I would like to thank you all for your comments. I have found the majority of your posts entertaining. If I had enough moderator points I'd give most of you a pat on the back. I would like to thank those of you who actually tried to answer my question and I'll be off now to try and implement some of these.
Those of you who actually read the discussion threads in my Journal might see how even amongst the intelligent and clueful people who post in my Journals that there is confusion on the "proper" way to handle a USB drive. Should it be automount? Should it be in the driver itself? Why bother with a script at all? How do you deal with the fact that the pen-drive can "move around" from sda1 to sdb1 if you have two pen-drives?
In writing this "user space driver"
I had to learn that usb-pen drives were usb masstorage devices. I had to learn about all the technologies involved, then I had to learn the "right way" to implement a "driver" that did what I wanted. I suspect that many people don't want to learn all those things. They shouldn't have to.
In closing, I admit I was a little sensational with my article... USB storage devices do work with Redhat 9 Linux but the default Gnome desktop and KDE don't offer any clues that the drive is working. I see this as a flaw in the distro.
It's a pitty this made it to Slashdot on the week I was on vacation.
I give you the gold medal. You actually tried to answer my question. Kudos to you!
The problem is that this is a "Redhat 9" problem. If you look at the Red Hat 9 standard "/etc/hotplug/usb" directory you'll find one script in there. It's the "usbcam" script written by... A programmer at Redhat. So I assume that the "correct" place for "/etc/hotplug/usb" scripts to come from are from distro people.
I will contact the hotplug people but I have a feeling that they will respond that this isn't "part of" hotplug and that it's a user add on. Further... I have to wonder why Redhat didn't commission anyone to write this script to begin with... it wasn't hard.
The only quirk was a "device busy"-bug that occured sometimes so the filesystem could not be cleanly unmounted. This was because some subsystems of Nautilus or any other filemananger component did not release the filesystem lock properly when closing its windows and the force-unmount had no effect.
My "driver" fixes this problem... to a certain extent. The boss didn't like having to drop to the command line to mount and unmount the device. So I was commissioned to write this script. The other problems I anticipated that would later cause headaches were the "Device busy" bug. You do an "umount -l" to fix this. "-l" is the "Lazy" unmount. I incorporated this command into my script at two points to doubly make sure that the users didn't screw their naughty nautilus sessions.
Red Hat 9, actually does work out of the box. It's just that to find what drive you should issue a "mount" command on you need to look in/var/log/messages, or at your scsi bus to figure out not only that a usb device has been hooked to the system but that it is running as... most commonly... sda1. Remember you aren't guaranteed sda1, if you have two pen drives one will be sda1 and another will be sdb1... or something.
You can add a line in fstab, or you can use my script to do it for you... yes, I know it's not a "real driver"...
I think you've completely missed the point. The point is that to the uninitiated user, they plug the drive in and nothing happens. The GUI doesn't automatically respond to the presence of the new drive and so... it doesn't "automagically" work. This proves that users are indeed stupid. For if they were smart they would merely read/var/log/message themselves and know which scsi emulated device was added and then mount it themselves either with su, sudo, or by hand editing their fstab.
My solution is not worthy of a sourceforge project, a kernel posting, or anything so grand. It's at best a small hack that fixes something in userspace. Redhat should know about this short coming... but how do I point it out to them? How do I send this to the right people? The right people being redhat 9 users with a clue.
I think it resembles one of those racing boats, the ones with jets on the back. Sort of like this: http://www.f1boat.com/02/portimao/images/ambiance/ flying_zepter.jpg
I suppose it make sense... so the plane can surf on the super-sonic shockwaves there by reducing their sound. I suppose simmilar wave physics are going on at super sonic speeds as they are at super-boat speeds.
A Naval Architect once told me that the speed of a boat is limited not by its motor so much as by it's hull length. This had something to do with the wave length that could be built up as the boat shoved water out of it's way. The way you got by this issue was by hydro-plaining... or hydro-plane-ing? So I suspect that this plane... hydro-planes on the super sonic shockwaves... I'll read the article now I guess.
