Moshe does more than a adequate job of explaining his stance on this issue. Between pointinging out the costs of making the kernel fully preemptible, citing his experiences with using it on personal machines (good) and servers (not so good), then noting the preemtible kernel breaks Mosix and LIDS I think he's got a right to his opinion. It's based upon at least as much fact as stating everyone loves the preemptible kernel.
His analysis is way wide of the mark. His main argument is that kernel preemption causes more context switches, and that's pure BS. Yes, there may be a few more quantum-expiry switches but these are infrequent enough as to be very difficult to notice. He misses that fact that for some loads, a preemption makes tasks run faster because a user task can continue sooner, following a completed disk read. He also misses the true cost of the preemption patch in that spinlocks are slightly more expensive.
The benefits of preemption in terms of reduction in latency on the other hand are large and measurable. For desktop use it's no contest, and I see non-preemptible kernel mainly being an option for certain types of servers carrying a kind of load that exposes the slight extra cost of the spinlocks, and a percent or two extra throughput somehow matters.
would it be all that crazy to adopt the VM system from the 'establishment' (bsd)? frequently the linux codebase DOES borrow from bsd. why is the VM system all that different?
BSD doesn't support >4 Gig memory on IA32, for one thing. Yes, Linux's VM could be more similar to BSD's and it would probably be better that way, but there is no way it can be identical.
Thus, the relative cost of a page fault has increased by a factor of 200. This makes VM far less attractive today than it used to be. It's not getting any better, either.
You completely failed to think about disk caching and memory-mapped files, your view of VM is way too limited, and that's how you argued yourself to an incorrect conclusion.
Finally, I went to www.ea.com a while ago. As usual, I tried with both Mozilla and Konqueror. Again, no good. They blocked them out, and suggested "upgrading" to IE.
They're being assholes. I set Opera to identify as IE and it handles the site just fine. I didn't bother to try with Mozilla, because I'd have to edit the config by hand. In a way, I'm glad to see that Mozilla makes it somewhat of a pain in the butt to spoof the identify, I want ea to know they're driving away traffic and why.
I seriously doubt Mozilla will have any problem with the site.
The answer is obvious. Microsoft are great at marketing. This was free publicity. Tons and tons and tons of free press....
Yeah right, nice theory, except that the press is all negative and probably dwells more on Microsoft's delay in removing the block than anything else. Plus makes Microsoft look weak because of having to give in to the demands of a bunch of angry geeks.
For the life of me, I can see no sense in Microsoft's strategy here, only that they seriously underestimated the strength of the backlash and the vulnerability of their position.
It makes even less sense to me that they delayed removing the block, which really heated things up. And finally, when they did get around to it, the damage was already done as far as effect on the anti-trust case goes.
They literally take the packets, serialize them, put in an RS-232 emulation protocol (control pins & stuff), stick Hayes AT modem commands atop that, and run packets atop THAT!
Thankyou very much. That's all I need to know about Bluetooth.
I read the paper, and it seems to basically say "it's pretty hard to get your dependencies right with recursive makes. If you don't get your dependencies right, then bad things happen. Therefore, recursive make is bad." It certainly is possible, if you're willing to put in any sort of effort, to get correct dependencies with recursive makes. I'm not going to comment on which method is better or takes less work, but the paper misrepresents just how bad recursive makes are.
I'm wondering, do kernel developers use tools like vmware/plex86 to debug their running kernels ? It seems like we've come a long way since debugging with strategically placed printfs
Vmware or plex86 could possibly be of some use, except that they're no good for debugging device drivers since the real devices are hidden behind a virtualization layer. For non-device driver work User Mode Linux is a more lightweight solution, and tracks the latest development kernels much more closely. An increasing number of the core developers are using User Mode Linux regularly.
For heavy debugging work on live kernels, kgdb is the perferred solution, with a serial cable link to a test machine. It takes a little more work to set this up and you need two machines. Kdb is a simpler debugger that can be patched into the kernel, useful for tracking down elusive kernel problems. It's included by default in SGI's XFS patch and pre-patched kernels.
There are some great tools available including LTT, the Linux Trace Toolkit and various lock-monitoring patches. Unfortunately, most driver development is still being done by the printk/reboot method. If this is your preferred method, make sure you install a journalling filesystem unless you like spending most of your time watching fsck work.
