I did something similar once (about five years ago).
We had a set of laptops which we used for training. It was always a pain in the a*** setting them up at the start of the course, and you could never be quite sure that you had done it properly.
There were also issues with IP addresses - for various reasons we would have liked to have had the same IP on each machine.
The solution I came up was to run linux as the base install with VMware running on that. A standard Windows NT 4 image ran inside the VM. Iptables was used to NAT the VM so that each VM thought it had the same IP address.
On boot up, linux would start X and then start VMware running full screen. If the user wasn't watching closely, they wouldn't even know they were running Linux. The only real give-away was that NT4 was a lot faster running virtualized than it was running natively on the same hardware. (This might have been due to more efficient disk caching by linux).
Resetting the training machines to their initial state was simply a matter of copying the VM file from a server to the laptop.
It was just a shame that the company suffered dot-com burnout not long after this was developed.
A recent Horizon Program on the BBC says that the final toll of deaths attributable to Chernobyl is more likely to be in the region of 56.
Apperently this is because low levels of radiation are far less dangerous than though to be the case even twenty years ago. Statistical evidence even suggest that low levels (equivalent to a high level of normal background radiation) may in fact have a positive effect on cancer rates.
Now how many people have died in the last year working in the oil extraction, coal mining, gas supply industries?
Where I live we have a 60MW/90KV DC transmission cable that runs over 30 miles under the sea to France. That is one way of sorting the inductance problem. I couldn't say how this compares with the cost of overhead transmission, but the total cost of the link was about £30million. (But that would include switching equipment at both ends, and the fact that the cable has been laid underwater and in a much more hostile environment than would be the case on land.
There are places that have an entirely underground power grid. I know because I live in one.
All of our high and low voltage supplies are underground (66/33KV transmission and 240V household supplies).
The benefits are the reduced susceptibility to wind damage, and the aesthetic issue of not having large, bulky cables handing from poles everywhere. (If only the telephone company could be persuaded to take the same approach!).
The downside is that the costs of any repair, taken in isolation, are probably higher. However the probability of a repair being required (due to wind damage, trees) is substantially lower.
The fact that I live in a small (and windy) island (Guernsey) in the Atlantic off the French coast is another factor. Relative prosperity a further factor - although it has been a policy here to put power lines underground since at least the second world war (when things were much less prosperous). Housing density must be another issue - Guernsey is relatively densely populated (1000 people per square mile on average). It is notable that in the less densely populated "country" areas there are a few overhead supplies to remote houses.
Are the costs of maintaining underground electric cables really any worse that those for other underground untilities - such as water or gas?
The same restrictions apply to SQL Server replication of database tables.
I think you are missing something fairly fundamental if you don't understand why you need a primary key on a database table in order to replicate it. How else do you know which row to update or delete? How do you handle multi-master replication? How do you deal with otherwise identical rows added to two replicated databases?
And I'm sure that this rumor has nothing to do with the news that IBM is trying to sell it's "Ibm PC" business. Could it be that it sees more margin in Apples than PCs?
The technology that Amazon use to make product recommendations is based on a process known as "collaborative filtering".
This was first developed at the University of Minnesota in the early/mid 90's as a way of recommending newsgroup articles that students might be interested in. It was found that the same methods could be used to usefully recommend movies and other products that customers might recommend.
The essence of collaborative filtering is "Customer A brought products X and Y and other customers who brought X and Y also brought Z, therfore recommend product Z to customer A". The hard part of collaborative filtering is developing algorithms that turn what my be hundreds of millions of lines of purchase history into useful set of rules, and doing it quickly enough to be usable inthe real world.
The technology from the U of Minnesota was developed by John Reidl and others through a company called Net Perceptions inthe mid 90's (www.netpercetpions.com). They developed the product called Grouplens. One of the first customers for this product was Amazon. Amazon actually brought a source code licence to Grouplens, and while they continued to be a Net Perceptions customer through into about 2001 they actually further developed the original Grouplens code internally. What Amazon currently use is probably quite different from the original NetP product, but would presumably be still covered by the source code license from NetP, which would also cover the patents held by NetP on collaborative filtering.
Since Netp was effectively wound up last year, their patent portfolio has been sold to Thalveg Data Flow LLC (see http://biz.yahoo.com/e/040401/netp.pk8-k.html)
I seriously doubt that Cendant have much to offer in terms of original work in this area. Certainly NetP had/has prior art going back into the ealy 90's. I can't see how the Cendant Patent can stand (and I can't see how it doesn't conflict with NetP patents either.)
