Your summary suggests that the ethical considerations involved here have to do with the availability of other cats, not anything to do with the actual cloning process. That said, do you support human cloning since the pool of adoptable babies is so small?
You don't get it. The poster's comment didn't oppose cloning qua cloning, but rather the questionable ethics of going to a lot of trouble to create a creature when so many others of that same sort are destroyed for lack of owners.
Of course, I don't think the point of the CopyCat project was to develop a revolutionary method of cat production. IDNRTA* but I'm guessing it had more to do with biology than somebody's dream of a chain of cat factories.
* I Did Not Read The Article.
Re:And why not turn lead into gold while we're at
on
Inside Intel
·
· Score: 2
A modern foundry could put as much cache as it wanted on to a chip (look at HP's most recent chip for an example) - but because of architectural tradeoffs, this isn't always a good idea.
Oh, I don't know about that. The PA-8700 has 2.25 MB of L2 cache, which is okay I suppose. The MIPS R14000 processors in the SGI Origin 3000 series have 8 MB of L2 cache per CPU, and they do pretty well, to put it mildly. I think your assertion that large secondary caches aren't always a good idea sounds a little weak.
Damn you people. You made me have to go look up that Powderhorn silo.
Wanna hear something funny? Our first quote on a Powderhorn (one silo, 6,000 slots) had something like $240,000 in it for the tapes. A quarter of a million dollars just for the tapes themselves. Unbelievable.
Needless to say, I haven't asked for itemized quotes from StorageTek since then. I just don't want to know.
For example, if you're a trademark holder and have evidence that a third party is selling counterfeit goods, a warrant can be obtained to send in the Marshalls for a seizure.
Which is, of course, a long way from what the original poster was saying. In your example, the seizure is of goods that themselves are illegal, and is to prevent the sale of them, which would be an illegal act. The seizure of Microsoft's source code for Windows (not an illegal good) to make it available for analysis (not a legally mandated act) would be completely unjustified and would never happen.
The exception that proves the rule. The judge who issued the warrant for the search of SJG offices and seizure of equipment therein overstepped his bounds. The government was found at fault and forced to pay damages and legal fees.
However, thanks to a bunch of zealots who associate themselves with the EFF, this case was blown completely out of proportion. It was a simple, albeit dramatic and traumatic, instance of illegal seizure. Advocates on the side of SJG and the EFF tried to turn it into a privacy/wiretap case by claiming that the seizure of SJG computers was tantamount to the illegal interception of electronic communications, prohibited by wiretap laws. The Fifth Circuit, of course, saw straight through that noise.
The point is that illegal seizure is illegal seizure. The fact that it has happened in the past doesn't mean it's policy.
As for the bench warrant, all it requires is that the judge believe that the most expedient way to resolve the matter is to seize those servers.
That's simply false. Judges can't go issuing warrants whenever they feel like it. There are legally and constituionally mandated guidelines that must be followed in order for the warrant to be legal.
There's no question that the most expedient way to get anything is just to take it. That doesn't mean our justice system operates that way.
If the court orders this and selects competent experts, they aren't going to wait while MS prepares a very special set of media. They will send in Federal Marshals to take control of the MS servers containing all source code for anything that ends up on the Windows OEM disc and copy *everything* on them. MS won't regain access to its systems until the experts can build the Windows OEM disc on their own systems.
You've been reading too many newsgroups. What you describe is seizure, and it's completely inappropriate in a civil matter. In order to get authorization to send in federal marshalls to seize property like that, a bench warrant must be issued. To get a bench warrant, the judge has to be convinced that there's evidence there that's relevant to a criminal investigation and that couldn't be gotten any other way.
In other words, if a judge believed that Microsoft's computers had information on them about who mailed Anthrax to those senators last fall, and that judge believed that Microsoft had been given an opportunity to turn the evidence over and had refused or that the evidence was in danger of being tampered with or destroyed, then and only then would you see a bench warrant issued for the sort of seizure you describe.
This is completely different from any action taken by the police in cooperation with the BSA. In those instances (like the Rotter raid last year), the police were convinced by the BSA that criminal activity was taking place, that the activity was very significant, and that any approach other than seizure would result in evidence being destroyed.
