Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

drspringfield

macrumors newbie
Dec 11, 2009
18
0
This only affects the Java plug-in, right? That being blocked I can deal with. If the entire JDK/JRE is blocked, that is more problematic.

Correct, and only in Safari.

----------

Is there any way to know exactly WHEN Apple makes these "background updates"? Like... does it happen any time I connect to the App Store under the (Checking for) Updates tab? I'm not as paranoid about this, but I am curious to know when files are modified on my Mac.

One per day. You may turn it off in the Security prefpane.
 

Azathoth

macrumors 6502a
Sep 16, 2009
659
0
Thanks for the fast action, Apple. Although it shows the tradeoff we've had to accept, that keeping up with the latest version can produce situations like this, with a discovered vulnerability for which there is no patch yet. Ironically, when Apple was a version behind, bleeding edge security issues would have been addressed by the time we Mac users got a Java release from Apple.

When Apple was managing Java on OS X, Apple did a piss-poor job and was weeks, months behind on security:
here is one example, though I think there were others in 2011:

https://krebsonsecurity.com/2012/06/apple-oracle-ship-java-security-updates/

"Oracle is the official producer of Java, but Apple maintains its own version, and it has consistently lagged months behind Oracle in fixing security bugs. This failure on Apple’s part finally caught up with Mac OS X users earlier this year and turned into a major embarrassment for Apple, when the Flashback malware infected more than 650,000 Mac systems using a vulnerability that Oracle (but not Apple) had patched roughly two months earlier."

The current blocking (seems to only work in Safari, not FF, but ok) is probably good enough for most users.

My bank(s) and the Germany IRS (needed every month to file my taxes as a freelancers), both require Java and on the Mac this has usually sucked badly, tending to run much better (i.e. just work) on my Windows VM.
 

jon08

macrumors 68000
Nov 14, 2008
1,885
104
I'm a bit confused.. Should I also disable the "Java Applet Plug-In" in my Firefox then?
 

iWe

macrumors regular
Jul 18, 2012
152
0
Apple has already addressed it, as long as you are connected to the internet.
Actually, no. Researching this issue myself, I found these instructions for determining when my Plugin Black List was last updated:

http://osxdaily.com/2011/06/02/check-mac-malware-definition-list-update/

Following these instructions, I came up with 12 Dec 2012. Following the instructions to force updating, it now results in 10 Jan 2013. I presume I will need to repeat this method of all of my Macs, since very clearly automatic is not the answer, at least not for everybody. Also left out generally from this discussion is that the automatic security is not present in Pre-Lion systems if all of the security updates have not been installed.
Thanks!
fing32.gif
 

sseaton1971

macrumors 6502
Feb 9, 2012
431
11
Java is the worst thing ever. Always buggy and slow. Oracle doesn't give a damn about Macs.

This Java vulnerability is not limited to just OS X, correct? I don't know much about the Java platform, but I am guessing many of the "buggy and slow" experiences you have with Java could be due to crappy developers rather than just the platform itself.
 

sseaton1971

macrumors 6502
Feb 9, 2012
431
11
On my mac, this change is in XProtect.meta.plist, not XProtect.plist.

You are correct.

For a work around, you can write a logout hook to disable the code that restricts the Java plugin, but you should be careful for the obvious security reasons.

Use plistbuddy or some other command in your script to remove the "com.oracle.java.JavaAppletPlugin" key. That key is in the following file:

/System/Library/CoreServices/CoreTypes.bundle/Contents/Resources/XProtect.meta.plist

You will need some sort of recurring script since XProtect periodically overwrites the contents of this file. If a user logs in and the Java plugin does not work (because of an XProtect update), the user can just logout and back in to invoke the script that removes the key. You could also have users run a command manually in Terminal, etc., when there is an Xprotect update and the Java plugin becomes disabled.

----------

How do I re-enable the Java plugin? I don't mind having a warning, and I'll disable it again after I've done what I'm trying to do, but I can't find how to work around this block. Any ideas?

See the post above:

https://forums.macrumors.com/posts/16650023/

If you need help with this, I can send you a Terminal script to enable the plugin and another to reenable the block.

----------

...which would help how? The vulnerability affects versions 4-7.

Why be so snarky? That bit of information was included in an update to the article. Perhaps the person asking about rolling back to version 6 posted comments prior to the update. Jeesh! :p
 

dokujaryu

macrumors 6502
May 3, 2011
359
12
Irvine, California
I've been a J2EE engineer for about 5 years and I was a C/C++ backend / PHP frontend engineer for about 5 before that. But none of that is relevant since this is about the Java browser plugin, which I do not endorse. I also don't endorse Silverlight or Active X or many of the technologies that seek to deliver application features via. a webpage by client program execution. I consider it to be too dangerous and I will always prefer the download and installation path. With all the misinformation in the responses to this article, I'm actually surprised no one has started blaming Oracle JavaScript.

