There are generally three points that need to be considered before upgrading an older system:
1. Is the older system still fulfilling the needs? (often jobs change over time and some systems are rigid thus require users to perform additional steps in the modern era to continue to make use of the older system).
2. Is support still available for the older system (eg if it breaks, is hardware available to replace it? Are engineers still available to maintain it? etc)
3. Will the new proposed system fix the aforementioned issues?
Often you'll find that the first two points are still or that the 3rd point isn't true. Sadly it's all too common that the reason someone pushes to upgrade older systems is either down to agism ("it's old. Old is bad!") or because someone needs to justify their own job / bulk out their CV ("I was responsible for migrating Company xyz from product a to product b.").
With the greatest of respect to yourself, as you get older I think you too will start to notice just how often upgrades go wrong and how some businesses would have been better off sticking with the old system. So there does need to be a carefully considered pragmatism with regards to just how robust some older systems can be and whether the upgrade is actually offering any benefit or whether the older system is genuinely just bad / no longer fulfilling the job requirements.
I later briefly worked there again after they had "upgraded" to a modern system. This ran a GUI on Windows on commodity hardware. The text was smaller and harder to read (on an early LCD monitor). Not all functions had keyboard shortcuts, and the keyboard itself felt worse. The interface noticeably paused sometimes.
The new system was an obvious downgrade, and the build quality of the old one was high enough that I'd expect it to have at least another 10 years of life. I can only guess that management saw the green CRT, assumed it was unacceptably primitive, and didn't bother asking the people who actually used it.
Thankfully, the newer touchscreen interfaces are much faster than before - the delays between selections went down.
The touchscreens still stink from a UI perspective - too many elements too close to each other, where inadvertent touches can result in unwanted behavior.
For example, a week or so ago I encountered an ATM that prompted me for my e-mail address to e-mail a receipt (or, as smaller, secondary options, buttons for printing a receipt or declining a receipt all together). It really rubbed me the wrong way. Who is trying to collect my e-mail address? How are they storing it? What else will they use it for? Why is this the default option?
While other mediums do exist that can act as a direct replacement in many instances, some require additional steps (eg sending signed correspondence via e-mail would require an electronic signature, photocopier with SMTP support or another means of scanning the document to PC); or offer little benefit for the additional migration costs (eg backing up to removable HDD over tape doesn't offer any significant improvements. Backing up to the cloud does, but it also offers some significant drawbacks as well, which might mitigate it's practicality for a specific business).
They're common in the US too. They're not unsupported equipment.
Yes, fax machines are still used. That doesn't make them not obsolete.
Fax is ridiculously slow, requires both participants to have fax machines (so you can't even realistically not have another solution implemented) and on top of all that super insecure.
Backup tapes are cassette tapes. Audio cassettes and backup tapes are different (not "completely" as the technology is actually quite similar), but backup tapes are magnetic tapes that come in cassette cartridges. They quite literally are cassette tapes.
> Yes, fax machines are still used. That doesn't make them not obsolete.
I guess that depends on your definition of "obsolete". Google defines it as "no longer produced or used" and fax machines are still manufactured and -as I described before- used quite common. So definitely not obsolete.
It should be noted that I wasn't advocating the use of fax, but I feel the need to address your other points since you've posted quite a specific and one sided opinion on facsimiles:
> Fax is ridiculously slow
Sending a fax isn't slower than having to manually scan your documents and then email the resulting PDF. Which is the main usage I see it still being used for. In fact in many instances, the alternatives to fax are more laboured even though the actually communication protocol for fax is slow and outdated.
> [Fax] requires both participants to have fax machines
An endpoint is required with all communication mediums. Even e-mail would require the recipient to have an e-mail address. Plus your point is moot when talking about technology that has already been adopted. If anything, the reason fax is still used is because people already have one so make use of it.
> [Fax is] super insecure.
This is sadly very true. It is probably the best reason not to use fax but it's also worth baring in mind that even for all it's flaws, fax is still no less secure than a telephone conversation (let's all replace telephones with something! :p) or the depressingly high number of SMTP servers who don't default to TLS. Thankfully that e-mail criticism is now changing at a significant pace.
---
Please be mindful that I'm not trying to advocate the usage of fax machine nor magnetic tapes. The tone of your arguments would suggest that you think I am promoting their usage, but I'm really not. I'm just being pragmatic about things. :)
You're trying to score "argument points". Stop that. The article was using it in the colloquial sense. They even had pictures. They're obviously talking about audio cassettes, not backup cassettes.
To elaborate: I was just commenting that the same technology is still in widespread use so it's still proving itself to be valuable in some infrastructures. From my perspective it seems odd to differentiate between audio and backup tapes since audio cassettes are really a subset of that technology (the magnetic tape was originally designed for computer data storage back in the mainframe days, albeit cassette form came quite a bit afterwards). So my point was that the technology itself is still widely used.
Personally I never liked audio cassettes, not even in the 80s. So I wasn't going to comment on why some people might be using that specific application of the technology.
Some time ago I worked on a modern travel reservation system frontend, to replace an aging 5250-based one. When I finally spent a day with the customer's end users to validate some things and learn how they worked (those things never shine through the specs) it was like a punch to the stomach. They were happy to see me and happy for the system replacement, but they worked with such lightning speed in the old system (where all 24 F-keys were bound and known by heart) they could keep up a conversation with the client while contantly pulling out the information they needed.
Just reaching for the mouse in that situation would have incurred noticeable latency for the customer. A snowball's chance in the proverbial hell would constitute much better odds than my web based system would have to improve their productivity. I didn't stay on that project until the end (for other reasons) so I don't know how it turned out, but it was my first real encounter with "legacy" systems and it humbled me like nothing since.
There are still high cost to maintaining these systems, and their batch-orientedness makes modern always-online mode very difficult. But productivity for end users should be very hard to match, especially since they were built for kilobit lines and megahertz computers in mind.