Directed by Howard Hawks and a William Faulkner screenplay adaption, it was a complicated story that was hard to translate to the screen -- claiming that Hawks and Faulkner didn't get it is a bit of a stretch. Raymond Chandler's books all seem to have enough twists and turns and strange motivational logic to drive any screenplay writer nutty just try to make them into understandable films. Practical realities like budgets, morality committees and so on and you wonder how some of them even made it to the screen.
I thought the movie was really true to the book, in fact too true for a movie. It wasn't he movie length but the fact that the story could and should have been streamlined or condensed somehow to make it a little less complex, but then the Tolkies would have had a snit and the movie would have gotten bad buzz.
As it stands now I thought the plotline was only slightly less complex than the Big Sleep -- too many characters, too many drastic scene transitions to make it flow smoothly.
All VP/CxO types are all about "strategy" (ie, going to meetings, drinking lots of Starbucks, and so on). "Tactics", or what OS to run in the data center, they could care less about unless it costs them money or gets their boss pissed off.
my parents who have no clue what the difference between a bit and a byte is can start rattling off MS Windows deficiencies, but it (will/does/is going to) take full fledge engineers to discover the same in the Unix systems.
Your parents probably have significant end-user experience with Microsoft systems (office, at home, etc) and therefor can list end-user deficiences. Your parents' lack of technical knowledge means they can't tell you about technical deficiencies, just things they don't like.
There are few large system sales that take place that don't involve a sales engineer, since the sales guy is often some geek in a suit that sold cars last week and will be selling copiers next week. The sales engineer is the one that gives you all the biased benchmarks and tells you it will fit your racks, etc.
I'd guess it means that they've hired some ex-Sun sales engineers and other unix sales people and paid them big piles of money to exaggerate the deficiencies of Sun and Linux systems relative to the strengths of MS systems.
It sounds like dumb direct-sequence is the problem, not frequency-hopping. Frequency hopping sounds like the *solution* to 2.4Ghz problems since the devices can hop around to find open channels or less interference.
It'd be nice if there had been an industry-wide bootup spec for looking to see if 2.4Ghz channels were in use and picking other channels if they were, or just make everything on 2.4ghz do FH SS.
Most smart managers want to fix a problem before it bites them. The fact that the name of the company ain't in the news has little to do with the amount of internal heat people are facing. You can bet your ass that the MSNBC called a lot of the company's management asking "Did you know how easy your gift cards are to rip off????" and the person in charge of the gift card program, who had probably touted its security previously, will be sitting in the boss' office on Jan 2 answering some hard questions.
At least that's how it'd work where I work.
Why not just assign PINs at purchase?
on
Gift Card Hacking
·
· Score: 2
Why not just assign a PIN number, stored in the store computer, not on the card, when the card is bought and charged?
Sure some yokels would write the number on the card and get it lifted or lose it, but the same could happen to cash.
Requiring extra information not available on the card would be ideal and would make the type of counterfeiting described in the article very difficult, as long as there was no simple way of resetting PINs. It wouldn't prevent inside jobs or people laundering stolen credit cards, but those types will always be hard to stop.
Minnesota Walmarts have them at the checkout
on
Gift Card Hacking
·
· Score: 1
Which is a good thing, because at the Walmart in my area "Customer Service" more closely resembles the customs area of an east-African country than a place where you go to get helped.
Swell. So there's no significant economic reason for that company to change their policies yet.
Sure there is, its the internal economic justification of the manager in charge of the gift card program. The boss is likely to hear about this, and when (s)he does (s)he will either change the program or get canned.
No one wants an easy-to-rip-off gift card system. It invites attack from other fraud artists (if this system is lax, then others likely are too), pisses off customers and ruins loyalty.
The larger problem is that there's little financial incentive for stores to fix the problem generally (other than being seen as generally lax), since the losses aren't their own, they're someone else's, and even hijacked cards are money made for the store.
Apple won't or can't compete in the commodity corporate desktop world and trying to expand that market would be a waste of time and money. The niche markets they do dominate, such as print production, are suffering to some extent from stagnation. The markets aren't growing bigger and with the general softness in the ad markets, I can tell you (as an ad industry employee) that budgets to replace B&W G3s with G4s wholesale aren't going to be there like they were 2-3 years ago when G3s rapidly replaced earlier PPC Macs -- there's little end user demand and ZERO management push.
