I would say that most people who compile their own kernel definitely optimize them for their own arch. Linux does a BETTER job of optimizing for each arch than Windows (ignoring ICCs ability to autovectorize some code, but that isn't exactly OS stuff for the most part). Nice that you mention x86-64, there are several completely x86-64 aware distros out there and only a beta-OS from Microsoft and very few x86-64 bit aware Windows apps versus pretty much everything for Linux...
The Debian reference is exactly what I mean, they can be really, really large...much larger than any Windows distribution.
Include a lot more data in the OS? What are you talking about? Every desktop version of Windows ever has been available at retail and fits on one CD, that's not a lot of data when you look at most of the Linux distros...obviously, the Linux distributions include a hell of a lot of stuff that no version of Windows does, but there is room for them to do any "optimizing" they need.
The Windows kernel is atleast, if not more, generic than the Linux kernel. Everything system specific is done in drivers on both platforms (I don't count "compiled for x86 versus PPC" to be system specific as obviously Windows is also compiled with a target CPU arch).
If you are implying that HP and Dell get to alter the XP kernel just so they can ship it on their systems, you are badly mistaken. They may ship their own "version" of Windows, but the alterations have almost all been done at the user-level and pre-installed drivers.
DVD install media only holds 4.7GB, there are several distros which are already approaching (or have already passed) this if you add up all the CDs...
While we are at it, what about cut and paste is broken in "Linux"? I know that X keeps its own copy buffer versus the window manager, but they are easy to sync in at least KDE...it is a simple click option in the klipper control screen.
He didn't mean the ACT of developing the software, he meant that it is entirely possible (even probable) that a user-space screwup (you know, those word processors and spreadsheets) will crash the whole OS, not just the GUI, not just the window manager, the whole thing can crash and burn.
Yea, but that isn't really the same thing. They aren't improving the kernel, then keeping the changes to themselves. They are allowing very limited support of their own hardware. They aren't in the Linux business in the same way that SGI, IBM, etc are. They just provide an interface to their own hardware, so if it doesn't work they are hurting themselves more than anyone else. These vendors are still primarily (99+%) Windows related, so they release a Linux driver in the same style as their Windows drivers, binary only.
The GPU vendors also have a decent (depending on your viewpoint) reason for doing this, their drivers have licensed software in them that they can't release as source. At least this was true at one time and all the modern graphics chips still support the covered algorithms (specifically ST3C if I'm not mistaken).
Also the "shims" aren't for GPL reasons, they are to allow their binary only driver to interface with at least a handful of different kernel versions, otherwise they would REALLY be busy updating their driver. With the current method, they only have to periodically release refreshes when something directly related to their driver or something major in the kernel changes.
You mean like the benefits IBM gives to Dell, RedHat, SGI, VA Linux, etc whenever it adds something to the kernel?
Or how about the benefits that RedHat gives IBM, Dell, SGI, VA Linux et al when IT adds something?
You are basically missing the real synergy being FOSS, one company contributes and helps its competition but those guys are also contributing and helping you. So in reality, you are helping yourself against the REAL competition (non-FOSS) and not your immediate "partners." Of course, there are leeches like Dell, but that's OK...
I like to change out my hardware a LOT, this is super easy to do in Linux and can vary from trivial to a re-install under Windows. I have tried various flavors of *nix over the years starting with an early Debian port to the Amiga...every time I tried a distro or two, they were a bit better than the ones six months before but still not quite "right." Until this past November when I tried out Gentoo, I figured it didn't have some of the problems I had with various other distros over the years and I was right. I love it, haven't booted back up into windows for anything other than games and even those reboots are getting more and more scarce.
I even recently convinced a non-techie to make the switch, he plays more games than I do but I think he'll stick it out and eventually be a true convert =)
This is easy to prove by looking at any one of millions of "optical illusions" which are basically instances where our brain makes faulty assumptions. If our eye were all it were cracked up to be, this wouldn't be happening but since most of the elegance is in the post-processing there are ways to "trick" vision...sometimes pretty substantially.
I highly doubt it is a physiological reason, simply an accumulation of knowledge. At the beginning of ones life, there is no knowledge but infinite information to be gained so we are extremely curious. As time progresses we learn more and more about the world but as we learn our drive to learn tends to decrease. Basically you know the answers to most of the questions you have about the world, not all, but enough to keep you from really exploring. Obviously these are generalizations, there are people who get MORE curious as they age (they realize that what they know is only the smallest tip of the iceberg of the vast amounts of knowledge that can be gained), but most do not. As a whole, people like the music they grew up with. That would mean that musical styles would pretty much become perpetually static and eventually there won't be "that crazy music the kids listen to" because there won't BE any kids.
