Among developed nations, the US is pretty mediocre on most quality of life indicators. The high point of the US is its high per capita GDP (which is more of a statistical oddity than a meaningful quality of life indicator), but it also has a number of pretty black marks (income disparity, crime, capital punishment, health care, infant mortality, etc.). On human happiness indicators, the US tends to score even less well in international comparisons because material wealth and happiness are only weakly correlated.
There's a line between good and evil. We're on one side of it. Nations like North Korea and Iran are on the other side of it.
That statement is pretty ironic given that it was the US that toppled Iran's democratically elected government and replaced it with a brutal dictatorship.
And this notion of "good" and "evil" nations is quite interesting anyway. Let's see: is it the people of a nation that are evil, or just its government? And when did the US stop being evil and become a good nation?
Nope, we're not racist.
Slavery existed here until the middle of the 19th century, legal inequality until the middle of the 20th century, and statistics as well as personal experience show that prejudice and discrimination are still rampant in many parts of the country and many populations.
But feel free to live in your fantasy world.
No, you seem to live in a fantasy world, but without knowing more about you, it's hard to diagnose why you know so little about what's going on in the world or in the US.
hmm, how come the GPL is never good enough whenever someone speaks about Qt or Trolltech?
That's because Qt is pretty much the only GUI toolkit for Linux that is GPL'ed. Just about everybody else realizes that a GPL'ed toolkit is not good for Linux or free software.
The LGPL only seems reasonable, if you like proprietary software....
You say that as if the LGPL were some nefarious scheme hatched by opponents of free software. But, in fact, the LGPL (and the GPL-with-exemption, which is now preferred) was created by the FSF in order to advance free software. It was created because sometimes it is necessary to accomodate proprietary software in order to advance the cause of free software. Commodity or system libraries, like GUI toolkits, are exactly the kind of situation in which the LGPL makes a lot more sense than the GPL.
but the plan is to replace the current libraries with LGPL replacements.
That would be great, but last time people tried that (Harmony), Troll Tech became pretty belligerent. Also, it doesn't address the other major problem with Qt/Embedded: everybody has to rewrite their apps.
I think it would be easier just to ship Opie on top of X11. That way, it doesn't matter that Opie is GPL'ed because people can write commercial apps in other X11 toolkits.
That's only the beta test; because the system works so well, it will be extended to everybody making less than $1,000,000/year (in 2000 dollars) within a year after initial deployment.
As far as I can tell, Opie is a dead end: being an open source fork of Qt, the only license under which it is available is the GPL. That makes Opie an impossible platform for many commercial applications.
Areas where a Linux PDA could shine are gaming, inventory support, hospitals, instrument control, etc. But the developers of that kind of software simply are not going to make it open source, nor are their customers going to care whether it's open source or not. So, a GPL-only toolkit has no chance. But even the commercial version of Qt isn't all that attractive because if those people are going to throw in their lot with a commercial vendor, they might as well go with Microsoft and PPC: it's cheaper than Qt and more widely used.
Now, Opie/X11 might have a chance because it reuses all that effort that went into Opie but also opens up the platform for commercial apps.
Ultimately, GPL'ed or dual-licensed GUI toolkits just won't work in the long run--there are too many free or LGPL'ed alternatives to choose from.
It looks atrocious (at this point) and doesn't have near the specs dell offers for the same price.
Actually, if it's similar to the old VR3, the hardware design is really nice: the rounded corners make it easy to carry around, the screen protector works great, and the device is quite small.
Palm or PPC hardware looks flashier and more high-tech, but hardware like the VR3 is more usable.
I think there is a more fundamental problem with this. A lot of this code, NUMA, XFS, etc., was dumped into the Linux kernel not because most Linux users need it, but because it served the interests of a few corporate contributors. Furthermore, the nearly monolithic structure of the Linux kernel means that the way this sort of stuff gets contributed is via patches and additions to the source tree.
So, quite apart from the fact that SCO's claims have no merit, I think this does point out a pretty fundamental problem with the Linux kernel: it accretes too much stuff. The kernel should be architected in such a way that IBM and SGI could distribute most of their code separately. Furthermore, the Linux mainstream user base should not have to suffer from the increased maintenance hassles (and, as it turns out, increased legal risk) of having millions of lines of code they don't want or need poured into the kernel.
