When's the last time you ran windows 98 or earlier. Remember how long it took to boot? My windows 98 box boots in less then 10 seconds (Sony vaio z505js) and linux on the same box boots in less then 5 seconds after POST (I'm not talking about a typical redhat install that starts every service known to man. It's quite minimal). Anyway, my point is this: 12 seconds for a "Fast restore" blows. It should be *booting* in less time then that, you *can* be booting in less time then that, and you don't have to deal with the annoyance that is windows XP to do it.
On your linux box you should be able to completely restore you session from boot in less time then windows XP is restoring from "hibernation". You just need to disable all the stuff that init is running that you don't use, and configure your window manager and applications correctly. There is no reason you shouldn't be able to log out and back in and have your session exactly as you've left it. Unfortunatly, if you want hibernation you need to get a PC with bios support for it, or you need to get a mac, on which linux will wake from sleep in less then 1 second.
Nope, but mine did come with one of those new fangled thermostat thingies that magically turns the compressor on only when it starts to get warm inside.
Most refrigerators aren't designed to run continuously. The one you linked to consumes ~100 watts, and could probably only cool a system that runs on less then 80 watts. Even then, the continuous running would probably destroy the refrigerator pretty quickly.
There's a bunch more too, Tundra, GMS, Force, just do a search on google. You'll likely find though that Apple has the best prices. If you want to play with a PPC (I'm assuming you want to do some low level stuff for fun or profit) you'll end up spending $1500 on just a board from somewhere else, or $1500 on a complete system from Apple. The Apple systems retain their value for a long time too.
Next time he sleeps, tie him to the bed with 20 pound test fishing line (Over a blanket so there's no bleeding when he tries to sit up). Feed him through a straw for a week, and by that time the withdrawl should be over and he'll be able to walk away. (I.E His character will be so far behind all his online buddies that he won't have fun anymore.)
I'm curious to know what kind of AIDS research you think could be done with a really fast computer? Most problems aren't suited to just throwing lots of computer power at them.
Furthermore, when equipment like this is built for military testing, the result almost always benifits other fields shortly afterward. After the DOE is done playing with it, time on the machine will be available for other research. You can rent CPU time on most of the ASCI machines the DOE has built in the past. Without the defense spending on nuclear simulation, this machine probably wouldn't have been funded in the first place.
Copyright law allows for independant creation. Blizard can only limit independant creation of similar code if they have a patent of some sort on it (Or I suppose nowadays if it violates the DMCA). All the bnetd group needs to do is prove that bnetd was created independantly of blizzard's battle.net server and they're all set. From what I've seen of their CVS repository, they'll have no trouble proving that.
Of cource blizzard could probably get all these guys on licence violations of their products...
It's not that there aren't open source alternatives, but that the open source alternatives have no marketing budget. Most people use what's crammed down their throat. They've been trained not to think for themselves.
You *CAN'T BUY* a dual P4 unless you get Xeon processors.
An equivalently priced machine from dell (2x1.5Ghz Xeons, DVD+RW but no cd-r support, 80Gb drive, GeForce4MX, Copper Gigabit, Modem, 512Mb RAM) costs $3869 after the $200 rebate.
The premium you pay for Macs is not that high anymore. 10-20%. The other thing that nobody tells you is that Macs retain their value. That $1.2K PC that you buy will be worth $300 in two years. The $1600 Mac will be worth $900 in two years.
$1635 - $900 = $735 $1200 - $300 = $900
You've payed more for the PC in the long run if you upgrade semi-regularly.
Actually, and I'm serious, I'd be impressed if he could move data to/dev/null that quickly. I don't know of any small computer bus architectures available right now that could handle that kind of speed.
More impressively, almost 4 million people could streem unique mp3's over this line simultaniously.
I certainly hope that your USB mouse driver is interrupt driven. There shouldn't be any polling going on, as that would be an enormous waste of system resources. Similarly the origional post was dubious at best, since PS/2 ports can also generate interrupts. A still mouse should use *zero* system resources. USB does have higher throughput then PS/2, so theoretically the mouse could send updates much more frequently then the PS/2 equivalent, and that would account for smoothness.
Unfortunatly, the test plan in the situation I was refering to was not designed by me and involved a human component in each step. Retesting was labor intensive. Either way, my point was that if the text section of the binary hasn't changed, then the code hasn't changed. Period. It doesn't matter if the source files are different, the code that the machine is running is still the same and will produce the same test results.
I'm not saying managers shouldn't want to gague progress. I'm saying it's the managers that pick the idiotic metric. Furthemore, it can be hard to break down achivements by week. Some projects can't be measured that finely, or worse, the week by week checkpoints will create extra work along the way. (see my reply to another comment in this thread for an example of this.) The aproach to measure progress needs to be decided on a per project basis after the design is complete. Using braindead monolithic policies to jedge project progress will get you nowhere alot of the time.
