The good thing about a Linux PDA is that it doesn't matter whether any particular vendor or manufacturer makes it: you can be certain that your code will continue to run on some PDA. As long as you write for X11 and the UNIX/Linux kernel APIs (and any open source libraries on top of those), you are fine. Those have been around for 15+ years and will continue to be around for many years to come. Just stay away from proprietary or rarely used libraries--those may become orphaned when the vendor hits hard times.
BTW, I hope the Yopy does make it. The Yopy looks like a nifty device, and I'm probably going to buy one.
It's kind of funny to hear the Apple marketing force extol the virtues of UNIX, when until OS9, they wouldn't have given a UNIX programmer the time of day and would have argued that the OS9 systems architecture was perfect. I also don't think OSX, while being UNIX-compatible, should be called a version of UNIX--it has almost none of the original UNIX source code, and it has little of the traditional UNIX architecture.
Still, despite these historical ironies, it's good to see Apple have an OS that is somewhat related to, and compatible with, UNIX. Apple desparately needed this, and going with a UNIX-like personality makes so much more sense than if they had cooked up another proprietary system for the core OS APIs.
Clinton, Giuliani, Bloomberg, and even Martha Stewart are rumored to only get only 4 hours of sleep on a normal night.
Well, then they have to make up for it on the weekend, or they have some serious brain disorder, and not the kind you want to have. Almost everybody needs about 8h of sleep per night. Some people need more. If you sleep less than what you need, you incur a sleep debt which you will have to repay. If the debt gets too large, you'll just keep falling asleep briefly throughout the day and not even notice (which can be rather dangerous). And if you are living with a large sleep debt, it's bad for your health.
Most Americans are already chronically sleep-deprived and suffering numerous health problems as a consequence.
One research group that has done excellent work on this and published a lot is Prof. Dement at Stanford (no, I'm not making up the name).
You are wrong: the 64k limit is intrinsic to TCP, both on the server and on the client side. When the server accepts a connection, it needs to keep track of which packets belong to which connection. Otherwise, the packets from multiple clients running on the same client machine couldn't be assigned to the correct stream. That bookkeeping is done with a 16bit number. Take a look at the protocols. If you actually complied with TCP/IP timeouts as originally specified, you can't even sustain anywhere near the connection rates that modern web servers achieve, and the shortened timeouts actually are rather bad from the point of view of robustness.
In any case, the fact that many kernels fall on their face for far smaller numbers of connections is a result of simplistic data structures (linear lists, bit arrays, etc.). Why do kernel developers choose simplistic data structures? Beats me. Perhaps it's related to the fact that implementing and reusing good data structure libraries in C is just such a pain, but it's hard to say whether that's the cause of the problem or merely the consequence of the general mindset of kernel developers. In any case, there is no point in whining about the poor abstraction in many operating system kernels--obviously, nobody else wants to do the work. As long as kernel developers fix these problems when they come up in whatever way they like, everybody is happy. And they do fix them. Several UNIX kernels changed over from lists to hash tables in their network-related data structures when they hit performance limits, so whatever is wrong in your favorite kernel can be fixed by your favorite kernel hackers as well.
Right now, there doesn't seem to be any way of getting a complete installation of GNUstep through the Debian package system (you get some libraries, but that's all as far as I can tell). I wouldn't mind giving GNUstep a try, but I'm not going to throw out my whole Linux installation. I suspect many others are in the same boat, and a complete set of easy-to-set-up GNUstep packages for Debian would probably reach more people at this point than a separate distribution.
The law was clearly intended to apply to breaking into computers, not cryptanalysis of published information. But, as worded, cryptanalysis might well fall within the scope of the law.
Still, the "having unjustly gained access to data" clause will probably save him--it seems a stretch to argue that cryptanalysis of a legally purchased medium whose decrypted contents are also available constitutes "unjustly gaining access".
