Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
This is actually an argument in favor of public disclosure of vulnerabilities. Lemi Orhan Ergin was catching a lot of criticism yesterday for posting it on twitter, but if this bug had been reported privately, it would have taken much longer to fix, while malicious actors would be able to exploit it all along.

It was "reported" about a month ago on Apple's support forum. So it does seem likely some people out there were aware of it before Apple became aware.
 
Too bad the guy who found the bug had to go public with it right away. I'm sure Apple would have fixed it quickly either way, but now the secret is out and all the updated machines will be vulnerable. Lots of people don't do regular updates.
 
You ever heard of Windows? Perhaps you should read up on that OS if you haven't.

Also, give me a break. Nobody finds everything, not even "Apple". Patched quickly and painlessly. Move along.

Really? You are 'what about' arguing this already? This isn't some complex, hard to find, hard to implement or questionable security hole.
 
That was quick

No, it wasn't. The bug got posted on the Apple forums two weeks ago already, as you can read here:
https://forums.developer.apple.com/thread/79235 (browse to november 13th, a post from chethan177). Maybe the issue got posted somewhere else before as well.

Unfortunately the Apple devs don't read those forums. And appearantly no one at Apple is reading those forums. Maybe this will change as having the bug fixed two weeks ago would have prevented a lot of damage / loss of goodwill.
 
So the fix changes the password (puts one in) and disables root. Okay. how did 13.x ship with that flaw?
Maybe reread the announcements. You seem confused on the basic details.
Effects 10.13.1 specifically.
The fix does not put in a password.
[doublepost=1511985506][/doublepost]
currently running MacOS 10.13.2 beta 4 public
now eagerly awaiting security patch

I tried this and was able after a couple of attempts to gain access to login options for example - though not sure how a third party would access this unless i handed over my laptop whist logged in and screen active.

That said, this is a surprise!

I wonder why it takes a few tries to succeed with ID root and a blank password? It hasn't worked first time in the few attempts I've made just now.

If you have screen sharing or remote management enabled...
 
Actually looking at links above @Westside guy posted a link to a discussion a few weeks ago on the Apple dev forum where the issue was discussed but people didn't really seem to have a handle on it. Interesting.

Meanwhile @tkermit link above shows a new discussion here saying AFP file sharing is not working for a couple of users, including him/herself.
This problem was so extraordinary stupid that hackers didn’t look for it, and that serious developers saw it as a way that could help them solving people’s problems. It’s like leaving your keys in the front door, and the postman is happy that he can put your packages indoors when you’re not at home - since he is an honest person, it doesn’t even occur to him that you could be robbed.
 
I will update, even tho this root issue not present with 10.13 Beta (17A362a)
Just checked it and no access with root, denied so must of been an update to 10.13 that introduce this bug but be updating to the latest 10.13.
 
Still, this mistake by Apple is amateurish. Checking for lack of passwords for pre-installed admin level accounts like this (that are not a part of the first power-on setup sequence) is one of the first things that any test bucket should include, automated or no.

The fact that a company like Apple was not checking for this is completely facepalm-worthy.
That’s what you say with hindsight. It’s a great idea after the fact. Nobody had this great idea before it happened.
 
I noticed last night after changing password for root permissions when copying files from NAS to Mac were locked. Didn’t happen 2 hours earlier when I copied some files over.
 
Seems a bit rushed. I hope they tested it thoroughly like they did the original vulnerability.

Seriously, your complaining that they fixed a major vulnerability too quickly? Had it taken an extra day or two, I’m sure you would complain about that as well. I imagine Apple can do no right in your eyes.
 
I didn't turn much up (anything, actually) looking at the Console just now but you can start by narrowing the list of "who" to anyone who had physical access and a login in the last 24 hours.
So you need physical access and a login password to activate this bug........ummmm. Not a problem in this household.
 
I patched the vulnerability manually a few minutes after the news broke. Permanent fix showed up the next day. Issue resolved for good. Quick and painless (no reboot required). Good work Apple.

And now back to our regularly scheduled trolls whining about this even though the real world has moved on.
 
No, it wasn't. The bug got posted on the Apple forums two weeks ago already, as you can read here:
https://forums.developer.apple.com/thread/79235 (browse to november 13th, a post from chethan177). Maybe the issue got posted somewhere else before as well.

Unfortunately the Apple devs don't read those forums. And appearantly no one at Apple is reading those forums. Maybe this will change as having the bug fixed two weeks ago would have prevented a lot of damage / loss of goodwill.
Fair enough, but it was fixed quickly after it became a big story
 
Android is a mobile operating system and you are also forgetting the fact that Google supports it's own devices for at least as long as Apple does.
Google fixes all of it's security issues, but most hardware producers don't fix their Android software as soon as the phone is off the market.

Why does a mobile OS make it okay to have major security flaws? There’s a large chunk of the population who’s only “computer” is their phone.

Also, Google only guarantees support for their Pixel line for 3 years - 2 years of feature updates and another year of security updates. Currently, Apple still supports the iPhone 5S, which is 4 years old - and it’ll probably get iOS 12 next years as Apple typically supports their phones for 5 years.
 
Firing good people for bad mistakes they've since learned from is just not smart, even if you discover multiple past mistakes.

They'll still be out there. They still have the same skills, and now they're more cautious. And someone else is benefiting from that experience.

Also, their mistakes are still in the code, and now nobody's around that will suddenly remember those mistakes.
This mistake is more on the end of people who gave everything final approval, not on the one guy who happened to write those particular lines. (Also, bugs don't get found by someone "remembering" them.) These two problems are BIG, they aren't something a company should shrug their shoulders about and say "now we know better for next time". These are two black eyes on Apple's security reputation.
 
Still, this mistake by Apple is amateurish. Checking for lack of passwords for pre-installed admin level accounts like this (that are not a part of the first power-on setup sequence) is one of the first things that any test bucket should include, automated or no.

The fact that a company like Apple was not checking for this is completely facepalm-worthy.
Apparently the bug is a bit more complicated than it looks. According to this preliminary analysis:

https://objective-see.com/blog/blog_0x24.html

it seems MacOS first tries to update the root account to a shadow password, but erroneously sets the user-supplied password (blank or otherwise) without proper verification. After that, you can log in with the supplied password. This explains why you have to click "unlock" twice (first the account is updated and activated, the second click actually logs you in), and why it works not just with a blank but with any password.
 
  • Like
Reactions: Mescagnus
This problem was so extraordinary stupid that hackers didn’t look for it, and that serious developers saw it as a way that could help them solving people’s problems. It’s like leaving your keys in the front door, and the postman is happy that he can put your packages indoors when you’re not at home - since he is an honest person, it doesn’t even occur to him that you could be robbed.

Indeed. It's so spectacularly unlikely you'd never think to test it, manually or automatically.
 
I'm glad this was addressed so quickly for those running HS, but is something that should not have happened, along with some other problems I read about in HS. They have been getting very sloppy. I haven't upgraded to High Sierra yet because of my lack of faith in current apple, for reasons like this, when I used to upgrade much sooner.

I hope what they said about auditing and reflecting more are sincere as I really feel they need some serious introspection in the whole company and not just MacOS development.
 
What kind of bs is that explanation that makes this update looks like a complex task? An exception to a login process is one line of code. Find the line, hit the delete button. Done. 45 s.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.