4 ms·

Leave it to HN engineers to hand wave away going 10x over budget and a decade late.

dahfizz 5y ago
Yeah, the limited lifespan would be much less concerning if we had any reason to believe we could iterate and crank out more / better telescopes in the future.
- jimkleiber 5y ago
  I guess I just assume they learned a lot in building this and that knowledge will transfer into new telescopes, launches, deployments, etc. Do you think they didn't learn enough by figuring this out?
  - emptyfile 5y ago
    IIRC the companies that made the beryllium main mirror went bankrupt because the material is much more tricky then anticipated. So you would have to do that again basically or come up with another mirror design.
    - jimkleiber 5y ago
      Ah I hadn't know. That being said, perhaps one thing they learned from this is which materials they may not want to use again in the future. I dunno, I just struggle to imagine that they didn't learn a lot from this and that most of us have really any idea whether the learnings are worth $10B, $1B, $100B, etc.
  - dahfizz 5y ago
    I am not confident that whatever caused JWT to take so long will be solved for the next project. NASA by its nature does not really benefit from "scale". Every important thing they do / produce is novel. I am sure they have learned alot from building JWT, as they certainly learned a lot from building Hubble or the space shuttle. I don't think that knowledge necessarily transfers forwards to make the next project easier.
    - jimkleiber 5y ago
      Well, as you seem to have said, NASA doesn't seem too interested in scaling but rather in doing things that stretch the bounds. If their goal isn't to scale the previous project but to push past what people thought were possible with experiments, that seems OK to me.
username_my1 5y ago
it's something that was never done before and one of the few things that will create all kind of new sciences. Your comment is really frustrating. few people in this life can do something that was never done before, much much fewer on this scale, much much much fewer with such impact on science and humanity.
oneoff786 5y ago
For one of the most complicated pieces of engineering ever achieved sure
onion2k 5y ago
The initial estimate turned out to be wrong. I hand wave that sort of thing away every day.
jimkleiber 5y ago
I'm curious as to why and whether some of the delay had to do with the public and lawmakers getting impatient and trying to cut it short, then actually forcing it longer, then blaming the engineers.
dredmorbius 5y ago
I find it interesting that planning overruns such as this seem to get pinned on implementation, rather than the planning and approval processes themselves. The evidence is that the time and cost it took to produce JWST are what they are. That planners lowballed these estimates, or that administrators and legislators wouldn't have approved greenlighting based on accurate estimates, speak to failures in planning and administration, rather than engineering. Much of the planning / large-project literature, which I've studied for some three decades, seems itself to be largely blind to this distinction. As to many comments found on the Internet mirroring such viewpoints.
- thehappypm 5y ago
  Maybe it’s by design. A 5 year project can be greenlit by Congress, and delays forgiven. 20 year project? Dies in committee.
  - dredmorbius 5y ago
    Nothing to show the folks back home, risk of being either primaried or challenged by the other party. Welcome to politics.
redisman 5y ago
Hilarious criticism coming from software engineers which is never late or over budget. Even when there’s nothing actually novel in the product
- adtac 5y ago
  I don't want to tell NASA/ESA/CSA how to operate, but I haven't heard of a single software product that launched 10 years late. Such a product would be immediately DOA because of how fast technology moves. Obviously a space telescope is literally rocket science, so it's nothing like the next popular JavaScript library, but I think NASA could adopt a little bit of the move-fast-break-things mentality. Apparently they're running Python 1.5 on the JWST ffs. All I'm saying is there's room to improve. JWST would be a monumental success for humanity even if it launched 20 years late.
  - qsmi 5y ago
    I can count so many I can't even decide where to begin. This being the classic: https://en.wikipedia.org/wiki/Development_of_Duke_Nukem_Forever https://en.wikipedia.org/wiki/Development_of_Duke_Nukem_Fore... "The video game Duke Nukem Forever spent fifteen years in development, from 1996 to 2011."
  - baq 5y ago
    yeah most get cancelled before that time, so you don't hear about them.
  - Okx 5y ago
    I don't think you should be risking anything when there's $5bn at stake and a single failure can cause you to lose all of it