Out of all the suggestions on this page, I like this(See the bottom of the page)
idea the best. The reason I like SchauzerGuy's solution best, is because it's actually TV. That means that All those school TV's can just tune to the right channel... no 802.11b muss and fuss. That means the solution can leverage the existing infrastructure of the school giving it larger effective coverage.
I know it's not free as in beer but it's definately the best solution for the ultimate application. Thirty-five bucks isn't bad though.
And think, nuggetman, you do this and you'll be the demi-god who brought your school it's own TV station. Woof! Later projects could incorporate sending that signal to a website or what not... encoding the school TV programs as movies that could be accessed via a web page. A webpage on the school's internal network maybe.
I like it because it's simple. It's simple because it's just TV.
just because I like linux so much I'll pretend that you have linux...
Check out:
http://www.videolan.org/
The player works on lots of platforms including windows but the server has to be linux. Now that's fine with me. I'm wondering about Digital Video and digital editing too...
You may also look at these links:
http://cs.uhh.hawaii.edu/~jeschke/links/video.html
This fella has made a pretty good collection of links that he's researched. Hey, I'm interested in this idea now... I think I may have an idea for a project.
This is a stupid question: where on earth did that phrase come from? Beer isn't free, is it? Have I been ripped off all along?
I was at a German block party a couple of weeks ago for "ascention sunday" or something. The beer was free as in beer there. It cost nothing, the semmel were three euro each, the steaks were three euro, but the beer was null euro... So I guess in Germany the Beer really is Free. But, you had to bring your own mug.
The new linksys does according to O'Reilly Net. You don't need much tho'... and I've not had trouble finding products to help with Broadband network bridging. All manner of PPPoE, RAS, and other kinds of Cable Modem, DSL Modem, and such tools are easy enough to find.
My problem was I was on ISDN or 56K and I wanted to share my internet connection.
This Actiontec
device would have been really handy to have had. As it was I ended up putting a modem and ISDN card into a Linux box and setting it up as a proxy server with Dial On Demand set to the ISDN provider.
I moved cities to get broadband though and I'm much happier now that I don't have to do such contortions anymore.
Heh. Maybe this means that the new Earth Paradigm will feature a lack of Ozone.
After reading your link it looks like the Ozone depleating reactions are believed to occur on the surface of polar stratospheric clouds. Even if the Shuttle only produces stratospheric water vapor it is evectively increasing the surface area on which the Ozone depleating chemical reactions can occur.
So I was about to foolishly point out that the shuttle clouds are only water vapor. But, that's enough to potentially contribute to Ozone loss even if there's no Chlorine and Bromine pollution. Amazing, something as harmless as water can result in toxic consequences when it's in the wrong place. That's astounding, you'd expect water vapor to be utterly harmless in all situations...
Here's an interesting problem. The spammers spam because in a billion e-mails one or two people give a positive response... that's enough to justify trillions of e-mails. The blocking of a paltry billion or so e-mails is enough to justify trillions of packets in a DDoS attack.
So one positive response in a trillion negative or null responses is enough to justify the wrath of a trillion DDoS packets on hundreds of servers.
How do you remove the desire to perform such a futile and repugnant act as spamming when such small positive reinforcement causes such profound tenacity for spewing forth volumes of wretched bile on innocent masses merely seeking to expunge said bile?
The crux of the problem is that communication is now so cheap that it is easy to waste. If communication were more valuable people would spend more time on it. If you had to pay for each e-mail in some way you wouldn't recklessly waste them.
I propose we start charging the sender for each e-mail... but not money. I propose we charge each sender for each e-mail they send a percentage of their soul.
Each e-mail you send would slowly drain away your soul until you were nothing but a dried out husk. Sort of like Slashdot.
and one year there was a power outage on campus over christmas break. The called out the national guard. No kidding. They had to kick in back up generators and such to keep stuff from freezing up because it was sixty below.
