A core Debian, consisting of just the essential packages, would be fantastic. Unfortunately I don't think it is going to happen. I think the internal politics have taken over Debian's agenda. It would seem that Debian has become a distrobution designed by committee. Forever locked in discussions. A stable Sarge was promised for Fall of 2003. I will be surprised if there is a stable release before then end of this year.
I hope I am wrong. Debian is a fantastic distro to work with. Debian has been my distro of choice for close to seven years.
*nux is my desktop of choice. The larger number of GUI configuration options available make it easier to overcome the UI annoyances. However *nix is a power user's OS. While Windows XP and MacOS are quite similar, *nix is in a category of its own with different strengths.
Notes to avoid the flames:
Not all power users use *nix. Some use Windows XP. Others use MacOS.
Windows XP and MacOS look completely different. Much of the functionality is also different. Both share the same goal. Unlike Gnome or KDE, Windows XP and MacOS also have an advantage in that both control the whole system.
As someone who uses a Mac I have to say that MacOS is not that special.
MacOS is a good operating system. The gui looks good, and is fast enough on good hardware. The BSD bit is great. Especially the compiler and the access to all that *nix open source software.
The problem is that MacOS has annoyances just like Windows has annoyances. The annoyances are different, but I don't think MacOS has any fewer then Windows. Mac users like to say "It just works." By the same criteria I think you would find that Windows XP "just works."
I find that my Mac screen gets cluttered quickly. Expose is a required feature for MacOS, otherwise it is too easy to lose track of application windows. Windows achieves similar results with the task bar showing every open window, but it breaks if too many windows are open. Alt-tab works better under Windows then the MacOS equivalent. I find that MacOS relies more heavily on the mouse then Windows. I have yet to find a way to maximize a window with the keyboard. The MacOS Maximize button dosen't maximize to full screen like I would expect it too. Windows applications have an annoying habit of grabbing keyboard focus when you least expect it. MacOS applications don't grab keyboard focus when you would expect it. The top application window in MacOS may not be the active application. I find that Windows is more customizable then MacOS. For example you can properly set the dot pitch of your screen under Windows. MacOS is locked to 75dpi (patheticly coarse). My monitor does 100dpi easily.
These are all petty little problems: or annoyances. Both Windows and MacOS have their fair share. Don't believe people who say MacOS is perfect. Don't believe the hype.
I think both work equally well. The only reason to choose one over the other is if one has specific features that you want. Features such as software or hardware.
One of the possible benefits of digital delivery (cable, Internet, sattelite, etc) is that the cost of running a "station" could be lowered to the point where it is feasible to only run one or two hours of programming a day. When this happens we can expect to see a huge explosion of specialty stations. A good example is local news in foreign languages for recent imigrants. We should also be able to see more high quality, niche television. The producers of shows like Firefly would love to be able to sell their shows directly to the audience instead of trying to sell the show to tv station executives. Too many good shows have been cancelled not because the show was losing money, but becuase the time slot could be used to make more money.
You forgot the part where this was not previously possible due to the way Linux handles consoles and keyboards. You need a kernel patch to make it all work.
The negative comments really bug me. There is a difference between being critical and dumping on someone's work.
ontopic: I had wanted to do this last year to share a home computer. At the time the needed patches were not stable enough. It is good to see that it works now.
It's the messages weaved into the story lines of games, movies and tv shows, the preaching under the guise of entertainment, that gets my hackles up.
Why? Are you worried that it will make people think and change their opinions? Some philosophers would state that the whole purpose of entertainment is to make people think.
Ever since the first story telling all froms entertainment has been coloured by the creator's (or performer's) opinions and politics. An obvious example is the best selling book of all time: The Bible. Likewise the Written Torah, and the Holy Qur'an. The Roman games were designed to placate the masses so that they would forget about the struggles of daily life. In Hamlet a play is performed to draw out Hamlet's Uncle the userper. Entertainment is a form of art. All art is basically a representation of the artist's world view. It is only recently that media executives have tried in vain to remove all messages from entertainment to "apeal to a broader market."
