Since evolution has no specific goal, there is really no meaningful definiiton of "devolving." Any change in allele frequency over time is evolution, whether or not you think it is a good change. Also, the part of the population which reproduces most successfully is, by definition, most suited... at least from an evolutionary persepective. There may be some sociological idea of how well suited a group is.. i dunno.
Hmm... now curing cancer is nice and all, but if/when applied to humans, does this mean they can smoke cigarettes w/o ill effect,
Cancer is hardly the only ill effect of smoking cigarettes. It is just hte one that gets the most press because, well, it is cancer. Smoking is one horrible, horrible thing you can do to your body. Even without cancer.
clean up nuclear waste with their bare hands,
Again, cancer isn't the only ill effect here...
r travel in space for extended duration w/o ill effect?
Speaking as someone with poor depth perception, I don't see a real difference between "real" 3D and what games call 3D. And if you really want to get pedantic, you could just as easily say that requiring special glasses to see the 3D effect isn't really 3D either. I mean, it is still just "tricking" the brain into seeing 3D. Just like adding shading and perspective to a rendering "tricks" the brain into thinking it is 3D. What is the essential difference? The only thing that would be truely 3D would be a holographic projection of a game.
Well, I would have to say there isn't anything "amazing" about Debian either. They both just work.
Indeed. I've use a good number of Linux distributions (and Solaris, HPUX, etc) and now FreeBSD and it is all unix (with a lower case "u") as far as I am concerned. As long as it has sane and robust package management and a straighforward upgrade path, I'm happy.
As for the ports being bleeding edge? No, I would say they stay caught up with the latest "stable" releases from the respective projects, but they sure aren't bleeding edge.
I've been a Linux user for the last 10 years or so... Debian for the latter half. I recently took a new job where they used FreeBSD servers rather than Linux. Instead of using my position to push a migration to Linux, I decided to just learn to use FreeBSD.
I can't say that I am particularly impressed with FreeBSD. There's nothing WRONG with it, per se, but there is also nothing amazing either. The only redeeming value I can think of off hand is having bleeding edge software available all the time through ports. Where with Debian I would get "stuck" with package versions dated from whenever the last stable release was and mixing unstable packages was not a good idea. Coming from Gentoo, I know you have something like ports and you are used to compiling every darn package you want to run (I hate it). You should probably give FreeBSD a try. You might like it.
On a well-designed system it is absolutely realistic to be able to restart it, especially since so much kernel code is drivers for non-essential hardware. If my sound card drivers crash, just block on any reads/writes until they're restarted. Same for my network card, cd drives, even hard disk should be alright for short periods. Heck, there is no reason this shouldn't work for losing a cpu on an SMP system, or a stick of ram provided the core kernel parts (scheduler etc.) weren't on that one.
Indeed, a well designed system should be fault tolerant. Solaris, for example, can recover from CPU failure and RAM failure. But that doesn't have much to do with microkernels. As for other drivers failing, well, I'm not so sure it is really that big of a problem to be talking about radical OS redesigns. Especially one that, traditionally, has significant performance penalties. I think most people would rather have a system that runs fast all the time vs. one that can handle that rare event when some faulty code in their sound driver writes to the wrong bit of RAM or whatever.
Well, duh, you load the "bootstrap" disk subsystem from ROM, obviously. Then you can load the "real" (full-featured, with permission checking, journalling) disk subsystem; start that up, and deactivate the bootstrap system.
PCs don't have filesystem aware ROM.
Why wouldn't I want to keep the rest of the system running as if nothing happened?
Because, as I said in the text you quoted, you have no idea what the crashed disk subsystem did or did not write to disk. If you continue as if nothing happened, you coudl REALY fuck up your data.
Anyway, I am not aware of "disk subsystem failure" to be a major problem in modern operating systems. Disks fail, sure, but how often does the subsystem just randomly cause a kernel panic?
Obviously, you would have to keep in mind that the assumptions have changed. It would no longer be acceptable for a critical program to crash if the whole file system was yanked out from under it.
Like others who have replied to me, I think you are going beyond what a microkernel can accomplish. If a program is in the middle of a disk write operation, either that program has a backup plan when that write times out or it crashes. The kernel isn't of much help.
Put that analogy to a server: if the SCSI system dies, it may be advantageous to inform all server processes that the filesystem is no longer accessible, so that they in turn can inform already connected or newly connecting clients that something is wrong, instead of just waiting for things to time out (remember those hanging NFS mounts with Zen-like time-out periods?)
So your major complaint, in this case, is with timeouts and client communication, not micro vs. monolithic kernels.
In short, gracefully degrading service while informing those affected is way better than just "disappearing from the radar" with no clue to clients whatsoever.
