THE reports from IDG make it sound as though FileZilla is a security threat [1,2] when it fact it is fakes that are a threat, as Sean pointed out to counter these allegations [3].
A basic tenant of open-source software security has long been the idea that since the code is open, anyone can look inside to see if there is something that shouldn't be there.
The cross-platform HEUR:Backdoor.Java.Agent.a, as reported in a blog post published Tuesday by Kaspersky Lab, takes hold of computers by exploiting CVE-2013-2465, a critical Java vulnerability that Oracle patched in June. The security bug is present on Java 7 u21 and earlier. Once the bot has infected a computer, it copies itself to the autostart directory of its respective platform to ensure it runs whenever the machine is turned on. Compromised computers then report to an Internet relay chat channel that acts as a command and control server.
Users clicking on some ads are redirected to sites armed with code that exploits vulnerabilities in Java and installs a variety of different malware.
The gist of the story is that "123456" is now the most commonly used weak password—surpassing the use of the word "password."
The OpenSSL Project confirmed that weak passwords used on the hosting infrastructure led to the compromise of its website, dispelling concerns...
Code repositories remained untouched in the December 29 hack, and the only outward sign of a breach was a defacement left on the OpenSSL.org home page. The compromise is nonetheless rattling some nerves. In a brief advisory last updated on New Year's Day, officials said "the attack was made via hypervisor through the hosting provider and not via any vulnerability in the OS configuration." The lack of additional details raised the question of whether the same weakness may have been exploited to target other sites that use the same service. After all, saying a compromise was achieved through a hypervisor vulnerability in the Web host of one of the Internet's most important sites isn't necessarily comforting news if the service or hypervisor platform is widely used by others.
Back when I wrote Perfect Passwords, I generated a list of the top 500 worst (aka most common) passwords which seems to have propagated quite a bit across the internet, including being mentioned on Gizomodo, Boing Boing, Symantec, Laughing Squid and many other sites. Since then I have collected a large number of new passwords bringing my current list to about 6,000,000 unique username/password combos, including many of those that have been recently made public*.
All that being said, and given that the Luddite solution of forsaking the Internet may not be terribly practical, this is another reason to encrypt technical data that you are sending by email even if the recipient is a U.S. person firmly planted on U.S. soil. No, the encryption isn’t a defense to the violation, but it is at least a mitigating factor. Remember, as I posted last May, that the U.S. military thinks it can put ITAR-controlled technical data on a Chinese satellite if it’s encrypted; so if you don’t have anything else to say in your defense when an email with export controlled data accidentally wanders through Lithuania, you will at least have that. And maybe one day in the distant future, BIS and DDTC will admit that the Internet exists and that encryption works.
Older versions of Safari for Mac store unencrypted user login credentials in a plain text file, according to security firm Kaspersky (via ZDNet). Safari saves the information in order to restore a previous browsing session, reopening all sites, even those that require authentication using the browser's "Reopen All Windows from Last Session" functionality.