Proof once again that Mr. Bush is making America safer! Sure, maybe we can't find Osama (guess that might take 12 years?) but we've finally located Bobby Fischer (thanks, Japan!) and all Americans can sleep better, no longer having to fear a mentally ill 61 year-old recluse. I know I feel better.
I don't really disagree with much of what you've said . . . I just don't follow how you choose to tack this on to my question about whether or not some of the terrorists in question were on expired visas - and not even answer the damned question. Do you not understand how message threading works?
Well then why make a comment like "just one more reason to use open source" in a thread about the use of proprietary code in tape libraries at all?
You do know that ES/9000's are like 7 or 8 generations removed from the current line, right?
But weren't some of them on expired visas? You know, those things we use to control the access of non-citizens into America? Who was it that said laws that aren't enforced are just advice? Truman?
Keep in mind, by taking away all those nail clippers the TSA in inadvertently causing Americans to have longer nails - and thus arming the very people they want to disarm! Let's face it, it's only a matter of time before another plane is hijacked by a group with overly long fingernails . . . all for want of a clipper.
When nail clippers are outlawed, only outlaws will have short nails!
Yes it's inconvenient, but don't you feel better knowing that after those searches there's a much lower likelihood that YOU have carried a bomb onto the plane?:^)
Please re-read the article. The judge specifically addresses that section and specifically indicates why, in his judgement, the defendants are in the wrong.
Yay! That's now, like what, 3 of us that actually READ the article instead of just seeing "DMCA" and turning into the Hulk?
It's still an issue to be addressed though, as competition in the maintenance market for enterprise hardware is a good thing(tm). I'd like to know a little more about the background of this case. I'm going to ask my local STK rep to see what he can tell me. All these firms (at least any you'd want to hire) are already paying the OEM's some money - they have to send their techs to training, purchase documentation, specific tools and diagnostics, etc. I'm wondering if it's also common to enter into a license agreement granting access to code like this and this firm had chosen to go a different way. There's way too many 3rd party firms out there maintaining STK/IBM/etc. equipment that I would have expected to hear a louder tree falling if this was a huge deal.
With logic like that one could then extend it further to say I have the right to run anything on my property - whether or not I have a license to do so. Unlimited copies of Windows, z/OS, Photoshop - you name it. Hey, it's on my property, no?
You own/lease whatever it says on the contract. In this case, STK owns the maintenance code and uses it (as directed by the customer) to maintain the equipment.
And, pray tell, where does one go obtain a large scale robotic tape library using open source? There's really only 3 players in the enterprise tape library market - STK, IBM, and Fujitsu. I haven't seen anyone say that the other 2 do NOT have similar license controls over their software, either (and I'm wondering if they are contemplating their options to similarly enforce their control, now). You think STK is playing dirty? You've probably never been in an enterprise data center that chooses a 3rd party provider to maintain their IBM equipment; it can get real ugly.
How many people do you think would have voted for a "Furthering Actions to Subvert, Compromise, and Impede the Schemes of Terrorists" Act?
Pretty much the entire Republican Party, I'd guess. Without somebody to sound out this stuff (think Hooked on Phonics) to King George and his royal court I really don't think they can figure out such sophisticated things as acronyms by themselves. I mean, FFS, our King can't even pronounce the name of the prison where his administration directed people to abuse prisoners for information. But, on a positive note, he WAS able to meet with the PM of Japan and not ralph on his lap . . . gotta look for those silver linings wherever you can.
Geez, you've already traded your fingerprint for a driver's license, you're not willing to make your reading list available to know that your neighbors aren't learning how to build bombs in their basement? Huh? Maybe YOU have traded your fingerprint for a driver's license, Mr. Anonymous Coward (gee, anonymity for library users is bad, anonymity for you is good?), but I sure haven't.
Now come on. You know very well that there's a huge difference between what is happening in America today and what the Soviets did. I don't know about you, but I do not have any fear of being woken up in the middle of the night, thrown into a van, and being shipped off to some Siberian gulag just because I surfed the wrong website last night.
No of course in the US you wouldn't be shipped off to a Siberian gulag. You'd be shipped off to Gitmo. Now go put your head back into the sand and keep repeating "It could never happen to me, It could never happen to me, . . . " Oh yeah, I do vote, btw - but I can't outrule a Supreme Court that decides to throw an election. Everybody hold up their hands that's wondering how Bush/Ridge parlay the "threat" to the election into a way to rig this one, too.
