Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Apple’s macOS Ventura – New Security Changes (sentinelone.com)
176 points by alwillis on June 19, 2022 | hide | past | favorite | 179 comments


Lets hope the open core project, namely the Open Core Legacy Patcher[1] will revive some older models to run Ventura. Personally, I'm running a 2014/15 Macbook Air 11" for 7 years now[2] and with "Open Core Legacy" on Monterrey with no issues at all.

[1] https://dortania.github.io/OpenCore-Legacy-Patcher/MODELS.ht...

[2] Except the mainboard, display and shell everything else thats modular (wifi card, ssd, battery and keyboard) was replaced/repaired at some point.


In pretty confident it will, the OCLP project has improved a lot and is now very capable and stable. My MacBook Pro 2012 runs Monterey, and it's really fast and stable. Better than any previous Mac OS in fact.

You can look on the Macrumors forum and Reddit to see if they're already getting the betas to work.

https://forums.macrumors.com/threads/macos-13-ventura-on-uns...

https://www.reddit.com/r/venturapatcher/

Edit: the OCLP team released an official statement, looks like they're having quite a few challenges. https://github.com/dortania/OpenCore-Legacy-Patcher/issues/9...


Interesting that you say that it's faster and more stable. My 2013 Air is on Monterey and it's "good enough" for my needs still, but slipping towards not quite good enough. Very tempted to give OCLP a try if it extends the life of the machine a bit more.


May be off topic but this had me thinking.

Will Safari 16 be available on older macOS? Assuming Apple doesn't break their tradition it should support two prior OS release, Monterey and Big Sur.

While I could do without an OS update on my MacBook Pro 2015. ( I cant record a single useful feature from all the previous OS update other than Universal Clipboard ) That means for MacBook Pro 2015 Model users they will only have two more Safari Update.


Will Safari 16 be available on older macOS?

Short answer: you'll get all of new web platform features (Container Queries, Subgrid, etc.); you may not get certain Safari-specific features that require a feature only present in macOS Ventura.

You can get Safari Technology Preview for Monterey and Big Sur that has those new web platform features right now [1].

[1]: https://developer.apple.com/safari/technology-preview/


The latest STP that runs on Monterey has the features from Ventura [1].

[1]: https://www.macrumors.com/2022/06/21/apple-releases-safari-t...


There was this guy called dosdude on macrumors that made these excellent patch sets to get them mostly working with unsupported versions.

But it involved a lot more than just firmware. And there were lots of known issues with some configurations.

I'll check that software out though, perhaps I can run Mojave on my 2010 mbp, I really miss the dark mode


At least two other security changes:

- userspace filesystems: the nail on the coffin for kernel extensions. Now we won't need to run in "reduced security" to use FUSE and that was the last kernel extension that remained popular. Probably kexts will be deprecated shortly - rapid security response

- it seems also to include changes in Xprotect and mrt


Do you have any links on the userspace filesystems? There's a few things I'd like to develop in that regard without getting into kexts


There is that: https://threedots.ovh/blog/2022/06/quick-look-at-user-mode-f...

But userspace filesystems are already present in iOS so you can find some reverse engineered info on that (e.g. in Jonathan Levin's books)


Thanks


Note that UVFS is currently not exposed to third parties.

For your own 3rd-party file systems, you'll still have to use a kernel extension for now. (or local NFS mount).


Very curious about userspace filesystems, would be awesome if there's finally a fast solution for this that's well-supported in the OS.


> the nail on the coffin for kernel extensions

The OpenZFS implementation on macOS also requires kernel extensions, and I don't suppose it can easily be ported to FUSE or that that would have desirable performance characteristics.

Special kernel extensions are also required to get some basic functionality working on macOS these days, like disabling pointer acceleration.


Too bad for them, Apple has already stated multiple times that the long term roadmap is that all third party stuff will only be available as userspace drivers.

The plan being, each kext has one year timeframe to migrate to the new userspace API after its introduction, and the year thereafter the kernel API gets dropped from the new OS release.


Maybe my intuition is wrong, and System Extensions support enough functionality that the ZFS port will still work well with the System Extensions interface, and my impression of userspace filesystems as slow is due to quirks of the FUSE implementation on Linux. I hope so.


> disabling pointer acceleration

Could you point me to the kext that does that? I’ve been trying to find a way to disable it for ages.


Under Miscellaneous, Exactmouse Tool. https://steelseries.com/engine

This is what everyone uses. Mouse acceleration disable is a dealbreaker for me- I won’t update unless something supports it.


That no longer worked, the last few times I tried it. The only thing that works now are these drivers that implements a different type of pointing device (CursorSense and SteerMouse) from Plentycom: http://www.plentycom.jp/en/index.html


The login items panel is such a good change, and also like 20 years overdue.


