Sounds like you should stick to FLOSS then. Anything you buy as SaaS can be canned, sold on or just become crap.
However if the maintainer happen to stop maintaining the service, I could setup the infrastructure and maintain it by myself with some learning, but that would take up more time I need to earn the money I'm paying to somebody else, who already done the groundwork and has the knowledge about the service. Also the service is worth $5-10 per month, but does not have enough value to worth the hassle of setting up the whole self hosted infra behind it and maintain it by myself.
I'm not sure why, but people recommending FLOSS as a solution tends to forget about the additional cost to setup and maintain infrastructure and software and a simple build pipeline (even if it's just a bash script, rsyncing the code to a server) when mentioning you could self-host it if they shutdown the service.
This is the "slightly annoying not to have this service, but not annoying enough to bother to replicate it even if I could" situation for me.
Freedom has a price.
I'm also tired of the trend of recommending FOSS when somebody is looking for a service. For personal solutions it works only if you are valuing your time really low or if you are having a lot of fun playing with it.
Frankly, if you think setting up your own FOSS is often a good solution you haven't listened to your advice enough times. It works if you have a few things. But if you try to do it with everything you will just give up after a few years and 50 things set up this way, or you will spend considerate amount of time on maintenance (which is fine as long that's what you want to do with your life).
But you inherently have less control over the situation when you pick this option. Sure, it's convenient to pay someone else to manage the service, but you expose yourself to the acquisition situation that this user did. Also, moving to another provider isn't usually an easy thing to do, it's not like deciding to shop at Costco instead of Wal-mart today. Beyond the most trivial of services, you usually pay a fairly hefty technical migration cost to move to another provider. Most companies can't just wake up on a Monday and decide to switch from AWS to GCP or another cloud provider.
I guess "Freedom" can mean different things to different people, but when you're at the mercy of a vendor like this, you're inherently less free than you would be if you ran your own service because you can't control what they do.
I am free to breathe, I have to physically do it though.
Some people have a machine to do it for them, they certainly aren't free -- https://www.theguardian.com/society/2020/may/26/last-iron-lu...
Sure, I sacrifice my freedom in the name of ease or expediency all the time - I'm not someone like Stallman, but I'm conscious of the tradeoff, and one of those is that services I use may well shut down, sell themselves, and indeed sell my data. That's part of the price of not being free.
If you want both, you could pay someone to manage a FOSS installation for you
This is exactly why you need the freedom of FLOSS. You can hire anyone to continue the development. You don't have to do it yourself.
Sure, and like everything else that has a price has to go through an cost benefit and an opportunity cost analysis.
Just because it "has a price" doesn't mean it's always worth it...
You just get the assurance that in the off chance somebody (including you) cares enough to maintain it, has the time, and is competent, they have have the license to do so.
But that's not that big a deal, except if its a project with many diverse backers and large diverse community (e.g. something like MySQL or SQLite or even Webkit will survive and be forked if needed - something like the average 1-3 main dev project not so much).
If an app can be a desktop app written in say JS (Java c# or similar term back-compat language) and not need “infra”. Or if it can be a Unix until that does one thing well. Then that is more likely to be a thing that can be forked and untouched for years and just work.
Granted anything that interacts with web services may have bit rot. However solid principles may allow those parts of code to be surfaced out to plugins etc.
Anyway let’s put it this way. I use YNAB 4 desktop budgeting software. No longer supported, not even FOSS and I’m not worried I can run that thing for the rest of my life, with little hassle, as long as I can run the last version of Windows that supports it in a VM.
If you just want a dependable service that's going to stay around, and you're willing to pay for it, what do you do?
As you say, anything that is SaaS could be canned/sold/etc. Anything that's FLOSS won't be as much of a service as you wanted, presuming you wanted to pay money on an ongoing basis to have a problem solved in a convenient way.