I for one would prefer if people would instead refer to it as 'the thing that shall not be named' as the title makes no assumptions, does not encroach on any several thousand year old technologies (I think that before the Romans put glass in them, windows were more of a hole in the wall than the true window experience they are now) and is dark and gloomy enough to reflect all those works forever lost by those forgetting to save every 10 minutes.
As to the virii, I wish Linux was as secure as all that, but as others have pointed out there are a fair number of exploitable suid-root programs with the average large distribution.
I had a read about HURD's security system a few months back and it looks a lot more promising than the traditional UNIX model (something about starting with no permisions and working your way up, rather than starting with all permisions and dropping them for your typical root service). It should be interesting to see if the new ideas work out in the long run, or whether the 30 year old security model will once again show that it got that old for a reason.
Slashdot Mirror
Indeed, it hardly operates at all.
I for one would prefer if people would instead refer to it as 'the thing that shall not be named' as the title makes no assumptions, does not encroach on any several thousand year old technologies (I think that before the Romans put glass in them, windows were more of a hole in the wall than the true window experience they are now) and is dark and gloomy enough to reflect all those works forever lost by those forgetting to save every 10 minutes.
As to the virii, I wish Linux was as secure as all that, but as others have pointed out there are a fair number of exploitable suid-root programs with the average large distribution.
I had a read about HURD's security system a few months back and it looks a lot more promising than the traditional UNIX model (something about starting with no permisions and working your way up, rather than starting with all permisions and dropping them for your typical root service). It should be interesting to see if the new ideas work out in the long run, or whether the 30 year old security model will once again show that it got that old for a reason.