13 ms·

All software has bugs. The problem here is the mismanagement, the crooked procurement, and the rush to flight because potential customers and diplomatic influe

by imglorp 11y ago

All software has bugs. The problem here is the mismanagement, the crooked procurement, and the rush to flight because potential customers and diplomatic influencees are fleeing.

mrweasel 11y ago
Mismanagement started very early to the F-35. I know nothing about planes, but I do know that if you want a software system to do EVERYTHING, it will never live up to expectation, if it's ever delivered. The rational of having just the one type of plane makes sense if you're Ryanair (who only fly Boeing 737-800). Of cause they only have ONE role for all their plane. Having just one type plane take the role of multiple others just gives you all of the complications of each of these plane all rolled up in one, plus a series of its own unique issues. I understand that Lockheed Martin just tries to deliver a product that request, but that's really the issue with many government contracts. You get what you request, and if someone tells you "Yeah, that might be a bad idea", the contract is just rewarded to a less honest company. It would be a bit shocking if no one with Lockheed Martin questioned if the F-35 was a good idea. At some point someone is going to dictate that the F-35 is done, true or not. The US will fly them for a short while and the other nations, that can't afford new planes for another 20-30 years will be stuck with this thing.
- masklinn 11y ago
  Note that you can have multiple types of plane with very similar controls & software, IIRC that's one of the advantages of Airbus the systems are pretty uniform so once a pilot's qualified on one plane it's much easier to qualify on others.
- rm_-rf_slash 11y ago
  The F-35 was a bad idea. In theory it made sense: fast, high tech, and stealthy. Then reality settled in: The plane can only really support a small handful of missiles, nothing close to the Arsenal the Navy's F-18 used to be able to haul into battle. It's fast, sure, but that comes at a cost to fuel efficiency, so while infantry in remote Afghanistan can get air support in less time than a pizza, "one pass, haul ass" means that if the air strike wasn't perfect, they need to call in a second strike group and wait again, not to mention the immense cost for the first few jets and bombs. The military is trying too hard to make a one-size-fits-all solution, when there has been one for close air support since Vietnam: the rugged and powerful A-10. In fact, with more and more of the fighters being forcibly retired, special ops teams are rightfully concerned, since they could previously rely on the A-10 to make 10+ passes before having to refuel, giving ground forces a significant advantage. The moral of the story is that more tech, more cost, and more complexity does not always culminate in more value. We have become so accustomed to the notion that things will always get more advanced, and therefore better, that we have forgotten to look to the past to replicate the same success in the future.
  - mrweasel 11y ago
    It makes you wonder what it would cost to built updated versions of the A-10 and F-18.
    - taf2 11y ago
      Pretty sure the F-18 has been getting updates - google f-18 stealth.
      - rm_-rf_slash 11y ago
        True, although the stealth capabilities are nowhere near the F-22 or F-35. Still, it helps to pack more than a couple missiles or bombs when the enemy inevitably discovers that you've dropped in to say hello.
      - Mikeb85 11y ago
        The EA-18 is probably the most interesting aircraft the air force has. At the rate technology is advancing, electronic warfare seems like it'll win over passive stealth.
        engi_nerd 11y ago
        The Navy operates the EA-18G, not the Air Force.
    - imglorp 11y ago
      Fairchild is long gone, having been bought years ago and lines closed and moved. You'd be starting almost at scratch.
    - Recurecur 11y ago
      The F-15 Silent Eagle is also interesting, and is a far better fighter than the F-18. I really like the B-1R concept, too bad it'll never be built.
  - spatulan 11y ago
    The A-10 is a disaster that needs to be retired. It's completely incapable of facing modern air defences, so is useless against enemies like China or Russia, and has operating costs 30 times higher than a dedicated COIN aircraft like the Super Tucano, meaning you have a lot fewer aircraft supporting your troops than you should have, or end up spending way too much money.
- gherkin0 11y ago
  > The rational of having just the one type of plane makes sense if you're Ryanair (who only fly Boeing 737-800). Of cause they only have ONE role for all their plane. Having just one type plane take the role of multiple others just gives you all of the complications of each of these plane all rolled up in one, plus a series of its own unique issues. In addition to that, they didn't manage to build just one type of plane. They actually built three types that look sort of the same and share a few parts (apparently not enough to actually save money). > At some point someone is going to dictate that the F-35 is done, true or not. The US will fly them for a short while and the other nations, that can't afford new planes for another 20-30 years will be stuck with this thing. I like the idea of building updated versions of existing designs as a hedge against the failure of this project. It looks like some of those are getting more interest: http://www.popularmechanics.com/military/weapons/news/a18052/boeings-stealthy-f-15-silent-eagle-israel/ http://www.popularmechanics.com/military/weapons/news/a18052...