The story sounds increasingly like some kind of RMS bashing (accidental pun). RMS actually made sure that GNU tools would be usable on proprietary systems like Suns. Besides, the issue isn't in RMS's hands: it's purely a question of what the correct legal interpretation of the GPL is.
Maybe if you are a twenty-something instant dot-com lottery winner and can afford private banking, you can get personal service. The rest of us have to make do with ATMs, telephone banking, and computers. Many banks now charge if you see a teller.
Besides, personal service doesn't work reliably either. I have been a victim of fraud twice, once when I was with a very reputable bank which offered personal service, and once with Internet banking. In both cases, lack of a secure authentication method was at the root of the problem.
A national ID will just be another ID for people to steal.
No, it will be the only ID for people to steal, and if the national ID is done right, it will be a lot harder to steal than what we have now.
The same fool that thought automatic executions of email attachments thought it would be a good idea to offer credit cards by mail.
Because that's what we are stuck with, we might as well make it as secure as we can.
I want a reliable way of identifying myself, to my computer, to my bank, or to an airline. That's in my own interest: it makes it harder for other people to steal from me.
The reason why I don't have anything like that and why identity theft is rampant is because people keep fighting it, under the misguided notion that it increases privacy. Fighting a national ID card doesn't increase privacy, it decreases it, because people still need to identify themselves, but there is no secure or legally protected way of doing that.
Fight for a secure national ID card and fight for legislation that limits how it can be used. That is much more valuable to your and my privacy than what you propose.
A card that you can use to prove that you are you is very useful to you and it doesn't hurt you. Furthermore, a globally unique "identifier", which should really be a collection of different digital signatures, is useful to you for all sorts of things. How else, for example, do you expect for your bank to do business with you and not run afoul of impostors? Ultimately, it comes down to biometric IDs and secrets, whether implemented by the neighborhood clerk you have known for 20 years or by a machine.
Problems arise when the "card" isn't just a card, but a set of back-end databases and records that are exchanged in non-transparent ways and that you have no control over. Problems also arise when the "cards" and ID numbers are designed and used poorly (e.g., when knowing your semi-public social security number potentially can be used to get access to your bank accounts).
The problem with using driver's licenses and all the other bogus ID documents and numbers that exist in the US is that they don't work well and are being used for things they were never designed for. Self-proclaimed civil libertarians are at fault here: we won't get any good, secure ID cards and numbers as long as any such effort is immediately torpedoed.
What we should do to protect our civil liberties is to design a robust, secure system of identification, and create privacy legislation that lets us get control of who stores what data about us. Or, in different words, the complete opposite of the agenda of the libertarians and the conservatives.
Probably. While WINE would be an excellent candidate for being GPL covered, it is not. And they don't need to hack existing Linux kernel code.
However, I suspect that they will feed improvements back into the WINE project. Otherwise, they end up having to track and integrate WINE code constantly, which would be a major hassle.
larger OS designed FOR desktop/tower/laptop computers.
These PDAs are a lot more powerful than the "desktop/tower/laptop" computers that Linux was originally designed for. KDE and Gnome may eat up oodles of memory, but Linux and X11 don't if you configure them right.
I'd rather see propertiary tight code on a upgradable eeprom
You mean like the PalmOS mess? It's tight alright, but it doesn't conform to any standards, and it's very limited. Or as "tight" as PocketPC, which manages to be slower than Linux even though it is less functional?
And along with proprietary operating systems come proprietary corporate strategies. Just look at what happened to the developers who foolishly bet on BeOS or NeXTStep-on-Intel. Even when a company is successful, like Microsoft, their changing product plans and strategies keep causing problems for developers.
Why have a Linux PDA? Because there are lots of Linux and X11 libraries out there that lets you write great handheld apps and connect well to all those corporate servers, because the development tools on Linux are great, and because Linux is a powerful operating system that runs well even on "limited" hardware like a 200MHz ARM with 32M of RAM. And, as a company or user, you don't have to worry about whether some MS strategist decides to take your OS in a direction you don't like: you have the sources and you decide how your product evolves.