Technically, none of that is a problem with modern programming language and software engineering technologies. Even though some kernel features may require a lot of cross-cutting concerns, we now know how to implement those efficiently and in a principled way. Of course, this won't happen with the Linux kernel, which seems to be solidly stuck in 1960's software engineering technologies.
Don't get me wrong: the Linux kernel is enormously useful and it has lots of drivers. In practice, it is still the most practical kernel around as far as I'm concerned. But these kinds of legal issues, the Bitkeeper mess, as well as the ever more sluggish pace at which new features get incorporated and new kernels get released, are symptoms of a deep problem, and that problem is not going to go away.
Do you really think I am being an "ideolog" when I believe that widespread torture is wrong? that imprisonment without trial of people exercising their right to freedom of expression is wrong?
You don't stop at believing it's wrong, you argue that people should act against the wrongdoing of foreign governments instead of participating in the democratic process at home. It is that part that I disagree with. We are responsible for the misdeeds of our own governments, and there are plenty of things our own government does wrong.
And you don't have an answer for what the Chinese government should do anyway. I mean a couple of centuries of democracy in the US haven't caused us to conform to international human rights standards--our government violates human rights, tortures and kills people, and engages in wars by popular choice. Many times in the 20th century, the US ousted democracies in order to institute US-friendly, human-rights violating dictatorships, so it doesn't seem like human rights and democracy are a big goal of US foreign policy either. So, tell us, except for pointing fingers, what is your plan for China?
People like you just like to use countries like China, Iraq, and North Korea as convenient excuses for dealing with the enormous social and economic problems here at home.
I have a Sharp Zaurus, and I think Qtopia just doesn't cut it.
Sure, it looks pretty nice and it has most of the functionality you might want in a PDA, but it is still significantly worse than either Palm or PPC. Some of the problems include badly thought out user interaction, wasteful use of the limited screen real estate (probably a result of being based on an adaptation of a desktop toolkit), and pretty excessive resource consumption by Qt/Embedded. And there is far less software available for Qtopia than for either Palm or PPC. If you want good PDA functionality, get a Palm.
On the other hand, as a Linux PDA for vertical apps, Qtopia-based PDAs also fall short: you are limited to the Qt toolkit and all the graphics and UI code from existing Linux apps require complete rewrites. You can't use any of the open source GUI tools you are likely used to (Tcl/Tk, Python/Gtk+, etc.). And if you want to write commercial apps, it's going to cost you (you can do commercial Palm development for free).
Linux PDAs will keep failing until their makers recognize that it is futile to compete with Palm and PPC head-on. Linux PDAs can thrive in the niche of providing portable little Linux machines, but that means not limiting the machines to running just a single GUI toolkit.
China is a country of a billion people, with many diverse ethnic groups. It's a country undergoing massive changes, and the whole coutry is walking a tightrope, always at risk of falling apart or rampant corruption.
I doubt that the current Chinese government is the best possible for China, and they are certainly far from a democracy yet, but anybody claiming to know that they know better how to run China is either an idiot or an ideolog. It took Europeans hundreds of years and many bloody wars to become modern democracies. Americans had to commit genocide and institute centuries of slavery before finally waking up to ideas of human rights and equality. Give the Chinese a break--they aren't doing all that badly in comparison, they are just a little late.
Oh, and as for Microsoft, the Chinese are just watching out for their economic interests: nurturing domestic high-tech expertise is a good thing for them, and replacing Microsoft software with domestically developed software just makes sense.
Now I only wish they would turn their minds on and realize that most of the world's ecological abuse goes on in third-world nations, as does most of the world's racism, sexism, and war.
US citizens are responsible for what happens in the US, not what happens in third world nations. Therefore, it makes perfect sense for US citizens to participate in the US political process, and part of that is rallies and demonstrations in the US against US government policies.
And, in any case, what kind of argumentation is that supposed to be anyway? You seem to be saying that US can be racist and destroy its environment because North Korea is even worse. If you want to engage in international comparisons at all, then the US should compare itself to the best nations in these areas, not to the worst, and in all the areas you mention, the US is in pretty bad shape in international comparisons.
