Just for the context, I am employed by a multi-billion company (which has more than 100k people)

Here, I'm in charge of some low level infrastructure components (the kind on which absolutely everything rely on, 5sec of downtime = 5sec of everything is down)

On one of my scope, I've inherited from a 15 years-old junkyard

The kind with a yearly support

The kind that costs millions

The kind that is so complex, that has seen so less evolutions other the years that nobody knows it anymore (even the people who were there 15y ago)

The kind that slows everybody else because it cannot meet other teams' needs

Long story short, I've got a flamethrower and we are purging everything

Management is happy, customers are happy too, my mates also enjoy working with sane tech (and not braindamaged shit)

Yes, this is the key distinction: old software that works vs old software that sucks.

The one that sucks was a so-so compromise back in the day, and became a worse and worse compromise as better solutions became possible. It's holding the users back, and is a source of regular headaches. Users are happy to replace it, even at the cost of a disruption. Replacing it costs you but not replacing it also costs you.

The one that works just works now, but used to, too. Its users are fine with it, feel no headache, and loathe the idea to replace it. Replacing it is usually costly mistake.

Or it doesn't. Because "software as an organic thing" like all analogies is an analogy, not truth. Systems can sit there and run happily for a decade performing the needed function in exactly the way that is needed with no 'rot'. And then maybe the environment changes and you decide to replace it with something new because you decide the time is right. Doesn't always happen. Maybe not even the majority of the time. But in my experience running high-uptime systems over multiple decades it happens. Not having somebody outside forcing you to change because it suits their philosophy or profit strategy is preferrable.

My guess is that most stuff is part of a bigger whole, and so it rots (unless it is adapted to that ever-changing whole)

Of course, you can have stuff running is constraint environment

Or more likely the 'whole' accesses the stable bit through some interface. The stable bit can happily keep doing it's job via the interface and the whole can change however it likes knowing that for that particular tasks (which hasn't changed) it can just call the interface.

Sounds like that is a different issue. I prefer to avoid spending a few weeks migrating software that i understand and support to a new OS when i dont have to. Some of it is 30 years old, but it has had all the bugs worked out.

This is exactly the same thing: OS is nothing but software. And, in this specific case, we are talking about appliance-like stuff, where the OS and the actual workloads are bundled together and sold by a third party

> I am employed by a multi-billion company (which has more than 100k people)

In my personal experience, this could mean that you're really good or that you're completely incompetent and unaware that computers need to be plugged to a power outlet to function.

Having started my IT career in manufacturing this 100%. We didn’t have a choice in some sometimes. Our support contracts would say Windows XP is the supported OS. We had lines that ran on DOS 5 because it would’ve been several million in hardware and software costs to replace and then not counting downtime of the line and would the new stuff even be compatible with the PLCs and other items.

> .. they don't want to waste time with an OS upgrade because the cost of the downtime is too high and none of the software they use is going to utilize any of the newly available features

Oopsie you got pwned and now your database or factory floor is down for weeks. Recovery is going to require specialists and costs will be 10 times what an upgrade would have cost with controlled downtime.

Not at all, it depends on the level of public exposure of the service.

In a factory, access is the primary barrier.

It's like an onion, outer surface has to be protected very well, but as you get deeper in the zone where less and less services have access then the risk / urgency is usually lowered.

Many large companies are consciously running with security issues (even Cloudflare, Meta, etc).

Yes, on the paper it's better to upgrade, in the real world, it's always about assessing the risk/benefits balance.

Sometimes updates can bring new vulnerabilities (e.g. if you upgrade from Windows 2000 to the "better and safer" Windows 11).

In your example, you have the guarantee to down the factory floor (for an unknown amount of time, what if PostgreSQL does not reboot as expected, or crashes during runtime in the updated version).

This is essentially an (hopefully temporary) self-inflicted DoS.

Versus an almost non-existent risk if the machine is well isolated, or even better, air-gapped.

> Versus an almost non-existent risk if the machine is well isolated, or even better, air-gapped.

Anyone else remember stuxnet?

I don't understand this comment. What exactly do you think LTS is?

Kernel live patching takes care of everything.

There's a difference between old software and old OS. Unless you've got new hardware, chances are you never really need a new OS.

I can't upvote this hard enough. It's nice to know there's at least one other person who feels this way out there.

Also, this is the most compelling reason I've seen so far to pay a subscription. For any business that merely relies upon software as an operations tool, it's far more valuable business-wise to have stuff that works adequately and is secure, than stuff that is new and fancy.

Getting security patches without having feature creep trojan-horsed into releases is exactly what I need!

I'm reminded of the services that will rebuild ancient electric motors to the exact spec so they can go back on the production line like nothing happened. For big manufacturing operations, it's not even worth the risk of replacing with aa new motor.