In 2004, OmniNerd examined the state of operating system security in response to studies demonstrating computers were compromised in less than 20 seconds. Three years later, the security scene has changed dramatically. Now, as the world becomes ever more dependent upon IT, the industry is holding manufacturers to a higher standard, demanding increased security from their products. Once again, OmniNerd has put the flagship products from operating system vendors to the test.
greetings,
while not the most popular desktop os, OpenBSD has some usage as a server os.
would it be possible for you to add this system to your selection of tested os'es?
bengt
OpenBSD is missing from this list! Please review OpenBSD also. I think it's a glaring omission given the fact that OpenBSD's primary goal is security.
Great paper though!
Sol10u3 aka Solaris 10 11/06 has serious security enhancements. It is meant to be secure by default out of the box and includes multi-level security like Trusted Solaris has.
If the nessus scans you ran were against an earlier release, it would be helpful to run the same tests against an 11/06 release.
So I was cruising through the comments left on Slashdot after the article hit the front page. It's evident that a lot of the readers there breezed the article, missed the content and then made hollow conclusions about my conclusion. How do I know they didn't RTFA? Google Analytics reports the average time spent on the page was three minutes or less.
For the less astute, ADD or sensationalist IT folk:
- the article was not designed to test security features
- the article was not looking at just what services are on out-of-box
- it was designed to test what binaries shipped "stock"
- it was designed to show the patchwork vendor's did over time on the product
- it was designed to illustrate how Mom & Pop haphazard use of an operating system will expose their computer
- it was designed to link those exposures back to how 2006's malware activity was also driven (not just through e-mail and webhacks)
- the OS's selected were the "most obvious" choices for non-obscure use
- old versions of XP were tested because frankly, Vista didn't exist for consumers in '06 meaning it was still the flagship product
As usual, IBM's iSeries (OS/400) was left out of another discussion of
operating systems. Is it sexy? Not by a long shot - but it will sit and run and run and run almost forever without a hiccup. Security is merely an afterthought - it's built in tightly. And speed of the larger models rivals a mainframe. These machines are for people who need real, reliable work done, not worrying about how the GUI is aesthetically pleasing or not.
I know I'm a little late to this conversation, but someone just linked me to this article yesterday. I noticed that the article cuts off in the middle of assessing SuSE Enterprise 10, and there is no link to continue reading. I'm curious to read the conclusion. Does anyone know where I can find it?



article
by 
Add a Comment (14)
Email This
Message Author
Statistics
RSS


Correspondence with popularity by Brandon :: NR9 :: Show
Forgive my ignorance (if it's showing), but how much of these results can be attributed (if any) simply to the popularity of each OS? In other words, is it possible that Windows and Mac OS are the most vulnerable, respectively, because they are the most popular and thus are targeted more often?