News   Oct 04, 2024
 2K     0 
News   Oct 04, 2024
 1.4K     0 
News   Oct 04, 2024
 3.9K     4 

A blog virus... oh no!!

I would be even more satisified when it'd say - "Go find him yourself, what do you think I am?"

AoD
 
Fine. Maybe it's obvious that I'm a conmputer scientist because I get excited about a new paradigm in OS design. It's kinda sad really, I was about as thrilled afterwards as I am after watching a really thought-provoking movie.
 
Minix sounds like the GNU Hurd project. Hurd was going to be "the" free open source OS that Linux ended up becoming. Hurd is still under development and slowly making progress. It is too bad there aren't more people working on it.
 
Minix is reasonably far along. It's a fully functional (if basic) OS. Of course there is practically no software compiled for it, but if you have source, it's not a problem. It's currently being developed mainly by university researchers in the Netherlands, but he mentioned he was trying to get funding for a small development team. They're currently porting x windows, so that'll be nice when it's done.

It's impressively stable. He demonstrated this by killing a driver process (say network) ten times a second with only a minor drop in performance in file transfer. For most OS's, killing a driver is more or less game over...
 
That's outstanding. Though, of course, it'd be preferable that you never have to kill a process :)

I've heard of minix before but the way you describe it doesn't ring any bells. Perhaps there's another *nix out there named minix.
 
Well, he did that to simulate a drastic flaw in a driver that caused it be highly unstable.

Some benefits he mentioned in terms of developing drivers as user programs is that they only get the permissions that are appropriate for them. Only the disk driver should have access to the disk, etc. whereas drivers that live in the kernel can get their messy hands on everything (I've had video drivers that mistakenly output sound, as well). It would also help with driver development, because if the driver dies, the OS remains stable (rather than needing to reload the OS). There are only three or so absolutely essential processes, and he emphasized that they must be kept small to reduce the likelihood/sheer number of bugs.

I found it really fascinating. IMO, if MS knows what's good for it, they should be prototyping their 5 or 10 year out OS on this basis. Windows has just been ballooning in size, such that it isn't really sustainable.
 
Well, he did that to simulate a drastic flaw in a driver that caused it be highly unstable.
Microkernels are great in embedded applications for this exact reason. It makes the difference between a minor annoyance and a product being returned or recalled.

This seems to be going out of style though in favour of higher level languages and writing the code as userland applications instead of drivers.

I think Apple is the only major vendor that (almost) uses a microkernel for standard computers.
 
Has any vendor been making OSs with small kernels for a 'standard' computer? Windows has had large kernels since the dawn of time, and linux kernels have been steadily increasing in size (and can no way be called small, besides). I don't know much about Mac OS, but I can't imagine it having a terribly small kernel.

Minix makes use of implementing drivers as user processes.
 
The Linux kernel can be made very small when the hardware it will run on is known... but it isn't a microkernel. Mac OS X's kernel XNU is a merger of BSD (monolithic) and Mach (microkernel) so the result is not a microkernel although being newer and since it hasn't run on many hardware configurations it is probably much smaller than the Windows or Linux kernel. Hurd is the OS I am most interested in seeing finished.
 

Back
Top