Fortunately the city of Fairbanks was still with power and in a few hours they re-routed power to the University. This wasn't just good for keeping people from freezing in the dorms (the poor sots like me who didn't go home for the holidays) but it was very good for the Cray which was being threatened with imminent condesation...
If the whole city of Fairbanks lost power I wonder how people would have coped. I suppose families would have moved in together. Houses with wood stoves would have been very crowded... I wonder if there's even enough housing with other means of heating to support the whole town?
Why is it we attribute so much more ability to evolve to robots just because they will have a humanoid form?
Because, we are humanoid. Humans tend to anthropomorphize objects and attribute human qualities to them. The more human looking the object (a doll for example) the stronger the urge to anthropomorphize. The upshot for humans is that if it looks human and acts human... it must be human.
Normal humans probably posses this trait to help with empathy. The ability to empathize with others makes it possible to form collective groups. The collective groups are stronger, more intelligent, and better able to survive than a lone individual.
The following assumptions are probably not too far fetched once mass production of humanoid robots begin:
Manufacturers will find clever ways for their robots to emulate empathy and other emotional responses to make them "easy to use": This is seen today in children's toys like Furby and Tickle Me Elmo.
Users will project personalities onto their humanoid machines: This is seen today with just about every complex machine uneducated end users use. How many people do you know that attribute personalities to old cars? How about their PC? Most of the time these are the "empathic thinkers" of society. They aren't empirically minded scientist types... they are the "down home" types. (I've often noted the uncanny ability for these people to intuit complex social interactions and navigate them easily without much thought. I've attributed this ability to another form of intelligence which I consider quite alien to myself.)
Manufacturers will compete for marketshare. This competition will yeild more and more complex humanoid robotic forms: This is seen today as Honda, Sony, and other companies still in R&D battle it out for more and more complex robots. The robots are "evolving" but it is a human directed evolution.
With these three factors in place over the course of hundreds of years, manufacturers will eventually begin to produce sentient-a-like robots. Robotic entities that emulate necessary human-like traits to do their jobs... jobs which will require learning and adaptability will eventually be included.
If it acts like a person, looks like a person, talks like a person... how do you know it's not a person? What if one day someone (like a linux-style geek) makes a robot that even pretends to object to the idea it doesn't have a free will. Maybe they even give it the ability to respond with murderous anger at the thought it isn't fully alive.
None of these ideas are new and most of them are at least as old as automation itself. Personally, I believe that it will take two hundred years to get to the point that we have to actually worry about machines overthrowing humanity. I only believe that the problems in creating something that acts sentient (but really isn't?) are enourmous... I may be wrong.
So, as a Philadelphia area resident can anyone get me a list of infected business/departments so I can fill the positions of the soon-to-be-fired IT Staff?
The note I just got said those jobs are being outsourced to India. Sorry you're still out of luck.:(
you absolutely need a paper-clip and a bic pen. You can't work on computers without a paper-clip and a bic pen!
You need the paper-clip to eject jammed disk drives and you need the bic pen to squeeze the plastic fish-hook-like motherboard fasteners together so you can pop them backwards off the motherboard.
I suppose you don't need the bic pen if you've got one of those snazzy new cases that use screws for everything.
Boy you kids have it easy today. When I was a kid we had to edit the inodes by hand on a giant kick drum. We had two large dumb-bell shaped magnets that we used, one for zeroes and the other for ones. So writing to the disk was like beating out a rythm... one *whang*, zero *whang*, one *whang*! If you wanted to erase the drive you just held one magnet down while kick-spinning the drive.
You'd get enourmous legs and fore-arms from this activity. Kinda like Popeye.
Instead of asking users to lock up their CD collections, which is unlikely to convince any court, why not charge a subscription and spend the money on multiple copies of every CD, to be stored in a central location? Each user will only have to pay a fraction of what they would have spent on CDs in the same period, because every CD will be used by somebody 100% of the time. The library will choose which discs to buy based on demand, and each user's subscription will be based on the amount of music they listen to.
