How many white people were terrorists and chose air travel as their means to spread terror? As a frequent flyer, all I have to say is thank goodness that the domestic terrorists to which you refer (Tim McVeigh, ELF and the like) haven't chosen to use airplanes to cause terror. Air travel would grind to a halt in the US.
You said: "well at least if the Chinese are going to ravage Tibet, they might help save the environment. So let us acknowledge and shed a tear for Tibet's sacrifice for all mankind." Honestly, the first sentence should be in the past tense. Have you ever been there? Monasteries which held 10,000 monks reduced to 6. Yes, 6. Most other monasteries obliterated entirely. Massive Han Chinese "settlement" in the Tibetan capital of Lhasa means it is barely Tibetan anymore. Breaks on the limit on number of children for Han Chinese that relocate to Tibet. And that is just recently. Go back a decade or two and you will find overwhelming military force used to subjugate and kill Tibetans. Mass rapes of Tibetan nuns. Plundering of the treasures of the potala. (The only reason the big gold bhudda remains is that it was too big/heavy to carry away.) Lets not forget the brilliant idea of forcing Tibetans to grow wheat instead of barley and the ensuing famine. Anyways, having been both there and Bhutan, I can say that in my observation, the Tibetan culture is dying. Contrasting it with the Bhutan culture, there is no way you could say the Tibetans have survived the Chinese onslaught. Even 10 years ago when I was there it was unmistakably a dying culture. The ravagement was more or less complete at that time. Now it is just time for the vultures to pick over the corpse (yes, a deliberate reference to sky burials).
I will second that. Hardest game I have every played (and I started with Wizardry on an Apple IIc). But it is also in the top 5 in terms of best/most rewarding.
Sorry, you are correct. But to clarify for those who don't care to click the link, the statistics themselves are from the National Center for Health, part of the CDC. So about as unbiased as one could reasonably expect without working through the numbers yourself.
Actually the number is about 4-5x times that. I am remembering 14,000, but you can find a breakdown on the NRA site (http://www.vpc.org/nrainfo/phil.html): 11,920 firearms homicides in the USA in 2003.
Aye. And while I know nothing about your writing skills, I am absolutely confident that you would do better than Travis Wright, based upon his work from Eagle Eye. Eagle Eye was a giant steaming pile of garbage with virtually nothing to recommend it in any respect. The dialogue was terrible, and the plot worse. Travis Wright should be locked up and never allowed access to a pen, computer, pencil, crayons, or anything else that he might use to right with ever again.
If I am not mistaken (and I might well be) most books these days are not manufactured on acid free paper. Given that, they will probably last 50-70 years at best. Unless you treat them to prevent decay, and store them in a temperature and humidity controlled environment, or sealed.
Having said that, I agree with you that longevity is key. We don't need material to be lost because data portability isn't available. However, given increasing power, the ready availability of tools today that let you strip out the text from proprietary formats, and virtualization (who cares if you can't read the file if you can virtualize a machine that can?) I am not terribly worried.
There are good and bad reasons for not buying an ebook reader, but I don't honestly think this is one of them. (And yes, I am a convert. I have had a Sony PRS-500 for 2.5 years, and just bought a 505.)
With respect to the "trust" issue: this is perception, not reality. Disk drive arrays of the type used for backup (virtual tape) have a reliability rating of 99.99% to 99.999% for uptime, and lose or corrupt data even less than that. The very best tape will approach 99.5% but only if it is not physically moved outside of a tape library--like it would, for example, it taken off site for disaster recovery purposes. The moment that humans start to handle it, bump it, get it dirty, etc, reliability drops below 99%, sometimes well below to the 95% level or so. So, very best case, disk is at least an order of magnitude more reliable than tape, and most often, several orders of magnitude.
full disclosure: I work for a company that sells virtual tape. Nevertheless, all the data is backed up by vendor neutral analysts like Gardner, and tape companies themselves.