Buying SGI would provide Apple with an entry into a world of higher-end computing than they currently have and would enable them to provide a much more vertically integrated solution to markets that are somewhat out of reach for them in terms of software and hardware -- high end film production, animation, and scientific visualization. From a technology perspective, it would give a credibility boost to Apple's nascent Unix and allow them to have hardware unified by a single OS.
It may be arguable that Apple's credibility in creative circles, early-to-market product offerings, and increasingly high performance machines will give it the bottom third of the video production market by default, and that SGIs technology is rapidly being obsoleteed by commodity hardware.
However, I don't think that there's nearly the growth prospect in desktop video that there was in desktop publishing or the huge edge over x86, either. And own its own, Apple still can't escape the low-end niche it sits in.
I thought that most of the cable and broadcast industry were "opposed" to HDTV rollout not because of the copy protection issues but because they wanted to slice 'n' dice the bandwidth for multiple channels.
As much an infrastructure problem for the CATV and a theres-only-so-many-free-MHZ problem for the broadcast people.
Since today's unused bandwidth cannot be used tomorrow, it stands to reason that any use of bandwidth up to whatever the congestion point is cannot be seen as waste. The only really wasted bandwidth is *unused* bandwidth.
You can make a value judgement about the quality of the information being transited but you cannot call that bandwidth wasted.
Some people buy "bandwidth", but what they're really buying is the right to transit X amount of bits per month for a fixed amount of money. But the same thing about congestion holds true for this kind of bandwidth -- as long as you don't go over your monthly cap, only used bits are wasted.
We've run into numerous service providers that use RFC1918 private address space. More than one has said, when setting up a routed connection, "just route 10.0.0.0/8 to us..." when we're already using 10.0.0.0 on some wan links as well as dealing with other providers who use 10.0.0.0. We've had to deal with providers that end up NATing connections two and three times because they and their vendors are all using the same private address space. The debug time on routing and access control glitches increases logarithmically for each NAT translation.
That address space works well for the WAN side of private links, or for testing or other stub networks that can't be connected elsewhere. But if you even think you might interconnect with other providers or other organizations, get real addresses.
Is the concept itself as flawed as it seems, or is this just yet another case of Microsoft's implementation of something being flawed?
I think the MS implementation is the problem, not the concept. Most people get a bee in their bonnet about this because they think it breaks the NAT "security" model.
Problem is, NAT provides security because it breaks routing, not because it is a security system by itself. That someone has come up with a routing/networking technique that keeps NAT's address translation ability *and* provides inbound connection capabiltiies is really pretty cool.
However, because NAT has traditionally provided the secondary benefit of security to the interior network, any system that implements a way to connect to interior networks through NAT should provide at least three security models:
No interior access. Should be the default setting as it most closely matches the behavior expected from traditional NAT
Interior access to specific defined machines. Like current static NAT mappings.
Full interior access. Should require manual intervention to achieve this state.
If I'm buying bandwidth from the same provider I'd be buying colo from, I guess I'm getting all the connectivity they have to offer at the colo, just rate limited by my link to them.
I think the physical infrastructure of many colos (remember, we're talking lower end hosting options for high bandwidth sites, not "build an infrastructure for Amazon.com") is way overkill for lower end sites. They need decent ventilation, not industrial refrigeration. They need a solid UPS, not their own 200KW power plants. They need a good lock on the door, not 3' of steel reinforced concrete and biometric id.
10-15% is an acceptable premium for decent AC, good UPS power and reasonable security.
I have a 2+ year old (in tech terms) ATI Rage 128 based card (AIW-128) running under XP and with the newest ATI drivers and the games I've played with it (most recently the Medal of Honor demo), performance is just fine by my eyes @ 1024x768 and 16 bit color.
I've seen nVidia GeForce2 cards going for $100 but I just don't see the point. There was a time when moving from a 2D card to a 3D card like the orginal Voodoo was really worth the $300 or so it cost -- performance and quality skyrocketed. Similarly the move from the voodoo I to the II, and from the II to that card's next generation (the ATI 128).
Past that point, unless you have some specific non-gaming application that really needs the 3D performance it seems like kind of a waste. 3D performance has been pushed beyond the point where it matters, even for gaming and the features being added seem trivial -- just TV out?