When you put it like that it makes you wonder why noone has tried this yet. The idea seems so simple. This sounds like a membership based library. You just check out a song and return it. Why just songs? Why not video, books, and magazines too? Hell... this is just a subscription based electronic library system.
The lynch-pin is the whole "checking out" system. At a physical library you physically check out a book... and you physically return it. If the fella who checks the book out uses a photocopy machine to make a thousand copies of the book and sells those copies or gives them away then the library isn't liable for copyright violation is it?
So you have to legally prove that this electronic library is equally...legally... innocent as a physical one. That's because you can't guarantee that someone won't take measures to get around your copyprotection schemes and any system that allows a machine to view material allows a machine to copy material... be it a book or an MP3.
So, why do libraries get away with having music collections and book collections but don't get sued all the time for copyright violations? Sure they have rules to keep IP from getting stolen but I can smuggle a recorder or scanner into a library... so why isn't that a problem for the library but it is for MP3.com?
The problem is legal not technological. That's what the most valuable lesson in Cringely's article is. This system is so economically and practically simple it's hard to believe noone's ever concieved of it before. I suspect it's more a matter of noone ever having taken the legal manuvers before instead...
FYI, the article seems to be something like a company "basic C coding standards" rather than anything to do with what I understand as "best practices"; maybe a coding tips sheet for an introductory C programming class.
From what I've seen in industry there are plenty of PHB's that need to read things like this. I've had PHB's argue that "Everything should be a global variable" because it "makes everything easier." Real businesses need to hear things like this... if it comes from IBM someone just might listen.
Slashdot Mirror
I would like to thank you all for your comments. I have found the majority of your posts entertaining. If I had enough moderator points I'd give most of you a pat on the back. I would like to thank those of you who actually tried to answer my question and I'll be off now to try and implement some of these.
Those of you who actually read the discussion threads in my Journal might see how even amongst the intelligent and clueful people who post in my Journals that there is confusion on the "proper" way to handle a USB drive. Should it be automount? Should it be in the driver itself? Why bother with a script at all? How do you deal with the fact that the pen-drive can "move around" from sda1 to sdb1 if you have two pen-drives?
In writing this "user space driver" I had to learn that usb-pen drives were usb masstorage devices. I had to learn about all the technologies involved, then I had to learn the "right way" to implement a "driver" that did what I wanted. I suspect that many people don't want to learn all those things. They shouldn't have to.
In closing, I admit I was a little sensational with my article... USB storage devices do work with Redhat 9 Linux but the default Gnome desktop and KDE don't offer any clues that the drive is working. I see this as a flaw in the distro.
It's a pitty this made it to Slashdot on the week I was on vacation.
Okay buster, so I should start a Sourceforge project for one lousy PERL script? I don't think so.
I give you the gold medal. You actually tried to answer my question. Kudos to you!
The problem is that this is a "Redhat 9" problem. If you look at the Red Hat 9 standard "/etc/hotplug/usb" directory you'll find one script in there. It's the "usbcam" script written by... A programmer at Redhat. So I assume that the "correct" place for "/etc/hotplug/usb" scripts to come from are from distro people.
I will contact the hotplug people but I have a feeling that they will respond that this isn't "part of" hotplug and that it's a user add on. Further... I have to wonder why Redhat didn't commission anyone to write this script to begin with... it wasn't hard.
The only quirk was a "device busy"-bug that occured sometimes so the filesystem could not be cleanly unmounted. This was because some subsystems of Nautilus or any other filemananger component did not release the filesystem lock properly when closing its windows and the force-unmount had no effect.
My "driver" fixes this problem... to a certain extent. The boss didn't like having to drop to the command line to mount and unmount the device. So I was commissioned to write this script. The other problems I anticipated that would later cause headaches were the "Device busy" bug. You do an "umount -l" to fix this. "-l" is the "Lazy" unmount. I incorporated this command into my script at two points to doubly make sure that the users didn't screw their naughty nautilus sessions.