Fluxbox should be fine. When you start Kopete and the Slit you will probably have to specify the second monitor with "-display:0.1". I believe multimonitor works with different resolutions and colour depths.
Not all systems use Delta. Some use Wye. Some Delta systems even include a neutral/ground to resolve balance issues. Delta with neutral seems popular in rural areas.
You can classify this as "when theory meets reality."
Some of the old ISA VGA cards would not work with another VGA card. A modern PCI card should not be a problem though. YMMV.
For the setup you are looking for I would skip xinerama and go straight for the old multimonitor support. The downside to multimonitor is that you can't move apps from one screen to the other. On the plusside both displays are treated seperately so fullscreen games and such should work as you want.
I found that KDE did not have good multimonitor support. Konq in particular did not like starting on the second display. KDE Xinerama support wasn't stellar either. Dialog boxes would apear between monitors, and fullscreen mode often meant fullscreen across both displays. Hopefully some of those issues have been resolved.
One of my pet peeves with OSX is the hard coded 72dpi display output. Both X and Windows are more flexible in this regard. I was shocked. For an OS that is supposed to have excellent font and graphic capability I had assumed that changeing the display's dpi would be basic functionality.
The US Air Force had been using modems since the late 1950s. However, anyone who had access to an Air Force modems probably also had access to the phone numbers. The first commercial modem was the Bell 103. It was available in 1962. I am guessing that hackers and phreakers have been wardailing ever since.
When we can find a 1.6GHz Duron for just over half the price of a 2.6GHz Celeron and get better performance consistently in almost every test we ran, the choice is clear.
Genetic Algorithms, or similar techniques, are probably used by some teams. Most teams have a huge database of the various setups used in the past. That data could be used to feed a complex model of a car's performance for a given driver. The models must be tuned to a given driver, as drivers have strong preferences for how their cars are tuned.
Believe it or not but some of the best antenna and radio designs of the past 50 years have been by amature radio operators. They are also responsible for an inovative rocket payload system so that they could piggy back their sattelites behind larger commercial loads. This should be a surprise to no one. These amatures have been driven by the same motivations that has made open source software some of the best in the world: they love what they do. Amatures need dedicated radio bands so they can work in and test their desgins without interference from or to others. These dedicated bands need to all over the spectrum as each band has its own advantages and challenges. Of course not every amature radio operator is a designer. But like open source software developers, RF designers need users to provide feedback. Amature operators provide imporant feedback such as signal strength and quality, and distance between stations.
It is also nice to have public radio bands that are not controled by commercial interests, in much the same way that it is nice to have public parks. Free of commercial interests amature radio bands are free (as in beer) to use. More importantly amature radio bands can be used in new and innovative ways that commercial interests are not interested in supporting.
Unfortunately these public radio bands are not as easily accessible as public parks. Licences and tests aren't the problem. RF is fragile and proveing that you will use it responsibly is important. A big problem is, ironically, the ARRL.
The ARRL has fought so hard to protect their radio bands that they risk loseing everything. The ARRL has lost its relavence to the general public. Amature operators around the world have been extremely reluctant to change. The old amatures have always welcomed new amatures, but they haven't gone out of there way to find new amatures. There has always been a huge source of new amatures in the hacker community. The ARRL needs to do more encourage these hackers to become radio amatures.
The current structure and activities of the ARRL does not encourage new participation. Young radio hackers are not interested in DX competitions and making 10 second contacts to fill out a QSL card. Young hackers are not interested in making contact with some grumpy old guy half way around the world just to hear what ailments he has. (This is a far too common occurance.)
Young hackers are interested in making world wide, community based, digital networks. They are interesetd in freedom of speech and privacy issues. They want to use encryption. Many of the old amatures are affraid that the young hackers want to move in and change everything. This is only partly true. The hackers do want to change a few things, but they are also more then willing to work with the community. Look at groups such as Seattle Wireless. These guys are essentially rogue freebanders. The ARRL needs to modernize themselves and the FCC to turn these freebanders into licenced amatures.
If the ARRL and similar groups don't do more to encourage new participation there won't be amature radio in few decades time, because there won't be many amature radio operators left alive. The recent easing of licence and band restrictions will help, but much more needs to be done.