This is typical Microsoft. Some organized big evil company might have orchestrated something at the highest level, used a variety of IP addresses from around the world, etc. But Microsoft doesn't have a clue--they don't rig this as part of some master plan, they just engage in some lousy mass mailing. It's like all the other areas where Microsoft wins through poor practices (bad quality control, incorrect claims of innovation, announcing and releasing before a product is ready, poor standards compliance, etc.). And while some people know the truth, the ZDNet figures are probably already being quoted in boardrooms around the world (yes, it's stupid, but that doesn't stop management from doing it). Let's hope this one will backfire for them and people will trust Microsoft-friendly statistics and reports less in the future.
Oh, God, another non-standard window system. What good is it if all these PDAs run Linux when you can't run the same GUI apps on them? Why do people keep doing this? A 200MHz ARM is 10x as fast as the high-end workstations X11 was developed for, and it runs fine on PDAs.
Well, the Creative Extigy may be nice, but it isn't exactly the first one to do this. "USB speakers" have a "sound card" built in. And companies like Tascam also make USB-based audio interfaces. The USB audio protocols are standardized, so this should work even for Linux (at least if they keep to the spec).
"The lift you get from that kind of profiling just isn't enough to pay all that extra data storage and process costs,"
Well, that's going to change. By analogy (to drag that up again), in 1981, USENET posters generally thought it would be impractical for a long time to come to put all USENET postings on the Internet. By the mid-90's, it had happened. You can bet that in the not too distant future, it will be so cheap to record and correlate all you on-line activities that no company will think twice about doing it--unless the law prevents them from doing it.
These kinds of compression claims are the perpetual motion machine of the information age. Actually, they are less plausible than perpetual motion. For perpetual motion, there is at least the (very remote) possibility that there is some kind of undiscovered physics. Impossibility statements in compression only hinge on mathematics, with no physics or experiments needed.
The major advantage of a Linux-based PDA is that it allows a huge amount of software to be ported quickly, and that it allows developers to use many different kinds of tools to develop for the PDA. What software and tools you say? All the stuff developed for the iPaq and AgendaVR3, all the stuff developed with cross-platform toolkits other than Qt, all the little WindowMaker apps, all the stuff developed for Windows (via compatibility libraries). Also, much of the X11 display code from big-screen apps can be reused on little screen devices, even if the GUI itself ends up having fewer buttons. X11-based displays also allow programmers to reuse their expertise and create new apps quickly.
Unfortunately, by using Qt/Embedded, the Zaurus partially loses that advantage: while you can muck around with VNC, in reality, the only apps any end user can run on it are Qt/Embedded apps--Qt takes over the screen. On the Zaurus, it's Qt or nothing. Software needs to be ported. Display code needs to be rewritten. Programmers who want to program the Zaurus must spend the time to learn it and use it and build new tools for it. While that isn't hard for an experienced programmer, it still is a lot of wasted time. Note that this is different from Qt on the desktop, where, through the magic of X11, Qt apps live happily side-by-side with other apps.
There is no real reason for this. If you download the Qt/Embedded and Qtopia demo and run it, you'll see that it uses up about 9Mbytes of RAM, considerably more than an X11 server and X11 apps running, say, on the AgendaVR3. And Troll Tech's own description of Qt/Embedded claims that its resource usage ("800k to 3M", depending on configuration) is comparable to that of an X11 server (which takes around 1M in a configuration suitable for handhelds).
What Sharp should do is create an X11 server for the device and recompile their Qt-based apps to use the X11 server. Then, the Sharp will be a standard Linux PDA. The way it is, the Sharp is, for practical purposes, a very slick looking but proprietary device. And that's not what Linux systems are supposed to be about.
In the medium term, I suspect "Gracenote" is going to switch to some proprietary audio fingerprinting technique, which is more reliable and robust than the current techniques used by FreeDB and the old CDDB. Furthermore, I suspect that there are numerous patents in the queue on doing this (never mind that it's pretty obvious). In the long term, of course, CDs will be about as important as wax cylinders.
You have spoken like a true anonymous coward. You make my point for me: anonymous moderation, anonymous responses, pseudonyms, group think. That's what happens when everything gets archived and scrutinized.
I want to be immortal, and writing is the best way.
Sure. That why people (including myself) publish under their own name. But USENET didn't use to be for publishing, it used to be a community forum. I'm not ashamed of anything I have ever written on USENET, I'm merely saddened that there is no forum anymore where people can both talk informally and by name.
It is our knowledge, collective and stretching backward in time. To ever lose the news archive would be a tragedy
It's already been lost. Oh, sure, you can cling to a bunch of articles from the 1980s. But what motivation is there to contribute anymore? USENET has become too big to be a community in the old sense, too much spam gets posted, and if you do participate in a discussion where you are willing to change your mind, you risk that people will find you half a century later and confront you with your every word.
I used to post messages on USENET under my own name. Some of them got picked up and republished in computer magazines. Now, I put all my technical advice on my web site, and I do all my "flaming" on Slashdot (pseudonymously). A venue where most people interact informally using their real names, where they get to know each other personally and establish reputations, doesn't exist anymore.
Slashdot Mirror
BTW, I hope the Yopy does make it. The Yopy looks like a nifty device, and I'm probably going to buy one.
Still, despite these historical ironies, it's good to see Apple have an OS that is somewhat related to, and compatible with, UNIX. Apple desparately needed this, and going with a UNIX-like personality makes so much more sense than if they had cooked up another proprietary system for the core OS APIs.
Well, then they have to make up for it on the weekend, or they have some serious brain disorder, and not the kind you want to have. Almost everybody needs about 8h of sleep per night. Some people need more. If you sleep less than what you need, you incur a sleep debt which you will have to repay. If the debt gets too large, you'll just keep falling asleep briefly throughout the day and not even notice (which can be rather dangerous). And if you are living with a large sleep debt, it's bad for your health.
Most Americans are already chronically sleep-deprived and suffering numerous health problems as a consequence.
One research group that has done excellent work on this and published a lot is Prof. Dement at Stanford (no, I'm not making up the name).
He has a guide specifically for students.
In any case, the fact that many kernels fall on their face for far smaller numbers of connections is a result of simplistic data structures (linear lists, bit arrays, etc.). Why do kernel developers choose simplistic data structures? Beats me. Perhaps it's related to the fact that implementing and reusing good data structure libraries in C is just such a pain, but it's hard to say whether that's the cause of the problem or merely the consequence of the general mindset of kernel developers. In any case, there is no point in whining about the poor abstraction in many operating system kernels--obviously, nobody else wants to do the work. As long as kernel developers fix these problems when they come up in whatever way they like, everybody is happy. And they do fix them. Several UNIX kernels changed over from lists to hash tables in their network-related data structures when they hit performance limits, so whatever is wrong in your favorite kernel can be fixed by your favorite kernel hackers as well.
Right now, there doesn't seem to be any way of getting a complete installation of GNUstep through the Debian package system (you get some libraries, but that's all as far as I can tell). I wouldn't mind giving GNUstep a try, but I'm not going to throw out my whole Linux installation. I suspect many others are in the same boat, and a complete set of easy-to-set-up GNUstep packages for Debian would probably reach more people at this point than a separate distribution.
Still, the "having unjustly gained access to data" clause will probably save him--it seems a stretch to argue that cryptanalysis of a legally purchased medium whose decrypted contents are also available constitutes "unjustly gaining access".
The story sounds increasingly like some kind of RMS bashing (accidental pun). RMS actually made sure that GNU tools would be usable on proprietary systems like Suns. Besides, the issue isn't in RMS's hands: it's purely a question of what the correct legal interpretation of the GPL is.