- jcadam 11y ago
  Several years ago I worked on one Lockheed project (Not the F-35, but part of another large defense program) as a subcontractor (at a Lockheed facility). The most dated development practices imaginable. Waterfall, no source control (they brought in subversion near the very end of the project in order to meet govt requirements but only the 'architect' was allowed to commit anything (or even to use the 'svn' command)), SLOC used to measure "progress", even the office furniture looked 70s vintage. The project was one giant clusterf*k from start to finish, and I swore an oath never to work on anything involving "Lockheed" ever again.
  - michael_h 11y ago
    I had my software installed in a Lockheed system. It was incredibly surreal - the amount of time spent assigning blame for problems was orders of magnitude more than the time spent solving problems. One of their sysads, in a meeting with the Important People, blamed my system for extremely high network and disk load. He showed graphs that spiked and plateaued a few minutes after system load, pointed at the spike, then pointed at my face and said, "It's his software that's making this whole thing slow. Get him out of here and we'll build something that actually works." I got unparalleled satisfaction when I got to say, "I haven't installed my software yet."
- alkonaut 11y ago
  Multi-role aircraft aren't a bad idea, so long as each role somewhat overlaps the other in requirements. Many well designed aircraft do 2-3 roles very well (e.g recon/attack/fighter roles). If you mix in another role that doest't match the hardware requirements of the first role that's when you are asking for problems. CAS, SToVL are such examples that won't blend with the first 3 without massive compromises. Full stealth is also a requirement that is rarely absolutely necessary, but adds to the cost and lowers performance of every other aspect of the plane. When these compromises had been discovered the project should have been scrapped, the SToVL elements thrown out and a separate plane should have been procured for that role. Lots of components could be shared, just not the airframe. I too believe the project is very near the edge now. It will survive as there is too much prestige invested, but if a few big players like Canada or Australia decide to get zero F-35's the unit development cost would skyrocket which would force other buyers to to back out. If was LM I'd be sending huge lobbying checks to Canada right now.
mannykannot 11y ago
I have a suspicion that the project (not just the software) came off the rails when 'and VTOL' was put in the requirements.
- p_eter_p 11y ago
  I've seen it mentioned in at least one article that the VTOL requirement is driven by the USMC's institutional memory of Guadalcanal. They don't want to get left on the beach without air support again, so they insist on having their own aircraft with VTOL capability. How valid a concern that is is..up for debate, in the era of nuclear supercarriers and jet aircraft.
  - ZanyProgrammer 11y ago
    Especially with American global air superiority, and the fact that the Marines haven't done opposed amphibious landings since...Inchon?
    - dmos62 11y ago
      Global air superiority? Not a geo-powers geek, but sounds implausible.
  - gaius 11y ago
    ... And the Royal Navy don't forget.
SFjulie1 11y ago
And russians still have planes with hydraulic commands from the cold war (50 yo) that outperform F35, F16, F15, F14, B2 :) They even flew near UK exactly one year ago with a turbine powered bomber from the 50s and a nuke inside and no planes intercepted it during its 5 hours flight over "modern air farce" doted countries. Notably the UK are lacking interceptors because they planned to replace theirs with the F35.... and the F35 is not there. Does not it sound ridiculous? And the stealth technology are nowhere to be stealth for long. It leaves heat signature, signal distortions. And why do US stealth planes have so much sharp angles when we all know it is bad for being stealth? Because their software don't let them do curbs... To avoid software bugs remove software from where it is not useful
- talmand 11y ago
  I don't understand the Russian bomber part, did they fail to detect the bomber or failed to intercept due to not having planes? Are you suggesting that the UK totally ditched all of their current interceptors and have their pilots sitting around waiting for the F-35 to show up? Are you saying the existing stealth planes are a money-making scam because we all know they built them wrong? The software doesn't allow... "curves"? Why doesn't it allow curves?
  - logfromblammo 11y ago
    Of course it allows curves! It's just that they can't have curvature less than a great circle....~ Due to the way the military produces software specifications, it is very likely that the unbroken chain of subject-matter-experts from the days of ochre pictograms on cave walls to today have ensured that the software program functions exactly the same as drawing lines on charts by hand. Since human navigators did not initially possess the capability to continuously recalculate courses, bearings, and headings, as would be required in a curved flight path, the ground-based and flight computers don't do it now. So planes still keep a fixed heading between a set of points, and any possible roundness is encoded into the turn data or a specialized, predefined pattern.
    - SFjulie1 11y ago

by imglorp 11y ago

All software has bugs. The problem here is the mismanagement, the crooked procurement, and the rush to flight because potential customers and diplomatic influencees are fleeing.