Well, I am going to toss this out there even though it may be a little inflammatory: Jordan was stylistically awful. Beyond belief really. The writing devolved over the course of the series from adequate to extremely painful. Your claims:
- "rarely used filler prose". If you read books 5-9, you will find that increasingly, that is all they were. Phrases were repeated a painfully large number of times. Characters expressed the same emotions and reactions to various situations over and over and over again.
- "his writing for more than 20 main charcters was exceptional" Well, there was little if any character growth for the majority of characters since book 4. If anything, it is shockingly repetitive in that the character's changed in no significant way for so long. And female characters in particular tended to the caricaturish in their unidimensionality.
No matter what excuse you care to use, it is obvious that a much firmer editorial hand was required. The number of people that simply stopped reading (based on comments here, on Amazon, and other forums) is very large. It just went on. Not only were questions not answered, and plot details not resolved, but new, seemingly irrelevant questions were raised, and new plot threads started. You may want to defend the work as "3rd generation hard fantasy" but it reads a lot more like Edward Gibbon than anything that is remotely interesting or compelling as a work of fiction.
Finally, I would not that Jordan's work has nothing redeeming from a literary point of view either--there are no compelling themes explored in interesting or insightful ways; no compelling use of metaphor or allegory; no deep (or even shallow) discovery of human nature and growth through conflict; nothing tragic about the conflicts; nothing at all. So without plot and character, there is simply nothing at all of interest.
And for reference, I have worked through Martin's books without losing interest. And Erikson's (who has more happen in a chapter than WoT had in whole books) even though he is up to 7000+ pages. Glen Cooks. Gene Wolfe. Etc. But that doesn't mean that I am uncritical or read uncritically. Jordan lost the plot and jumped the shark a long time ago, and those problems are compounded by dreadful style, awful characterization, and the total absence of compelling plot developments. (I am also pretty critical of Goodkind for similar reasons--the prose is simply awful, as is the characterization.)
In addition to the link provided by the other poster, let me add my personal experience: e-ink really is as good as print, and much much better than LCD. No eyestrain at all. I have had a Sony ereader for over a year, and have read 20,000+ pages on it. It simply rocks. The resolution is stellar, the clarity of print (even at size small) is excellent, and it is very light. I have read at the beach, in a plane, in bed, on the couch, etc. There is no place where it is not as legible or more so than print, and vastly better than LCD. Yes, I have also tried LCD for reading--first a HPaq and then a PSP--and I intensely dislike the "glowing page" of backlit LCD. E-ink is just better. Finally, I would also say that the Sony product and format is so close to the experience of an actual book that it really does "draw me in" like a real book would. For those who have not seen e-ink or experienced how different it is from LCD, you owe it to yourselves to try it or go see a device that uses it.
Erm, no, it doesn't. Sorry. It uses Memory Stick, which is a Sony proprietary format. You can buy third party versions, but it is obvious that they pay a licensing fee to Sony. Compare prices for 2 GB memory stick to 2 GB SD both from Lexar, for example.
Is anybody else feeling the irony? A theory doesn't really seem to be "very old" if it is only 30 years only when we are discussing a universe over a trillion years old...
Well, I will reply to the top post in the vain hope that this gets moderated somewhere (positive). Amidst the vast number of posts on this subject, the overwhelming consensus seems to be that this is impossible. Unfortunately the overwhelming consensus in dead wrong.
Here is how it works. One, we are not talking about compression of individual files; we are talking about "compressing" a data stream composed of *a lot* of data. The more the better. 10s or 100s of TB (that's terrabytes, kids, 1000x a GB, 1000000 x a MB) is idea. The technology then compares blocks that the data is composed of, and retains pointers to the blocks that are the same, rather than the entire block. Therefore, it doesnt really matter if your data is composed of an Oracle database, a bunch of mp3s, and a few million Excel spreadsheets, it is all good.