According to the update at CNet, Microsoft specifically stated that they are watching for Opera versions because they aren't fully compliant with XHTML 1.0, and want the user to use only compliant XHTML browsers to view MSN. (Microsoft's claims)
Microsoft are spreading FUD. Both Opera and Mozilla do a far better job at rendering XHTML 1.0 correctly. IE has several problems with XHTML.
They claim that browsers should be standards compliant to access MSN.com, when the page itself doesn't even validate as valid XHTML. Anyone can verify this by running it through W3C's HTML Validator:
http://validator.w3.org/
In other words, Microsoft are spreading misinformation, and yet again ignoring existing open standards.
This is worrying, but hardly surprising, considering Microsoft's track record.
This is stooping to an all-time low - if that is even possible. I would go so far as to call Microsoft "XHTML standard compliance" claims blatant lies, and an insult to people's intelligence. They appear to think that anything they say goes.
No, Microsoft has stooped lower. When the Microsoft was a member of the W3C working group drafting the CSS standard, they took the working group's ideas and patented them.
If only one had the money to sue Microsoft. Sadly, they have the money, so suing them won't help. They will walk all over you.
If you think what Microsoft did is wrong, you can say why, here.
They're doing it for the same reason that dictionaries and thesauri targeted at school children exclude the very same words. You want an unabridged reference? Fine. Buy one.
The flip side of that is, you want a censored reference, fine, buy from Microsoft.
Personally, I don't give a fsck whether ATI optimizes for Quake or not, what I care about is who is making their specs open. It's ATI, not NVidia, so all I can say is: go ATI. Sell lots of cards.
That might help with the spool file, but it still doesn't solve the problem of being unable to deliver into individual mailboxes in a user's home dir. (For example, to sort mailing lists.)
I didn't realize that was also a problem. Presumably this should be done by the user's mail client.
Erm... adduser needs to know about the mta so it can execute a procedure supplied by the mta at the time a new user is added. Obvious, no? Non-trivial to implemement: yes. Because we have no defined way for high level facilities to cooperate in such a way. Put on thinking cap.
Keep these SQL apps behind the firewall.. turn off all IIS features on the sql boxes.. and at least Nimda should not be able to get at it.
Your attacker could still use
some other exploit that doesn't rely on IIS. I hope you don't think we've seen the last of these.
Note that an exploit like the above wouldn't turn into a Ro0t on a Linux/Unix box because the database server typically doesn't run with system privilege.
If you aren't allowed to patch your server, then you should isolate it behind a firewall of some sort, so that the chances of infection are minimized. This may not work well for IIS (beyond simply not running it), but it will serve you well in the general case.
So, you're suggesting security by obscurity? Hmm, best of luck to you.
Some exploits work just fine through the firewall, so then you've got a compromised server insider your firewall and a false sense of security. There's no substitute for being secure in the first place. If it's not secure, don't connect it to your network.
Users do want simplicity and ease of use. And it is also true that Linux can't give them this right now.
No, that's not true. Have you tried a recent Mandrake for example? Dead simple, right from the install through to the apps. I'd say Linux has the edge on install, because much of the hardware in the known universe is supported by default, no screwing around with driver disks.
It used to be kind of esoteric, and the reward was, you'd get something stable. Now it's not esoteric, and, hmm, it's free, updates are free, support is free, so it looks like Windows is going to have a tough time competing, huh?
StarOffice is really no different than Microsoft Office from the user's point of view, most really don't know or care what the differences are. These days you really have to stretch into nooks and crannies to find places where Linux's ease of use is any different from Windows. And Linux is still moving. Windows is kind of moving too, but not very fast, you need time-lapse photography to detect the motion. Now, most Windows development is aimed at finding new ways to restrict users, not ways to make their computers more more useful.
While not the perfect midi sequencer, jazz++ is robust and GPL'd. Since jazz++ only appeared twice in the postings (each moderated at +1...) related to the earlier story, it would seem this fine product has low visibility among the/. crowd. The only viable GNU/Linux midi solution died the same week./ had a call for audio solutions on Linux. Gotta love irony..."
So long as there's one copy of the source in the wild, GPL software never dies. I downloaded the CVS for safekeeping, and when I've got time I'll work on it.
I didn't see any mention of internationalization (I18N) or localization (L10N) in any part of this list.
I didn't have to tell Mozilla anything or do anything special, and I can read sites in Chinese, Japanese and Korean. Check it out, here's a four-language site. I don't know how I'd go about typing CJK text though.
Gov't jobs don't really add anything to the economy, though. Gov't money comes from taxpayers, so more gov't jobs means more drag on the private sector that is actually producing something.
