Seems like most people aren'y thinking this through all the way.
First, it's likely that Intel would create a chip specifically for Apple. Volumes would still be high enough to make this possible. Intel does have experience with RISC chips. It seems very unlikely that Intel would migrate the P4 to Mac as that seems impractical. Pentium-M may be more likely, but I would still expect to see a RISC based chip (which Intel likely already has multiple design possibilities for). Intel has tried several times to break away from x86, this may factor into those attempts.
Second.. heat issues?! I keep hearing that Intel chips run too hot for Macs. Since when has the G5 been any cooler?! In any case, Intel is likely to create a RISC based chip which means a new design and most likely less heat.
Talks between Intel and Apple go way back. This isn't a new discussion. For various reasons Apple has never decided to go with Intel. Motorola was hopeless, and even IBM is a bit behind the game in terms of performance. The cost/performance ratio is still much better on the PC (by what seems like a factor of 2). If Apple is trying to drive prices down (which it seems like they are), this is an important factor.
Their best options would be AMD or Intel and AMD (IIRC) doesn't have much experience with RISC.
You can get away without writing much documenation these days. Usually the larger distros have how-tos for all common software. I just recently set up Shorewall on Gentoo and used the Gentoo documentation to do it. I looked around on Shorewall's site, had a hard time figuring it out.. and then found a Gentoo how-to that had a step by step guide on how to do it on my distro of choice (which is easier than a generic how-to anyway).
Let the documenation go, and just post the source code on the site.
kernel.org isn't exactly a documenation cornucopiea after all.
To get to The Moon you need to reach escape velocity. To get to Mars you also must reach escape veolocity. The only thing that makes getting to Mars more difficult is the longer journey. There really ISN'T that much point in stopping on the Moon if you're going to Mars.
Most of the liftoff weight for a spacecraft is fuel and it's up and the air wether or not The Moon is any minable form of usable fuel on it. We only know of traces at this point. If that remains the case than there really isn't much point for using The Moon as a stepping stone to Mars (it's more like a deviation).
Comments like this crack me up. For some reason people think that compiling Gentoo only involves sitting in front of a black screen watching compiliation lines scroll by. Actually, the system is quite usable while compiling. Occasionally (when I get the itch to play unreal or doom) I even suspend compilation, go play a few games, and then restart compilation (ctrl-z is your friend). You can even ctrl-c out of compilation and restart w/ emerge --resume. It really is that easy. Waiting for things to compile does not have to be painful at all. Pretty much anything other than playing heavy games can be done while compiling, and you can always resume your emerge if you get the gaming itch. If you do the install w/ something like Knoppix, you can even do almost anything during the initial install.
We're planning on moving to 5 in the near future, and I have been doing my development with it for a couple of months (obviously just using the php 4 syntax). I have noticed very few problems with our existing code. A few programs spit out error messages where there were none before, but they were actual coding problems. So far I would say 5 is nice and I'm looking forward to the more strict OOP features. PHP isn't the most well designed langauge out there, but it's better than most and 5 is a step in the right direction I think.
I better start migrating our cluster to Windows!!! Our linux servers get cracked open all the time and and everytime I add a new server it takes like an hour to set up!! It's a nightmare!! I trust Windows because it is rock solid and known for it's excellent security!!
So let's see.. it will cost $50,000 in license fees to move to Windows.
Not to beat a dead stick, but maybe if enough people say it, Netscape will realize that this browser is FUGLY!!!!
FUUUGGGLLY!!!
In fact it's so fugly, that I had an involuntary WTF reaction before the screenshot even fully loaded. I don't care if this was the fastest browser on the planet (unlikely), I would never use it based on ugliness alone.
Newsflash Netscape. It's the 2000s and busy browsers, and indeed busy anything are definitely NOT the way to try to look "fashionable." Even though I'm slightly embarassed at writing such a statement, it's definitely true.