I don't expect to be left alone for 6 months. What I'm trying to say I guess is that the best way to get an acurate measure of how engineers are progressing is to have the project managed by an engineer. I'm also a big fan of having sales contracts for custom engineering work reviewed by the engineers before it's signed. I've seen too much dumb shit agreed to in custom engineering contracts, and for too little money. (You can't pay me enough to break the laws of physics, or to say, support USB on a device with no USB hardware).
I'm working on a project right now. It's a OS kernel layer. The schedule is written exactly as you say. Get a shipable product right away with a minimal feature set, and then add features. The trouble is that it would have taken me just as long to write the entire product as it is for me to write each of three feature checkpoints* (I've completed two of them at this point). The feature completion dates were written even before I had completed the design (the design document was the first date). Be careful with your method. Decisions about scheduling feature completion dates cannot be made by people without intimate knowledge of a design.
Also, managers that require complete retesting after whitespace/comment changes piss me off. It's not always possible to document how you're going to do something as you do it. The subtlties that you want documented in comments come out during the implementation, not during the design. Commenting after the fact (to augment existing comments) is a necissity for creating good documentation. If the change doesn't affect the binary output of the compiler then you don't need a compete retest. I have worked with companies where I've had to prove to them that the text section of the binary was identical in order to prevent a restart of the 3 month testing cycle (and delaying the arival of my check:), but I've also worked with companies where no argument I made would convince them that nothing has really changed even though I added documentation. It's managers like that who encourage people to not do the extra documentation work at the end of a project.
*(BTW, I couldn't just implement all of the features right away and finish early, some of the check points include manual invocation of features that are supposed to work automatically. Since in the final version there is no provision for causing certain events manually alot of extra work had to be done to make that possible. Much of that code has to be removed, and is wasted, becuase manual invocation of these features would cause undesierable results in the final version.)
Would those be simple problems solved, or major engineering problems solved? Would the programmer who made a button change the values in a form to the correct ones have accomplished as much as someone who developed a more scalable network protocol, or increased the paralellism of a complex algorithim, or came up with a new mathematical theorm? Is the stupid manager going to decide that the button problem was harder to solve because the IDE generated 500 lines of code for his button click, while the computer scientist only generated a few pages of scribbling in his notebook?
Yeah, that's great, but what manager works like that? No manager would ever say to me "John, how long would this take you?" for a project with a 6 month or 1 year time frame, and then leave me alone for 6 months while I write code. Managers want progress reports and statistical measures to reduce the risk that that they might get fired when the project they're managing fails. That's where the boneheaded requests for the "number of lines of code you've written this week", or "number of bugs you've fixed this week" come from. It's hard to look at an unfinished piece of software and know how it's coming along, but it's easy to look at a painting or a statue that's in progress and see work being done.
Unless you're limited by cost, don't use host based RAID. It will always be less reliable then a dedicated RAID controller. If you must use host based RAID, try and find a card that supports RAID 0/1 because it's faster and more reliable. I've had good experiences with MegaRAID cards, and the IBM host based raid controllers, but by good experience I mean that I've only had a few problems. There is always a chance that something will get screwed up when you change your setup.
I know you typed for a long time, but what exactly are you trying to say here. You seem to agree and disagree with this article, and you provide a myriad of incorrect facts to "back up" your "point". For example:
A chemical rocket manned mission was accomplished in 1969 - over 30 years ago. In fact, if I recall, they made several trips and nobody died.
Actually, three people died, and we almost lost three more.
8. Clean? Let's start with safety, which is, IMHO, a prerequisite for cleanliness. At...
Here you compare an absolute with a relative descriptor. Just because something isn't completely clean doesn't mean it isn't the cleanest thing available.
Also, you compare everything he says with what Zubrin thinks. Now I'm not saying that Zubrin is flat out wrong, but he hasn't sent any more people to mars then the rest of us. His ideas and opinions are no more then that, simply ideas and opinions. He is definatly one to admit that the problem is too complex to solve all at once. Even he doesn't know if he's right. You can't write off what anyone else says about getting to mars because it's not the same as what Zubrin said. There's likely more than one way to do it.
Then at the end you say: Yes, and in order to acquire more solar energy, we need advanced propulsion systems to set up collectors further out in the solar system. So, are you for or against nuclear propultion? Do you agree with the author dispite all the 'flaws' you found in his article, or disagree?
Slashdot Mirror
When's the last time you ran windows 98 or earlier. Remember how long it took to boot? My windows 98 box boots in less then 10 seconds (Sony vaio z505js) and linux on the same box boots in less then 5 seconds after POST (I'm not talking about a typical redhat install that starts every service known to man. It's quite minimal). Anyway, my point is this: 12 seconds for a "Fast restore" blows. It should be *booting* in less time then that, you *can* be booting in less time then that, and you don't have to deal with the annoyance that is windows XP to do it.
