If the last one is true, devs take note. If FAA can rely on slow, small storage for safety critical applications, you probably can reduce the resource requirements for your app and improve your customers’ experience and extend the life of their hardware.
There’s been an incredible expansion of computer hardware and networking in every dimension since the heyday of the floppy. And the floppy itself was a huge step up from technology of even a few years prior.
Your comment could have been made in any year since 1950, and it would have been both correct and irrelevant outside the areas most resistant to hardware upgrades. FAA is conservative, and should be, but the rest of us have supercomputers in our pockets.
Have software developers abused the hardware capabilities they have been granted? Of course they have. I’d love to see some care put into that. But beyond the basics, hand-twiddled hyperoptimizations are now an exercise for craftspeople. Or…AI?
The issue is they don't feel like supercomputers, because of how much they've been taken for granted.
My computer and phone today feel no faster or better to use than they did 20 years ago. They might be vastly more capable for niche applications, but for the day-to-day stuff it seems like we keep buying new hardware, so we can run the latest software at the same old speeds, mostly for security updates.
I have an M1 MBP... a bit old, but still does everything I need. When I click Calendar in the dock (when it is already open, but just haven't used it in a while), it can take 3-5 seconds to come to the front as the active app. It feels broken. This wait wasn't a thing 5 years ago on this hardware, nor was it an issue on a PPC Mac from 25 years ago. Yet now its' a problem... why? So the UI can calculate the refraction of the background image in the liquid glass? Should I be spending $2,500 on a new MBP just to switch more quickly from my browser to the Calendar? It doesn't make much sense.
Unfortunately web search has gotten so bad that it's hard to find past information about it.
Another article with some more details: https://www.pcmag.com/news/the-faa-is-still-using-floppy-dis...
I know they are saying IDS-4 is a "1980s" system, but it appears IDS-4 was actually developed in the early 1990s; it uses MS-DOS and 3.5 inch floppies: https://sysatl.com/systems-atlanta-inc-awarded-faa-ids4-sust...
Workers of the day were used to paper first routines; PCs on every desk were not common until the 90s when home computers, like phones in 2026, hit a fast enough for most tasks in an office constant upgrades are not necessary.
There was a lot of ducks to align back then.
While we see improvements in hardware and software today, it's NOTHING like the late 80's and 90's. Our first home pc was out of date in less than a year (thank you for that recommendation, Tiger Direct). As a child, I couldn't run games that came out more than about 10-12 months after we upgraded our machine. It really was stunning to be a part of.
And now my home pc is about 4 years old, and replaced a machine that was 9. I plan to run this one for at least another 6 or 7 years unless something amazing happens that nobody sees coming.
What does this even… mean?
https://www.oig.dot.gov/sites/default/files/library-items/FA...
I can't find information about how the disks are used specifically, how often they are read, written, what data they are used for, if they can be replaced by floppy emulators, etc.
12 billion for a new system that might not actually be an improvement is a lot of money.
Floppies aren't a bad technology, but they do have limits.
I doubt that SD cards will last that long.
I never had a bad sector on 5.25-inch Double-sided, Double-Density (360K) diskettes, but I did all the time with 3.5-inch HD ones (1.44MB). I'm seeing some references to 8-inch and some to 3.5-inch disks used in the FAA, but I'm not sure what source has reliable information about that.
The weather records have to be written to the floppy from an online device at some point in the pipeline.
The workflow has likely been compromised at some time, but due to lack of data integrity checks, it was never detected.
If we give them more money, this is sure to improve! Monopolies and all ...
They could've replaced floppies multiple times, with all the certifications and hardware replacements and new procedures and regulations, and we'd still be using (a different, but still) "obsolete" medium for this now.
But the cost of switching -- mainly thoroughly certifying the replacement system, retraining, and the risks of an error happening during one of those processes causing a catastrophe -- that's real.
There's no point in upgrading the storage mechanisms of arbitrary systems, or completely redoing the whole system from scratch on new hardware which would be the case here, unless it yields measurable improvements in fitness for purpose to balance against the switching costs.