Now for the catches: you need a lot of data, as mentioned. It helps if this is backup data, a lot, because you are highly likely to capture redundant files (either across different machines, or over time) and therefore rendundant blocks. Finally, it is slow. Very slow. When compared to an enterprise backup that might be capable of generating 1000 MB/s on aggregate to a few dozen tape drives, it is show-stoppingly slow. Think sub 100 MB/s.
For other competitors to datamojo, try Copan, Sepaton, and DataDomain.
And I will be the first to admit that I did not RTFA, and I do have the advantage of being familiar with a lot of backup technologies. So I don't know for sure, but if the article mentions any of the points I raised, it is depressing at best that out of the dozens of posts moderated positively that hardly a one mentions any of the facts above. Holy moly.
Will this mean that you can share storage more easily, maybe. It certainly seems to reduce sharks/ESS into an expensive interface for attaching discs.
Well, it is one thing to send output from thousands of nodes to thousands of nodes, and achieve very high performance, as seems to be the case here. It is another to send output from a single node to a single node as you describe, where one node is a storage controller (ESS, whatever) and the other is a large database server that is probably not running a file system at all, and may be generating as many (small) I/O requests as hundreds of those nodes in the first case are. Add on top of that resiliency, replication, caching, etc. and there is a great difference between a massively (horizontally) scaleable GPFS and a masssively (vertically) scaleable like a Shark. All of which is not necessarily to say that the two approaches won't merge at some point, but I suspect that point is quite far off--say 10 or 20 years--given the very different requirements that the two approaches attempt to meet.
Time to burn some karma. How the holy fuck does this nonsense get moderated insightful. This topic has been discussed repeatedly on/. Flogged to death in a rather ham fisted manner, one might say, if one were inclined to mix metaphors. And each and every time some number of posters need to be reminded, as does the parent here, that Google is not offering up entire copyrighted works for browsing, but rather just snippets. So, I ask, again: why does factually inaccurate nonsense get moderated as insightful? Anybody?
Yes. To force you to buy a new display device. New TV, new monitor for your computer(s), new video iPod/PSP/whatever. Just keep you buying. It is insulting and offensive to the consumer, but almost certainly true.
Another point: why not come up with a hardware standard that is sufficiently extensible that it supports future technologies? If consumer electronics is going to get onto the personal computer growth and revision curve, with new technologies every few months, it has to become more modular. I am not, and I suspect few others are, willing to upgrade everything every year or two because one component is upgraded. Just my opinion, but this is a critical weakness in the strategy of every major consumer electronics manufacturer at the moment.
Two more points, on the subject of cell phone costs in Canada. More like additional evidence on the impact of regulation, rather than saying you are wrong (because you are clearly not!).
1) One of the biggest reasons Telus bought Clearnet was the spectrum Clearnet owned. Just like in the US, the Canadian government had auctioned off spectrum rights, and Clearnet had a nice piece of spectrum. The value to Telus of Clearnet = spectrum rights + subscribers + reduced competition. At least the spectrum was the reason discussed internally at Telus at the time.
2) An article in the Globe and Mail about the cost of wireless: land lines are subsidized. Again, regulation has had the impact of increasing consumer prices; not to say that regulation is right and wrong, but more that regulation has wide-reaching and profound impacts in today's communications marketplace. And sadly, most regulators have no idea what they are monkeying with--I find it plenty believable that the CRTC is just not smart enough to figure out that regulating the price of land lines also impacts the price of cell service.
Well, full disclosure first: this is what I do all day, every day. Backup and recovery, that is.
So a few comments, because some of your thoughts worry me a bit:
Backup is expensive. Backup is insurance. Ideally you never need to recover, which means that all that money spent on backup is "wasted". That is life. It is an operational cost of doing business, that many businesses pay because they recognize that the risk of not having one outweighs the expense of doing it. But there it is. And I do recognize that smaller businesses (less than 1000 employees) usually have a harder time understanding the costs and justifying them. So how do you define "cost effective" in backup? There is no conceivable way that it can be run as a profit center and make money. Backups cost the business money, just like their insurance policy. How much is your data worth? That is the core question. How much is it worth? How much will you lose if you don't have access to it for some period of time?
