Warez are an economic problem: warez distribute expensive software to people who wouldn't otherwise buy it, giving it more market share without eating into profits. It's a great mechanisms for differential pricing. The losers are competitors trying to enter the market with a comparable product at a lower price because few people care: those that pirate are going to pirate #1, and the people with money pay the extra 30%.
Your conclusion does not follow from the premise. The GPL is not an EULA -- you do not have to agree to it to use the software. The GPL only depends on copyright law. The EULA's depend on contract law. Some think that existing contract law does not or should not permit EULA's.
I think your analysis is wrong. It is copyright that allows GPL software authors to keep others from copying the software. But the GPL then establishes a contract that says "in return for giving you permission to use this software, you agree to these terms".
A EULA basically works the same way, although companies like to pretend that copyright isn't involved at all, and they may be able to corrupt the legal system to the point that it isn't.
So, if you play something like Half Life or The Thing, your MacDonalds hamburger can sprout legs, start oozing blood, and attack you. I always thought they were made from alien meat anyway.
Shine a laser at a mountain a hundred miles away and rotate at modest speed--the spot of light will move faster than light. From the fluffy description in the New Scientist, it sounds as if they roughly did an electrical version of that--what moves is something you construct in your mind, not anything tangible or anything you could use to "send signals faster than light". And, unlike the "complicated setups" they are referring to, their effect is purely classical.
Mistakes and fraud will happen, and they will slip through peer review--that's inevitable. The problem is not that this happens, but that science, and physics in particular, have a celebrity culture kind of like Hollywood does so that these things end up hurting other people--a popular fraud can attract more funding and attention than a dozen people coming up with less glamorous results. And many of the most hyped results turn out to be more good PR than breakthroughs when things have calmed down.
While scientists only recently started promising getting bigger penises in a serious way, they have been announcing get rich quick schemes and a cure for cancer for a century, and people keep falling for it. Science even has its tabloid press, of which The New Scientist and certain section of Nature are a good example (but Nature at least also contains a lot of good science).
I kind of doubt US government agencies could give an accurate accounting of how many phones they have tapped if they wanted to, and they probably don't even want to. And "tapped" probably doesn't take into account any kind of monitoring and audio keyword search that isn't aimed at a specific person.
I just don't think that's such a large problem--people tend not to want to injure themselves, and the really reckless ones already have plenty of choices, both legal and illegal.
This is a personal choice--it doesn't endanger anybody else. I think ride operators should be required to label and state clearly what is known about the dangers, but the state shouldn't prohibit such operations--people should be free to hurt or kill themselves in whatever way they like. But, then, I think the same about both recreational and medical drugs.
No it was a zaurus, I just looked at it again at cicuit city, though they don't have it on their webpage(like many things). It was something like 450 bucks. It's the most expensive PDA they have in the store.
I'm sorry, but you are just making that up. Every single PocketPC that Circuit City sells costs more than that. Check their web page.
It is slow!
Not compared to PocketPC machines.
It is Expensive!
At $330, it's 2/3 the price of the NR70, the closest Sony model. And for that, you get a much more powerful handheld.
in the end a sony clie now sits on my desk
The Clies are very nice organizers. An SJ20 or SJ30 is a really good deal. If a Clie gets your job done, be happy. But as handheld platforms, the Palm-based systems are just much more limited in what they can do.
If you need IPsec, then you can certainly recompile your kernel.
So, you admit it, you just don't think it's a problem. Well, wake up and smell the coffee: for real-world installations, this is a problem. Both sys admins and users have better things to do than recompile kernels.
If you're using a modern distribution,
I'm using Debian and RedHat.
everything should be supported out of box.
Well, it isn't on the majority of machines that I have installed. It's things like ACPI, Mosix, audio cards, on-board networks, Bluetooth, FireWire, multimedia, USB devices, and file system types.
Just as Windows doesn't support certain hardware out of box. Both take some work to get running.
On Windows or MacOS, I can download a ready-made driver package and install it. On Linux, I should be able to do an "apt-get install bluetooth-drivers" or "apt-get install ipsec-kernel-module", and it should download a few hundred kilobytes at most, but no distribution has figured out how to make that work. And that's the problem.
Money-for-teaching deals are bad whether Microsoft or Sun does them. Did you report a story about any instances when Sun did this? If not, I don't see why you would complain about "Slashdot myopia".
