Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
macOS Ventura 13.5 Breaks Location Services Permissions for Apps
- Thread starter MacRumors
- Start date
- Sort by reaction score
That's a judgement call. In my opinion, it's late. I'm the one who raised the issue with the editors (indirectly through their site feedback forum), since I was getting annoyed that they hadn't given it some visibility. There's a very active thread on these forums complaining about the bug since 7/24; it was constantly visible on the front page. It's such a significant bug that Apple should have had a fix pushed out a day or two after it was discovered.
I am appreciative that they followed through when they were made aware. In the end it was my fault; I should have just messaged them rather than getting annoyed.
My first thought when I saw the headline was “how is this just being posted now”
I thought my feed was out of order or something
It’s definitely late
So late it was already fixed in Sonoma so I didn’t realize it wasn’t fixed in Ventura
So they released an obvious bug. So what; I know the development team who delivered the new System Settings are incompetent (as a group). So, so, so, so much more significant is that they haven't fixed it. They've coded themselves into some corner and can't find a way out. (I've been there.) Their inability to handle this bug should leave everyone very afraid.
The big story isn't the bug; it's that they haven't fixed it and what that implies. That's the story that I wish had made the front page.
The big story isn't the bug; it's that they haven't fixed it and what that implies. That's the story that I wish had made the front page.
Not much. It's been three weeks; they don't usually release macOS updates more frequently than that.
Imagine Apple pushed out an update that disabled their office apps: Numbers, Pages, and Keynote. Then imagine there was no update for three weeks. Would you say there's not much implication in that? Would you say that's just normal and expected behavior? I suspect they would have pushed out an update in the following couple of days.
So my intuition is based on how serious I think the bug is. I consider this a pretty serious bug. You might think it's just a minor bug, deferred to the usual update time. Applications are malfunctioning and key features of their privacy value proposition are not working.
But, there's no way to argue that without getting inside Apple. I think they're in panic mode. But, maybe they just fixed it quickly and are just waiting for the next scheduled patch.
They are relying too much on automated testing and public beta testers. Just like Microsoft. Microsoft famously, publicly fired all testers and now all testing is done by “insiders.”
I don’t remember seeing a similar story about Apple but clearly testing has been identified as an area to cut costs.
Automated testing and volunteers can’t be expected to be up to the same quality as trained, experienced software testers.
Sonoma seems pretty solid, maybe they improved the process for that but at this point Apple updates have just as bad a reputation as Microsoft updates.
If the bug were more severe than it is, then yes, that would put more pressure to do a release.
Nevertheless, they generally don't rush non-security releases, probably for various bureaucratic reasons.
Well, only the configuration of privacy settings is broken. The privacy settings still apply.
Probably.
This is arguably a case of too little automated testing.
A unit test like Assert.that(displayedApps.count, is.greaterThan(0)), say.
(Now, that's easy to say after the fact, and from the outside looking in. OTOH, they just completely rewrote the settings app. Hopefully, they designed the new one for easier test automation?)
I doubt it.
It's puzzling how this has slipped through.
I guess all but one engineer is working on the glasses.
Location prefs - sorry settings - completely borked on my MBP.
And News keeps on asking me to give it permission to access my location, even though I keep on telling it no.
I guess all but one engineer is working on the glasses.
Location prefs - sorry settings - completely borked on my MBP.
And News keeps on asking me to give it permission to access my location, even though I keep on telling it no.
A user's ability to control their privacy settings is the advertised feature.
ð microsoft fires testers 2014 - Google Search
www.google.com
Don’t doubt it. Just Google it. Did you think I made it up?
And yeah this probably could have been caught by an automated test, but the version of System Settings they shipped was not exactly a paragon of refined software development in the first place. Kind of shows they were mandated to make it that way and are still figuring out how to actually make that work.
ð microsoft fires testers 2014 - Google Search
www.google.com
Don’t doubt it. Just Google it. Did you think I made it up?
No. I’m aware of the rumor.
But sure, let’s take one of those articles. “Before 2014, Microsoft had a team devoted to testing Windows in real-world situations and analyzing the results. [..] it didn’t bring back the team.”
Microsoft no longer having a team to test scenarios is a very different assertion than ”Microsoft famously, publicly fired all testers”. No, they infamously, privately, supposedly fired one specific team of testers.
Well, it’s late in my opinion. This bug has been around since the 13.5 release which came out more than two weeks ago, and hasn’t been fixed yet. And there has already been a whole active thread about it on this forum which is the one linked in the article. It’s very strange how it hasn’t made it to a front article on this site until now, but at least it’s now there.