Point number two is that the massively dominant mechanism for backup today is tape. I would venture to guess that 99.9% of business with more than 1000 employees backup to tape. And it is easy to break, burn, and steal. Again, that is life. We implement reasonable measures to ensure that it doesn't happen, but best practices say to make two copies of any tape: one for onsite, one for off. And yes, if you are concerned about data security (credit card transaction processors, banks, etc.) encrypt the tape. Yes it is expensive, and yet it is worth it. It is not "cost effective" except when you think that you may be out of business if your courier blows it, looses the tape, and the world finds out; just ask CardSystems Solutions. That is the context in which cost effective needs to be understood.
Further, what interfaces let you copy data at those speeds? Well there are more options than I can outline in this post: multiple SAN interfaces? Multiple LAN segments? RAID array snapshots? Host based mirroring? etc. Lots of options, all with good and bad things about them. But again, RAID arrays are used by 99.9% of businesses with more than 1000 employees. Server duplication is *not* backup. It may suffice for disaster recovery, but it has two enormous problems: it does not preserve data as it used to be at a given point in time (say, at financial quarter end), and it does propogate errors. Corrupted tablespace? You just replicated corruption. User just deleted files? While they just got deleted remotely too. Redundancy accomplishes several things, but none of them are backup. No auditor would be satisfied with that, and if they company is traded publicly, the CIO would likely be fired if that strategy was revealed to the public.
Well, full disclosure: I work for EMC. Having said that, let me respond to your statement about "losing grip" on 90s dominance.
EMC has achieved 11 straight quarters of double digit growth. Last quarter revenue grew 19%. EMC is clearly #1 by market share. In fact, there is an article here: http://www.businessweek.com/technology/content/aug 2005/tc2005084_5865_tc024.htm that well sums up the solid financial and market share position EMC is in.
I don't want to debate the merits of EMC technology here--this isn't the time or place. But your statement seemed to imply a dying on the vine (or mediocre achievement at best) for EMC, which is factually not the case.
And it is precisely becuase of this unjustified, indeed unjustifiable, rhetorical nonsense that I no longer purchase Wired. C'mon guys, that style of writing is not appropriate for a magazine that is pretending to be credible or meaningful. A comic book maybe (and I say that with no disrespect to comics). Apparently Wired has not figured out that the unjustified exuberance they exuded about all things techno prior to the dot bomb crash is no longer relevant.
Erm, not really true. A good RAID array will choose the drive with the head positioned closest to the data. Now I have no idea if this is standard on RAID controllers you would find in a small server, but it is certainly common on shared storage arrays.
However I can say from years of personal experience that database data (much of it) really does compress that well. Pictures don't compress worth a durn. Most file space stuff does about 1.5:1 to 2:1.
And I have no idea what algorithm LTO1/2 and SDLT have built into the drives, but it does work *really* well. And because it is firmware, it is pretty darn fast. As are the drives. Assuming you can sustain the stream on the system side, we find it pretty common to exceed 10 MB/s (and yes I mean bytes!) per drive.
The bottleneck really does become the ability to sustain fast transfers. For example, LTO2, 9980, 3592 are all rated at about 34 MB/sec, and then if you give them compressable data, can actually run at 65-70 MB/sec. (Bear in mind these drives cost $20-40k new). Now think about trying to get a dozen of them running effeciently at the same time!/rant off, I know you weren't talking about throughput. And your points about compression are dead on (except for the fact that it doesn't take any time, because it is done at the tape drive level).