The world is by definition ruled by the majority, over time people would wear down and conform then there would be no majority, only dim representations of ourselves. Life is only interesting because it is short, if we could live forever we would all wait for the best time to do something instead of living for the day!
Maybe that would lead to a more knowledgable populace, but not more creative. To put it bluntly, people in general become more conservative as time progresses in their lives and they are most creative when the world is still "new" to them. With a static population we would solve a lot of our current problems, but more importantly not think up any new ones for the future...in other words, the world would stagnate.
Well, the "floor reporter" is the host of Fine Living's _Thirsty Traveler_ which probably had a lot more to do with his placement than his food knowledge. While I like Traveler a LOT, it isn't a show about food and he seems to have only a moderate grasp of the cooking which IS done on the show.
I would actually say you are on the thin side a bit, without seeing you I won't pass judgement though =)
I know your pain in trying to buy jeans though, I am 6'4" and 185 pounds with a 34 inch waist. It seems that most places stock the waist size, just in jeans made for 5'8" people... I've actually bought several pair at American Eagle, they are normally fairly expensive by at any one time half of their inventory is on sale and I've gotten some cheap jeans there that I like a LOT.
Unfortunately that doesn't always work. I recently signed up for DirecTV and when I told the CSR on the phone my address she took down apartment 86 instead of apartment A6 but everything else was correct. I never received my first two statements and I live in a small apartment complex with roughly 60 others, have lived in the same apartment for several years (quite a bit longer than most...college town and I'm a grad student), have had the same mailman the entire time, and am the ONLY person in the phone book (town has roughly 60K people) with my last name.
Luckily I figured out the problem by checking online and caught it the last day before they started tacking on late charges. I corrected the address online and have received statements ever since.
RTFA? How about you RTWC (read the whole comment), he says exactly that a few lines down.
Public, means accessible and for the public good, not for YOUR good alone. For instance, a community park has a fence surrounding a dangerous sink hole. You are NOT allowed to go in and remove a section of the fence just because it is on public land even if you intended to go back in and replace the section.
I couldn't agree more...most of the problems with the USPS can be tracked down to individual depots. Specifically, I've noticed that anything which goes through Memphis, TN basically gets an extra 2-3 DAYS added to its journey. This is obviously anecdotal, but it has happened consistantly numerous times. Mail to my location goes through one of three nearby cities before reaching me, stuff through the other two is much faster even if it is a longer distance. Typically once I track something to Memphis with their online tracking, it doesn't even move for at least two days, sometimes three.
I imagine all other shipping companies have the same problems, the Dallas-Fort Worth depot of FedEx for instance. Most other depots are VERY fast, turnaround times of several HOURS...things almost invariably sit overnight at DFW, sometimes not leaving for nearly 48 hours.
Actually, I would say "Chernobyl on steroids" implies "like Chernobyl, but much larger and nastier" not "like Chernobyl, but cleaner and more environmently friendly"...
I would say that a coal, oil, or natural gas plant has a hell of a lot more fire potential and if the reactant leaks it isn't really that big a deal. The quantity of reactant used in this type of system is miniscule, deuterium isn't a problem at ALL and tritium isn't really all that bad either. I would say that a leak of ANY of the reactants or products would be better than the overwhelming majority of chemical spill type problems, then when the quantity of reactants and products is considered it becomes almost a non-issue. COOLANT leaks would probably be a lot worse than reactant leaks and that sort of thing happens at almost EVERY type of industrial facility.
Maybe no new plants have been built here, but Japan and France have definitely built new reactors since this time. Scientific evolution doesn't stop just because two countries stop building new reactors...hell a lot of the innovation has come from research here, even though we haven't deployed a new reactor here in a while.
I didn't...I claimed "[American's] played a huge role in the evolution of the modern computer" not that they invented it. Like I said in other posts, I was more highlighting American contributions to the field than stating every single invention or concept leading up to them.
Slashdot Mirror
I would say that most people who compile their own kernel definitely optimize them for their own arch. Linux does a BETTER job of optimizing for each arch than Windows (ignoring ICCs ability to autovectorize some code, but that isn't exactly OS stuff for the most part). Nice that you mention x86-64, there are several completely x86-64 aware distros out there and only a beta-OS from Microsoft and very few x86-64 bit aware Windows apps versus pretty much everything for Linux...
