If you are competent enough to build your own system, and you have the time, it may be cheaper to build one yourself. Don't forget to factor in your labor costs though. Obviously, you'll have labor costs either way, installing a new server. But when buying from parts, you've got to spend time picking out individual pieces and ordering them from different places. Then you get to make them all work together.;)
I just ordered my own system. I got a motherboard with an Intel 815E chipset, which comes with built-in video, sound, and sometimes even 10/100 Ethernet. All I had to buy was motherboard, CPU, RAM, HD, floppy, CD-ROM, and case.
I got a great deal on a rackmount case. Only $150 + $30 shipping at TechStore. It's got 3 5.25" and 2 3.5" external bays plus 1 internal 3.5" bay. Most places want to charge you at least $300 for a rackmount case. This case is really quite nice too. I got most of my other components from Transcend PC.
I think that people fail to realize that ipchains does not need to be put on a separate firewall box. You can run it on any box, including a workstation. The only reason that it normally runs on a single box is so that it can protect a whole network with all the configuration done at one location.
I think things like GLADE are the future. You describe what the UI is supposed to look like. Then you add the functionality using a different language.
Actually, with languages like Haskell and Prolog, you can describe the result that you want, and give the system enough rules to figure out how to implement it. These languages are more descriptive than imperative. They have a lot of power, but I don't think they'll ever catch on much.
The AmigaDOS command-line language implemented something very close to this, perhaps even better than the way you describe. All the following commands are equivalent:
COPY src dest
COPY TO dest src
COPY FROM src TO dest
COPY TO dest FROM src
The command has a default argument order, but also has keywords that can be used to rearrange the order. I really wish a couple UNIX commands had this option. Like grep -- I can never remember if the search string goes at the beginning or at the end.
When creating a cluster, you want to have a clear goal in mind. A cluster combines the resources of several machines, making them appear as one machine. What resources do you want to combine? Do the machines have to appear as a single machine to reach that goal?
Beowulf would be a good solution if you want to combine computational power, but your programs must be written for a parallel processing API.
Linux Virtual Servers is a decent solution if you want to implement network services such as a web server. There are commercial products that can do this as well. (Alteon boxes do it in hardware, TurboLinux has a software product.)
If you're trying to combine the systems into one big server that you can log into, I don't think that there is anything that can do that.
Don't make the mistake that Mozilla made. They released it as Open Source when it was not a viable working program. That's one reason why it has been making slow progress. It's hard to get people interested in something that doesn't work.
As far as waiting until later, that's probably not the best idea either. As long as the program is good and useful, you might as well get the extra hype by making it Open Source right away, to get more users.
What is Basilisk II? Was that the Mac emulator that ran on the Amiga?
Individuals have values; groups do not.
on
Selfish Society
·
· Score: 1
Morals and values are decided by individuals, not groups. The "values of society" is a misnomer. It is what the majority believes. That doesn't mean that I have to believe it. Democracy is supposed to protect the rights and beliefs of the minority.
The point is that there is no "common value system" for geeks. If anything, geeks are more likely to think for themselves and come up with their own beliefs. Don't try to pigeon-hole us all into one demographic.
I think you may have cause and effect mixed up here. Just because intersections with traffic lights have more accidents does not mean that traffic lights cause more accidents. More likely, the fact that more accidents happen with traffic lights is because traffic lights are installed at intersections with more traffic.
The round-about circle intersection works pretty well if you know how to use it. But most of the people in my part of the world all want to use the outside lane.
Well, the first task is to get it looking at least as good as the competition. Then add features to make it better than the competition. Then we can work on the UI. Unless there are some UI experts, the GNOME guys tend to concentrate on functionality and then usability. The fact is, Outlook's layout works well. Changes to something that works that well should be minimal unless we're sure that they will improve things.
I can't recall too many performance increases smaller than this, percentage-wise. There were smaller increments between 60/66, 90/100, and 120/133, but I'm pretty sure that the faster models came out at the same time as the slower ones.
A 11% increase doesn't sound that exciting to me. Then again, a 113MHz increase doesn't sound so bad.
Everyone keeps referring to the site at msstate.edu as being the University of Missouri. Check your state abbreviations, people. That would be Mississippi State.