Slashdot Mirror
How many white people were terrorists and chose air travel as their means to spread terror? As a frequent flyer, all I have to say is thank goodness that the domestic terrorists to which you refer (Tim McVeigh, ELF and the like) haven't chosen to use airplanes to cause terror. Air travel would grind to a halt in the US.
You said: "well at least if the Chinese are going to ravage Tibet, they might help save the environment. So let us acknowledge and shed a tear for Tibet's sacrifice for all mankind." Honestly, the first sentence should be in the past tense. Have you ever been there? Monasteries which held 10,000 monks reduced to 6. Yes, 6. Most other monasteries obliterated entirely. Massive Han Chinese "settlement" in the Tibetan capital of Lhasa means it is barely Tibetan anymore. Breaks on the limit on number of children for Han Chinese that relocate to Tibet. And that is just recently. Go back a decade or two and you will find overwhelming military force used to subjugate and kill Tibetans. Mass rapes of Tibetan nuns. Plundering of the treasures of the potala. (The only reason the big gold bhudda remains is that it was too big/heavy to carry away.) Lets not forget the brilliant idea of forcing Tibetans to grow wheat instead of barley and the ensuing famine. Anyways, having been both there and Bhutan, I can say that in my observation, the Tibetan culture is dying. Contrasting it with the Bhutan culture, there is no way you could say the Tibetans have survived the Chinese onslaught. Even 10 years ago when I was there it was unmistakably a dying culture. The ravagement was more or less complete at that time. Now it is just time for the vultures to pick over the corpse (yes, a deliberate reference to sky burials).
I will second that. Hardest game I have every played (and I started with Wizardry on an Apple IIc). But it is also in the top 5 in terms of best/most rewarding.
Sorry, you are correct. But to clarify for those who don't care to click the link, the statistics themselves are from the National Center for Health, part of the CDC. So about as unbiased as one could reasonably expect without working through the numbers yourself.
Actually the number is about 4-5x times that. I am remembering 14,000, but you can find a breakdown on the NRA site (http://www.vpc.org/nrainfo/phil.html): 11,920 firearms homicides in the USA in 2003.
Well, I am not trying to get work as a screenwriter in Hollywood, am I? Grammar Nazis.
Aye. And while I know nothing about your writing skills, I am absolutely confident that you would do better than Travis Wright, based upon his work from Eagle Eye. Eagle Eye was a giant steaming pile of garbage with virtually nothing to recommend it in any respect. The dialogue was terrible, and the plot worse. Travis Wright should be locked up and never allowed access to a pen, computer, pencil, crayons, or anything else that he might use to right with ever again.
If I am not mistaken (and I might well be) most books these days are not manufactured on acid free paper. Given that, they will probably last 50-70 years at best. Unless you treat them to prevent decay, and store them in a temperature and humidity controlled environment, or sealed.
Having said that, I agree with you that longevity is key. We don't need material to be lost because data portability isn't available. However, given increasing power, the ready availability of tools today that let you strip out the text from proprietary formats, and virtualization (who cares if you can't read the file if you can virtualize a machine that can?) I am not terribly worried.
There are good and bad reasons for not buying an ebook reader, but I don't honestly think this is one of them. (And yes, I am a convert. I have had a Sony PRS-500 for 2.5 years, and just bought a 505.)
Just to clarify: the performance expectation is 30x the number of IOPS than an equivalent capacity hard drive, 1/10th the response time.
With respect to the "trust" issue: this is perception, not reality. Disk drive arrays of the type used for backup (virtual tape) have a reliability rating of 99.99% to 99.999% for uptime, and lose or corrupt data even less than that. The very best tape will approach 99.5% but only if it is not physically moved outside of a tape library--like it would, for example, it taken off site for disaster recovery purposes. The moment that humans start to handle it, bump it, get it dirty, etc, reliability drops below 99%, sometimes well below to the 95% level or so. So, very best case, disk is at least an order of magnitude more reliable than tape, and most often, several orders of magnitude.
full disclosure: I work for a company that sells virtual tape. Nevertheless, all the data is backed up by vendor neutral analysts like Gardner, and tape companies themselves.
