Considering that Steve Ballmer once pointed out that "revving software is insanely profitable", there is no reason that companies couldn't use the model of selling new revisions every year or so, which the users are free to purchase upgrades or skip. It also provides a better incentive to write better software (so the users don't skip more revs).
The only reason to not do this is that subscriptions are even more profitable, more rent-seeking, and more extractive
That works fine for purely client-side software, such as was common in the 90s. Photoshop would be a great model of that. For such software, the only, or main, cost is R&D, and you pay for their R&D for each new version.
It doesn't work so well for software like this, where one of the main selling points is that all your notes are stored in the cloud and can be retrieved on any new device (or by opening the website in a tab). That's a marginal cost which is directly proportional to your actual use of the tool itself - not an R&D cost which relates to developing the code.
I understand objecting to software where the SaaS/cloud aspect is tacked on as an afterthought, to justify a subscription model – again Photoshop is an example – but that's not what's going on here. It's eminently reasonable for this software to be subscription-based.
In short: It's reasonable for a car to be a one-off purchase, but it doesn't work the same way for the gas.
Even with Photoshop, eventually the product was complete enough that most users felt little need to upgrade even after multiple releases (I’m one such user — PS 7/CS1/CS2 is more than adequate and almost every feature added since is superfluous). They basically ran out of compelling features to add, which pushed them in the direction of subscriptions.
Not that I like subscriptions (quite the opposite) but I think most client only software will eventually hit that saturation point where sales dip and ongoing support becomes questionable.
I think this is more true of indie/small devs than of gigacorps, though. Adobe probably could’ve massages their strategy to make things work with single time purchases, but an indie doesn’t have nearly as much latitude.
Yes, if cloud storage & access is a key feature, there are some costs that scale with usage. But with a note-taking app like this, both bandwidth and storage from any user should be down in the $0.01-per-month or even per-year range.
I think Obsidian has a much better model - I can run it independently as much as I like, and if I find it useful (or want to give them more support), I can sign up for their cloud syncing service.
Having to use it from the outset, or after I get 40 cards, no thanks.
Or you could look at it as the better software they write, the longer a user may go between revs... Thinking about how long our agency used Adobe CS2... Unless you are just referring to flashy features, which I fear would not ultimately always be about improving the product. The less sexy stuff is probably where I'd rather the developers focus if the feature set is already solid.
The only reason to not do this is that subscriptions are even more profitable, more rent-seeking, and more extractive