Yes, I have been baffled that it was always so difficult for the user to manage what’s auto-launched on start-up. So many apps try to bury into start-up so they can keep collecting data and lightly spamming the user.

Can anyone shed light on why it took so long? I had always figured the non-existence of a login items panel was a purposeful choice.


Maybe because it was not a common feature in the market. In Windows for example it's even more of a mess because there's a ton of places for autostarting software to run from. There this excellent tool from sysinternals called autostarts but it's not an OS feature despite sysinternals now being owned by Microsoft.

And in Linux? No great way of managing systemd via the GUI either afaik.

And Apple has always been one for hiding technical complexity from the user. It's only that security became prio #1 that they're doing this. 2010 Apple would not have presented the user with these popups for example.


PS: Why the 0 points? :) They even made fun of Vista's popups at that time: https://www.youtube.com/watch?v=VuqZ8AqmLPY

Apple at that time really didn't care as much about security as they do now, it was before the fappening, before wannacry etc. They were more obscure as an OS so there was not as much malware. Security was not as much on the radar as it is now.


It's actually been there for a long time. It's a separate tab in "Users & Groups" pre-Ventura.


This view only includes actual apps that launch on startup, it doesn't include agents, daemons etc. Many popular apps have one, if not many, that the users are usually not even aware of and can't turn off via the UI. To see what I mean, try running `launchctl list | grep -v "com.apple"` as the user you're logged in with. It will list jobs loaded into launchctl not owned by Apple, and that isn't even the only way to make things run at startup.


No, the “Login Items” panel that was previously under Users & Groups did not include Launch Agents or Launch Daemons.

Launch Daemons in particular are managed directly by launchd and can have more sophisticated triggers including periodic execution.


I see. Thank you for the correction.


Apps aren’t required to use that screen though. Some of them will, but a lot of apps not in the App Store can and do register themselves with launchd on their own.

Apple should be proactive and extract those items automatically, but in practice, they don’t.


Launch agents and daemons didn't even show up there. It was more for the user themselves to register programs they wanted to auto open.

However that feature became a bit stale since macOS started reopening all apps anyway.


> However, the Gatekeeper check here is overridable by users.

This is presented as a flaw, but I'm not sure they are thinking through the alternatives. It's hard to give too much credence to security experts who are't thinking holistically. Perhaps there is a flaw, but I'm curious to know what it is.


Security folks tend to have a very myopic view on things. Ever wondered why your computer got less and less useable? Security people pushing their agenda.


Yeah, because we get many sad users when their hard drive gets encrypted by ransomware, and even more so if it is a shared drive.

So the less toys to play, the better.


> Ever wondered why your computer got less and less useable

Would disagree.

I think the security changes have made the OS more usable since I now get visibility into what apps are doing.

And I love the idea that security people pushing their agenda of making devices more secure and more private is painted as a bad thing.


I've always learned:

  security = 1/convenience


They explain their reasoning right after that statement. Their concern is social engineering is still a way to convince people to override this.


It is but there needs to be a way for the user to keep full control of the system they own of they choose to have it.

And enterprises already have a way to turn this override off so I don't really understand their beef here.


Things the user can override are things social engineers can convince users to override.


Exactly. Technical measures are important, but if someone wants to play a game or do something that's been banned on Apple's stores and finds a site that claims to have an installer (which is actually malware) with instructions to disable Gatekeeper or SIP or what not, social engineering can work. Their goal is to do the thing they wanted to do, probably not thinking of security in the meanwhile. Popup alerts are going to be interpreted as something to get rid of so they can do the thing.

It's a difficult balance. Power users, engineers, developers - we can (usually) tell when warnings need to be heeded. People who use their devices to achieve a goal without really understanding or caring about what's happening usually won't.


I still recall when that viral elf bowling game was showing up on everyone's computers, and it struck me that we were all quite fortunate it wasn't secretly malware.


It's not a balance. There must be a way to override it.

I really don't want Apple to decide what I can have on my computer like they do with iOS. It'll be more secure but also a lot less free and functional.


To which my answer is — being alive is risky, get over it and stop treating all people like idiots that need constant oversight lest they do something potentially dangerous.


"... and that's why I don't wear a helmet."


This is overboard. Defaults matter a LOT. Requiring users to override defaults to deliver ad/malware is a losing proposition and ad/malware creators know this.


Yes. Note that thus is the technical argument for disallowing sideloading.


Let's also ban antifreeze because someone could kill you by socially engineering you into drinking it.


I’m sure quite a few MBP 2015 users are going to be a little sad the end is near.

