4 ms·

That is to assume that those customers care about economic benefit. Frankly, if cost were their only concern, the US military would have abandoned ULA and gone

That is to assume that those customers care about economic benefit. Frankly, if cost were their only concern, the US military would have abandoned ULA and gone with SpaceX long ago. I'm reminded of the first calls (90s->2002, not the later programs) for bids to replace the KC-135. There were only two real contenders, Airbus and Boeing. Airbus's first reaction was essentially: "why even bother."

greglindahl 12y ago
When you're launching a $2B satellite, saving $60mm on the launch is not a slam-dunk. The Air Force had a series of launch failures that made them extremely conservative. Sure, I think they ought to be moving faster with SpaceX, but you should give them credit for having some serious concerns.
- avmich 12y ago
  A lot of that $2B cost comes from requirement that everything should just work, and therefore thoroughly tested, because customers don't want to have it broken in space where no one can fix things and have to send another one... which costs $2B... see cyclical reasoning? If launches would be cheaper, customers could afford to make less expensive payloads - after all, if it breaks, they can just send another one. A primary reason why payloads are so expensive is because launch services - used to be - so expensive. With lowering the launch price payloads can become cheaper.
  - greglindahl 12y ago
    The situation today is that the Air Force has a $2B satellite to launch soon that they ordered a decade ago. You're right that tomorrow, a different one could be built with the new launch cost in mind. The entire "new space" movement revolves around lowering launch costs and building much cheaper satellites to take advantage of lower launch costs.
  - branchan 12y ago
    That's not entirely correct. The bottom line is that military satellites are just expensive - period. They will always be expensive because of their capabilities.
    - meric 12y ago
      Agreed. They will always have more capability than a non-military satellite, and so will be more expensive. If it didn't have more capability than a non-military satellite, then it isn't a military satellite.
      - eru 12y ago
        Why? The military might go far a carefully selected subset of capabilities, and get a cheaper price or more reliability out of it.
        meric 12y ago
        I meant it had to be different to non-military satellites in some important ways. Otherwise it'd be just a non-military satellite. Imagine buying a consumer-grade computer. It's cheap to get one with 16gb ram, but if you want a 64gb ram one, it would probably vastly more than the 64/16=4 times as expensive. Each step you go out of the non-military satellite capability curve (with multiple dimension including features, reliability), it gets exponentially more expensive. If the a research-grade computer wasn't different from a consumer-grade computer in some aspect (by having a lot of ram, cores, processing speed), it would just be a consumer-grade computer. A computer that is substantially cheaper would be hobbyist-grade computer. Of course I'm only talking about traditional military rather than forces like A* Qda which might make do with whatever they can get their hands on, which is probably what you're referring to. I suppose a very cheap satellite could be terrorist-grade. ;)
        eru 12y ago
        You could argue the same the other way 'round. Just assume that military is the standard, and in order to qualify as consumer, you need to add extra features. So far your argument is symmetric.

greglindahl 12y ago

When you're launching a $2B satellite, saving $60mm on the launch is not a slam-dunk. The Air Force had a series of launch failures that made them extremely conservative. Sure, I think they ought to be moving faster with SpaceX, but you should give them credit for having some serious concerns.

avmich 12y ago
A lot of that $2B cost comes from requirement that everything should just work, and therefore thoroughly tested, because customers don't want to have it broken in space where no one can fix things and have to send another one... which costs $2B... see cyclical reasoning? If launches would be cheaper, customers could afford to make less expensive payloads - after all, if it breaks, they can just send another one. A primary reason why payloads are so expensive is because launch services - used to be - so expensive. With lowering the launch price payloads can become cheaper.
- greglindahl 12y ago
  The situation today is that the Air Force has a $2B satellite to launch soon that they ordered a decade ago. You're right that tomorrow, a different one could be built with the new launch cost in mind. The entire "new space" movement revolves around lowering launch costs and building much cheaper satellites to take advantage of lower launch costs.
- branchan 12y ago
  That's not entirely correct. The bottom line is that military satellites are just expensive - period. They will always be expensive because of their capabilities.
  - meric 12y ago
    Agreed. They will always have more capability than a non-military satellite, and so will be more expensive. If it didn't have more capability than a non-military satellite, then it isn't a military satellite.
    - eru 12y ago
      Why? The military might go far a carefully selected subset of capabilities, and get a cheaper price or more reliability out of it.
      - meric 12y ago
        I meant it had to be different to non-military satellites in some important ways. Otherwise it'd be just a non-military satellite. Imagine buying a consumer-grade computer. It's cheap to get one with 16gb ram, but if you want a 64gb ram one, it would probably vastly more than the 64/16=4 times as expensive. Each step you go out of the non-military satellite capability curve (with multiple dimension including features, reliability), it gets exponentially more expensive. If the a research-grade computer wasn't different from a consumer-grade computer in some aspect (by having a lot of ram, cores, processing speed), it would just be a consumer-grade computer. A computer that is substantially cheaper would be hobbyist-grade computer. Of course I'm only talking about traditional military rather than forces like A* Qda which might make do with whatever they can get their hands on, which is probably what you're referring to. I suppose a very cheap satellite could be terrorist-grade. ;)
        eru 12y ago
        You could argue the same the other way 'round. Just assume that military is the standard, and in order to qualify as consumer, you need to add extra features. So far your argument is symmetric.