engineering feat and a turning point, especially re:software. At some point, a "Snow Leopard" version of the F-35 will turn out as a very fine piece of equipment, only possibly w…

Huh? You can't boot any recent system without running code from software that totals many lines more than a mere 8 million. That they keep quoting it really just makes me scared for the people working there, their worth measured by the lines they contributed to the behemoth as dictated by people that started in the 70s and never did learn anything new.

Yes, but not made in one go. It's sad that it's a giant piece of C++ fail, but you can't fault them for ambition.

Re: The F-35's $400k helmet

#113

Post by JohnBooty »

mkaziz wrote: post #73
JohnBooty wrote: post #65

Earlier quoted context omitted.

The good news is that war kills less people than it used to. Paradoxically, these fantastically lethal killing machines might be actually preventing deaths. The US/Iraq war and occupation lasted about a decade and there were less than 200,000 direct combat deaths. Contrast with WWII, in which there were quite a few individual battles with greater death tolls... between one and two million people died in the Battle of…

Pedantic but: it won't ruin the planet - the planet is remarkably resistant. But we might succeed in killing ourselves and a bunch of other species off.

Good point.

Re: The F-35's $400k helmet

#114

Post by fapjacks »

njharman wrote: post #99
sschueller wrote: post #5

All this technology just to kill people. It is a sad world we live in.

Humans kill people. Always have, always will. Only the most facile could possibly believe this or any war machine is "just to kill people".

Yes, but there's a reason the US government won't let me drive an M1 to work. Which is unfortunate, because I would love to. I'm just saying, let's not make inaccurate assertions about the purpose of some of our tools.

Re: The F-35's $400k helmet

#115

Post by fapjacks »

jonsen wrote: post #21
flyinglizard wrote: post #3

The nice thing about the public scrutiny of the F-35 program, is all this normally-classified capability information that's released out to the public to get them excited about the project. Such level of publicity of cutting edge weapon systems simply didn't exist a decade ago.

It's a smoke screen. In the skunks they are doing a pilotless version.

Exactly. And probably one that can stay airborne for weeks or months and drop ordinance autonomously.

Re: The F-35's $400k helmet

#116

Post by trhway »

nazgob wrote: post #55
trhway wrote: post #51

Earlier quoted context omitted.

>>>Modernized versions of T-72 is still in wide service with Russian army. >Oh, that's funny. SKS is in wide service. Along with Mosin-Nagant. In active service Russia has 2000 T-72, 2000 T-90, 3000 T-80. And 3 or 4 of T-14. So, according to you, which tank is in wide service in Russia?

Where did you get data about 2k T-90s in service?

you're right, Russia has only about 1K of these, with 2K being the total production run so far (ie. another half was exported)

Re: The F-35's $400k helmet

#117

Post by bunderbunder »

njharman wrote: post #97
furyg3 wrote: post #25

It's a very 'agile' project. When will it be done? Don't ask! How much will it cost? Who knows! It's going to be awesome, though! But... you don't have to wait to see how powerful this weapons system is. The F-35 completely took out the Dutch government before one even rolled off the assembly line! Note to governments: when ordering your weapons systems it may be a good idea to get a cost-per-unit quote before signin…

Agile has fixed delivery date and variable features delivered on that date. Features are prioritized by customer at start of every iteration. Done right, Agile delivers exactly the things customer has chosen to afford. In my experience unfortunately, exactly 0% of customers (internal or external) are ablet to accept/wrap their head around a variable feature deliverable and demand a fix set of features decided at star…

You can do fixed delivery, variable features or fixed features, variable delivery. It doesn't matter which, the point is just to acknowledge that you can't fix both so you better understand which is more important to you.

That said, it's perfectly reasonable for customers to expect to have some promise about what they're paying for. That's where you trot out the idea of a minimum viable product: Make some tough decisions about what features are actually required, and then agree to a timeline that makes it a virtual certainty that you'll be able to deliver all of the required ones. When you hit that mark ahead of schedule and start burning through the list of nice-to-haves they'll be elated.

Re: The F-35's $400k helmet

#118

Post by monknomo »

rayiner wrote: post #104
monknomo wrote: post #96

Earlier quoted context omitted.