I loved that machine. I was able to skip the 2017 MBP and go to 2019, but honestly I miss the smaller trackpad.


Along with other comments, we have at least 10 fully working mbp 2015s. Many with upgraded 2tb storage. All used for audio work. Running 10.13 - 10.16. Little to no issues and I still buy up any I find online. Absolutely fantastic machines.


I have a 2015 MBP, and still think this machine has legs to last a bit longer. Maybe I'll put arch linux on it later on


It’s what I’m gonna be doing!!


Still using the 2013 and 2015 MBPs here, had no reason to upgrade. Love them to death, still no hardware issues whatsoever except a new battery. I also prefer the smaller trackpad. Guess its time to gift them to my parents.

Luckily, the M1/M2 is finally a worthy upgrade, after years of keyboard issues and unwanted features.


My 2013 mbp is being retired for a mac studio as I accidentally updated ios to a version that needs xcode13, which i cannot install.

It had a good run! (My 2011 imac died last year)


agreed the larger trackpad is a net minus. I get all kinds of spurious input because of it sensing my palms


Butterfly Keyboard ( And arguably the new Magic Keyboard ) with little to no Key travel distance, along with Larger Trackpad which create false positive input were two key minus design features.

Unfortunately every time I pointed this out most of HN were quick to answer this is an user issue and not a design flaw.


I can see why most on HN (and actually outside of HN too) would say a larger trackpad is great, so I think you might indeed be in the minority opinion on this one. Even mainstream reviewers tended to list it as a positive.

But butterfly keyboard had been pretty much universally decried as a terrible mistake almost everywhere, including HN.


>But butterfly keyboard had been pretty much universally decried as a terrible mistake almost everywhere, including HN.

That was certainly not the case until the reliability problem got magnified in 2018. When the problem has been there since 2016. Before that Butterfly was somehow the holy grail for touch typist.


I don't believe it ever was the holy grail of touch typist. First, it was incredibly loud, and most people complained about it. The key travel distance was mostly cited as a con, not a pro. And only then the reliablity issues started to arise. But it didn't take 2018. This article (https://news.ycombinator.com/item?id=15496745) made it to the top of HN in Oct 2017, but people had already been complaining for months (See this article from February https://forums.macrumors.com/threads/some-2016-macbook-pro-o...).


A lot of us hated the butterfly keyboard from the beginning. Both here on HN and in other community situations. The noise interfered with calls when taking notes. I made a lot more typing mistakes due to the space being reduced between keys. I’ve typed on lots of keyboards over the years and got used to most of them but never could get used to that.

When they skipped or doubled keypresses started it was just the cherry on top.


I remember reactions being more mixed. I and others always wanted more key travel, but I also knew people who loved it.


You're not alone. I hated the butterfly keyboard, it was like tapping on stone.

Even the "returned to normal" keyboard on the new mbp is not nearly as good as the 2015 was in terms of tactile feel.


Will non-Ventura devices still get vulnerabilities patched, or are they dead in the water?


Apple has established a pattern of haphazardly offering partial and late security updates for the two more releases (e.g. Big Sur, Monterey, Ventura).


Looking from the outside, it sure does look like every security patch is assessed on its difficulty to port to older OSes, its severity, and its reach.

It can be maddening.


Security updates are N-2 I believe (so current OS and last two major versions).


> "Passkeys aim to solve the problems with passwords"

So are you locked in with Apple if you use this, or can you switch all your existing passwords to another "passkey provider/service" ?

> "Wave Goodbye to CAPTCHAS"

I assume that's Safari only... so this is bad news for Firefox?



> In collaboration with Google, Microsoft and other industry players, Apple has been working on a new logon technology for web and other remote services called ‘passkeys’.

I don't think it will be Apple only. However, I am wondering what will happen to services like bitwarden [1] if it is available on other OSes as well.

[1] https://bitwarden.com/


What's your concern regarding Bitwarden?


Not the original poster, but I think the worry is that how can a password manager survive in a passwordless future.

Will major players be too powerful that no competing solutions will realistically exist.


That's exactly my concern. I am the OP you refer to. BitWarden is a great product and an open source one at that.

I suppose that passwords will always be a thing for the paranoids amongst us. But i don't know whether that is enough to sustain the product.


You still need to have a wallet to store all your passkeys.

I’d rather use bitwarden than Apples wallet.


> So are you locked in with Apple if you use this, or can you switch all your existing passwords to another "passkey provider/service" ?

Eventually, yes. Not now, but the goal is eventually, yes. It requires support to come to Android and at that point, they'll build the bridge to bring them together. I don't think the system currently exists for this, but they've said Passkeys will be a "multi-year industry-wide transition" so I'm inclined to believe it'll ship in the coming years.