..but what about recovery plans for catastrophic events? Those backup tapes sitting in a filing cabinet next to the server are useless when the building burns down or is flooded. I suppose you could just ship the tapes to another location, but then restoration becomes and even longer ordeal.
That has nothing to do with tape, that has to do with idiots being allowed to manage/design the backup infrastructure. First - you contradict yourself in some ways when you mention that if the building burns down but the tapes are offsite restoration is impaired. If the building burned down there ain't no restoration!
Any sane company using tape as their primary backup strategy (or even secondary strategy) is already either writing those tapes directly to a second site (that is also their recovery site or is connected to a recovery site) or, with higher risk and lower cost, cycling those tapes offsite via CTAM (Chevy Truck Access Method) or similar. The latter strategy isn't very prevalent in enterprise data centers anymore but still may have value to SMB's.
Frankly with the volume of data that any enterprise data center has these days wholesale restoration to disk from tape is no longer a viable strategy. It's a basic speeds and feeds numbers game that has anybody in the 100+ TB range being down for far too long while the restores churn along. Most folks of this size employ a combined approach of mirroring the disk data to a second site AND writing tape backups to a second site. This provides for rapid recovery of the majority of the Tier 1 servers (whatever the platform - mainframe, Unix, Windows, AS/400, etc.) and still provides the longer term retention of tape backups for the one-off restores that are still needed.
Please actually read what I wrote. I did not say that Ferrari's RACING budget wasn't larger - I said they didn't have the most money, period. I don't understand how other people seemed to have totally misinterpreted that really simple statement.
Stack Ferrari's annual revenue stream against that of Ford or BMW and tell me Ferrari has more money - I laugh. Ford has WAY more money than Ferrari and chooses not to increase their F1 racing budget. Their call. But they can hardly cry "poor" and that they are losing (badly)because they can't compete financially with Ferrari.
Ferrari isn't winning because they have the most money. BMW-Williams, Toyota, Ford (Jaguar-Cosworth), etc. aren't exactly collecting foodstamps you know.
Ferrari is winning because they have assembled a fantastic team. From the engineering talent that designed the car to the logistic support to the pit crew and, of course, one of the greatest drivers in the history of the sport - Michael Schumaker.
I'm also trying to remember who won the World Series last year. The Yankees, right, cause they have the biggest payroll? Ooops, no, it was the Florida Marlins with a payroll at the other end of the spectrum.
Analogies are great except when they don't hold up.
If you have a problem, you remove the bad drive, replace it and reinitialize the RAID arrive.
And in so doing wipe out all the data? If you are going to bitch at the guy for not googling at least give him good info. Initialize/reinitialize = start from ground zero. Kind of de-feets the porpoise, no? After replacing a drive in a RAID-5 array one typically rebuilds the array.
This year: Make a robot that will successfully find a parking spot at the mall!
These already exist, and are especially prevalent in Southern California. They are called "Bimbots". They aren't much on high-level conversation but have been designed to be pleasing in appearance and have the ability to pilot vehicles to and from shopping establishments.
Managing a 300+ TB DASD farm I know for a fact that IBM does NOT ship 4 TB systems where a customer has ordered 1 TB. Especially now that they don't even manufacture HDA's anymore and are either buying them (typically from HDS) for their Shark or reselling LSI (FastT); in either case there's an actual cash payment IBM must make for the manufactured goods back to the HDA suppliers. There may be situations where a specific, ordered, amount doesn't jibe well with how the HDA & RAID config will map it out, in which case a rounding-up will be done, but they certainly are NOT deliberately shipping 4:1 ratios of actual:paid capacity. It also would be simply bad business to overbuy now for a potential sale down the road, as the cost of goods continues to decline. The sellers know it and the buyers know it. Why but 3 cent per MB storage today in the hopes of selling it 1 or 2 fiscal quarters later at 2 cents per MB? Crazy.
The true cost of the initial storage investment is in the controller itself - where the logic, cache, etc. is housed and - increasingly - s/w licenses for storage features. After that the subsequent sales of additional capacity is a very simple, non-disruptive, process and there's simply no advantage for either seller or buyer to oversupply and undersell the initial configuration. Additional capacity is very much a just-in-time commodity. I can decide on a Monday I need more capacity and have it online before the end of the week.
That's not to say there aren't dumb sellers or buyers out there that would fall for this ("my price is going to be a tad high initially to help cover the margin on increased cost of goods, but later on when you need more capacity it's already there for easy upgrading") but no serious enterprise storage manager would fall for it, and no serious enterprise storage seller would try it unless they didn't care about the account.
I have to respond to this one if you mean, by "big iron", mainframe.
1) Any mainframe shipped within the last few years has either every possible processor it can handle included or half of that. It's a manufacturing issue. The method that IBM uses to manufacture and package the processors makes it ultimately cheaper to do it this way, not to mention that recent advances in processor performance require this type of on-board/on-chip arrangement. The days of individual "engines" being installed are long-gone.
In this situation, the customer buys - say - an 8-way. At some point in the future when demand has risen to the point where capacity needs to be increased the customer can purchase an "upgrade" (let's say to a 10-way). Rather than some encrypted number on a disk, the CE enables a microcode feature that then enables 2 more engines.
There is a different type of "governor" that can also be applied. Any given engine runs at a specific speed resulting in a certain amount of processor capacity (MIPS is still a popular unit for mainframes though true perf. and capacity types hate it). Standard capacity for a z990 mainframe is 450 MIPS/engine. The recently introduced z890 mainframe, though, has a "variable" engine speed - anywhere from 26 MIPS/engine to 366 MIPS/engine.
Why would somebody want a "lower speed" engine? Because mainframe software vendors typically index their licensing costs to installed processor capacity. By being able to better match their capacity to their requirements (instead of having to upgrade in 450 MIP chunks) an installation get better control their software costs - which on the mainframe far outweigh the hardware costs.
So here we have an example where "underclocking" is actually welcome.
2) As to shippping multiple terabytes of disk storage even though one 1 TB was ordered, never seen that. HDA's are a manufactured good, and shipping more HDA's then required eats into what is already a business run on razor-thin margins. I once heard an exec of a storage manufacturer joke "What's the difference between pizza and disk drives? Well, they are both round but you can make a profit selling pizza!"
Slashdot Mirror
Proof once again that Mr. Bush is making America safer! Sure, maybe we can't find Osama (guess that might take 12 years?) but we've finally located Bobby Fischer (thanks, Japan!) and all Americans can sleep better, no longer having to fear a mentally ill 61 year-old recluse. I know I feel better.
Why not outsource to me in West Virginia...? Probably afraid of the language barrier. :)
I don't really disagree with much of what you've said . . . I just don't follow how you choose to tack this on to my question about whether or not some of the terrorists in question were on expired visas - and not even answer the damned question. Do you not understand how message threading works?
Well then why make a comment like "just one more reason to use open source" in a thread about the use of proprietary code in tape libraries at all? You do know that ES/9000's are like 7 or 8 generations removed from the current line, right?
But weren't some of them on expired visas? You know, those things we use to control the access of non-citizens into America? Who was it that said laws that aren't enforced are just advice? Truman?
Keep in mind, by taking away all those nail clippers the TSA in inadvertently causing Americans to have longer nails - and thus arming the very people they want to disarm! Let's face it, it's only a matter of time before another plane is hijacked by a group with overly long fingernails . . . all for want of a clipper. When nail clippers are outlawed, only outlaws will have short nails!
Yes it's inconvenient, but don't you feel better knowing that after those searches there's a much lower likelihood that YOU have carried a bomb onto the plane? :^)
"Billions Served . . . countless more fucked over!"
Please re-read the article. The judge specifically addresses that section and specifically indicates why, in his judgement, the defendants are in the wrong.
Yay! That's now, like what, 3 of us that actually READ the article instead of just seeing "DMCA" and turning into the Hulk?
It's still an issue to be addressed though, as competition in the maintenance market for enterprise hardware is a good thing(tm). I'd like to know a little more about the background of this case. I'm going to ask my local STK rep to see what he can tell me. All these firms (at least any you'd want to hire) are already paying the OEM's some money - they have to send their techs to training, purchase documentation, specific tools and diagnostics, etc. I'm wondering if it's also common to enter into a license agreement granting access to code like this and this firm had chosen to go a different way. There's way too many 3rd party firms out there maintaining STK/IBM/etc. equipment that I would have expected to hear a louder tree falling if this was a huge deal.