Provideing a highspeed network to a rural area is a hard problem. There are few customers to offset the operating expenses.
A local start-up was working on a highspeed network for rural areas. It used HF in a licenced band so interferance would not have been an issue. Because the system used HF one tower could cover quite a large area. The speeds were not lightning fast but were faster then modems. I believe the project goal was just a little faster then sattelite.
Unfortunately the project was killed for two reasons. The first was patents. There are some (arguably obvious) patents that cover highspeed networks over HF. The patents owners were not interested in developing the technology themselves, rather they wanted to charge exhorbitant fees to licence the patents. Given enough money this issue could have been resolved, but when coupled with the second problem project was canceled. The second problem was lack of a market.
From the start the system was designed to serve sparsely populated rural areas. This system could not compete with DSL, cable or 802.11 based systems. The bandwidth was slower, and more the system was more expensive. The setup costs were high as a client station needed a good HF transciever and antenna. The service fees were high as the base stations were designed to only handle a few customers. The system had to be heavily optimized for rural areas in order to achieve the large distances required. The optimizations were such that it could not even be scaled back to compete in the quasi-rural suburban environments. The system was expensive. While an end customer might be willing to pay $1000 to setup a station, plus $100/month for highspeed no provider was willing to take the risk when a base tower could easily cost $100k just to install.
I suspect that highspeed of power lines is going to face similar challenges and suffer the same fate. The setup costs are deffinately lower, but the system is still faced with some of the same technical problems. Long distances cause more noise, which lowers bandwidth, which reduce the number of customers on a given segment. With fewer customers there is less chance of a profit.
A pin insert would connect to pads on the mainboard and pads on the cpu. I worked with some SGI systems that used this kind of setup. It is a good idea.
I am a little disapointed that Intel did not go with a pin insert. However it would have cost more which would have been hard to justify to the mainboard makers and their razor thin margins. In the long run I think that a pin insert would have been a smart move. Judgeing by the reported fragility of the socket 775 I won't be surprised if Intel moves to a new socket as early as next year. I am sure we will see an outcry from the (wanabee) enthusiast crowd due to the shear number of socket 775 mainboards that they will trash. A pin insert may have eased that problem.
Pin inserts could even have been a lucrative revenue stream. It would not take much to convince the enthusiasts that a new pin insert was needed whenever a cpu was upgraded to insure maximum performance. In some respects computer enthusiasts are as bad as audiophiles.
Every feature you list depends on the MAC. It is trivial to spoof a MAC.
A managed VPN would achieve the same results as Campus Manager with the addition of strong authentication and security. A VPN sounds big and scary, but a modern one isn't. Many VPN appliances even have point and drool interfaces.
The popularity of Linux is in large part because of the AT&T suit which stopped 386BSD dead in its tracks. When Linus released his first kernel there was a huge demand for an affordable *nix that ran on the 386 and there was nothing to feed that demand. There was great uncertainty over the outcome of BSD. This delay of BSD for the 386 allowed Linux to gain a strong following of developers. When BSD was eventually freed of AT&T code many of the developers that would have worked on BSD where now working on Linux, and saw no reason to switch.
While the AT&T suit against BSD allowed Linux to attract those initial developers it is Linus who kept them. Linus is a nice, easy going person. Many kernel developers have stated that it is becuase Linus is easy to work with that they have continued to work on Linux. Even when telling someone off, Linus does it politely. Linux has been able to attract new developers because Linus has always allowed anyone to submit a patch.
A third factor has been huge part of the sucess of Linux. It has nothing to do with technical merit or leadership. Linux was able to ride the DotCom hype in a way that BSD couldn't. BSD was viewed as being old. Which it was. Linux was the "hot new thing." Being new did not make Linux better then BSD, but that didn't matter. In the hype and hysteria of the dot.bombs being new was more important than anything. This hype really popularized Linux attracting new users and more importantly new developers. Being popular is not always viewed as a benefit but for open source projects popularity is important. Open source projects need to attract and keep developers.