Any CS or EE curriculum that aims at teaching mastery of a language is not worth taking anyway. To the degree that CS or EE should teach programming at all, it should teach general principles of programming, not the idiosyncracies of specific languages.
I agree with you that "industrial" languages like Java, C#, C, or C++ are particularly bad choices for introductory teaching. If they are used in advanced courses, that's because the libraries and support to teach the subject matter only exists in them, not to teach those languages per se. Pascal actually still isn't a bad choice for teaching, although I think Scheme and a few others are probably better. The fact that they are commercially irrelevant is an advantage as far as I'm concerned.
Before you go faulting the design of the kernel, I'd ask you, what do you know that everyone hacking on the kernel doesn't. Face it, hardware changes, the goals of the OS change.
Microsoft, Apple, Amiga, and BeOS manage to make it work. It's only Linux that, in practice, seems to require kernel recompilation for many installations and distributions.
In the real world, people don't use sissy [...]
Look, I'm speaking as a user. I don't care about your (mis-)conceptions about software engineering or systems programming, I'm not suggesting any specific solutions for Linux. I'm telling you: the Linux kernel is my biggest headache in maintaining Linux desktops and servers. All the other stuff is handled wonderfully by the standard packaging and configuration systems.
On my Inspiron 8200 laptop, for example, every single gadget I have, [...]
Yes, the rallying cry of a software developer who doesn't care about users: "I like the way it works". And that's fine. If the Linux kernel developers don't want to fix that, that's their choice, that's the way open source works. But if they want continue to see widespread usage by others, I predict they need to fix this, because a kernel without this deficiency (from the point of many users) will come along sooner or later.
Example: IPsec. Not included in the standard kernel. In order to get it working, I'll have to patch, configure, and recompile kernels for half a dozen different machines. For handhelds running Linux, this will be even more of a chore.
And because there is no reason to do that, it makes no sense to limit the kernel developers just so the 3 people that distribute seperate drivers can have a stable ABI.
Yes, and that is one of the problems: rather than fixing the kernel, kernel developers just stick more and more drivers into the kernel source tree.
People don't go, "I don't use Linux because its not a microkernel written in Scheme," people say "I don't use Linux because it doesn't have the software I need."
I agree 100%. And the software they need that isn't working is the drivers and other kernel modules they need to get their hardware working and communicate with the rest of the world.
I have tried a Zaurus though i belive a differant model.
AFAIK, there effectively has been only one commercially released model (the developer model is similar but has only 32 Mbytes).
It's slow as hell!
You must be confusing the Zaurus with something else. There have been three Linux PDAs: the Agenda/VR, the Yopy, and the Zaurus. Linux also runs in the iPaqs. The Agenda/VR runs on a 66MHz MIPS chip, and it is a bit sluggish--but once applications are loaded, it's OK. I haven't tried the Yopy, but I don't think you could have either--as far as I know, it's only available mail order.
While it has some flaws (see above), the Zaurus UI is very fast and responsive, comparable to Palm and better than Windows CE, in my experience. In fact, the quality of the Zaurus UI is at least comparable to Windows CE, not just in my opinion, but also in many reviews. On the iPaq, we can even make a side-by-side comparison of the performance on identical hardware, and X11 running on the iPaq beats Windows CE hands down.
the price wasn't there, though much cheaper than the linux models.
The Zaurus costs $340 mail order, for a 206MHz device with keyboard, CF, MMC, MP3, and other features, and it comes with a huge amount of software. It is, and has always been, cheaper than comparable Windows CE machines.
I know people with WinCE models and love them
To each their own. To me, the Windows CE machines aren't even in the running: they are expensive, slow, and lack most of the software that I would want to run.
I think the Zaurus is a great little machine: it runs like a charm and it's easy to port non-GUI software to it. The packaging system works great. The expansion slots and support for CF-based hardware is great. But I think the system is really held back by its UI.