On the debate->

Java is a fine language for both enduser applications and backend server web applications. The problem isn't really the language or the architecture, it's the programmers. Java is like the (airquotes) "new" Visual Basic. It's very easy to access and it breaks down a lot of barriers of entry for new programmers. Don't know what endian your processor is? Who cares! Don't understand memory management? No problem! Need a library? Throw it in a folder YOLO!!! These are just basic examples, but what it boils down to is people who have no business programming will succeed in writing somewhat functional software with it.

In my opinion, the argument that Java is slow is fairly dead in common usages of the language. People complain about things like performance when they've never really profiled well written code. In my tests, Java algorithms perform close to the same speed as C given a sufficiently long running process. In most cases the JVM can optimize many complex operations after it has been running for a few iterations (code warmup). In addition, few people take advantage of the performance features of Java such as NIO file access.

In addition, using Java in an enterprise context generally gives you something more important than performance: velocity. Today processors are sufficiently fast, memory is sufficiently cheap, and clustered blades scale sufficiently to allow companies (in general) to buy their way out of the performance argument for a lot less than hiring more engineers to optimize code. And besides, why would you spend money on engineers to optimize features that are already written when there's a huge backlog of features that haven't been written yet? So much of software isn't about actual software engineering, and only a subset of of the software engineering concerns revolve around performance.

Granted, there is still no cure for stupid. Bad programming can ruin any language. The reason these criticisms about Java persist is that even poorly engineered code will probably still run. C/C++ would have a lot more compilation issues, stack overflows, segfaults, and other inescapable "crash" problems. Java's strict OO, exception system, and garbage collection allows bad engineers to ignore flaws more easily.
 
Last edited:

JosephAW

macrumors 603
May 14, 2012
5,958
7,913
Assuming the this affects PPC Macs as well.

I'm assuming this vulnerability affects Power Macs running OS X as well. Although the Java exploit back in April of last year that tried to hack the PPC Mac using a binary it downloaded failed because it was an i386 only code and not universal binary resulting in a non-execution of the binary.
According to my Java enabled Safari 5.0 running OS X 10.5.8 it has "Java 1.5.0_30 from Apple" installed by default and is easily disabled by the preference panel in Safari.
I tried to enable Java in TenFourFox Firefox 17 config console but wasn't even able to do that, so I left all the extensions and plugins disabled by default. I'm sure Camino and Opera as easy to switch off as well.
The real question remains if someone can exploit the Java that is physically installed in the operating system by running a java app through the command line vulnerability when files download and try to open automatically.
Eventually Java may need to be physically removed from the hard disk for a complete solution.
 

LOrion

macrumors newbie
Jan 12, 2013
1
0
California
com.oracle.java.JavaAppletPlugin = Browser plug-in.

Apple has not blocked Java 7 on OS X.

Please correct the headline ASAP before this thread becomes a major flamewar.

VERY confusing... What was original headline, what is 'current' one.

----------

Both show that you block JAVA by using Safari Preferences Security
and disable the JAVA box. Not the other 3 RIGHT?

Please explain these things in NON TECHY Talk.

Mine was enabled and I have seen no warnings...what is all that warning mumbo jumbo about.

What do we do next and please specifiy for Mountain Lion and Pre ML OSX.
 

Mike1984

macrumors member
Oct 21, 2010
39
15
Java is the worst thing ever. Always buggy and slow. Oracle doesn't give a damn about Macs.

you're full of ****.

----------

I've been a J2EE engineer for about 5 years and I was a C/C++ backend / PHP frontend engineer for about 5 before that. But none of that is relevant since this is about the Java browser plugin, which I do not endorse. I also don't endorse Silverlight or Active X or many of the technologies that seek to deliver application features via. a webpage by client program execution. I consider it to be too dangerous and I will always prefer the download and installation path. With all the misinformation in the responses to this article, I'm actually surprised no one has started blaming Oracle JavaScript.

On the debate->

Java is a fine language for both enduser applications and backend server web applications. The problem isn't really the language or the architecture, it's the programmers. Java is like the (airquotes) "new" Visual Basic. It's very easy to access and it breaks down a lot of barriers of entry for new programmers. Don't know what endian your processor is? Who cares! Don't understand memory management? No problem! Need a library? Throw it in a folder YOLO!!! These are just basic examples, but what it boils down to is people who have no business programming will succeed in writing somewhat functional software with it.

In my opinion, the argument that Java is slow is fairly dead in common usages of the language. People complain about things like performance when they've never really profiled well written code. In my tests, Java algorithms perform close to the same speed as C given a sufficiently long running process. In most cases the JVM can optimize many complex operations after it has been running for a few iterations (code warmup). In addition, few people take advantage of the performance features of Java such as NIO file access.