Linux did benefit from the DotCom boom, but Linux did not necessarily require the hype. Prior to the boom there were already some commercial interest in Linux such as Red Hat. In acedemic circles Linux had gained some interest. Projects such as Beowulf clustering would probably have happened regardless of the DotCom hype. It is hard to say if IBM would have taken an interst in Linux had it not been for the the DotCom boom. Likewise it is hard to say if a site like/. would exist.
Some would argue that the Gnu Licence has been important to the sucess of Linux. While the GPL may have attracted some developers it problaby has repelled just as many. Users largely don't care as both licences are free from an average users point of view. The BSD licence is deffinately more friendly to commercial interests. A number of companies have been playing fast and loose with the GPL, such as some of the embedded routers. A few of the commercial Linux distrobutions have been less then open with their source. While in some cases this has been acceptable, it does show that those companies would probably be more comfortable with BSD. These commerical interests use Linux due to its popularity. There is a market for "Linux." The embedded developers have used Linux more often then not out of convinience. Linux for embedded systems is easier to find then BSD.
The popularity of Linux continues to attract new users, and new developers. Not only kernel developers, but also user space developers such as the KDE and Gnome teams. This popularity is self fueling and keeps Linux and the user space advanceing technically. Linus keeps the kernel developers happy and directs the continued technical improvements in the kernel. The importance of Linus himself can't be overstated. If he had not encouraged development the way he did, Linux would not have made it to version 1. But none of this would have happened if BSD had not been caught up in the AT&T lawsuit.
I use the multiple columns mode as well. I did find it very hard to use at first. Multiple column mode is quite different. I am glad I stuck with it though.
I think this mode could also use some tweaking, as on occation I find it acts unexpectedly with respect to the rest of MacOS.
Slashdot Mirror
A core Debian, consisting of just the essential packages, would be fantastic. Unfortunately I don't think it is going to happen. I think the internal politics have taken over Debian's agenda. It would seem that Debian has become a distrobution designed by committee. Forever locked in discussions. A stable Sarge was promised for Fall of 2003. I will be surprised if there is a stable release before then end of this year.
I hope I am wrong. Debian is a fantastic distro to work with. Debian has been my distro of choice for close to seven years.
*nux is my desktop of choice. The larger number of GUI configuration options available make it easier to overcome the UI annoyances. However *nix is a power user's OS. While Windows XP and MacOS are quite similar, *nix is in a category of its own with different strengths.
Notes to avoid the flames:
Not all power users use *nix. Some use Windows XP. Others use MacOS.
Windows XP and MacOS look completely different. Much of the functionality is also different. Both share the same goal. Unlike Gnome or KDE, Windows XP and MacOS also have an advantage in that both control the whole system.
And what kinds of changes do you want in cmd-tab?
Switching between windows not apps. Here is a solution.
Any specific examples of the not grabbing keyboard focus when you expect it?
Right click on the Camino. Choose New. The Camino app, and new window won't be in focus. It is an odd bug.
I've seen a lot of Macs running at more than 75dpi.
It is possible to run MacOS at a resolution that results in a display better then 75dpi, but the gui will not compensate. The fonts will be too small.
thanks.
As someone who uses a Mac I have to say that MacOS is not that special.
MacOS is a good operating system. The gui looks good, and is fast enough on good hardware. The BSD bit is great. Especially the compiler and the access to all that *nix open source software.
The problem is that MacOS has annoyances just like Windows has annoyances. The annoyances are different, but I don't think MacOS has any fewer then Windows. Mac users like to say "It just works." By the same criteria I think you would find that Windows XP "just works."
I find that my Mac screen gets cluttered quickly. Expose is a required feature for MacOS, otherwise it is too easy to lose track of application windows. Windows achieves similar results with the task bar showing every open window, but it breaks if too many windows are open. Alt-tab works better under Windows then the MacOS equivalent. I find that MacOS relies more heavily on the mouse then Windows. I have yet to find a way to maximize a window with the keyboard. The MacOS Maximize button dosen't maximize to full screen like I would expect it too. Windows applications have an annoying habit of grabbing keyboard focus when you least expect it. MacOS applications don't grab keyboard focus when you would expect it. The top application window in MacOS may not be the active application. I find that Windows is more customizable then MacOS. For example you can properly set the dot pitch of your screen under Windows. MacOS is locked to 75dpi (patheticly coarse). My monitor does 100dpi easily.
These are all petty little problems: or annoyances. Both Windows and MacOS have their fair share. Don't believe people who say MacOS is perfect. Don't believe the hype.
I think both work equally well. The only reason to choose one over the other is if one has specific features that you want. Features such as software or hardware.
then someone applied the patches for you.
one method
a second method
One of the possible benefits of digital delivery (cable, Internet, sattelite, etc) is that the cost of running a "station" could be lowered to the point where it is feasible to only run one or two hours of programming a day. When this happens we can expect to see a huge explosion of specialty stations. A good example is local news in foreign languages for recent imigrants. We should also be able to see more high quality, niche television. The producers of shows like Firefly would love to be able to sell their shows directly to the audience instead of trying to sell the show to tv station executives. Too many good shows have been cancelled not because the show was losing money, but becuase the time slot could be used to make more money.
You forgot the part where this was not previously possible due to the way Linux handles consoles and keyboards. You need a kernel patch to make it all work.
Pretty sad isn't it.
Slashdot needs new moderation options such as:
-1 Didn't RTFA.
-1 Don't be so negative.
The negative comments really bug me. There is a difference between being critical and dumping on someone's work.
ontopic: I had wanted to do this last year to share a home computer. At the time the needed patches were not stable enough. It is good to see that it works now.
It's the messages weaved into the story lines of games, movies and tv shows, the preaching under the guise of entertainment, that gets my hackles up.
Why? Are you worried that it will make people think and change their opinions? Some philosophers would state that the whole purpose of entertainment is to make people think.
Ever since the first story telling all froms entertainment has been coloured by the creator's (or performer's) opinions and politics. An obvious example is the best selling book of all time: The Bible. Likewise the Written Torah, and the Holy Qur'an. The Roman games were designed to placate the masses so that they would forget about the struggles of daily life. In Hamlet a play is performed to draw out Hamlet's Uncle the userper. Entertainment is a form of art. All art is basically a representation of the artist's world view. It is only recently that media executives have tried in vain to remove all messages from entertainment to "apeal to a broader market."
Fluxbox should be fine. When you start Kopete and the Slit you will probably have to specify the second monitor with "-display :0.1". I believe multimonitor works with different resolutions and colour depths.
Not all systems use Delta. Some use Wye. Some Delta systems even include a neutral/ground to resolve balance issues. Delta with neutral seems popular in rural areas.
You can classify this as "when theory meets reality."
Some of the old ISA VGA cards would not work with another VGA card. A modern PCI card should not be a problem though. YMMV.
For the setup you are looking for I would skip xinerama and go straight for the old multimonitor support. The downside to multimonitor is that you can't move apps from one screen to the other. On the plusside both displays are treated seperately so fullscreen games and such should work as you want.
I found that KDE did not have good multimonitor support. Konq in particular did not like starting on the second display. KDE Xinerama support wasn't stellar either. Dialog boxes would apear between monitors, and fullscreen mode often meant fullscreen across both displays. Hopefully some of those issues have been resolved.
One of my pet peeves with OSX is the hard coded 72dpi display output. Both X and Windows are more flexible in this regard. I was shocked. For an OS that is supposed to have excellent font and graphic capability I had assumed that changeing the display's dpi would be basic functionality.
This case was not a waste of time. This is how Bristish style legal systems are supposed to work.
The US Air Force had been using modems since the late 1950s. However, anyone who had access to an Air Force modems probably also had access to the phone numbers. The first commercial modem was the Bell 103. It was available in 1962. I am guessing that hackers and phreakers have been wardailing ever since.
An introduction to modem history
From a December 2003 article:
Genetic Algorithms, or similar techniques, are probably used by some teams. Most teams have a huge database of the various setups used in the past. That data could be used to feed a complex model of a car's performance for a given driver. The models must be tuned to a given driver, as drivers have strong preferences for how their cars are tuned.
Believe it or not but some of the best antenna and radio designs of the past 50 years have been by amature radio operators. They are also responsible for an inovative rocket payload system so that they could piggy back their sattelites behind larger commercial loads. This should be a surprise to no one. These amatures have been driven by the same motivations that has made open source software some of the best in the world: they love what they do. Amatures need dedicated radio bands so they can work in and test their desgins without interference from or to others. These dedicated bands need to all over the spectrum as each band has its own advantages and challenges. Of course not every amature radio operator is a designer. But like open source software developers, RF designers need users to provide feedback. Amature operators provide imporant feedback such as signal strength and quality, and distance between stations.
It is also nice to have public radio bands that are not controled by commercial interests, in much the same way that it is nice to have public parks. Free of commercial interests amature radio bands are free (as in beer) to use. More importantly amature radio bands can be used in new and innovative ways that commercial interests are not interested in supporting.
Unfortunately these public radio bands are not as easily accessible as public parks. Licences and tests aren't the problem. RF is fragile and proveing that you will use it responsibly is important. A big problem is, ironically, the ARRL.
The ARRL has fought so hard to protect their radio bands that they risk loseing everything. The ARRL has lost its relavence to the general public. Amature operators around the world have been extremely reluctant to change. The old amatures have always welcomed new amatures, but they haven't gone out of there way to find new amatures. There has always been a huge source of new amatures in the hacker community. The ARRL needs to do more encourage these hackers to become radio amatures.
The current structure and activities of the ARRL does not encourage new participation. Young radio hackers are not interested in DX competitions and making 10 second contacts to fill out a QSL card. Young hackers are not interested in making contact with some grumpy old guy half way around the world just to hear what ailments he has. (This is a far too common occurance.)
Young hackers are interested in making world wide, community based, digital networks. They are interesetd in freedom of speech and privacy issues. They want to use encryption. Many of the old amatures are affraid that the young hackers want to move in and change everything. This is only partly true. The hackers do want to change a few things, but they are also more then willing to work with the community. Look at groups such as Seattle Wireless. These guys are essentially rogue freebanders. The ARRL needs to modernize themselves and the FCC to turn these freebanders into licenced amatures.
If the ARRL and similar groups don't do more to encourage new participation there won't be amature radio in few decades time, because there won't be many amature radio operators left alive. The recent easing of licence and band restrictions will help, but much more needs to be done.
Provideing a highspeed network to a rural area is a hard problem. There are few customers to offset the operating expenses.
A local start-up was working on a highspeed network for rural areas. It used HF in a licenced band so interferance would not have been an issue. Because the system used HF one tower could cover quite a large area. The speeds were not lightning fast but were faster then modems. I believe the project goal was just a little faster then sattelite.
Unfortunately the project was killed for two reasons. The first was patents. There are some (arguably obvious) patents that cover highspeed networks over HF. The patents owners were not interested in developing the technology themselves, rather they wanted to charge exhorbitant fees to licence the patents. Given enough money this issue could have been resolved, but when coupled with the second problem project was canceled. The second problem was lack of a market.
From the start the system was designed to serve sparsely populated rural areas. This system could not compete with DSL, cable or 802.11 based systems. The bandwidth was slower, and more the system was more expensive. The setup costs were high as a client station needed a good HF transciever and antenna. The service fees were high as the base stations were designed to only handle a few customers. The system had to be heavily optimized for rural areas in order to achieve the large distances required. The optimizations were such that it could not even be scaled back to compete in the quasi-rural suburban environments. The system was expensive. While an end customer might be willing to pay $1000 to setup a station, plus $100/month for highspeed no provider was willing to take the risk when a base tower could easily cost $100k just to install.
I suspect that highspeed of power lines is going to face similar challenges and suffer the same fate. The setup costs are deffinately lower, but the system is still faced with some of the same technical problems. Long distances cause more noise, which lowers bandwidth, which reduce the number of customers on a given segment. With fewer customers there is less chance of a profit.
Fog, rain, snow, trees, and hills all "interfere" with lasers. Laser comms are great for parts of Arizona though. :-)
A pin insert would connect to pads on the mainboard and pads on the cpu. I worked with some SGI systems that used this kind of setup. It is a good idea.
