Programming | What programming language should I start with?

Page 2 of 3 [ 35 posts ]  Go to page Previous  1, 2, 3  Next


Was this helpful on you programming language selection?
Poll ended at 12 Jun 2014, 7:40 pm
Yes 33%  33%  [ 1 ]
No 67%  67%  [ 2 ]
Total votes : 3

MolinaMegaTech
Hummingbird
Hummingbird

User avatar

Joined: 12 Mar 2014
Age: 23
Gender: Male
Posts: 18
Location: In your CPU

27 Mar 2014, 3:08 pm

I know right... They keep on going back and forth



sliqua-jcooter
Veteran
Veteran

User avatar

Joined: 25 Jan 2010
Age: 36
Gender: Male
Posts: 1,488
Location: Burke, Virginia, USA

27 Mar 2014, 8:25 pm

Kurgan wrote:
sliqua-jcooter wrote:
Wrong *again*. Not only does every major linux distribution include SElinux, RHEL/CentOS/Fedora turn it on *by default* with a permissive configuration, and locking it down is as simple as installing a different configuration that the distributions also provide.


Fedora, CentOS and RHEL, I have no experience with. Nevertheless, they all have less than 10% of the Linux desktop share combined.


Why are we limiting ourselves to the "desktop" market. From a security perspective, there is 0 difference between a machine that is a desktop and a machine that is a server. They're all boxes on the network running services for users. It doesn't make any difference whether the system has a GUI or not. And, if you're going to limit the discussion to desktops, take back everything you said about Windows Server.

Quote:
It wasn't present in neither Ubuntu or Slackware back when I was using it--and thus, simply knowing the root password gave me full power of a computer. Because of this, making malware was basically a DYI activity.


In other words, you don't know enough about Linux security to comment about it. Got it.

Quote:
Technically, with your root password, I could disable SELinux, but I don't think it's easy to make a virus or a worm that does it for you.
1) No, you can't - unless you have physical access or remote KVM. 2) You could say the exact same thing about UAC. 3) If you have physical access to a system, you have total access. Period.

Quote:
The problem with Linux is what a virus could potentially do; in Windows, it's limited to terminating user-space processes without my consent or encrypting files that have copies on GitHub.


The entirety of the user's computing experience is in the user land - so this is an entirely moot point. The only thing kernel access gets you is the ability to read whatever is in memory from whatever application. Not useful in the vast majority of cases.

Quote:
The link you posted just explained how the hashes worked; it didn't show any C/C++ code on how they were implemented.


Right, because it's not like you can't do a google search for "NTLM hash generator" and see if anyone has been able to write software to generate NTLM hashes. Oh, wait... http://lmgtfy.com/?q=NTLM+hash+generator

Hey, on the first page there's even a project that does *exactly* that. C/C++ code that implements the NTLM hash mechanism: http://www.codeproject.com/Articles/328 ... -Generator


_________________
Nothing posted here should be construed as the opinion or position of my company, or an official position of WrongPlanet in any way, unless specifically mentioned.


Kurgan
Veteran
Veteran

User avatar

Joined: 6 Apr 2012
Age: 35
Gender: Male
Posts: 4,132
Location: Scandinavia

27 Mar 2014, 8:59 pm

sliqua-jcooter wrote:
Why are we limiting ourselves to the "desktop" market. From a security perspective, there is 0 difference between a machine that is a desktop and a machine that is a server. They're all boxes on the network running services for users. It doesn't make any difference whether the system has a GUI or not. And, if you're going to limit the discussion to desktops, take back everything you said about Windows Server.


From a security perspective, banks, institutions, etc. that run Linux, all have proprietary security extensions.

I never said I prefered Windows Server to Linux Server (the latter is more convenient to use than the former); I prefer Windows 7 and 8 over desktop versions of Linux.

Quote:
In other words, you don't know enough about Linux security to comment about it. Got it.


You may not realize it, but your claims have been tested in Virtual Box. I can put malicious drivers in the kernel (with the same privileges as the OS itself, and better privileges than most anti-virus tools), I can make it communicate with user-space files, and I can have it access any file on a common Linux computer because of this.

In Windows, nothing that isn't signed by Microsoft gets in the kernel.

Quote:
1) No, you can't - unless you have physical access or remote KVM.


Or via ssh, if you allow users to execute root commands in it.

Quote:
2) You could say the exact same thing about UAC.