mrweasel 11y ago

Mismanagement started very early to the F-35. I know nothing about planes, but I do know that if you want a software system to do EVERYTHING, it will never live up to expectation, if it's ever delivered. The rational of having just the one type of plane makes sense if you're Ryanair (who only fly Boeing 737-800). Of cause they only have ONE role for all their plane. Having just one type plane take the role of multiple others just gives you all of the complications of each of these plane all rolled up in one, plus a series of its own unique issues. I understand that Lockheed Martin just tries to deliver a product that request, but that's really the issue with many government contracts. You get what you request, and if someone tells you "Yeah, that might be a bad idea", the contract is just rewarded to a less honest company. It would be a bit shocking if no one with Lockheed Martin questioned if the F-35 was a good idea. At some point someone is going to dictate that the F-35 is done, true or not. The US will fly them for a short while and the other nations, that can't afford new planes for another 20-30 years will be stuck with this thing.

masklinn 11y ago
Note that you can have multiple types of plane with very similar controls & software, IIRC that's one of the advantages of Airbus the systems are pretty uniform so once a pilot's qualified on one plane it's much easier to qualify on others.
rm_-rf_slash 11y ago
The F-35 was a bad idea. In theory it made sense: fast, high tech, and stealthy. Then reality settled in: The plane can only really support a small handful of missiles, nothing close to the Arsenal the Navy's F-18 used to be able to haul into battle. It's fast, sure, but that comes at a cost to fuel efficiency, so while infantry in remote Afghanistan can get air support in less time than a pizza, "one pass, haul ass" means that if the air strike wasn't perfect, they need to call in a second strike group and wait again, not to mention the immense cost for the first few jets and bombs. The military is trying too hard to make a one-size-fits-all solution, when there has been one for close air support since Vietnam: the rugged and powerful A-10. In fact, with more and more of the fighters being forcibly retired, special ops teams are rightfully concerned, since they could previously rely on the A-10 to make 10+ passes before having to refuel, giving ground forces a significant advantage. The moral of the story is that more tech, more cost, and more complexity does not always culminate in more value. We have become so accustomed to the notion that things will always get more advanced, and therefore better, that we have forgotten to look to the past to replicate the same success in the future.
- mrweasel 11y ago
  It makes you wonder what it would cost to built updated versions of the A-10 and F-18.
  - taf2 11y ago
    Pretty sure the F-18 has been getting updates - google f-18 stealth.
    - rm_-rf_slash 11y ago
      True, although the stealth capabilities are nowhere near the F-22 or F-35. Still, it helps to pack more than a couple missiles or bombs when the enemy inevitably discovers that you've dropped in to say hello.
    - Mikeb85 11y ago
      The EA-18 is probably the most interesting aircraft the air force has. At the rate technology is advancing, electronic warfare seems like it'll win over passive stealth.
      - engi_nerd 11y ago
        The Navy operates the EA-18G, not the Air Force.
  - imglorp 11y ago
    Fairchild is long gone, having been bought years ago and lines closed and moved. You'd be starting almost at scratch.
  - Recurecur 11y ago
    The F-15 Silent Eagle is also interesting, and is a far better fighter than the F-18. I really like the B-1R concept, too bad it'll never be built.
- spatulan 11y ago
  The A-10 is a disaster that needs to be retired. It's completely incapable of facing modern air defences, so is useless against enemies like China or Russia, and has operating costs 30 times higher than a dedicated COIN aircraft like the Super Tucano, meaning you have a lot fewer aircraft supporting your troops than you should have, or end up spending way too much money.
gherkin0 11y ago
> The rational of having just the one type of plane makes sense if you're Ryanair (who only fly Boeing 737-800). Of cause they only have ONE role for all their plane. Having just one type plane take the role of multiple others just gives you all of the complications of each of these plane all rolled up in one, plus a series of its own unique issues. In addition to that, they didn't manage to build just one type of plane. They actually built three types that look sort of the same and share a few parts (apparently not enough to actually save money). > At some point someone is going to dictate that the F-35 is done, true or not. The US will fly them for a short while and the other nations, that can't afford new planes for another 20-30 years will be stuck with this thing. I like the idea of building updated versions of existing designs as a hedge against the failure of this project. It looks like some of those are getting more interest: http://www.popularmechanics.com/military/weapons/news/a18052/boeings-stealthy-f-15-silent-eagle-israel/ http://www.popularmechanics.com/military/weapons/news/a18052...
jcadam 11y ago
Several years ago I worked on one Lockheed project (Not the F-35, but part of another large defense program) as a subcontractor (at a Lockheed facility). The most dated development practices imaginable. Waterfall, no source control (they brought in subversion near the very end of the project in order to meet govt requirements but only the 'architect' was allowed to commit anything (or even to use the 'svn' command)), SLOC used to measure "progress", even the office furniture looked 70s vintage. The project was one giant clusterf*k from start to finish, and I swore an oath never to work on anything involving "Lockheed" ever again.
- michael_h 11y ago
  I had my software installed in a Lockheed system. It was incredibly surreal - the amount of time spent assigning blame for problems was orders of magnitude more than the time spent solving problems. One of their sysads, in a meeting with the Important People, blamed my system for extremely high network and disk load. He showed graphs that spiked and plateaued a few minutes after system load, pointed at the spike, then pointed at my face and said, "It's his software that's making this whole thing slow. Get him out of here and we'll build something that actually works." I got unparalleled satisfaction when I got to say, "I haven't installed my software yet."
alkonaut 11y ago
Multi-role aircraft aren't a bad idea, so long as each role somewhat overlaps the other in requirements. Many well designed aircraft do 2-3 roles very well (e.g recon/attack/fighter roles). If you mix in another role that doest't match the hardware requirements of the first role that's when you are asking for problems. CAS, SToVL are such examples that won't blend with the first 3 without massive compromises. Full stealth is also a requirement that is rarely absolutely necessary, but adds to the cost and lowers performance of every other aspect of the plane. When these compromises had been discovered the project should have been scrapped, the SToVL elements thrown out and a separate plane should have been procured for that role. Lots of components could be shared, just not the airframe. I too believe the project is very near the edge now. It will survive as there is too much prestige invested, but if a few big players like Canada or Australia decide to get zero F-35's the unit development cost would skyrocket which would force other buyers to to back out. If was LM I'd be sending huge lobbying checks to Canada right now.