Sitting here now in 2026 I have a box containing almost 3000 3.5" FHD floppy disks.
Each about 1.544 Megabytes, unless you cut the notch.
I literally dumpster dived this resource after being invited in to do so at a local electronics shop closure, some years ago now, on the proviso that I did not leave empty handed.
There are about 12 devices in my living room that can use these things, but I'll never use more than 30.
Some day, I'll see the last one go. That'll be weird.
Equally likely: they'll see YOU go. Pape... uhm I mean floppies are patient.
I assume they just mean _old_ computers here. Unless they have some specialized analog air traffic computers? That would be pretty cool.
I have heard experienced pilots lamenting that we're still using voice on a scratchy radio, with garbled audio and people accidentally talking over other transmissions both being constant problems, for essentially all vital comms.
While I have no problem with keeping that as a fallback, there's zero reason why we shouldn't have, at least for airliners that have two pilots, a text-based communication system. We can still convert the text to speech in the cockpit to keep awareness the same, but we could play overlapping transmissions one after the other. And think of the other advantages. People wouldn't get callsigns wrong, could be alerted specially when their callsign was being talked to...
It's complicated because we will need to cooperate with all countries, but it's not like we don't have ICAO to work within.
Every 5 years or so, a different publication will write an in-depth story about how difficult they are to replace, because nothing works nearly as well, and the controllers go back to using paper every time a new system is trialed. I think I've even read about them in a cognitive psychology textbook at uni.
There is something to be said for knowing how to deal with critical systems when the automation breaks.
A few decades later, similar story, he came back to work part time, when desktop publishing was all the rage.
Welp, of course the computers went down.
So they fired up the old hot metal typesetting machines and he manually set all 100,000 words of that Newspaper by hand, and then they went to print using 19th century steam powered print lines.
Except, that story is total bullshit that I made up. Because "the computer went down" is something we can mitigate much easier than doing everything by hand.
Sometimes the old technology actually is objectively worse, inefficient, and obsolete. Like floppy disks, or manually having humans close the books at a bank.
In that system you can tell what is going on by looking at in a glance and works with no power. Sometimes simple is a great answer.
However, I can think of one reason the 'no power' argument fails. They do a lot by radio. That needs power. Also you would need to retrain everyone. Oh and you probably would get locked into one company that wants to control the whole software stack and sell tons of support.
When my home was built in 1999, I poured over all the home automation systems. I opted for conventional wiring. But I did wire every room (myself) with cat5 cable and RG6 coax, in a star configuration. It has served me very well, no problems at all.
I don't use the RG6 anymore, as my TV service comes over the intertoobs now. The cat5 isn't supposed to work with gigabit equipment, but it works fine. Wifi has improved by leaps and bounds, but I still rely on hardwired ethernet, as it is stupid reliable.
I think the best idea is to install conduit as the house is built, so that outdated wires can be replaced with newer wires by just tying the end of the old wire to the end of the new wire and pulling.
Just be sure to inspect the house as it is being built, because I think one user on HN wrote that conduit was required by law, but when he went to swap the wires, he couldn't pull them - he discovered there was only pipe in the visible terminations, not the whole route.
the digital version would require 10+ endpoints with 1000x points of failure.
Do they seriously not have a backup connection? I would have expected them to have a backup stelite or maybe even a 4G/5G mobile data system in place.
If I can use my phone as a Hotspot when Comcast is doing some service so I can work, I don't understand what idiot designed an airport without backup connections.
That was the backup connection. The primary failed, and concurrently the backup had been severed by a construction crew.
No one notices the thousands of times the redundant links works just fine. We only notice when the primary has an outage and a construction crew cuts a fiber line at the same time.
A more useful comment might have been to say something about the depth of the backup, having only one backup connection not being enough or something.
I have literally spraypainted the pavement and wall - "One fibre goes here, and down on the corner there, the other fibre goes there" - just to ensure a good separation of routes. I want the pinch point to be within the rack that the routers sit in.
I don't get why an airport can't do similar.
So they didnt have a redundant connection. They had 1 live one that was severed.