I have used Windows, Solaris/CDE, Linux (many distros and UIs), and OSX. As a coder/sys admin I feel most comfortable with linux, Gentoo and KDE in particular. The reason for this is the ease of system administration and setup for the software I need while maintaining the ability to tweak it entirely to my likely. No Gentoo was not productive for the first few days. As an added bonus, there is a ton of free software available that just doesn't work as well on OSX or Windows, which can also raise my productivity (Umbrello, Kate, KDevelop, Gimp). It runs best (if it all) in Linux, and getting something equivelant (and native) on Windows or OSX is likely to cost a fair amount.
I recently got a new PowerBook through work, and while I love it, there are limitations when compared to linux. Previous users stated that setting up OSX as a server is a breeze. Sure, this is true if you're only using a basic server, but if you need an advanced configuration, OSX can be a true nightmare. The tools are available under the hood, but it is NOT designed to be tinkered with. My reasoning for choosing OSX for work was the gaurantee for driver compatibility and full support (it was either put linux on a Dell laptop or get the PowerBook). I figured I would take a slight productivity dip, and I did. Everyone has a different routine, and to maximize productivity an interface must be highly configurable. Of course, the average user is not savy enough to deal with such a high level of configuration. OSX is a compromise. It is the best all around, but Linux has greater potential when configured for specific tasks.
I guess Windows is great if you measure productivity in Frames Per Second.
To sum up my perspective:
Linux productivity sucks in most "out of the box" configurations, but has the highest potential.
Windows is just "OK." It gets you by and lets you play games. It is also good for office apps as OpenOffice.org really isn't that great IMO (sorry - I have to question it's design), which makes MS Office the best office software by default. KOffice and the Gnome based apps have a chance at changing this in the long run.
OSX is highly productive out of the box, but has little capacity for getting that maximum tweaking. It is also VERY different than Windows. Most Windows users will feel more productive in a generic linux install than OSX. It is also a big advantage to be able to run MS Office on OSX (at least until it crashes).
We can argue about what data means all day long, but really this just comes down to common sense. It's a reasonable assumption that pumping an atmostphere full of greenhouse gasses is going to cause global warming. Whether or not we currently are experiencing the results of this warming may be hard to say, but I would have to question the intelligence/motives of someone who said that doing this is not dangerous.
If you relate this to the concepts of warming another planet like Mars to make it more habitable, there is virtually complete agreement that pumping huge amounts of greenhouse gasses into the atmosphere would allow the planet to become significantly warmer.
Now.. how can anyone argue that doing this on Earth would not produce the same results? It's like telling someone that they don't have any proof that you would die if you jumped out of an airlock in space without a suit. No one has done that before (to my knowledge), but it's a safe assumption that it would make you very dead. In this instance, ignoring common sense would be beyond foolish.
It's even worse when you do it on a global level. Once we wreck our environment, then I guess we'll have some hard data on global warming. Hopefully we survive our stupidity, but it doesn't seem like we even deserve to survive as a species if we are so collectively foolish.
Right after I switch my entire environment from linux to windows due to the amazing overwhelmingly dangerous linux kernel exploit from a few days ago and the huge number of linux security issues that were said to be listed this year... this happens. Now my Win2003 server uptimes will get all jacked up.. and they just booted for the first time 2 days ago. Now I'm going to have to switch to OS/1337 in order to preserve my sys admin ego flames.
Don't use windows. Don't use linux. They both suck. OS/1337 has not once had a single crack or exploit discovered. Plus.. it has never crashed. Not once. OS/1337 will rock your world.
Well now that the movie has nothing to do with the game, we can still argue that the title is correct because this movie is definitely doomed.. Any hope that it would even be worth renting has now been swept away. I'd rather just edit all of the game's FMV together than subject myself to this excercise in mediocrity.
I use KDE on gentoo and have a gig of RAM. With several windows open KDE uses 10% (100MB) of that RAM. I have another machine that uses KDE with 512 MB of RAM and KDE hogs a little less on that (60-80 MB). I hardly ever hit swap on either machine. Application data often sits in RAM after being closed, and will just get re-allocated down the line if RAM is needed.
Actually the worst memory hog (by far) is Java. Java seems to be the only thing that makes me hit swap on a regular basis.