First, Qt is clearly is a desktop widget set that has been converted to a handheld: the shape and size of many components simply takes up way too much space, completely unnecessarily, and the overall layout of applications is also quite wasteful of screen space. The Zaurus Opera browser illustrates this: at the "tiny" settings: you still get the 3D decorations for buttons and text entry boxes, but the text for those widgets is completely gone; now, which is more important--wasting pixels on a 3D look or being able to see the actual text? There are lots of ways of indicating buttons that do not require a several pixel wide 3D border. Of course, even at larger scales, when you can read the text, those pixels take up unnecessary space. If you only have 240x320 pixels to play around with, this kind of thing needs to be highly optimized, and there are better ways of doing this. What we really need is a special-purpose handheld toolkit, not a port from a desktop. Altogether, with its 240x320 screen, Zaurus applications feel more space constrained than a Palm Pilot; it's the same problem that Windows CE and PocketPC have.
Second, Qt/QPE software takes up lots of memory to run. The QPE process alone takes up nearly 8 Mbytes of memory, with a terminal application taking an additional 3.8 Mbytes (all RSS while being used). A full, running X server (Xvnc) running on the same system takes 1.3 Mbytes, and that includes additional support for the VNC protocol, and a handheld version of rxvt can be run in a few hundred kbytes of memory. The Agenda/VR applications also took up a fraction of the amount of memory of what the equivalent Zaurus applications take up. It's a myth that Qt/Embedded is memory-efficient.
Third, the use of Qt/Embedded cuts the Zaurus off from a lot of other Linux handheld development: there are lots of neat, small, efficient applications written for other widget sets, some of them straight to X11, some of them using FLTK and others. While you can run them using Xvnc on a Zaurus, you end up with two completely unintegrated environments--that is not acceptable for day-to-day usage.
Sharp and Lineo should recompile the Zaurus applications to use X11; given that they are written in Qt, that should be easy, although they may have to do some performance tuning on the X11 version of Qt. Then, people could pick and choose which applications they like to run among a larger variety of software. I'd replace some of the built-in applications with FLTK-based ones, saving both memory and getting an interface I prefer. And then there would be a lot more software available for it and we wouldn't need reviews of four tiny applets at the amazing price of $10 each.
That's more attributable to the fact that Linus doesn't want to freeze the driver API rather than any fault of the design.
If, after 10 years of hacking, it's not possible to provide a basic set of APIs for drivers, file systems, and other common kernel components, then the design is at fault. If not anything else, the Linux kernel could have two sets of APIs: stable and experimental. Other kernel architectures, involving message passing, RPC, or objects, also force people to think about this rather than keep changing things around haphazardly.
Umm, the same thing happens with most OSs.
I don't know what "most" means, but there are certainly many ways of avoiding that problem. For example, if you write the kernel in something other than assembly or C/C++, it gets much harder to crash the kernel accidentally. If you build a message passing kernel, you can transparently move drivers in and out of kernel address space, trading off performance and safety as needed. It's only monolithic kernels written in an unsafe language that are this sensitive.
Don't get me wrong: Linux has been a reliable workhorse for many years, and the functionality in it is wonderful. But I think these issues are really becoming the biggest obstacle to its more widespread adoption and use on the desktop, and it's only going to get worse. If people don't seriously start thinking about addressing this now, some other kernel will take over in a few years, and that would mean more hassle for everybody. There are many ways of fixing it (see above), but first the patient has to admit that there is a problem, and I don't see that happening yet.
You don't see that much commercial software for Linux because Linux has many of the mainstream software categories reasonably well covered with free software. No, you don't exactly get MS Office or Adobe Photoshop, but you get applications that are functionally pretty close. It's primarily niche and specialty software for which it makes sense to make a Linux port--and that software is being ported--software like Matlab, design software, embedded tools, etc.
No, but you could, if you needed it to be a shell builtin for performance or other reasons.
Indeed. When I do need the performance, it would be nice to be able to load modules dynamically. But for something like IPsec, PPP, UFS, ISO9660, CMOS, etc., I don't need the performance; maybe you do on your big server, but I don't, on my little laptop. As for "other reasons", there shouldn't be any reason other than performance to load something dynamically.
You still have to write and compile something - why not a kernel module?
Because, empirically, kernel modules seem to end up being very dependent on kernel versions; if they weren't, distributions like Debian wouldn't ship with different collections of most kernel modules for each kernel, they would ship with one kernel module per package for each function/driver, without much of a notion of a "kernel version".
Another reason is that one bug in one kernel modules brings down the whole thing. That's unnecessary and makes driver development a huge pain.
Not that OS X is actually a microkernel OS!
I made no claims about what it is or even whether it is a good architecture. What I claimed was that Jobs correctly identified a problem and tried to address it as best as possible with the software available at the time.
