- Related Stories
-
Apple plugs security hole in iTunes
May 10, 2005 -
Apple's Tiger on the prowl
May 4, 2005 -
Apple patches a batch of Mac OS X flaws
May 4, 2005
Still, only a handful of the vulnerabilities are of major concern, according to security analysts. The package of fixes was released Monday.
"This one is a big update. I don't recall seeing as many updates as we see today," said Thomas Kristensen, Secunia's chief technology officer.
By comparison, Apple last May released an update for 20 vulnerabilities and in March distributed an update for a dozen flaws.
But Kristensen noted that, with the new update, only a few of the 44 vulnerabilities are of great concern. He also said that 25 percent of the patches involve older vulnerabilities that have yet to lead to exploit code being developed by attackers. Still, Secunia is rating the overall update as "highly critical."
Apple declined to comment on the vulnerabilities and referred all questions to its security update.
The flaws affect Apple's Mac OS 10.3.9 and 10.4.2 operating system software and related server software.
Kristensen said that some vulnerabilities involving AppKit and Safari are critical.
AppKit, which is used to open RTFs (rich text files) and Word documents, has flaws that allow a remote attacker to create a malicious file that results in a buffer overflow. That in turn can lead to arbitrary code being executed on a user's system.
Apple, however, notes that only some applications use AppKit, and that Microsoft Word for Mac OS X is not vulnerable.
Flaws in Safari, meanwhile, can allow an attacker to bypass the browser's security checks and execute arbitrary commands, when the user clicks on a maliciously crafted rich text file.
Another flaw, a vulnerability in Apple's Sever Manager D, a modified version of Apache, is also being considered critical by some.
That flaw can result in a buffer overflow and remote execution of code by an attacker, with no user interaction, said Frank Nagle, assistant director of vulnerability aggregation for iDefense, a VeriSign company.
Although Apple lists other security flaws that could be exploited by a remote attacker, they are "less critical," according to Secunia.
For example, two vulnerabilities in Apache 2 could be exploited by a remote attacker to either bypass security restrictions or launch a denial-of-service attack.
But Apple did not set Apache 2 by default, so it is less of an issue than it would be if the same vulnerabilities affected Apache 1.3, Nagle said.
See more CNET content tagged:
Apache Software Foundation, flaw, attacker, vulnerability, Apple Computer
Perhaps you meant Mac OS X 10.3.9?
Perhaps you meant Mac OS X 10.3.9?
Answer: There are only 4 Macs in the whole world connected to the internet....so it's not worth the effort to write an exploit. HA HA HA
Sorry, when you said to prepare for the flames I couldn't resist trolling and posting :):)
I like Macs. I have nothing against them. I am the first to admit that MSFT has a horrible security track record...Just had to drop in and push a few buttons. :)
Does that mean the MS supporters are less committed or that the Mac supporters need to be committed? :)
How does that make a difference? As is proven (again) by this article, both systems have flaws. So how is it Microsoft's fault that their flaws happen to be the most exploited? It makes perfect sense that the most used system would be the most attacked. Generally an attacker's goal is to cause as much damage and trouble as possible, and to do that, it makes sense to exploit the flaws found in the system that the majority of people are using, not the minority. If two products have flaws, one isn't automatically better than the other just because the most used of the two gets attacked.
Answer: There are only 4 Macs in the whole world connected to the internet....so it's not worth the effort to write an exploit. HA HA HA
Sorry, when you said to prepare for the flames I couldn't resist trolling and posting :):)
I like Macs. I have nothing against them. I am the first to admit that MSFT has a horrible security track record...Just had to drop in and push a few buttons. :)
Does that mean the MS supporters are less committed or that the Mac supporters need to be committed? :)
How does that make a difference? As is proven (again) by this article, both systems have flaws. So how is it Microsoft's fault that their flaws happen to be the most exploited? It makes perfect sense that the most used system would be the most attacked. Generally an attacker's goal is to cause as much damage and trouble as possible, and to do that, it makes sense to exploit the flaws found in the system that the majority of people are using, not the minority. If two products have flaws, one isn't automatically better than the other just because the most used of the two gets attacked.
"Virus writers will be gone soo-nia,
because we're the company named Secu-nia,
we have no software or vaccu-nias,
because we are Secunia."
them before, and I don't expect to again - unless someone can tell
me why I should.
"Virus writers will be gone soo-nia,
because we're the company named Secu-nia,
we have no software or vaccu-nias,
because we are Secunia."
them before, and I don't expect to again - unless someone can tell
me why I should.
virus writer's tattered underwear. TRY to be relevant when you
post.
I can tell you never used Windows 1.0
virus writer's tattered underwear. TRY to be relevant when you
post.
I can tell you never used Windows 1.0
If there is an autoupdate feature in OS X I am not aware of it because I do not care to shell out $500 for the lowest end mac that has crappy hardware and cant be upgraded.
You post blah- blah but you lack a true grip on reality when
slamming Apple's OS. You just repeat IT watercooler myths.
Grow up and educate yourself before you offer an opinion.