I recently picked up Advanced PHP Programming by George Schlossnagle just the other day and (although I haven't had a chance to dig deeply into it) am very impressed with the content. Unlike so many PHP books, it shows some of the best methods for large scale PHP designs. It covers all of the new features of PHP 5 and explains other advanced topics , such as caching. I was already doing many of these things, but just seeing a different perspective opened a world of possibilities to me (and pointed out where my methodology could use improvement). I'm sure there will be much more to learn as I work through it.
Each point to the correspoding scripts in/etc/init.d/
This may seem assinine, but it's reality for portage based OS's. You can't manually modify this stuff when you continue to emerge changes! Your modifications will get overwritten. That's why Gentoo had rc-update and etc-update.
It's similar to fBSD.. although I never figured out why everyone bashes Gentoo and praises fBSD (I think they're both great).
My first Gentoo install ended in a similar way. You have to approach a portage (notice I didn't say source) based distro with the assumption that it will take some time to figure out the differences.
With Gentoo, it takes awhile to get comfortable with the changes to how init works.
Much of these speed increases are mostly a result of shrinking die sizes. Most archetectural changes revolved around the introduction of new instructions (SSE). A lot of work was also done to improve the effeciency of the PIII for the coppermine release (which saw a signficant speed increase). The PIV project, which worked in parellel and was doing a much more radical redesign, wasn't able to benefit from this work. The archetecture became different enough that new and much more thorough R&D would have to go into improving PIV effeciency. Unfortunately, the PIV design is one of brute force and these types of design improvements have limited returns for such designs.
Just focussing on the PIII: the first to be introduced was the Katmai, which had a.25 micron core. With efficiency improvents and a drop to.18 micron, the Coppermine was able to achieve excellent results with a max speed in the 1-1.13 ghz range (although 1.13 required nice cooling). Finally, the Tualatin didn't offer many changes other than moving the die size down to.13 and adding some improved heat dissapation technology. These babies got up to 1.4 ghz off the shelf and could clock up to 1.6 ghz in practice.
The reason for the 1.6ghz ceiling? No it wasn't the CPU! Memory bandwidth was the reason these things couldn't go past 1.6. A PIII running at 1.6 ghz can effectively compete with 2.4-3.0 ghz PIVs!! If you could couple it with some high speed RAM, these things could have easily soared past 2.0ghz while remaining on a.13 micron die. For these reasons, many view the PIII coppermine/tualatin as one of the best made/designed CPUs of all time. Shrink them down to.9 micron and they would beat the crap out of PIV!
Alas, the Pentium M is a PIII with MORE efficiency improvements. The capabilities of this design have to be WAY beyond the PIV. It's a discredit to Intel's leadership that they aren't marketing their best product!
The space shuttle is a piece of crap; it should be grounded because its too expensive.
The shuttle is, without a doubt, too expensive; but saying its a piece of crap seems off the deep end. It is still cutting edge technology as far as space flight is concerned. It's also the only craft ever created up to this point that can construct a station in orbit (Russian stations were not able to be built this way, thus had much more limitations).
The shuttle may be dramatic overkill for getting people and cargo to orbit, but don't insult the machine or engineers who built it because of that! It's quite amazing that they were able to build such a thing with technology developed in the 60s and 70s. We still can't do much better today, and our next generation vehicle will almost surely be more limited than the shuttle. The shuttle excels at what it was designed to do, the problem is that the US government wanted it to do too much.
Your argument seems to blast gcc more than gentoo. It is more likely that your benchmarking method is flawed. -02 optimizations will only show increases for anything that is computationaly intensive. It is better to use integer intensive programs (pretty much everything but games and scientific apps) with 02. Simple programs will not have much opportunity for optimization, so 02 will most likely not make a difference for them. 03 is best used for anything that is heavy on floating point calculations. Many of these opimizations will acutally slow down (or even break - cause crashes n some instances) integer performance, so you want to be careful with 03. It is best used with games and scientific apps. You also must consider any linked libraries as well. They would need to be compiled with identical flags. Many people compile something like kdepim with 02, while kdebase and kdelibs are compiled with 0. They don't get a performance increase and say 02 is useless.
It looks like your test program doesn't take this into consideration.