And I think he actually succeeded much more than Linux did in this particular regard: kernel extensions on OS X work much better than on Linux.
(Jobs also correctly identified the problem with C/C++ GUI toolkits and his solution, Objective-C with DisplayPostscript, probably also was the best technical compromise at the time, but I think that choice hasn't turned out as well as his choice for kernel--OpenStep and Cocoa ended up with most of the same problems as other GUI toolkits.)
Most laptops have IR built in, which lets you set up a network and gives you up to 4Mbps. If it's not built in, you can get a small and cheap USB dongle.
Bluetooth is another choice: tiny USB dongles and standardized. Up to 1Mbps. The software isn't quite as mature yet, but it's getting there.
You should be suspicious of companies that give away commercial software for free. Such a move is generally aimed at eliminating competition, and you'll be paying a lot more in the long run. The same also holds for some companies that "give away" dual licensed software: they may be trying to use open source as a way to drive out competitors.
Accept something for free if it is clear that the giver has no commercial interest in giving it to you or if the relationship is such that the giver can't exercise control over you or the product later on. Otherwise, be very suspicious and try to avoid the "gift" if you can.
The *BSD kernels may be a little more reliable and simple, and the Linux kernel may support more drivers, but it seems to me the differences are pretty much lost in the noise.
I think Jobs had the right idea when he picked Mach as the basis for NeXTStep: he wanted a kernel that looked like UNIX from the outside but that was much more componentized than the UNIX kernels of the time, or BSD/Linux today. I don't know whether Mach/Darwin is the best choice for that, but in general, I think it's where open source needs to go.
After all, we don't recompile Bash or dynamically load libraries into Bash every time someone comes out with a new command line program. We shouldn't have to do that either for a new file system type, networking protocol, or driver. And expending much time on a BSD/Linux rivalry isn't going to address such issues.
Which is fine until 'what the wealthy enjoy' becomes basic necessities like food and energy.
Energy (and water and food) consumption at US levels is not a "basic necessity", it is an absurd and completely unnecessary waste. Apart from any arguments about growth, it is just obscene.
Now I don't get your argument. You say that a few people dying from radiation is such a catastrophe that we need to go massively overboard on our safety standards. On the other hand, millions dead in food riots doesn't seem to phase you.
No, I'm saying that your reasoning is sensible, simple, and, unfortunately, completely wrong. If only things were as simple as you think they are: people are hungry, so you produce more food to feed them, problem solved.
But populations and consumption adjust to resources: we create more available and they grow. You end up facing a worse problem when you are done than when you started. What needs to be fixed is the influence of available resources on growth and consumption.
A good start towards that goal is to stop growing. Not only does that stop the problem from getting bigger and bigger, it also takes away the illusion that if we just grow a little more, our problems will be solved. Once that illusion is gone, people can perhaps start thinking about making the necessary hard choices.
It's like highways: building more doesn't reduce congestion, it actually makes it worse.
Oh, this will happen without taxation. Scarcity will increase the price.
Unfortunately, our capacity to consume energy and produce carbon dioxide or nuclear waste is not limited by available resources. There is probably enough carbon in the ground to increase atmospheric carbon dioxide by orders of magnitude, and we can mine and breed nearly unlimited amounts of nuclear fuel. Sooner or later, we either die out as a species or we have to decide to stop increasing consumption. We might as well face the facts now--it only gets worse the longer we go on.
Am I being trolled here?
Not by me, but you probably are being trolled by your favorite politician and 1960's style economist.
Have you ever seen what happens to the poor (in free markets) when basic necessities like energy and food become expensive? Sure, demand is reduced, but large segments of the population will often go without while the wealthy continue to enjoy abundance.
Large segments of the population always "go without" what the wealthy enjoy. Large segments of the US population "go without" retirement benefits, job security, privacy, decent health care, extended vacations, and a lot of other things. Those are a lot more important than the ability to cruise around in a gas guzzling SUV, compared to a small, efficient vehicle. If we wanted to reduce those inqualities, that would be very easy: through high taxes on the top few percent of earners, just like income tax was intended to do initially.
Energy prices should be gradually but steadily increased, through taxation on both non-renewable fuels and inefficient consumption. Society would adapt to this by reducing energy usage. That's pretty much the only way, and it works.