All new cards it seem should come not only with good 3D, but video in and out, TV tuners, and the ability to do hardware MPEG2 compression of full-frame video at zero cost to the CPU. At that point the video card arms race would make more sense..
It seems that it's not so much the cost of bandwidth, but the charging for every {giga,mega}byte of bandwidth at colocations in addition to charging for rackspace.
For example, for about $2k per month I can get a MPPP dual T1 internet connection, 3Mbit/sec, and run that link for all the wire will deliver all month long. Even at 80% constant utilization that's 800Gbyte per month.
Most colo schemes seem to want to charge more than this -- at least 30 to 50% more, in fact.
What accounts for that? Presume a typical site has the disk/ram/cpu to deliver enough content to burn through 800Gbyte/month, is A/C, a router port and electricity worth the extra money? Is it the "strain" on hosting center uplinks that a popular site, connected at high bandwidth to the provider's network?
Assuming that high-bandwidth direct connections that colo customers have yields high demands on network uplinks, why not give customers a break and provide rate-limited network links for a substantial savings? Or if you're in the market for running a high bandwidth site, buy connectivity rather than colo space and run it for all its worth.
Depends on the apartment power system. Really old apartments (circa 1920 or so) like the one I used to live in appeared to have a single huge 115V backplane (each apartment had a fuse box, with two 15A fuses and a larger 30A fuse). I'm guessing that in reality there was some attempt to modernize a little, like a 200A main breaker and the load split over two legs of a 230V circuit, but overall kind of stone age. It might work for those apartments on the same leg of the 220.
Larger, modern places with modern electrical systems may, due to the power and code requirements, have a more sophisticated electrical system. High voltage feed to the building run on a riser bus with transformers on each floor running a 220V line to each apartments own panel, or something like that. I'm guessing that these places might not be so friendly to this.
I am not an electrician, but I seem to spend a lot of time in the electrical vault.
Whenever people talk about how much money is LOST to piracy, I always am left thinking that the money was never there in the first place- that those "pirates" would never have purchased the item anyway... so protection does more to piss off honest consumers than to increase revenue. How many ordinary people actually take the time or effort to mod a console (or overclock a PC) ?
I agree completely with your analysis -- the assertion that every pirated copy of a program is equal to a sale of that program is a total fallacy. The dollar figures quoted by people asserting that each copy pirated is a lost sale are even more ludicrous -- how many of those copies would have been sold at full retail?
I do think that there is some argument for losses due to piracy, though. I think there's a point on the disposable income curve where people would spend the money if they had to but if its simple enough (eg, Napster) to not do it, they won't.
I think the bigger problem for electronic entertainment is the plethora of outlets for disposable income. If piracy is easy, I might make other spending decisions (sports, bars, etc) and still get the games I might otherwise have paid for. It's much more difficult to sneak into stadiums, cheat bar tabs and so on. If piracy is just difficult enough, I might forgo those other options and buy the games outright. The problem with most corporate piracy analysis is that they assume that the pirated copies would always be sales.
Just got done watching a so-so documentary on the seige of Khe Sanh. Some of the footage was bomb bay camera shots of bomb strikes, and it was pretty awe-inspiring -- bomb after bomb with HUGE shockwaves visible from the air.
What practical difference is there between not knowing you own something and not owning it?
If I steal your car when you're on vacation for a month, is there a practical difference between not having a car and not knowing it was stolen?
My guess is that the major gaming companies (EA, Activision, etc) have huge IP portfolios from the early gaming days, including IP that was second and third generation to the companies they bought. They may not have accurate records, know they own it, or, more importantly, someone wisely has determined that the money necessary to chase after a few hundred people around the world using ROMs for some dead gaming system dramatically outweighs the financial return they'd ever get from that IP.
The rule makes sense, but how many gaming companies truly stopped business, settled their debts with cash or assets other than their intellectual property?
It seems to me that the most common scenario would involve a company going bust and their most valuable asset, their intellectual property, being sold off to debtors. Second most common would be the company getting bought out, including IP, by some other company.
The least likely event is that the company just gave up and stopped. Even in that case, there was somebody who can claim ownership of the company's assets.