In addition, using Java in an enterprise context generally gives you something more important than performance: velocity. Today processors are sufficiently fast, memory is sufficiently cheap, and clustered blades scale sufficiently to allow companies (in general) to buy their way out of the performance argument for a lot less than hiring more engineers to optimize code. And besides, why would you spend money on engineers to optimize features that are already written when there's a huge backlog of features that haven't been written yet? So much of software isn't about actual software engineering, and only a subset of of the software engineering concerns revolve around performance.

Granted, there is still no cure for stupid. Bad programming can ruin any language. The reason these criticisms about Java persist is that even poorly engineered code will probably still run. C/C++ would have a lot more compilation issues, stack overflows, segfaults, and other inescapable "crash" problems. Java's strict OO, exception system, and garbage collection allows bad engineers to ignore flaws more easily.

Let's not get hysterical.
Javascript itself, that Apple depends on for a host of application coding, is vulnerable to many security attacks.

http://www.slideshare.net/orysegal/clientside-javascript-vulnerabilities
 

Mike1984

macrumors member
Oct 21, 2010
39
15
bad java. baaaad java

If you want REAL SECURITY, you DISABLE all Client Side code, including JavaSCRIPT.
As Cross Site Scripting is the Worst Security Vulnerability out there.

Doing that, however, loses all the "cool" features.

The most secure sites run just JavaEE or Windows ASPX, with No client side libraries.
Nothing.
 

Huntn

macrumors Core
May 5, 2008
23,483
26,600
The Misty Mountains
If you want REAL SECURITY, you DISABLE all Client Side code, including JavaSCRIPT.
As Cross Site Scripting is the Worst Security Vulnerability out there.

Doing that, however, loses all the "cool" features.

The most secure sites run just JavaEE or Windows ASPX, with No client side libraries.
Nothing.

What kind of cool features do you lose? I've been using NoScript forever. Can you control java with that?
 

yg17

macrumors Pentium
Aug 1, 2004
15,027
3,002
St. Louis, MO
I've been a J2EE engineer for about 5 years and I was a C/C++ backend / PHP frontend engineer for about 5 before that. But none of that is relevant since this is about the Java browser plugin, which I do not endorse. I also don't endorse Silverlight or Active X or many of the technologies that seek to deliver application features via. a webpage by client program execution. I consider it to be too dangerous and I will always prefer the download and installation path. With all the misinformation in the responses to this article, I'm actually surprised no one has started blaming Oracle JavaScript.

On the debate->

Java is a fine language for both enduser applications and backend server web applications. The problem isn't really the language or the architecture, it's the programmers. Java is like the (airquotes) "new" Visual Basic. It's very easy to access and it breaks down a lot of barriers of entry for new programmers. Don't know what endian your processor is? Who cares! Don't understand memory management? No problem! Need a library? Throw it in a folder YOLO!!! These are just basic examples, but what it boils down to is people who have no business programming will succeed in writing somewhat functional software with it.

In my opinion, the argument that Java is slow is fairly dead in common usages of the language. People complain about things like performance when they've never really profiled well written code. In my tests, Java algorithms perform close to the same speed as C given a sufficiently long running process. In most cases the JVM can optimize many complex operations after it has been running for a few iterations (code warmup). In addition, few people take advantage of the performance features of Java such as NIO file access.

In addition, using Java in an enterprise context generally gives you something more important than performance: velocity. Today processors are sufficiently fast, memory is sufficiently cheap, and clustered blades scale sufficiently to allow companies (in general) to buy their way out of the performance argument for a lot less than hiring more engineers to optimize code. And besides, why would you spend money on engineers to optimize features that are already written when there's a huge backlog of features that haven't been written yet? So much of software isn't about actual software engineering, and only a subset of of the software engineering concerns revolve around performance.

Granted, there is still no cure for stupid. Bad programming can ruin any language. The reason these criticisms about Java persist is that even poorly engineered code will probably still run. C/C++ would have a lot more compilation issues, stack overflows, segfaults, and other inescapable "crash" problems. Java's strict OO, exception system, and garbage collection allows bad engineers to ignore flaws more easily.

Thank you. I'm a Java/J2EE developer too, and have had experience with other languages (as well as C and PHP, which you mentioned). It can be clunky for desktop applications, certainly not going to deny that, and poor programming only makes the problem worse. But for web, it's an excellent platform and I wouldn't want to code with anything else right now.
 

munkery

macrumors 68020
Dec 18, 2006
2,217
1
When Apple was managing Java on OS X, Apple did a piss-poor job and was weeks, months behind on security:
here is one example, though I think there were others in 2011:

https://krebsonsecurity.com/2012/06/apple-oracle-ship-java-security-updates/

"Oracle is the official producer of Java, but Apple maintains its own version, and it has consistently lagged months behind Oracle in fixing security bugs. This failure on Apple’s part finally caught up with Mac OS X users earlier this year and turned into a major embarrassment for Apple, when the Flashback malware infected more than 650,000 Mac systems using a vulnerability that Oracle (but not Apple) had patched roughly two months earlier."