Red Hat 9, actually does work out of the box. It's just that to find what drive you should issue a "mount" command on you need to look in /var/log/messages, or at your scsi bus to figure out not only that a usb device has been hooked to the system but that it is running as ... most commonly ... sda1. Remember you aren't guaranteed sda1, if you have two pen drives one will be sda1 and another will be sdb1 ... or something.
...
You can add a line in fstab, or you can use my script to do it for you... yes, I know it's not a "real driver"
I think you've completely missed the point. The point is that to the uninitiated user, they plug the drive in and nothing happens. The GUI doesn't automatically respond to the presence of the new drive and so ... it doesn't "automagically" work. This proves that users are indeed stupid. For if they were smart they would merely read /var/log/message themselves and know which scsi emulated device was added and then mount it themselves either with su, sudo, or by hand editing their fstab.
My solution is not worthy of a sourceforge project, a kernel posting, or anything so grand. It's at best a small hack that fixes something in userspace. Redhat should know about this short coming... but how do I point it out to them? How do I send this to the right people? The right people being redhat 9 users with a clue.
It's a user-space driver or a hotplug script.
My boss didn't like that he had to click the icon.
It will not always be sda1, different systems running the same distro may make the drive a different scsi device. Don't you people read the articles?
Somebody better let out a bunch of CueCats to catch all those mice!!!
I think it resembles one of those racing boats, the ones with jets on the back. Sort of like this: http://www.f1boat.com/02/portimao/images/ambiance/ flying_zepter.jpg
I suppose it make sense... so the plane can surf on the super-sonic shockwaves there by reducing their sound. I suppose simmilar wave physics are going on at super sonic speeds as they are at super-boat speeds.
A Naval Architect once told me that the speed of a boat is limited not by its motor so much as by it's hull length. This had something to do with the wave length that could be built up as the boat shoved water out of it's way. The way you got by this issue was by hydro-plaining... or hydro-plane-ing? So I suspect that this plane... hydro-planes on the super sonic shockwaves... I'll read the article now I guess.
The beer was free only at this particular block party. It was just ironic that the beer was indeed free as in beer.
Out of all the suggestions on this page, I like this(See the bottom of the page) idea the best. The reason I like SchauzerGuy's solution best, is because it's actually TV. That means that All those school TV's can just tune to the right channel... no 802.11b muss and fuss. That means the solution can leverage the existing infrastructure of the school giving it larger effective coverage.
I know it's not free as in beer but it's definately the best solution for the ultimate application. Thirty-five bucks isn't bad though.
And think, nuggetman, you do this and you'll be the demi-god who brought your school it's own TV station. Woof! Later projects could incorporate sending that signal to a website or what not... encoding the school TV programs as movies that could be accessed via a web page. A webpage on the school's internal network maybe.
I like it because it's simple. It's simple because it's just TV.
just because I like linux so much I'll pretend that you have linux...
l
Check out: http://www.videolan.org/
The player works on lots of platforms including windows but the server has to be linux. Now that's fine with me. I'm wondering about Digital Video and digital editing too...
You may also look at these links:
http://cs.uhh.hawaii.edu/~jeschke/links/video.htm
This fella has made a pretty good collection of links that he's researched. Hey, I'm interested in this idea now... I think I may have an idea for a project.
Are there any other free (as beer) solutions?
This is a stupid question: where on earth did that phrase come from? Beer isn't free, is it? Have I been ripped off all along?
I was at a German block party a couple of weeks ago for "ascention sunday" or something. The beer was free as in beer there. It cost nothing, the semmel were three euro each, the steaks were three euro, but the beer was null euro... So I guess in Germany the Beer really is Free. But, you had to bring your own mug.
The new linksys does according to O'Reilly Net. You don't need much tho' ... and I've not had trouble finding products to help with Broadband network bridging. All manner of PPPoE, RAS, and other kinds of Cable Modem, DSL Modem, and such tools are easy enough to find.
