Intel might claim that they are going for the throat with the PIII Xeon, but they have a huge gap to close before marketing catches up with reality.
As far as I can tell, the most the Xeon can scale to is 8 processors. At least that is the largest machine I can find for Xeon. Sun's midrange machines _start_ at 4 processors and, for now, go up to 64. The planned maximum for UltraSparc III is 1024. Sun 1, Intel 0.
UltraSparc is 64 bit, Xeon is 32. Sun 2, Intel 0
UltraSparcs have had integrated caches for quite a while, as far back as SuperSparc I. On the current cpus, the desktop boxes can have up 2MB cache, the midrange servers 4MB, and the large Enterprise servers get 8MB cache. The PIII Xeon's claim to fame is the extra electronics they add to the processor card which allows some hardware admin of the processor. The cache size is 256KB...same as a standard PIII.
Sun 3, Intel 0.
Until Intel produces a cpu which can scale well and has an OS which supports that scaling, I don't think Sun has much to worry about.
And for those of you who think MHz is a consideration, bzzzt! In the world of 24x7 server farms, torque counts more than horsepower. The UltraSparc cpu can handle a much greater load without sweat than most other cpus. The UltraSparc was designed to handle massive loads. So, while the UltraSparc will likely lag behind an equivalently rated Intel cpu, which would you rather use to make a cross country houshold move: A Kenworth or a few dozen pick-up trucks?
I may choose to read or ignore a book about murder. I happen to like that genre actually, nearly as much as I like film noir. But still, the difference is that I *choose* to read that subject.
On the other hand a virus is basically a hit and run _crime_. As one of the other respondents above remarked, modern art, by my standards would not be considered art. I disagree because I can ignore it. I may not call it art, but someone else may. I cannot ignore a virus, even if I am running a nearly immune system.
To compound the issue, not only do viruses steal your ability ignore them, the nastier ones tend to cost money. Either the viruses destroy work product or they create work for the admin who then has to fix his network.
If you want to call it art, fine, as long as the creator makes it performance art. Like boxing...so I can add to the performance.
You can't possibly consider a virus writer to be an artist? I'm sure that some of code they produce is elegant, or at least quite advanced and technical. But to call the result of that work 'art' is just fallacy.
Next thing: Murder as art, The art of the Heist, and the all time favorite, Keying Cars as an Expression of Angst - the Artists Perspective.
A virus is just destructive code. To me that means it is no different than a molotov cocktail. Yes there are differing degrees of harm created, but whether that harm is physical or results in the loss of work product, there is harm involved.
Having read a few of the responses so far, I am in agreement that these are particularly valid questions to pose. What I don't see in this letter is a defense for/. action (or inaction, really)
Apart from stating that Andover.net is loath to censor user comments, nothing in this letter backs up any claim that these are protected speech posts. (As opposed to merely revealing 'proprietary' information.)
Which is clearly missing the point. Of course Java was never designed to replace an OS, be it Windows, OS/2, Linux...whatever.
So where is the threat to Windows if Java were to become successful? Look at it this way; If $OS of choice has a fast, stable and compliant JVM, and the application you want to run is fully Java, then why buy Windows?
And if one company can be successful with a major Java app, then others might become successful. After a while, the OS becomes irrelevant...or at least not a cash cow for $COMPANY.
It is for this reason that Windows embracing and extending MS's JVM was a bad thing. Remove the interoperability and you remove one of the primary benefits to the technology.
I'm curious about the whole slot thing if this is true. I recall that the reason the slot was introduced was because Sockets were considered incapable of running >400MHz. Heck, I even remember reading statements such as that as late as the introduction of the AMD K6-300.
Now we're going back to the good ole socket because the slot can't go above 1GHz? The word 'boondoggle' comes to mind.
Not to speak badly of the police, but the police are not law experts really. Sure they have a fantastic working knowledge of a fairly wide array of laws, when it gets down to specifics the police that I know personally err on the side of caution.
In other words; If an officer sees you doing something or possessing something he/she thinks might be illegal, you will likely be arrested/ticketed and the courts will figure it out.
I tend to discount advice saying 'a cop friend of mine says that that is illegal'. Better to look at the local ordinances yourself to determine the status of $SOMETHING.
Sure thing. Makes the job of removing deeply ingrained (socialized and inborn) morals that much easier, since you have to remove less in order to create an efficient killer.
