Sounds good! But you're probably top-ending your machines (buying them fully packed rather than leaving headroom for expansion). When you run out of logical partitioning options adding nodes will be less effective, so for really big operations you still have to escape the Intel architecture and go with a Sun or HP or IBM or what-have-you. Especially us poor SAP SOBs.
With the possible exeption of IBM you'll probably not see Linux running DB clusters on those platforms, as the companies have put so much time and money into optimizing their own cluster technology. The "big iron" is still proprietary space for now, and the big databases need big iron unless they're the digital equivalent of scrapyards (where the most common activity is rusting).
That's not to say Linux isn't ideal for the space you're in -- as a better alternative to NT on commodity boxes. Many of today's "small-to-average" databases run in the 100-500GB range, and dwarf the "large databases" of not-so-long-ago. It's just that today's VLDB are measured in terabytes, headed for petabytes, and expecations for response times are shrinking. Our ambition grows with our grasp.
This (if it's reliable:) answers my bluetooth issue: why do I want a wireless networking protocol that *doesn't* talk to my home network? (ans: I don't)
Why do I want another wireless card in my laptop or on my home network router, if I can stick with the WLAN stuff I have now? And don't quote the "voice" business from the article, I'm not in the habit of talking into my PDA, and voice-over-IP routes better anyway. Sure there might be home-intercom applications, but voice-over-ip at 10x the speed should handle that nicely:-).
I support the investigation of criminals and their close associates, keep that in mind. I do not support the investigation of "unmoral activity" -- ask Billy Wilder about "unmoral activity" investgations by the House in the 50's.
The FBI, Secret Service, CIA and other state and federal agencies have been slapped for abusing their search and seizure powers more times than you can shake a stick at, for reasons varying from partisan politics to racial discrimination to "discussing dangerous ideas" to "just because". Ask Steve Jackson how long it will take to get your computer back if you follow a URL to a terrorist-supporting site because you wanted to check the balance in your IRA account....remember, the proposals let them thoroughly investigate anyone "relevant", with or without their consent, and your opinion of your relevance doesn't count. And with the new law, neither does the Court's.
This is not to say that the FBI and others don't do a stellar job most of the time, but hey -- they really have all the data collection powers they need (or very, very nearly). What they need to fight crime is better analysis, more clueful analysts, and more informants or undercover agents. Anything more is pretty much useless for fighting crime, but is useful for the next McCarthys. It might be useful to permit law enforcement to proceed with wiretapping for a limited amount of time, to permit speedy response, but taking court oversight out of the process practically ensures abuse.
Oh, and the only good that will come from a national ID is that you'll get more targeted ads. And it'll make database design easier. And forgers can reduce the amount of card stock they run through (since only one blank form will be needed to steal an ID). Terrorists will get and exercise their national ID's and the FBI will use the numbers to backtrack them after the fact. Unless of course, the terrorists have half a brain and get more than one ID.
Strange, but the investigators seem to be doing an efficient job of backtracking them now.... Why do we need it again?
Please explain how the use of hydrogen as a fuel would "decrease the dependancy on foriegn oil"?
Simple -- use nuclear power to crack the water. This does actually mean we need to build some efficent recyclers for spent fuel (we don't have any commercial ones AFAIK). That would seriously reduce the output of nuclear waste from these plants, but even if we don't the nuke plant doesn't have to fly over my house at 300+ knots.
Ok, those of you who don't beleive in nuclear power can consider your favorite power source, but the point is that H2 combustion permits you to move the power from a fixed site, to the airplane (just like we move fossil fuels from where they are), and in general it's a lot more feasible than using those other sources in the plane. Anybody care to comment on the efficiency of h2 as a power transmission medium?
When people talk about Hydrogen powered aircraft, they are talking about using rocket engines, not fuel cells.
Not necessarily. A jet turbine could be built with just about any flammable fuel, although its geometry and compression ratios would be different for anything but Jet-A... Jets already compress the incoming air to improve combustion efficiency. I'd like to see the math. Anyway, newer fuel cells are getting smaller and lighter...