Either way, the clients can't write to the server so what difference does it make? I know if an NFS server disappears, I get kernel logs saying such. How would a microkernel help?
If you store a copy of the disk driver relating to whichever disk holds further drivers in a portion of memory marked read only, then you can restart any driver. Same would apply for the network driver on a diskless system.
I don't think there is generally a driver per disk. It would be the whole subsystem. It seem a little excessive to be storing "backup" copies of drivers in memory so you can restart them. I'm no kernel hacker, but I do know that hardware can get into some strange states that are difficult to recover from short of a reboot.
Running the system as if nothing happened would be disasterous. But choosing appropriate action based on the problem of one subsystem, which may or may not be reboot it anyway. If the hardware is faulty, then no amount of software will fix it. But if a driver dies, and the system can remain usable, the driver's author could be sent debugging information. Notifiying the user may also be important. It sucks when the kernel crashes and document I'm writing gets lost. But if there was a hard drive crash, and I could still print it, or save it to a network device, that would be improvement.
It think this goes beyond what a microkernel can accomplish depending on what subsystem has a problem. The printing subsystem in your favorite OS is going to require the ablity to write to disk (spooling). Part of your word processing program could be swapped to disk which contains the code needed to save a document. How to swap that back into RAM? What if it is the paging system the frieks out? Your program is going to hang whether you have a microkernel or a monolithic kernel.
And if it wasn't the HDD, at the very least the system could do a clean shutdown. Both are better alternatives than just killing everything.
Depends on the problem. I've had several Linux kernel problems, for example, that didn't hang the machine. There are kernel panics, and then there are just "OOPS"s. I don't know how much better a microkernel system would fare. Again, depends on what goes wrong.
The goal is to have a system such that you maximize the segregation of the parts. If the SCSI subsystem crashes -- for example -- you flush it and restart it.
If the disk subsystem crashes, how exactly do you get at the on disk program to start it with? Are you going to try to restart it without actually reloading it? Would you trust it?
While it may not be possible to totally isolate every subsystem, with a microkernel subsystems should be more robust than in monolithic kernels.
The different between theory and reality is that, in theory, there is no difference.
For all of Linus' scorn of microkernels, Linux borrows heavily from the concept, if not from the theory. One could almost say that Linux implements a microkernel poorly through the kernel module interface. It fails to be a true microkernel in a number of ways, though, not least of which is the low degree to which it isolates modules.
Likewise, you could say that OS X implements a microkernel poorly (or is it a poor microkernel?). But in the end, is it any faster or more stable or more modular than your average monolithic kernel? I don't mean to bring up OS X as an example of microkernel's fatal flaws, but is there another example of a microkernel based system that is much better? Wasn't NT4 microkernel based? Did that help it?
In any case, your nervousness about a system where a "fundamental" subsystem craps out is understandable in someone who's main experience is with monolithic kernels, because the corruption of one subsystem often nfects other systems. For example, IME when the Linux SCSI module starts barfing (which happens with distressing regularity), if you're lucky, you can unload and reload the SCSI modules, but eventually, you're going to have to reboot, because it never quite works well after a reload.
What makes you think that a microkernel would necessarily fair much better in this case? In my experience, SCSI problems are as much hardware as software.
A corruption in one subsystem shouldn't lead to corruptions in any other subsystem.
The question remains, what is the value in running a system with one of its basic subsystems (disk I/O in the above example) crashed? If the engine room of a ship gets breached, the ship stays afloat. Great. Hardly anyone dies. If the disk subsystem of an OS craps out, well, thats bad. You're probably going to have to reboot it anyway and you certainly don't want the rest of the system to run as if nbothing happened because you have no idea what the disk subsystem did or did not write to disk before it crashed. Continuing as if nothing happened would be disasterous.
I'm not questioning the compartmentalized ship idea, I'm questioning the compartmentalize kernel. You can reboot a kernel. You can't just reboot a ship. A ship NEEDS to stay afloat no matter what. An operating system doesn't. Sometimes it is better that a kernel panic (sink) rather than continue running in a crippled and possibly corrupt state. There is a reason why the kernel kills processes that segfault rather than try to keep them running. If you keep them running, you risk (more) data corruption.. and that is generally considered worse than a little downtime.
That said, modern operating systems are already pretty well compartmentalized with the majority of code running in user space. The kernel is actually a relatively small part of the whole system. At some point it becomes a matter of diminishing returns. You COULD break down the kernel into smaller parts and maybe that would be a little more elegant, but in the real world it probably isn't going to make much difference.
