Hey, we fingerprint all Canadians entering the US, soon to demand (RFID) passports from all of them, while our NSA is tapping their phones and email. Who knows what our CIA does up there.
Since Americans are allergic to Canadian pennies now worth almost as much as ours, and dump them whenever we see them on our side of the border, these RFID trackers are relatively pretty benign.
Maybe if we just all wound down the BS simcurity that pretends to protect us, and instead actually just destroyed us some Qaeda recruitment cells, the US dollar would become strong enough again that we wouldn't bother schlepping their Canadian coins back home, like pocket lint.
I tend to agree. But how does putting your Java into stored procedures compare to the long/inefficient SQL? And is the Java stored procedure slowdown due to a worse JVM, more demand on the same host for both DB and JVM, or some other architectural bottleneck?
Does the PS3 API to the SPUs automatically (at compile time, or runtime) scale to the number of SPUs available? And is switching networking from PS3 to CBE infiniband completely modular?
In other words, can the same code run faster against more data if coded to scale to resources, without the equivalent of lots of #IFDEF parallel code paths?
The Cell's techniques for exploiting the parallel pipeline of DSPs/PPCs is notoriously novel and complex. The IBM Linux/gcc toolchain is supposed to wrap that in a familiar API and build process.
The question is what's the effort/cost in porting PS3 to CBE? Is the savings of prototyping on PS3 worth the expense of porting to CBE, if they're that different? Considering the differential gradient of CBE cost by the time the PS3 prototype is ready to port.
Apparently, you can now develop Linux apps on PS3. High-end Cell machines, like a Cell Blade CBE from IBM, cost about $20-50K. And those CBEs are not really finished, stable HW architectures, and are in short supply, making their OS/SW environment very changeable.
A good strategy is to start developing on a PS3, while CBE HW catches up. If development takes 6-9 months, by then the CBEs will be cheaper, more stable, better understood by both their vendors and the community for getting support and working around weak links. And that time can be used to fundraise and team recruit around a PS3 prototype.
But the $64,000 Question (literally) is what it takes to port a PS3 Linux app to CBE Linux. Does anyone know yet? Whitepaper? Walking/talking expert for hire/bribe? If porting a PS3/Linux app to CBE/Linux is harder than porting an embedded x86 app to a Xeon, or an embedded R6000 to a multiproc R6000 server, then it might be worth it just to wait to start on the CBEs when they're ready. Though a PS3 running a supercomputer DSP app prototype could be cool enough to be worth the whole trouble, anyway.
Objects don't have to be C++ objects. They can be just class blueprints inherited from other classes, for instantiated objects, which are just related logic and the data accessed.
Your SMEQL looks a lot like lisp.
Something like object lisp for large collections of multidimensional (even asymmetric) objects could bring benefits of encapsulation/reuse and relations to a syntax that better reflects both the data model and the sequence of operations, in rules like policies. A dataflow version would be easy to read, debug and maintain.
SW platform development always features a tradeoff between general purpose APIs and optimized performance engines. Databases are like this. The economic advantages for everyone in using an API even as awkward and somewhat inconsistent as SQL are more valuable than the lost performance in the fundamental relational/query model.
But it doesn't have to be that way. SQL can be retained as an API, but different storage/query engines can be run under the hood to better fit different storage/query models for different kinds of data/access. A better way out would be a successor to SQL that is more like a procedural language for objects with all operators/functions implicitly working on collections like tables. Yes, something like object lisp, best organized as a dataflow with triggers and events. So long as SQL can be automatically compiled into the new language, and back, for at least 5 years of peaceful coexistence.
When you get some progress completed, open the project on SourceForge, maybe pointing to it on one of the Mindstorms programmer sites. The rest of us want to help.
"Flamebait" because I want MS to pay the costs for its profits, instead of the public paying it? Or because I said the NSA is violating our privacy rights, or that they're reading these messages? Only the trollMods know.
Or maybe they don't - maybe they're just NSA trollModBots suppressing as much as they can without tipping us off. Either way they're traitors to America.
I know that Stallman champions zero-price software, and only tangentially open source software. So? I prefer the open source to the zero price. And my suggestion operates according to both, unless you can find somewhere I suggested charging the public to listen to the audio they're transcribing for free. I don't know whether Stallman would prefer to get emails asking him to correct what he thinks he said in an old audio recording over a wiki. Why don't you email him and ask him?