So you think it's better to just ship the money out of the country? Right.
Gov't is a necessary evil, and should be no larger than necessary. The private sector can usually do the same job better, cheaper, and faster.
Most probably, the private sector *is* going to do this Linux support. So what's your point?
Slashdot Mirror
So what happens? The kernel just paints itself into a corner until the machine freezes. Only way to recover is to power cycle.
No, there's a task killing option available under the SysRq key now. And you're wrong, the oom_kill is still in there, it hasn't been removed.
Moshe does more than a adequate job of explaining his stance on this issue. Between pointinging out the costs of making the kernel fully preemptible, citing his experiences with using it on personal machines (good) and servers (not so good), then noting the preemtible kernel breaks Mosix and LIDS I think he's got a right to his opinion. It's based upon at least as much fact as stating everyone loves the preemptible kernel.
His analysis is way wide of the mark. His main argument is that kernel preemption causes more context switches, and that's pure BS. Yes, there may be a few more quantum-expiry switches but these are infrequent enough as to be very difficult to notice. He misses that fact that for some loads, a preemption makes tasks run faster because a user task can continue sooner, following a completed disk read. He also misses the true cost of the preemption patch in that spinlocks are slightly more expensive.
The benefits of preemption in terms of reduction in latency on the other hand are large and measurable. For desktop use it's no contest, and I see non-preemptible kernel mainly being an option for certain types of servers carrying a kind of load that exposes the slight extra cost of the spinlocks, and a percent or two extra throughput somehow matters.
would it be all that crazy to adopt the VM system from the 'establishment' (bsd)? frequently the linux codebase DOES borrow from bsd. why is the VM system all that different?
BSD doesn't support >4 Gig memory on IA32, for one thing. Yes, Linux's VM could be more similar to BSD's and it would probably be better that way, but there is no way it can be identical.
Thus, the relative cost of a page fault has increased by a factor of 200. This makes VM far less attractive today than it used to be. It's not getting any better, either.
You completely failed to think about disk caching and memory-mapped files, your view of VM is way too limited, and that's how you argued yourself to an incorrect conclusion.
Rik van Reil's (-ac series kernels) is somewhat modeled off the FreeBSD VM. How much so I don't know.
Loosely
Finally, I went to www.ea.com a while ago. As usual, I tried with both Mozilla and Konqueror. Again, no good. They blocked them out, and suggested "upgrading" to IE.
They're being assholes. I set Opera to identify as IE and it handles the site just fine. I didn't bother to try with Mozilla, because I'd have to edit the config by hand. In a way, I'm glad to see that Mozilla makes it somewhat of a pain in the butt to spoof the identify, I want ea to know they're driving away traffic and why.
I seriously doubt Mozilla will have any problem with the site.
The answer is obvious. Microsoft are great at marketing. This was free publicity. Tons and tons and tons of free press....
Yeah right, nice theory, except that the press is all negative and probably dwells more on Microsoft's delay in removing the block than anything else. Plus makes Microsoft look weak because of having to give in to the demands of a bunch of angry geeks.
For the life of me, I can see no sense in Microsoft's strategy here, only that they seriously underestimated the strength of the backlash and the vulnerability of their position.
It makes even less sense to me that they delayed removing the block, which really heated things up. And finally, when they did get around to it, the damage was already done as far as effect on the anti-trust case goes.
Somebody in the executive suite is on crack.
They literally take the packets, serialize them, put in an RS-232 emulation protocol (control pins & stuff), stick Hayes AT modem commands atop that, and run packets atop THAT!
Thankyou very much. That's all I need to know about Bluetooth.
I read the paper, and it seems to basically say "it's pretty hard to get your dependencies right with recursive makes. If you don't get your dependencies right, then bad things happen. Therefore, recursive make is bad." It certainly is possible, if you're willing to put in any sort of effort, to get correct dependencies with recursive makes. I'm not going to comment on which method is better or takes less work, but the paper misrepresents just how bad recursive makes are.
It's slow too.
I'm wondering, do kernel developers use tools like vmware/plex86 to debug their running kernels ? It seems like we've come a long way since debugging with strategically placed printfs
Vmware or plex86 could possibly be of some use, except that they're no good for debugging device drivers since the real devices are hidden behind a virtualization layer. For non-device driver work User Mode Linux is a more lightweight solution, and tracks the latest development kernels much more closely. An increasing number of the core developers are using User Mode Linux regularly.