Now this is just crap. Do you honestly believe that the military is staffed with golems which only strive to perform whatever task is set forth by commanders?
Think about that for a second. What you are basically positing here is that whenever you join the military you give up whatever identity you have.
What it all boils down to is that the military is a job. Nothing more, nothing less. Having been in the Marines for 11 years I can honestly say that I never met any Marine who's sole goal was to kill defenseless women and children. I can also say that nearly all of the Marines I have ever known were moral and honorable people.
I suppose that if you have no faith in Humanity it would be easy to think that a video game can make anyone into an amoral killing machine.
Fortunately all of the people I know far exceed your expectations.
Solaris up through 8 does not come with any type of real journaled fs. I leave out 8 because I have yet to see it.
What it does come with is Disksuite. Capable in and of itself, but not journaled.
The first thing any sysadmin does after installing a new Sun box? Install Veritas. Why on Earth Sun is still in the dark ages in that regard, I have no clue.
Agreed. Want to run a high speed image manipulation cluster, use xfs. Need absolutely secure data warehousing in the data center, use jfs. (secure as in 'no loss' not encryption) Having used HP-UX with its LVM and Solaris with Veritas VxFS (which I understand to be the same thing actually) I really miss the ability to change filesystems on the fly. Now if we could just get rid of '/dev/sda1' I'd be even happier. It looks clunky, but/dev/c0t0d0s1 makes more sense in the long run. Just MHO
I recall watching Cringeley's first special on computers a few years ago. In order to claim that the reverse engineering was clean, there were two sets of engineers. The first set of engineers had access to the device being reversed (IBMs PC BIOS) and compiled a set of rules. The second set of engineers worked _only_ from the rules obtained by the first set and supposedly never came into contact with the device being reversed. I get the impression from your posting that there really isn't even a product available yet to be reversed, but that there is enough information to design a parallel process. Whether or not this can stand as 'clean room' reversing would depend on how the data you obtained was presented. I agree with one of the prior posters; If it was obtained from patent documents, you might have some problems. I am curious whether or not, if the data was obtained from published commercial/scientific documents, copyrights might also be claimed?
Not being a programmer, I am curious how long it would really take to implement a d.net client. I am not sure how the four different d.net clients now operate, but I assume that some portion of the code is reuseable and another portion of the code would be specific to the client. Assuming that NASA would be willing to open source the signal processing code, is it possible that the d.net folks could 'plug' it into the clients? Since the d.net folks have already coded client switching into the current clients, they could conceivably createa fifth client...say MLDR. Then whenever the clients connect, and there are new data to crunch, those people who have setup their clients properly would automagically work on a few lander blocks.
Not being a programmer, I am curious how long it would really take to implement a d.net client. I am not sure how the four different d.net clients now operate, but I assume that some portion of the code was reuseable and another portion of the code would be specific to the client.
Assuming that NASA would be willing to open source the signalprocessing code, is it possible that the d.net folks could 'plug' it into the clients? Since the d.net folks have already coded client switching into the current clients, they could conceivably createa fifth client...say MLDR. Then whenever the clients connect, and there are new data to crunch, those people who have setup their clients properly would automagically work on a few lander blocks.
Enlight makes at least one rackmount 'file server' case. Model #EN-8950. It is likely the -8960 will have a rackmount kit also, but the 8960 is $750 whereas the 8950 can be had for $160.
I haven't checked what the prices are for the rackmount kits for these cases though.
Funny you should mention that. In fact, my friend, that is exactly what the Linux kernel is. Yes, the Linux kernel is a huge, monolithic program (thus the term "monolithic kernel") that contains a good deal of Linux's device support as well as a zillion other things.
Huge? To use paraphrase another post: 'You haven't used linux before, have you?' Unless, of course, you consider ~700k huge. Nor are there 'zillions' of other things in the kernel.
Slashdot Mirror
Intel might claim that they are going for the throat with the PIII Xeon, but they have a huge gap to close before marketing catches up with reality.
...same as a standard PIII.
As far as I can tell, the most the Xeon can scale to is 8 processors. At least that is the largest machine I can find for Xeon. Sun's midrange machines _start_ at 4 processors and, for now, go up to 64. The planned maximum for UltraSparc III is 1024. Sun 1, Intel 0.