You know, I just remembered why we don't launch nuclear waste out into space. It would -really- suck if it blew up before it cleared orbit, wouldn't it?
Nuclear waste is very different from nuclear fuel, and there are nuclear fuels that are not all that hazardous. Uranium, for example, is not very strongly radioactive. What is dangerous is all the short-lived by-products produced during fission reactions.
Just imagine if something went wrong like chernobyl. Except this time it's 30 miles in the air where it can travel around the globe quite nicely.
No, it's not like Chernobyl. Chernobyl was a running nuclear reactor, which had already created lots of fission products. It's also not like Plutonium-based nuclear batteries--Uranium is far less poisonous than Plutonium. So, in terms of risk, the objections may be less strong.
But people may object to any nation sending nuclear reactors into space. The US and other nations would love to use nuclear reactors to power space-based weapons, and this sort of civilian project might create the infrastructure for, and legitimize, such uses of nuclear power in space.
Whether it is ecology issues, racism, sexism or peace marches, most of the action is here in the USA where well-to-do clueless college kids are in abundance.
So, you much rather prefer turning the US into a fascist state, military dictatorship, or making it so poor that nobody can protest? Well, the way things are going, you may yet get your wish.
I am grateful that I live in a country in which college kids and environmentalists both have the freedom and sufficient economic resources to worry about ecology, racism, sexism, and peace.
It is about time mankind gets out of low-earth-orbit, either by giving up on space altogether
Giving up on manned space exploration is not the same as giving up on space. Manned space exploration is worthless from a scientific point of view, and it is a bottomless sink for money and a boon for governmental pork barrel spending. Unmanned space exploration, on the other hand, is highly valuable scientifically and much cheaper. And all money-making commercial use of space is unmanned as well.
and we probably would be if we'd kept up what we were doing in the 60's.
By "what we were doing in the 60's", you mean "driven by the cold war and the threat of nuclear annihilation, pursue senseless vanity space projects to demonstrate to the world one's technological superiority"?
Well, I am glad that we didn't keep this up. In fact, I don't give a damn whether people set foot on Mars during my lifetime, and I wouldn't want another dollar wasted on manned space exploration. I'd much rather see that money go to unmanned probes, which yield much more interesting scientific data.
There are three and four CCD digital cameras; this one isn't one of them. Instead, it uses four different color filters to get more information about the spectral composition of the incoming light. That lets it correct for color aberrations better than 3 color CCDs. This is an old trick, and you can actually get film that does the same thing.
Palm hardware is fairly nice: compact, light, long battery life, and the applications are pretty decent as well.
But PalmOS is becoming a drag on the company--PalmOS was OK for underpowerd 68k-based handhelds, but the 175-400MHz RISC handhelds of today would be much better off running a more modern OS; none of the restrictions, flakiness, and limitations of PalmOS are justified by the hardware it runs on anymore. Neither PalmOS 5 nor PalmOS 6 look like they are going to fix the grave deficiencies in PalmOS.
One can only hope that, with this split, the Palm hardware group can look to other options in terms of software: Symbian, Linux, etc. They can even keep running the existing Palm applications, which people are used to. But the OS has to go.
xfree86 is mostly derived from a free reference implementation. It may fail to offer hardware acceleration for many reasons. For example, the necessary hw specs may just not he published. Blit was hardware accelerated in commercial X11 servers pretty much as soon as the hardware was capable of supporting it.
Keep in mind that there are several different, independent X11 implementations from different vendors.
XRender is a new extension with only a reference implementation in XFree86. The point is to experiment with an API prior to freezing it. I know this may come as news to people who have grown up on Microsoft software, but real software developers first try out various ideas and then later start hacking it for speed. It would be quite surprising, actually, if it were faster than a hand-tuned client-side software implementation.
It will be a while until XRender beats client-side software implementations. Furthermore, you can't just take a client-side renderer and hack in XRender calls and expect it to run fast--code that works efficiently with a client-server window system like X11 needs to be written differently than something that moves around pixels locally.
Macintosh systems are pretty easy to administer for individuals. It's also fairly easy to set up ad-hoc workgroups with Macintosh.