When you sign in with a passkey, you have the option of scanning a QR code from a locally present device running any software that can speak the standard (e.g., Android). This means that you can login using any software that supports Passkeys using any devices that support Passkeys. For example, Chrome on Windows (chrome://flags, turn on passkey support) with an iPhone is a valid pair.


Passkey's (and webauth2's) major value proposition for software/service providers is for people to stop sharing accounts. This will result in more sales/subscriptions, and better tracking of preferences/activity.

That it also happens to provide better security is just more cheese on the trap.


If they don’t allow a way to extract the private key from the wallet, do you really own that account?

I wouldn’t touch it with a 10 yard stick unless I have full control of the private key.


curious too as someone that uses 7 computers running 3 OSes, are my passkeys accessable, syncable, across OSes?


One obvious solution is to use a cross-platform password manager that supports Passkeys [1].

[1]: https://blog.1password.com/1password-is-joining-the-fido-all...


This is based on a standard that Apple, Google and Microsoft have all agree to and have agreed to a method to transfer keys between devices.


three US Corporations will be in charge of the implementation of this newly required security ?


Considering that the three control all of the popular operating systems for computers and devices, and most popular browsers who else would it be? It’s an open standards.


Assumedly any service that implements this will let you reset your password away from Passkeys, but it's still the soft lockin of "Ughhhh I don't want to reset everything".

Sites will never go full Passkeys because that obviously falls over if you want to access it from any other device or computer, support request costs would go through the ceiling


I wouldn’t be surprised if banks go passkeys only.


By 2040.


I don't think number 2 (specifically the complaint that the user can override gatekeeper) in their list is really an issue for enterprises. If they want they can already turn off the gatekeeper user override through a profile. I do exactly that on our Macs.

And for consumers I don't think it's a good idea to remove this option altogether. The owner should have full control if they desire it.

The constant check for modifications is great though. I'm surprised that wasn't the case before.


Most concerning is Gatekeeper, as I do still routinely run into scenarios where it harassess me about applications I am trying to run and on the odd occasion I have to manually codesign things.

It will be super annoying if this now starts making developer's life hell because it is nannying binaries they are building, sharing or working with as part of their development work.


You can completely disable Gatekeeper if it annoys you:

$ sudo spctl --master-disable


I’ve done this a few times but still get nagged when launching unsigned apps for the first time. Is that part of a different security mechanism?


Probably, I would need to see the message to be sure what you're referring to. But LaunchServices will for example still warn you the first time you open a downloaded application. You used to be able to get rid of that with `defaults write com.apple.LaunchServices LSQuarantine -bool false` but I don't know if the syntax has changed.


If you're building, you can designate something as a developer tool and gatekeeper will ignore it.


Am I the only one increasingly frustrated with macOS's naming scheme? I have no idea what the latest version is. Ubuntu versioning gets this right; you can parse their codenames alphabetically to derive the semantic version. But Apple's heuristic here seems to be "throw a dart at a map of California".


The marketing names for macOS and OSX have always been random other than having a general theme to it.

The OS in most cases just refers to itself by the number and is what it will show in a lot of scenarios in addition to the marketing name.


> The marketing names for macOS and OSX have always been random other than having a general theme to it.

Not entirely true: two of the cats were name variations of their predecessors to express an intent of limited end-user / feature updates and a focus on refinement (even though taxonomically the cats have basically no relationships outside of being cats, mountain lions aren’t even in the same genus as lions)


In a sense they sort of did continue this style of convention with Yosemite and El Capitan (the latter being in the former), Sierra and High Sierra.

It’s only in the last few releases that the dart board has come out


Since they moved macOS off version 10.x (finally), "Apple's crack marketing team" left the desert (Mojave) for the Pacific coast. Unfortunately, they didn't plan the trip carefully, so they started at Big Sur with macOS 11, went north to Monterey for macOS 12, then turned around and headed back south to Ventura for macOS 13. At least those locations are in alphabetical order -- but with Ventura they seem to have painted themselves into a corner.

So will macOS 14 be further south (Carlsbad?) or back north (Eureka?) -- stay tuned...


That's fair. In a similar vein, there was also Sierra and High Sierra.


Ubuntu has both a codename and a version number. macOS has a version number too (Ventura is 13) but Apple frustratingly don’t use it prominently.


The version number and build number are on the "About this Mac" screen and in the output of `sw_vers` - where else would you like to see it used?


In marketing material, such as here:

https://www.apple.com/uk/macos/macos-ventura-preview/


I personally doubt you'll ever see that, though it's a possibility since iOS uses version numbers.


Alliterating Antlion


Just use years. You are already releasing shit yearly. Name it macOS 22.

These places mean nothing to me.