If it was just two modern computers and transferring data between them, the floppy disk would have been dropped in a heartbeat. But it's the old computers still in the critical path that are holding everything back. And those are really hard to maintain. Last year we phased out a system that was dependent on computers from the mid-1990s, getting replacement parts was practically impossible, they mostly just don't exist. You'd buy old systems off ebay and hope that they had enough working parts to scavenge, and the right working parts.
EDIT: Light editing because I had some extra words.
There was a time when the company I was working for was selling service contracts for DEC TZ30 tape drives to DEC itself because their own internal service organization was unwilling or unable to support them.
The application was software distribution, most likely to customers they were contractually obligated to support, and no newer drives were able to write the required tape format, so upgrading to newer drives wasn't an option.
Here's a software analogy - required support for very old OS versions often interferes with the scope, simplicity, and security of new feature work, and even if the work is still feasible at that point it is slowed down greatly, eating up attention and time that could go towards addressing other important issues. That accumulates over time, and cumulative loss of improvements that could have otherwise been made has a real effect in unprevented disasters and wasted controller availability.
It's better to test a replacement while you still have the time, than to risk rushing something else.
The fact that it has gone x days without an incident doesnt upgrade or downgrade its reliability. The engineering and processes are function in an equation that calculates mean time to failure. Life-critical processes have numbers placed to this... like a failure every 50 years or 100 years... or 100000 years, and you work towards solving that.
So they make their systems run on the internet now?
What could possibly go wrong?
This one, where the tour stops are ATC data buildings, where they only stop in to swap out floppy disks for USB drives.
They are signalling Trumps involvement so I will assume it must be another vanity project he wants to stamp his name on. I look forward to hundreds of grounded planes due to cloud based weather reporting not being available because of a borked server update or misconfigured wifi adapter.
I have never heard of a sploit that roots your machine from inserting a floppy.
Just... modernize the thing already. thats your job. It's not impressive that you did this. You're supposed to do this. You dont get extra kudos for doing the thing you signed up to do...
and dont forget to put Trump's name at the beginning. Or he might get cranky
1. The party leader is a fragile, dying narcissist who is desperate for his name to be attached to anything other than the misery he's inflicted upon the world.
2. Their _actual_ historic achievements are all war crimes, human rights violations, and policy changes that would terrify their average supporter if they actually understood the long-term consequences.
The inability to keep the air traffic control system effective and maintainable is far more embarrassing than the inability to update the IRS software or build a high-speed rail.
Engineers of today aren't better than engineers of yesterday, they just have access to new technology (and new footguns).
I'd rather have a well-designed legacy system that does the job well, than an incompetently designed "modernized" system that doesn't.
Our ability to wield technology has gone up, while our ability to engineer and understand that technology (as a whole) has gone down.
Age should be the most suspect of arguments against the design of a system. Enumerate specifically its flaws and not just that it is old. System design doesn't wear out, it is abstract, like mathematics.
Most of those were extremely optimized for their specific purposes - for instance, screens were carefully designed to fit all the necessary information to complete a task without any scrolling. They had extremely low latency despite all the hardware being potato-grade compared to what we have today.
When latency existed, simple things like the keyboard buffer allowed the skilled user to get ahead of the machine and move onto another task, while the processing caught up.
Today, it's all a webapp with a ton of React, probably 5,000 npm modules, and you get to watch skeletal loading rectangles pulsating after every click (or branded animations on the "customer-facing" flavor). Instead of having muscle memory of hands on a keyboard (usually a specialized one which gave a dedicated key to every important function), the user must move their arm to find touch targets placed at arbitrary places across a 24" touchscreen, and often hunting for functions inside ••• menus.
Obviously, no usability testing has been done on any of these. Or if it has, the only thing tested was "Can a 17-year-old with zero skills ploddingly complete a transaction (the happy path only!), after only 15 minutes of training?"
So yeah, I never listen to any "newer = better" takes.