But that does not translate into easy administration of a large network of machines. The tools and support needed for administering large numbers of machines are completely different from those needed for individual machines or small workgroups.
Microsoft customers make the same mistake: they think that because XP Home Edition sort-of works out of the box, that networks of XP machines must be pretty easy to set up and administer, too.
Your best bet for reducing administration costs for large sites is still UNIX and Linux systems. It isn't perfect, and you require a skilled IT staff to be able to deal with it, but in the end, it's more effective than having to hire dozens of people fiddling with OS X or Windows machines like they were home computers.
And, no, Mac OS X does not qualify as a UNIX or Linux system for the purposes of administration because its administrative tools and configuration system is quite different from that found on UNIX machines and because many software packages on OS X require GUI interaction and even reboots for their installations.
If you want secure remote access through a web page, you can use an unsigned Java ssh and/or Java VNC applet. But that's the only thing you can do without significant problems (well, to the degree that Java itself isn't already a problem).
For a true VPN, code running in the browser or invoked by the browser needs to get access to low-level networking, and there is no way in which that can be done without additional user intervention.
In the simplest case, you may be able to create a signed Java applet, but that will not simply run, it will at least require permission from the user to do something unsafe, and telling your users to give Java applets that kind of permission is not a good idea. Also, the Java APIs for building this kind of software are pretty limited, and I doubt you can do a true VPN implementation in Java.
Another choice would be ActiveX components. Again, they need to be signed and they require permission from the user. They could work in principle, but they only work for Windows. And you are probably going to have lots of support hassles anyway, with incompatibilities and software conflicts.
I think applet-based SSH and VNC are useful. But if you want a real VPN, your best bet is to go with an open standard like IPsec and to use a simple, stand-alone application with a good GUI to both start up the network connection and the VPN. That way, your users get secure access with a single click. What more do you eant?
Slashdot Mirror
Well, we don't have to guess about such things because they have been extensively studied:
Among developed nations, the US is pretty mediocre on most quality of life indicators. The high point of the US is its high per capita GDP (which is more of a statistical oddity than a meaningful quality of life indicator), but it also has a number of pretty black marks (income disparity, crime, capital punishment, health care, infant mortality, etc.). On human happiness indicators, the US tends to score even less well in international comparisons because material wealth and happiness are only weakly correlated.
There's a line between good and evil. We're on one side of it. Nations like North Korea and Iran are on the other side of it.
That statement is pretty ironic given that it was the US that toppled Iran's democratically elected government and replaced it with a brutal dictatorship.
And this notion of "good" and "evil" nations is quite interesting anyway. Let's see: is it the people of a nation that are evil, or just its government? And when did the US stop being evil and become a good nation?
Nope, we're not racist.
Slavery existed here until the middle of the 19th century, legal inequality until the middle of the 20th century, and statistics as well as personal experience show that prejudice and discrimination are still rampant in many parts of the country and many populations.
But feel free to live in your fantasy world.
No, you seem to live in a fantasy world, but without knowing more about you, it's hard to diagnose why you know so little about what's going on in the world or in the US.
hmm, how come the GPL is never good enough whenever someone speaks about Qt or Trolltech?
That's because Qt is pretty much the only GUI toolkit for Linux that is GPL'ed. Just about everybody else realizes that a GPL'ed toolkit is not good for Linux or free software.
The LGPL only seems reasonable, if you like proprietary software....
You say that as if the LGPL were some nefarious scheme hatched by opponents of free software. But, in fact, the LGPL (and the GPL-with-exemption, which is now preferred) was created by the FSF in order to advance free software. It was created because sometimes it is necessary to accomodate proprietary software in order to advance the cause of free software. Commodity or system libraries, like GUI toolkits, are exactly the kind of situation in which the LGPL makes a lot more sense than the GPL.
but the plan is to replace the current libraries with LGPL replacements.
That would be great, but last time people tried that (Harmony), Troll Tech became pretty belligerent. Also, it doesn't address the other major problem with Qt/Embedded: everybody has to rewrite their apps.
I think it would be easier just to ship Opie on top of X11. That way, it doesn't matter that Opie is GPL'ed because people can write commercial apps in other X11 toolkits.
