The outages, attacks, and meltdowns will continue until business leadership stops thinking they know more about IT than their actual IT Engineers, and start supporting them accordingly.
> “This is exactly what happens when billion-dollar companies refuse to modernize,” one aviation analyst tweeted. “You wouldn’t trust a 30-year-old car to drive cross-country. Why are we trusting it to fly planes?”
While i sympathize, the world does rely on high quality 30+ year old software. I think it's time, as an industry, to stop seeing software as disposable and start designing for longevity.
There are several regulatory, financial, and other business processes that must occur for commercial jets to fly. They can't just put fuel, pax, luggage, and crew on-board and send them on their way without recording data in various systems.
It's no longer possible to not be a nightmare. Companies no longer employ devs for life to keep the system functional and up-to-date, and (most probably) no devs would stay in such a role for their entire career.
When nobody knows how the system works, it can only be a nightmare: they keep it barely functional with marginal patches, accumulating debt until it finally crashes in such a way that it can't be fixed, then a new team comes in, designs and builds a new system, patches some major flaws, and then leave, and the situation repeats for another 10 years.
New software system can be even less reliable. The old/new distinction is not really important, what important is if reliability is a priority whoever owns the system. If there is a sustained effort to find and fix bug and other issues.
In the old days, airlines and power grids were incentivized by regulations to spend money to build extra capacity into their systems to increase reliability in times of crisis.
Now they're all deregulated and they are highly incentivized by their shareholders not to overbuild capacity because it costs money and provides zero return on investment.
The upshot is that these industries operate with zero excess capacity. They work fine when everything goes as planned, but the minute something breaks or a storm happens, everything cascades into a steaming pile of shit.
In the case of the airlines, competitive market forces previously provided some incentive to invest in reliability but so many airline mergers have been allowed to happen that there's effectively no airline competition in the US any more.
Plenty of 30-year old cars can and do drive across the country. The analogy makes me wonder: in 30 years will 30-year-old cars be able to drive across the country? I suspect 60-year-old cars will still be able to make that trip, but will the current crop of cars with all of the electronics and onboard computers? I suspect they won't, but that's an opinion formed by bias and not facts.
This country bailed out this industry multiple times in the past 20-25 years. Yet they continue to ignore investment in their own operations, and instead use it towards executive bonuses and stock buybacks.
Anybody else tired of this broken hypercapitalistic system that only serves to benefit the few? Enshittification is a symptom of a much more disgusting disease.
Is United the one suing Microsoft and everyone and pointing fingers everywhere else about how CrowdStrike broke them worse than the other airlines last year?
> “You wouldn’t trust a 30-year-old car to drive cross-country. Why are we trusting it to fly planes?”
Because software doesn't show wear. On the contrary - a system that's been running stable for 30 years probably has fewer bugs than one that's been running since last week. They are running old software on new computers.
17 comments
[ 6.6 ms ] story [ 37.8 ms ] threadWhile i sympathize, the world does rely on high quality 30+ year old software. I think it's time, as an industry, to stop seeing software as disposable and start designing for longevity.
It's no longer possible to not be a nightmare. Companies no longer employ devs for life to keep the system functional and up-to-date, and (most probably) no devs would stay in such a role for their entire career.
When nobody knows how the system works, it can only be a nightmare: they keep it barely functional with marginal patches, accumulating debt until it finally crashes in such a way that it can't be fixed, then a new team comes in, designs and builds a new system, patches some major flaws, and then leave, and the situation repeats for another 10 years.
Just wait till I find out how old this SABRE is!
https://en.m.wikipedia.org/wiki/Sabre_(travel_reservation_sy...
There are plenty of real sources for this story:
https://www.wsj.com/business/airlines/united-airlines-halts-...
https://www.nytimes.com/2025/08/06/business/united-airlines-...
https://www.reuters.com/world/us/united-airlines-flights-res...
Why would a 30-year-old car be flying a plane? That makes no sense.
Agreed.
Frequently the actual issue related to these big system outages is cost-cutting/outsourcing/lack of maintenance.
Now they're all deregulated and they are highly incentivized by their shareholders not to overbuild capacity because it costs money and provides zero return on investment.
The upshot is that these industries operate with zero excess capacity. They work fine when everything goes as planned, but the minute something breaks or a storm happens, everything cascades into a steaming pile of shit.
In the case of the airlines, competitive market forces previously provided some incentive to invest in reliability but so many airline mergers have been allowed to happen that there's effectively no airline competition in the US any more.
Anybody else tired of this broken hypercapitalistic system that only serves to benefit the few? Enshittification is a symptom of a much more disgusting disease.
Because software doesn't show wear. On the contrary - a system that's been running stable for 30 years probably has fewer bugs than one that's been running since last week. They are running old software on new computers.