I think the most important aspect of Linux growth is that it does not seem to have ever gone down. Linux has gradually grown year after year. Now that it is much more mature, it should continue to do the same.
Even if it maintains it's present rate, it seems likely that Linux will achieve at least significant market share, if not dominance, in many of our lifetimes.
Also, the article does not mention that growth rates are almost always exponential. If Linux continues to grow market share like it is, it will begin to grow at a much more rapid rate. The reason for this is word of mouth. Linux share is now in the 3-4% range. That means 3 or 4 people out of 100 use it (on the desktop). That means chances are that the average person doesn't know anyone who actually USES Linux. But once you hit the 10% range, that's 1 in 10 people that use Linux. Suddenly many more people know someone that uses Linux, and many will probably be willing to give a try. Suddenly you're shooting up to 25% share and you're in the trend/fad range. At this point momentum will usually swing completely in your favour. Not are you the trend, you're comeptition is looking poor for losing so much share. Chances are that at this point, market share will flip-flop and the underdog will be the leader. The Desktop OS market isn't like most other markets where there is a lot room for competition. It's very much a single product dominated market. It will likely always be that way. If GNU/Linux succeeds, however, it iss likely to also share a fair amount of success w/ other open source OSes like BSD.
Similar, but an exception allows for a gracefull exit before the *crash* occurs. This is useful for apps because it will save data and exit normally (without memory leaks) where a true crash couldn't avoid this. But at the kernel level both will result in the same thing (at least I would think). I suppose a kernel exception might be able to go as far as shut down your computer rather than lock it..
Slashdot Mirror
I believe intelligent life exists beyond Earth.
The real debate is whether intelligent life exists ON Earth.
Seems like most people aren'y thinking this through all the way.
First, it's likely that Intel would create a chip specifically for Apple. Volumes would still be high enough to make this possible. Intel does have experience with RISC chips. It seems very unlikely that Intel would migrate the P4 to Mac as that seems impractical. Pentium-M may be more likely, but I would still expect to see a RISC based chip (which Intel likely already has multiple design possibilities for). Intel has tried several times to break away from x86, this may factor into those attempts.
Second.. heat issues?! I keep hearing that Intel chips run too hot for Macs. Since when has the G5 been any cooler?! In any case, Intel is likely to create a RISC based chip which means a new design and most likely less heat.
Talks between Intel and Apple go way back. This isn't a new discussion. For various reasons Apple has never decided to go with Intel. Motorola was hopeless, and even IBM is a bit behind the game in terms of performance. The cost/performance ratio is still much better on the PC (by what seems like a factor of 2). If Apple is trying to drive prices down (which it seems like they are), this is an important factor.
Their best options would be AMD or Intel and AMD (IIRC) doesn't have much experience with RISC.
You can get away without writing much documenation these days. Usually the larger distros have how-tos for all common software. I just recently set up Shorewall on Gentoo and used the Gentoo documentation to do it. I looked around on Shorewall's site, had a hard time figuring it out.. and then found a Gentoo how-to that had a step by step guide on how to do it on my distro of choice (which is easier than a generic how-to anyway).
Let the documenation go, and just post the source code on the site.
kernel.org isn't exactly a documenation cornucopiea after all.
To get to The Moon you need to reach escape velocity. To get to Mars you also must reach escape veolocity. The only thing that makes getting to Mars more difficult is the longer journey. There really ISN'T that much point in stopping on the Moon if you're going to Mars.
Most of the liftoff weight for a spacecraft is fuel and it's up and the air wether or not The Moon is any minable form of usable fuel on it. We only know of traces at this point. If that remains the case than there really isn't much point for using The Moon as a stepping stone to Mars (it's more like a deviation).
Comments like this crack me up. For some reason people think that compiling Gentoo only involves sitting in front of a black screen watching compiliation lines scroll by. Actually, the system is quite usable while compiling. Occasionally (when I get the itch to play unreal or doom) I even suspend compilation, go play a few games, and then restart compilation (ctrl-z is your friend). You can even ctrl-c out of compilation and restart w/ emerge --resume. It really is that easy. Waiting for things to compile does not have to be painful at all. Pretty much anything other than playing heavy games can be done while compiling, and you can always resume your emerge if you get the gaming itch. If you do the install w/ something like Knoppix, you can even do almost anything during the initial install.