True, and this would leave the backdoor open to attacks. With that being said, malware on Windows won't crash your entire system, and Windows 8.1 comes with a preinstalled cloud where documents are saved by default—which a virus to my knowledge can't reach.
.
Quote:
3) If you have physical access to a system, you have total access. Period.


A virus installed on my computer has physical access. Because of the security meassures in Windows 8.1, it can't do anything beyond hijacking my web-browser unless I give it permissions to do it.

Quote:
The entirety of the user's computing experience is in the user land - so this is an entirely moot point. The only thing kernel access gets you is the ability to read whatever is in memory from whatever application. Not useful in the vast majority of cases.


Kernel access grants you access to the hardware, such as the BIOS. Access to userspace with kernel (unrestricted) privileges, allows you to do absolute anything. Kernel malware in Windows 4.0 proved how dangerous this could be.

Quote:
Right, because it's not like you can't do a google search for "NTLM hash generator" and see if anyone has been able to write software to generate NTLM hashes. Oh, wait... http://lmgtfy.com/?q=NTLM+hash+generator

Hey, on the first page there's even a project that does *exactly* that. C/C++ code that implements the NTLM hash mechanism: http://www.codeproject.com/Articles/328 ... -Generator


I did a Google search on it. Couldn't find a single decoder for NTLMv2.



sliqua-jcooter
Veteran
Veteran

User avatar

Joined: 25 Jan 2010
Age: 36
Gender: Male
Posts: 1,488
Location: Burke, Virginia, USA

28 Mar 2014, 7:58 am

Kurgan wrote:
From a security perspective, banks, institutions, etc. that run Linux, all have proprietary security extensions.


No, they don't. I've worked at banks, and I held a top secret clearance and worked for government contractors at the pentagon - they use AppArmor or SELinux, respectively. SELinux was designed by the NSA to implement a true MAC layer in the linux kernel - it's mandated that all TS government systems implement SELinux.

Quote:
I never said I prefered Windows Server to Linux Server (the latter is more convenient to use than the former); I prefer Windows 7 and 8 over desktop versions of Linux.


This isn't a conversation about preference. This is a conversation about security systems available in the OS.

Quote:
You may not realize it, but your claims have been tested in Virtual Box. I can put malicious drivers in the kernel (with the same privileges as the OS itself, and better privileges than most anti-virus tools), I can make it communicate with user-space files, and I can have it access any file on a common Linux computer because of this.

In Windows, nothing that isn't signed by Microsoft gets in the kernel.


Linux has module signing as well. Difference being, you can choose which organization you want to trust for signatures. There are distributions available that make this required. http://www.linuxjournal.com/article/7130

Quote:
Quote:
1) No, you can't - unless you have physical access or remote KVM.


Or via ssh, if you allow users to execute root commands in it.


No. If you try to modify the selinux configuration on my box, it won't let you - even as root. If you try to use the setenforce command, it won't let you. This is a standard part of hardened SELinux configurations. You have to reboot the system and disable SELinux at the boot loader, either by booting into single user mode, or by passing a kernel parameter that makes SELinux permissive.

Quote:
True, and this would leave the backdoor open to attacks. With that being said, malware on Windows won't crash your entire system, and Windows 8.1 comes with a preinstalled cloud where documents are saved by default—which a virus to my knowledge can't reach.


Why would I want to crash your system? That is of no benefit to the person writing the malware. Either I want to spy on you, grab your files, look for bit coin, etc. We're well past the point where crashing a system is the ultimate goal of anything.

.
Quote:
A virus installed on my computer has physical access. Because of the security meassures in Windows 8.1, it can't do anything beyond hijacking my web-browser unless I give it permissions to do it.


No, it doesn't - someone sitting at a keyboard has physical access. A virus can't boot a livecd and mount the local disk to make changes, it can't boot into safe mode, it can't modify BIOS settings, or mess with the physical configuration of the box.

Quote:
I did a Google search on it. Couldn't find a single decoder for NTLMv2.


Hashes aren't "decoded". They are inherently one-way. Sounds like somebody failed Computer Science 101.


_________________
Nothing posted here should be construed as the opinion or position of my company, or an official position of WrongPlanet in any way, unless specifically mentioned.