Your argumentation is like that of a drug addict who says "I know it's bad for me, but stopping is unpleasant". Your solution to one dangerous drug (oil) is to switch to another dangerous one (nuclear energy), which just happens to not make you look quite as bad.
We (western nations) happen to be the ones exporting food to the world. If we stopped or raised prices significantly, many third world countries would no longer be able to afford enough to keep their large (and growing) populations fed.
Yup, I have to admit it's a neat package. We destroy indigenous cultures and farming methods, effectively encourage population growth, interfere with family planning, and cause other nations to become dependent on our food and indebted to us. The latest twist is to make farmers dependent on even more expensive, genetically modified crops that may or may not work in our industrial farms but make no sense in other countries.
Creating large masses of poor and hungry people is great for getting powerful. The Catholic church discovered this a long time ago, but we are doing it so much better these days.
In different news, scientists also have discovered a 100 million year old penis (two, actually), the oldest one ever found. Is it just the time of year for penises in the news? Does the end-of-summer news hole need to get filled?
Incidentally, the other article also says that the one grown in the lab works more like that of a 60 year old--rabbit? Doesn't sound too attractive, unless you happen to be an 80 year old rabbit, I suppose.
Slashdot Mirror
Warez are an economic problem: warez distribute expensive software to people who wouldn't otherwise buy it, giving it more market share without eating into profits. It's a great mechanisms for differential pricing. The losers are competitors trying to enter the market with a comparable product at a lower price because few people care: those that pirate are going to pirate #1, and the people with money pay the extra 30%.
I think your analysis is wrong. It is copyright that allows GPL software authors to keep others from copying the software. But the GPL then establishes a contract that says "in return for giving you permission to use this software, you agree to these terms".
A EULA basically works the same way, although companies like to pretend that copyright isn't involved at all, and they may be able to corrupt the legal system to the point that it isn't.
Come on, you know it's a steam powered pyramid water pump :-)
So, if you play something like Half Life or The Thing, your MacDonalds hamburger can sprout legs, start oozing blood, and attack you. I always thought they were made from alien meat anyway.
Shine a laser at a mountain a hundred miles away and rotate at modest speed--the spot of light will move faster than light. From the fluffy description in the New Scientist, it sounds as if they roughly did an electrical version of that--what moves is something you construct in your mind, not anything tangible or anything you could use to "send signals faster than light". And, unlike the "complicated setups" they are referring to, their effect is purely classical.
While scientists only recently started promising getting bigger penises in a serious way, they have been announcing get rich quick schemes and a cure for cancer for a century, and people keep falling for it. Science even has its tabloid press, of which The New Scientist and certain section of Nature are a good example (but Nature at least also contains a lot of good science).
I kind of doubt US government agencies could give an accurate accounting of how many phones they have tapped if they wanted to, and they probably don't even want to. And "tapped" probably doesn't take into account any kind of monitoring and audio keyword search that isn't aimed at a specific person.
I just don't think that's such a large problem--people tend not to want to injure themselves, and the really reckless ones already have plenty of choices, both legal and illegal.
This is a personal choice--it doesn't endanger anybody else. I think ride operators should be required to label and state clearly what is known about the dangers, but the state shouldn't prohibit such operations--people should be free to hurt or kill themselves in whatever way they like. But, then, I think the same about both recreational and medical drugs.
I'm sorry, but you are just making that up. Every single PocketPC that Circuit City sells costs more than that. Check their web page.
It is slow!
Not compared to PocketPC machines.
It is Expensive!
At $330, it's 2/3 the price of the NR70, the closest Sony model. And for that, you get a much more powerful handheld.
in the end a sony clie now sits on my desk
The Clies are very nice organizers. An SJ20 or SJ30 is a really good deal. If a Clie gets your job done, be happy. But as handheld platforms, the Palm-based systems are just much more limited in what they can do.
So, you admit it, you just don't think it's a problem. Well, wake up and smell the coffee: for real-world installations, this is a problem. Both sys admins and users have better things to do than recompile kernels.
If you're using a modern distribution,
I'm using Debian and RedHat.
everything should be supported out of box.
Well, it isn't on the majority of machines that I have installed. It's things like ACPI, Mosix, audio cards, on-board networks, Bluetooth, FireWire, multimedia, USB devices, and file system types.
Just as Windows doesn't support certain hardware out of box. Both take some work to get running.
