Yes, you can. But the blackberry has only limited functionality on the client device. The device has a browser, but the more complex things (like cookies) are implemented on a middleware server. Custom applications may also be implemented on the middleware server. No retail consumer would ever be interested in this device, but its ease of use and customization-capability makes it somewhat attractive to corporations.
What a surprise it wasn't mentioned that this was patched months ago, right?
Did you not notice that a link to the security update bulletin was provided? The Slashdot posting clearly indicates that Microsoft wants users to download the update. What more did you want?
Linux distros issue security patches for their vulnerabilities weekly and nobody complains, but when Microsoft releases a patch, suddenly it's this huge issue to run a tiny executable that plugs security flaws, and then people bitch at Windows two months later when a virus comes out to exploit it...
While I won't deny that most folks look for any reason to bash Microsoft, neither this Slashdot posting nor the referenced article presented in an anti-MS tone. Nobody was complaining.
If you haven't patched after two months, you're just the same as all those people who got hit with Blaster...
One thing to keep in mind is that server patches for productive systems need to be carefully tested and downtime planned in advance. Two months for this is really not that unreasonable. I think someone else already posted that the patch for this virus (and SASSER) caused problems on some productive systems.
it's not like those of us who are actually trying to walk anywhere aren't already constantly dodging distracted pedestrians who's eyes are glued to the adverts on the bus shelters
I personally would have no guilty feelings about intentionally walking right over the top of any pedestrian who is trying to walk and watch TV at the same time, but isn't really watching where they are going. My lack of guilt would be similar with regard to those people standing still in the middle of the sidewalk and blocking everybody's path.
For misspellings in particular, he states "Misspellings end up having higher spam probabilities than the words they're intended to conceal."
I realize this doesn't quite answer your question, but I suggest you continue to train your filter.
I continued to train my filter for a while but I've nearly given up. Paul's assessment of the situation is not entirely accurate. Although a misspelled word would have a higher spam probability than correctly spelled words in common use, the RANDOM misspellings means that almost all of the "words" in the email messages are being seen by the spam filter for the first time, and therefore don't contribute towards classifying the message as spam. Although there is a finite number of ways each word may be misspelled, it's a very high number that any amount of training just can't keep up with.
Then I moved to Thunderbird, and suddenly obvious spam is regularly ending up in my inbox, despite several weeks' training.
I wonder if this could be timing. I use Mozilla Mail as my client at home, and I turned on spam filtering for my wife's email account (because she was silly and gave her email address to Publisher's Clearinghouse). After a couple weeks of training, it was catching almost all the spam, but in the last few months the spammers have been intentionally misspelling words in random ways, which reduces the effectiveness considerably. Does anybody know if SpamBayes addresses this issue?
I take from the cost difference that you probably live in/near a major city. Does your city not have some kind of commuter rail or light rail system that could reduce your drive?
Actually, about 45 miles from the city, but unfortunately the rail and bus systems do not extend very far on my side of the city. If they did, I would strongly consider public transit, even if it were a bit more costly.
It would be far better for the environment if you shortened your commute.
I agree. However, that's not always an option. These days, employers are not very loyal to their employees and it doesn't make sense to pick up and move each time you change jobs. Other reasons include two spouses who have jobs in opposite directions and have to settle on a place to live somewhere between them, quality of school districts, and the prohibitive cost to live closer to the area where the job is. In my case, it is cost-prohibitive to live closer to the job. The costs for homes near where I work are 250 percent of the one I live in now.
I have a long commute. My current vehicle which is nearing the end of its lifespan gets about 20 mpg and at the moment costs me $35 / week in gasoline. This works out to more than $150 / month. With the current situation in gasoline prices, I expect it to rise to as much as $200 / month. If I had a vehicle that achieved 40 mpg, that would give me an extra $100 monthly to save or spend. Yes, it's worth it to me to sacrifice the "fun" car for $100 per month.
Now for the not-so-selfish argument. Imagine the environmental and resource impact if everybody changed to a new car that consumed half the fuel of their old one. For the United States alone, it would mean a drastic reduction in the need for foreign oil, and I submit that it would have a dramatic impact on foreign policy -- particularly that in the middle east.
I can't wait to see what happens when they discover newsgroups.
This has already happened to one of the Philadelphia news channels, although I must say that they have no clue what newsgroups really are. They ran a special feature about Voicenet, accusing them of supporting child pornography and all kinds of things. They showed the police going into the office and seizing Voicenet assets. I was shocked when it first came on. About twenty minutes into it, I became surpised at just how idiotic the whole thing was. It was all about the "Quickvue" search tool that can basically thumbnail internet content, in particular Usenet newsgroups. Apparently, a number of people were using the tool to thumbnail some of the alt.binaries.*.erotica.* newsgroups with child pornography. The news made it sound like all of this was the fault of Voicenet, and that they were doing something sinister. When Voicenet responded that they were not really able to police the content of the newsgroups, the TV station asserted that this was ridiculous, making it sound like an easy task to monitor every single post that comes into every single of the 120,000+ newsgroups out there for banned content. Just for the record, the servers were seized in January and no charges have yet been filed against Voicenet. I think the authorities are looking for subscriber lists to go directly after people viewing the content. I'm not sure if the seizure was really legal, though.
Although this article is entertaining, the search results are meaningless. The power of Google is in the fact that I can find something I'm looking for in just a few minutes without having to leave my desk or engage in social pleasantries. To really make the results accurate, you must at least include all the factors that impact the time it takes to arrive at an answer. For calling someone on the phone, you need to add in the time it takes to figure out who to call (maybe they did that). For the Library, you need to add: drive time, time spent waiting for the online (or paper) card catalog to be available, and the time it takes to find an appropriate reference in the catalog and then go get the book or periodical you need. You can't count on people knowing exactly what source of information to use for the facts they want to know.
I believe in using the right tool for the job. If you are in the middle of something at work or at school and need to check on a fact real quick, use Google. If you are doing in-depth research on a topic, you are probably better off first going to the library because it's easier to determine the quality of your source material there. Afterwards, you can supplement with a bit of Googling and you'll probably know whether your search results are useful or pure hogwash. The phone call method? Use that if you're lonely.
You probably don't remember because when it comes to Windows, all's forgiven when things go wrong. It's expected things will go wrong, but that's alright. It's Windows.
Although I really like Linux, I have to disagree with you. Most of the time, Windows does just fine with installing new hardware. The only time when Windows is a real pain is when it fails to correctly identify and install hardware. At that point, it becomes more difficult to deal with than Linux because it does such a good job of hiding the mess it just made.
But the initial argument is still valid. Whether people are switching to Linux, or are long-time Linux users, it's unreasonable to expect them to have to fiddle around with drivers or search the Internet for HOW-TO documentation each time they want to add new hardware or when they upgrade their computer. Yes, I'm fully aware that new computers work on Windows because the hardware manufacturers specifically build and test them for Windows. But Linux certification should mean exactly the same thing for Linux. It's just crazy that a Linux machine would be "certified" and then come loaded with a distribution that doesn't fully work. And if the certified machine will only work properly with newer distributions, the certification should be specific about what distributions should be expected to work. Otherwise, it's meaningless.
I think the Linux community does as good a job as can be expected making Linux work with new hardware, especially given that most things have to be done without any guidance from the manufacturer. However, the actual procedure for installing new hardware that DOES work with Linux could be much easier. I'm certainly no fan of "plug-n-pray", so I would never suggest re-implementing that travesty. But it should be possible to generate distribution-specific install scripts or packages for devices that eliminate the "guessing game" that people have to play to get their hardware working. A perfect example is a sound card in my old computer. It's a Ensoniq Soundscape VIVO, which lists itself as Gateway (because it's a Gateway computer). I have yet to have a distribution recognize and properly configure this device correctly during the install. But the device works. I just have to run a few programs, collect some information, edit a file or two and then it works fine. An install script could easily be written to do this.
And now could SOMEBODY please tell me why Knoppix works on pretty much EVERY machine I throw at it, but full distributions seem to be unable to do this?
Windows update rarely requires more than a click on an "I accept" button, and one reboot.
Assuming that you only have one patch to install... Where I work, they create a periodic image with the OS and all the other goodies we use. When a machine has to be re-imaged, the user gets to spend the next several HOURS installing patches. This is because there are about 50 billion patches since last September, and must of them must be installed exclusively of any others, followed by a reboot. Although each patch only takes about 10 minutes or so to download and install, you have to add at least another 5 minutes per reboot (longer for the Windows 2000 machines that spend nearly ten minutes "saving your settings..."), and it really takes a LONG time to get through the process. Anybody that has to have their machine re-imaged prepares for a full day of lost productivity.
I admit that I haven't RTFA, but that sure looks like what you say they've patented.
I'm not an expert, but this definitely looks like it either provides prior art or violates the patent. The patent was filed in 2002. Any idea when QLaunch first came on the scene? That would make a big difference.
And in a spammer's case, moving over seas doesn't even involve literally moving himself / herself and family over there.
Actually, it would in this case. If a spammer is remotely operating machines overseas, they are still breaking the law by sending the unsolicited email to recipients in the United States. If they are caught in the US, they will be prosecuted here. So, they get to choose. They can either enjoy their life as a spammer and never ever set foot in the US again, or they can cease spamming. They may also choose to spam more covertly, but there are no guarantees there.
As some of our friends in Europe have already pointed out, most of the spam messages are advertising "products" available for people in the United States. While that doesn't guarantee that the money paying for the spam is coming from the US, it gives a strong indicator. Therefore, US federal laws WILL do a pretty good job to at least alter the way these people do "business." The end result remains to be seen.
The biggest challenge is tracking down and successfully prosecuting the perps. It will be interesting to see how this trial goes and whether the Feds can make the charges stick.
Your Palm power button is safe (not on the list), as are the buttons on your wristwatch, Mac mouse, car radio and garden tractor (totally different field of application).
To add to this -- contrary to popular belief, this patent has nothing to do with operations performed by a stylus or mouse. It is carefully restricted to focus on HARDWARE-based buttons of a PDA. Anybody who has a Pocket PC should know what I'm talking about. A similar frame of reference for the rest of you: Most everybody has seen (or owns) one of those "Internet" keyboards that has separate buttons for Email, Search, Browse and Play. These are similar to the application buttons on a Pocket PC. What Microsoft has patented is a method to detect how long these buttons are held down and respond in different ways, depending on the length of the button press. They've carefully limited the scope of their patent to only these "application buttons."
Now, you MIGHT be able to claim prior art with a calculator watch (aka "limited-resource computing device") that did different things depending on how long the buttons on the side were pressed. But I doubt it.
It does however make it difficult to collect damages in case of infringment since the defendant can claim that he was unaware that the images were copyrighted.
Now that really doesn't make sense. If the Berne Convention makes copyright notices unnecessary, then it should be assumed that any image you find is copyrighted unless other notice is given. I don't think you can use cluelessness as a defence.
Definately not a good thing for me. I am currently thinking of learning some programming languages to maybe start in that field as i have a growing interest in that.
I really don't want to discourage you, but I honestly can't recommend just picking up a few programming languages and hoping to net a job with them in today's market. There are already enough people with Computer Science degrees and experience that can't find decent jobs, and unless you really luck out and land in an area that is just DYING to get some programmers, you probably won't have any better success.
What I CAN recommend is that you look at some of the local newspapers in the area to which you are moving as well as the internet job boards (like Monster) to see what kind of employement has the highest availability and compensation, and then consider retraining yourself for that field. As always, however, make sure it is something you think you'll enjoying doing every weekday for 8+ hours. I can also give you a little reassurance -- salaries are in general lower in the south, but so is cost of living. You just need to make sure any cut you take in pay is compensated by cuts in living expenses.
This is entirely off-topic, so any moderator should feel free to mod the thread down....
Last time I ran Linux on the desktop, which was Redhate 7.1, it usually didn't make it for 'days'. I had plenty of kernel panics.
I'm not trying to brush the issue aside, but this really does sound like some sort of hardware problem. I ran Redhat 7.1 for over 2 years with no kernel panics at all. In fact, I've never had a kernel panic on Redhat. Maybe you are doing some other task on your machine that I never do. Anyway, my uptime was generally measured in months, and downtime was only due to power failures because I have no UPS.
Windows XP is intended as a workstation, not a server, and thus an uptime measured in days is sufficient.
While I agree that a workstation doesn't need the uptime of a server, my big concern is that even a workstation shouldn't be crashing. If uptime can only be measured in days (and I actually think it's probably better than that -- weeks), then there is some event (or set of them) that periodically causes the crash, making the machine unreliable. It's a good practice to hit the save button frequently, but you shouldn't have to do so just because you never know when the machine will crash.
Actually I haven't managed to crash or bluescreen my windows XP system in a long time, but what DOES happen after a few days is that explorer.exe (the one providing the task bar, system tray, etc) will die.
My laptop is running Windows 2000, but I have similar experiences. Actually, my biggest problem stems from one of the security patches that was installed a month or so ago. Right after the patch, I started having trouble with my external mouse and keyboard. If I powered up from a hibernate state and connected the mouse and keyboard, it would never recognize them. After that, it also wouldn't perform a shut down either, forcing the need to power the thing off. Not good. I know for a fact now that it was the security patch because I had to get a new laptop just last week, and the image that was placed on it was a Windows 2000 image from late last year. When the security update patches started getting applied, all of them caused no problems until a critical one applied itself while I was at lunch. That was when the problem started up again. This past weekend, I applied SP4, so I'll test today to see if the problem goes away.
In this respect Linux is a superior desktop system, because if X gets munged, I can just ctrl-alt-backspace and X will die, then respawn. Windows really could use a method of completely restarting the GUI without rebooting.
Agreed. This is one of the things that makes me like Linux. Although there are speed advantages to running things in kernel space, there are definitely some bigger advantages to separating out the GUI like this.
Barring that, it could use a system tray that isn't really fucking stupid.
And make the registry less prone to corruption, too. Or get rid of it.
I agree that there weren't many reasons, but there was one: ClearType. That's pretty much the only thing in Windows XP that interests me, thus the primary reason why I'm still on Windows 2000.
Slashdot Mirror
Yes, you can. But the blackberry has only limited functionality on the client device. The device has a browser, but the more complex things (like cookies) are implemented on a middleware server. Custom applications may also be implemented on the middleware server. No retail consumer would ever be interested in this device, but its ease of use and customization-capability makes it somewhat attractive to corporations.
Did you not notice that a link to the security update bulletin was provided? The Slashdot posting clearly indicates that Microsoft wants users to download the update. What more did you want?
Linux distros issue security patches for their vulnerabilities weekly and nobody complains, but when Microsoft releases a patch, suddenly it's this huge issue to run a tiny executable that plugs security flaws, and then people bitch at Windows two months later when a virus comes out to exploit it...
While I won't deny that most folks look for any reason to bash Microsoft, neither this Slashdot posting nor the referenced article presented in an anti-MS tone. Nobody was complaining.
If you haven't patched after two months, you're just the same as all those people who got hit with Blaster...
One thing to keep in mind is that server patches for productive systems need to be carefully tested and downtime planned in advance. Two months for this is really not that unreasonable. I think someone else already posted that the patch for this virus (and SASSER) caused problems on some productive systems.
I personally would have no guilty feelings about intentionally walking right over the top of any pedestrian who is trying to walk and watch TV at the same time, but isn't really watching where they are going. My lack of guilt would be similar with regard to those people standing still in the middle of the sidewalk and blocking everybody's path.
I realize this doesn't quite answer your question, but I suggest you continue to train your filter.
I continued to train my filter for a while but I've nearly given up. Paul's assessment of the situation is not entirely accurate. Although a misspelled word would have a higher spam probability than correctly spelled words in common use, the RANDOM misspellings means that almost all of the "words" in the email messages are being seen by the spam filter for the first time, and therefore don't contribute towards classifying the message as spam. Although there is a finite number of ways each word may be misspelled, it's a very high number that any amount of training just can't keep up with.
I wonder if this could be timing. I use Mozilla Mail as my client at home, and I turned on spam filtering for my wife's email account (because she was silly and gave her email address to Publisher's Clearinghouse). After a couple weeks of training, it was catching almost all the spam, but in the last few months the spammers have been intentionally misspelling words in random ways, which reduces the effectiveness considerably. Does anybody know if SpamBayes addresses this issue?
If he did, that might also ensure a trip to Club Fed.
Actually, about 45 miles from the city, but unfortunately the rail and bus systems do not extend very far on my side of the city. If they did, I would strongly consider public transit, even if it were a bit more costly.
I agree. However, that's not always an option. These days, employers are not very loyal to their employees and it doesn't make sense to pick up and move each time you change jobs. Other reasons include two spouses who have jobs in opposite directions and have to settle on a place to live somewhere between them, quality of school districts, and the prohibitive cost to live closer to the area where the job is. In my case, it is cost-prohibitive to live closer to the job. The costs for homes near where I work are 250 percent of the one I live in now.
I have a long commute. My current vehicle which is nearing the end of its lifespan gets about 20 mpg and at the moment costs me $35 / week in gasoline. This works out to more than $150 / month. With the current situation in gasoline prices, I expect it to rise to as much as $200 / month. If I had a vehicle that achieved 40 mpg, that would give me an extra $100 monthly to save or spend. Yes, it's worth it to me to sacrifice the "fun" car for $100 per month.
Now for the not-so-selfish argument. Imagine the environmental and resource impact if everybody changed to a new car that consumed half the fuel of their old one. For the United States alone, it would mean a drastic reduction in the need for foreign oil, and I submit that it would have a dramatic impact on foreign policy -- particularly that in the middle east.
Thanks! I've been looking for that for a couple months! I couldn't remember which station it was.
This has already happened to one of the Philadelphia news channels, although I must say that they have no clue what newsgroups really are. They ran a special feature about Voicenet, accusing them of supporting child pornography and all kinds of things. They showed the police going into the office and seizing Voicenet assets. I was shocked when it first came on. About twenty minutes into it, I became surpised at just how idiotic the whole thing was. It was all about the "Quickvue" search tool that can basically thumbnail internet content, in particular Usenet newsgroups. Apparently, a number of people were using the tool to thumbnail some of the alt.binaries.*.erotica.* newsgroups with child pornography. The news made it sound like all of this was the fault of Voicenet, and that they were doing something sinister. When Voicenet responded that they were not really able to police the content of the newsgroups, the TV station asserted that this was ridiculous, making it sound like an easy task to monitor every single post that comes into every single of the 120,000+ newsgroups out there for banned content. Just for the record, the servers were seized in January and no charges have yet been filed against Voicenet. I think the authorities are looking for subscriber lists to go directly after people viewing the content. I'm not sure if the seizure was really legal, though.
I believe in using the right tool for the job. If you are in the middle of something at work or at school and need to check on a fact real quick, use Google. If you are doing in-depth research on a topic, you are probably better off first going to the library because it's easier to determine the quality of your source material there. Afterwards, you can supplement with a bit of Googling and you'll probably know whether your search results are useful or pure hogwash. The phone call method? Use that if you're lonely.
This might come in useful during the winter, but the reduced battery life could leave you with cold feet...
Although I really like Linux, I have to disagree with you. Most of the time, Windows does just fine with installing new hardware. The only time when Windows is a real pain is when it fails to correctly identify and install hardware. At that point, it becomes more difficult to deal with than Linux because it does such a good job of hiding the mess it just made.
But the initial argument is still valid. Whether people are switching to Linux, or are long-time Linux users, it's unreasonable to expect them to have to fiddle around with drivers or search the Internet for HOW-TO documentation each time they want to add new hardware or when they upgrade their computer. Yes, I'm fully aware that new computers work on Windows because the hardware manufacturers specifically build and test them for Windows. But Linux certification should mean exactly the same thing for Linux. It's just crazy that a Linux machine would be "certified" and then come loaded with a distribution that doesn't fully work. And if the certified machine will only work properly with newer distributions, the certification should be specific about what distributions should be expected to work. Otherwise, it's meaningless.
I think the Linux community does as good a job as can be expected making Linux work with new hardware, especially given that most things have to be done without any guidance from the manufacturer. However, the actual procedure for installing new hardware that DOES work with Linux could be much easier. I'm certainly no fan of "plug-n-pray", so I would never suggest re-implementing that travesty. But it should be possible to generate distribution-specific install scripts or packages for devices that eliminate the "guessing game" that people have to play to get their hardware working. A perfect example is a sound card in my old computer. It's a Ensoniq Soundscape VIVO, which lists itself as Gateway (because it's a Gateway computer). I have yet to have a distribution recognize and properly configure this device correctly during the install. But the device works. I just have to run a few programs, collect some information, edit a file or two and then it works fine. An install script could easily be written to do this.
And now could SOMEBODY please tell me why Knoppix works on pretty much EVERY machine I throw at it, but full distributions seem to be unable to do this?
Assuming that you only have one patch to install... Where I work, they create a periodic image with the OS and all the other goodies we use. When a machine has to be re-imaged, the user gets to spend the next several HOURS installing patches. This is because there are about 50 billion patches since last September, and must of them must be installed exclusively of any others, followed by a reboot. Although each patch only takes about 10 minutes or so to download and install, you have to add at least another 5 minutes per reboot (longer for the Windows 2000 machines that spend nearly ten minutes "saving your settings..."), and it really takes a LONG time to get through the process. Anybody that has to have their machine re-imaged prepares for a full day of lost productivity.
I'm not an expert, but this definitely looks like it either provides prior art or violates the patent. The patent was filed in 2002. Any idea when QLaunch first came on the scene? That would make a big difference.
Agreed. Good point.
Actually, it would in this case. If a spammer is remotely operating machines overseas, they are still breaking the law by sending the unsolicited email to recipients in the United States. If they are caught in the US, they will be prosecuted here. So, they get to choose. They can either enjoy their life as a spammer and never ever set foot in the US again, or they can cease spamming. They may also choose to spam more covertly, but there are no guarantees there.
As some of our friends in Europe have already pointed out, most of the spam messages are advertising "products" available for people in the United States. While that doesn't guarantee that the money paying for the spam is coming from the US, it gives a strong indicator. Therefore, US federal laws WILL do a pretty good job to at least alter the way these people do "business." The end result remains to be seen.
The biggest challenge is tracking down and successfully prosecuting the perps. It will be interesting to see how this trial goes and whether the Feds can make the charges stick.
To add to this -- contrary to popular belief, this patent has nothing to do with operations performed by a stylus or mouse. It is carefully restricted to focus on HARDWARE-based buttons of a PDA. Anybody who has a Pocket PC should know what I'm talking about. A similar frame of reference for the rest of you: Most everybody has seen (or owns) one of those "Internet" keyboards that has separate buttons for Email, Search, Browse and Play. These are similar to the application buttons on a Pocket PC. What Microsoft has patented is a method to detect how long these buttons are held down and respond in different ways, depending on the length of the button press. They've carefully limited the scope of their patent to only these "application buttons."