So HP have "chosen" to use iTunes and the iPod. Didn't David Fester just say that Windows was all about "choice"...or does that comment only apply to "choices" approved by Big Brother Bill^H^H^H^H^H^H Microsoft
I was recently at a school open day at my old school. I noticed that they had the original Commodore PET 8032 on display that they acquired in, oh, about 1978. (This is a 1MHz 6502, 32K of RAM, external casette tape drive) I remember it's arrivial when I was about 14 or 15. No idea if it still works (I guess not), but memories of playing space invaders on that tiny screen....
This may seem like a stuipd question, but in the US do you generally pay to send or to receive SMS?
In Europe (and most of the rest of the world that uses GSM networks AFAIK) you generally pay a small fee (about 10c) to send each SMS message. This makes sending SMS spam far more expensive than with e-mail, as a result of which there is no widespread problem of SMS spam.
As I understand the business of the ships that take the elves into the west (to Valinor), an elven ship bearing elves that sailed into the west would sail on the Straight Road to Valinor (as they would have before the drowning of Numenor), whereas a ship bearing mortals would follow the curvature of the earth and so never come to Valinor. (As described in the Silmarillion at the very end of the Akallabeth)
The important words are There is now no ship that would bear me hence. I think that what Arwen means by this is that having chosen a mortal life, and being mortal herself, were she to sail on any ship going into the west then it would follow the mortal path and never reach Valinor.
The only mortals ever permitted to pass into the West on elvish ships (by special favour of the Valar) were the ringbearers (Bilbo, Frodo, Samwise)
Of course this begs the question of where Legolas and Gimli ended up (given that dwarves, while being far more long-lived than men, are still essentially mortal)
There is the UK air traffic control systems, which were based on 1960-era systems. This doesn't quite count as they were replaced by new state of the art systems last year. I hear that they may even have finished debugging the new systems;-)
I do "in house" tech support for my wife. One of today's gems (bless her)
"How do I get a single quote?" I point (patiently?) at the "/' key. "I thought that was a comma!" "What's this key?" (pointing #$*patiently$W%$# at the/, key.
I think it all depends... on the VCR and the hard drive in question - and how the hard drive is mounted.
I used to have a Dell laptop with an embarassingly noisy hard drive that stood comparison with a circular saw. I replaced it with a new (bigger) drive a few months ago, and the same laptop is almost silent. The funny thing is that when I put the old drive in to the laptop as a removable second hard drive it isn't particularly noisy, which makes me think that there was a problem with the way it was mounted before.
I wouldn't see hard drives a being any less reliable than VCR's. Of course Hard Drives probably have different failure modes (VCR's seem to fail in the tape head: tapes get eaten (by the VCR) or tapes degrade over time.) A hard drive does have the potential to take out all of your recordings with one failure - but that may not be a huge problem if you use the HD mainly for time-shifting but buy other stuff that you want to keep on DVD.
The idea of making product recommendations based on click-stream history is not a new concept. I know of several customers of my employer that have implemented, or are considering implementing it as a "fall-back" product recommendation strategy.
The idea of click-stream based recommendations was under discussion in the collaborative filtering field at least three years ago (if not longer). I fail to see how Amazon can make this application fly past claims of prior art.
This is a problem of the immaturity of the technology.
The first writing was almost certainly not very future-compatible either (marks on rocks, impressions on wax tablets) By the time of the original Domesday book the "technology" of writing was several millennia old and quite mature (i.e. even then they knew how to make inks and papers that would last at least a few decades if not longer)
The Domesday book project on the BBC micro should be put in context. It was built to run on a technology that was about 6 years old (first microcomputer 1976 - first BBC micro 1982). The video disks that it used were not terribly common even in the early 80's - they may have been better quality than VHS cassettes, but you couldn't *record* to them. Even the BBC Micro that it ran on looks laughable by todays standards (a 4MHz 6502 with 32KB of RAM (yes KB), no hard drive, 5 1/4 inch floppies, at most 16KB of graphics memory. (Of course a 3GHz PC with 1GB Ram and a 120GB hard drive will seem equally inadequate in another 20 years time)
We will have much more future-proof technology in 50-100 years time when things have had time to settle down and mature a little. By then we should have a much more realistic idea of what the life expectancy of a CD is (for example) and what steps we need to take to create media that is more long-lasting.