On Windows or MacOS, I can download a ready-made driver package and install it. On Linux, I should be able to do an "apt-get install bluetooth-drivers" or "apt-get install ipsec-kernel-module", and it should download a few hundred kilobytes at most, but no distribution has figured out how to make that work. And that's the problem.
Money-for-teaching deals are bad whether Microsoft or Sun does them. Did you report a story about any instances when Sun did this? If not, I don't see why you would complain about "Slashdot myopia".
Any CS or EE curriculum that aims at teaching mastery of a language is not worth taking anyway. To the degree that CS or EE should teach programming at all, it should teach general principles of programming, not the idiosyncracies of specific languages.
I agree with you that "industrial" languages like Java, C#, C, or C++ are particularly bad choices for introductory teaching. If they are used in advanced courses, that's because the libraries and support to teach the subject matter only exists in them, not to teach those languages per se. Pascal actually still isn't a bad choice for teaching, although I think Scheme and a few others are probably better. The fact that they are commercially irrelevant is an advantage as far as I'm concerned.
Microsoft, Apple, Amiga, and BeOS manage to make it work. It's only Linux that, in practice, seems to require kernel recompilation for many installations and distributions.
In the real world, people don't use sissy [...]
Look, I'm speaking as a user. I don't care about your (mis-)conceptions about software engineering or systems programming, I'm not suggesting any specific solutions for Linux. I'm telling you: the Linux kernel is my biggest headache in maintaining Linux desktops and servers. All the other stuff is handled wonderfully by the standard packaging and configuration systems.
On my Inspiron 8200 laptop, for example, every single gadget I have, [...]
Yes, the rallying cry of a software developer who doesn't care about users: "I like the way it works". And that's fine. If the Linux kernel developers don't want to fix that, that's their choice, that's the way open source works. But if they want continue to see widespread usage by others, I predict they need to fix this, because a kernel without this deficiency (from the point of many users) will come along sooner or later.
Example: IPsec. Not included in the standard kernel. In order to get it working, I'll have to patch, configure, and recompile kernels for half a dozen different machines. For handhelds running Linux, this will be even more of a chore.
And because there is no reason to do that, it makes no sense to limit the kernel developers just so the 3 people that distribute seperate drivers can have a stable ABI.
Yes, and that is one of the problems: rather than fixing the kernel, kernel developers just stick more and more drivers into the kernel source tree.
People don't go, "I don't use Linux because its not a microkernel written in Scheme," people say "I don't use Linux because it doesn't have the software I need."
I agree 100%. And the software they need that isn't working is the drivers and other kernel modules they need to get their hardware working and communicate with the rest of the world.
AFAIK, there effectively has been only one commercially released model (the developer model is similar but has only 32 Mbytes).
It's slow as hell!
You must be confusing the Zaurus with something else. There have been three Linux PDAs: the Agenda/VR, the Yopy, and the Zaurus. Linux also runs in the iPaqs. The Agenda/VR runs on a 66MHz MIPS chip, and it is a bit sluggish--but once applications are loaded, it's OK. I haven't tried the Yopy, but I don't think you could have either--as far as I know, it's only available mail order.
While it has some flaws (see above), the Zaurus UI is very fast and responsive, comparable to Palm and better than Windows CE, in my experience. In fact, the quality of the Zaurus UI is at least comparable to Windows CE, not just in my opinion, but also in many reviews. On the iPaq, we can even make a side-by-side comparison of the performance on identical hardware, and X11 running on the iPaq beats Windows CE hands down.
the price wasn't there, though much cheaper than the linux models.
The Zaurus costs $340 mail order, for a 206MHz device with keyboard, CF, MMC, MP3, and other features, and it comes with a huge amount of software. It is, and has always been, cheaper than comparable Windows CE machines.
I know people with WinCE models and love them
To each their own. To me, the Windows CE machines aren't even in the running: they are expensive, slow, and lack most of the software that I would want to run.