My problem was I was on ISDN or 56K and I wanted to share my internet connection.
This Actiontec device would have been really handy to have had. As it was I ended up putting a modem and ISDN card into a Linux box and setting it up as a proxy server with Dial On Demand set to the ISDN provider.
I moved cities to get broadband though and I'm much happier now that I don't have to do such contortions anymore.
Heh. Maybe this means that the new Earth Paradigm will feature a lack of Ozone.
After reading your link it looks like the Ozone depleating reactions are believed to occur on the surface of polar stratospheric clouds. Even if the Shuttle only produces stratospheric water vapor it is evectively increasing the surface area on which the Ozone depleating chemical reactions can occur.
So I was about to foolishly point out that the shuttle clouds are only water vapor. But, that's enough to potentially contribute to Ozone loss even if there's no Chlorine and Bromine pollution. Amazing, something as harmless as water can result in toxic consequences when it's in the wrong place. That's astounding, you'd expect water vapor to be utterly harmless in all situations...
Here's an interesting problem. The spammers spam because in a billion e-mails one or two people give a positive response... that's enough to justify trillions of e-mails. The blocking of a paltry billion or so e-mails is enough to justify trillions of packets in a DDoS attack.
So one positive response in a trillion negative or null responses is enough to justify the wrath of a trillion DDoS packets on hundreds of servers.
How do you remove the desire to perform such a futile and repugnant act as spamming when such small positive reinforcement causes such profound tenacity for spewing forth volumes of wretched bile on innocent masses merely seeking to expunge said bile?
The crux of the problem is that communication is now so cheap that it is easy to waste. If communication were more valuable people would spend more time on it. If you had to pay for each e-mail in some way you wouldn't recklessly waste them.
I propose we start charging the sender for each e-mail... but not money. I propose we charge each sender for each e-mail they send a percentage of their soul.
Each e-mail you send would slowly drain away your soul until you were nothing but a dried out husk. Sort of like Slashdot.
and one year there was a power outage on campus over christmas break. The called out the national guard. No kidding. They had to kick in back up generators and such to keep stuff from freezing up because it was sixty below.
Fortunately the city of Fairbanks was still with power and in a few hours they re-routed power to the University. This wasn't just good for keeping people from freezing in the dorms (the poor sots like me who didn't go home for the holidays) but it was very good for the Cray which was being threatened with imminent condesation...
If the whole city of Fairbanks lost power I wonder how people would have coped. I suppose families would have moved in together. Houses with wood stoves would have been very crowded... I wonder if there's even enough housing with other means of heating to support the whole town?
Because, we are humanoid. Humans tend to anthropomorphize objects and attribute human qualities to them. The more human looking the object (a doll for example) the stronger the urge to anthropomorphize. The upshot for humans is that if it looks human and acts human... it must be human.
Normal humans probably posses this trait to help with empathy. The ability to empathize with others makes it possible to form collective groups. The collective groups are stronger, more intelligent, and better able to survive than a lone individual.
The following assumptions are probably not too far fetched once mass production of humanoid robots begin:
-
Manufacturers will find clever ways for their robots to emulate empathy and other emotional responses to make them "easy to use":
-
Users will project personalities onto their humanoid machines:
-
Manufacturers will compete for marketshare. This competition will yeild more and more complex humanoid robotic forms:
With these three factors in place over the course of hundreds of years, manufacturers will eventually begin to produce sentient-a-like robots. Robotic entities that emulate necessary human-like traits to do their jobs... jobs which will require learning and adaptability will eventually be included.This is seen today in children's toys like Furby and Tickle Me Elmo.
This is seen today with just about every complex machine uneducated end users use. How many people do you know that attribute personalities to old cars? How about their PC? Most of the time these are the "empathic thinkers" of society. They aren't empirically minded scientist types... they are the "down home" types. (I've often noted the uncanny ability for these people to intuit complex social interactions and navigate them easily without much thought. I've attributed this ability to another form of intelligence which I consider quite alien to myself.)