The strength of the California brand across the globe is crazy good. It’s all a bit of marketing.


I think you overestimate the knowledge most people have of California. I've been to Cali a bunch of times and none of the names they picked mean anything to me except Big Sur, but that's pretty obscure. I wouldn't expect anyone around me to know what or where it is unless they happened to be really big into the tech scene.

US tech firms have a long history of using US place names as code names for operating system releases. Windows 95 was Chicago, if I recall correctly.


It wasn't much better with big cats. There were two sort of semantically related releases, Leopard/Snow Leopard and Lion/Mountain Lion. Of those on the Leopard/Snow Leopard I thought made sense as Snow Leopard was a "oh shit fix all the bugs" release. SL was the first full OS release after the Intel transition and 64-bit kernel.


I can picture a big cat in my mind. That helped to peg OS X releases and enabled me to mentally distinguish and order them.

I can't picture "Monterey" or "Ventura" or any other macOS names, they have no meaning to me.


Ventura has a meaning to me.... Ace Ventura, a cartoon from my childhood about a egotistical questionably competent detective.

I'm not sure that's the image Apple wants to give off though...


Places in CA have a lot of references in music and movies [1]. Hollywood loves to pimp where their people are from.

[1] https://en.wikipedia.org/wiki/Ventura_Boulevard#References


"Ventura Highway, in the sunshine..." [0]

I may be dating myself somewhat here.

[0] https://en.wikipedia.org/wiki/Ventura_Highway


Ventura was also the home of Chouinard Equipment, later renamed to Black Diamond Equipment, and is the base of Patagonia, a clothing company.


Sunny beach town, Spanish mission, etc.


Big Sur, Monterey, Yosemite, El Capitan, Catalina... all worked for me because I've either been there or there was a screensaver/wallpaper to associate them with their locales. I really don't know a thing about Ventura.


Yes but it starts to be difficult to keep 18 names in the head, be them cats or California places


I just miss the Big Cat names.


Yes. In Big Cat era, you could at least memorise those names which have some meaning to nearly everyone around the world, and it always had a version number.

Mac OS X 10.7 "Lion" – 2010

Mac OS X 10.8 "Mountain Lion" – 2012

Now it is only a name I guess only people in US / California will know or understand. The same joke From Apple's "crack marketing team" and played out by Craig Federighi for something like 10 years[1].

But I guess that is post Steve Jobs's Apple for you.

[1] Just guessing since I remember they started using this line after Forstall left.


Let's go back to cats. Since macOS is becoming more and more like iOS, we don't have to limit ourselves to big cats anymore. Small cats are on the table too. There's gotta be at least 100 cat breeds, that should last us a while.


Small correction:

Mac OS X 10.7 Lion was released in 2011.

Also, there's no "Mac" in OS X 10.8 Mountain Lion.


Hasn't it been this way the whole time? Why is your frustration increasing?


I imagine as the number of versions increase, there is more tracking you have to do when someone doesn't explicitly mention the version number.


That's definitely my issue: over the years, we have accumulated more and more and more of these names. When someone talks about iOS 4 or iOS 7 in some article I am reading, I know what they are talking about and the extent to which the version matters; but, when someone talks about macOS Gaviota, I have to think "wait, was that the one that just came out, or was that one of the ones I haven't had to think about in a decade? oh shit... maybe it's the one that got announced today and I just haven't heard the name yet?!".

(That said, I will also note that frustration is not inherently constant even when something is truly static: sometimes you get used to something over time and it stops bothering you, while other times it slowly drives you mad.)


Haven't you not been getting food this whole time? Why is your frustration increasing?


Help me understand how being starved for food is analogous to having OS version names that don’t reflect the order of OS releases.


The irritant becomes with as time goes on as long as its present. More and more meaningless names to memorize...


Ventura is MacOS 13.0.0.


Ah thx, was thinking about Lino Ventura so a place in Italy I guess; could be worse eg Ponte Vecchio


Future versions will be named Bakersfield, Weed, and Needles, not to mention Oxnard?


nodejs has joined the conversation


Do the new DNSSEC changes imply anything for local DNS overrides? Would Apple refuse to block a site if DNSSEC is enabled and PiHole returns a blocking response?


Probably not. I haven't been able to dig into what they're doing, but I watched the presentation about the feature, and it's an opt-in API for applications. I'm not clear on this but I have to assume that the macOS/iOS resolver code is still leaning on your external recursor to do the DNSSEC validation stuff (otherwise, it's going to generate _a lot_ of extra lookup traffic), which means it's going to trust whatever your PiHole tells it anyways.


So is Rogue Amoeba essentially screwed then?