Kurgan
Veteran
Veteran

User avatar

Joined: 6 Apr 2012
Age: 35
Gender: Male
Posts: 4,132
Location: Scandinavia

28 Mar 2014, 1:25 pm

sliqua-jcooter wrote:
No, they don't. I've worked at banks, and I held a top secret clearance and worked for government contractors at the pentagon - they use AppArmor or SELinux, respectively. SELinux was designed by the NSA to implement a true MAC layer in the linux kernel - it's mandated that all TS government systems implement SELinux.


Several banks have modified the source code (and made sure that it doesn't fall into the wrong hands). http://www.networkmagazineindia.com/200303/cover5.shtml (IDBI Bank).

Actually, configuring Linux to meet their needs (among others, their security needs) is more expensive than doing the same with Windows Server, according to Member's Equity Bank, but this statement should be taken with a pinch of salt.

Quote:
This isn't a conversation about preference. This is a conversation about security systems available in the OS.


Then Windows 8.1 wins easily if we're talking about pre-installed configurations. If we account for all available extensions, then it's impossible to make a comparison. If you frequent torrent sites and other places where you'd potentially get malware, you're less likely to with Linux, but only because it's a less interesting platform to make viruses for.

I'm not saying that Linux is a bad OS (because unlike the Linux fanboys, I can still appreciate an OS while prefering another OS).

Quote:
Linux has module signing as well. Difference being, you can choose which organization you want to trust for signatures. There are distributions available that make this required. http://www.linuxjournal.com/article/7130


You need to specify this yourself in most distros. There are also distributions that allow you to test the drivers before you implement them, but this is besides the point. With the most popular distros out there, all you need to get an unsigned driver into the kernel is approval from the root user. Moreover, most casual computer users have no idea how to change these settings.

Quote:
No. If you try to modify the selinux configuration on my box, it won't let you - even as root. If you try to use the setenforce command, it won't let you. This is a standard part of hardened SELinux configurations. You have to reboot the system and disable SELinux at the boot loader, either by booting into single user mode, or by passing a kernel parameter that makes SELinux permissive.


This depends on your configuration. I have no experience with Fedora, but at least a few years ago, it was entirely possible with the root password, unless you placed restrictments on the ssh.

SELinux wasn't even included when I installed Mint (and I don't think it would be in Ubuntu or Debian either, given that Mint is an improved version of the aforementioned).

Quote:
Why would I want to crash your system? That is of no benefit to the person writing the malware. Either I want to spy on you, grab your files, look for bit coin, etc. We're well past the point where crashing a system is the ultimate goal of anything.


Crashing a system would be beneficial to gain infamy, and any viruses I've had on my current computer, were programmed for rather childish and petty reasons--and were easy to remove. If you crash the systems at a business, you'll also make sure that they'll lose money (I'll asume that you already know the price of downtime). Few viruses have any actual use these days beyond spreading havock or proving a point--and stuff like CryptoLocker (easy meassures can be taken to make sure you never get any ransomware in the first place), is almost exclusively reserved for Phishing mails. Valuable projects are almost always stored on an SVN or Git; Windows 8.1 also comes with a very tightly integrated cloud where unathorized malware can't just walk in and get what they need (unless they infect the servers). Regardless, making ransomware can be done the exact same way in Linux.

Malware used for spying isn't nearly as popular today as it was back in the heydays of KaZaa, DC++, and similar applications. Stealing bitcoins just by using viruses is difficult, and it's far more convenient to use use scams and pyramid schemes in the same manner you would use to steal real money.

Quote:
No, it doesn't - someone sitting at a keyboard has physical access. A virus can't boot a livecd and mount the local disk to make changes, it can't boot into safe mode, it can't modify BIOS settings, or mess with the physical configuration of the box.


You can make malware that behaves exactly like someone with physical access if you have the root password.

If it has access to the kernel, it can modify BIOS settings. There were a few examples of this before drivers had to be signed, and when user-space applications had direct hardware access. Take CIH, for instance.

Quote:
Hashes aren't "decoded". They are inherently one-way. Sounds like somebody failed Computer Science 101.


Arguing whether they're "decrypted" (something that can be done on all systems by brute-force given enough time) or "decoded" is splitting hairs. I'm among the top students, the only one in class who finished with extra credits (beyond what's required for a degree in computer science) in mathematics in prescribed time, and even got access to off-the-shelf technology from Cisco for my thesis. Come to think of it, the company that I'm doing the assignment for (the largest software developer in the Nordic countries), uses Windows for programming.