The current blocking (seems to only work in Safari, not FF, but ok) is probably good enough for most users.

My bank(s) and the Germany IRS (needed every month to file my taxes as a freelancers), both require Java and on the Mac this has usually sucked badly, tending to run much better (i.e. just work) on my Windows VM.

Prior to Flashback, malware leveraging Java in OS X wasn't exploiting Java without user interaction but required users to accept running an unsigned or self-signed Java applet.

In relation to Java exploiting malware without user interaction, Java in OS X was never targeted until Java on other platforms became heavily targeted as well. This is despite the Apple released Java lagging behind the mainstream Java.

The only protection in Java is the Java sandbox, which doesn't use the sandbox of the host OS, given the purpose of Java is to execute code; the host OS memory protections don't apply in relation to Java. Java exploiting malware that doesn't require user interaction targets bypassing the Java sandbox to install payloads.

I think that this type of Java malware has become an issue more recently because malware developers have become more proficient at bypassing the Java sandbox. This type of Java exploiting malware wasn't an issue during the time period where Java for OS X was supplied by Apple at least until Flashback.

The final payload of Flashback was only installed on 10000 of the infected targets despite the number of users infected with the initial payload.

Flashback was an ad-click hijacking malware where the attackers were actually serving the ads instead of altering legitimate ads on websites. So the attackers had bandwidth considerations to deal with which limited the amount of targets that could be served ads until more profit was earned. The threat wasn't in the wild long enough for the attackers to profit enough to inject the final payload and serve ads to everyone in the botnet.

Also being an ad-click hijacking malware, Flashback didn't cause any financial loss to individual users but only to organizations that serve ads.

But, Apple has removed the vector used to allow ad-click hijacking in OS X without password authentication so malware targeting this functionality is no longer able to install silently in OS X.

This threat, like all OS X malware, was overblown by those with a vested interest, such as anti-virus developers, and the media.

But this recent Javapocalypse shows that Java functionality within browsers represents a security issue to the point that users should avoid requiring Java functionality within the browser.

Switch to banks with web apps that don't require Java and lobby your government to no longer use Java for web services as well.
 

SteinMaster

macrumors 6502
Feb 28, 2009
260
0
USA
Is there anything I need to do to address this issue? I went into Safari>Preferences>Security and disabled (unchecked) Java. Do I also need to disable JavaScript? When I disable JavaScript, my web pages do not display properly. I know nothing about Java. I also searched the forum and didn't find recommendations.
 

Weaselboy

Moderator
Staff member
Jan 23, 2005
34,136
15,597
California
Is there anything I need to do to address this issue? I went into Safari>Preferences>Security and disabled (unchecked) Java. Do I also need to disable JavaScript? When I disable JavaScript, my web pages do not display properly. I know nothing about Java. I also searched the forum and didn't find recommendations.

All you need to do is uncheck Java. Leave the Javascript on as it is not related to this threat.
 

PJMAN2952

macrumors regular
May 22, 2011
133
0
So what should I do? I downgraded to Java 7 so it would work with Google Chrome since it's not a 64 bit app. Am I safe? Should I update to the newest version?
 

munkery

macrumors 68020
Dec 18, 2006
2,217
1
So what should I do? I downgraded to Java 7 so it would work with Google Chrome since it's not a 64 bit app. Am I safe? Should I update to the newest version?

The public releases of Java versions 4 to 7 are not safe.

Apparently, the developer release of JDK 7u12 is safe.
 

jaster2

macrumors member
Jun 21, 2010
99
111
Java only blocked in Safari?

I ran into this problem yesterday before hearing about the Java vulnerability news. I was trying to run something within our institution which requires java and it kept giving me an error. I launched Firefox and it worked fine. I just tested this on the java.com site and confirmed java is disabled in Safari but running in Firefox on the same machine. That's much different than Apple completely "disabling" java.
 

munkery

macrumors 68020
Dec 18, 2006
2,217
1
I ran into this problem yesterday before hearing about the Java vulnerability news. I was trying to run something within our institution which requires java and it kept giving me an error. I launched Firefox and it worked fine. I just tested this on the java.com site and confirmed java is disabled in Safari but running in Firefox on the same machine. That's much different than Apple completely "disabling" java.

Apple is just disabling Java in Safari until a fix is provided from Oracle. Apple is not disabling Java in general.

This has already been stated in this thread.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.