There are larger tape libraries than the Powderhorn. The ADIC AML/E can store 70,000+ tapes, all accessible via robotics.
You're thinking of the AML/2. The AML/E was discontinued some time ago, I believe. It topped out at 10,386 tapes. The AML/2 can store up to 76,608 tapes, or 5,184 TB. Very big, true.
The Powderhorn with ACSLS control can hold up to 144,000 tapes. That's 9,640 TB. Biggest tape library in the world, as far as I know.
The PetaSite from Sony is pretty frigging big, too, but I don't know exactly how big. I'm not sure how many racks you can chain together with their extension robotics.
True story: originally Sony Broadcast and Professional was planning to market the DMS 8000 series of tape systems in the US and Canada under the name "PetaFile." Say it out loud and you'll know why they didn't.
First, you only have a single generation of backup.
The system we're talking about is itself an archive. It doesn't get backed up, per se; it just gets replicated.
Given that our archive is going to be about 80 TB when installed, your "multiple generations on an off-site rotation" idea, besides being totally unnecessary, is frankly impractical in the extreme.
The other is the single failure point of the telecom line.
At about US$53,000 per month for dark fibre, the penalty clauses in the contract with the telco are such that if the connection goes down for more than one minute, we get paid so much money that we can pay off our penalty clauses and, believe it or not, actually make a small profit.
You make a good point about tape, but it's not really relevant to our situation. For instance, we could use a StorageTek Powderhorn LSM to replicate our archive, but we'd either have to put on staff to manage thousands of 9840 tapes off line (the LSM only holds 6,000 tapes) or buy multiple LSMs. The amount of floor space that would require would cost us much more than the dark fibre.
And he said... I'll email you the details as soon as we end the call.
If I have to interact with other humans, I'm most productive when I'm at my office with my headset on, and my email and IM programs open. It's like having three channels open at once, and it's amazing how well it works. Any phone that would let me talk or email, but not both at the same time, would eventually be frustrating. Maybe not much and maybe not often, but frustrating anyway.
If they are inert then how are O2 & CO2 soluble in it?
You know... that's a good question. I'm not sure. Maybe calling PFC "inert" is an overstatement. Maybe it's better to call it nonconductive and noncorrosive. Or maybe gases are only soluble in emulsions of PFC.
My favourite was always the Cray 2 which used artificial human blood plasma as a coolant....
Oh, man, do I even need to take the time to correct this?
The Cray 2, like several successors including the C90 and T90, used liquid fluorocarbon as a coolant. This is true.
To say that liquid fluorocarbon is artificial blood plasma is simply false. There are several commercial products that can be used as blood substitutes-- such as Oxycyte-- but these are oxygen-carrying perfluorocarbon (PFC) emulsions. These products are about as similar to the liquid coolant used in the Cray 2 as scrambled eggs are to mayonnaise.
The coolants used in the various Crays, plus lots of other electronic systems, were all pure perfluorocarbon liquids, like Fluorinert, which is a commercial product produced and sold by 3M. They're good choices for immersion cooling because they're chemically inert. Ironically, most of them (like FC-87, fer instance) have boiling points well below that of water; FC-87's is around 30C. They're useful anyway because they're dense and have specific heats about four times less than water's.
The various artificial blood products, though, are PFC emulsions, in which microscopic droplets of PFC are suspended in a saline solution. These liquids can be used because gases like oxygen and CO2 are highly soluble in PFC. Since the droplets of PFC are about 70 times smaller than red blood cells, it's very easy for them to act like RBCs in the blood stream for gas transfer.
The fact that Cray's are liquid cooled is neat enough without messing it up with misinformation.
If your IEEE 1394 computer/laptop does not supply power and your device requires it to operate, there are cables which are 4pin + power(converter) to 6pin which solve this problem.
While this may be true, my point is still important: if you buy a laptop hoping to use portable FireWire devices, you may be very disappointed by the Sony model. Portable FireWire devices must either run on batteries or take power over the FireWire cable, and Sony's iLink implementation can't provide power. Caveat emptor.