OSX takse up it's fair share of RAM. More than XP or any other OS by my experience.
Don't forget XFCE. I'm not sure Gnome is much more popular on Gentoo than KDE.. I'd be willing to bet that it isn't.
We're planning on moving to 5 in the near future, and I have been doing my development with it for a couple of months (obviously just using the php 4 syntax). I have noticed very few problems with our existing code. A few programs spit out error messages where there were none before, but they were actual coding problems. So far I would say 5 is nice and I'm looking forward to the more strict OOP features. PHP isn't the most well designed langauge out there, but it's better than most and 5 is a step in the right direction I think.
I better start migrating our cluster to Windows!!! Our linux servers get cracked open all the time and and everytime I add a new server it takes like an hour to set up!! It's a nightmare!! I trust Windows because it is rock solid and known for it's excellent security!!
So let's see.. it will cost $50,000 in license fees to move to Windows.
Nevermind.
Not to beat a dead stick, but maybe if enough people say it, Netscape will realize that this browser is FUGLY!!!!
FUUUGGGLLY!!!
In fact it's so fugly, that I had an involuntary WTF reaction before the screenshot even fully loaded. I don't care if this was the fastest browser on the planet (unlikely), I would never use it based on ugliness alone.
Newsflash Netscape. It's the 2000s and busy browsers, and indeed busy anything are definitely NOT the way to try to look "fashionable." Even though I'm slightly embarassed at writing such a statement, it's definitely true.
I have used Windows, Solaris/CDE, Linux (many distros and UIs), and OSX. As a coder/sys admin I feel most comfortable with linux, Gentoo and KDE in particular. The reason for this is the ease of system administration and setup for the software I need while maintaining the ability to tweak it entirely to my likely. No Gentoo was not productive for the first few days. As an added bonus, there is a ton of free software available that just doesn't work as well on OSX or Windows, which can also raise my productivity (Umbrello, Kate, KDevelop, Gimp). It runs best (if it all) in Linux, and getting something equivelant (and native) on Windows or OSX is likely to cost a fair amount.
I recently got a new PowerBook through work, and while I love it, there are limitations when compared to linux. Previous users stated that setting up OSX as a server is a breeze. Sure, this is true if you're only using a basic server, but if you need an advanced configuration, OSX can be a true nightmare. The tools are available under the hood, but it is NOT designed to be tinkered with. My reasoning for choosing OSX for work was the gaurantee for driver compatibility and full support (it was either put linux on a Dell laptop or get the PowerBook). I figured I would take a slight productivity dip, and I did. Everyone has a different routine, and to maximize productivity an interface must be highly configurable. Of course, the average user is not savy enough to deal with such a high level of configuration. OSX is a compromise. It is the best all around, but Linux has greater potential when configured for specific tasks.
I guess Windows is great if you measure productivity in Frames Per Second.
To sum up my perspective:
Linux productivity sucks in most "out of the box" configurations, but has the highest potential.
Windows is just "OK." It gets you by and lets you play games. It is also good for office apps as OpenOffice.org really isn't that great IMO (sorry - I have to question it's design), which makes MS Office the best office software by default. KOffice and the Gnome based apps have a chance at changing this in the long run.
OSX is highly productive out of the box, but has little capacity for getting that maximum tweaking. It is also VERY different than Windows. Most Windows users will feel more productive in a generic linux install than OSX. It is also a big advantage to be able to run MS Office on OSX (at least until it crashes).
We can argue about what data means all day long, but really this just comes down to common sense. It's a reasonable assumption that pumping an atmostphere full of greenhouse gasses is going to cause global warming. Whether or not we currently are experiencing the results of this warming may be hard to say, but I would have to question the intelligence/motives of someone who said that doing this is not dangerous.
If you relate this to the concepts of warming another planet like Mars to make it more habitable, there is virtually complete agreement that pumping huge amounts of greenhouse gasses into the atmosphere would allow the planet to become significantly warmer.