First, Qt is clearly is a desktop widget set that has been converted to a handheld: the shape and size of many components simply takes up way too much space, completely unnecessarily, and the overall layout of applications is also quite wasteful of screen space. The Zaurus Opera browser illustrates this: at the "tiny" settings: you still get the 3D decorations for buttons and text entry boxes, but the text for those widgets is completely gone; now, which is more important--wasting pixels on a 3D look or being able to see the actual text? There are lots of ways of indicating buttons that do not require a several pixel wide 3D border. Of course, even at larger scales, when you can read the text, those pixels take up unnecessary space. If you only have 240x320 pixels to play around with, this kind of thing needs to be highly optimized, and there are better ways of doing this. What we really need is a special-purpose handheld toolkit, not a port from a desktop. Altogether, with its 240x320 screen, Zaurus applications feel more space constrained than a Palm Pilot; it's the same problem that Windows CE and PocketPC have.
Second, Qt/QPE software takes up lots of memory to run. The QPE process alone takes up nearly 8 Mbytes of memory, with a terminal application taking an additional 3.8 Mbytes (all RSS while being used). A full, running X server (Xvnc) running on the same system takes 1.3 Mbytes, and that includes additional support for the VNC protocol, and a handheld version of rxvt can be run in a few hundred kbytes of memory. The Agenda/VR applications also took up a fraction of the amount of memory of what the equivalent Zaurus applications take up. It's a myth that Qt/Embedded is memory-efficient.
Third, the use of Qt/Embedded cuts the Zaurus off from a lot of other Linux handheld development: there are lots of neat, small, efficient applications written for other widget sets, some of them straight to X11, some of them using FLTK and others. While you can run them using Xvnc on a Zaurus, you end up with two completely unintegrated environments--that is not acceptable for day-to-day usage.
Sharp and Lineo should recompile the Zaurus applications to use X11; given that they are written in Qt, that should be easy, although they may have to do some performance tuning on the X11 version of Qt. Then, people could pick and choose which applications they like to run among a larger variety of software. I'd replace some of the built-in applications with FLTK-based ones, saving both memory and getting an interface I prefer. And then there would be a lot more software available for it and we wouldn't need reviews of four tiny applets at the amazing price of $10 each.
If, after 10 years of hacking, it's not possible to provide a basic set of APIs for drivers, file systems, and other common kernel components, then the design is at fault. If not anything else, the Linux kernel could have two sets of APIs: stable and experimental. Other kernel architectures, involving message passing, RPC, or objects, also force people to think about this rather than keep changing things around haphazardly.
Umm, the same thing happens with most OSs.
I don't know what "most" means, but there are certainly many ways of avoiding that problem. For example, if you write the kernel in something other than assembly or C/C++, it gets much harder to crash the kernel accidentally. If you build a message passing kernel, you can transparently move drivers in and out of kernel address space, trading off performance and safety as needed. It's only monolithic kernels written in an unsafe language that are this sensitive.
Don't get me wrong: Linux has been a reliable workhorse for many years, and the functionality in it is wonderful. But I think these issues are really becoming the biggest obstacle to its more widespread adoption and use on the desktop, and it's only going to get worse. If people don't seriously start thinking about addressing this now, some other kernel will take over in a few years, and that would mean more hassle for everybody. There are many ways of fixing it (see above), but first the patient has to admit that there is a problem, and I don't see that happening yet.
You don't see that much commercial software for Linux because Linux has many of the mainstream software categories reasonably well covered with free software. No, you don't exactly get MS Office or Adobe Photoshop, but you get applications that are functionally pretty close. It's primarily niche and specialty software for which it makes sense to make a Linux port--and that software is being ported--software like Matlab, design software, embedded tools, etc.
Indeed. When I do need the performance, it would be nice to be able to load modules dynamically. But for something like IPsec, PPP, UFS, ISO9660, CMOS, etc., I don't need the performance; maybe you do on your big server, but I don't, on my little laptop. As for "other reasons", there shouldn't be any reason other than performance to load something dynamically.
You still have to write and compile something - why not a kernel module?
Because, empirically, kernel modules seem to end up being very dependent on kernel versions; if they weren't, distributions like Debian wouldn't ship with different collections of most kernel modules for each kernel, they would ship with one kernel module per package for each function/driver, without much of a notion of a "kernel version".
Another reason is that one bug in one kernel modules brings down the whole thing. That's unnecessary and makes driver development a huge pain.
Not that OS X is actually a microkernel OS!
I made no claims about what it is or even whether it is a good architecture. What I claimed was that Jobs correctly identified a problem and tried to address it as best as possible with the software available at the time.
And I think he actually succeeded much more than Linux did in this particular regard: kernel extensions on OS X work much better than on Linux.