It has integrated firewire, in addition to all the other goodies.
No, it has integrated iLink, not integrated FireWire. FireWire (the name, I mean) applies to the six-pin implementation of IEEE 1394; FireWire can carry power as well as data on the cable.
On the other hand, iLink applies only to the 4-pin implementation. Data only on that cable, no power. That's a big problem for laptops; why else have an IEEE 1394 connection at all if you can't send power over it? It's useless unless you're near a wall socket.
Kind of a shame, really. Other than this fact, it's a fairly decent looking laptop. Bluetooth has definitely got "ooh, neat!" going for it.
EMC, Hitachi, HP, and SUN all offer ways to create an internal point in time disk copy of production data.
Off topic (again), but FYI the HP and Sun storage solutions are OEM'd by Hitachi.
The Sun StorEdge 9910 and StorEdge 9960 are HDS 9910 and 9960 with different skins on them. Likewise, the HP XP48 is an HDS 9910 and the XP512 is the HDS 9960, so named for the number of disks in each (48 and 512).
Floods, tornadoes, fires, etc happen. Sometimes people fly planes into buildings. When that happens, tapes are the only thing that keeps your business in business.
I know this is completely off topic, but sometimes tape just isn't cost effective, particularly when you figure in the costs of manually storing and maintaining a library of data tapes in a vault somewhere. (Most of that cost is in head count: you have to pay somebody to do that work, and that's not a $19,000 a year job.)
We're presently doing the cost analysis on a kind of radical idea. We're storing many terabytes of data in a data center in San Jose, California. The data center is as good as it can be, but there's still the danger (however unlikely) of earthquake or some other drastic event.
Rather than trying to back everything up to data tape, we've gotten pricing from a telco on a dark fiber link between the San Jose data center and another data center somewhere in Colorado-- can't remember where. Since we're already putting an HDS 9960 in the San Jose data center, we can put an identical one in Colorado and use the 9960's internal "NanoCopy" software to keep them in sync.
Believe it or not, it's working out to be more cost effective. One of the big reasons is that keeping that much tape on-line in a data center would require a StorageTek PowderHorn silo, and data center floor space is expensive. The difference in cost between the floor space and the dark fiber is so small that they cancel out.
Like I said, I realize this is light-years away from what the poster was originally asking about, but it's kinda neat nonetheless.
And neither is Cray. Just a few days ago Cray announced over $9 million in orders for their SV1ex, and their SV2 project is coming along nicely. They're not exactly making a profit, but they're not in a crazy death spiral yet, either.
There's a big difference between being a niche player and being a dead company.
I keep wondering why big companies like HP and Sun choose linux, instead of freeBSD.
I can't speak about Sun or HP, but some time ago SGI started working on tons of stuff for Linux, including but not limited to their XFS filesystem. More info: http://oss.sgi.com.
It's pretty clear, when you think about it, why they chose to release their valuable technologies for Linux rather than BSD: the GPL. GPL is, contrary to what Microsoft might say, a pretty business-friendly license. If a business spends billions of dollars over decades developing, say, XFS, then releases it under a BSD-style license, then anybody can incorporate that technology into their commercial products for free.
On the other hand, releasing XFS for Linux under the GPL means SGI gets to say they have XFS on IRIX and also on Linux, but it does not mean that Sun can put XFS in Solaris or whatever.
You can't make any money, directly, off of producing GPL'd code, but you can at least prevent your competitors from benefiting from your work.
I just cannot see Sun replacing Solaris on their high-end multi-processor machines... Or at least not until Linux scales equally well:-)
It's not just scalability. There are features in Solaris-- although I'm not familiar enough with them to get specific-- that allow the customer to partition out failed hardware components for repair or replacement without taking the system down. Linux has no features even remotely like that.
Of course, the big problem was always getting the system to recognize the new hardware component again. If I remember right, that still requires a reboot....
You're being too simplistic, probably because of my hurried and incomplete example. Semantic searching is most useful in the "general to more specific" instances. While searching for "orange" shouldn't necessarily search for "fruit," as you point out, it's very likely that a search for "fruit" should key off of "orange."
This is really much more applicable to concept searching than it is to simple text indexing. (Mmm... SimpleText...)
For example, if a catalogger is describing a painting, she might use the word "orange" to describe the subject of a still life. In that instance, the painting's metadata structures would reflect the fact that "orange" in the subject field must be the concrete noun "orange," which is a specific instance of the concrete noun "fruit." The idea being that a user could search for "fruit" and get a hit on a painting that has been described as "still life with oranges."
So maybe my idea isn't all that applicable to Google after all. Hell, I'm not even sure I'm on topic any more.;-)
So now when I search for "Ornage" it asks me "Did you mean orange." I guess Google could extend this if it was hooked up to WordNet - "Did you mean orange the fruit, orange the color, orange the tree, or orange the river"
EXACTLY! Like that scene in 2010:
CHANDRA
I would like to open a new file. Here is the name for it. [types "phoenix"] Do you know what that means?
SAL
There are twenty five references in the current encyclopedia.
CHANDRA
Which one do you think is relevant?
SAL
The tutor of Achilles?
CHANDRA
That's very interesting, I didn't know that one. Try again.
SAL
A fabulous bird, re-born from the ashes of its earlier life.
How about adding the option to have google understand what I *mean* to search for, not what I tell it to search for.
You might have been kidding, but you've got a really good idea there.
How about semantic searching: equip Google with a database that organizes words in a relational hierarchy from the general to the specific. For example, "orange" is a more specific form of "fruit," and also a more specific form of "color."
When you search for "orange," Google might also have the ability to search for "fruit" and "color," depending on how broad you want your search to be.
Slashdot Mirror
Your summary suggests that the ethical considerations involved here have to do with the availability of other cats, not anything to do with the actual cloning process. That said, do you support human cloning since the pool of adoptable babies is so small?
You don't get it. The poster's comment didn't oppose cloning qua cloning, but rather the questionable ethics of going to a lot of trouble to create a creature when so many others of that same sort are destroyed for lack of owners.
Of course, I don't think the point of the CopyCat project was to develop a revolutionary method of cat production. IDNRTA* but I'm guessing it had more to do with biology than somebody's dream of a chain of cat factories.
* I Did Not Read The Article.
A modern foundry could put as much cache as it wanted on to a chip (look at HP's most recent chip for an example) - but because of architectural tradeoffs, this isn't always a good idea.
Oh, I don't know about that. The PA-8700 has 2.25 MB of L2 cache, which is okay I suppose. The MIPS R14000 processors in the SGI Origin 3000 series have 8 MB of L2 cache per CPU, and they do pretty well, to put it mildly. I think your assertion that large secondary caches aren't always a good idea sounds a little weak.
Damn you people. You made me have to go look up that Powderhorn silo.
Wanna hear something funny? Our first quote on a Powderhorn (one silo, 6,000 slots) had something like $240,000 in it for the tapes. A quarter of a million dollars just for the tapes themselves. Unbelievable.
Needless to say, I haven't asked for itemized quotes from StorageTek since then. I just don't want to know.
For example, if you're a trademark holder and have evidence that a third party is selling counterfeit goods, a warrant can be obtained to send in the Marshalls for a seizure.
Which is, of course, a long way from what the original poster was saying. In your example, the seizure is of goods that themselves are illegal, and is to prevent the sale of them, which would be an illegal act. The seizure of Microsoft's source code for Windows (not an illegal good) to make it available for analysis (not a legally mandated act) would be completely unjustified and would never happen.
Steve Jackson Games?
The exception that proves the rule. The judge who issued the warrant for the search of SJG offices and seizure of equipment therein overstepped his bounds. The government was found at fault and forced to pay damages and legal fees.
However, thanks to a bunch of zealots who associate themselves with the EFF, this case was blown completely out of proportion. It was a simple, albeit dramatic and traumatic, instance of illegal seizure. Advocates on the side of SJG and the EFF tried to turn it into a privacy/wiretap case by claiming that the seizure of SJG computers was tantamount to the illegal interception of electronic communications, prohibited by wiretap laws. The Fifth Circuit, of course, saw straight through that noise.