Well, I am going to toss this out there even though it may be a little inflammatory: Jordan was stylistically awful. Beyond belief really. The writing devolved over the course of the series from adequate to extremely painful. Your claims:
- "rarely used filler prose". If you read books 5-9, you will find that increasingly, that is all they were. Phrases were repeated a painfully large number of times. Characters expressed the same emotions and reactions to various situations over and over and over again.
- "his writing for more than 20 main charcters was exceptional" Well, there was little if any character growth for the majority of characters since book 4. If anything, it is shockingly repetitive in that the character's changed in no significant way for so long. And female characters in particular tended to the caricaturish in their unidimensionality.
No matter what excuse you care to use, it is obvious that a much firmer editorial hand was required. The number of people that simply stopped reading (based on comments here, on Amazon, and other forums) is very large. It just went on. Not only were questions not answered, and plot details not resolved, but new, seemingly irrelevant questions were raised, and new plot threads started. You may want to defend the work as "3rd generation hard fantasy" but it reads a lot more like Edward Gibbon than anything that is remotely interesting or compelling as a work of fiction.
Finally, I would not that Jordan's work has nothing redeeming from a literary point of view either--there are no compelling themes explored in interesting or insightful ways; no compelling use of metaphor or allegory; no deep (or even shallow) discovery of human nature and growth through conflict; nothing tragic about the conflicts; nothing at all. So without plot and character, there is simply nothing at all of interest.
And for reference, I have worked through Martin's books without losing interest. And Erikson's (who has more happen in a chapter than WoT had in whole books) even though he is up to 7000+ pages. Glen Cooks. Gene Wolfe. Etc. But that doesn't mean that I am uncritical or read uncritically. Jordan lost the plot and jumped the shark a long time ago, and those problems are compounded by dreadful style, awful characterization, and the total absence of compelling plot developments. (I am also pretty critical of Goodkind for similar reasons--the prose is simply awful, as is the characterization.)
In addition to the link provided by the other poster, let me add my personal experience: e-ink really is as good as print, and much much better than LCD. No eyestrain at all. I have had a Sony ereader for over a year, and have read 20,000+ pages on it. It simply rocks. The resolution is stellar, the clarity of print (even at size small) is excellent, and it is very light. I have read at the beach, in a plane, in bed, on the couch, etc. There is no place where it is not as legible or more so than print, and vastly better than LCD. Yes, I have also tried LCD for reading--first a HPaq and then a PSP--and I intensely dislike the "glowing page" of backlit LCD. E-ink is just better. Finally, I would also say that the Sony product and format is so close to the experience of an actual book that it really does "draw me in" like a real book would. For those who have not seen e-ink or experienced how different it is from LCD, you owe it to yourselves to try it or go see a device that uses it.
Erm, no, it doesn't. Sorry. It uses Memory Stick, which is a Sony proprietary format. You can buy third party versions, but it is obvious that they pay a licensing fee to Sony. Compare prices for 2 GB memory stick to 2 GB SD both from Lexar, for example.
Is anybody else feeling the irony? A theory doesn't really seem to be "very old" if it is only 30 years only when we are discussing a universe over a trillion years old...
Well, I will reply to the top post in the vain hope that this gets moderated somewhere (positive). Amidst the vast number of posts on this subject, the overwhelming consensus seems to be that this is impossible. Unfortunately the overwhelming consensus in dead wrong.
Here is how it works. One, we are not talking about compression of individual files; we are talking about "compressing" a data stream composed of *a lot* of data. The more the better. 10s or 100s of TB (that's terrabytes, kids, 1000x a GB, 1000000 x a MB) is idea. The technology then compares blocks that the data is composed of, and retains pointers to the blocks that are the same, rather than the entire block. Therefore, it doesnt really matter if your data is composed of an Oracle database, a bunch of mp3s, and a few million Excel spreadsheets, it is all good.