It would seem to me that there would *always* be someone who owned the IP, although it might not be the original company. Awareness of ownership and desire to enforce copyright are probably in question, but ownership?
Slashdot Mirror
Directed by Howard Hawks and a William Faulkner screenplay adaption, it was a complicated story that was hard to translate to the screen -- claiming that Hawks and Faulkner didn't get it is a bit of a stretch. Raymond Chandler's books all seem to have enough twists and turns and strange motivational logic to drive any screenplay writer nutty just try to make them into understandable films. Practical realities like budgets, morality committees and so on and you wonder how some of them even made it to the screen.
I thought the movie was really true to the book, in fact too true for a movie. It wasn't he movie length but the fact that the story could and should have been streamlined or condensed somehow to make it a little less complex, but then the Tolkies would have had a snit and the movie would have gotten bad buzz.
As it stands now I thought the plotline was only slightly less complex than the Big Sleep -- too many characters, too many drastic scene transitions to make it flow smoothly.
All VP/CxO types are all about "strategy" (ie, going to meetings, drinking lots of Starbucks, and so on). "Tactics", or what OS to run in the data center, they could care less about unless it costs them money or gets their boss pissed off.
my parents who have no clue what the difference between a bit and a byte is can start rattling off MS Windows deficiencies, but it (will/does/is going to) take full fledge engineers to discover the same in the Unix systems.
Your parents probably have significant end-user experience with Microsoft systems (office, at home, etc) and therefor can list end-user deficiences. Your parents' lack of technical knowledge means they can't tell you about technical deficiencies, just things they don't like.
There are few large system sales that take place that don't involve a sales engineer, since the sales guy is often some geek in a suit that sold cars last week and will be selling copiers next week. The sales engineer is the one that gives you all the biased benchmarks and tells you it will fit your racks, etc.
I'd guess it means that they've hired some ex-Sun sales engineers and other unix sales people and paid them big piles of money to exaggerate the deficiencies of Sun and Linux systems relative to the strengths of MS systems.
It sounds like dumb direct-sequence is the problem, not frequency-hopping. Frequency hopping sounds like the *solution* to 2.4Ghz problems since the devices can hop around to find open channels or less interference.
It'd be nice if there had been an industry-wide bootup spec for looking to see if 2.4Ghz channels were in use and picking other channels if they were, or just make everything on 2.4ghz do FH SS.
Most smart managers want to fix a problem before it bites them. The fact that the name of the company ain't in the news has little to do with the amount of internal heat people are facing. You can bet your ass that the MSNBC called a lot of the company's management asking "Did you know how easy your gift cards are to rip off????" and the person in charge of the gift card program, who had probably touted its security previously, will be sitting in the boss' office on Jan 2 answering some hard questions.
At least that's how it'd work where I work.
Why not just assign a PIN number, stored in the store computer, not on the card, when the card is bought and charged?
Sure some yokels would write the number on the card and get it lifted or lose it, but the same could happen to cash.
Requiring extra information not available on the card would be ideal and would make the type of counterfeiting described in the article very difficult, as long as there was no simple way of resetting PINs. It wouldn't prevent inside jobs or people laundering stolen credit cards, but those types will always be hard to stop.
Which is a good thing, because at the Walmart in my area "Customer Service" more closely resembles the customs area of an east-African country than a place where you go to get helped.
Swell. So there's no significant economic reason for that company to change their policies yet.
Sure there is, its the internal economic justification of the manager in charge of the gift card program. The boss is likely to hear about this, and when (s)he does (s)he will either change the program or get canned.
No one wants an easy-to-rip-off gift card system. It invites attack from other fraud artists (if this system is lax, then others likely are too), pisses off customers and ruins loyalty.
The larger problem is that there's little financial incentive for stores to fix the problem generally (other than being seen as generally lax), since the losses aren't their own, they're someone else's, and even hijacked cards are money made for the store.
I've always liked Apple as a suitor to SGI.
Apple won't or can't compete in the commodity corporate desktop world and trying to expand that market would be a waste of time and money. The niche markets they do dominate, such as print production, are suffering to some extent from stagnation. The markets aren't growing bigger and with the general softness in the ad markets, I can tell you (as an ad industry employee) that budgets to replace B&W G3s with G4s wholesale aren't going to be there like they were 2-3 years ago when G3s rapidly replaced earlier PPC Macs -- there's little end user demand and ZERO management push.