> Gatekeeper’s role is to ensure that when users execute some code, that code meets the local system policy. The policy includes checks such as whether the code is validly signed and whether it has been tampered with in certain ways.

Weasel-word alert. I never thought I’d see the day when technologists would applaud the gradual death of general-purpose computing, but here we are. A decade from now Apple probably won’t even ship a local version of Xcode, and the transformation will be complete as all new development happens in Xcode Cloud where no line of code goes unscrutinized by the watchful eye of the mother ship. At least we’ll be Safe™.


It is funny to me to read this because I recently joined the Xcode Cloud team to precisely work on this, thinking that I could help Apple make developer’s life easier in the near future, but according to your comment, there are people out there who will consider my team’s work a regression.


There is no problem with a remote compilation. If anything, it's good to have options.

The problem if someone decides that it's going to be the only option. And another problem is that they can.

So, your work is not regression, and it definitely has a positive use case. It's just that it can also make certain unethical things possible.


> It is funny to me to read this because I recently joined the Xcode Cloud team to precisely work on this, thinking that I could help Apple make developer’s life easier in the near future, but according to your comment, there are people out there who will consider my team’s work a regression.

how exactly will it be easier than my current workflow of:

- Boot computer

- Press win-key + d

- type the letters "qtc"

- hit enter

- ctrl-alt-shift-<index the project I'm working on>

- ready to code


Not having to download 10g to edit 1 line of code on a computer where you haven't set up XCode should be a win?

Having said that, I agree, the biggest problem here is that even if it doesn't seem obvious now, once the cloud offering is there the control it offers will make it very appealing for Apple to expand its use and eventually offer features there that aren't in the real XCode. It can fast be a slippery slope to the non-cloud app being deprecated.


I disagree. Before Gatekeeper there was no way to define a policy about what code could run, now with Gatekeeper there is. Currently, Apple define a default policy. Users can edit this themselves if they acknowledge the risks. Admins of Macs can also set their own policies.

The ability to have policies is very different from enforcing overly strong policies. Apple seems quite clear that they see iOS as being a platform with a stronger policy, and macOS as being a platform with at least the ability to run a weaker policy.

Edit: also Xcode Cloud isn't what you imply it is/could be, and Apple's moves with Swift being developed in the open suggest to me a very different direction for development. I can't see this ever being locked down, either in terms of technology or policy.


Gatekeeper was introduced in 2011 in Lion. It has already been a decade. I didn't see that transformation happening.


The existence of Gatekeeper already causes a huge privacy violation by “requiring” my computer to phone home to verify the signature the first time it launches an application. Everyone should have realized this when Apple’s OCSP responder went down in November 2020 and nobody could launch anything that wasn’t built in to the OS.

According to TFA this kind of verification will now occur every time an application is launched to deter post-verification “tampering” by you, the user. How big of a privacy violation would it have to become to bother you, out of curiosity? If we let this continue we will end up in a future where full “Remote Attestation” of every hardware and software component is required to participate in the Internet. This isn’t hypothetical doom-saying, either: game consoles already work like this. I remember my XBOX360 could detect modified DVD drive firmware, launches of individual pieces of software (e.g. Halo 3 Delta leak), and other types of system modifications, then it would permanently ban that machine from XBOX LIVE. And that was all 15+ years ago.

Just imagine what a gift this will be to law enforcement, for example, once they can go to Apple all like “Hey, Siri, show me all users of Tor Browser around the time of ${BITCOIN_TRANSACTION_ID}”.


I thought all UNIX heads longed for the days we used to telnet or startx into the UNIX development server, everyone had their $HOME configured with noexec, and the tooling configured by the BOFH team.


> A decade from now Apple probably won’t even ship a local version of Xcode, and the transformation will be complete as all new development happens in Xcode Cloud where no line of code goes unscrutinized by the watchful eye of the mother ship.

Any young folks wondering: yes, this exact same thing was being posted ten years ago, all the time.

"Apple's gonna totally lock down macOS without any way around it, they hate general purpose computing" and the related "Apple's gonna merge iOS and macOS" are the apocalypse cult of computer geek forums. They might be right eventually, but only after being wrong a hundred times. And they never get the timeline right.


Do you not see that every new macOS release is a few more steps towards that dystopia?


Started kinda late on macOS/OSX (in my career, and in the life of Apple) but have still been using it for 11ish years, and developing, off and on, for iOS.

Nope, I don't.


There is no more prescient of a take on this news as this one.

Once the singularity is nearly complete you'll know: macOS and iOS will merge into one monolithic OS.


Somewhere between Mountain Lion and High Sierra, it became impossible to delete or even mark non-executable various annoying built-in applications which I never use, e.g. iTunes.app and Safari.app, which often open without me asking them to.