You would think Jobs would have learned his lesson
on
Apple Cube Confirmed
·
· Score: 2
Jobs already made the mistake of releasing a cube-shaped computer once in his career. The original NeXT computer was a black cube, 12 inches on a side. You don't see too many of them around any more. Of course part of the reason could be that the only removable storage they had was a 256MB optical disk, which were really expensive.
And I'd be willing to bet that the Cobalt Raq systems sell better than the Cobalt Qubes.
Here are the specs of the NeXT cube, according to "The NeXT Book", by Bruce Webster, first edition, 1989. 68030 CPU at 25MHz, 68882 FPU, 56001 DSP, 8MB RAM, expandable to 64MB (16 SIMM slots!), 330MB or 660MB 5.25" hard drive, 256MB optical drive (the only removable storage!), built-in SCSI, built-in Ethernet, 1120x832 gray-scale (2 bit-plane) 17" display, 256K video buffer, passive backplane with 4 slots (similar to NuBus, but the cards are about 11"x11"; the main board takes up one of the slots), nice keyboard (includes power, volume, and brightness control), decent mouse, black cast magnesium case.
Jobs already made the mistake of releasing a cube-shaped computer once in his career. The original NeXT computer was a black cube, 12 inches on a side. You don't see too many of them around any more.
And I'd be willing to bet that the Cobalt Raq systems sell better than the Cobalt Qubes.
Here are the specs of the NeXT cube, according to "The NeXT Book", by Bruce Webster, first edition, 1989. 68030 CPU at 25MHz, 68882 FPU, 56001 DSP, 8MB RAM, expandable to 64MB (16 SIMM slots!), 330MB or 660MB 5.25" hard drive, 256MB optical drive (the only removable storage!), built-in SCSI, built-in Ethernet, 1120x832 gray-scale (2 bit-plane) 17" display, 256K video buffer, passive backplane with 4 slots (similar to NuBus, but the cards are about 11"x11"; the main board takes up one of the slots), nice keyboard (includes power, volume, and brightness control), decent mouse, black cast magnesium case.
The hardware was definitely impressive for its day (1989) but a bit pricey. Actually, much of it was ahead of its time. Nobody had 17" monitors back then; nobody had that kind of resolution. And the optical drive was a flop. The color slabs were the really nice boxes. They came out a year or so later than the cube.
The really impressive part of the NeXT was the software though. You can see its legacy in a lot of some Linux GUI software, and the Mac OS X is based on NeXTStep. One of the best things about NeXTStep is its use of Display PostScript. WSYIWYG, because the display and the printer are written to in exactly the same way.
Other posts have pointed out that it is cheaper to build telescopes on the earth, but orbital telescopes eliminate problems with the atmosphere causing interference. Might placing a telescope on the moon be more cost-effective than an orbiting telescope, while solving some of the problems with orbiting telescopes? Or would it end up being the worst of both worlds instead of the best of both worlds?
I respect the work of the developers, but I don't see anything on the web site backing up the claim that the format is completely free of patents. Have they done a patent search?
The company selling the code can't retain all rights, or else you wouldn't receive anything from them. The contract clearly had to indicate that there was some type of deliverable. The delivery of that product or service implicitly has to be usable by the recipient. In other words, even if they retain a non-exclusive right to the work they create, they cannot exclude you from using what they sold you. There is no way that they can win, unless the company never paid them at all for their partial work.
We use Maxspeed stations at work. They're not too bad. They turn on instantly, and you can run at least 4 off a decent Linux system without problems. You just install a proprietary serial card in the server and hang the clients off the server. Then install Red Hat on the server and add the Maxspeed package on top of it. One problem is that the server needs to be in the same room as the clients pretty much. I'm guessing the boxes cost about $400 apiece. (No prices on the web page.) You can just add an old 15" SVGA monitor for another $50 or less.
I see a lot of headlines about there being too many tech jobs and not enough qualified candidates to fill them. The real problem is that companies think they can find the perfect employee and not have to train them at all. This is ludicrous. If an employee wanted to keep doing the same thing that she did in her last job, she would have stayed there.