I am a little disapointed that Intel did not go with a pin insert. However it would have cost more which would have been hard to justify to the mainboard makers and their razor thin margins. In the long run I think that a pin insert would have been a smart move. Judgeing by the reported fragility of the socket 775 I won't be surprised if Intel moves to a new socket as early as next year. I am sure we will see an outcry from the (wanabee) enthusiast crowd due to the shear number of socket 775 mainboards that they will trash. A pin insert may have eased that problem.
Pin inserts could even have been a lucrative revenue stream. It would not take much to convince the enthusiasts that a new pin insert was needed whenever a cpu was upgraded to insure maximum performance. In some respects computer enthusiasts are as bad as audiophiles.
Every feature you list depends on the MAC. It is trivial to spoof a MAC.
A managed VPN would achieve the same results as Campus Manager with the addition of strong authentication and security. A VPN sounds big and scary, but a modern one isn't. Many VPN appliances even have point and drool interfaces.
The popularity of Linux is in large part because of the AT&T suit which stopped 386BSD dead in its tracks. When Linus released his first kernel there was a huge demand for an affordable *nix that ran on the 386 and there was nothing to feed that demand. There was great uncertainty over the outcome of BSD. This delay of BSD for the 386 allowed Linux to gain a strong following of developers. When BSD was eventually freed of AT&T code many of the developers that would have worked on BSD where now working on Linux, and saw no reason to switch.
/. would exist.
While the AT&T suit against BSD allowed Linux to attract those initial developers it is Linus who kept them. Linus is a nice, easy going person. Many kernel developers have stated that it is becuase Linus is easy to work with that they have continued to work on Linux. Even when telling someone off, Linus does it politely. Linux has been able to attract new developers because Linus has always allowed anyone to submit a patch.
A third factor has been huge part of the sucess of Linux. It has nothing to do with technical merit or leadership. Linux was able to ride the DotCom hype in a way that BSD couldn't. BSD was viewed as being old. Which it was. Linux was the "hot new thing." Being new did not make Linux better then BSD, but that didn't matter. In the hype and hysteria of the dot.bombs being new was more important than anything. This hype really popularized Linux attracting new users and more importantly new developers. Being popular is not always viewed as a benefit but for open source projects popularity is important. Open source projects need to attract and keep developers.
Linux did benefit from the DotCom boom, but Linux did not necessarily require the hype. Prior to the boom there were already some commercial interest in Linux such as Red Hat. In acedemic circles Linux had gained some interest. Projects such as Beowulf clustering would probably have happened regardless of the DotCom hype. It is hard to say if IBM would have taken an interst in Linux had it not been for the the DotCom boom. Likewise it is hard to say if a site like
Some would argue that the Gnu Licence has been important to the sucess of Linux. While the GPL may have attracted some developers it problaby has repelled just as many. Users largely don't care as both licences are free from an average users point of view. The BSD licence is deffinately more friendly to commercial interests. A number of companies have been playing fast and loose with the GPL, such as some of the embedded routers. A few of the commercial Linux distrobutions have been less then open with their source. While in some cases this has been acceptable, it does show that those companies would probably be more comfortable with BSD. These commerical interests use Linux due to its popularity. There is a market for "Linux." The embedded developers have used Linux more often then not out of convinience. Linux for embedded systems is easier to find then BSD.
The popularity of Linux continues to attract new users, and new developers. Not only kernel developers, but also user space developers such as the KDE and Gnome teams. This popularity is self fueling and keeps Linux and the user space advanceing technically. Linus keeps the kernel developers happy and directs the continued technical improvements in the kernel. The importance of Linus himself can't be overstated. If he had not encouraged development the way he did, Linux would not have made it to version 1. But none of this would have happened if BSD had not been caught up in the AT&T lawsuit.
I use the multiple columns mode as well. I did find it very hard to use at first. Multiple column mode is quite different. I am glad I stuck with it though.
I think this mode could also use some tweaking, as on occation I find it acts unexpectedly with respect to the rest of MacOS.