I think different person has different standpoint.This was a part of my work at the university. Manufacturers are plain stupid about how they integrate computers into instruments.
For example, we had to maintain a PDP-11/70 in order to keep an old but very important microscope operational. A replacement for its functionality would have run in the high 6-digit price tags, so we had incentive. The first challenge was transplanting the system to our network. It required learning about the IP stack configuration for an OS that was long “dead”—but not for us.
Another example was a spectral analysis instrument for a SEM that was about $200K worth of kit. It used a DOS program to collect and analyze the data from the sensor. It as running on what had become an ancient PC, under, of all things, OS/2 Warp. (Yes, it was a DOS program). We built a machine at least 10x faster than the original from spares and put DOS 6 on it. The program would not run!
I downgraded the DOS through every version down to 2.11 and it still wouldn’t run. Though I had access to everything Microsoft through MSDN I did not have install media for OS/2 Warp, which, apparently, was the only environment it would run in. We rolled back to the old machine and it lived out its life—as predicted, a short one—and then became so much very expensive scrap.
And here is the problem with the vendors. The manufacturer of the device was still in business but they had EoL‘d this particular device and so even with its very high cost there was no one to provide updates to make it work. The instrument itself was beautifully made and would probably have operated indefinitely but the computer it relied on was not going to make it past several years, and replacements (which is a niche business) for older motherboards with AT bus slots, etc., cost considerably more than a much faster, better modern version.
And, don’t get me started on Tektronix selling super-expensive scopes, network analyzers, and the like running Windows 95! All of that expensive hardware, nowhere near out of date, chained to an operating system that couldn’t be placed on the network safely! I had to design an isolated network for our labs so we could allow for Windows 95, Windows NT, and even Windows 3.11 in some cases where exposing them would have meant instant compromise.
But, we had to keep the benchtop computing running, and different members of my staff had different specialities to help do that. Legacy hardware, legacy OS, particular products, etc. It was a vital function for labs that couldn’t afford to throw away a $250,000 instrument because the $2,500 PC it relied on was no longer viable.
As an end user, such as you, just focus on performance, quality and competitive price.
As manufacturer, they not only focus on performance and quality, but also focus on cost, marketing, competitor, productive cycle and so on. So they will care about whether it is worthy to change it.
but sometime manufacturers don't know user's requirement in time, maybe you can suggest them to improve it.