With logic like that one could then extend it further to say I have the right to run anything on my property - whether or not I have a license to do so. Unlimited copies of Windows, z/OS, Photoshop - you name it. Hey, it's on my property, no?
I don't think so.
You own/lease whatever it says on the contract. In this case, STK owns the maintenance code and uses it (as directed by the customer) to maintain the equipment.
And, pray tell, where does one go obtain a large scale robotic tape library using open source? There's really only 3 players in the enterprise tape library market - STK, IBM, and Fujitsu. I haven't seen anyone say that the other 2 do NOT have similar license controls over their software, either (and I'm wondering if they are contemplating their options to similarly enforce their control, now). You think STK is playing dirty? You've probably never been in an enterprise data center that chooses a 3rd party provider to maintain their IBM equipment; it can get real ugly.
How many people do you think would have voted for a "Furthering Actions to Subvert, Compromise, and Impede the Schemes of Terrorists" Act?
Pretty much the entire Republican Party, I'd guess. Without somebody to sound out this stuff (think Hooked on Phonics) to King George and his royal court I really don't think they can figure out such sophisticated things as acronyms by themselves. I mean, FFS, our King can't even pronounce the name of the prison where his administration directed people to abuse prisoners for information. But, on a positive note, he WAS able to meet with the PM of Japan and not ralph on his lap . . . gotta look for those silver linings wherever you can.
Geez, you've already traded your fingerprint for a driver's license, you're not willing to make your reading list available to know that your neighbors aren't learning how to build bombs in their basement? Huh? Maybe YOU have traded your fingerprint for a driver's license, Mr. Anonymous Coward (gee, anonymity for library users is bad, anonymity for you is good?), but I sure haven't.
Now come on. You know very well that there's a huge difference between what is happening in America today and what the Soviets did. I don't know about you, but I do not have any fear of being woken up in the middle of the night, thrown into a van, and being shipped off to some Siberian gulag just because I surfed the wrong website last night. No of course in the US you wouldn't be shipped off to a Siberian gulag. You'd be shipped off to Gitmo. Now go put your head back into the sand and keep repeating "It could never happen to me, It could never happen to me, . . . " Oh yeah, I do vote, btw - but I can't outrule a Supreme Court that decides to throw an election. Everybody hold up their hands that's wondering how Bush/Ridge parlay the "threat" to the election into a way to rig this one, too.
..but what about recovery plans for catastrophic events? Those backup tapes sitting in a filing cabinet next to the server are useless when the building burns down or is flooded. I suppose you could just ship the tapes to another location, but then restoration becomes and even longer ordeal. That has nothing to do with tape, that has to do with idiots being allowed to manage/design the backup infrastructure. First - you contradict yourself in some ways when you mention that if the building burns down but the tapes are offsite restoration is impaired. If the building burned down there ain't no restoration! Any sane company using tape as their primary backup strategy (or even secondary strategy) is already either writing those tapes directly to a second site (that is also their recovery site or is connected to a recovery site) or, with higher risk and lower cost, cycling those tapes offsite via CTAM (Chevy Truck Access Method) or similar. The latter strategy isn't very prevalent in enterprise data centers anymore but still may have value to SMB's. Frankly with the volume of data that any enterprise data center has these days wholesale restoration to disk from tape is no longer a viable strategy. It's a basic speeds and feeds numbers game that has anybody in the 100+ TB range being down for far too long while the restores churn along. Most folks of this size employ a combined approach of mirroring the disk data to a second site AND writing tape backups to a second site. This provides for rapid recovery of the majority of the Tier 1 servers (whatever the platform - mainframe, Unix, Windows, AS/400, etc.) and still provides the longer term retention of tape backups for the one-off restores that are still needed.
PL/1 is in there. However PL/S isn't.
Please actually read what I wrote. I did not say that Ferrari's RACING budget wasn't larger - I said they didn't have the most money, period. I don't understand how other people seemed to have totally misinterpreted that really simple statement.
Stack Ferrari's annual revenue stream against that of Ford or BMW and tell me Ferrari has more money - I laugh. Ford has WAY more money than Ferrari and chooses not to increase their F1 racing budget. Their call. But they can hardly cry "poor" and that they are losing (badly)because they can't compete financially with Ferrari.
Ferrari isn't winning because they have the most money. BMW-Williams, Toyota, Ford (Jaguar-Cosworth), etc. aren't exactly collecting foodstamps you know.