Yeah, the limited lifespan would be much less concerning if we had any reason to believe we could iterate and crank out more / better telescopes in the future.

jimkleiber 5y ago
I guess I just assume they learned a lot in building this and that knowledge will transfer into new telescopes, launches, deployments, etc. Do you think they didn't learn enough by figuring this out?
- emptyfile 5y ago
  IIRC the companies that made the beryllium main mirror went bankrupt because the material is much more tricky then anticipated. So you would have to do that again basically or come up with another mirror design.
  - jimkleiber 5y ago
    Ah I hadn't know. That being said, perhaps one thing they learned from this is which materials they may not want to use again in the future. I dunno, I just struggle to imagine that they didn't learn a lot from this and that most of us have really any idea whether the learnings are worth $10B, $1B, $100B, etc.
- dahfizz 5y ago
  I am not confident that whatever caused JWT to take so long will be solved for the next project. NASA by its nature does not really benefit from "scale". Every important thing they do / produce is novel. I am sure they have learned alot from building JWT, as they certainly learned a lot from building Hubble or the space shuttle. I don't think that knowledge necessarily transfers forwards to make the next project easier.
  - jimkleiber 5y ago
    Well, as you seem to have said, NASA doesn't seem too interested in scaling but rather in doing things that stretch the bounds. If their goal isn't to scale the previous project but to push past what people thought were possible with experiments, that seems OK to me.

username_my1 5y ago

it's something that was never done before and one of the few things that will create all kind of new sciences. Your comment is really frustrating. few people in this life can do something that was never done before, much much fewer on this scale, much much much fewer with such impact on science and humanity.

oneoff786 5y ago

For one of the most complicated pieces of engineering ever achieved sure

onion2k 5y ago

The initial estimate turned out to be wrong. I hand wave that sort of thing away every day.

jimkleiber 5y ago

I'm curious as to why and whether some of the delay had to do with the public and lawmakers getting impatient and trying to cut it short, then actually forcing it longer, then blaming the engineers.

dredmorbius 5y ago

I find it interesting that planning overruns such as this seem to get pinned on implementation, rather than the planning and approval processes themselves. The evidence is that the time and cost it took to produce JWST are what they are. That planners lowballed these estimates, or that administrators and legislators wouldn't have approved greenlighting based on accurate estimates, speak to failures in planning and administration, rather than engineering. Much of the planning / large-project literature, which I've studied for some three decades, seems itself to be largely blind to this distinction. As to many comments found on the Internet mirroring such viewpoints.

thehappypm 5y ago
Maybe it’s by design. A 5 year project can be greenlit by Congress, and delays forgiven. 20 year project? Dies in committee.
- dredmorbius 5y ago
  Nothing to show the folks back home, risk of being either primaried or challenged by the other party. Welcome to politics.

redisman 5y ago

Hilarious criticism coming from software engineers which is never late or over budget. Even when there’s nothing actually novel in the product

adtac 5y ago
I don't want to tell NASA/ESA/CSA how to operate, but I haven't heard of a single software product that launched 10 years late. Such a product would be immediately DOA because of how fast technology moves. Obviously a space telescope is literally rocket science, so it's nothing like the next popular JavaScript library, but I think NASA could adopt a little bit of the move-fast-break-things mentality. Apparently they're running Python 1.5 on the JWST ffs. All I'm saying is there's room to improve. JWST would be a monumental success for humanity even if it launched 20 years late.
- qsmi 5y ago
  I can count so many I can't even decide where to begin. This being the classic: https://en.wikipedia.org/wiki/Development_of_Duke_Nukem_Forever https://en.wikipedia.org/wiki/Development_of_Duke_Nukem_Fore... "The video game Duke Nukem Forever spent fifteen years in development, from 1996 to 2011."
- baq 5y ago
  yeah most get cancelled before that time, so you don't hear about them.
- Okx 5y ago
  I don't think you should be risking anything when there's $5bn at stake and a single failure can cause you to lose all of it