Limited to keyboard and 24x80 screen, but yes, the apps were lean & responsive -- sub-second response time, and keyboard muscle memory. And some shops I worked in capped your DB calls to <25 per transaction. In stark contrast to some of the Laravel apps I've worked in recent years where the SQL logs showed me 300-400+ DB calls on a given request :(
Instead, the United States government should have continued the policies and values that led to Next Gen, ERAM, STARS, and AAS.
I would urge you not to consider this modernization from the perspective you may be familiar, as someone who may work on the bleeding edge and sometimes sees a lack of caution from engineers pushing for change.
The FAAs safety risk management is built around a rigorous process, but that is also a massive slowdown to reasonable progress.
One of the eng I worked with was 80+ years old and just didnt want to retire. He did not understand modern systems at all and the risks we were looking at were not (imo) realistically balanced.
I get the idea that the current administration is pushing safety to the side, but there is another risk where safety becomes a blocker to reasonable change.
When I left that work (more than 15 years ago) it was clearly a blocker.
Magnetic degradation, wear from read/write heads, and environmental sensitivity mean disks have a natural expiration date and it's not long. And they are also easy to corrupt from all kinds of magnetic sources around us.
I get other stuff, even DOS like system. Safety critical software/hardware that was battle tested makes sense. Not floppy. Sometime it's just bureaucracy.
They absolutely mog the cancerous garbage that is NAND flash memory in terms of data retention. The fact that you can pick an 1980s floppy from the junk room and read it is telling. Meanwhile your average TLC SSD will probably get wiped in 3 years if kept in the drawer, and a QLC will get wiped in 1 year.
Reliability wise floppies are also better: flash memory leaks electrons naturally and gets shredded by radiation events, with temperaure accelerating these processes. Meanwhile magnetic disks are not altered by any normal magnetic field you could produce at home.
More info here: https://www.urbandictionary.com/define.php?term=mog
YMMV. I’m sure a lot of people were shoving these things into backpacks, and this was post-2000, so the quality of disks for sale was poorer than it was in the 90s and 80s.
I’ve also (since then, for myself) repaired drives. The drives fail because they’re mechanical and have motors, and anything that moves or has a motor in it will need some kind of service. Servicing a floppy drive requires specialized equipment which is hard to find (or requires a lot of patience—it’s possible to repair a floppy drive with just a screwdriver, but it’s not fun; guess how I know?)
This is probably my "I'm getting old" moment. Good lord.
That they are still used in a critical infra item is a bit surprising. Getting new stock would be a major hassle.
https://github.com/spullara/pcterm128
Without having details on the specific floppy disks in question there’s no way you can authoritatively say they are unreliable. They absolutely produced floppy disks that were far more reliable than the ones sourced from the cheapest vendor possible some of us used in school back in the day. Ignoring the fact I’m guessing the disks in question aren’t having to survive life in a students backpack.
Is it because "long-term archival" strategies aren't updated frequently?
You really have no idea what you are talking about. It's normal for floppies, especially 3.5" to last decades just sitting on a shelf. I've got 5.25" floppies from 1984, and they boot just fine on my old apple //e.
42 years in my parent's dank midwest basement.
If this is the case, why was an airport knocked offline due to a cut fiber optic? A safety focused system would have a backup system say satellite or mobile 4G/5G. But they didn't have such.
This isn't about safety. Its about bureaucracy wrapped up as safety.
There were two independent systems.
You can argue thats not a sufficient third redundancy, I don't have enough context to throw an opinion out there on the safety of this particular infrastructure project.
I am confident that the fallback process was considered and kept in place as the tertiary backup.
Excellent point though. It took me some thinking to put myself back in SRM mindset and what they'd have considered here.
Not really. Triplex redundancy is based on a failure model where failures can produce arbitrary output. Three systems allows majority voting (often, median select) to provide correct output in the presence of a single arbitrary failure. It allows detection of double-fail-silent failures and uncorrelated double failures, but that's not really the point, since continued operation under these conditions is at risk, and failing nonoperational isn't often acceptable for these systems. (You'll also see quadruplex systems that use the same voting mechanism to detect and disable failed channels, allowing handling two arbitrary non-simultaneous failures, for systems that need a longer window of continued operation after failure.)