The Debian reference is exactly what I mean, they can be really, really large...much larger than any Windows distribution.
That's wierd, I copy and paste similiarly a LOT. How were you copying? and how were you pasteing?
Include a lot more data in the OS? What are you talking about? Every desktop version of Windows ever has been available at retail and fits on one CD, that's not a lot of data when you look at most of the Linux distros...obviously, the Linux distributions include a hell of a lot of stuff that no version of Windows does, but there is room for them to do any "optimizing" they need.
The Windows kernel is atleast, if not more, generic than the Linux kernel. Everything system specific is done in drivers on both platforms (I don't count "compiled for x86 versus PPC" to be system specific as obviously Windows is also compiled with a target CPU arch).
If you are implying that HP and Dell get to alter the XP kernel just so they can ship it on their systems, you are badly mistaken. They may ship their own "version" of Windows, but the alterations have almost all been done at the user-level and pre-installed drivers.
DVD install media only holds 4.7GB, there are several distros which are already approaching (or have already passed) this if you add up all the CDs...
While we are at it, what about cut and paste is broken in "Linux"? I know that X keeps its own copy buffer versus the window manager, but they are easy to sync in at least KDE...it is a simple click option in the klipper control screen.
He didn't mean the ACT of developing the software, he meant that it is entirely possible (even probable) that a user-space screwup (you know, those word processors and spreadsheets) will crash the whole OS, not just the GUI, not just the window manager, the whole thing can crash and burn.
Yea, but that isn't really the same thing. They aren't improving the kernel, then keeping the changes to themselves. They are allowing very limited support of their own hardware. They aren't in the Linux business in the same way that SGI, IBM, etc are. They just provide an interface to their own hardware, so if it doesn't work they are hurting themselves more than anyone else. These vendors are still primarily (99+%) Windows related, so they release a Linux driver in the same style as their Windows drivers, binary only.
The GPU vendors also have a decent (depending on your viewpoint) reason for doing this, their drivers have licensed software in them that they can't release as source. At least this was true at one time and all the modern graphics chips still support the covered algorithms (specifically ST3C if I'm not mistaken).
Also the "shims" aren't for GPL reasons, they are to allow their binary only driver to interface with at least a handful of different kernel versions, otherwise they would REALLY be busy updating their driver. With the current method, they only have to periodically release refreshes when something directly related to their driver or something major in the kernel changes.
I just didn't want to write "Free Open Source Software"
It was originally intended to be a short one liner...then it grew into a real comment.
You mean like the benefits IBM gives to Dell, RedHat, SGI, VA Linux, etc whenever it adds something to the kernel?
Or how about the benefits that RedHat gives IBM, Dell, SGI, VA Linux et al when IT adds something?
You are basically missing the real synergy being FOSS, one company contributes and helps its competition but those guys are also contributing and helping you. So in reality, you are helping yourself against the REAL competition (non-FOSS) and not your immediate "partners." Of course, there are leeches like Dell, but that's OK...
Just think of all the extra people walking around saying "whistles say whoo"
I like to change out my hardware a LOT, this is super easy to do in Linux and can vary from trivial to a re-install under Windows. I have tried various flavors of *nix over the years starting with an early Debian port to the Amiga...every time I tried a distro or two, they were a bit better than the ones six months before but still not quite "right." Until this past November when I tried out Gentoo, I figured it didn't have some of the problems I had with various other distros over the years and I was right. I love it, haven't booted back up into windows for anything other than games and even those reboots are getting more and more scarce.
I even recently convinced a non-techie to make the switch, he plays more games than I do but I think he'll stick it out and eventually be a true convert =)
I've seen more cyclists on the ped trail than peds on the bike trail to be honest...especially around Olive Park...
This is easy to prove by looking at any one of millions of "optical illusions" which are basically instances where our brain makes faulty assumptions. If our eye were all it were cracked up to be, this wouldn't be happening but since most of the elegance is in the post-processing there are ways to "trick" vision...sometimes pretty substantially.
I'm pretty sure Paramount's Berman and Bragga are really a test case for this sort of AI script writing...