How can the average user see this? When "Software Update" runs, almost any update to the system (not updates to an Apple application like iTunes) will require a restart of the whole machine. In a true Microkernel design you might need to relaunch the Finder or restart the communications architecture, but unless something changes kernel space code you wouldn't need to restart the whole computer.
The uptime command would give Apple proponents much more to brag about if it were a true microkernel,
Well, that is very much oversimplified. Just because Apple requires a reboot doesn't mean that they necessarily updated active kernel space code. It is just that it is safer and easier to reboot the machine than to try to restart just the htings that were updated and hope everything goes OK. Perhaps if Apple was really all that concerned about uptime, they would do that. When I update my Linux box, on the other hand, I rarely require a reboot because the kernel is not updated very often and Debian is pretty good about restarting updated services. And if it doesn't, I can do it. I can even reload kernel modules if I have to. So why would I want to run a microkernel?
but beyond hardware abstraction I don't think Apple has the same needs for a microkernel architecture as others. Since that's the case, I don't think it's fair to hold it up as an example of the fatal sins of microkernels in general.
I'm not holding up OS X as an example of the fatal sins of microkernel, I'm trying to get at the real (not theoretical) advantages and some good examples. So far it is mostly theory.
Nor do I think dragging in your personal valuations of speed and stability are rigorous indictments of Mac OS X's performance either.
It is hardly a secret that OS X lags behind in various server oriented benchmarks.
The problem I have with the compartmentalized ship metaphor is that I question the value of being able to run a system that has one compartment "breached." The system may technically still run, but is it goign to be of any use in such a state? Aren't you going to want to reboot it anyway or is the theory that you can restart a component without rebooting? Is this realistic? Seems to me that a system would get into a pretty funky state depending on which component failed.
If microkernels are so great and hte idea has been around for more than 15 years, why is there yet an example of one that lives up to everything Mr. Tannenbaum talks about? What true microkernel are/were there? NT4? Yeah, we all know how that went. Mach? OS X? Ok, I guess, but it is slow and no more stable or secure than any other Unix. In fact, I've had OS X machines crash far more than any Linux or FreeBSD systems I've used.
Linux, FreeBSD, and Solaris (examples of modern monolithic kernels) are pretty damn good. Not perfect, obviously, but you can easily get the 5 nines uptime out of these given good enough hardware and applications. As for security... how often are systems exploited at the kernel level? Sure, Linux had a module loading bug, but aren't the vast majority of exploits application based? Buffer overflows? Cross-site scripting flaws?
You (I understand) would add that faith can go much deeper, into areas that are not subject to reason, but I would say that this is of no value.
Well, when you are on your deathbed and cannot rationalize any life after death, you'll probably see just how valuable faith beyond reason is.
As for faith being a gift from God - yes, I think that is right. Faith in God increases as you get to know him better, and the fact that we can get to know him at all is purely down to his initiative. So I'd say that very much makes faith a gift.
This totally contradicts your idea of faith being based on reason. Either faith is a gift from God or it is something you create based on reason and experience. Which one is it?
Then why do scientists support this expansion theory with the red shift. It seems like this red shift is only representative of expansion in the three dimensions we are familiar with?
Yeah, so?
Which leads back to the question, where is the center of this expansion?
Since every point is moving away from every other point, there is no center. Or another way of looking at it: every point is a (relative) center. But there is no THE center.
To answer your first question (and I haven't really thought this through a great deal), I'd say that 'belief' is always provisional and often tentative. If I say "I believe he'll be here in time" I'm expressing a small element of doubt. It's subjective, not objective. My belief might be wrong. If I say "He'll be here in time", I'm leaving no room for doubt. If I say "I have faith that he'll be here in time" I mean that I know him as a punctual sort who won't let us down.
But you said earlier that you would consider modifying your faith if it contradicted experience. You seem to be leaving some room for doubt regarding your faith. It is conditional and the only real different between belief and faith is degree, neither being absolute.
I've actually heard that faith is a gift from God and not something that one creates oneself. Would God give you conditional, tenative faith?
Yes, when many people say "faith" they mean holding something to be true for no real reason. In fact, that is the second definition for the word here [reference.com]. However, I would agree with the grandparent that such faith is intellectually dishonest, and I would add pointless. Who cares what you have faith in if you have no defendable reason for that faith?
Indeed, it is wise to keep one's faith to one's self where it belongs.
To use that meaning of the word makes it something only worthy of ridicule.
Ridiculing people for considering a deeper meaning of faith only make you look ridiculous.
This artificial separation of religious faith and reason baffles me. Unless one posits a perverse God who purposely structures the universe so that observation would lead away from the truth about Him, then one would expect true assertions about God to be at least consistent with reason applied to observation.
I've expected lots of things that just didn't pan out.
The grandparent mentions his Christian faith. Whether you think Christianity is true or not, Christian faith does not have to be blind belief in something one cannot support. Rather, the Christian's faith is trust in a God who the Christian has concluded has demonstrated Himself to be trustworthy.
Ideally, love, trust, and faith would be unconditional. I look to that ideal.
The Christian's reasons for making this conclusion can be examined critically. For example, Christianity makes strong claims of fact which can be the subject of historical investigation. If Jesus Christ was resurrected from the dead as described in the Gospels, then that is strong evidence that there is something to this Christianity thing. If not, then even the apostle Paul says it is all pointless.
Nothing against Christianity, but I find such reliance on emperical evidence and historical accuracy to be rather shallow and unsatisfying. Maybe it is pointless. Maybe that is the point. Maybe we do what we do for no other purpose than to do it. Like art for art's sake. Maybe the point is to have faith for no reason at all... just to have faith.
Now, apply that to hate and mistrust and malice. If you dropped all reasons, would they exist? Would you hold on to hate if you had no reason for it? Would you be angry for the sake of anger?
Sorry to disagree, but in my book faith is always based on experience. It's strange how as soon as we start talking about religion, the word 'faith' seems to get used - mainly by the non-religious - to mean 'blind faith', quite to the contrary of how we normally use it.
Then how do you differential between beliefs and faith? Is there no difference?
If I said I had faith in my boss at work, you would assume that I'd seen good things from him in the past, hence I'd coem to trust him.
If I were to say something like that, I would mean that I thought my boss would do th eright thing DESPITE what i know of him or her. It is an statement of unconditional trust. If I say I believe in my boss, that just means he or she has a good history and I predict (althoght I'm not sure) that this pattern will continue. Although in practice I would probalby tend to mix terms out of intellectual laziness.
If I said I had lost faith in someone, you would assume that something happened, not that I had randomly decided not to trust that person any more for no reason. But as soon as I say I have faith in God, people assume that for it to be true faith it cannot be based on evidence or it wouldn't be faith! How silly.
From the beginning of the Wikipedia article on Faith:
The word faith has various uses; its central meaning is similar to "belief", "trust" or "confidence", but unlike these terms, "faith" tends to imply a transpersonal rather than interpersonal relationship - with God or a higher power.
I'm going with the latter meaning here. In interpresonal relationships, faith, belief, trust, are indeed nearly synonomous. But when it comes to spirituality, faith takes on a whole new meaning which theologians and philosophers have been pondering for ages. For me, faith is an (trans)personal think. I'll go on and on about beliefs and ideas and stuff. All that is worldly and relative. But faith is soemthing else. I won't try to tell you what it should mean for you, but suffice it to say that I think that equating belief and faith when it comes to religion cheapens your religion. I also think that if you get your faith out of a book (Bible?), you are cheating yourself. Sacred texts are a good read and all, but when it comes right down to it, it is just words with various possible interpretations.
Although some may find it controversial, some women breastfeed their offspring for as many as 3 to (rarely) 7 years from birth.
Even ignorning that, the question still stands, what are we protecting children from? What is it about nudity that might harm a child psychologically?
What is even more stupid is how we differential between bare skin and clothing when the clothing is skin tight. Few people have any problem with bathing suits, right? You take you kid to teh beach and there are men and women walking aroudn with barely a spandex stitch covering things. What else is there to see? A nipple? A butt crack? Some pubic hair! Oooo! The horror! Not that I necessarily WANT to see everyone naked, because I don't. But that is only because I find that most people look ugly without cloths on. It certainly isn't shameful or offensive. And it most certainly isn't a negative influence on a child.
What makes you think the universe has a size in any meaninful sense of the word? Do you think there is an edge of the universe that you will fall off of if you go too far?
And if there is a "size," what makes you think that the current universe would be the basis of a new one?
Except that faith has to be based on reality, otherwise it would be intellectually dishonest.
No, truth faith is based on nothing otherwise it is just belief.
From my own point of view as a Christian, if something that the Bible appeared to hold as true flatly contradicted what I knew to be true from my own experience then I would have to seriously re-examine either my understanding of the Bible, or my understanding of my experience. If the two are in contradiction, then one is wrong.
Those are just beliefs. You don't get faith from a book. Beliefs change with interprtations. Beliefs evolve. Beliefs are relative to culture. Faith is something else.
Totally, kinda like when scientists act all surprised when they discover that the "elementary partical" that they thought was fundamental and indivisable turns out to be made of even smaller particles. Its a fractal, duh!