That's only the beta test; because the system works so well, it will be extended to everybody making less than $1,000,000/year (in 2000 dollars) within a year after initial deployment.
As far as I can tell, Opie is a dead end: being an open source fork of Qt, the only license under which it is available is the GPL. That makes Opie an impossible platform for many commercial applications.
Areas where a Linux PDA could shine are gaming, inventory support, hospitals, instrument control, etc. But the developers of that kind of software simply are not going to make it open source, nor are their customers going to care whether it's open source or not. So, a GPL-only toolkit has no chance. But even the commercial version of Qt isn't all that attractive because if those people are going to throw in their lot with a commercial vendor, they might as well go with Microsoft and PPC: it's cheaper than Qt and more widely used.
Now, Opie/X11 might have a chance because it reuses all that effort that went into Opie but also opens up the platform for commercial apps.
Ultimately, GPL'ed or dual-licensed GUI toolkits just won't work in the long run--there are too many free or LGPL'ed alternatives to choose from.
It looks atrocious (at this point) and doesn't have near the specs dell offers for the same price.
Actually, if it's similar to the old VR3, the hardware design is really nice: the rounded corners make it easy to carry around, the screen protector works great, and the device is quite small.
Palm or PPC hardware looks flashier and more high-tech, but hardware like the VR3 is more usable.
I think there is a more fundamental problem with this. A lot of this code, NUMA, XFS, etc., was dumped into the Linux kernel not because most Linux users need it, but because it served the interests of a few corporate contributors. Furthermore, the nearly monolithic structure of the Linux kernel means that the way this sort of stuff gets contributed is via patches and additions to the source tree.
So, quite apart from the fact that SCO's claims have no merit, I think this does point out a pretty fundamental problem with the Linux kernel: it accretes too much stuff. The kernel should be architected in such a way that IBM and SGI could distribute most of their code separately. Furthermore, the Linux mainstream user base should not have to suffer from the increased maintenance hassles (and, as it turns out, increased legal risk) of having millions of lines of code they don't want or need poured into the kernel.
Technically, none of that is a problem with modern programming language and software engineering technologies. Even though some kernel features may require a lot of cross-cutting concerns, we now know how to implement those efficiently and in a principled way. Of course, this won't happen with the Linux kernel, which seems to be solidly stuck in 1960's software engineering technologies.
Don't get me wrong: the Linux kernel is enormously useful and it has lots of drivers. In practice, it is still the most practical kernel around as far as I'm concerned. But these kinds of legal issues, the Bitkeeper mess, as well as the ever more sluggish pace at which new features get incorporated and new kernels get released, are symptoms of a deep problem, and that problem is not going to go away.
Do you really think I am being an "ideolog" when I believe that widespread torture is wrong? that imprisonment without trial of people exercising their right to freedom of expression is wrong?
You don't stop at believing it's wrong, you argue that people should act against the wrongdoing of foreign governments instead of participating in the democratic process at home. It is that part that I disagree with. We are responsible for the misdeeds of our own governments, and there are plenty of things our own government does wrong.
And you don't have an answer for what the Chinese government should do anyway. I mean a couple of centuries of democracy in the US haven't caused us to conform to international human rights standards--our government violates human rights, tortures and kills people, and engages in wars by popular choice. Many times in the 20th century, the US ousted democracies in order to institute US-friendly, human-rights violating dictatorships, so it doesn't seem like human rights and democracy are a big goal of US foreign policy either. So, tell us, except for pointing fingers, what is your plan for China?
People like you just like to use countries like China, Iraq, and North Korea as convenient excuses for dealing with the enormous social and economic problems here at home.
I have a Sharp Zaurus, and I think Qtopia just doesn't cut it.
Sure, it looks pretty nice and it has most of the functionality you might want in a PDA, but it is still significantly worse than either Palm or PPC. Some of the problems include badly thought out user interaction, wasteful use of the limited screen real estate (probably a result of being based on an adaptation of a desktop toolkit), and pretty excessive resource consumption by Qt/Embedded. And there is far less software available for Qtopia than for either Palm or PPC. If you want good PDA functionality, get a Palm.