For heavy debugging work on live kernels, kgdb is the perferred solution, with a serial cable link to a test machine. It takes a little more work to set this up and you need two machines. Kdb is a simpler debugger that can be patched into the kernel, useful for tracking down elusive kernel problems. It's included by default in SGI's XFS patch and pre-patched kernels.
There are some great tools available including LTT, the Linux Trace Toolkit and various lock-monitoring patches. Unfortunately, most driver development is still being done by the printk/reboot method. If this is your preferred method, make sure you install a journalling filesystem unless you like spending most of your time watching fsck work.
--
Daniel
Microsoft are spreading FUD. Both Opera and Mozilla do a far better job at rendering XHTML 1.0 correctly. IE has several problems with XHTML.
They claim that browsers should be standards compliant to access MSN.com, when the page itself doesn't even validate as valid XHTML. Anyone can verify this by running it through W3C's HTML Validator:
http://validator.w3.org/
In other words, Microsoft are spreading misinformation, and yet again ignoring existing open standards.
This is worrying, but hardly surprising, considering Microsoft's track record.
This is stooping to an all-time low - if that is even possible. I would go so far as to call Microsoft "XHTML standard compliance" claims blatant lies, and an insult to people's intelligence. They appear to think that anything they say goes.
No, Microsoft has stooped lower. When the Microsoft was a member of the W3C working group drafting the CSS standard, they took the working group's ideas and patented them.
If only one had the money to sue Microsoft. Sadly, they have the money, so suing them won't help. They will walk all over you.
If you think what Microsoft did is wrong, you can say why, here.
They're doing it for the same reason that dictionaries and thesauri targeted at school children exclude the very same words. You want an unabridged reference? Fine. Buy one.
The flip side of that is, you want a censored reference, fine, buy from Microsoft.
I see your point.
Since it's being funded by the government, it should use the cross-platform OpenGL standard, not Direct3D.
Personally, I don't give a fsck whether ATI optimizes for Quake or not, what I care about is who is making their specs open. It's ATI, not NVidia, so all I can say is: go ATI. Sell lots of cards.
That might help with the spool file, but it still doesn't solve the problem of being unable to deliver into individual mailboxes in a user's home dir. (For example, to sort mailing lists.)
I didn't realize that was also a problem. Presumably this should be done by the user's mail client.
You forgot xfce.
The procedure would create the user's mail spool file, using adduser's elevated privilege.
Erm... adduser needs to know about the mta so it can execute a procedure supplied by the mta at the time a new user is added. Obvious, no? Non-trivial to implemement: yes. Because we have no defined way for high level facilities to cooperate in such a way. Put on thinking cap.
"Using extended stored procedures, the attacker could essentially gain complete control over the server itself."
Your attacker could still use some other exploit that doesn't rely on IIS. I hope you don't think we've seen the last of these.
Note that an exploit like the above wouldn't turn into a Ro0t on a Linux/Unix box because the database server typically doesn't run with system privilege.
So, you're suggesting security by obscurity? Hmm, best of luck to you.
Some exploits work just fine through the firewall, so then you've got a compromised server insider your firewall and a false sense of security. There's no substitute for being secure in the first place. If it's not secure, don't connect it to your network.
No, that's not true. Have you tried a recent Mandrake for example? Dead simple, right from the install through to the apps. I'd say Linux has the edge on install, because much of the hardware in the known universe is supported by default, no screwing around with driver disks.
It used to be kind of esoteric, and the reward was, you'd get something stable. Now it's not esoteric, and, hmm, it's free, updates are free, support is free, so it looks like Windows is going to have a tough time competing, huh?
StarOffice is really no different than Microsoft Office from the user's point of view, most really don't know or care what the differences are. These days you really have to stretch into nooks and crannies to find places where Linux's ease of use is any different from Windows. And Linux is still moving. Windows is kind of moving too, but not very fast, you need time-lapse photography to detect the motion. Now, most Windows development is aimed at finding new ways to restrict users, not ways to make their computers more more useful.
So long as there's one copy of the source in the wild, GPL software never dies. I downloaded the CVS for safekeeping, and when I've got time I'll work on it.
I didn't have to tell Mozilla anything or do anything special, and I can read sites in Chinese, Japanese and Korean. Check it out, here's a four-language site. I don't know how I'd go about typing CJK text though.
So you think it's better to just ship the money out of the country? Right.
Gov't is a necessary evil, and should be no larger than necessary. The private sector can usually do the same job better, cheaper, and faster.
Most probably, the private sector *is* going to do this Linux support. So what's your point?