Operating systems are starting to make that impossible.

I used to use an old version of Quicken for Mac (this was back in the MacOS 10.2 days). It was "finished" in that it did absolutely everything I needed and there were no bugs that affected me. As time marched on, Apple switched from PowerPC to Intel which was fine because they had technology (Rosetta) which allowed me to run PowerPC applications. Then they dropped Rosetta, so now I had to choose whether to upgrade MacOS or continue to use Quicken.

I now use software called Banktivity to replace Quicken. Much like Quicken, Banktivity is "finished" as far as my requirements go. The version I have is Intel-only and Apple is switching to Arm processors. How long will Rosetta 2 last before it's pulled and I'm forced to choose between upgrading my OS and using this software?

I think there are similar things in the mobile world too. I think Google requires apps to use a minimum API version to maintain a listing in Google Play. I believe Apple does something similar with the App Store too. So short of regularly recompiling, testing, and publishing your app, everything in the stores has a finite lifetime.

>Operating systems are starting to make that impossible.

Microsoft operating systems have always strived to make that impossible.

Apple has to compete somehow.

When you push unfinished, not-fully-tested code out the door, even if you can confidently patch it later, it's just never been the most professional thing to do.

Especially for something that could be mission-critical like an OS.

Getting it into people's hands sooner is no excuse.

This kind of thing really lowers the bar when it comes to the rest of the development ecosystem constantly engaged in making temporary solutions to permanent problems.

This ripples out and overflows far enough to effectively be a major anti-recycling effort for megatonnes of hardware itself.

>> "Microsoft operating systems have always strived to make that impossible."

That's the one thing I've never seen someone accuse Microsoft of. They're famous for their obsession with backward compatibility.

It sounds like you've had a different experience, and now I'm curious. Care to share?

Microsoft's backward compatibility was always impressive until Windows 10 (ignoring the death of 16-bit on 64-bit Windows). S Mode and the store are great examples of how Microsoft has lost interest in supporting third-party binaries. They want to own Android, not Windows.

I've had software break in the forced feature upgrades of 10, and I had BSODs when printing on every machine I supported at work thanks to a security patch.

It's possible they're breaking some stuff so other things work better. A lot of developers are still using the registry, for example. The choice is basically leave it wide open for developers to break or push them to AppData and friends. All the programs I use also use the current standard, and I don't have any issues.

This might actually explain why 10 and 11 have been the most flawless experience I've had with Windows while so many people have endless trouble. It might be similar to the transition to NT in XP: everything broke, but almost everything made after was more stable and secure.

That was exactly my reaction! The latest visual studio still runs on Windows 2000, and I've yet to find anything that doesn't work on Windows 7.

A much different kind of backward compatibility would have been needed for it to have been more useful instead of unnecesarily wasteful.

Each version of Windows was designed to ideally trigger a purchase of new hardware as a concession to the hardware partners, while each purchase of more powerful hardware was designed to be prohibitively difficult for users to operate using previous Windows versions, triggering a new OS purchase.

Non-Microsoft software suffers more from this disadvantage, compared to the benefit of having some of the Microsoft non-OS software function fairly well across a few OS versions.

I understand how embarassing it would be for users in general to have become aware of how much higher performance they could have gotten from using a previous version of Windows on the next generation of hardware, so this was always made unnecessarily difficult if not impossible.

Also how much hardware could have been saved from scrap heaps if maximum utilization of hardware was given top priority?

When the main reason for growth in manufacturing was consumers purchasing their first computer, this is not as much of a problem.

Once computers are ubiquitous, each new purchase triggers a corresponding waste declaration event.

Accelerating growth pressure under these conditions, while failing to even provide archives for "outdated" code needed to operate remaining vintage hardware, amounts to an overall anti-reuse/anti-recycling strategy.

Would you consider running an emulator to keep using the original software?

In gaming this has become mainstream as GOG and Steam now leverage DOSBox for this purpose.

Personally I think vendors beyond a certain size should be regulated so they must maintain BC for some minimum period like 20y. Regressions should incur fines unless it's proven they are required for security fixes.