Slashdot Mirror
Since evolution has no specific goal, there is really no meaningful definiiton of "devolving." Any change in allele frequency over time is evolution, whether or not you think it is a good change. Also, the part of the population which reproduces most successfully is, by definition, most suited... at least from an evolutionary persepective. There may be some sociological idea of how well suited a group is.. i dunno.
-matthew
Hmm... now curing cancer is nice and all, but if/when applied to humans, does this mean they can smoke cigarettes w/o ill effect,
;)
Cancer is hardly the only ill effect of smoking cigarettes. It is just hte one that gets the most press because, well, it is cancer. Smoking is one horrible, horrible thing you can do to your body. Even without cancer.
clean up nuclear waste with their bare hands,
Again, cancer isn't the only ill effect here...
r travel in space for extended duration w/o ill effect?
Again...
-matthew
Speaking as someone with poor depth perception, I don't see a real difference between "real" 3D and what games call 3D. And if you really want to get pedantic, you could just as easily say that requiring special glasses to see the 3D effect isn't really 3D either. I mean, it is still just "tricking" the brain into seeing 3D. Just like adding shading and perspective to a rendering "tricks" the brain into thinking it is 3D. What is the essential difference? The only thing that would be truely 3D would be a holographic projection of a game.
-matthew
Well, I would have to say there isn't anything "amazing" about Debian either. They both just work.
;-)
Indeed. I've use a good number of Linux distributions (and Solaris, HPUX, etc) and now FreeBSD and it is all unix (with a lower case "u") as far as I am concerned. As long as it has sane and robust package management and a straighforward upgrade path, I'm happy.
As for the ports being bleeding edge? No, I would say they stay caught up with the latest "stable" releases from the respective projects, but they sure aren't bleeding edge.
Coming from Debian, that *is* bleeding edge.
-matthew
I've been a Linux user for the last 10 years or so... Debian for the latter half. I recently took a new job where they used FreeBSD servers rather than Linux. Instead of using my position to push a migration to Linux, I decided to just learn to use FreeBSD.
I can't say that I am particularly impressed with FreeBSD. There's nothing WRONG with it, per se, but there is also nothing amazing either. The only redeeming value I can think of off hand is having bleeding edge software available all the time through ports. Where with Debian I would get "stuck" with package versions dated from whenever the last stable release was and mixing unstable packages was not a good idea. Coming from Gentoo, I know you have something like ports and you are used to compiling every darn package you want to run (I hate it). You should probably give FreeBSD a try. You might like it.
-matthew
On a well-designed system it is absolutely realistic to be able to restart it, especially since so much kernel code is drivers for non-essential hardware. If my sound card drivers crash, just block on any reads/writes until they're restarted. Same for my network card, cd drives, even hard disk should be alright for short periods. Heck, there is no reason this shouldn't work for losing a cpu on an SMP system, or a stick of ram provided the core kernel parts (scheduler etc.) weren't on that one.
Indeed, a well designed system should be fault tolerant. Solaris, for example, can recover from CPU failure and RAM failure. But that doesn't have much to do with microkernels. As for other drivers failing, well, I'm not so sure it is really that big of a problem to be talking about radical OS redesigns. Especially one that, traditionally, has significant performance penalties. I think most people would rather have a system that runs fast all the time vs. one that can handle that rare event when some faulty code in their sound driver writes to the wrong bit of RAM or whatever.
-matthew
Well, duh, you load the "bootstrap" disk subsystem from ROM, obviously. Then you can load the "real" (full-featured, with permission checking, journalling) disk subsystem; start that up, and deactivate the bootstrap system.
PCs don't have filesystem aware ROM.
Why wouldn't I want to keep the rest of the system running as if nothing happened?
Because, as I said in the text you quoted, you have no idea what the crashed disk subsystem did or did not write to disk. If you continue as if nothing happened, you coudl REALY fuck up your data.
Anyway, I am not aware of "disk subsystem failure" to be a major problem in modern operating systems. Disks fail, sure, but how often does the subsystem just randomly cause a kernel panic?
Obviously, you would have to keep in mind that the assumptions have changed. It would no longer be acceptable for a critical program to crash if the whole file system was yanked out from under it.
Like others who have replied to me, I think you are going beyond what a microkernel can accomplish. If a program is in the middle of a disk write operation, either that program has a backup plan when that write times out or it crashes. The kernel isn't of much help.
-matthew
Put that analogy to a server: if the SCSI system dies, it may be advantageous to inform all server processes that the filesystem is no longer accessible, so that they in turn can inform already connected or newly connecting clients that something is wrong, instead of just waiting for things to time out (remember those hanging NFS mounts with Zen-like time-out periods?)