UltraSparc is 64 bit, Xeon is 32.
Sun 2, Intel 0
UltraSparcs have had integrated caches for quite a while, as far back as SuperSparc I. On the current cpus, the desktop boxes can have up 2MB cache, the midrange servers 4MB, and the large Enterprise servers get 8MB cache. The PIII Xeon's claim to fame is the extra electronics they add to the processor card which allows some hardware admin of the processor. The cache size is 256KB
Sun 3, Intel 0.
Until Intel produces a cpu which can scale well and has an OS which supports that scaling, I don't think Sun has much to worry about.
And for those of you who think MHz is a consideration, bzzzt! In the world of 24x7 server farms, torque counts more than horsepower. The UltraSparc cpu can handle a much greater load without sweat than most other cpus. The UltraSparc was designed to handle massive loads. So, while the UltraSparc will likely lag behind an equivalently rated Intel cpu, which would you rather use to make a cross country houshold move: A Kenworth or a few dozen pick-up trucks?
You are forgetting one major factor: Choice.
...so I can add to the performance.
I may choose to read or ignore a book about murder. I happen to like that genre actually, nearly as much as I like film noir. But still, the difference is that I *choose* to read that subject.
On the other hand a virus is basically a hit and run _crime_. As one of the other respondents above remarked, modern art, by my standards would not be considered art. I disagree because I can ignore it. I may not call it art, but someone else may. I cannot ignore a virus, even if I am running a nearly immune system.
To compound the issue, not only do viruses steal your ability ignore them, the nastier ones tend to cost money. Either the viruses destroy work product or they create work for the admin who then has to fix his network.
If you want to call it art, fine, as long as the creator makes it performance art. Like boxing
You can't possibly consider a virus writer to be an artist? I'm sure that some of code they produce is elegant, or at least quite advanced and technical. But to call the result of that work 'art' is just fallacy.
Next thing: Murder as art, The art of the Heist, and the all time favorite, Keying Cars as an Expression of Angst - the Artists Perspective.
A virus is just destructive code. To me that means it is no different than a molotov cocktail. Yes there are differing degrees of harm created, but whether that harm is physical or results in the loss of work product, there is harm involved.
Having read a few of the responses so far, I am in agreement that these are particularly valid questions to pose. What I don't see in this letter is a defense for /. action (or inaction, really)
Apart from stating that Andover.net is loath to censor user comments, nothing in this letter backs up any claim that these are protected speech posts. (As opposed to merely revealing 'proprietary' information.)
Punt?
Which is clearly missing the point. Of course Java was never designed to replace an OS, be it Windows, OS/2, Linux...whatever.
...or at least not a cash cow for $COMPANY.
So where is the threat to Windows if Java were to become successful? Look at it this way; If $OS of choice has a fast, stable and compliant JVM, and the application you want to run is fully Java, then why buy Windows?
And if one company can be successful with a major Java app, then others might become successful. After a while, the OS becomes irrelevant
It is for this reason that Windows embracing and extending MS's JVM was a bad thing. Remove the interoperability and you remove one of the primary benefits to the technology.
Too many watersports Blue....
I'm curious about the whole slot thing if this is true. I recall that the reason the slot was introduced was because Sockets were considered incapable of running >400MHz. Heck, I even remember reading statements such as that as late as the introduction of the AMD K6-300.
Now we're going back to the good ole socket because the slot can't go above 1GHz? The word 'boondoggle' comes to mind.
Seems like this might be one time that you could
have a slashpoll.
I saw devices that would print from videotape and laserdisc when I was in Japan 10 years ago. IIRC, Matsushita was making them.
So what makes this news?
Actually, Stewart played the Master at Arms, or somesuch.
Not to speak badly of the police, but the police are not law experts really. Sure they have a fantastic working knowledge of a fairly wide array of laws, when it gets down to specifics the police that I know personally err on the side of caution.
In other words; If an officer sees you doing something or possessing something he/she thinks might be illegal, you will likely be arrested/ticketed and the courts will figure it out.
I tend to discount advice saying 'a cop friend of mine says that that is illegal'. Better to look at the local ordinances yourself to determine the status of $SOMETHING.
Sure thing. Makes the job of removing deeply ingrained (socialized and inborn) morals that much easier, since you have to remove less in order to create an efficient killer.
