Any actual interesting changes under the hood other than UI changes? I cant remember the last time macOS release that actually brings any useful feature I use.
It's been so long since Apple has released anything in either iOS or macOS that excited me as a user. I don't seem to be their target customer anymore.
The only reason I even have to "upgrade" to a higher version number is how quickly app developers (including Apple themselves) drop support for older OS's. My iPhone which is stuck on iOS 15 runs just as well as the day I bought it, but every other app I download tells me (in essence) "LOL your phone is too old and our developers are too lazy to keep our software running on it. Upgrade your OS or get lost loser".
That's literally the only thing motivating me to upgrade anymore: The treadmill of software compatibility. Apple doesn't have to innovate--they just need to make sure the ecosystem is broken after ~5-10 years or so.
Isn't that true for pretty much every OS? The feature set we need to be able to do our jobs and computing hobbies have been available for two decades.
Operating systems like Debian is sufficiently boring that I can just upgrade and continue computing. macOS upgrades have become a small gamble, the stuff that I depend on may not continue to work, or at least it will take a good deal of work. There are however no reason to upgrade, so the risk isn't really worth the hassle of upgrading and breaking Java or Python.
You can still get software that installs and works perfectly on Windows 7 (released 16 years ago). Good luck finding software that even installs on Snow Leopard (released 16 years ago), let alone works well.
The flip side of this is that every attempt at advancing the Windows UI framework story beyond win32/MFC and WPF has failed and the platform itself is steeped neck deep in technical debt.
You can't blame this all on backwards compatibility. Microsoft can't just stick to one successor technology and keeps obsoleting every attempt after a few years. With that approach, of course no one is going to move over.
That’s one issue, but another is that they don’t make into successors full replacements. They’ve stuck to WinAppSDK aka WinUI for a while now, but it’s missing such basic elements as a datagrid/tableview which makes it difficult to take seriously as a desktop UI framework.
Also, they didn't even migrate their own internal dialogs with each replacement. You can still uncover dialog styles all the way back to Windows 95-style if you poke around Windows 11 long enough. Whenever they came out with yet another successor, they haphazardly migrated a handful of P1 components and then called it a day.
I mean OS updates are necessary, from security updates to support for new hardware. Stuff like init scripts were a suboptimal solution to begin with, it is not like that old stuff was a good solution. Xorg also doesn't fit how modern computers work, and is merely a collection of bandaids that is inherently unable to reach sane security by todays standards. So its not like progress is entirely superficial. And also with Flatpak there has emerged a way of shipping stuff such that used libraries can be shared but do not have to be, and every app can move at its own pace without conflicting with other apps or the OS. So at least in Linux land, especially in the last couple years, there has been great advances from a technical point of view. And those tackle also the problems arising when a huge number of indipendent parts come together, which were naturally very pronounced for Linux in the past.
Sometimes it’s Apple and Google that are forcing developers. The system is perfectly capable of running the app (you’re not using any new API) but store policies force you to add the restriction anyway.
Yeah we are in this situation right now with an App, we literally can't update it unless we target a more modern version of the SDK, which introduces breaking changes
This problem could be mitigated by Apple making older versions of software available. Then you could continue to release updates, and users on older devices could continue to use earlier versions of your app on their devices.
Apple actually partially solves this: as a user, if I have EVER downloaded Older Version X of an app, and then go to download it again, they let me. However, if I have never downloaded the old version and go to download it, they just say “this app is not compatible with your device.” and don't give me the chance to get the older, compatible version. I don’t know why they make this distinction.
Worse are the third party apps where the old version still actually runs, but the developer deliberately blocks you with a full-screen “go away” dialog (I’m looking at you, FlightAware).
Does it mean that you potentially have multiple version of the same software released if you are choosing to support devices that have been dropped from the latest SDK?
I got my first MacBook at Catalina, and still miss it. For a while, I downgraded my Intel Mac to Catalina again; I love the aesthetic compared to the newer releases, and it’s fast and snappy.
But the situation now is: No recent apps work on Catalina since it’s considered obsolete (except open-source apps you compile yourself). But Big Sur and higher are ridiculously slow on Intel hardware, to the point where it’s unusable. I now have an otherwise perfectly good 2019 Intel MacBook that has been gathering dust for the past years.