The point is that illegal seizure is illegal seizure. The fact that it has happened in the past doesn't mean it's policy.
As for the bench warrant, all it requires is that the judge believe that the most expedient way to resolve the matter is to seize those servers.
That's simply false. Judges can't go issuing warrants whenever they feel like it. There are legally and constituionally mandated guidelines that must be followed in order for the warrant to be legal.
There's no question that the most expedient way to get anything is just to take it. That doesn't mean our justice system operates that way.
MS was sued under the civil provisions of Sherman. There are criminal provisions, but they haven't been exercised.
Oh, and me without any mod points.
If the court orders this and selects competent experts, they aren't going to wait while MS prepares a very special set of media. They will send in Federal Marshals to take control of the MS servers containing all source code for anything that ends up on the Windows OEM disc and copy *everything* on them. MS won't regain access to its systems until the experts can build the Windows OEM disc on their own systems.
You've been reading too many newsgroups. What you describe is seizure, and it's completely inappropriate in a civil matter. In order to get authorization to send in federal marshalls to seize property like that, a bench warrant must be issued. To get a bench warrant, the judge has to be convinced that there's evidence there that's relevant to a criminal investigation and that couldn't be gotten any other way.
In other words, if a judge believed that Microsoft's computers had information on them about who mailed Anthrax to those senators last fall, and that judge believed that Microsoft had been given an opportunity to turn the evidence over and had refused or that the evidence was in danger of being tampered with or destroyed, then and only then would you see a bench warrant issued for the sort of seizure you describe.
This is completely different from any action taken by the police in cooperation with the BSA. In those instances (like the Rotter raid last year), the police were convinced by the BSA that criminal activity was taking place, that the activity was very significant, and that any approach other than seizure would result in evidence being destroyed.
There are larger tape libraries than the Powderhorn. The ADIC AML/E can store 70,000+ tapes, all accessible via robotics.
You're thinking of the AML/2. The AML/E was discontinued some time ago, I believe. It topped out at 10,386 tapes. The AML/2 can store up to 76,608 tapes, or 5,184 TB. Very big, true.
The Powderhorn with ACSLS control can hold up to 144,000 tapes. That's 9,640 TB. Biggest tape library in the world, as far as I know.
The PetaSite from Sony is pretty frigging big, too, but I don't know exactly how big. I'm not sure how many racks you can chain together with their extension robotics.
True story: originally Sony Broadcast and Professional was planning to market the DMS 8000 series of tape systems in the US and Canada under the name "PetaFile." Say it out loud and you'll know why they didn't.
First, you only have a single generation of backup.
The system we're talking about is itself an archive. It doesn't get backed up, per se; it just gets replicated.
Given that our archive is going to be about 80 TB when installed, your "multiple generations on an off-site rotation" idea, besides being totally unnecessary, is frankly impractical in the extreme.
The other is the single failure point of the telecom line.
At about US$53,000 per month for dark fibre, the penalty clauses in the contract with the telco are such that if the connection goes down for more than one minute, we get paid so much money that we can pay off our penalty clauses and, believe it or not, actually make a small profit.
You make a good point about tape, but it's not really relevant to our situation. For instance, we could use a StorageTek Powderhorn LSM to replicate our archive, but we'd either have to put on staff to manage thousands of 9840 tapes off line (the LSM only holds 6,000 tapes) or buy multiple LSMs. The amount of floor space that would require would cost us much more than the dark fibre.
And he said... I'll email you the details as soon as we end the call.
If I have to interact with other humans, I'm most productive when I'm at my office with my headset on, and my email and IM programs open. It's like having three channels open at once, and it's amazing how well it works. Any phone that would let me talk or email, but not both at the same time, would eventually be frustrating. Maybe not much and maybe not often, but frustrating anyway.
If they are inert then how are O2 & CO2 soluble in it?
;-)
You know... that's a good question. I'm not sure. Maybe calling PFC "inert" is an overstatement. Maybe it's better to call it nonconductive and noncorrosive. Or maybe gases are only soluble in emulsions of PFC.