So even if you're more capable at programming in Java than some people with 2 years experience with it, some companies are too dumb to figure that out. On the bright side, you wouldn't want to work for a company that dumb anyway. Unfortunately, there are a lot of dumb companies out there, making it difficult to get a job sometimes.
The best jobs I've gotten are ones that challenge me to learn new things. I think I do better at those jobs too, so I don't understand why companies just don't get it.
Don't try to save money when buying a carrying case for your notebook. I tried 2 cheap cases for my notebook and they both broke after about 2 trips out of town. I ended up with a Targus and wish I had bought it in the first place. Whatever you buy, make sure the carrying strap is connected to the bag really really well.
A fine time to be oppressed.
on
CFP 2000 Wrapup
·
· Score: 1
We live in a fine time to be an oppressed employee, at least relative to, say, 1880, 1920 or 1955.
I got a chuckle out of that. But I would have to agree with it 100%.
I've been saying the same thing about poor people in this country. What do I have that they don't? Not all that much. Which leads me to the conclusion that they really aren't poor, but they don't realize it.
And it's the same with the middle class. They don't realize that they can take time to just enjoy and relax. That's what the Industrial Revolution, the Consumer Revolution, and the Techology Revolution have given us, but we've failed to realize how to take advantage of it.
If I were paranoid, I'd say that someone is keeping us from making that realization. If so, I'd guess that it is corporations. They need us to continue working for them and buying their products that we don't really need.
Stephenson continues to impress.
on
CFP 2000 Wrapup
·
· Score: 3
Just the summary of Stephenson's speech touches on so many things. I find it very interesting that he contrasts his ideas about threats to privacy with the Big Brother model. I think there's a good chance he will end up writing a work that will be seen as the "1984" of the 21st century. And I think he's already hit the nail on the head.
I think that corporatism will have a large part to play in this domination system. That seems to be where the biggest threat is coming from already. I've already seen a book or two on how opression has moved from the racial minorities to the working class in general, and how employers use various methods to keep workers from becoming too powerful.
Regarding corporatism, I find it rather sad -- I see so many workers with good ideas that work in corporate cultures where their ideas have no way to be expressed. If the Open Source community can somehow get their merit-based methods to be applied to the business world, I'll have a much more optimistic outlook on the future of this country (and ones like it).
Slashdot Mirror
If you are competent enough to build your own system, and you have the time, it may be cheaper to build one yourself. Don't forget to factor in your labor costs though. Obviously, you'll have labor costs either way, installing a new server. But when buying from parts, you've got to spend time picking out individual pieces and ordering them from different places. Then you get to make them all work together. ;)
I just ordered my own system. I got a motherboard with an Intel 815E chipset, which comes with built-in video, sound, and sometimes even 10/100 Ethernet. All I had to buy was motherboard, CPU, RAM, HD, floppy, CD-ROM, and case.
I got a great deal on a rackmount case. Only $150 + $30 shipping at TechStore. It's got 3 5.25" and 2 3.5" external bays plus 1 internal 3.5" bay. Most places want to charge you at least $300 for a rackmount case. This case is really quite nice too. I got most of my other components from Transcend PC.
I think that people fail to realize that ipchains does not need to be put on a separate firewall box. You can run it on any box, including a workstation. The only reason that it normally runs on a single box is so that it can protect a whole network with all the configuration done at one location.
I think things like GLADE are the future. You describe what the UI is supposed to look like. Then you add the functionality using a different language.
Actually, with languages like Haskell and Prolog, you can describe the result that you want, and give the system enough rules to figure out how to implement it. These languages are more descriptive than imperative. They have a lot of power, but I don't think they'll ever catch on much.
The command has a default argument order, but also has keywords that can be used to rearrange the order. I really wish a couple UNIX commands had this option. Like grep -- I can never remember if the search string goes at the beginning or at the end.
When creating a cluster, you want to have a clear goal in mind. A cluster combines the resources of several machines, making them appear as one machine. What resources do you want to combine? Do the machines have to appear as a single machine to reach that goal?
Beowulf would be a good solution if you want to combine computational power, but your programs must be written for a parallel processing API.