What about just some generic artificial platelets that clot bloody wounds like the natural platelets? Platelets are by far the most short-supplied blood product that constrains blood products. Every serious trauma patient who gets into a hospital quickly exhausts their own platelets, and consumes easily a half-dozen donors, usually triple or more the amount of red blood cells they consume.
Anticancer targeted platelets are a great advance. But many times as many people need the simpler generic stuff. Before pharmacos get paid lots of public money for the anticancer platelets they'll surely patent for maximum profit (after heavily subsidized and risk-mitigated development), they should produce the generic platelets that aren't as profitable, but help save many more people.
There's Bluetooth for Roomba, and the Lego Mindstorms uses it too. How about some swarm tech that uses say, a single Perl module to control a mixed swarm of Mindstorm and Roomba devices?
Better get the command security right, or robotwars will have a whole network battle raging among the clashing droids for control of all the marbles.
I'm unhappy about the corporate welfare system. The (no) taxes are part of it. The subsidies are another part. That's what I wrote my post about. The other big part is the monopoly protection, like when MS is found a monopoly in a giant court case, but continues to operate as a monopoly with practically no changes.
"The government" likes it that way because of the bribes and lobbying. Jack Abramoff, the face card in the criminal Republican lobbyist deck representing the dozen years of the Republican rule, including the tax code (updated from the 1986 Reagan tax code), got started at Preston Gates , the law firm run by Bill Gates III's father, Bill Gates Jr. If you don't think that "coincidence" has anything to do with Microsoft not paying taxes in the 1990s Bubble, then not being broken up as a monopoly in Bush's 2000s, then the voodoo is working on you.
Sure, I provided a bunch of articles. Some were in fact written over 6 years ago, like the one I just quoted, telling the facts about the period just passed, in which MS paid no taxes. By your logic, Bill Clinton was not the president 1997-2000, because articles documenting that he was were written over 6 years ago.
While "no taxes" would be an exaggeration about the last 3 years (though I wouldn't be surprised if they didn't pay those taxes either, except in a report for suckers), because 8.6% does not in fact equal zero, you are claiming they paid something like 1/3, when they paid less than 1/4 of what you claim. What
So while "no taxes" is an exaggeration of the past 3 years, the point I made is that MS is not paying it's way. Since 8.6% for 3 years and 0% (or negligibly close to it) for the late 1990s is so much closer to the "no taxes" I said, and I'm talking about MS getting fat corporate subsidies instead of even just paying their fair share, I'm not going to argue with you any more. Because you're splitting hairs when you're right, and making no sense when you're wrong.
The new twist Microsoft (Nasdaq: MSFT) added several years ago was to deduct from its taxable corporate income the difference between the amount employees paid it to buy the shares and the amount the shares are worth on the open market. The company's employees do get taxed on this amount (when they exercise their options and buy the stock), so according to the IRS they received taxable income from their employer, and the company can deduct it as a salary expense. Even though it wasn't a cash expense, it's still deductible. Issue enough stock, and a company can shift its entire corporate tax burden to its employees and wind up paying no taxes on its own income.
Microsoft was the first company to achieve tax-free status.
My unhappiness at filling in your pit of credulity is shrinking with every post. Try reading some of the citations before arguing with my summary using only Microsoft's cover story.
After all this time, money spent by IBM defending/pursuing, and all the defining issues raised, I don't want SCO dying before a precedent verdict is set. The best justice will be for SCO to spend itself bankrupt pursuing this frivolous lawsuit, its frivolous lawyers getting stiffed and wasting more time as creditors in bankruptcy court, and Linux proven free of the FUD SCO has produced as its flagship product. Either way, watching the speculators betting on SCO's stock rising on blackmail is fun, but satisfaction lies in proving the facts about how Linux is free.
I hate to burn your satisfaction, but MS had annual income much more than $15.7BILLION the past 4 years. It took in an average over $40B, paying under $4.7B, or under 8.6% in taxes. In the late 1990s, MS paid practically no taxes.
How about posting audio streams/downloads of all Stallman recordings, and accepting publicly submitted transcripts on a Wiki? Let the community decide what Stallman said, including comments by Stallman. Such a project could be completed for cheap, fairly quickly - the open source way.
Of course I want the NSA I pay for and depend on to protect me working to make Vista safer. Because Vista is part of the security environment, eventually the biggest part. It's such a threat to Americans' security that NSA should be able to require MS to let NSA help secure it.
The problem is that NSA costs money to operate. Tax money. Tax money that Microsoft doesn't pay. Microsoft cuts costs by ignoring security whenever it can (most of the time). While raking in literally untold $BILLIONS in profits. Now their security work is being subsidized by free work by NSA. So I'm paying for Microsoft to be able to brag that "Vista is secured by the NSA", which will increase its sales. I doubt that JoeSchmoeLinux, Inc, gets Microsoft's attention from the NSA, even proportionate to the benefit it would provide.
Of course I want NSA helping secure Vista, because that makes us safer, NSA's job. I just want Microsoft to pay its share of providing it that benefit. That would mean paying at least its share of taxes, something like $30BILLION or more a year, plus a fee for "special treatment" by the NSA that other companies don't get. Or at least make NSA's Vista work public, so the public can benefit, including MS competitors even if they're small.
That would also make the work even more secure, instead of relying on security through obscurity which MS prefers to protect its profits more than its operations. The secret/proprietary work also lets the NSA hide literal spyware in Vista, which is absolutely unacceptable. Since the NSA is busy spying on Americans right now, including filtering this message, I am now in the business of paying the NSA to spy on me, in violation of my inalienable rights, by putting their spyware into commercial code subsidized by my "NSA fees".
When MS pays NSA to produce "open security" that others can share, I'll be safer. Until then, it's all part of the increasingly unsustainable threat to my security at every level, from economics to privacy to tyranny.
Mention "liar warmongers", even referring to WWII, and the trollMods dispatch their brownshirts to gestapo your posts. I guess when law is outlawed, then only outlaws make war. And outlaw wannabes make war on Slashdot posts. War in which the first casualty is the truth.
Any reasonable security model protecting online currency has to be proof against tampering with the client, and instead secure the server. Even before the client's source was opened, people were reverse-engineering the network protocol, against which work there is practically no protection (especially with an unencrypted client/server protocol). Currency and money supply must be controlled at the server, or counterfeiting can't be prevented. Even in the real world, the only proof against counterfeit is the extreme complexity of physical currency features, which are not only virtually impossible in digital currency, but directly contradict digital currency's best features.
The real question about Second Life's "Treasury security" with this release is whether Linden Labs has first secured their servers against currency attacks. That sounds like a fascinating research paper for the current term of a CS or even economics, student or researcher.
Slashdot Mirror
Hey, we fingerprint all Canadians entering the US, soon to demand (RFID) passports from all of them, while our NSA is tapping their phones and email. Who knows what our CIA does up there.
Since Americans are allergic to Canadian pennies now worth almost as much as ours, and dump them whenever we see them on our side of the border, these RFID trackers are relatively pretty benign.
Maybe if we just all wound down the BS simcurity that pretends to protect us, and instead actually just destroyed us some Qaeda recruitment cells, the US dollar would become strong enough again that we wouldn't bother schlepping their Canadian coins back home, like pocket lint.
That all sounds like work for the optimization phase of the compiler. Is the IBM Cell gcc that lamebrained?
I tend to agree. But how does putting your Java into stored procedures compare to the long/inefficient SQL? And is the Java stored procedure slowdown due to a worse JVM, more demand on the same host for both DB and JVM, or some other architectural bottleneck?
That's reassuring.
Does the PS3 API to the SPUs automatically (at compile time, or runtime) scale to the number of SPUs available? And is switching networking from PS3 to CBE infiniband completely modular?
In other words, can the same code run faster against more data if coded to scale to resources, without the equivalent of lots of #IFDEF parallel code paths?
The Cell's techniques for exploiting the parallel pipeline of DSPs/PPCs is notoriously novel and complex. The IBM Linux/gcc toolchain is supposed to wrap that in a familiar API and build process.
The question is what's the effort/cost in porting PS3 to CBE? Is the savings of prototyping on PS3 worth the expense of porting to CBE, if they're that different? Considering the differential gradient of CBE cost by the time the PS3 prototype is ready to port.
But will it accelerate my basic desktop by offloading rendering from CPU to GPU?
Apparently, you can now develop Linux apps on PS3. High-end Cell machines, like a Cell Blade CBE from IBM, cost about $20-50K. And those CBEs are not really finished, stable HW architectures, and are in short supply, making their OS/SW environment very changeable.
A good strategy is to start developing on a PS3, while CBE HW catches up. If development takes 6-9 months, by then the CBEs will be cheaper, more stable, better understood by both their vendors and the community for getting support and working around weak links. And that time can be used to fundraise and team recruit around a PS3 prototype.
But the $64,000 Question (literally) is what it takes to port a PS3 Linux app to CBE Linux. Does anyone know yet? Whitepaper? Walking/talking expert for hire/bribe? If porting a PS3/Linux app to CBE/Linux is harder than porting an embedded x86 app to a Xeon, or an embedded R6000 to a multiproc R6000 server, then it might be worth it just to wait to start on the CBEs when they're ready. Though a PS3 running a supercomputer DSP app prototype could be cool enough to be worth the whole trouble, anyway.
Fuck you. You have no brain.
Collections don't (necessarily) have an order.
Objects don't have to be C++ objects. They can be just class blueprints inherited from other classes, for instantiated objects, which are just related logic and the data accessed.
Your SMEQL looks a lot like lisp.
Something like object lisp for large collections of multidimensional (even asymmetric) objects could bring benefits of encapsulation/reuse and relations to a syntax that better reflects both the data model and the sequence of operations, in rules like policies. A dataflow version would be easy to read, debug and maintain.
SW platform development always features a tradeoff between general purpose APIs and optimized performance engines. Databases are like this. The economic advantages for everyone in using an API even as awkward and somewhat inconsistent as SQL are more valuable than the lost performance in the fundamental relational/query model.
But it doesn't have to be that way. SQL can be retained as an API, but different storage/query engines can be run under the hood to better fit different storage/query models for different kinds of data/access. A better way out would be a successor to SQL that is more like a procedural language for objects with all operators/functions implicitly working on collections like tables. Yes, something like object lisp, best organized as a dataflow with triggers and events. So long as SQL can be automatically compiled into the new language, and back, for at least 5 years of peaceful coexistence.
When you get some progress completed, open the project on SourceForge, maybe pointing to it on one of the Mindstorms programmer sites. The rest of us want to help.
Moderation 0
50% Flamebait
50% Interesting
"Flamebait" because I want MS to pay the costs for its profits, instead of the public paying it? Or because I said the NSA is violating our privacy rights, or that they're reading these messages? Only the trollMods know.
Or maybe they don't - maybe they're just NSA trollModBots suppressing as much as they can without tipping us off. Either way they're traitors to America.
I know that Stallman champions zero-price software, and only tangentially open source software. So? I prefer the open source to the zero price. And my suggestion operates according to both, unless you can find somewhere I suggested charging the public to listen to the audio they're transcribing for free. I don't know whether Stallman would prefer to get emails asking him to correct what he thinks he said in an old audio recording over a wiki. Why don't you email him and ask him?
Will this project produce a driver that will let my Inspiron 8000 offload window rendering from the CPU to the nVidia GPU, so Ubuntu runs faster?
Or is it just a way to get higher FPS on 3D games running on nVidia HW?
What about just some generic artificial platelets that clot bloody wounds like the natural platelets? Platelets are by far the most short-supplied blood product that constrains blood products. Every serious trauma patient who gets into a hospital quickly exhausts their own platelets, and consumes easily a half-dozen donors, usually triple or more the amount of red blood cells they consume.
Anticancer targeted platelets are a great advance. But many times as many people need the simpler generic stuff. Before pharmacos get paid lots of public money for the anticancer platelets they'll surely patent for maximum profit (after heavily subsidized and risk-mitigated development), they should produce the generic platelets that aren't as profitable, but help save many more people.
There's Bluetooth for Roomba, and the Lego Mindstorms uses it too. How about some swarm tech that uses say, a single Perl module to control a mixed swarm of Mindstorm and Roomba devices?
Better get the command security right, or robotwars will have a whole network battle raging among the clashing droids for control of all the marbles.
I'm unhappy about the corporate welfare system. The (no) taxes are part of it. The subsidies are another part. That's what I wrote my post about. The other big part is the monopoly protection, like when MS is found a monopoly in a giant court case, but continues to operate as a monopoly with practically no changes.
"The government" likes it that way because of the bribes and lobbying. Jack Abramoff, the face card in the criminal Republican lobbyist deck representing the dozen years of the Republican rule, including the tax code (updated from the 1986 Reagan tax code), got started at Preston Gates , the law firm run by Bill Gates III's father, Bill Gates Jr. If you don't think that "coincidence" has anything to do with Microsoft not paying taxes in the 1990s Bubble, then not being broken up as a monopoly in Bush's 2000s, then the voodoo is working on you.
Sure, I provided a bunch of articles. Some were in fact written over 6 years ago, like the one I just quoted, telling the facts about the period just passed, in which MS paid no taxes. By your logic, Bill Clinton was not the president 1997-2000, because articles documenting that he was were written over 6 years ago.