Now, you MIGHT be able to claim prior art with a calculator watch (aka "limited-resource computing device") that did different things depending on how long the buttons on the side were pressed. But I doubt it.
Now that really doesn't make sense. If the Berne Convention makes copyright notices unnecessary, then it should be assumed that any image you find is copyrighted unless other notice is given. I don't think you can use cluelessness as a defence.
Yes, this fits well into my theory. Sometime before November 2nd, we will have one or more of the following:
Anybody agree with my theory?
I really don't want to discourage you, but I honestly can't recommend just picking up a few programming languages and hoping to net a job with them in today's market. There are already enough people with Computer Science degrees and experience that can't find decent jobs, and unless you really luck out and land in an area that is just DYING to get some programmers, you probably won't have any better success.
What I CAN recommend is that you look at some of the local newspapers in the area to which you are moving as well as the internet job boards (like Monster) to see what kind of employement has the highest availability and compensation, and then consider retraining yourself for that field. As always, however, make sure it is something you think you'll enjoying doing every weekday for 8+ hours. I can also give you a little reassurance -- salaries are in general lower in the south, but so is cost of living. You just need to make sure any cut you take in pay is compensated by cuts in living expenses.
Last time I ran Linux on the desktop, which was Redhate 7.1, it usually didn't make it for 'days'. I had plenty of kernel panics.
I'm not trying to brush the issue aside, but this really does sound like some sort of hardware problem. I ran Redhat 7.1 for over 2 years with no kernel panics at all. In fact, I've never had a kernel panic on Redhat. Maybe you are doing some other task on your machine that I never do. Anyway, my uptime was generally measured in months, and downtime was only due to power failures because I have no UPS.
Windows XP is intended as a workstation, not a server, and thus an uptime measured in days is sufficient.
While I agree that a workstation doesn't need the uptime of a server, my big concern is that even a workstation shouldn't be crashing. If uptime can only be measured in days (and I actually think it's probably better than that -- weeks), then there is some event (or set of them) that periodically causes the crash, making the machine unreliable. It's a good practice to hit the save button frequently, but you shouldn't have to do so just because you never know when the machine will crash.
Actually I haven't managed to crash or bluescreen my windows XP system in a long time, but what DOES happen after a few days is that explorer.exe (the one providing the task bar, system tray, etc) will die.
My laptop is running Windows 2000, but I have similar experiences. Actually, my biggest problem stems from one of the security patches that was installed a month or so ago. Right after the patch, I started having trouble with my external mouse and keyboard. If I powered up from a hibernate state and connected the mouse and keyboard, it would never recognize them. After that, it also wouldn't perform a shut down either, forcing the need to power the thing off. Not good. I know for a fact now that it was the security patch because I had to get a new laptop just last week, and the image that was placed on it was a Windows 2000 image from late last year. When the security update patches started getting applied, all of them caused no problems until a critical one applied itself while I was at lunch. That was when the problem started up again. This past weekend, I applied SP4, so I'll test today to see if the problem goes away.
In this respect Linux is a superior desktop system, because if X gets munged, I can just ctrl-alt-backspace and X will die, then respawn. Windows really could use a method of completely restarting the GUI without rebooting.
Agreed. This is one of the things that makes me like Linux. Although there are speed advantages to running things in kernel space, there are definitely some bigger advantages to separating out the GUI like this.
Barring that, it could use a system tray that isn't really fucking stupid.
And make the registry less prone to corruption, too. Or get rid of it.
I agree that there weren't many reasons, but there was one: ClearType. That's pretty much the only thing in Windows XP that interests me, thus the primary reason why I'm still on Windows 2000.
You do know that a statement like this is just bait, right?