Buying SGI would provide Apple with an entry into a world of higher-end computing than they currently have and would enable them to provide a much more vertically integrated solution to markets that are somewhat out of reach for them in terms of software and hardware -- high end film production, animation, and scientific visualization. From a technology perspective, it would give a credibility boost to Apple's nascent Unix and allow them to have hardware unified by a single OS.
It may be arguable that Apple's credibility in creative circles, early-to-market product offerings, and increasingly high performance machines will give it the bottom third of the video production market by default, and that SGIs technology is rapidly being obsoleteed by commodity hardware.
However, I don't think that there's nearly the growth prospect in desktop video that there was in desktop publishing or the huge edge over x86, either. And own its own, Apple still can't escape the low-end niche it sits in.
I thought that most of the cable and broadcast industry were "opposed" to HDTV rollout not because of the copy protection issues but because they wanted to slice 'n' dice the bandwidth for multiple channels.
As much an infrastructure problem for the CATV and a theres-only-so-many-free-MHZ problem for the broadcast people.
How do you "waste" bandwidth?
Since today's unused bandwidth cannot be used tomorrow, it stands to reason that any use of bandwidth up to whatever the congestion point is cannot be seen as waste. The only really wasted bandwidth is *unused* bandwidth.
You can make a value judgement about the quality of the information being transited but you cannot call that bandwidth wasted.
Some people buy "bandwidth", but what they're really buying is the right to transit X amount of bits per month for a fixed amount of money. But the same thing about congestion holds true for this kind of bandwidth -- as long as you don't go over your monthly cap, only used bits are wasted.
We've run into numerous service providers that use RFC1918 private address space. More than one has said, when setting up a routed connection, "just route 10.0.0.0/8 to us..." when we're already using 10.0.0.0 on some wan links as well as dealing with other providers who use 10.0.0.0. We've had to deal with providers that end up NATing connections two and three times because they and their vendors are all using the same private address space. The debug time on routing and access control glitches increases logarithmically for each NAT translation.
That address space works well for the WAN side of private links, or for testing or other stub networks that can't be connected elsewhere. But if you even think you might interconnect with other providers or other organizations, get real addresses.
The age of the film should be a weighting factor in its ranking. The fact that the Godfather was a 25 year old film should be meaningful.
I think the MS implementation is the problem, not the concept. Most people get a bee in their bonnet about this because they think it breaks the NAT "security" model.
Problem is, NAT provides security because it breaks routing, not because it is a security system by itself. That someone has come up with a routing/networking technique that keeps NAT's address translation ability *and* provides inbound connection capabiltiies is really pretty cool.
However, because NAT has traditionally provided the secondary benefit of security to the interior network, any system that implements a way to connect to interior networks through NAT should provide at least three security models:
Frankly, it can sometimes turn out some pretty puzzling results
And the PF commentary from native speakers isn't sometimes puzzling?
If I'm buying bandwidth from the same provider I'd be buying colo from, I guess I'm getting all the connectivity they have to offer at the colo, just rate limited by my link to them.
I think the physical infrastructure of many colos (remember, we're talking lower end hosting options for high bandwidth sites, not "build an infrastructure for Amazon.com") is way overkill for lower end sites. They need decent ventilation, not industrial refrigeration. They need a solid UPS, not their own 200KW power plants. They need a good lock on the door, not 3' of steel reinforced concrete and biometric id.
10-15% is an acceptable premium for decent AC, good UPS power and reasonable security.
I have a 2+ year old (in tech terms) ATI Rage 128 based card (AIW-128) running under XP and with the newest ATI drivers and the games I've played with it (most recently the Medal of Honor demo), performance is just fine by my eyes @ 1024x768 and 16 bit color.
I've seen nVidia GeForce2 cards going for $100 but I just don't see the point. There was a time when moving from a 2D card to a 3D card like the orginal Voodoo was really worth the $300 or so it cost -- performance and quality skyrocketed. Similarly the move from the voodoo I to the II, and from the II to that card's next generation (the ATI 128).
Past that point, unless you have some specific non-gaming application that really needs the 3D performance it seems like kind of a waste. 3D performance has been pushed beyond the point where it matters, even for gaming and the features being added seem trivial -- just TV out?