If you change the failure modes, you change the architecture. These days you often see dual-dual systems where each pair of systems cross-checks and disables itself on mismatch, converting a single arbitrary failure to a single silent failure; which then allows downstream systems to accept inputs from two such fail-silent systems and trust either that is providing output, removing the need for a median select.
What’s a better way? We still don’t have state-regulation/certification for software in safety critical systems, which terrifies me when I think about it.
The pilot no longer had flight controls. He desperately tried to control the airplane using variable thrust, but it just wasn't good enough, and crashed into a mountainside.
The fix was to add automatic valves in the lines so loss of pressure wouldn't bleed out the whole system.
In a perfect world, we would move at a steady pace, as to keep up with the world but without rushing, and people would be accountable without a heavy bureaucratic processes.
Unfortunately, that's not how it works in practice, at least not for long, so between the two, we usually favor the slow and bureaucratic process when safety and reliability matter.
I can understand not changing a system for change sake, but the floppy disk has been obsolete for nearly 20 years now. It's been hard to find a floppy disk reader for the last 10 years. Going so slow that you can't respond to that change is far FAR too slow.
As you pointed out, you can't find floppy discs or disc manufacturers any more. I do remember one of the VP's of this company saying as long as they keep teaching Fortran to college engineers, they have no reason to upgrade yet. That got a few raised eyebrows from our team. We relented and upgraded what they were comfortable with - but man oh man it was quite the eye opener for me as a young dev how the thought processes of the people running these companies really are.
As long as those manufacturing systems were air-gapped and working, I can understand being conservative with changes. Of course at some point you're at risk of hardware failures too, for things that are likely NLA.
Yeah, that's really the biggest issue with these systems. Were I one of these manufacturers I'd invest in getting the current software running in a VM specifically to make it easier to transition to new hardware. The last thing they'll want to deal with is that their software was written for Sun Spark and that no longer exists.
That said, hard to go wrong porting over the x86 at this point.
Some of the signal control infrastructure for the Melbourne rail network was ported over to an emulated PDP-11 System for this reason. This was previously discussed on HN here: https://news.ycombinator.com/item?id=42514553
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/lin...
https://www.youtube.com/watch?v=GVHEFiLQ3GU
These airplanes are flying all over the world, the inspections need ultrasonic data, and the FAA has to validate that all of them are done. Even identifying this issue required lots of bureaucracy.
Multiply that by the number of airplanes and airplane models and it's easy to see why they move so slowly: when you are putting millions of people in the air each day, enforcing absurdly specific rules with an iron fist requires a monster of a bureaucracy.
But creating a bureaucracy doesn't mean you've created a rigorous process to ensure safety.
'Move fast and break things' is fine when the result of failure is someone failing to make a billion dollars at their IPO. It's not acceptable in the world where failure means dams break, reactors go supercritical, planes collide, or other similarly apocalyptic scenarios.
Aviation involves constant unpredictable human inputs and physical states. Hundreds of vehicles, not installations designed to have as few moving parts as possible. It is not yet solved, the status quo merely has a societally tolerable death rate. There can be new sensor hardware, more safety redundancies, and info synthesis/notification interfaces that are currently blocked by old hardware bottlenecks. Controller and pilot time and attention is limited, and it can be wasted less.
And inversely... when pilots leave too much to automation, their skills may degrade to the point that by the time you actually need a capable human crew because something has gone terribly wrong, the capability of that crew has degraded too much, and the FAA has warned about this in the past [1].
[1] https://medium.com/faa/the-dangers-of-overreliance-on-automa...
Also, there's a big difference between a small towered airport in a more rural area, and a major international airport with 10x or more the commercial traffic. So there's probably some difference of buildout depending on the facility usage.
Quick edit: looks like there were two fiber cut incidents recently, one at an airport last month, and the TRACON issue that happened 1.5 months ago!
Backhoes can disrupt connections, but so can failures to pay bills.
Systems like the radar in question often cannot work with broadband and satellite links.
Engineering will not save you from incompetent governance.
In one role, I worked on writing software for drones.