This is seen today as Honda, Sony, and other companies still in R&D battle it out for more and more complex robots. The robots are "evolving" but it is a human directed evolution.
If it acts like a person, looks like a person, talks like a person... how do you know it's not a person? What if one day someone (like a linux-style geek) makes a robot that even pretends to object to the idea it doesn't have a free will. Maybe they even give it the ability to respond with murderous anger at the thought it isn't fully alive.
None of these ideas are new and most of them are at least as old as automation itself. Personally, I believe that it will take two hundred years to get to the point that we have to actually worry about machines overthrowing humanity. I only believe that the problems in creating something that acts sentient (but really isn't?) are enourmous... I may be wrong.
Clearly the Founding Father's intended to protect the freedom of speech but not speech such as crying "Fire" in a crowded theater.
Sarcasm: Obviously, crying "Stinker" or "Dud!" is just as bad since both actions cause people to stampede from the theaters.
So, as a Philadelphia area resident can anyone get me a list of infected business/departments so I can fill the positions of the soon-to-be-fired IT Staff?
:(
The note I just got said those jobs are being outsourced to India. Sorry you're still out of luck.
you absolutely need a paper-clip and a bic pen. You can't work on computers without a paper-clip and a bic pen!
You need the paper-clip to eject jammed disk drives and you need the bic pen to squeeze the plastic fish-hook-like motherboard fasteners together so you can pop them backwards off the motherboard.
I suppose you don't need the bic pen if you've got one of those snazzy new cases that use screws for everything.
Boy you kids have it easy today. When I was a kid we had to edit the inodes by hand on a giant kick drum. We had two large dumb-bell shaped magnets that we used, one for zeroes and the other for ones. So writing to the disk was like beating out a rythm... one *whang*, zero *whang*, one *whang*! If you wanted to erase the drive you just held one magnet down while kick-spinning the drive.
You'd get enourmous legs and fore-arms from this activity. Kinda like Popeye.
Instead of asking users to lock up their CD collections, which is unlikely to convince any court, why not charge a subscription and spend the money on multiple copies of every CD, to be stored in a central location? Each user will only have to pay a fraction of what they would have spent on CDs in the same period, because every CD will be used by somebody 100% of the time. The library will choose which discs to buy based on demand, and each user's subscription will be based on the amount of music they listen to.
...legally... innocent as a physical one. That's because you can't guarantee that someone won't take measures to get around your copyprotection schemes and any system that allows a machine to view material allows a machine to copy material... be it a book or an MP3.
When you put it like that it makes you wonder why noone has tried this yet. The idea seems so simple. This sounds like a membership based library. You just check out a song and return it. Why just songs? Why not video, books, and magazines too? Hell... this is just a subscription based electronic library system.
The lynch-pin is the whole "checking out" system. At a physical library you physically check out a book... and you physically return it. If the fella who checks the book out uses a photocopy machine to make a thousand copies of the book and sells those copies or gives them away then the library isn't liable for copyright violation is it?
So you have to legally prove that this electronic library is equally
So, why do libraries get away with having music collections and book collections but don't get sued all the time for copyright violations? Sure they have rules to keep IP from getting stolen but I can smuggle a recorder or scanner into a library... so why isn't that a problem for the library but it is for MP3.com?
The problem is legal not technological. That's what the most valuable lesson in Cringely's article is. This system is so economically and practically simple it's hard to believe noone's ever concieved of it before. I suspect it's more a matter of noone ever having taken the legal manuvers before instead...
FYI, the article seems to be something like a company "basic C coding standards" rather than anything to do with what I understand as "best practices"; maybe a coding tips sheet for an introductory C programming class.
From what I've seen in industry there are plenty of PHB's that need to read things like this. I've had PHB's argue that "Everything should be a global variable" because it "makes everything easier." Real businesses need to hear things like this... if it comes from IBM someone just might listen.