I’m torn on this. For both apps I mentioned there is a reasonable path forward, it’s simply a matter of buying a newer version. But it really amounts to spending extra money to get functionality I’ve already paid for once, after all both of these programs are 100% complete and bug free relative to my requirements.

I’ve got a few GOG games and they’ve done an incredible packaging job with DOSBox - if there was a way to take an old PPC Mac app and package it up to run in a similar manner, I’d have probably taken a serious look at it back when Rosetta was discontinued.

No, shitty operating systems make that impossible.

What’s an operating system that makes it possible?

I'm willing to bet that Windows 10 or 11 would have zero trouble running some random version of Quicken from 2002. Most PC games from that era are still easily runnable without too much hassle and those have way more complexity and weird crap to deal with.

Linux might be doable too if it doesn't have a bunch of shared libraries or if you at least know what versions it needs.

That’s about what I was thinking. I made and distributed a small application on Windows, Mac, and Linux back in the late 2000s and only the windows version still works. Getting the Linux version wouldn’t take much work, but the mac version would need to be rewritten.

Docker?

Security vulnerabilities are regularly found in many softwares (both online and offline), so that's one reason to upgrade.

Sure, but in some products that's really a non-issue.

Like, take photoshop for an example. If you're a photographer that only ever edits your own photos, the odds of a needing a security update for photoshop are really minimal.

If you regularly edit photos from clients, maybe it's worth considering.

I'd say definitely: https://www.f-secure.com/v-descs/exploit_w32_jpg_vulnerabili...

Eh, it still depends.

Is it riskier than keeping an updated copy? Sure. Is the expected return period of a hack given your circumstances above a threshold that is acceptable to you? Maybe, maybe not.

> so that's one reason to upgrade.

That seems like "hey, we develop our software the worst way so that it's always full of bugs, but if you want less bugs, buy our subscription". Complete conflict of interest vs making quality and well-tested software in the first place.

Well except most useful software (not all) interacts with other systems. I loved my old mail client until it stopped working with the latest encryption schemes. A nicer way around this is like what plan9 has, where programs don't worry about handling secure connections etc, so you don't need to patch every program.

some people still use a 90's version of wordperfect, and that's fine, but if you have software that depends on other software - example Qt3 was dropped everywhere, even though there's plenty of useful Qt3 software around, it needs to be upgraded.

So even if your system wasn't buggy, it may be incompatible or the dependencies no longer are updated (like with Qt3, which had networking and other helper functions), you wouldn't want to keep using it.

Security vulnerabilities should be fixed by patches/security fixes w/o forcing users to go up the ladder to "za niu anz shaini version" + "our new and enhanced x$/mo"

Should be = not happening in a world where not doing it means more money.

With that said it can get problematic fixing old third party packages that are no longer supported.

Then the money for doing this should be provided to developers by the universe.

Proprietary software should come with a warranty that covers security vulnerabilities for N years.

Nothing is stopping you from requesting such a warranty from your vendors, and refusing to buy from those who fail to offer such a warranty.

Coordination problems do.

Imagine a developer who has to spend 50-90 hours per month just to provide the promised security patches for each of the 5 products they have created and sold at least once.

No one is in a better position than that developer to make this maintenance burden lighter, so they're the one who should be incentivised to do that.

https://news.ycombinator.com/item?id=32286651

This used to be the case.

Photoshop, for instance. I don't need it to have new features; there are certain things I need it to do for me, and that's it. I don't need constant new features. But I do expect security updates.

Same model we have with cars (that BMW is trying hard to upend): it has the features I need, and if there's a systemic defect that poses a safety risk across those models sold, they're expected to fix it.

I agree that subscriptions have become toxic in their prevalence.

>> I don't need constant new features. But I do expect security updates.

Isn't updates to software in response to emerging security issues a "feature"?

Nope lol it's most closely relatable to a product recall/repair.

Usually, repair shops are not free.

Cars are nothing like software, they clearly deteriorate with time to the point parts eventually need replacing. You could argue there's a better case for a subscription model for buying cars than for software.

> Why can't software ever be "finished" or "sufficient" to meet your needs?

For a very long time the driver was operating system and hardware features.