I highly doubt it is a physiological reason, simply an accumulation of knowledge. At the beginning of ones life, there is no knowledge but infinite information to be gained so we are extremely curious. As time progresses we learn more and more about the world but as we learn our drive to learn tends to decrease. Basically you know the answers to most of the questions you have about the world, not all, but enough to keep you from really exploring. Obviously these are generalizations, there are people who get MORE curious as they age (they realize that what they know is only the smallest tip of the iceberg of the vast amounts of knowledge that can be gained), but most do not. As a whole, people like the music they grew up with. That would mean that musical styles would pretty much become perpetually static and eventually there won't be "that crazy music the kids listen to" because there won't BE any kids.
The world is by definition ruled by the majority, over time people would wear down and conform then there would be no majority, only dim representations of ourselves. Life is only interesting because it is short, if we could live forever we would all wait for the best time to do something instead of living for the day!
Maybe that would lead to a more knowledgable populace, but not more creative. To put it bluntly, people in general become more conservative as time progresses in their lives and they are most creative when the world is still "new" to them. With a static population we would solve a lot of our current problems, but more importantly not think up any new ones for the future...in other words, the world would stagnate.
Well, the "floor reporter" is the host of Fine Living's _Thirsty Traveler_ which probably had a lot more to do with his placement than his food knowledge. While I like Traveler a LOT, it isn't a show about food and he seems to have only a moderate grasp of the cooking which IS done on the show.
It is also value-less, so he is somewhat correct even if accidentally ;)
I would actually say you are on the thin side a bit, without seeing you I won't pass judgement though =)
I know your pain in trying to buy jeans though, I am 6'4" and 185 pounds with a 34 inch waist. It seems that most places stock the waist size, just in jeans made for 5'8" people... I've actually bought several pair at American Eagle, they are normally fairly expensive by at any one time half of their inventory is on sale and I've gotten some cheap jeans there that I like a LOT.
And everyone said Moore's law was dead...
Unfortunately that doesn't always work. I recently signed up for DirecTV and when I told the CSR on the phone my address she took down apartment 86 instead of apartment A6 but everything else was correct. I never received my first two statements and I live in a small apartment complex with roughly 60 others, have lived in the same apartment for several years (quite a bit longer than most...college town and I'm a grad student), have had the same mailman the entire time, and am the ONLY person in the phone book (town has roughly 60K people) with my last name.
Luckily I figured out the problem by checking online and caught it the last day before they started tacking on late charges. I corrected the address online and have received statements ever since.
RTFA? How about you RTWC (read the whole comment), he says exactly that a few lines down.
Public, means accessible and for the public good, not for YOUR good alone. For instance, a community park has a fence surrounding a dangerous sink hole. You are NOT allowed to go in and remove a section of the fence just because it is on public land even if you intended to go back in and replace the section.
I couldn't agree more...most of the problems with the USPS can be tracked down to individual depots. Specifically, I've noticed that anything which goes through Memphis, TN basically gets an extra 2-3 DAYS added to its journey. This is obviously anecdotal, but it has happened consistantly numerous times. Mail to my location goes through one of three nearby cities before reaching me, stuff through the other two is much faster even if it is a longer distance. Typically once I track something to Memphis with their online tracking, it doesn't even move for at least two days, sometimes three.
I imagine all other shipping companies have the same problems, the Dallas-Fort Worth depot of FedEx for instance. Most other depots are VERY fast, turnaround times of several HOURS...things almost invariably sit overnight at DFW, sometimes not leaving for nearly 48 hours.
Actually, I would say "Chernobyl on steroids" implies "like Chernobyl, but much larger and nastier" not "like Chernobyl, but cleaner and more environmently friendly"...
I would say that a coal, oil, or natural gas plant has a hell of a lot more fire potential and if the reactant leaks it isn't really that big a deal. The quantity of reactant used in this type of system is miniscule, deuterium isn't a problem at ALL and tritium isn't really all that bad either. I would say that a leak of ANY of the reactants or products would be better than the overwhelming majority of chemical spill type problems, then when the quantity of reactants and products is considered it becomes almost a non-issue. COOLANT leaks would probably be a lot worse than reactant leaks and that sort of thing happens at almost EVERY type of industrial facility.
Maybe no new plants have been built here, but Japan and France have definitely built new reactors since this time. Scientific evolution doesn't stop just because two countries stop building new reactors...hell a lot of the innovation has come from research here, even though we haven't deployed a new reactor here in a while.
I didn't...I claimed "[American's] played a huge role in the evolution of the modern computer" not that they invented it. Like I said in other posts, I was more highlighting American contributions to the field than stating every single invention or concept leading up to them.