Linux Virtual Servers is a decent solution if you want to implement network services such as a web server. There are commercial products that can do this as well. (Alteon boxes do it in hardware, TurboLinux has a software product.)
If you're trying to combine the systems into one big server that you can log into, I don't think that there is anything that can do that.
Don't make the mistake that Mozilla made. They released it as Open Source when it was not a viable working program. That's one reason why it has been making slow progress. It's hard to get people interested in something that doesn't work.
As far as waiting until later, that's probably not the best idea either. As long as the program is good and useful, you might as well get the extra hype by making it Open Source right away, to get more users.
What is Basilisk II? Was that the Mac emulator that ran on the Amiga?
Morals and values are decided by individuals, not groups. The "values of society" is a misnomer. It is what the majority believes. That doesn't mean that I have to believe it. Democracy is supposed to protect the rights and beliefs of the minority.
The point is that there is no "common value system" for geeks. If anything, geeks are more likely to think for themselves and come up with their own beliefs. Don't try to pigeon-hole us all into one demographic.
I think you may have cause and effect mixed up here. Just because intersections with traffic lights have more accidents does not mean that traffic lights cause more accidents. More likely, the fact that more accidents happen with traffic lights is because traffic lights are installed at intersections with more traffic.
The round-about circle intersection works pretty well if you know how to use it. But most of the people in my part of the world all want to use the outside lane.
Well, the first task is to get it looking at least as good as the competition. Then add features to make it better than the competition. Then we can work on the UI. Unless there are some UI experts, the GNOME guys tend to concentrate on functionality and then usability. The fact is, Outlook's layout works well. Changes to something that works that well should be minimal unless we're sure that they will improve things.
I can't recall too many performance increases smaller than this, percentage-wise. There were smaller increments between 60/66, 90/100, and 120/133, but I'm pretty sure that the faster models came out at the same time as the slower ones.
A 11% increase doesn't sound that exciting to me. Then again, a 113MHz increase doesn't sound so bad.
Everyone keeps referring to the site at msstate.edu as being the University of Missouri. Check your state abbreviations, people. That would be Mississippi State.
Jobs already made the mistake of releasing a cube-shaped computer once in his career. The original NeXT computer was a black cube, 12 inches on a side. You don't see too many of them around any more. Of course part of the reason could be that the only removable storage they had was a 256MB optical disk, which were really expensive.
And I'd be willing to bet that the Cobalt Raq systems sell better than the Cobalt Qubes.
Here are the specs of the NeXT cube, according to "The NeXT Book", by Bruce Webster, first edition, 1989. 68030 CPU at 25MHz, 68882 FPU, 56001 DSP, 8MB RAM, expandable to 64MB (16 SIMM slots!), 330MB or 660MB 5.25" hard drive, 256MB optical drive (the only removable storage!), built-in SCSI, built-in Ethernet, 1120x832 gray-scale (2 bit-plane) 17" display, 256K video buffer, passive backplane with 4 slots (similar to NuBus, but the cards are about 11"x11"; the main board takes up one of the slots), nice keyboard (includes power, volume, and brightness control), decent mouse, black cast magnesium case.
Jobs already made the mistake of releasing a cube-shaped computer once in his career. The original NeXT computer was a black cube, 12 inches on a side. You don't see too many of them around any more.
And I'd be willing to bet that the Cobalt Raq systems sell better than the Cobalt Qubes.
Here are the specs of the NeXT cube, according to "The NeXT Book", by Bruce Webster, first edition, 1989. 68030 CPU at 25MHz, 68882 FPU, 56001 DSP, 8MB RAM, expandable to 64MB (16 SIMM slots!), 330MB or 660MB 5.25" hard drive, 256MB optical drive (the only removable storage!), built-in SCSI, built-in Ethernet, 1120x832 gray-scale (2 bit-plane) 17" display, 256K video buffer, passive backplane with 4 slots (similar to NuBus, but the cards are about 11"x11"; the main board takes up one of the slots), nice keyboard (includes power, volume, and brightness control), decent mouse, black cast magnesium case.
The hardware was definitely impressive for its day (1989) but a bit pricey. Actually, much of it was ahead of its time. Nobody had 17" monitors back then; nobody had that kind of resolution. And the optical drive was a flop. The color slabs were the really nice boxes. They came out a year or so later than the cube.