The most important time period is probably the next half century when we need to work on preserving those archives of dead data, rolling them forward onto new media and (possibly) into new data formats. Just hope that in 50 years time someone has a better idea than we do now what makes for a long-term digital archive format;-)
E.E. "Doc" smith was definitely great in his day, and is still a good read. The fact much of his work is still available when so much else from that period is forgotten is prehaps an indication of how good it is.
Considering that the first Lensman books are closer in time to H.G. Wells and Jules Verne than to the present day, they are still remarkably readable and even the science in them dosen't seem too out of date..
Definitely worth spending time on if you have some to spare.
Amen to that. There are very few authors where I will buy a new book on the strength of the author's name alone and without asking if I will actually have time to read thge book. Iain Banks has to stand at the head of that list (which would also have to include Peter F Hamilton, Vernor Vinge, Jo Clayton, and Stephen Donaldson)
Of course there are many other good authors I may buy if I think I will have the time to read them and I like the look of the book ( but I already have one shelf full of SF I haven't found time to read yet !-()
What is the cost of a closed source solution which 5 years down the line is no longer supported by the original supplier?
(who has since been taken over twice by companies that have no interest in your original product.)
Their response to your problem? - "give us $100K to upgrade to the next version of our product".. which you won't be able to use without major re-engineering because it is incompatible with your original software.
An everyday story of the closed source software industry....
Slashdot Mirror
I did something similar once (about five years ago).
We had a set of laptops which we used for training. It was always a pain in the a*** setting them up at the start of the course, and you could never be quite sure that you had done it properly.
There were also issues with IP addresses - for various reasons we would have liked to have had the same IP on each machine.
The solution I came up was to run linux as the base install with VMware running on that. A standard Windows NT 4 image ran inside the VM. Iptables was used to NAT the VM so that each VM thought it had the same IP address.
On boot up, linux would start X and then start VMware running full screen. If the user wasn't watching closely, they wouldn't even know they were running Linux. The only real give-away was that NT4 was a lot faster running virtualized than it was running natively on the same hardware. (This might have been due to more efficient disk caching by linux).
Resetting the training machines to their initial state was simply a matter of copying the VM file from a server to the laptop.
It was just a shame that the company suffered dot-com burnout not long after this was developed.
A recent Horizon Program on the BBC says that the final toll of deaths attributable to Chernobyl is more likely to be in the region of 56.
Apperently this is because low levels of radiation are far less dangerous than though to be the case even twenty years ago. Statistical evidence even suggest that low levels (equivalent to a high level of normal background radiation) may in fact have a positive effect on cancer rates.
Now how many people have died in the last year working in the oil extraction, coal mining, gas supply industries?
DC does have its place.
e 21stcentury.asp/
Where I live we have a 60MW/90KV DC transmission cable that runs over 30 miles under the sea to France. That is one way of sorting the inductance problem. I couldn't say how this compares with the cost of overhead transmission, but the total cost of the link was about £30million. (But that would include switching equipment at both ends, and the fact that the cable has been laid underwater and in a much more hostile environment than would be the case on land.
see http://www.electricity.gg/about/companyhistory/th
I believe there is a similar 2000MW link between the UK and France acorss the English Channel.
And if you think backhoe fade is a problem, just think what a trawling fishing boat could do to your power cable.
There are places that have an entirely underground power grid. I know because I live in one.
All of our high and low voltage supplies are underground (66/33KV transmission and 240V household supplies).
The benefits are the reduced susceptibility to wind damage, and the aesthetic issue of not having large, bulky cables handing from poles everywhere. (If only the telephone company could be persuaded to take the same approach!).
The downside is that the costs of any repair, taken in isolation, are probably higher. However the probability of a repair being required (due to wind damage, trees) is substantially lower.
The fact that I live in a small (and windy) island (Guernsey) in the Atlantic off the French coast is another factor. Relative prosperity a further factor - although it has been a policy here to put power lines underground since at least the second world war (when things were much less prosperous).
Housing density must be another issue - Guernsey is relatively densely populated (1000 people per square mile on average). It is notable that in the less densely populated "country" areas there are a few overhead supplies to remote houses.