Also, fuel cells don't have a high enough power-to-weight ratio to get an airplane off the ground. [...] Electric motors are nowhere close to
having the kind of efficiency needed for an aircraft engine.
Again, let's see the math. Electric motors are extremely efficient, but I tend to agree that the weight of the electric motors and their energy supply has made electrical propusion undesirable for airplanes (although you can get one off the ground, it won't go far or lift much). But I'd be interested to see whether we can do better with current or nearly-here tech. That fuel cell we heard about earlier this week was about half the weight of equivalent batteries, wasn't it?
I suspect that a turbofan is still more likely, but who knows, if we can get the fuel cell membrane incorporated into the skin or structure of the plane (to take advantage of free oxygen and reduce added weight) and maybe supplement with solar power (planes spend a lot of time above the clouds, especially going west), maybe it would be do-able.
Re:Emacs emulation in vim?
on
VIM 6.0 is Out
·
· Score: 5, Funny
No, it's like asking a Ferrari to emulate a 60-story office complex, an employment agency, an aircraft manufacturer, a luxury yacht, and a Ferrari.
Perhaps the EFF should set up a facility for monitoring all politicians' surfing and email habits... surely some of them go online. If they have no expectation of privacy, it wouldn't be illegal, any more than publicizing their voting record or public appearance schedule, right? Sauce for the goose.
Sometimes I wonder if the people proposing these laws plan to emigrate when they retire...
If the industry is losing billions to copying, and they've made it impossible, we can expect to see prices fall to say $4.99, right? Or were they lying about napster....
Hrm... yes, but when smuggling weapons onto the plane (infiltrating to cause havoc), it's best to go to airports with poor security.
If MS continues to stealth-install IIS (so that admins don't patch it cause they don't know it's there) and if they continue to leave holes in browsers (like always executing.eml files) then their OS will keep being hijacked.
Re:Not a reliable weapons system yet
on
Robots Go To War
·
· Score: 2
Friendly fire was the first thing I thought of -- it's finally possible for a pilot to take himself out with an air-to-ground weapon.
"Look, Bert, two guys hunched over a card table! Let's shoot 'em..."
The radio stations get paid by the eardrum. Anything that makes the audience turn the channel or turn off the radio is bad (by their lights). This is just the pablum-ization of the airwaves to prevent loss of market share. Clearly they don't care about the tastefulness or message of the music, only the likelihood that someone, somewhere, will find a reason to stop listening to the commercials.
Of course, that should tell you something about the quality of programming on the network...
Multiple fans mean multiple inspections & failures. An array of 100 3-meter fans would probably cause more problems than one well-maintained 30 meter fan, and cost more than a nuclear pile in the basement:-).
w/r/t the spillage problem, you could do this more simply with one large multivaned turbine.
The problem is that a design that maximizes use of the airflow means more minimum drag (when the fans are feathered to spill air). If you're too efficient you make a sail that drags your building s over.
Better than snow generators -- the buildings will want to bend on windy days, since the cooling will be localized to the turbine side... Now there's a fun engineering problem!
The thing to do is to twist the arms of the Taliban without getting
violent, have them turn over bin Laden by convincing them that he's a liability, and let the man rot for the rest of his life in solitary in Colorado.
Right! Then we can go bomb Iraq....
OBL is already wanted for conspiracy in other events, but this amount of international ire is large enough to spread to more than one target.
I suppose my rambling came around to the argument that if, like in civil engineering, there were
- professional licensing boards for Software Engineers
- certification of designs, plans and built code by those engineers
- professional liability for the items they certify
- and requirements for professional certification for commercial applications (building codes)
then we'd have more reliable software. Standards and regulations for what may and may not be certified ("sealed" in traditional parlance) can come later, but lack of accountability is the problem. Any method that legally requires accountabilty for quality and reliability in a class of software will force an improvement, and a corresponding rise in the cost of making that software (which is why it must be a legal requirement for all competitors, or it won't happen).
So we need building codes for software. Because right now people are willing to pay for a cardboard refrigerator box, and expect to get a single family dwelling in return, and it's not illegal to give them the box. That's how you get to be the richest man in the world:-).
But what codes are appropriate? Simple merchantability, or some certification of tested-ness and guarantee of defect repair? And how much of that are we willing to pay for? Certainly the current crop of EULAs would have to go.
So what do you think -- a Software Engineering Board? We have professsional engineering boards for civil, mechanical and electrical design. Of course, this would put severe limitations on the hobbyist and any non-engineer-reviewed software projects. Careful what you ask for...
Slashdot Mirror
With the possible exeption of IBM you'll probably not see Linux running DB clusters on those platforms, as the companies have put so much time and money into optimizing their own cluster technology. The "big iron" is still proprietary space for now, and the big databases need big iron unless they're the digital equivalent of scrapyards (where the most common activity is rusting).
That's not to say Linux isn't ideal for the space you're in -- as a better alternative to NT on commodity boxes. Many of today's "small-to-average" databases run in the 100-500GB range, and dwarf the "large databases" of not-so-long-ago. It's just that today's VLDB are measured in terabytes, headed for petabytes, and expecations for response times are shrinking. Our ambition grows with our grasp.
Why do I want another wireless card in my laptop or on my home network router, if I can stick with the WLAN stuff I have now? And don't quote the "voice" business from the article, I'm not in the habit of talking into my PDA, and voice-over-IP routes better anyway. Sure there might be home-intercom applications, but voice-over-ip at 10x the speed should handle that nicely :-).
I support the investigation of criminals and their close associates, keep that in mind. I do not support the investigation of "unmoral activity" -- ask Billy Wilder about "unmoral activity" investgations by the House in the 50's.
The FBI, Secret Service, CIA and other state and federal agencies have been slapped for abusing their search and seizure powers more times than you can shake a stick at, for reasons varying from partisan politics to racial discrimination to "discussing dangerous ideas" to "just because". Ask Steve Jackson how long it will take to get your computer back if you follow a URL to a terrorist-supporting site because you wanted to check the balance in your IRA account....remember, the proposals let them thoroughly investigate anyone "relevant", with or without their consent, and your opinion of your relevance doesn't count. And with the new law, neither does the Court's.
This is not to say that the FBI and others don't do a stellar job most of the time, but hey -- they really have all the data collection powers they need (or very, very nearly). What they need to fight crime is better analysis, more clueful analysts, and more informants or undercover agents. Anything more is pretty much useless for fighting crime, but is useful for the next McCarthys. It might be useful to permit law enforcement to proceed with wiretapping for a limited amount of time, to permit speedy response, but taking court oversight out of the process practically ensures abuse.
Oh, and the only good that will come from a national ID is that you'll get more targeted ads. And it'll make database design easier. And forgers can reduce the amount of card stock they run through (since only one blank form will be needed to steal an ID). Terrorists will get and exercise their national ID's and the FBI will use the numbers to backtrack them after the fact. Unless of course, the terrorists have half a brain and get more than one ID.
Strange, but the investigators seem to be doing an efficient job of backtracking them now.... Why do we need it again?
I thought that was seedy street lighting (or radioactive acne).
Yes, generally things that beat a nice large anvil are, in fact, hitting a nice large anvil.
Simple -- use nuclear power to crack the water. This does actually mean we need to build some efficent recyclers for spent fuel (we don't have any commercial ones AFAIK). That would seriously reduce the output of nuclear waste from these plants, but even if we don't the nuke plant doesn't have to fly over my house at 300+ knots.
Ok, those of you who don't beleive in nuclear power can consider your favorite power source, but the point is that H2 combustion permits you to move the power from a fixed site, to the airplane (just like we move fossil fuels from where they are), and in general it's a lot more feasible than using those other sources in the plane. Anybody care to comment on the efficiency of h2 as a power transmission medium?