I’ve got a MacBook and Mac Mini stuck on Monterey (12), and an iMac stuck on Big Sur (11). I’m pretty much dead in the water when it comes to software compatibility, unless I want to put Linux on them. Even homebrew gives me a warning that they’ve stopped support and to expect everything to break. It’s a sad state of affairs.
This[1] worked well to upgrade old Macs that were stuck on old versions of macOS for me, if you're not choosing to stay on older versions for other reasons.
Support rapidly being dropped happens mostly with smaller devs, because when resources are limited in the Apple platform world you can either adopt newer APIs and language features or you can support old OSes 3+ versions back. Trying to do both lands you in feature check conditional hell and requires a large matrix of test devices to ensure that nothing is being broken.
It’s less of a burden for corporate giants which is why you see much longer support timelines from e.g. Google.
The immutable and atomic movement in Linuxland is very exciting. Cloud native distro building with docker/podman in CI pipelines is just insane: Building+testing+deployment with ci/cd can now not only be done with some python package, but a whole distribution. Also Wayland, Pipewire, Flatpak and Btrfs are great stuff. Of course they don't get developed in one release cycle, but in recent years they made large leaps and became default on many distros.
Not Linux, but I still look forward to window managers and Neovim releases. The Cosmic desktop also looks promising, though I’m not using it until it has a scrolling window manager available for it.
Spotlight got a major upgrade. It’s notably faster and deeply integrates with Shortcuts (letting you specify input variables, for example) among other things.
I’ve got Spotlight configured to index nothing but my applications (which is surprisingly difficult to configure and breaks with every major OS upgrade). Disabling all its default indexing has alleviated 95% of unexplainable CPU spikes and autocomplete pollution, so now I can finally use it for what it’s meant to be: the most overengineered fuzzy finder application launcher.
I actually preferred the pre-tahoe spotlight. The information density was higher and while it did not always give me the most relevant result atleast it was consistent and I could scroll down to find it. New spotlight is less dense and jumbles everything together.
Anyone using Raycast has had these features forever. Nice to see some attention on Spotlight but it's still nowhere close to the functionality you get from Raycast.
I've been using Raycast for a couple months but I'm hoping I can uninstall it if Spotlight is responsive enough in Tahoe. What bothers me about Raycast is the monthly subscription for certain features. I don't mind paying for Mac software – I'm quite happy to do that – but I do mind paying monthly subscriptions for Mac software with seemingly no justification for it (i.e. what monthly resources does running a "window command" use on Raycast that justifies locking it behind a monthly subscription?)
I thought Pro was only for AI features as well (that's what it said when I installed Raycast), but this dialog is saying Pro is required for custom window layouts as well. I only discovered this today when I was trying to create a new command to paste the screenshot from my clipboard into Preview for OCR.
Raycast is interesting but I’m not going to touch it so long as VC funding is involved. Alfred has been doing the job well enough, only requires me to buy a new version a couple times per decade, and isn’t going to become enshittified because there’s no VCs to come knocking looking for a profit.
+1 for Alfred. I'm a proud Power Pack / lifetime-license holder from the beginning. Very few outfits anymore have the chops to both offer and make good on a single-payment, long-lasting product with frequent and excellent substantive updates.
It’s insanely tiny and efficient for what it does, too. One of the only apps that’s so small that updates are done downloading within a second or two of clicking “Download”, even on a mediocre connection!
I'm the solo user of my device & have rarely if ever launched Photos or Photo Booth, but I launch Photoshop a handful of times everytime I work.
For the last few years, it is a total gamble of what Spotlight's first suggestion will be when I trigger it and type "Pho"
Often, Photos.app
Weirdly, about once a week, Photo Booth.app
80% of the time, it correctly guesses Photoshop.app
So I can't just trigger spotlight, type "pho", and hit return. I have to wait for Spotlight's guess, sometimes arrow to Photoshop, or sometimes just hit enter when it guesses correctly.
So many macOS microFrustrations like this, and now Tahoe looks like it does.
The fact that so much of the page is devoted to this liquid glass feature pretty much tells you the answer is no. Plus the fact that the "And so much more" section lists 10 different updates in the same space as their poster with a link to a PDF instead of building out a larger webpage speaks volumes.
I have 5000+ git repos locally that I grep through regularly and I have to admit my hopes for ASIF being faster were busted. It is faster than sparsebundle, but nowhere near native, unfortunately.
TextEdit has a styling toolbar now which I appreciate. The new spotlight has more functionality and seems faster (and less likely to pull up a website instead of the app I'm trying to launch)