Does anyone know how to re-enable this functionality?


As of Big Sur it's impossible. The entire OS is now a single cryptographically signed image that's verified on boot, which is great for security but it includes some things that don't particularly need to be locked down.


5-6 year support for hardware is just too short. I have a Mid 2015 Macbook Pro as a "backup" and that computer is still quite decent.


See reply below about Open Core Legacy Patcher[1] which enables to use older Intel Models to use modern OS Versions. And for even older hardware, check out the patchers from Dosdude[2]. Most likely this will not be possible with the M-Class Processors from Apple, which is a shame.

[1]https://dortania.github.io/OpenCore-Legacy-Patcher/MODELS.ht...

[2]http://dosdude1.com/


I wonder if it’s worse this year than usual because of the switchover to AppleSilicon? I have a 2013 Mac Pro that I use daily and it’s gotten all updates until this one. 9 years is pretty good. But I can understand wanting to end support for as many Intel machines as quickly as possible.


When Apple announced the transition from PPC to Intel, the PPC Macs only got two major updates before being EOLed. Even the Power Mac G5 (2005) only got two major updates before being cut off from support.


It should be noted, however, that those major updates had longer lifespans then. Snow Leopard, the first Intel-only version of Mac OS, came out in 2009.


MacOS versions tend to receive security updates for 2 additional years after they’re supplanted, so it’s more like 7-8 years. Plus, all of the Macs that aren’t eligible for Ventura can use Bootcamp to install Windows or dual boot into Linux if you’re not happy with MacOS anymore.


The newest machine that can't run the new version would be a Mac Pro purchased a scant 4 years ago. Potentially for several thousand dollars.

Meanwhile there are 10 year old ~$1000 Thinkpads running Windows 11 or Linux. If they just wanted to run Linux on it they could have saved themselves some money.

8 years of updates to current version and 10 years of security updates should be the absolute minimum for every expensive hardware.


I question anyone purchasing a machine with generations old hardware in 2018. Especially considering the Mac pro had been supplanted by newer macs at that point.


In 2018 CPUs were still stagnating, so a 2013 CPU would likely be almost as performant as the latest one.


It's not really about the CPU power, more that by 2018 it was well known that the Trashcan Mac Pro was a dead-end design and a soon to have radical refresh.


Absolutely correct it would have been a shitty purchase but one which ought to be supported none the less.

Companies ought to love people who give them thousands for outdated hardware worth hundreds.


They haven’t yet released a machine worth upgrading to if you’re on a 2015 MBP. I recently got a 14” M1 and I think I’ll just reinstall my 2015 and move back to it.


You can’t be serious? The m1 is so good, great performance, much quieter and doesn’t get hot, and the battery life is amazing. Also there’s not many compatibility issues now


…absolutely not. I jumped from a 2015 to an M1 and never looked back. It’s 100% worth the upgrade.


The 14" M1 is the spiritual successor to the 2015 MBP I think. Ports, magsafe charger (with higher quality cable), larger screen within a marginally larger footprint, keyboard is good unlike the last few years, solidly built, no touch bar, finger print login, performance, etc. No complaints personally.


I upgraded from a 2015 15" MBP to a 2021 16" MBP. I'm loving the upgrade. Fast. Better display. Much better battery life. Still has magsafe. The keyboard is acceptable. What do you think is missing and/or bad about the new models?


I’d wager that the vast majority of people who have made that upgrade would disagree with that view; it’s a substantial step-change in most ways, if you use the device for software development or content creation.


That’s certainly a hot take. I’m not sure many would share that opinion.


Battery life and speed alone are totally noticeable. This doesn't mention that Intel fires up the fans anytime a CPU calculation is done. I think you're being disingenuous here.


I upvoted you because you're entitled your opinion, but I whole-heartedly disagree.


Nothing terribly major here, it sounds like. Making login items visible is a long-overdue change... but none of these are going to have any particular impact on average users or average apps.

On a pettier note, can we get a better source than a website that's using JS to change its title when it doesn't have focus to try to gain attention? (It toggles about every second between "macOS Ventura | 7 New Security Changes to Be Aware Of" and "Message from SentinelOne". https://imgur.com/ynPqpvK - it's pretty awful.) I don't normally complain about scummy websites on here, but this is just annoying.

Interestingly, I went looking for alternative sources for the content, and found that identical content is on other sites [1] which are also doing the same title-flicker technique. So presumably this is part of some content network...

[1]: https://phxtechsol.com/2022/06/13/apples-macos-ventura-7-new...