On your linux box you should be able to completely restore you session from boot in less time then windows XP is restoring from "hibernation". You just need to disable all the stuff that init is running that you don't use, and configure your window manager and applications correctly. There is no reason you shouldn't be able to log out and back in and have your session exactly as you've left it. Unfortunatly, if you want hibernation you need to get a PC with bios support for it, or you need to get a mac, on which linux will wake from sleep in less then 1 second.
Sorry, but the first FPS for a console that didn't suck was Goldeneye for N64.
Nope, but mine did come with one of those new fangled thermostat thingies that magically turns the compressor on only when it starts to get warm inside.
Most refrigerators aren't designed to run continuously. The one you linked to consumes ~100 watts, and could probably only cool a system that runs on less then 80 watts. Even then, the continuous running would probably destroy the refrigerator pretty quickly.
Many people would be happy to sell you a board with a G4 on it. Maybe even 2 G4s!
Marvell makes ATX boards with 1 or 2 7450s.
MotorolaMakes a very nice ATX board with 2 7450's on it. They also have the Sandpoint platform which you can use with many different PPC chips.
Merlancia seems to have some good stuff.
There's a bunch more too, Tundra, GMS, Force, just do a search on google. You'll likely find though that Apple has the best prices. If you want to play with a PPC (I'm assuming you want to do some low level stuff for fun or profit) you'll end up spending $1500 on just a board from somewhere else, or $1500 on a complete system from Apple. The Apple systems retain their value for a long time too.
Next time he sleeps, tie him to the bed with 20 pound test fishing line (Over a blanket so there's no bleeding when he tries to sit up). Feed him through a straw for a week, and by that time the withdrawl should be over and he'll be able to walk away. (I.E His character will be so far behind all his online buddies that he won't have fun anymore.)
Warning, this may be illegal in some countries.
I'm going to go out on a limb here, but perhaps you could give them money by buying books from them....
I'm curious to know what kind of AIDS research you think could be done with a really fast computer? Most problems aren't suited to just throwing lots of computer power at them.
Furthermore, when equipment like this is built for military testing, the result almost always benifits other fields shortly afterward. After the DOE is done playing with it, time on the machine will be available for other research. You can rent CPU time on most of the ASCI machines the DOE has built in the past. Without the defense spending on nuclear simulation, this machine probably wouldn't have been funded in the first place.
Copyright law allows for independant creation. Blizard can only limit independant creation of similar code if they have a patent of some sort on it (Or I suppose nowadays if it violates the DMCA). All the bnetd group needs to do is prove that bnetd was created independantly of blizzard's battle.net server and they're all set. From what I've seen of their CVS repository, they'll have no trouble proving that.
Of cource blizzard could probably get all these guys on licence violations of their products...
It's not that there aren't open source alternatives, but that the open source alternatives have no marketing budget. Most people use what's crammed down their throat. They've been trained not to think for themselves.
Crap. I meant equivalently equipped.
You *CAN'T BUY* a dual P4 unless you get Xeon processors.
An equivalently priced machine from dell (2x1.5Ghz Xeons, DVD+RW but no cd-r support, 80Gb drive, GeForce4MX, Copper Gigabit, Modem, 512Mb RAM) costs $3869 after the $200 rebate.
PowerMac G4: $1599
256Mb DIMM: $36
Total: $1635
What are you talking about again?
The premium you pay for Macs is not that high anymore. 10-20%.
The other thing that nobody tells you is that Macs retain their value. That $1.2K PC that you buy will be worth $300 in two years. The $1600 Mac will be worth $900 in two years.
$1635 - $900 = $735
$1200 - $300 = $900
You've payed more for the PC in the long run if you upgrade semi-regularly.
That's exactly what I'm trying to say.
Actually, and I'm serious, I'd be impressed if he could move data to /dev/null that quickly. I don't know of any small computer bus architectures available right now that could handle that kind of speed.
More impressively, almost 4 million people could streem unique mp3's over this line simultaniously.
I certainly hope that your USB mouse driver is interrupt driven. There shouldn't be any polling going on, as that would be an enormous waste of system resources. Similarly the origional post was dubious at best, since PS/2 ports can also generate interrupts. A still mouse should use *zero* system resources. USB does have higher throughput then PS/2, so theoretically the mouse could send updates much more frequently then the PS/2 equivalent, and that would account for smoothness.
Unfortunatly, the test plan in the situation I was refering to was not designed by me and involved a human component in each step. Retesting was labor intensive. Either way, my point was that if the text section of the binary hasn't changed, then the code hasn't changed. Period. It doesn't matter if the source files are different, the code that the machine is running is still the same and will produce the same test results.