(Jobs also correctly identified the problem with C/C++ GUI toolkits and his solution, Objective-C with DisplayPostscript, probably also was the best technical compromise at the time, but I think that choice hasn't turned out as well as his choice for kernel--OpenStep and Cocoa ended up with most of the same problems as other GUI toolkits.)
Bluetooth is another choice: tiny USB dongles and standardized. Up to 1Mbps. The software isn't quite as mature yet, but it's getting there.
Accept something for free if it is clear that the giver has no commercial interest in giving it to you or if the relationship is such that the giver can't exercise control over you or the product later on. Otherwise, be very suspicious and try to avoid the "gift" if you can.
I think Jobs had the right idea when he picked Mach as the basis for NeXTStep: he wanted a kernel that looked like UNIX from the outside but that was much more componentized than the UNIX kernels of the time, or BSD/Linux today. I don't know whether Mach/Darwin is the best choice for that, but in general, I think it's where open source needs to go.
After all, we don't recompile Bash or dynamically load libraries into Bash every time someone comes out with a new command line program. We shouldn't have to do that either for a new file system type, networking protocol, or driver. And expending much time on a BSD/Linux rivalry isn't going to address such issues.
Energy (and water and food) consumption at US levels is not a "basic necessity", it is an absurd and completely unnecessary waste. Apart from any arguments about growth, it is just obscene.
Now I don't get your argument. You say that a few people dying from radiation is such a catastrophe that we need to go massively overboard on our safety standards. On the other hand, millions dead in food riots doesn't seem to phase you.
No, I'm saying that your reasoning is sensible, simple, and, unfortunately, completely wrong. If only things were as simple as you think they are: people are hungry, so you produce more food to feed them, problem solved.
But populations and consumption adjust to resources: we create more available and they grow. You end up facing a worse problem when you are done than when you started. What needs to be fixed is the influence of available resources on growth and consumption.
A good start towards that goal is to stop growing. Not only does that stop the problem from getting bigger and bigger, it also takes away the illusion that if we just grow a little more, our problems will be solved. Once that illusion is gone, people can perhaps start thinking about making the necessary hard choices.
It's like highways: building more doesn't reduce congestion, it actually makes it worse.
Oh, this will happen without taxation. Scarcity will increase the price.
Unfortunately, our capacity to consume energy and produce carbon dioxide or nuclear waste is not limited by available resources. There is probably enough carbon in the ground to increase atmospheric carbon dioxide by orders of magnitude, and we can mine and breed nearly unlimited amounts of nuclear fuel. Sooner or later, we either die out as a species or we have to decide to stop increasing consumption. We might as well face the facts now--it only gets worse the longer we go on.
Am I being trolled here?
Not by me, but you probably are being trolled by your favorite politician and 1960's style economist.
Large segments of the population always "go without" what the wealthy enjoy. Large segments of the US population "go without" retirement benefits, job security, privacy, decent health care, extended vacations, and a lot of other things. Those are a lot more important than the ability to cruise around in a gas guzzling SUV, compared to a small, efficient vehicle. If we wanted to reduce those inqualities, that would be very easy: through high taxes on the top few percent of earners, just like income tax was intended to do initially.
Energy prices should be gradually but steadily increased, through taxation on both non-renewable fuels and inefficient consumption. Society would adapt to this by reducing energy usage. That's pretty much the only way, and it works.
Your argumentation is like that of a drug addict who says "I know it's bad for me, but stopping is unpleasant". Your solution to one dangerous drug (oil) is to switch to another dangerous one (nuclear energy), which just happens to not make you look quite as bad.
We (western nations) happen to be the ones exporting food to the world. If we stopped or raised prices significantly, many third world countries would no longer be able to afford enough to keep their large (and growing) populations fed.
Yup, I have to admit it's a neat package. We destroy indigenous cultures and farming methods, effectively encourage population growth, interfere with family planning, and cause other nations to become dependent on our food and indebted to us. The latest twist is to make farmers dependent on even more expensive, genetically modified crops that may or may not work in our industrial farms but make no sense in other countries.
Creating large masses of poor and hungry people is great for getting powerful. The Catholic church discovered this a long time ago, but we are doing it so much better these days.
Incidentally, the other article also says that the one grown in the lab works more like that of a 60 year old--rabbit? Doesn't sound too attractive, unless you happen to be an 80 year old rabbit, I suppose.