I'm not surprised they're using scummy tactics. Their actual software runs like crap, so gotta do whatever they can to get users. I had a work MBP and personal MBP with exactly the same specs, main difference being the Sentinel One agent. The work one was constantly spinning up the fans, S1 was gobbling up memory, and support was completely useless in diagnosing. Their Linux agent isn't much better with constant memory leaks.


There’s a lot of changes under the hood that this post doesn’t mention.


I noticed it even did that in Reader Mode in Safari. Really annoying.


Mostly off-topic: have Apple and Microsoft completely given up on non-trivial changes to desktop operating systems? Will MacOS look basically the same in 40 years? Or is the idea that everything will be AR/VR by then and there is no use innovating in this domain?


I guess you'd need to set a goal post of what counts as trivial?

Many of the things mentioned in the article aren't trivial. They may be smaller in scope, but size (large / small) are different than complexity.

You can take a look at what's new in Ventura https://www.apple.com/macos/macos-ventura-preview/features/ but that's not even getting into the under pinnings.

Similarly Microsoft made fairly significant changes to Windows between 10 and 11, and several times to 10 within its life cycle.

Unless you're talking purely visual design, in which case what kind of changes would you expect without upending people's workflow?


We have gotten so use to these trivial changes that our expectations have renormalized. Desktop OSs are asymptotically approaching a fixed point.

Examples of modest but non-trivial changes:

- eliminate the folder-file system (or at least make it completely invisible to the user)

- remove UI distinction (but not necessarily the sandbox distinction) between web apps and normal apps.

- seamless mobile-desktop integration, so the user views them as just different form factors for accessing the same resources.

(There are of course much more radical changes than these that one could imagine.)


I'm curious as to your background if you consider any of the things mentioned in the articles as "trivial" changes. Have you worked in systems development before?

I similarly question your definition of "modest". The first one alone is incredibly radical, and has been tried several times in the past but people keep asking for hierarchical file systems. It's far from modest.

1. How do you propose users organize things?

2. Already exists today with electron and webview. What would you propose an OS provide here? Many apps you use today on macOS are web apps within a native context.

3. This is already growing on macOS with features like continuity handofd, universal control, being able to run mobile apps on desktop, iCloud sync of projects etc.. Each year they've clearly moved towards unifying things.

If these are what you consider modest though, I fear what you consider radical without throwing out decades of learned user interaction in the process


I’m not using “trivial” as a measure of ease of back-end implementation, but rather of how it actually changes user experience and productivity. There is no limit to how hard it can become to implement trivial changes behind the scenes; it would be silly to ignore or downplay the ossification of desktop OS capability just because software developers continue to expend more and more effort to make smaller and smaller improvements.

My reading of your comment is that you aren’t actually interested in thinking about non-trivial changes here. “Didn’t you know people have tried eliminating folder systems before? It’s hard and hasn’t succeeded yet” is obvious and does not seriously engage with the possibility. (“Didn’t you know people have been attempting to make stylus input work for decades without success?”) Likewise, the fact that web apps can be disguised as native apps is not the same thing as eliminating the distinction at the user level, and I don’t think you would have conflated these if you were really interested in it.

So I don’t think it will be productive to continue this conversation.


Again, that's why I'm delineating between scale and complexity. Trivial implies complexity, but you seem to keep going back to scale of the change.

Saying something is trivial, by definition, implies its a simple change. Nothing mentioned so far is simple. None of your suggestions were modest.

I understand you're using the word according to how you think of it, but I'm trying to point out that you're incorrect, and that many of the things you say are modest are not so.

You're actively down playing the amount of work and it either feels disingenuous to make your point, or divorced from the reality of implementation.


Swap “everything will be AR/VR” to “everyone uses mobile as primary devices” and I think it’s a better guess. Or maybe “everyone have at least two computer devices”. I feel like the goal is to have the most seamless experience between tablets, smartphones and desktop, and impactful changes that don’t work towards that goal are just discarded.


Window management has continued to evolve in nontrivial ways, imo. More fundamental interactions probably won't and probably shouldn't change; those idioms are mature and deeply engrained at this point. It would alienate swathes of users to rock such an established boat.


Microsoft tried with Windows on their phones. Look where that led them.


To 10% market share in Europe right before they decided to drop it all.

It was starting to become the alternative to Android for many of us.

Had they provided a proper migration path from Windows Phone 7 into 8, and then from 8.x UA model into 10 UWP, and more Win developers would have followed along, instead of hating them for all the rewrites.


If you look at independent developers working on the Mac OS, you find it's pretty much dead. Only the name programs get updated nowadays whereas everybody else has moved to iPhone/iPad as that's where the money is. Open source still chugs along though.


Do you have specific improvements in mind, or are you saying you want them to change stuff just for the sake of changing it?


The security engineering division at Apple intends to destroy this product.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: