> If you're a professional you have a machine dedicated to that work. If a software update will break software you've paid for and need to use, you don't update that machine.
Yes, that's exactly the right thing to do. Only then the security folks will start whining about factories and shipping terminals being controlled by ancient PCs with WinXP (not to mention power plants with hardware and software older than most of us on this site). In fact, the security folks and the business folks align enough on it that professional software is force-feeding you updates too, and you can't do anything about it unless you're a multinational megacorp and can afford to make bespoke deals with OS vendors.
> Between 5 and 10 years is a perfectly reasonable run time for paid pro software.
5 is the minimum. Legal minimum for some documents, in some cases.
Still, the problem usually isn't upgrades per se, it's that universally these days, newer versions of products are almost always inferior in terms of functionality, performance and ergonomics. So, I might be easily able to afford refreshing my software tools after 5 ways of using them to earn a living, but then I discover they all went to shit and new versions are worse than the versions I have (and even worse, half of the software is now subscription-only).
> So, I might be easily able to afford refreshing my software tools after 5 ways of using them to earn a living,
That's just how all of tech works. Some pockets of software may be able to be fundamentally unchanged over decades, but the fact is that the software I used 5 years ago is not the same as the software I used now. Your choices are the same as ever:
- Don't upgrade for as long as possible
- Upgrade and eat the inefficiency cost for a while until replacements are found/made
- go the FOSS route and either do it yourself or rely on the goodwill of the community until its inevitable next schism.
There is no perfect solution unless you're willing to become a domain expert in that specific kind of tech and roll your own.
The only silver lining is that most of the fundamentals remain the same so you're not starting from square one. Whatever new web stack is being used today is still probably based upon React, which is based on JQuery, which is based Javascript. ES6 isn't a complete recvolution from ES3 (even if there are new major concepts to learn over those 15 years). So you have some knowledge transfer of seeing where and how things.
Yes, that's exactly the right thing to do. Only then the security folks will start whining about factories and shipping terminals being controlled by ancient PCs with WinXP (not to mention power plants with hardware and software older than most of us on this site). In fact, the security folks and the business folks align enough on it that professional software is force-feeding you updates too, and you can't do anything about it unless you're a multinational megacorp and can afford to make bespoke deals with OS vendors.
> Between 5 and 10 years is a perfectly reasonable run time for paid pro software.
5 is the minimum. Legal minimum for some documents, in some cases.
Still, the problem usually isn't upgrades per se, it's that universally these days, newer versions of products are almost always inferior in terms of functionality, performance and ergonomics. So, I might be easily able to afford refreshing my software tools after 5 ways of using them to earn a living, but then I discover they all went to shit and new versions are worse than the versions I have (and even worse, half of the software is now subscription-only).