Now.. how can anyone argue that doing this on Earth would not produce the same results? It's like telling someone that they don't have any proof that you would die if you jumped out of an airlock in space without a suit. No one has done that before (to my knowledge), but it's a safe assumption that it would make you very dead. In this instance, ignoring common sense would be beyond foolish.
It's even worse when you do it on a global level. Once we wreck our environment, then I guess we'll have some hard data on global warming. Hopefully we survive our stupidity, but it doesn't seem like we even deserve to survive as a species if we are so collectively foolish.
It is quite hypocritical of Sun to be saying this when so little of their software runs on anything but Solaris.
XFree86 rings a bell
Right after I switch my entire environment from linux to windows due to the amazing overwhelmingly dangerous linux kernel exploit from a few days ago and the huge number of linux security issues that were said to be listed this year... this happens. Now my Win2003 server uptimes will get all jacked up.. and they just booted for the first time 2 days ago. Now I'm going to have to switch to OS/1337 in order to preserve my sys admin ego flames.
Don't use windows. Don't use linux. They both suck. OS/1337 has not once had a single crack or exploit discovered. Plus.. it has never crashed. Not once. OS/1337 will rock your world.
Well now that the movie has nothing to do with the game, we can still argue that the title is correct because this movie is definitely doomed.. Any hope that it would even be worth renting has now been swept away. I'd rather just edit all of the game's FMV together than subject myself to this excercise in mediocrity.
I use KDE on gentoo and have a gig of RAM. With several windows open KDE uses 10% (100MB) of that RAM. I have another machine that uses KDE with 512 MB of RAM and KDE hogs a little less on that (60-80 MB). I hardly ever hit swap on either machine. Application data often sits in RAM after being closed, and will just get re-allocated down the line if RAM is needed.
Actually the worst memory hog (by far) is Java. Java seems to be the only thing that makes me hit swap on a regular basis.
I recently picked up Advanced PHP Programming by George Schlossnagle just the other day and (although I haven't had a chance to dig deeply into it) am very impressed with the content. Unlike so many PHP books, it shows some of the best methods for large scale PHP designs. It covers all of the new features of PHP 5 and explains other advanced topics , such as caching. I was already doing many of these things, but just seeing a different perspective opened a world of possibilities to me (and pointed out where my methodology could use improvement). I'm sure there will be much more to learn as I work through it.
Here's a link:
http://www.amazon.com/exec/obidos/tg/detail/-/067
As for the top books on Amazon.. I'm not sure many self-respecting developers would have many of them...
You may not have been adding the ethernet scripts correctly. Did you use rc-update for each of those cards? You should have something like:
/etc/runlevels/default/net.*
/etc/runlevels/default/net.eth0@ /etc/runlevels/default/net.eth1@ /etc/runlevels/default/net.eth2@ /etc/runlevels/default/net.eth3@ /etc/runlevels/default/net.eth4@
/etc/init.d/
ls
Each point to the correspoding scripts in
This may seem assinine, but it's reality for portage based OS's. You can't manually modify this stuff when you continue to emerge changes! Your modifications will get overwritten. That's why Gentoo had rc-update and etc-update.
It's similar to fBSD.. although I never figured out why everyone bashes Gentoo and praises fBSD (I think they're both great).
My first Gentoo install ended in a similar way. You have to approach a portage (notice I didn't say source) based distro with the assumption that it will take some time to figure out the differences.
With Gentoo, it takes awhile to get comfortable with the changes to how init works.
Much of these speed increases are mostly a result of shrinking die sizes. Most archetectural changes revolved around the introduction of new instructions (SSE). A lot of work was also done to improve the effeciency of the PIII for the coppermine release (which saw a signficant speed increase). The PIV project, which worked in parellel and was doing a much more radical redesign, wasn't able to benefit from this work. The archetecture became different enough that new and much more thorough R&D would have to go into improving PIV effeciency. Unfortunately, the PIV design is one of brute force and these types of design improvements have limited returns for such designs.