Now this is just crap. Do you honestly believe that the military is staffed with golems which only strive to perform whatever task is set forth by commanders?
Think about that for a second. What you are basically positing here is that whenever you join the military you give up whatever identity you have.
What it all boils down to is that the military is a job. Nothing more, nothing less. Having been in the Marines for 11 years I can honestly say that I never met any Marine who's sole goal was to kill defenseless women and children. I can also say that nearly all of the Marines I have ever known were moral and honorable people.
I suppose that if you have no faith in Humanity it would be easy to think that a video game can make anyone into an amoral killing machine.
Fortunately all of the people I know far exceed your expectations.
Of course no-one would attack Canada. That would piss off the Mounties. Then all heck would break loose.
Solaris up through 8 does not come with any type of real journaled fs. I leave out 8 because I have yet to see it.
What it does come with is Disksuite. Capable in and of itself, but not journaled.
The first thing any sysadmin does after installing a new Sun box? Install Veritas. Why on Earth Sun is still in the dark ages in that regard, I have no clue.
Make that: /dev/[r]dsk/c0t0d0s1
Agreed. Want to run a high speed image manipulation cluster, use xfs. Need absolutely secure data warehousing in the data center, use jfs. (secure as in 'no loss' not encryption) /dev/c0t0d0s1 makes more sense in the long run. Just MHO
Having used HP-UX with its LVM and Solaris with Veritas VxFS (which I understand to be the same thing actually) I really miss the ability to change filesystems on the fly.
Now if we could just get rid of '/dev/sda1' I'd be even happier. It looks clunky, but
I recall watching Cringeley's first special on computers a few years ago. In order to claim that the reverse engineering was clean, there were two sets of engineers. The first set of engineers had access to the device being reversed (IBMs PC BIOS) and compiled a set of rules. The second set of engineers worked _only_ from the rules obtained by the first set and supposedly never came into contact with the device being reversed.
I get the impression from your posting that there really isn't even a product available yet to be reversed, but that there is enough information to design a parallel process. Whether or not this can stand as 'clean room' reversing would depend on how the data you obtained was presented.
I agree with one of the prior posters; If it was obtained from patent documents, you might have some problems.
I am curious whether or not, if the data was obtained from published commercial/scientific documents, copyrights might also be claimed?
Anyone got a link? The surf nazis at work here have playboy blocked.
Gee, and I was really going to read the article...honest.
Dammit...
Sorry, I thought I hit preview....
Not being a programmer, I am curious how long it would really take to implement a d.net client. I am not sure how the four different d.net clients now operate, but I assume that some portion of the code is reuseable and another portion of the code would be specific to the client. ...say MLDR. Then whenever the clients connect, and there are new data to crunch, those people who have setup their clients properly would automagically work on a few lander blocks.
Assuming that NASA would be willing to open source the signal processing code, is it possible that the d.net folks could 'plug' it into the clients? Since the d.net folks have already coded client switching into the current clients, they could conceivably createa fifth client
Just an idea....
Not being a programmer, I am curious how long it would really take to implement a d.net client. I am not sure how the four different d.net clients now operate, but I assume that some portion of the code was reuseable and another portion of the code would be specific to the client.
...say MLDR. Then whenever the clients connect, and there are new data to crunch, those people who have setup their clients properly would automagically work on a few lander blocks.
Assuming that NASA would be willing to open source the signalprocessing code, is it possible that the d.net folks could 'plug' it into the clients? Since the d.net folks have already coded client switching into the current clients, they could conceivably createa fifth client
Just an idea....
Another option:
OK>set defualts
Enlight makes at least one rackmount 'file server' case. Model #EN-8950. It is likely the -8960 will have a rackmount kit also, but the 8960 is $750 whereas the 8950 can be had for $160.
I haven't checked what the prices are for the rackmount kits for these cases though.
Funny you should mention that. In fact, my friend, that is exactly what the Linux kernel is. Yes, the Linux kernel is a huge, monolithic program (thus
the term "monolithic kernel") that contains a good deal of Linux's device support as well as a zillion other things.
Huge? To use paraphrase another post: 'You haven't used linux before, have you?' Unless, of course, you consider ~700k huge. Nor are there 'zillions' of other things in the kernel.
Then again, as a sysadmin my I'm likely biased.