Are the costs of maintaining underground electric cables really any worse that those for other underground untilities - such as water or gas?
For some reason I read it this way when I first saw it. Do I need to get out more?
The same restrictions apply to SQL Server replication of database tables.
I think you are missing something fairly fundamental if you don't understand why you need a primary key on a database table in order to replicate it. How else do you know which row to update or delete? How do you handle multi-master replication?
How do you deal with otherwise identical rows added to two replicated databases?
Amazon originally acquired their recommendation technology froma company called Group Lens (later Net Perceptions ) in about 1997.
The whole Collaborative Filtering Recommender system was developed at the University of Minnesota by John Reidl and others in the Mid 90's.
I can't see Amazon ever getting round the extensive prior art on this subject. The patent can only be purely defensive.
And I'm sure that this rumor has nothing to do with the news that IBM is trying to sell it's "Ibm PC" business. Could it be that it sees more margin in Apples than PCs?
The technology that Amazon use to make product recommendations is based on a process known as "collaborative filtering".
.
This was first developed at the University of Minnesota in the early/mid 90's as a way of recommending newsgroup articles that students might be interested in. It was found that the same methods could be used to usefully recommend movies and other products that customers might recommend.
The essence of collaborative filtering is "Customer A brought products X and Y and other customers who brought X and Y also brought Z, therfore recommend product Z to customer A". The hard part of collaborative filtering is developing algorithms that turn what my be hundreds of millions of lines of purchase history into useful set of rules, and doing it quickly enough to be usable inthe real world.
The technology from the U of Minnesota was developed by John Reidl and others through a company called Net Perceptions inthe mid 90's (www.netpercetpions.com). They developed the product called Grouplens. One of the first customers for this product was Amazon. Amazon actually brought a source code licence to Grouplens, and while they continued to be a Net Perceptions customer through into about 2001 they actually further developed the original Grouplens code internally. What Amazon currently use is probably quite different from the original NetP product, but would presumably be still covered by the source code license from NetP, which would also cover the patents held by NetP on collaborative filtering
Since Netp was effectively wound up last year, their patent portfolio has been sold to Thalveg Data Flow LLC (see http://biz.yahoo.com/e/040401/netp.pk8-k.html)
I seriously doubt that Cendant have much to offer in terms of original work in this area. Certainly NetP had/has prior art going back into the ealy 90's. I can't see how the Cendant Patent can stand (and I can't see how it doesn't conflict with NetP patents either.)
10 years abo I can recall paying out over $300 for 8MB memory modules. I don't think we have a problem yet....
So HP have "chosen" to use iTunes and the iPod. Didn't David Fester just say that Windows was all about "choice" ...or does that comment only apply to "choices" approved by Big Brother Bill^H^H^H^H^H^H Microsoft
I was recently at a school open day at my old school. I noticed that they had the original Commodore PET 8032 on display that they acquired in, oh, about 1978. (This is a 1MHz 6502, 32K of RAM, external casette tape drive) I remember it's arrivial when I was about 14 or 15. No idea if it still works (I guess not), but memories of playing space invaders on that tiny screen....
This may seem like a stuipd question, but in the US do you generally pay to send or to receive SMS?
In Europe (and most of the rest of the world that uses GSM networks AFAIK) you generally pay a small fee (about 10c) to send each SMS message. This makes sending SMS spam far more expensive than with e-mail, as a result of which there is no widespread problem of SMS spam.
As I understand the business of the ships that take the elves into the west (to Valinor), an elven ship bearing elves that sailed into the west would sail on the Straight Road to Valinor (as they would have before the drowning of Numenor), whereas a ship bearing mortals would follow the curvature of the earth and so never come to Valinor. (As described in the Silmarillion at the very end of the Akallabeth)
The important words are There is now no ship that would bear me hence. I think that what Arwen means by this is that having chosen a mortal life, and being mortal herself, were she to sail on any ship going into the west then it would follow the mortal path and never reach Valinor.
The only mortals ever permitted to pass into the West on elvish ships (by special favour of the Valar) were the ringbearers (Bilbo, Frodo, Samwise)
Of course this begs the question of where Legolas and Gimli ended up (given that dwarves, while being far more long-lived than men, are still essentially mortal)
There is the UK air traffic control systems, which were based on 1960-era systems. ;-)
This doesn't quite count as they were replaced by new state of the art systems last year. I hear that they may even have finished debugging the new systems
I do "in house" tech support for my wife. One of today's gems (bless her)
/, key.