Not necessarily. A jet turbine could be built with just about any flammable fuel, although its geometry and compression ratios would be different for anything but Jet-A... Jets already compress the incoming air to improve combustion efficiency. I'd like to see the math. Anyway, newer fuel cells are getting smaller and lighter...
Also, fuel cells don't have a high enough power-to-weight ratio to get an airplane off the ground. [...] Electric motors are nowhere close to having the kind of efficiency needed for an aircraft engine.
Again, let's see the math. Electric motors are extremely efficient, but I tend to agree that the weight of the electric motors and their energy supply has made electrical propusion undesirable for airplanes (although you can get one off the ground, it won't go far or lift much). But I'd be interested to see whether we can do better with current or nearly-here tech. That fuel cell we heard about earlier this week was about half the weight of equivalent batteries, wasn't it?
I suspect that a turbofan is still more likely, but who knows, if we can get the fuel cell membrane incorporated into the skin or structure of the plane (to take advantage of free oxygen and reduce added weight) and maybe supplement with solar power (planes spend a lot of time above the clouds, especially going west), maybe it would be do-able.
No, it's like asking a Ferrari to emulate a 60-story office complex, an employment agency, an aircraft manufacturer, a luxury yacht, and a Ferrari.
Sometimes I wonder if the people proposing these laws plan to emigrate when they retire...
Not unless you then say "You should use some kind of UNIX instead...", and the person is a terrorist. Then you're "advising or aiding".
If the industry is losing billions to copying, and they've made it impossible, we can expect to see prices fall to say $4.99, right? Or were they lying about napster....
If MS continues to stealth-install IIS (so that admins don't patch it cause they don't know it's there) and if they continue to leave holes in browsers (like always executing .eml files) then their OS will keep being hijacked.
"Look, Bert, two guys hunched over a card table! Let's shoot 'em..."
"Hey, Ernie, what's that whooshing noise?"
Sorry, I just haven't met the creationist that I can take seriously (of course, the universe is only 30 seconds old, so maybe there's time...).
Of course, that should tell you something about the quality of programming on the network...
w/r/t the spillage problem, you could do this more simply with one large multivaned turbine.
The problem is that a design that maximizes use of the airflow means more minimum drag (when the fans are feathered to spill air). If you're too efficient you make a sail that drags your building s over.
Better than snow generators -- the buildings will want to bend on windy days, since the cooling will be localized to the turbine side... Now there's a fun engineering problem!
Right! Then we can go bomb Iraq....
OBL is already wanted for conspiracy in other events, but this amount of international ire is large enough to spread to more than one target.
One more reason to go metric....
yes, it has been a long day. Thanks for asking :-).
That explains all the sections that seem to contain no information -- parentheses.....
Naturally -- they're just a blind for CURE.
The next night Dr. Wiseman tried again: "-126!"
There was complete silence, then the crowd laughed and cheered until tears ran down their faces.
They'd never heard that one before.
I am a Number! Not a Free Man! --- No. 444a65726d616e
- professional licensing boards for Software Engineers
- certification of designs, plans and built code by those engineers
- professional liability for the items they certify
- and requirements for professional certification for commercial applications (building codes)
then we'd have more reliable software. Standards and regulations for what may and may not be certified ("sealed" in traditional parlance) can come later, but lack of accountability is the problem. Any method that legally requires accountabilty for quality and reliability in a class of software will force an improvement, and a corresponding rise in the cost of making that software (which is why it must be a legal requirement for all competitors, or it won't happen).
But what codes are appropriate? Simple merchantability, or some certification of tested-ness and guarantee of defect repair? And how much of that are we willing to pay for? Certainly the current crop of EULAs would have to go.
So what do you think -- a Software Engineering Board? We have professsional engineering boards for civil, mechanical and electrical design. Of course, this would put severe limitations on the hobbyist and any non-engineer-reviewed software projects. Careful what you ask for...