Most government agencies that do 'research' funnel grant and contract money to private entities. Having scientists or building things directly is 'communism' and we're against that. After all, a government employee manages to be both expensive and incompetent, so why not save money by hiring contractors to do their work? /s I think going all in on the F-35 has to do with the some kind of power struggle between the ai…

At least with DARPA it's not about communism but wanting to avoid bureaucracy. They don't even own their own building and program managers are kicked out after four years. Basically they don't want to become NASA.

It's true that ending up like NASA would be unfortunate, but it seems like there is a better way. Also, my crack about communism was not meant to be from DARPS's viewpoint. I see it more that it is easier to fund something that spreads money to private business across many districts than it is to fund generic government activity.

Re: The F-35's $400k helmet

#119

Post by cm127 »

ghostberry wrote: post #47
cm127 wrote: post #36

Earlier quoted context omitted.

Not really. Without such technology, war wouldn't be profitable. Without the profits, there would be no incentive to fight. But as long as we carry the bigger gun, we'll always have an incentive to fight.

"Without such technology, war wouldn't be profitable. Without the profits, there would be no incentive to fight." This is without a doubt one of the most backwards ignorant things I have ever read online. It might come as a shock to you, but humans have been killing other humans since the stone age. And, as time has gone on, it's become less and less likely for a person to die violently, at the same time that arms te…

You can tell yourself we fight for honor or pride or all that bullshit, but it's always been about money. The question before going into war has always been: how many lives is it worth? i.e. what's the profit?

Re: The F-35's $400k helmet

#120

Post by RogerL »

enraged_camel wrote: post #80
rayiner wrote: post #72

Earlier quoted context omitted.

Defense systems go over-budget for the same reason software projects do--it's all custom and custom development is unpredictable. It's not like these companies can sell the U.S. stuff they had lying around having developed it for another customer. I used to do development for a DOD project. At the outset of the project, it's often not clear what you're trying to do will even work. It might not even be clear what you…

>>I used to do development for a DOD project. At the outset of the project, it's often not clear what you're trying to do will even work. It might not even be clear what you should be trying to do. That's interesting, and also scary. I work in IT consulting where it's extremely rare for customers to pay for a professional services project before seeing a proof of concept demonstration of what will be delivered. In fa…

Okay, extrapolate this to a weapons system. You can't expect a company to come up with their own version of the F-35 (which is extremely advanced) with the hope that, if you are satisfied, you'll buy 100 or so. If the government will not go at risk for this, there is no way a company, with vastly fewer resources, is going to do it.

I've bid weapons systems and built weapons systems. I can honestly say that you make your very best effort to come up with a realistic proposal (we aren't trying to rip off the government. I'm just as much a taxpayer as you are). But there are limits. Detailed requirements are only going to be known well after you start. The government doesn't know how to specify the requirements - in large part that is your job. When you bid you want to be sure that you can be profitable, but then you also have competition, and a blind process (you don't know what your competitor is proposing, and at what price). This limits collusion, but the lack of information makes it hard to write a good proposal. Company A might think feature X is important, and put that in the contract. You didn't. Who is right? You'll find out when the contract is awarded. Opps if you got it wrong.

Once you win the real "fun" begins. If you are lucky the program office running your contract understands the nature of research and the unknowability of the future, and works with you to set a reasonable scope for the project, and works with you as more information is found. OTOH you can have a bean counter trying to protect their career by counting "shalls" in the document (a legally binding term) and ensuring that each one can be checked off with what you deliver. Doesn't matter if it is rational or not. And then there are copious documentation requirements. The intent is that these requirements be tailored to the contract and needs, but again bean counters can intervene and have you doing massive amounts of busy work. A great program officer will stream line it and have you doing effective work.

But, you know, there is not infinite money to prototype, build, test, scrap, and build again. You have 1 billion moving parts (figuratively) that you are trying to track and optimize for, and you kind of work your way to a solution. that is, unless you can afford the project to cost 10x as much (you are already complaining about cost, so that is a nonstarter), or take 5x as long (why don't I have my jet yet?!?!). The result is, unsurprisingly, unoptimal.

tl;dr: weapons systems are an extremely expensive mix of state of the art research/development/deployment all in one. It is not going to go smoothly. It is the worst possible way to do a project, except for all the other ways, which are worse.

Post reply on HN