Hacker News new | past | comments | ask | show | jobs | submit login

That can't be true. If NT-based versions of Windows implemented a system call mechanism that protected the kernel from users, XP wouldn't have been ridden with viruses, and there would have been no purpose in giving Vista and 7 the access control mechanism to warn users of potentially harmful system calls. By the way, Cheatah just refers to the original Mac OS X. Your phrasing "stopped shipping OSes with unprotected kernels ... [starting with] Cheetah" makes it sound like Mac OS X initially didn't have this protection, which is not the case.

First, Cheetah wasn't the first Mac OS X. There was Mac OS X Server 1.0 in 1999 (see: Wikipedia). Cheetah was the first desktop-oriented version of Mac OS X.

Second, I didn't imply that prior versions Mac OS X didn't have kernel protection, I implied that prior versions of Mac OS didn't have kernel protection. This is indisputably true (see: Mac OS 9). Personally, I find Windows / Mac OS parallel surprisingly close here: Windows ME is to Windows XP as Mac OS 9 is to Mac OS X Cheetah.

Third, UAC (User Account Control), the access control introduced with Windows Vista, is almost entirely unrelated to kernel protection (except that UAC would probably be pointless without it). The problem UAC tries to solve is "users running as an administrator too often", not "the kernel isn't protected from user programs". In other words, it is Windows' answer to sudo, not a fundamental change to the Windows kernel.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact