Assume it costs $10 to make a GPU, high end or low end, should all GPUs be made high end? No.
Lets say the low end GPU can be put on a PCI card and sold for $50 at $10 profit. You can sell a lot of these because they are cheap, but your margins are low.
Now put the high end GPU on a slightly more expensive PCI card and sold for, lets say, $300. Profit is what, somewhere around $200-$250? Sure you won't sell as many, but the margins are quite impressive.
Now it's not quite this simple, but that's the general idea behind tiered product lines. A company wants a product in every price group so as to attract consumers that are willing to spend that particular amount of money. If they only made high end, they'd either loose out on customers only willing to spend $50, or they'd have to lower the price of their cards. Either way, they make less profit in the end.
It's not really a legend if it actually happened, if even for a time. The same thing has happened with the celeron line more than once as well. Overclockers have often hit spurts where a particular celeron was simple a pentium with cache disabled and underclocked a little because intel was trying to empty their warehouses without devaluing their higher end product.
I know this because I was raised with Windows and came away from it with almost nothing.
I don't think it was the computer's fault you ended up with nothing. There's a lot of ugly internals to the Windows OS, and if you came away with nothing, that's completely because you weren't trying hard enough. People don't learn "much of anything" because they are uninterested. This will happen whether you give them Windows, Linux, MacOS or OS/360: users will learn what they need to in order to do what they want to and that's it. Those who think the computer should do more than is obvious that it can do will take the time to really learn things, and they will take something away from the experience.
That said, I'd agree with the linux route for a different reason. There's so much software for linux that you can just apt-get, emerge, yum, etc. Sure, there's a lot of free software to get most things done on windows, too, but only if you know how to use google already. Esp with Ubuntu and the like: just click the boxes in aptitude, etc and software is downloaded and installed, all in one step. Linspire's Click-n-Run is super cool, too. Sure beats the Search on google, download the file, find the file, run the file, click next on all of the dialog boxes process that's required on windows.
If you preinstall an OS with Web browser, Office Products, some games, etc. you're fine. They're charity computers right? With the money you save on windows licenses you can get even more computers. That's the greater good, IMHO, even if some of the recipients are less than happy with the OS.
Shouldn't be, unless something is coded poorly. It's not open source, so you can't check it, but it does use strong encryption for the tunneling. I haven't come across any vulnerability reports for it.
RAM- you need enough RAM so you don't hit swap. For today's games, thats 1 GB. After that, the number 1 thing you can do to improve system performance is to get low latency RAM. Your CPU will be waiting for RAM, minimize the time that it is.
Well, ok, it's not trash-junk, but it's not up to the hype either. There was a review on TechRepublic a while ago that I'm pretty sure made it to slashdot (if not, then digg). Basically it showd that Low Latency RAM in itself made little to no difference and more RAM was always the way to go.
That said, Low Latency ram is not entirely a waste of money. Low Latency ram has a better shot at overclocking (like turning DDR 400 into DDR 450 by relaxing the timings and pumping the clock rate.) It's also more likely to be higher quality and thus less likely to go bad on you. As an asside, if you're interested in RAM in general, I've found this site very informative/userful.
But I would minimize the value of Low Latency in and of itself.
I use Gentoo Linux at home and either Slackware Linux or WinXP at work. Hamachi has been a really easy way to access my local network resources withing having to worry about leaving ports forwarded and I know everything is secure. If you search on the hamachi forums, you can find a nice GUI that controlls the commandline linux client.
Unfortunately, my home linux client recently stopped working and I can't get it to connect:( I haven't really spent any time to figure out what's wrong. I'm also a little concerned with the perminant 5.x.x.x ip address given to new clients--eventually they're going to start running out, if not just from people uninstalling and reinstalling without backing up their configs and from people trying it but never really using it, wasting an IP.
Other than it not connecting at home for the past couple of weeks and my concerns about the address scheme, I find it to be a really good system.
You could use Hamachi to make a secure link between two machines on which to run VNC (rather than SSH tunneling) but hamachi is not a remote control utility, just a Virtual Private Network utility.
The choice to shell out money for what's essentially VNC?
Or, what's the difference?
If you google LanDesk you'll see it's a full desktop support package, along the lines of Novell's ZenWorks product line: remote control, application deployment, desktop imaging, etc, etc, etc. VNC only fills one piece of that puzzle.
Of course, it doesn't have a good way to wrap ssh around it, but nothing can be perfect.
Simple! Just install an SSH server on your computer and create an account for them to connect to. 1) Have them download putty 2) Send them a PDF showing exactly what to configure (for the port forwarding) 3) have them connect with the username/password you created 4) Have them send the request to local host.
You could blend steps 1 and 2 togther by creating an MSI or something that pre-configures putty with a connection for your computer with the proper port forwards.
Oh wait... you wanted a good way, not just a way...
If only there were a windows vnc that bundled the ssh somehow...
I don't understand why remote X is brought up every time someone mentions VNC... VNC runs on windows, too. I'm sure his company probably has primarily (if not all) Windows machines. Remote X doesn't do so well on windows (by nature of the lack of X).
I don't have any experience with LanDesk, but I think remote management/remote control software in general isn't so bad. If it's just remote control, that really isn't any big deal and comes in quite handy if you ever do have to call them for help.
If they completely lock down the machines and take away your admin privilges, well that's life and it can be good or bad. Most often this is only a problem if need to install software and once this has been deployed for a short time and things are running more smoothly again this, too, should be relatively painless; just call or send an e-mail and someone can type in the password and install it. This kinda depends on the strength of your IT department, though. When I was in highschool the instructors machines were secured tightly and there wasn't enough staff to assist in installing software, preventing teachers from getting work done occasionaly. That was an extreme case, though (1 guy, hired as the Video Productions instructor, doing IT for the whole building...) I would expect that in your case it shouldn't be too painful.
As a disclaimer, I am an IT guy and our engineering college at the university has it's own IT group that engineering student fees pay for. I know our professors (and students) were less happy when IT was managed by the main campus group; we're more responsive and less politically hampered.
Republicans and Democrats argue about stupid things like wars, taxes, pollution and crimes against humanity. We're arguing about stuff that matters: "My computer is faster than yours"
This post's subject is misleading: - suspend2 project is not suspended - Nigel Cunningam will keep working on it - it's only inclusion in vanilla kernel (mainline) which is unclear
No, it is suspended. If it weren't suspended he would try to further the project. Instead he's going to maintain the patches at their current status --> No new features, no new hardware bug fixes. If it doesn't work for you now, it still won't so don't bother waiting.
He's going to update the patch to work with new kernels and that's it. That's suspended.
In the US the same person can't disassemble and examine as the person whose writing the new code.
If all you did was look at the file system, you're ok (but you have to prove that.) If you disassembled their windows driver, then all you can do is make a spec like the parent said.
It's the same way Compaq cloned the IBM PC Bios. They had to setup a clean room environment for the actual developers handing them only the spec written by the disassemblers.
Mod parent down for thread bare worn out lame attempt at "sarcasm". Parent has nothing to add to discussion
I don't know... I found his comment humourous and dead accurate. He said the story was stupid, and it is. Seems on topic to me... Now you're comment on the other hand...
I don't see any reason why you couldn't apply the same sort of patch to the linux client. However, I don't think the linux client supports 10-way yet, even if you have an Intel Core-Duo. Once they come out with a 10-way capable client, one should be able to apply the exact same method as what Maxxus described. The only reason I made the Windows only comment was because the actual patch has only been applied to the windows only binary thus far. It will be repeated with any linux client that supports 10-way with an Intel Only restriction if such a client is released.
Also, I'm not sure a kernel patch would work anyway. I'm not convinced the OS is ever queried, the processor is asked directly by accessing a special register. You'd have to have skype running on a fake processor like Bochs or Qemu and then it might not be fast enough for 10-way anyhow...
They admit their mistakes and *do* fix the framework and mark the old stuff as deprecated, so anyone that thinks you have to run frameworks side-by-side is smoking something.
I guess that leads to the questions "Why are both 1.1 and 2.0 in Windows update as installable options? Why doesn't the installation of 2.0 trump the installation of 1.1? (Why does it let me install both?) And why have I seen applications that require 1.1 installed, and not 2.0?" For example, BizTalk server 2004does not run with the.Net 2.0.
I've never seen a Java app that says "This will work on Java 1.3 BUT NOT NEWER." He a microsoft KB says this is the case with BizTalk Server 2004, and the ability to install both VMs side by side shows that they expect it to be the case with other software as well.
I guess you got lucky with your Apps in that they only did things in 1.1 that can still be done exactly the same way in 2.0. Good for you.
OSS groups have a lot of mailing lists. Are they going to have the funding to pay AOL? Probably not. Should anyone recieving the e-mail be an AOL subscriber anyway? Definately not, but there are some and the OSS group will want to reach their audience.
Slashdot Mirror
Assume it costs $10 to make a GPU, high end or low end, should all GPUs be made high end? No.
Lets say the low end GPU can be put on a PCI card and sold for $50 at $10 profit. You can sell a lot of these because they are cheap, but your margins are low.
Now put the high end GPU on a slightly more expensive PCI card and sold for, lets say, $300. Profit is what, somewhere around $200-$250? Sure you won't sell as many, but the margins are quite impressive.
Now it's not quite this simple, but that's the general idea behind tiered product lines. A company wants a product in every price group so as to attract consumers that are willing to spend that particular amount of money. If they only made high end, they'd either loose out on customers only willing to spend $50, or they'd have to lower the price of their cards. Either way, they make less profit in the end.
It's not really a legend if it actually happened, if even for a time. The same thing has happened with the celeron line more than once as well. Overclockers have often hit spurts where a particular celeron was simple a pentium with cache disabled and underclocked a little because intel was trying to empty their warehouses without devaluing their higher end product.
I know this because I was raised with Windows and came away from it with almost nothing.
I don't think it was the computer's fault you ended up with nothing. There's a lot of ugly internals to the Windows OS, and if you came away with nothing, that's completely because you weren't trying hard enough. People don't learn "much of anything" because they are uninterested. This will happen whether you give them Windows, Linux, MacOS or OS/360: users will learn what they need to in order to do what they want to and that's it. Those who think the computer should do more than is obvious that it can do will take the time to really learn things, and they will take something away from the experience.
That said, I'd agree with the linux route for a different reason. There's so much software for linux that you can just apt-get, emerge, yum, etc. Sure, there's a lot of free software to get most things done on windows, too, but only if you know how to use google already. Esp with Ubuntu and the like: just click the boxes in aptitude, etc and software is downloaded and installed, all in one step. Linspire's Click-n-Run is super cool, too. Sure beats the Search on google, download the file, find the file, run the file, click next on all of the dialog boxes process that's required on windows.
If you preinstall an OS with Web browser, Office Products, some games, etc. you're fine. They're charity computers right? With the money you save on windows licenses you can get even more computers. That's the greater good, IMHO, even if some of the recipients are less than happy with the OS.
No way, man. These are moon kids. No computers on the moon, man.
russians != chinese
2006 != 1998
nuff said?
That comment was about as obligatory as getting drunk and beating your wife.
"But Judge! I couldn't help myself!"
Shouldn't be, unless something is coded poorly. It's not open source, so you can't check it, but it does use strong encryption for the tunneling. I haven't come across any vulnerability reports for it.
Seriously though, if you want to know about it, get it from the horses mouth
RAM- you need enough RAM so you don't hit swap. For today's games, thats 1 GB. After that, the number 1 thing you can do to improve system performance is to get low latency RAM. Your CPU will be waiting for RAM, minimize the time that it is.
Well, ok, it's not trash-junk, but it's not up to the hype either. There was a review on TechRepublic a while ago that I'm pretty sure made it to slashdot (if not, then digg). Basically it showd that Low Latency RAM in itself made little to no difference and more RAM was always the way to go.
That said, Low Latency ram is not entirely a waste of money. Low Latency ram has a better shot at overclocking (like turning DDR 400 into DDR 450 by relaxing the timings and pumping the clock rate.) It's also more likely to be higher quality and thus less likely to go bad on you. As an asside, if you're interested in RAM in general, I've found this site very informative/userful.
But I would minimize the value of Low Latency in and of itself.
I use Gentoo Linux at home and either Slackware Linux or WinXP at work. Hamachi has been a really easy way to access my local network resources withing having to worry about leaving ports forwarded and I know everything is secure. If you search on the hamachi forums, you can find a nice GUI that controlls the commandline linux client.
:( I haven't really spent any time to figure out what's wrong. I'm also a little concerned with the perminant 5.x.x.x ip address given to new clients--eventually they're going to start running out, if not just from people uninstalling and reinstalling without backing up their configs and from people trying it but never really using it, wasting an IP.
Unfortunately, my home linux client recently stopped working and I can't get it to connect
Other than it not connecting at home for the past couple of weeks and my concerns about the address scheme, I find it to be a really good system.
You could use Hamachi to make a secure link between two machines on which to run VNC (rather than SSH tunneling) but hamachi is not a remote control utility, just a Virtual Private Network utility.
The choice to shell out money for what's essentially VNC?
Or, what's the difference?
If you google LanDesk you'll see it's a full desktop support package, along the lines of Novell's ZenWorks product line: remote control, application deployment, desktop imaging, etc, etc, etc. VNC only fills one piece of that puzzle.
Of course, it doesn't have a good way to wrap ssh around it, but nothing can be perfect.
Simple! Just install an SSH server on your computer and create an account for them to connect to.
1) Have them download putty
2) Send them a PDF showing exactly what to configure (for the port forwarding)
3) have them connect with the username/password you created
4) Have them send the request to local host.
You could blend steps 1 and 2 togther by creating an MSI or something that pre-configures putty with a connection for your computer with the proper port forwards.
Oh wait... you wanted a good way, not just a way...