The bug should have been at least reported earlier by the site so that it gets more recognition, regardless of whether or not it’d be fixed in a rapid release. Just like how the issue with the Meta sites on the previous (a) releases was quickly reported by this site on the same day, which resulted in Apple quickly pulling out the (a) update and releasing the (c) update instead to fix the issue.
Split hairs I guess. They announced that they were doing it. The specific team was Windows testers. Here’s one of the articles from that search from someone there who says this. Plus my experience with Windows this entire time leads me to believe that all this smoke has some fire to it.
Former Microsoft Employee explains why bugs in Windows updates increased - gHacks Tech News
Has the number of bugs in Windows updates increased in the past couple of years? If so, what is the reason for the increase in bugs?
www.ghacks.net
There will be some Macs that can't be upgraded to Sonoma die to their age so I'd imagine it'll be a 13.5.1 fix otherwise it's going to leave a lot of people stuck.
They announced that they "publicly fired all testers"?
Here’s one of the articles from that search from someone there who says this. Plus my experience with Windows this entire time leads me to believe that all this smoke has some fire to it.
Former Microsoft Employee explains why bugs in Windows updates increased - gHacks Tech News
Has the number of bugs in Windows updates increased in the past couple of years? If so, what is the reason for the increase in bugs?
So, a disgruntled former staffer is disgruntled with their former employer. Film at eleven.
And, again, that article doesn't say that they "fired all testers". It says that they — supposedly — no longer apply one specific testing approach. We don't know if that's true, nor do we know that it means that all Windows QA teams are gone.
That's not "splitting hairs".
You really got me thinking. I'm curious what others think. If they had solved this bug 3 weeks ago, do people think they would have sat on the fix until some scheduled patch?
Short answer is yes.
There is of course the scenario where there are no significant bugs in 13.5 at all, in which case a 13.5.1 isn't scheduled. I presume one is scheduled now, for various reasons, including this bug. But judging from a) how they seem to be doing release management and b) how I would be doing it, yes, I think they'll wait to collect enough fixes to make a release worth it.
The exception to this is security patches. Even for those, they tend to wait until they have several in one go, but if a security issue is severe and actively exploited, they may make an out-of-band release, now that they have their RSR mechanism. (Which doesn't work for all scenarios. So, for example, if a bug is severe, and actively exploited, but in the kernel itself, they won't do an RSR. Instead, they'll wait for the next regular release.)
Do they expedite releases if they contain important fixes? Maybe, but I think not by much. I rarely see them do "stop the presses; we gotta release something tonight" stuff.
I wonder if your opinion in this is generally shared by others.
When I called Apple support a week after the problem started, the person said that they were aware of the issue and they were hard at work trying to solve it. He might have been making that up. But still, your assertion that I can't conclude anything from a three-week delay (so far) in isolation is independent of that.
We don't do release management in the way you describe. High user impact gets expedited. We usually get a patch out within two days. Perhaps Apple concluded this wasn't high user impact.
In any case, again, I'm quite curious if others share your opinion, that Apple might just be sitting on the fix.
Yeah, they better fix it on a Ventura release, or else the older Macs stuck on Ventura will continue to have this bug. I hope Apple will fixes it very soon.
The public Sonoma release is coming in less than a month though, so I’m not really sure when they could possibly release a 13.5.1 fix update..
I don’t read those investors mind. If they vote for AI QA bot, no matter how bad it is, Apple will go for it. It’s this simple. I‘m not hinting AI is better. I’m thinking maybe AI can save some cost.
The specific testing approach is to approach it with far fewer dedicated testing employees. They did publicly announce that when they publicly announced the layoffs. I didn’t say all QA teams are gone. You glossed over this point, that’s why I said that.
I will admit that I didn’t realize this was nearly ten years ago now, and I didn’t do enough research to find a definitive quote to give you.
I don’t think they started just shoving it out the door totally untested, but I do think they cut back on the best testing method in order to save money, and I suspect Apple did the same. That’s all I’m saying.
That could well be true, yes.
I'm less sure whether we can draw any conclusions between that and this particular bug. Yes, testing should've caught it — whether internal or external, automated or manual — and either it didn't, or it did but the release manager said the bug wasn't severe enough to block the release.
But, regardless of that, I'm more curious why this regression occurred, and in a minor release, to boot. It suggests that 13.5 contains a significant changes of TCC. Perhaps a failed backport from 14.0.