You caught me. Good question.
you should know that mayonnaise are made from eggs!
Yes... and thus endeth the analogy. Scrambled eggs are eggs. Mayonnaise is an emulsion of eggs and air suspended in an oil matrix.
Fluorinert is made of PFC. Oxycyte and other artificial blood compounds are emulsions of PFC in a saline matrix.
I choose my comparison pretty carefully.
My favourite was always the Cray 2 which used artificial human blood plasma as a coolant....
Oh, man, do I even need to take the time to correct this?
The Cray 2, like several successors including the C90 and T90, used liquid fluorocarbon as a coolant. This is true.
To say that liquid fluorocarbon is artificial blood plasma is simply false. There are several commercial products that can be used as blood substitutes-- such as Oxycyte-- but these are oxygen-carrying perfluorocarbon (PFC) emulsions. These products are about as similar to the liquid coolant used in the Cray 2 as scrambled eggs are to mayonnaise.
The coolants used in the various Crays, plus lots of other electronic systems, were all pure perfluorocarbon liquids, like Fluorinert, which is a commercial product produced and sold by 3M. They're good choices for immersion cooling because they're chemically inert. Ironically, most of them (like FC-87, fer instance) have boiling points well below that of water; FC-87's is around 30C. They're useful anyway because they're dense and have specific heats about four times less than water's.
The various artificial blood products, though, are PFC emulsions, in which microscopic droplets of PFC are suspended in a saline solution. These liquids can be used because gases like oxygen and CO2 are highly soluble in PFC. Since the droplets of PFC are about 70 times smaller than red blood cells, it's very easy for them to act like RBCs in the blood stream for gas transfer.
The fact that Cray's are liquid cooled is neat enough without messing it up with misinformation.
If your IEEE 1394 computer/laptop does not supply power and your device requires it to operate, there are cables which are 4pin + power(converter) to 6pin which solve this problem.
While this may be true, my point is still important: if you buy a laptop hoping to use portable FireWire devices, you may be very disappointed by the Sony model. Portable FireWire devices must either run on batteries or take power over the FireWire cable, and Sony's iLink implementation can't provide power. Caveat emptor.
It has integrated firewire, in addition to all the other goodies.
No, it has integrated iLink, not integrated FireWire. FireWire (the name, I mean) applies to the six-pin implementation of IEEE 1394; FireWire can carry power as well as data on the cable.
On the other hand, iLink applies only to the 4-pin implementation. Data only on that cable, no power. That's a big problem for laptops; why else have an IEEE 1394 connection at all if you can't send power over it? It's useless unless you're near a wall socket.
Kind of a shame, really. Other than this fact, it's a fairly decent looking laptop. Bluetooth has definitely got "ooh, neat!" going for it.
I run a major financial bank institution on a bunch of overclocked Athlon XP's. we use IDE RAID and linux 2.4.10.
That sound you hear is the rustling of ten million "withdrawl" slips being hurriedly filled out.
EMC, Hitachi, HP, and SUN all offer ways to create an internal point in time disk copy of production data.
Off topic (again), but FYI the HP and Sun storage solutions are OEM'd by Hitachi.
The Sun StorEdge 9910 and StorEdge 9960 are HDS 9910 and 9960 with different skins on them. Likewise, the HP XP48 is an HDS 9910 and the XP512 is the HDS 9960, so named for the number of disks in each (48 and 512).
Floods, tornadoes, fires, etc happen. Sometimes people fly planes into buildings. When that happens, tapes are the only thing that keeps your business in business.
I know this is completely off topic, but sometimes tape just isn't cost effective, particularly when you figure in the costs of manually storing and maintaining a library of data tapes in a vault somewhere. (Most of that cost is in head count: you have to pay somebody to do that work, and that's not a $19,000 a year job.)
We're presently doing the cost analysis on a kind of radical idea. We're storing many terabytes of data in a data center in San Jose, California. The data center is as good as it can be, but there's still the danger (however unlikely) of earthquake or some other drastic event.