So your major complaint, in this case, is with timeouts and client communication, not micro vs. monolithic kernels.
In short, gracefully degrading service while informing those affected is way better than just "disappearing from the radar" with no clue to clients whatsoever.
Either way, the clients can't write to the server so what difference does it make? I know if an NFS server disappears, I get kernel logs saying such. How would a microkernel help?
-matthew
If you store a copy of the disk driver relating to whichever disk holds further drivers in a portion of memory marked read only, then you can restart any driver. Same would apply for the network driver on a diskless system.
I don't think there is generally a driver per disk. It would be the whole subsystem. It seem a little excessive to be storing "backup" copies of drivers in memory so you can restart them. I'm no kernel hacker, but I do know that hardware can get into some strange states that are difficult to recover from short of a reboot.
Running the system as if nothing happened would be disasterous. But choosing appropriate action based on the problem of one subsystem, which may or may not be reboot it anyway. If the hardware is faulty, then no amount of software will fix it. But if a driver dies, and the system can remain usable, the driver's author could be sent debugging information. Notifiying the user may also be important. It sucks when the kernel crashes and document I'm writing gets lost. But if there was a hard drive crash, and I could still print it, or save it to a network device, that would be improvement.
It think this goes beyond what a microkernel can accomplish depending on what subsystem has a problem. The printing subsystem in your favorite OS is going to require the ablity to write to disk (spooling). Part of your word processing program could be swapped to disk which contains the code needed to save a document. How to swap that back into RAM? What if it is the paging system the frieks out? Your program is going to hang whether you have a microkernel or a monolithic kernel.
And if it wasn't the HDD, at the very least the system could do a clean shutdown. Both are better alternatives than just killing everything.
Depends on the problem. I've had several Linux kernel problems, for example, that didn't hang the machine. There are kernel panics, and then there are just "OOPS"s. I don't know how much better a microkernel system would fare. Again, depends on what goes wrong.
-matthew
The goal is to have a system such that you maximize the segregation of the parts. If the SCSI subsystem crashes -- for example -- you flush it and restart it.
If the disk subsystem crashes, how exactly do you get at the on disk program to start it with? Are you going to try to restart it without actually reloading it? Would you trust it?
While it may not be possible to totally isolate every subsystem, with a microkernel subsystems should be more robust than in monolithic kernels.
The different between theory and reality is that, in theory, there is no difference.
For all of Linus' scorn of microkernels, Linux borrows heavily from the concept, if not from the theory. One could almost say that Linux implements a microkernel poorly through the kernel module interface. It fails to be a true microkernel in a number of ways, though, not least of which is the low degree to which it isolates modules.
Likewise, you could say that OS X implements a microkernel poorly (or is it a poor microkernel?). But in the end, is it any faster or more stable or more modular than your average monolithic kernel? I don't mean to bring up OS X as an example of microkernel's fatal flaws, but is there another example of a microkernel based system that is much better? Wasn't NT4 microkernel based? Did that help it?
In any case, your nervousness about a system where a "fundamental" subsystem craps out is understandable in someone who's main experience is with monolithic kernels, because the corruption of one subsystem often nfects other systems. For example, IME when the Linux SCSI module starts barfing (which happens with distressing regularity), if you're lucky, you can unload and reload the SCSI modules, but eventually, you're going to have to reboot, because it never quite works well after a reload.
What makes you think that a microkernel would necessarily fair much better in this case? In my experience, SCSI problems are as much hardware as software.
A corruption in one subsystem shouldn't lead to corruptions in any other subsystem.
The question remains, what is the value in running a system with one of its basic subsystems (disk I/O in the above example) crashed? If the engine room of a ship gets breached, the ship stays afloat. Great. Hardly anyone dies. If the disk subsystem of an OS craps out, well, thats bad. You're probably going to have to reboot it anyway and you certainly don't want the rest of the system to run as if nbothing happened because you have no idea what the disk subsystem did or did not write to disk before it crashed. Continuing as if nothing happened would be disasterous.
-matthew
I'm not questioning the compartmentalized ship idea, I'm questioning the compartmentalize kernel. You can reboot a kernel. You can't just reboot a ship. A ship NEEDS to stay afloat no matter what. An operating system doesn't. Sometimes it is better that a kernel panic (sink) rather than continue running in a crippled and possibly corrupt state. There is a reason why the kernel kills processes that segfault rather than try to keep them running. If you keep them running, you risk (more) data corruption.. and that is generally considered worse than a little downtime.
That said, modern operating systems are already pretty well compartmentalized with the majority of code running in user space. The kernel is actually a relatively small part of the whole system. At some point it becomes a matter of diminishing returns. You COULD break down the kernel into smaller parts and maybe that would be a little more elegant, but in the real world it probably isn't going to make much difference.