The really impressive part of the NeXT was the software though. You can see its legacy in a lot of some Linux GUI software, and the Mac OS X is based on NeXTStep. One of the best things about NeXTStep is its use of Display PostScript. WSYIWYG, because the display and the printer are written to in exactly the same way.
Pardon my ignorance, but who is Bruce Campbell? What has he been in? Why is it ironic that he would be chosen?
Other posts have pointed out that it is cheaper to build telescopes on the earth, but orbital telescopes eliminate problems with the atmosphere causing interference. Might placing a telescope on the moon be more cost-effective than an orbiting telescope, while solving some of the problems with orbiting telescopes? Or would it end up being the worst of both worlds instead of the best of both worlds?
Why is the documentation in EXE files? That's the stupidest thing I've seen all day.
I respect the work of the developers, but I don't see anything on the web site backing up the claim that the format is completely free of patents. Have they done a patent search?
Other than that, it looks great.
The company selling the code can't retain all rights, or else you wouldn't receive anything from them. The contract clearly had to indicate that there was some type of deliverable. The delivery of that product or service implicitly has to be usable by the recipient. In other words, even if they retain a non-exclusive right to the work they create, they cannot exclude you from using what they sold you. There is no way that they can win, unless the company never paid them at all for their partial work.
We use Maxspeed stations at work. They're not too bad. They turn on instantly, and you can run at least 4 off a decent Linux system without problems. You just install a proprietary serial card in the server and hang the clients off the server. Then install Red Hat on the server and add the Maxspeed package on top of it. One problem is that the server needs to be in the same room as the clients pretty much. I'm guessing the boxes cost about $400 apiece. (No prices on the web page.) You can just add an old 15" SVGA monitor for another $50 or less.
Um, gel?
Yes, I think maybe they're made of some kind of gel.
I see a lot of headlines about there being too many tech jobs and not enough qualified candidates to fill them. The real problem is that companies think they can find the perfect employee and not have to train them at all. This is ludicrous. If an employee wanted to keep doing the same thing that she did in her last job, she would have stayed there.
So even if you're more capable at programming in Java than some people with 2 years experience with it, some companies are too dumb to figure that out. On the bright side, you wouldn't want to work for a company that dumb anyway. Unfortunately, there are a lot of dumb companies out there, making it difficult to get a job sometimes.
The best jobs I've gotten are ones that challenge me to learn new things. I think I do better at those jobs too, so I don't understand why companies just don't get it.
Don't try to save money when buying a carrying case for your notebook. I tried 2 cheap cases for my notebook and they both broke after about 2 trips out of town. I ended up with a Targus and wish I had bought it in the first place. Whatever you buy, make sure the carrying strap is connected to the bag really really well.
I got a chuckle out of that. But I would have to agree with it 100%.
I've been saying the same thing about poor people in this country. What do I have that they don't? Not all that much. Which leads me to the conclusion that they really aren't poor, but they don't realize it.
And it's the same with the middle class. They don't realize that they can take time to just enjoy and relax. That's what the Industrial Revolution, the Consumer Revolution, and the Techology Revolution have given us, but we've failed to realize how to take advantage of it.
If I were paranoid, I'd say that someone is keeping us from making that realization. If so, I'd guess that it is corporations. They need us to continue working for them and buying their products that we don't really need.
Just the summary of Stephenson's speech touches on so many things. I find it very interesting that he contrasts his ideas about threats to privacy with the Big Brother model. I think there's a good chance he will end up writing a work that will be seen as the "1984" of the 21st century. And I think he's already hit the nail on the head.
I think that corporatism will have a large part to play in this domination system. That seems to be where the biggest threat is coming from already. I've already seen a book or two on how opression has moved from the racial minorities to the working class in general, and how employers use various methods to keep workers from becoming too powerful.
Regarding corporatism, I find it rather sad -- I see so many workers with good ideas that work in corporate cultures where their ideas have no way to be expressed. If the Open Source community can somehow get their merit-based methods to be applied to the business world, I'll have a much more optimistic outlook on the future of this country (and ones like it).