While "no taxes" would be an exaggeration about the last 3 years (though I wouldn't be surprised if they didn't pay those taxes either, except in a report for suckers), because 8.6% does not in fact equal zero, you are claiming they paid something like 1/3, when they paid less than 1/4 of what you claim. What
So while "no taxes" is an exaggeration of the past 3 years, the point I made is that MS is not paying it's way. Since 8.6% for 3 years and 0% (or negligibly close to it) for the late 1990s is so much closer to the "no taxes" I said, and I'm talking about MS getting fat corporate subsidies instead of even just paying their fair share, I'm not going to argue with you any more. Because you're splitting hairs when you're right, and making no sense when you're wrong.
As explained in one of the first of the many articles to which I linked in the post to which you first responded, Microsoft avoided paying taxes in the late 1990s:
My unhappiness at filling in your pit of credulity is shrinking with every post. Try reading some of the citations before arguing with my summary using only Microsoft's cover story.
After all this time, money spent by IBM defending/pursuing, and all the defining issues raised, I don't want SCO dying before a precedent verdict is set. The best justice will be for SCO to spend itself bankrupt pursuing this frivolous lawsuit, its frivolous lawyers getting stiffed and wasting more time as creditors in bankruptcy court, and Linux proven free of the FUD SCO has produced as its flagship product. Either way, watching the speculators betting on SCO's stock rising on blackmail is fun, but satisfaction lies in proving the facts about how Linux is free.
I hate to burn your satisfaction, but MS had annual income much more than $15.7BILLION the past 4 years. It took in an average over $40B, paying under $4.7B, or under 8.6% in taxes. In the late 1990s, MS paid practically no taxes.
Outrageous.
How about posting audio streams/downloads of all Stallman recordings, and accepting publicly submitted transcripts on a Wiki? Let the community decide what Stallman said, including comments by Stallman. Such a project could be completed for cheap, fairly quickly - the open source way.
Of course I want the NSA I pay for and depend on to protect me working to make Vista safer. Because Vista is part of the security environment, eventually the biggest part. It's such a threat to Americans' security that NSA should be able to require MS to let NSA help secure it.
The problem is that NSA costs money to operate. Tax money. Tax money that Microsoft doesn't pay. Microsoft cuts costs by ignoring security whenever it can (most of the time). While raking in literally untold $BILLIONS in profits. Now their security work is being subsidized by free work by NSA. So I'm paying for Microsoft to be able to brag that "Vista is secured by the NSA", which will increase its sales. I doubt that JoeSchmoeLinux, Inc, gets Microsoft's attention from the NSA, even proportionate to the benefit it would provide.
Of course I want NSA helping secure Vista, because that makes us safer, NSA's job. I just want Microsoft to pay its share of providing it that benefit. That would mean paying at least its share of taxes, something like $30BILLION or more a year, plus a fee for "special treatment" by the NSA that other companies don't get. Or at least make NSA's Vista work public, so the public can benefit, including MS competitors even if they're small.
That would also make the work even more secure, instead of relying on security through obscurity which MS prefers to protect its profits more than its operations. The secret/proprietary work also lets the NSA hide literal spyware in Vista, which is absolutely unacceptable. Since the NSA is busy spying on Americans right now, including filtering this message, I am now in the business of paying the NSA to spy on me, in violation of my inalienable rights, by putting their spyware into commercial code subsidized by my "NSA fees".
When MS pays NSA to produce "open security" that others can share, I'll be safer. Until then, it's all part of the increasingly unsustainable threat to my security at every level, from economics to privacy to tyranny.
Moderation 0
50% Troll
50% Insightful
Mention "liar warmongers", even referring to WWII, and the trollMods dispatch their brownshirts to gestapo your posts. I guess when law is outlawed, then only outlaws make war. And outlaw wannabes make war on Slashdot posts. War in which the first casualty is the truth.
Any reasonable security model protecting online currency has to be proof against tampering with the client, and instead secure the server. Even before the client's source was opened, people were reverse-engineering the network protocol, against which work there is practically no protection (especially with an unencrypted client/server protocol). Currency and money supply must be controlled at the server, or counterfeiting can't be prevented. Even in the real world, the only proof against counterfeit is the extreme complexity of physical currency features, which are not only virtually impossible in digital currency, but directly contradict digital currency's best features.
The real question about Second Life's "Treasury security" with this release is whether Linden Labs has first secured their servers against currency attacks. That sounds like a fascinating research paper for the current term of a CS or even economics, student or researcher.