-matthew
How can the average user see this? When "Software Update" runs, almost any update to the system (not updates to an Apple application like iTunes) will require a restart of the whole machine. In a true Microkernel design you might need to relaunch the Finder or restart the communications architecture, but unless something changes kernel space code you wouldn't need to restart the whole computer.
The uptime command would give Apple proponents much more to brag about if it were a true microkernel,
Well, that is very much oversimplified. Just because Apple requires a reboot doesn't mean that they necessarily updated active kernel space code. It is just that it is safer and easier to reboot the machine than to try to restart just the htings that were updated and hope everything goes OK. Perhaps if Apple was really all that concerned about uptime, they would do that. When I update my Linux box, on the other hand, I rarely require a reboot because the kernel is not updated very often and Debian is pretty good about restarting updated services. And if it doesn't, I can do it. I can even reload kernel modules if I have to. So why would I want to run a microkernel?
but beyond hardware abstraction I don't think Apple has the same needs for a microkernel architecture as others. Since that's the case, I don't think it's fair to hold it up as an example of the fatal sins of microkernels in general.
I'm not holding up OS X as an example of the fatal sins of microkernel, I'm trying to get at the real (not theoretical) advantages and some good examples. So far it is mostly theory.
Nor do I think dragging in your personal valuations of speed and stability are rigorous indictments of Mac OS X's performance either.
It is hardly a secret that OS X lags behind in various server oriented benchmarks.
-matthew
Why in the world is this modded flaimbait? It is a totally sincere and valid inquiry! I'm seriously looking for real world examples here.
-matthew
The problem I have with the compartmentalized ship metaphor is that I question the value of being able to run a system that has one compartment "breached." The system may technically still run, but is it goign to be of any use in such a state? Aren't you going to want to reboot it anyway or is the theory that you can restart a component without rebooting? Is this realistic? Seems to me that a system would get into a pretty funky state depending on which component failed.
-matthew
If microkernels are so great and hte idea has been around for more than 15 years, why is there yet an example of one that lives up to everything Mr. Tannenbaum talks about? What true microkernel are/were there? NT4? Yeah, we all know how that went. Mach? OS X? Ok, I guess, but it is slow and no more stable or secure than any other Unix. In fact, I've had OS X machines crash far more than any Linux or FreeBSD systems I've used.
Linux, FreeBSD, and Solaris (examples of modern monolithic kernels) are pretty damn good. Not perfect, obviously, but you can easily get the 5 nines uptime out of these given good enough hardware and applications. As for security... how often are systems exploited at the kernel level? Sure, Linux had a module loading bug, but aren't the vast majority of exploits application based? Buffer overflows? Cross-site scripting flaws?
-matthew
You (I understand) would add that faith can go much deeper, into areas that are not subject to reason, but I would say that this is of no value.
Well, when you are on your deathbed and cannot rationalize any life after death, you'll probably see just how valuable faith beyond reason is.
As for faith being a gift from God - yes, I think that is right. Faith in God increases as you get to know him better, and the fact that we can get to know him at all is purely down to his initiative. So I'd say that very much makes faith a gift.
This totally contradicts your idea of faith being based on reason. Either faith is a gift from God or it is something you create based on reason and experience. Which one is it?
-matthew
Then why do scientists support this expansion theory with the red shift. It seems like this red shift is only representative of expansion in the three dimensions we are familiar with?
Yeah, so?
Which leads back to the question, where is the center of this expansion?
Since every point is moving away from every other point, there is no center. Or another way of looking at it: every point is a (relative) center. But there is no THE center.
-matthew
It is relative expansion. Everything is expanding relative to everything else (for the most part).
To answer your first question (and I haven't really thought this through a great deal), I'd say that 'belief' is always provisional and often tentative. If I say "I believe he'll be here in time" I'm expressing a small element of doubt. It's subjective, not objective. My belief might be wrong. If I say "He'll be here in time", I'm leaving no room for doubt. If I say "I have faith that he'll be here in time" I mean that I know him as a punctual sort who won't let us down.
But you said earlier that you would consider modifying your faith if it contradicted experience. You seem to be leaving some room for doubt regarding your faith. It is conditional and the only real different between belief and faith is degree, neither being absolute.
I've actually heard that faith is a gift from God and not something that one creates oneself. Would God give you conditional, tenative faith?
-matthew
Yes, when many people say "faith" they mean holding something to be true for no real reason. In fact, that is the second definition for the word here [reference.com]. However, I would agree with the grandparent that such faith is intellectually dishonest, and I would add pointless. Who cares what you have faith in if you have no defendable reason for that faith?