If only there were a windows vnc that bundled the ssh somehow...
I don't understand why remote X is brought up every time someone mentions VNC... VNC runs on windows, too. I'm sure his company probably has primarily (if not all) Windows machines. Remote X doesn't do so well on windows (by nature of the lack of X).
I don't have any experience with LanDesk, but I think remote management/remote control software in general isn't so bad. If it's just remote control, that really isn't any big deal and comes in quite handy if you ever do have to call them for help.
If they completely lock down the machines and take away your admin privilges, well that's life and it can be good or bad. Most often this is only a problem if need to install software and once this has been deployed for a short time and things are running more smoothly again this, too, should be relatively painless; just call or send an e-mail and someone can type in the password and install it. This kinda depends on the strength of your IT department, though. When I was in highschool the instructors machines were secured tightly and there wasn't enough staff to assist in installing software, preventing teachers from getting work done occasionaly. That was an extreme case, though (1 guy, hired as the Video Productions instructor, doing IT for the whole building...) I would expect that in your case it shouldn't be too painful.
As a disclaimer, I am an IT guy and our engineering college at the university has it's own IT group that engineering student fees pay for. I know our professors (and students) were less happy when IT was managed by the main campus group; we're more responsive and less politically hampered.
Republicans and Democrats argue about stupid things like wars, taxes, pollution and crimes against humanity. We're arguing about stuff that matters: "My computer is faster than yours"
As for point 2, AMD has yet to announce any improvement on the FPU/SSE2 part of their processors.
I wouldn't expect to see anything like that. AMD pretty much JUST added SSE2. I still don't have an AMD in any of my systems that supports SSE2...
This post's subject is misleading:
- suspend2 project is not suspended
- Nigel Cunningam will keep working on it
- it's only inclusion in vanilla kernel (mainline) which is unclear
No, it is suspended. If it weren't suspended he would try to further the project. Instead he's going to maintain the patches at their current status --> No new features, no new hardware bug fixes. If it doesn't work for you now, it still won't so don't bother waiting.
He's going to update the patch to work with new kernels and that's it. That's suspended.
Well, with AT&T/SBC buying Bell South, it's getting more likely to be the telco will win...
Looking back, I think someone might have modded just because of your signature, which really is funny ;)
In the US the same person can't disassemble and examine as the person whose writing the new code.
;)
If all you did was look at the file system, you're ok (but you have to prove that.) If you disassembled their windows driver, then all you can do is make a spec like the parent said.
It's the same way Compaq cloned the IBM PC Bios. They had to setup a clean room environment for the actual developers handing them only the spec written by the disassemblers.
But most importantly, talk to a lawyer
Mod parent down for thread bare worn out lame attempt at "sarcasm". Parent has nothing to add to discussion
I don't know... I found his comment humourous and dead accurate. He said the story was stupid, and it is. Seems on topic to me... Now you're comment on the other hand...
I don't see any reason why you couldn't apply the same sort of patch to the linux client. However, I don't think the linux client supports 10-way yet, even if you have an Intel Core-Duo. Once they come out with a 10-way capable client, one should be able to apply the exact same method as what Maxxus described. The only reason I made the Windows only comment was because the actual patch has only been applied to the windows only binary thus far. It will be repeated with any linux client that supports 10-way with an Intel Only restriction if such a client is released.
Also, I'm not sure a kernel patch would work anyway. I'm not convinced the OS is ever queried, the processor is asked directly by accessing a special register. You'd have to have skype running on a fake processor like Bochs or Qemu and then it might not be fast enough for 10-way anyhow...
They admit their mistakes and *do* fix the framework and mark the old stuff as deprecated, so anyone that thinks you have to run frameworks side-by-side is smoking something.
.Net 2.0.
I guess that leads to the questions "Why are both 1.1 and 2.0 in Windows update as installable options? Why doesn't the installation of 2.0 trump the installation of 1.1? (Why does it let me install both?) And why have I seen applications that require 1.1 installed, and not 2.0?" For example, BizTalk server 2004does not run with the
I've never seen a Java app that says "This will work on Java 1.3 BUT NOT NEWER." He a microsoft KB says this is the case with BizTalk Server 2004, and the ability to install both VMs side by side shows that they expect it to be the case with other software as well.
I guess you got lucky with your Apps in that they only did things in 1.1 that can still be done exactly the same way in 2.0. Good for you.
OSS groups have a lot of mailing lists. Are they going to have the funding to pay AOL? Probably not. Should anyone recieving the e-mail be an AOL subscriber anyway? Definately not, but there are some and the OSS group will want to reach their audience.
It runs on a Sega CD, no wonder it never took off.
It was never sold, no wonder it never took off.