mannykannot 11y ago

I have a suspicion that the project (not just the software) came off the rails when 'and VTOL' was put in the requirements.

p_eter_p 11y ago
I've seen it mentioned in at least one article that the VTOL requirement is driven by the USMC's institutional memory of Guadalcanal. They don't want to get left on the beach without air support again, so they insist on having their own aircraft with VTOL capability. How valid a concern that is is..up for debate, in the era of nuclear supercarriers and jet aircraft.
- ZanyProgrammer 11y ago
  Especially with American global air superiority, and the fact that the Marines haven't done opposed amphibious landings since...Inchon?
  - dmos62 11y ago
    Global air superiority? Not a geo-powers geek, but sounds implausible.
- gaius 11y ago
  ... And the Royal Navy don't forget.

SFjulie1 11y ago

And russians still have planes with hydraulic commands from the cold war (50 yo) that outperform F35, F16, F15, F14, B2 :) They even flew near UK exactly one year ago with a turbine powered bomber from the 50s and a nuke inside and no planes intercepted it during its 5 hours flight over "modern air farce" doted countries. Notably the UK are lacking interceptors because they planned to replace theirs with the F35.... and the F35 is not there. Does not it sound ridiculous? And the stealth technology are nowhere to be stealth for long. It leaves heat signature, signal distortions. And why do US stealth planes have so much sharp angles when we all know it is bad for being stealth? Because their software don't let them do curbs... To avoid software bugs remove software from where it is not useful

talmand 11y ago
I don't understand the Russian bomber part, did they fail to detect the bomber or failed to intercept due to not having planes? Are you suggesting that the UK totally ditched all of their current interceptors and have their pilots sitting around waiting for the F-35 to show up? Are you saying the existing stealth planes are a money-making scam because we all know they built them wrong? The software doesn't allow... "curves"? Why doesn't it allow curves?
- logfromblammo 11y ago
  Of course it allows curves! It's just that they can't have curvature less than a great circle....~ Due to the way the military produces software specifications, it is very likely that the unbroken chain of subject-matter-experts from the days of ochre pictograms on cave walls to today have ensured that the software program functions exactly the same as drawing lines on charts by hand. Since human navigators did not initially possess the capability to continuously recalculate courses, bearings, and headings, as would be required in a curved flight path, the ground-based and flight computers don't do it now. So planes still keep a fixed heading between a set of points, and any possible roundness is encoded into the turn data or a specialized, predefined pattern.
  - SFjulie1 11y ago