I'm not saying managers shouldn't want to gague progress. I'm saying it's the managers that pick the idiotic metric. Furthemore, it can be hard to break down achivements by week. Some projects can't be measured that finely, or worse, the week by week checkpoints will create extra work along the way. (see my reply to another comment in this thread for an example of this.) The aproach to measure progress needs to be decided on a per project basis after the design is complete. Using braindead monolithic policies to jedge project progress will get you nowhere alot of the time.
I don't expect to be left alone for 6 months. What I'm trying to say I guess is that the best way to get an acurate measure of how engineers are progressing is to have the project managed by an engineer. I'm also a big fan of having sales contracts for custom engineering work reviewed by the engineers before it's signed. I've seen too much dumb shit agreed to in custom engineering contracts, and for too little money. (You can't pay me enough to break the laws of physics, or to say, support USB on a device with no USB hardware).
I'm working on a project right now. It's a OS kernel layer. The schedule is written exactly as you say. Get a shipable product right away with a minimal feature set, and then add features. The trouble is that it would have taken me just as long to write the entire product as it is for me to write each of three feature checkpoints* (I've completed two of them at this point). The feature completion dates were written even before I had completed the design (the design document was the first date). Be careful with your method. Decisions about scheduling feature completion dates cannot be made by people without intimate knowledge of a design.
:), but I've also worked with companies where no argument I made would convince them that nothing has really changed even though I added documentation. It's managers like that who encourage people to not do the extra documentation work at the end of a project.
Also, managers that require complete retesting after whitespace/comment changes piss me off. It's not always possible to document how you're going to do something as you do it. The subtlties that you want documented in comments come out during the implementation, not during the design. Commenting after the fact (to augment existing comments) is a necissity for creating good documentation. If the change doesn't affect the binary output of the compiler then you don't need a compete retest. I have worked with companies where I've had to prove to them that the text section of the binary was identical in order to prevent a restart of the 3 month testing cycle (and delaying the arival of my check
*(BTW, I couldn't just implement all of the features right away and finish early, some of the check points include manual invocation of features that are supposed to work automatically. Since in the final version there is no provision for causing certain events manually alot of extra work had to be done to make that possible. Much of that code has to be removed, and is wasted, becuase manual invocation of these features would cause undesierable results in the final version.)
Would those be simple problems solved, or major engineering problems solved? Would the programmer who made a button change the values in a form to the correct ones have accomplished as much as someone who developed a more scalable network protocol, or increased the paralellism of a complex algorithim, or came up with a new mathematical theorm? Is the stupid manager going to decide that the button problem was harder to solve because the IDE generated 500 lines of code for his button click, while the computer scientist only generated a few pages of scribbling in his notebook?
Yeah, that's great, but what manager works like that? No manager would ever say to me "John, how long would this take you?" for a project with a 6 month or 1 year time frame, and then leave me alone for 6 months while I write code. Managers want progress reports and statistical measures to reduce the risk that that they might get fired when the project they're managing fails. That's where the boneheaded requests for the "number of lines of code you've written this week", or "number of bugs you've fixed this week" come from. It's hard to look at an unfinished piece of software and know how it's coming along, but it's easy to look at a painting or a statue that's in progress and see work being done.
Unless you're limited by cost, don't use host based RAID. It will always be less reliable then a dedicated RAID controller. If you must use host based RAID, try and find a card that supports RAID 0/1 because it's faster and more reliable. I've had good experiences with MegaRAID cards, and the IBM host based raid controllers, but by good experience I mean that I've only had a few problems. There is always a chance that something will get screwed up when you change your setup.
Most PERC boards are AMI MegaRAID cards rebranded.
I know you typed for a long time, but what exactly are you trying to say here. You seem to agree and disagree with this article, and you provide a myriad of incorrect facts to "back up" your "point". For example:
A chemical rocket manned mission was accomplished in 1969 - over 30 years ago. In fact, if I recall, they made several trips and nobody died.
Actually, three people died, and we almost lost three more.
8. Clean? Let's start with safety, which is, IMHO, a prerequisite for cleanliness. At...
Here you compare an absolute with a relative descriptor. Just because something isn't completely clean doesn't mean it isn't the cleanest thing available.
Also, you compare everything he says with what Zubrin thinks. Now I'm not saying that Zubrin is flat out wrong, but he hasn't sent any more people to mars then the rest of us. His ideas and opinions are no more then that, simply ideas and opinions. He is definatly one to admit that the problem is too complex to solve all at once. Even he doesn't know if he's right. You can't write off what anyone else says about getting to mars because it's not the same as what Zubrin said. There's likely more than one way to do it.
Then at the end you say: Yes, and in order to acquire more solar energy, we need advanced propulsion systems to set up collectors further out in the solar system.
So, are you for or against nuclear propultion? Do you agree with the author dispite all the 'flaws' you found in his article, or disagree?