On the other hand, as a Linux PDA for vertical apps, Qtopia-based PDAs also fall short: you are limited to the Qt toolkit and all the graphics and UI code from existing Linux apps require complete rewrites. You can't use any of the open source GUI tools you are likely used to (Tcl/Tk, Python/Gtk+, etc.). And if you want to write commercial apps, it's going to cost you (you can do commercial Palm development for free).
Linux PDAs will keep failing until their makers recognize that it is futile to compete with Palm and PPC head-on. Linux PDAs can thrive in the niche of providing portable little Linux machines, but that means not limiting the machines to running just a single GUI toolkit.
China is a country of a billion people, with many diverse ethnic groups. It's a country undergoing massive changes, and the whole coutry is walking a tightrope, always at risk of falling apart or rampant corruption.
I doubt that the current Chinese government is the best possible for China, and they are certainly far from a democracy yet, but anybody claiming to know that they know better how to run China is either an idiot or an ideolog. It took Europeans hundreds of years and many bloody wars to become modern democracies. Americans had to commit genocide and institute centuries of slavery before finally waking up to ideas of human rights and equality. Give the Chinese a break--they aren't doing all that badly in comparison, they are just a little late.
Oh, and as for Microsoft, the Chinese are just watching out for their economic interests: nurturing domestic high-tech expertise is a good thing for them, and replacing Microsoft software with domestically developed software just makes sense.
Trust me-I have been there-all you get is flamewars with bad spelling. People used to flame using even handwriting and snail mail.
They are the government of an autonomous country--they make the laws and they can do whatever they want when it comes to copyright.
Now I only wish they would turn their minds on and realize that most of the world's ecological abuse goes on in third-world nations, as does most of the world's racism, sexism, and war.
US citizens are responsible for what happens in the US, not what happens in third world nations. Therefore, it makes perfect sense for US citizens to participate in the US political process, and part of that is rallies and demonstrations in the US against US government policies.
And, in any case, what kind of argumentation is that supposed to be anyway? You seem to be saying that US can be racist and destroy its environment because North Korea is even worse. If you want to engage in international comparisons at all, then the US should compare itself to the best nations in these areas, not to the worst, and in all the areas you mention, the US is in pretty bad shape in international comparisons.
You know, I just remembered why we don't launch nuclear waste out into space. It would -really- suck if it blew up before it cleared orbit, wouldn't it?
Nuclear waste is very different from nuclear fuel, and there are nuclear fuels that are not all that hazardous. Uranium, for example, is not very strongly radioactive. What is dangerous is all the short-lived by-products produced during fission reactions.
Just imagine if something went wrong like chernobyl. Except this time it's 30 miles in the air where it can travel around the globe quite nicely.
No, it's not like Chernobyl. Chernobyl was a running nuclear reactor, which had already created lots of fission products. It's also not like Plutonium-based nuclear batteries--Uranium is far less poisonous than Plutonium. So, in terms of risk, the objections may be less strong.
But people may object to any nation sending nuclear reactors into space. The US and other nations would love to use nuclear reactors to power space-based weapons, and this sort of civilian project might create the infrastructure for, and legitimize, such uses of nuclear power in space.
Whether it is ecology issues, racism, sexism or peace marches, most of the action is here in the USA where well-to-do clueless college kids are in abundance.
So, you much rather prefer turning the US into a fascist state, military dictatorship, or making it so poor that nobody can protest? Well, the way things are going, you may yet get your wish.
I am grateful that I live in a country in which college kids and environmentalists both have the freedom and sufficient economic resources to worry about ecology, racism, sexism, and peace.
It is about time mankind gets out of low-earth-orbit, either by giving up on space altogether
Giving up on manned space exploration is not the same as giving up on space. Manned space exploration is worthless from a scientific point of view, and it is a bottomless sink for money and a boon for governmental pork barrel spending. Unmanned space exploration, on the other hand, is highly valuable scientifically and much cheaper. And all money-making commercial use of space is unmanned as well.
So it needs people on Mars to run it, and people on Mars to take advantage of it. Do they actually have any firm plans for getting people to Mars?
No, they'll just hire locals--it's cheaper.
and we probably would be if we'd kept up what we were doing in the 60's.