Indeed, it is wise to keep one's faith to one's self where it belongs.
To use that meaning of the word makes it something only worthy of ridicule.
Ridiculing people for considering a deeper meaning of faith only make you look ridiculous.
This artificial separation of religious faith and reason baffles me. Unless one posits a perverse God who purposely structures the universe so that observation would lead away from the truth about Him, then one would expect true assertions about God to be at least consistent with reason applied to observation.
I've expected lots of things that just didn't pan out.
The grandparent mentions his Christian faith. Whether you think Christianity is true or not, Christian faith does not have to be blind belief in something one cannot support. Rather, the Christian's faith is trust in a God who the Christian has concluded has demonstrated Himself to be trustworthy.
Ideally, love, trust, and faith would be unconditional. I look to that ideal.
The Christian's reasons for making this conclusion can be examined critically. For example, Christianity makes strong claims of fact which can be the subject of historical investigation. If Jesus Christ was resurrected from the dead as described in the Gospels, then that is strong evidence that there is something to this Christianity thing. If not, then even the apostle Paul says it is all pointless.
Nothing against Christianity, but I find such reliance on emperical evidence and historical accuracy to be rather shallow and unsatisfying. Maybe it is pointless. Maybe that is the point. Maybe we do what we do for no other purpose than to do it. Like art for art's sake. Maybe the point is to have faith for no reason at all... just to have faith.
Now, apply that to hate and mistrust and malice. If you dropped all reasons, would they exist? Would you hold on to hate if you had no reason for it? Would you be angry for the sake of anger?
-matthew
Then how do you differential between beliefs and faith? Is there no difference?
If I said I had faith in my boss at work, you would assume that I'd seen good things from him in the past, hence I'd coem to trust him.
If I were to say something like that, I would mean that I thought my boss would do th eright thing DESPITE what i know of him or her. It is an statement of unconditional trust. If I say I believe in my boss, that just means he or she has a good history and I predict (althoght I'm not sure) that this pattern will continue. Although in practice I would probalby tend to mix terms out of intellectual laziness.
If I said I had lost faith in someone, you would assume that something happened, not that I had randomly decided not to trust that person any more for no reason. But as soon as I say I have faith in God, people assume that for it to be true faith it cannot be based on evidence or it wouldn't be faith! How silly.
From the beginning of the Wikipedia article on Faith: I'm going with the latter meaning here. In interpresonal relationships, faith, belief, trust, are indeed nearly synonomous. But when it comes to spirituality, faith takes on a whole new meaning which theologians and philosophers have been pondering for ages. For me, faith is an (trans)personal think. I'll go on and on about beliefs and ideas and stuff. All that is worldly and relative. But faith is soemthing else. I won't try to tell you what it should mean for you, but suffice it to say that I think that equating belief and faith when it comes to religion cheapens your religion. I also think that if you get your faith out of a book (Bible?), you are cheating yourself. Sacred texts are a good read and all, but when it comes right down to it, it is just words with various possible interpretations.
-matthew
Even ignorning that, the question still stands, what are we protecting children from? What is it about nudity that might harm a child psychologically?
What is even more stupid is how we differential between bare skin and clothing when the clothing is skin tight. Few people have any problem with bathing suits, right? You take you kid to teh beach and there are men and women walking aroudn with barely a spandex stitch covering things. What else is there to see? A nipple? A butt crack? Some pubic hair! Oooo! The horror! Not that I necessarily WANT to see everyone naked, because I don't. But that is only because I find that most people look ugly without cloths on. It certainly isn't shameful or offensive. And it most certainly isn't a negative influence on a child.
-matthew
What makes you think the universe has a size in any meaninful sense of the word? Do you think there is an edge of the universe that you will fall off of if you go too far?
And if there is a "size," what makes you think that the current universe would be the basis of a new one?
-matthew
Except that faith has to be based on reality, otherwise it would be intellectually dishonest.
No, truth faith is based on nothing otherwise it is just belief.
From my own point of view as a Christian, if something that the Bible appeared to hold as true flatly contradicted what I knew to be true from my own experience then I would have to seriously re-examine either my understanding of the Bible, or my understanding of my experience. If the two are in contradiction, then one is wrong.
Those are just beliefs. You don't get faith from a book. Beliefs change with interprtations. Beliefs evolve. Beliefs are relative to culture. Faith is something else.
-matthew
Totally, kinda like when scientists act all surprised when they discover that the "elementary partical" that they thought was fundamental and indivisable turns out to be made of even smaller particles. Its a fractal, duh!
-matthew