Ferrari is winning because they have assembled a fantastic team. From the engineering talent that designed the car to the logistic support to the pit crew and, of course, one of the greatest drivers in the history of the sport - Michael Schumaker.
I'm also trying to remember who won the World Series last year. The Yankees, right, cause they have the biggest payroll? Ooops, no, it was the Florida Marlins with a payroll at the other end of the spectrum.
Analogies are great except when they don't hold up.
FC.
If you have a problem, you remove the bad drive, replace it and reinitialize the RAID arrive. And in so doing wipe out all the data? If you are going to bitch at the guy for not googling at least give him good info. Initialize/reinitialize = start from ground zero. Kind of de-feets the porpoise, no? After replacing a drive in a RAID-5 array one typically rebuilds the array.
This year: Make a robot that will successfully find a parking spot at the mall! These already exist, and are especially prevalent in Southern California. They are called "Bimbots". They aren't much on high-level conversation but have been designed to be pleasing in appearance and have the ability to pilot vehicles to and from shopping establishments.
Managing a 300+ TB DASD farm I know for a fact that IBM does NOT ship 4 TB systems where a customer has ordered 1 TB. Especially now that they don't even manufacture HDA's anymore and are either buying them (typically from HDS) for their Shark or reselling LSI (FastT); in either case there's an actual cash payment IBM must make for the manufactured goods back to the HDA suppliers. There may be situations where a specific, ordered, amount doesn't jibe well with how the HDA & RAID config will map it out, in which case a rounding-up will be done, but they certainly are NOT deliberately shipping 4:1 ratios of actual:paid capacity. It also would be simply bad business to overbuy now for a potential sale down the road, as the cost of goods continues to decline. The sellers know it and the buyers know it. Why but 3 cent per MB storage today in the hopes of selling it 1 or 2 fiscal quarters later at 2 cents per MB? Crazy.
The true cost of the initial storage investment is in the controller itself - where the logic, cache, etc. is housed and - increasingly - s/w licenses for storage features. After that the subsequent sales of additional capacity is a very simple, non-disruptive, process and there's simply no advantage for either seller or buyer to oversupply and undersell the initial configuration. Additional capacity is very much a just-in-time commodity. I can decide on a Monday I need more capacity and have it online before the end of the week.
That's not to say there aren't dumb sellers or buyers out there that would fall for this ("my price is going to be a tad high initially to help cover the margin on increased cost of goods, but later on when you need more capacity it's already there for easy upgrading") but no serious enterprise storage manager would fall for it, and no serious enterprise storage seller would try it unless they didn't care about the account.
I have to respond to this one if you mean, by "big iron", mainframe.
1) Any mainframe shipped within the last few years has either every possible processor it can handle included or half of that. It's a manufacturing issue. The method that IBM uses to manufacture and package the processors makes it ultimately cheaper to do it this way, not to mention that recent advances in processor performance require this type of on-board/on-chip arrangement. The days of individual "engines" being installed are long-gone.
In this situation, the customer buys - say - an 8-way. At some point in the future when demand has risen to the point where capacity needs to be increased the customer can purchase an "upgrade" (let's say to a 10-way). Rather than some encrypted number on a disk, the CE enables a microcode feature that then enables 2 more engines.
There is a different type of "governor" that can also be applied. Any given engine runs at a specific speed resulting in a certain amount of processor capacity (MIPS is still a popular unit for mainframes though true perf. and capacity types hate it). Standard capacity for a z990 mainframe is 450 MIPS/engine. The recently introduced z890 mainframe, though, has a "variable" engine speed - anywhere from 26 MIPS/engine to 366 MIPS/engine.
Why would somebody want a "lower speed" engine? Because mainframe software vendors typically index their licensing costs to installed processor capacity. By being able to better match their capacity to their requirements (instead of having to upgrade in 450 MIP chunks) an installation get better control their software costs - which on the mainframe far outweigh the hardware costs.
So here we have an example where "underclocking" is actually welcome.
2) As to shippping multiple terabytes of disk storage even though one 1 TB was ordered, never seen that. HDA's are a manufactured good, and shipping more HDA's then required eats into what is already a business run on razor-thin margins. I once heard an exec of a storage manufacturer joke "What's the difference between pizza and disk drives? Well, they are both round but you can make a profit selling pizza!"