Rather than trying to back everything up to data tape, we've gotten pricing from a telco on a dark fiber link between the San Jose data center and another data center somewhere in Colorado-- can't remember where. Since we're already putting an HDS 9960 in the San Jose data center, we can put an identical one in Colorado and use the 9960's internal "NanoCopy" software to keep them in sync.
Believe it or not, it's working out to be more cost effective. One of the big reasons is that keeping that much tape on-line in a data center would require a StorageTek PowderHorn silo, and data center floor space is expensive. The difference in cost between the floor space and the dark fiber is so small that they cancel out.
Like I said, I realize this is light-years away from what the poster was originally asking about, but it's kinda neat nonetheless.
Cray may be dead, but supercomputers aren't.
And neither is Cray. Just a few days ago Cray announced over $9 million in orders for their SV1ex, and their SV2 project is coming along nicely. They're not exactly making a profit, but they're not in a crazy death spiral yet, either.
There's a big difference between being a niche player and being a dead company.
I keep wondering why big companies like HP and Sun choose linux, instead of freeBSD.
I can't speak about Sun or HP, but some time ago SGI started working on tons of stuff for Linux, including but not limited to their XFS filesystem. More info: http://oss.sgi.com.
It's pretty clear, when you think about it, why they chose to release their valuable technologies for Linux rather than BSD: the GPL. GPL is, contrary to what Microsoft might say, a pretty business-friendly license. If a business spends billions of dollars over decades developing, say, XFS, then releases it under a BSD-style license, then anybody can incorporate that technology into their commercial products for free.
On the other hand, releasing XFS for Linux under the GPL means SGI gets to say they have XFS on IRIX and also on Linux, but it does not mean that Sun can put XFS in Solaris or whatever.
You can't make any money, directly, off of producing GPL'd code, but you can at least prevent your competitors from benefiting from your work.
I just cannot see Sun replacing Solaris on their high-end multi-processor machines... Or at least not until Linux scales equally well :-)
It's not just scalability. There are features in Solaris-- although I'm not familiar enough with them to get specific-- that allow the customer to partition out failed hardware components for repair or replacement without taking the system down. Linux has no features even remotely like that.
Of course, the big problem was always getting the system to recognize the new hardware component again. If I remember right, that still requires a reboot....
You're being too simplistic, probably because of my hurried and incomplete example. Semantic searching is most useful in the "general to more specific" instances. While searching for "orange" shouldn't necessarily search for "fruit," as you point out, it's very likely that a search for "fruit" should key off of "orange."
;-)
This is really much more applicable to concept searching than it is to simple text indexing. (Mmm... SimpleText...)
For example, if a catalogger is describing a painting, she might use the word "orange" to describe the subject of a still life. In that instance, the painting's metadata structures would reflect the fact that "orange" in the subject field must be the concrete noun "orange," which is a specific instance of the concrete noun "fruit." The idea being that a user could search for "fruit" and get a hit on a painting that has been described as "still life with oranges."
So maybe my idea isn't all that applicable to Google after all. Hell, I'm not even sure I'm on topic any more.
So now when I search for "Ornage" it asks me "Did you mean orange." I guess Google could extend this if it was hooked up to WordNet - "Did you mean orange the fruit, orange the color, orange the tree, or orange the river"
EXACTLY! Like that scene in 2010:
CHANDRA
I would like to open a new file. Here is the name for it. [types "phoenix"] Do you know what that means?
SAL
There are twenty five references in the current encyclopedia.
CHANDRA
Which one do you think is relevant?
SAL
The tutor of Achilles?
CHANDRA
That's very interesting, I didn't know that one. Try again.
SAL
A fabulous bird, re-born from the ashes of its earlier life.
CHANDRA
That is correct.
How about adding the option to have google understand what I *mean* to search for, not what I tell it to search for.
You might have been kidding, but you've got a really good idea there.
How about semantic searching: equip Google with a database that organizes words in a relational hierarchy from the general to the specific. For example, "orange" is a more specific form of "fruit," and also a more specific form of "color."
When you search for "orange," Google might also have the ability to search for "fruit" and "color," depending on how broad you want your search to be.
Just a thought.