That very rigorous process was very reassurring to me. Reasonable progress was absolutely occurring.
> One of the eng I worked with was 80+ years old and just didnt want to retire. He did not understand modern systems at all and the risks we were looking at were not (imo) realistically balanced.
What kind of risks were you looking at with what kinds of balances, if you're able to share?
The SRM board knew the legacy system but much less the modern one. Its hard to compare safety when you only clearly understand one systems safeguards.
I appreciate the process being there but at the time it felt much more like (some) people protecting their jobs than the public.
The risks I felt were overblown (iirc) were around likelihood of catastrophic failure in the system.
I actually loved the overall process and it informed how I look at software, it was the application of the process that bothered me.
Insomuch as people protecting their jobs: that's always a thing. But I honestly felt that people wanted to protect their jobs because they wanted to protect the public. Some people definitely wanted to keep their work and personal life separate with solid 9-to-5 boundaries. That's perfectly OK too. It was helpful to be inclusive of them even if their time was already filled.
I worked with numerous people who didn't understand all of the components or systems. We took it upon ourselves to be very open to questions and to thoroughly explain how each thing works. I've found if someone had good understanding of one component but didn't understand another component, it was very useful to schedule time with them to walk through the components they didn't understand to try to alleviate their concerns. They might not know what questions to ask, so I would instead ask questions about their well-known components to try to understand and correlate to newer components. Perhaps their concerns were even justified but they didn't know how to articulate them!
Iirc there were 1 or 2 people who could hold the process up and were not open to persuasion.
I was also young and less equipped to handle that.
I worked in a few other spaces and liked a slightly more balanced approach.
Silicon Valley style approach is not what id strive for at the FAA. I dont think that is the extreme they have swung to either.
If you don't make a change and then there's a plane crash that's not your problem.
Basically when accident rates are already very low, even a small amount of transitional risk can add up to offset years of marginal improvements. So the burden of proof rests heavily on the proposer of change to a system that already is working reliably.
Can you recall anything that that engineer said or did which turned out to be credible? Or which you perhaps learned from?
But this kind of precaution is quite necessary in the development of AI systems. Current regulations have turned a blind eye to accelerate software and AI progress. Now we have security bugs and misalignments, and we will have AI misalignments as well.
Knowing how computing landscape has changed in SV, I can't imagine modern day SV helping fix FAA systems. Alternatively we could go with multiple incarnations of the same system, being battle tested and proven with independent concurrences, alongside current FAA system. Yeah, it is a boatload of money, and probably wasting 90 percent of it in redundancy.
Version control was always something that was adopted more by process-heavy organizations until the popularization of GitHub. Before that, it was common for the move-fast type of developers to eschew version control.
Earlier in my career, people didn't bother with the chore of using version control unless they were working on an important project that demanded it.
Some slight rigor shows you are not just doing whatever. I found most projects once you scale past 2 or 3 people and your 3rd release you basically are forced into some sort source control. Or it is a game of 'what file did they break today and I have to spend 3 hours fixing before I can add my feature'. Then you end up with merge meetings and very weird things happening.
Usually missing source control in an older org means there is a large swath of tech debt. Watch for the weeds. You will find other things very quickly.
And then there are the safety issues.
Can you provide any examples you've come across where the current administration has pushed safety to the side with respect to the FAA? I'm curious to read about those. I know there have been some changes to regulations and there are sometimes disagreements around those but I'm not sure if the ones I've seen (which aren't comprehensive) would be considered pushing safety to the side.
we can look at this more holistically than a specific IT project. like in china, which is not a place i want to live, it is still valid to say, their process of steamrolling past local interests means that airport expansion and transit to new airports can be built quickly, which helps meet demand for air travel better than working at the absolute limits of safety and efficiency tradeoffs.
local control in the US infests everything. it causes airports to be too small, guilds to grow too slowly, it opposes infrastructure, it makes politicians too vulnerable and too weak to make changes that upset minor but powerful real estate interests.
so in one POV, safety is a blocker to reasonable change, but everywhere else in the world, they just increase capacity, and the huge stresses on safety limits are reduced.