All new cards it seem should come not only with good 3D, but video in and out, TV tuners, and the ability to do hardware MPEG2 compression of full-frame video at zero cost to the CPU. At that point the video card arms race would make more sense..
It seems that it's not so much the cost of bandwidth, but the charging for every {giga,mega}byte of bandwidth at colocations in addition to charging for rackspace.
For example, for about $2k per month I can get a MPPP dual T1 internet connection, 3Mbit/sec, and run that link for all the wire will deliver all month long. Even at 80% constant utilization that's 800Gbyte per month.
Most colo schemes seem to want to charge more than this -- at least 30 to 50% more, in fact.
What accounts for that? Presume a typical site has the disk/ram/cpu to deliver enough content to burn through 800Gbyte/month, is A/C, a router port and electricity worth the extra money? Is it the "strain" on hosting center uplinks that a popular site, connected at high bandwidth to the provider's network?
Assuming that high-bandwidth direct connections that colo customers have yields high demands on network uplinks, why not give customers a break and provide rate-limited network links for a substantial savings? Or if you're in the market for running a high bandwidth site, buy connectivity rather than colo space and run it for all its worth.
Depends on the apartment power system. Really old apartments (circa 1920 or so) like the one I used to live in appeared to have a single huge 115V backplane (each apartment had a fuse box, with two 15A fuses and a larger 30A fuse). I'm guessing that in reality there was some attempt to modernize a little, like a 200A main breaker and the load split over two legs of a 230V circuit, but overall kind of stone age. It might work for those apartments on the same leg of the 220.
Larger, modern places with modern electrical systems may, due to the power and code requirements, have a more sophisticated electrical system. High voltage feed to the building run on a riser bus with transformers on each floor running a 220V line to each apartments own panel, or something like that. I'm guessing that these places might not be so friendly to this.
I am not an electrician, but I seem to spend a lot of time in the electrical vault.
Whenever people talk about how much money is LOST to piracy, I always am left thinking that the money was never there in the first place- that those "pirates" would never have purchased the item anyway... so protection does more to piss off honest consumers than to increase revenue. How many ordinary people actually take the time or effort to mod a console (or overclock a PC) ?
I agree completely with your analysis -- the assertion that every pirated copy of a program is equal to a sale of that program is a total fallacy. The dollar figures quoted by people asserting that each copy pirated is a lost sale are even more ludicrous -- how many of those copies would have been sold at full retail?
I do think that there is some argument for losses due to piracy, though. I think there's a point on the disposable income curve where people would spend the money if they had to but if its simple enough (eg, Napster) to not do it, they won't.
I think the bigger problem for electronic entertainment is the plethora of outlets for disposable income. If piracy is easy, I might make other spending decisions (sports, bars, etc) and still get the games I might otherwise have paid for. It's much more difficult to sneak into stadiums, cheat bar tabs and so on. If piracy is just difficult enough, I might forgo those other options and buy the games outright. The problem with most corporate piracy analysis is that they assume that the pirated copies would always be sales.
Just got done watching a so-so documentary on the seige of Khe Sanh. Some of the footage was bomb bay camera shots of bomb strikes, and it was pretty awe-inspiring -- bomb after bomb with HUGE shockwaves visible from the air.
What practical difference is there between not knowing you own something and not owning it?
If I steal your car when you're on vacation for a month, is there a practical difference between not having a car and not knowing it was stolen?
My guess is that the major gaming companies (EA, Activision, etc) have huge IP portfolios from the early gaming days, including IP that was second and third generation to the companies they bought. They may not have accurate records, know they own it, or, more importantly, someone wisely has determined that the money necessary to chase after a few hundred people around the world using ROMs for some dead gaming system dramatically outweighs the financial return they'd ever get from that IP.
The rule makes sense, but how many gaming companies truly stopped business, settled their debts with cash or assets other than their intellectual property?
It seems to me that the most common scenario would involve a company going bust and their most valuable asset, their intellectual property, being sold off to debtors. Second most common would be the company getting bought out, including IP, by some other company.
The least likely event is that the company just gave up and stopped. Even in that case, there was somebody who can claim ownership of the company's assets.
It would seem to me that there would *always* be someone who owned the IP, although it might not be the original company. Awareness of ownership and desire to enforce copyright are probably in question, but ownership?