Now for the catches: you need a lot of data, as mentioned. It helps if this is backup data, a lot, because you are highly likely to capture redundant files (either across different machines, or over time) and therefore rendundant blocks. Finally, it is slow. Very slow. When compared to an enterprise backup that might be capable of generating 1000 MB/s on aggregate to a few dozen tape drives, it is show-stoppingly slow. Think sub 100 MB/s.
For other competitors to datamojo, try Copan, Sepaton, and DataDomain.
And I will be the first to admit that I did not RTFA, and I do have the advantage of being familiar with a lot of backup technologies. So I don't know for sure, but if the article mentions any of the points I raised, it is depressing at best that out of the dozens of posts moderated positively that hardly a one mentions any of the facts above. Holy moly.
Will this mean that you can share storage more easily, maybe. It certainly seems to reduce sharks/ESS into an expensive interface for attaching discs.
Well, it is one thing to send output from thousands of nodes to thousands of nodes, and achieve very high performance, as seems to be the case here. It is another to send output from a single node to a single node as you describe, where one node is a storage controller (ESS, whatever) and the other is a large database server that is probably not running a file system at all, and may be generating as many (small) I/O requests as hundreds of those nodes in the first case are. Add on top of that resiliency, replication, caching, etc. and there is a great difference between a massively (horizontally) scaleable GPFS and a masssively (vertically) scaleable like a Shark. All of which is not necessarily to say that the two approaches won't merge at some point, but I suspect that point is quite far off--say 10 or 20 years--given the very different requirements that the two approaches attempt to meet.
Time to burn some karma. How the holy fuck does this nonsense get moderated insightful. This topic has been discussed repeatedly on /. Flogged to death in a rather ham fisted manner, one might say, if one were inclined to mix metaphors. And each and every time some number of posters need to be reminded, as does the parent here, that Google is not offering up entire copyrighted works for browsing, but rather just snippets. So, I ask, again: why does factually inaccurate nonsense get moderated as insightful? Anybody?
"is there any other reason to introduce UDI?"
Yes. To force you to buy a new display device. New TV, new monitor for your computer(s), new video iPod/PSP/whatever. Just keep you buying. It is insulting and offensive to the consumer, but almost certainly true.
Another point: why not come up with a hardware standard that is sufficiently extensible that it supports future technologies? If consumer electronics is going to get onto the personal computer growth and revision curve, with new technologies every few months, it has to become more modular. I am not, and I suspect few others are, willing to upgrade everything every year or two because one component is upgraded. Just my opinion, but this is a critical weakness in the strategy of every major consumer electronics manufacturer at the moment.
1) One of the biggest reasons Telus bought Clearnet was the spectrum Clearnet owned. Just like in the US, the Canadian government had auctioned off spectrum rights, and Clearnet had a nice piece of spectrum. The value to Telus of Clearnet = spectrum rights + subscribers + reduced competition. At least the spectrum was the reason discussed internally at Telus at the time.
2) An article in the Globe and Mail about the cost of wireless: land lines are subsidized. Again, regulation has had the impact of increasing consumer prices; not to say that regulation is right and wrong, but more that regulation has wide-reaching and profound impacts in today's communications marketplace. And sadly, most regulators have no idea what they are monkeying with--I find it plenty believable that the CRTC is just not smart enough to figure out that regulating the price of land lines also impacts the price of cell service.
Well, full disclosure first: this is what I do all day, every day. Backup and recovery, that is.
So a few comments, because some of your thoughts worry me a bit:
Backup is expensive. Backup is insurance. Ideally you never need to recover, which means that all that money spent on backup is "wasted". That is life. It is an operational cost of doing business, that many businesses pay because they recognize that the risk of not having one outweighs the expense of doing it. But there it is. And I do recognize that smaller businesses (less than 1000 employees) usually have a harder time understanding the costs and justifying them. So how do you define "cost effective" in backup? There is no conceivable way that it can be run as a profit center and make money. Backups cost the business money, just like their insurance policy. How much is your data worth? That is the core question. How much is it worth? How much will you lose if you don't have access to it for some period of time?