By "what we were doing in the 60's", you mean "driven by the cold war and the threat of nuclear annihilation, pursue senseless vanity space projects to demonstrate to the world one's technological superiority"?
Well, I am glad that we didn't keep this up. In fact, I don't give a damn whether people set foot on Mars during my lifetime, and I wouldn't want another dollar wasted on manned space exploration. I'd much rather see that money go to unmanned probes, which yield much more interesting scientific data.
No, that can't be it: the Linux implementations of these features are much, much smaller than that.
There are three and four CCD digital cameras; this one isn't one of them. Instead, it uses four different color filters to get more information about the spectral composition of the incoming light. That lets it correct for color aberrations better than 3 color CCDs. This is an old trick, and you can actually get film that does the same thing.
Palm hardware is fairly nice: compact, light, long battery life, and the applications are pretty decent as well.
But PalmOS is becoming a drag on the company--PalmOS was OK for underpowerd 68k-based handhelds, but the 175-400MHz RISC handhelds of today would be much better off running a more modern OS; none of the restrictions, flakiness, and limitations of PalmOS are justified by the hardware it runs on anymore. Neither PalmOS 5 nor PalmOS 6 look like they are going to fix the grave deficiencies in PalmOS.
One can only hope that, with this split, the Palm hardware group can look to other options in terms of software: Symbian, Linux, etc. They can even keep running the existing Palm applications, which people are used to. But the OS has to go.
xfree86 is mostly derived from a free reference implementation. It may fail to offer hardware acceleration for many reasons. For example, the necessary hw specs may just not he published. Blit was hardware accelerated in commercial X11 servers pretty much as soon as the hardware was capable of supporting it.
Keep in mind that there are several different, independent X11 implementations from different vendors.
XRender is a new extension with only a reference implementation in XFree86. The point is to experiment with an API prior to freezing it. I know this may come as news to people who have grown up on Microsoft software, but real software developers first try out various ideas and then later start hacking it for speed. It would be quite surprising, actually, if it were faster than a hand-tuned client-side software implementation.
It will be a while until XRender beats client-side software implementations. Furthermore, you can't just take a client-side renderer and hack in XRender calls and expect it to run fast--code that works efficiently with a client-server window system like X11 needs to be written differently than something that moves around pixels locally.
Macintosh systems are pretty easy to administer for individuals. It's also fairly easy to set up ad-hoc workgroups with Macintosh.
But that does not translate into easy administration of a large network of machines. The tools and support needed for administering large numbers of machines are completely different from those needed for individual machines or small workgroups.
Microsoft customers make the same mistake: they think that because XP Home Edition sort-of works out of the box, that networks of XP machines must be pretty easy to set up and administer, too.
Your best bet for reducing administration costs for large sites is still UNIX and Linux systems. It isn't perfect, and you require a skilled IT staff to be able to deal with it, but in the end, it's more effective than having to hire dozens of people fiddling with OS X or Windows machines like they were home computers.
And, no, Mac OS X does not qualify as a UNIX or Linux system for the purposes of administration because its administrative tools and configuration system is quite different from that found on UNIX machines and because many software packages on OS X require GUI interaction and even reboots for their installations.
If you want secure remote access through a web page, you can use an unsigned Java ssh and/or Java VNC applet. But that's the only thing you can do without significant problems (well, to the degree that Java itself isn't already a problem).
For a true VPN, code running in the browser or invoked by the browser needs to get access to low-level networking, and there is no way in which that can be done without additional user intervention.
In the simplest case, you may be able to create a signed Java applet, but that will not simply run, it will at least require permission from the user to do something unsafe, and telling your users to give Java applets that kind of permission is not a good idea. Also, the Java APIs for building this kind of software are pretty limited, and I doubt you can do a true VPN implementation in Java.
Another choice would be ActiveX components. Again, they need to be signed and they require permission from the user. They could work in principle, but they only work for Windows. And you are probably going to have lots of support hassles anyway, with incompatibilities and software conflicts.
I think applet-based SSH and VNC are useful. But if you want a real VPN, your best bet is to go with an open standard like IPsec and to use a simple, stand-alone application with a good GUI to both start up the network connection and the VPN. That way, your users get secure access with a single click. What more do you eant?