Maybe if you are a twenty-something instant dot-com lottery winner and can afford private banking, you can get personal service. The rest of us have to make do with ATMs, telephone banking, and computers. Many banks now charge if you see a teller.
Besides, personal service doesn't work reliably either. I have been a victim of fraud twice, once when I was with a very reputable bank which offered personal service, and once with Internet banking. In both cases, lack of a secure authentication method was at the root of the problem.
A national ID will just be another ID for people to steal.
No, it will be the only ID for people to steal, and if the national ID is done right, it will be a lot harder to steal than what we have now.
The same fool that thought automatic executions of email attachments thought it would be a good idea to offer credit cards by mail.
Because that's what we are stuck with, we might as well make it as secure as we can.
The reason why I don't have anything like that and why identity theft is rampant is because people keep fighting it, under the misguided notion that it increases privacy. Fighting a national ID card doesn't increase privacy, it decreases it, because people still need to identify themselves, but there is no secure or legally protected way of doing that.
Fight for a secure national ID card and fight for legislation that limits how it can be used. That is much more valuable to your and my privacy than what you propose.
Problems arise when the "card" isn't just a card, but a set of back-end databases and records that are exchanged in non-transparent ways and that you have no control over. Problems also arise when the "cards" and ID numbers are designed and used poorly (e.g., when knowing your semi-public social security number potentially can be used to get access to your bank accounts).
The problem with using driver's licenses and all the other bogus ID documents and numbers that exist in the US is that they don't work well and are being used for things they were never designed for. Self-proclaimed civil libertarians are at fault here: we won't get any good, secure ID cards and numbers as long as any such effort is immediately torpedoed.
What we should do to protect our civil liberties is to design a robust, secure system of identification, and create privacy legislation that lets us get control of who stores what data about us. Or, in different words, the complete opposite of the agenda of the libertarians and the conservatives.
I really doubt that; do you have any references? More likely, Apple picked tcsh because of the BSD heritage.
However, I suspect that they will feed improvements back into the WINE project. Otherwise, they end up having to track and integrate WINE code constantly, which would be a major hassle.
These PDAs are a lot more powerful than the "desktop/tower/laptop" computers that Linux was originally designed for. KDE and Gnome may eat up oodles of memory, but Linux and X11 don't if you configure them right.
I'd rather see propertiary tight code on a upgradable eeprom
You mean like the PalmOS mess? It's tight alright, but it doesn't conform to any standards, and it's very limited. Or as "tight" as PocketPC, which manages to be slower than Linux even though it is less functional?
And along with proprietary operating systems come proprietary corporate strategies. Just look at what happened to the developers who foolishly bet on BeOS or NeXTStep-on-Intel. Even when a company is successful, like Microsoft, their changing product plans and strategies keep causing problems for developers.
Why have a Linux PDA? Because there are lots of Linux and X11 libraries out there that lets you write great handheld apps and connect well to all those corporate servers, because the development tools on Linux are great, and because Linux is a powerful operating system that runs well even on "limited" hardware like a 200MHz ARM with 32M of RAM. And, as a company or user, you don't have to worry about whether some MS strategist decides to take your OS in a direction you don't like: you have the sources and you decide how your product evolves.
This is typical Microsoft. Some organized big evil company might have orchestrated something at the highest level, used a variety of IP addresses from around the world, etc. But Microsoft doesn't have a clue--they don't rig this as part of some master plan, they just engage in some lousy mass mailing. It's like all the other areas where Microsoft wins through poor practices (bad quality control, incorrect claims of innovation, announcing and releasing before a product is ready, poor standards compliance, etc.). And while some people know the truth, the ZDNet figures are probably already being quoted in boardrooms around the world (yes, it's stupid, but that doesn't stop management from doing it). Let's hope this one will backfire for them and people will trust Microsoft-friendly statistics and reports less in the future.
