Indeed. In 1980s I managed development of Unix systems where we ported and optimized it for new hardware we had built. During that time, the first set of reports of viruses came for the OS. The level of breaches were mind boggling:
1. The main email background process had a debugging interface. If you logged in into it and typed "wizard" it would give you a root (super user) login and you could do anything to the system! It was put in by the key developer of the OS!
2. You could type a special command at the text editor (vi) and get root privileges.
3. There was a command called "finger" that would let you update your own personal information (office location, etc.). This was added to the end of the line in the password file for your user name. Well, all you had to do was type a newline and that way, create a new root user with a password of your choice on the next line!
4. Someone wrote a program which just passed on random arguments to the OS system calls (the main way you tasked the operating system to do anything). Well, there was little syntax checking, causing the OS to crash left and right as soon as you ran the program. I remember we fixed something like 150 different bugs because of this.
I could go on but you get the picture. #3 and #4 are some of the core techniques utilized today to breach modern operating systems.
Umm, none of the above (including the Sendmail Wizard hack which is decades old!) are viruses. Bugs and security holes, yes. But not viruses.