Otherwise you will continue to sound silly.
If there is an autoupdate feature in OS X I am not aware of it because I do not care to shell out $500 for the lowest end mac that has crappy hardware and cant be upgraded.
You post blah- blah but you lack a true grip on reality when
slamming Apple's OS. You just repeat IT watercooler myths.
Grow up and educate yourself before you offer an opinion.
Otherwise you will continue to sound silly.
I wonder how many...43 maybe? :)
I wonder how many...43 maybe? :)
> malicious file that results in a buffer overflow. That
> in turn can lead to arbitrary code being executed
> on a user's system.
I see this statement all the time applied to UNIX systems. I don't
think it's true. Buffers in UNIX programs are always allocated in a
data segment (as opposed to a text segment where code lives).
This includes the stack frames used for local procedural storage.
It is not even possible to write to the text segments of a
program; UNIX in conjunction with the hardware MMU make sure
that it can't happen -an attempt to do so will immediately fault
the app.
By the same token, it is impossible to execute data in the data
segment, unless the application calls a special
function ::MakeDataExecutable() (at least on the Mac). If the
programmer never calls ::MakeDataExecutable(), which is by far
the norm, then you can overflow buffers all day long and it will
NEVER cause any harm. It simply can't by definition. That is one
of the inherent security features of UNIX and one of the big
reasons it is virtually virus free.
So to reiterate, executable code lives in text segments and can
only be read and executed -never wriiten to. Buffers and all
other data live in data segments, which can only be read and
written -never executed. All of this is strictly enforced by
hardware with immediate death to any and all violaters.
Hence data in a buffer, which by definition must be in the data
segment -even if the buffer overflows, cannot be executed.
Overflow or not.
So, you've either been misinformed, or you're making things up.
And furthermore, why would Apple bother releasing a patch for a security hole if there was no security hole?
> malicious file that results in a buffer overflow. That
> in turn can lead to arbitrary code being executed
> on a user's system.
I see this statement all the time applied to UNIX systems. I don't
think it's true. Buffers in UNIX programs are always allocated in a
data segment (as opposed to a text segment where code lives).
This includes the stack frames used for local procedural storage.
It is not even possible to write to the text segments of a
program; UNIX in conjunction with the hardware MMU make sure
that it can't happen -an attempt to do so will immediately fault
the app.
By the same token, it is impossible to execute data in the data
segment, unless the application calls a special
function ::MakeDataExecutable() (at least on the Mac). If the
programmer never calls ::MakeDataExecutable(), which is by far
the norm, then you can overflow buffers all day long and it will
NEVER cause any harm. It simply can't by definition. That is one
of the inherent security features of UNIX and one of the big
reasons it is virtually virus free.
So to reiterate, executable code lives in text segments and can
only be read and executed -never wriiten to. Buffers and all
other data live in data segments, which can only be read and
written -never executed. All of this is strictly enforced by
hardware with immediate death to any and all violaters.
Hence data in a buffer, which by definition must be in the data
segment -even if the buffer overflows, cannot be executed.
Overflow or not.
So, you've either been misinformed, or you're making things up.
And furthermore, why would Apple bother releasing a patch for a security hole if there was no security hole?
missing? All I am saying is that a buffer overflow cannot result in
unauthorized code being executed in UNIX. Shouting at me,
calling me "clueless" (extremely unprofessional behavior for a
newsdotcom representative by the way) does not in any way
refute the assertion.
The "you just don't know what you are talking about" argument"
might be good enough for cable but somewhat lacking in a
technical forum. Working in the ROM and kernel for 10 years on
OS X and 20 years or so on OS 9 has given me some insight into
how these things work.
I did not realize what kind of people were behind this
organization, but I suppose I was too busy looking at the facts
(sorry "trees") wasn't I?
previous thread. It ended up out here instead. Oh well.
- News.com Shows Its True Colors.
- by Mystigo August 18, 2005 7:53 PM PDT
- I certainly can't follow your logic. What part of the thread am I
- Like this Reply to this comment
-
-
- Reply Misdirected
- by Mystigo August 18, 2005 7:59 PM PDT
- I meant to reply to the news.com "commentator"'s comments in the
- Like this
-
- Impressive
- by Andrew J Glina August 19, 2005 12:41 AM PDT
- Ten years working on MacOS X? How? It came out in 2001, and even the Beta was not out until 2000. Methinks you are trying too hard to impress.
- Like this
-
(100 Comments)missing? All I am saying is that a buffer overflow cannot result in
unauthorized code being executed in UNIX. Shouting at me,
calling me "clueless" (extremely unprofessional behavior for a
newsdotcom representative by the way) does not in any way
refute the assertion.
The "you just don't know what you are talking about" argument"
might be good enough for cable but somewhat lacking in a
technical forum. Working in the ROM and kernel for 10 years on
OS X and 20 years or so on OS 9 has given me some insight into
how these things work.
I did not realize what kind of people were behind this
organization, but I suppose I was too busy looking at the facts
(sorry "trees") wasn't I?
previous thread. It ended up out here instead. Oh well.