.25 micron core. With efficiency improvents and a drop to .18 micron, the Coppermine was able to achieve excellent results with a max speed in the 1-1.13 ghz range (although 1.13 required nice cooling). Finally, the Tualatin didn't offer many changes other than moving the die size down to .13 and adding some improved heat dissapation technology. These babies got up to 1.4 ghz off the shelf and could clock up to 1.6 ghz in practice.
.13 micron die. For these reasons, many view the PIII coppermine/tualatin as one of the best made/designed CPUs of all time. Shrink them down to .9 micron and they would beat the crap out of PIV!
Just focussing on the PIII: the first to be introduced was the Katmai, which had a
The reason for the 1.6ghz ceiling? No it wasn't the CPU! Memory bandwidth was the reason these things couldn't go past 1.6. A PIII running at 1.6 ghz can effectively compete with 2.4-3.0 ghz PIVs!! If you could couple it with some high speed RAM, these things could have easily soared past 2.0ghz while remaining on a
Alas, the Pentium M is a PIII with MORE efficiency improvements. The capabilities of this design have to be WAY beyond the PIV. It's a discredit to Intel's leadership that they aren't marketing their best product!
The space shuttle is a piece of crap; it should be grounded because its too expensive.
The shuttle is, without a doubt, too expensive; but saying its a piece of crap seems off the deep end. It is still cutting edge technology as far as space flight is concerned. It's also the only craft ever created up to this point that can construct a station in orbit (Russian stations were not able to be built this way, thus had much more limitations).
The shuttle may be dramatic overkill for getting people and cargo to orbit, but don't insult the machine or engineers who built it because of that! It's quite amazing that they were able to build such a thing with technology developed in the 60s and 70s. We still can't do much better today, and our next generation vehicle will almost surely be more limited than the shuttle. The shuttle excels at what it was designed to do, the problem is that the US government wanted it to do too much.
Your argument seems to blast gcc more than gentoo. It is more likely that your benchmarking method is flawed. -02 optimizations will only show increases for anything that is computationaly intensive. It is better to use integer intensive programs (pretty much everything but games and scientific apps) with 02. Simple programs will not have much opportunity for optimization, so 02 will most likely not make a difference for them. 03 is best used for anything that is heavy on floating point calculations. Many of these opimizations will acutally slow down (or even break - cause crashes n some instances) integer performance, so you want to be careful with 03. It is best used with games and scientific apps. You also must consider any linked libraries as well. They would need to be compiled with identical flags. Many people compile something like kdepim with 02, while kdebase and kdelibs are compiled with 0. They don't get a performance increase and say 02 is useless.
It looks like your test program doesn't take this into consideration.
I think the most important aspect of Linux growth is that it does not seem to have ever gone down. Linux has gradually grown year after year. Now that it is much more mature, it should continue to do the same. Even if it maintains it's present rate, it seems likely that Linux will achieve at least significant market share, if not dominance, in many of our lifetimes.
Also, the article does not mention that growth rates are almost always exponential. If Linux continues to grow market share like it is, it will begin to grow at a much more rapid rate. The reason for this is word of mouth. Linux share is now in the 3-4% range. That means 3 or 4 people out of 100 use it (on the desktop). That means chances are that the average person doesn't know anyone who actually USES Linux. But once you hit the 10% range, that's 1 in 10 people that use Linux. Suddenly many more people know someone that uses Linux, and many will probably be willing to give a try. Suddenly you're shooting up to 25% share and you're in the trend/fad range. At this point momentum will usually swing completely in your favour. Not are you the trend, you're comeptition is looking poor for losing so much share. Chances are that at this point, market share will flip-flop and the underdog will be the leader. The Desktop OS market isn't like most other markets where there is a lot room for competition. It's very much a single product dominated market. It will likely always be that way. If GNU/Linux succeeds, however, it iss likely to also share a fair amount of success w/ other open source OSes like BSD.
Similar, but an exception allows for a gracefull exit before the *crash* occurs. This is useful for apps because it will save data and exit normally (without memory leaks) where a true crash couldn't avoid this. But at the kernel level both will result in the same thing (at least I would think). I suppose a kernel exception might be able to go as far as shut down your computer rather than lock it..
True. If an exception is thrown, it's mostly likely a program-halt condition. Which you obviously don't want in a kernel.