Oh, God, another non-standard window system. What good is it if all these PDAs run Linux when you can't run the same GUI apps on them? Why do people keep doing this? A 200MHz ARM is 10x as fast as the high-end workstations X11 was developed for, and it runs fine on PDAs.
Well, the Creative Extigy may be nice, but it isn't exactly the first one to do this. "USB speakers" have a "sound card" built in. And companies like Tascam also make USB-based audio interfaces. The USB audio protocols are standardized, so this should work even for Linux (at least if they keep to the spec).
The old work was using cold gases--kind of difficult to put into an optical computer. This is using crystals, which is more likely to be practical.
Well, that's going to change. By analogy (to drag that up again), in 1981, USENET posters generally thought it would be impractical for a long time to come to put all USENET postings on the Internet. By the mid-90's, it had happened. You can bet that in the not too distant future, it will be so cheap to record and correlate all you on-line activities that no company will think twice about doing it--unless the law prevents them from doing it.
These kinds of compression claims are the perpetual motion machine of the information age. Actually, they are less plausible than perpetual motion. For perpetual motion, there is at least the (very remote) possibility that there is some kind of undiscovered physics. Impossibility statements in compression only hinge on mathematics, with no physics or experiments needed.
Unfortunately, by using Qt/Embedded, the Zaurus partially loses that advantage: while you can muck around with VNC, in reality, the only apps any end user can run on it are Qt/Embedded apps--Qt takes over the screen. On the Zaurus, it's Qt or nothing. Software needs to be ported. Display code needs to be rewritten. Programmers who want to program the Zaurus must spend the time to learn it and use it and build new tools for it. While that isn't hard for an experienced programmer, it still is a lot of wasted time. Note that this is different from Qt on the desktop, where, through the magic of X11, Qt apps live happily side-by-side with other apps.
There is no real reason for this. If you download the Qt/Embedded and Qtopia demo and run it, you'll see that it uses up about 9Mbytes of RAM, considerably more than an X11 server and X11 apps running, say, on the AgendaVR3. And Troll Tech's own description of Qt/Embedded claims that its resource usage ("800k to 3M", depending on configuration) is comparable to that of an X11 server (which takes around 1M in a configuration suitable for handhelds).
What Sharp should do is create an X11 server for the device and recompile their Qt-based apps to use the X11 server. Then, the Sharp will be a standard Linux PDA. The way it is, the Sharp is, for practical purposes, a very slick looking but proprietary device. And that's not what Linux systems are supposed to be about.
In the medium term, I suspect "Gracenote" is going to switch to some proprietary audio fingerprinting technique, which is more reliable and robust than the current techniques used by FreeDB and the old CDDB. Furthermore, I suspect that there are numerous patents in the queue on doing this (never mind that it's pretty obvious). In the long term, of course, CDs will be about as important as wax cylinders.
You have spoken like a true anonymous coward. You make my point for me: anonymous moderation, anonymous responses, pseudonyms, group think. That's what happens when everything gets archived and scrutinized.
Sure. That why people (including myself) publish under their own name. But USENET didn't use to be for publishing, it used to be a community forum. I'm not ashamed of anything I have ever written on USENET, I'm merely saddened that there is no forum anymore where people can both talk informally and by name.
It's already been lost. Oh, sure, you can cling to a bunch of articles from the 1980s. But what motivation is there to contribute anymore? USENET has become too big to be a community in the old sense, too much spam gets posted, and if you do participate in a discussion where you are willing to change your mind, you risk that people will find you half a century later and confront you with your every word.
I used to post messages on USENET under my own name. Some of them got picked up and republished in computer magazines. Now, I put all my technical advice on my web site, and I do all my "flaming" on Slashdot (pseudonymously). A venue where most people interact informally using their real names, where they get to know each other personally and establish reputations, doesn't exist anymore.