"How do I get a single quote?"
I point (patiently?) at the "/' key.
"I thought that was a comma!"
"What's this key?" (pointing #$*patiently$W%$# at the
I can just see it now... petrol stations just outside the German border advertising cheap petrol and PC's at 16% off!
I would rephrase that...
The best bugs never get specified
The next best bugs never get designed
The next best bugs never get written...and so on.
The i-Magic virtual reality cycle turbo trainer has started appearing in the cycling press over the last year or so. Details at:-
Tacx web site
This is not aimed at the game player, and rather more at the hard core cyclist who wants to train in the winter, but the principle is the same.
I think it all depends... on the VCR and the hard drive in question - and how the hard drive is mounted.
I used to have a Dell laptop with an embarassingly noisy hard drive that stood comparison with a circular saw. I replaced it with a new (bigger) drive a few months ago, and the same laptop is almost silent. The funny thing is that when I put the old drive in to the laptop as a removable second hard drive it isn't particularly noisy, which makes me think that there was a problem with the way it was mounted before.
I wouldn't see hard drives a being any less reliable than VCR's. Of course Hard Drives probably have different failure modes (VCR's seem to fail in the tape head: tapes get eaten (by the VCR) or tapes degrade over time.) A hard drive does have the potential to take out all of your recordings with one failure - but that may not be a huge problem if you use the HD mainly for time-shifting but buy other stuff that you want to keep on DVD.
The idea of making product recommendations based on click-stream history is not a new concept. I know of several customers of my employer that have implemented, or are considering implementing it as a "fall-back" product recommendation strategy.
The idea of click-stream based recommendations was under discussion in the collaborative filtering field at least three years ago (if not longer). I fail to see how Amazon can make this application fly past claims of prior art.
This is a problem of the immaturity of the technology.
;-)
The first writing was almost certainly not very future-compatible either (marks on rocks, impressions on wax tablets) By the time of the original Domesday book the "technology" of writing was several millennia old and quite mature (i.e. even then they knew how to make inks and papers that would last at least a few decades if not longer)
The Domesday book project on the BBC micro should be put in context. It was built to run on a technology that was about 6 years old (first microcomputer 1976 - first BBC micro 1982). The video disks that it used were not terribly common even in the early 80's - they may have been better quality than VHS cassettes, but you couldn't *record* to them. Even the BBC Micro that it ran on looks laughable by todays standards (a 4MHz 6502 with 32KB of RAM (yes KB), no hard drive, 5 1/4 inch floppies, at most 16KB of graphics memory.
(Of course a 3GHz PC with 1GB Ram and a 120GB hard drive will seem equally inadequate in another 20 years time)
We will have much more future-proof technology in 50-100 years time when things have had time to settle down and mature a little. By then we should have a much more realistic idea of what the life expectancy of a CD is (for example) and what steps we need to take to create media that is more long-lasting.
The most important time period is probably the next half century when we need to work on preserving those archives of dead data, rolling them forward onto new media and (possibly) into new data formats. Just hope that in 50 years time someone has a better idea than we do now what makes for a long-term digital archive format
E.E. "Doc" smith was definitely great in his day, and is still a good read. The fact much of his work is still available when so much else from that period is forgotten is prehaps an indication of how good it is.
Considering that the first Lensman books are closer in time to H.G. Wells and Jules Verne than to the present day, they are still remarkably readable and even the science in them dosen't seem too out of date..
Definitely worth spending time on if you have some to spare.
Amen to that. There are very few authors where I will buy a new book on the strength of the author's name alone and without asking if I will actually have time to read thge book. Iain Banks has to stand at the head of that list (which would also have to include Peter F Hamilton, Vernor Vinge, Jo Clayton, and Stephen Donaldson)
Of course there are many other good authors I may buy if I think I will have the time to read them and I like the look of the book ( but I already have one shelf full of SF I haven't found time to read yet !-()
What is the cost of a closed source solution which 5 years down the line is no longer supported by the original supplier?
(who has since been taken over twice by companies that have no interest in your original product.)
Their response to your problem? - "give us $100K to upgrade to the next version of our product".. which you won't be able to use without major re-engineering because it is incompatible with your original software.
An everyday story of the closed source software industry....