Point number two is that the massively dominant mechanism for backup today is tape. I would venture to guess that 99.9% of business with more than 1000 employees backup to tape. And it is easy to break, burn, and steal. Again, that is life. We implement reasonable measures to ensure that it doesn't happen, but best practices say to make two copies of any tape: one for onsite, one for off. And yes, if you are concerned about data security (credit card transaction processors, banks, etc.) encrypt the tape. Yes it is expensive, and yet it is worth it. It is not "cost effective" except when you think that you may be out of business if your courier blows it, looses the tape, and the world finds out; just ask CardSystems Solutions. That is the context in which cost effective needs to be understood.
Further, what interfaces let you copy data at those speeds? Well there are more options than I can outline in this post: multiple SAN interfaces? Multiple LAN segments? RAID array snapshots? Host based mirroring? etc. Lots of options, all with good and bad things about them. But again, RAID arrays are used by 99.9% of businesses with more than 1000 employees. Server duplication is *not* backup. It may suffice for disaster recovery, but it has two enormous problems: it does not preserve data as it used to be at a given point in time (say, at financial quarter end), and it does propogate errors. Corrupted tablespace? You just replicated corruption. User just deleted files? While they just got deleted remotely too. Redundancy accomplishes several things, but none of them are backup. No auditor would be satisfied with that, and if they company is traded publicly, the CIO would likely be fired if that strategy was revealed to the public.
Maybe, just maybe, because the same comanies that own the entertainment conglomerates also own most of the mainstream media as well? Just a thought...
Well, full disclosure: I work for EMC. Having said that, let me respond to your statement about "losing grip" on 90s dominance. EMC has achieved 11 straight quarters of double digit growth. Last quarter revenue grew 19%. EMC is clearly #1 by market share. In fact, there is an article here: http://www.businessweek.com/technology/content/aug 2005/tc2005084_5865_tc024.htm that well sums up the solid financial and market share position EMC is in.
I don't want to debate the merits of EMC technology here--this isn't the time or place. But your statement seemed to imply a dying on the vine (or mediocre achievement at best) for EMC, which is factually not the case.
And it is precisely becuase of this unjustified, indeed unjustifiable, rhetorical nonsense that I no longer purchase Wired. C'mon guys, that style of writing is not appropriate for a magazine that is pretending to be credible or meaningful. A comic book maybe (and I say that with no disrespect to comics). Apparently Wired has not figured out that the unjustified exuberance they exuded about all things techno prior to the dot bomb crash is no longer relevant.
Erm, not really true. A good RAID array will choose the drive with the head positioned closest to the data. Now I have no idea if this is standard on RAID controllers you would find in a small server, but it is certainly common on shared storage arrays.
All more or less true.
/rant off, I know you weren't talking about throughput. And your points about compression are dead on (except for the fact that it doesn't take any time, because it is done at the tape drive level).
However I can say from years of personal experience that database data (much of it) really does compress that well. Pictures don't compress worth a durn. Most file space stuff does about 1.5:1 to 2:1.
And I have no idea what algorithm LTO1/2 and SDLT have built into the drives, but it does work *really* well. And because it is firmware, it is pretty darn fast. As are the drives. Assuming you can sustain the stream on the system side, we find it pretty common to exceed 10 MB/s (and yes I mean bytes!) per drive.
The bottleneck really does become the ability to sustain fast transfers. For example, LTO2, 9980, 3592 are all rated at about 34 MB/sec, and then if you give them compressable data, can actually run at 65-70 MB/sec. (Bear in mind these drives cost $20-40k new). Now think about trying to get a dozen of them running effeciently at the same time!