I think you may have missed my point. When in London or Paris, an eight mile commute means jumping on the Tube/Metro for a quick trip then a walk from the station. Unless you are in an area protected from urban sprawl or that has had a substantial amount of federal highway funds dumped into mass transit (not common in the States' political climate, it normally goes into widening roads or building bypasses), then mass transit is not a good option
As an aside, it's unreasonable to say that people (most of whom are neither young nor fit) should commute under their own power for distances of more than a a mile or so. Take a look at American demographics. Saying "man up" shows a lack of consideration for the larger problem.
In many areas there is good public transportation that only takes slightly longer to get where you are going.
If we are discussing the US, then this is a misleading statement. A more accurate one would be "In certain urban locations and a few suburban locales where geography has prevented urban sprawl, there is mediocre to barely acceptable public transportation." The best the States have to offer (San Francisco or Washington D.C.) compares poorly with, say, London or Paris. (I've used all four of these and won't speak to those I haven't.) I'm not willing to convert my 8-mile/13-minute commute into an 75 minute bus ordeal because it is "environmentally friendly."
While I don't disagree with most of the conclusions in the article, the financial reasoning is wrong. The purchase cost of a computer is the smallest part of the lifetime costs. If you take his theoretical 100 PC environment, you can expect to need a full-time support position (industry averages are between 100 and 200 systems per support tech). If you pay him/her peanuts ($15/hour) then your cost just went up by $30k/year for the (let's say) four year life expectancy of the equipment. That makes the numbers $177k vs $234k; the percentage savings just dropped through the floor.
It gets worse if you look at the real cost of support people. Windows support techincians are easy to find. Linux support technicians are not (I say this as someone who has tried to hire dozens of support staff at levels from Mac-Hand-Holder-3rd-Class to Unix-Minor-Deity). If you factor in the cost differences of the two skill sets, the savings gets even slimmer.
Finally, the analysis ignores the reality that people like to use the software with which they are familiar. I've run experiments with family and while I (mostly) like OpenOffice, it makes some people crazy because it is different. Non-geek users don't want to learn new software, they want to get their target task completed then go do something else. If you add the additional training/hand-holding to the mix you can pretty much kiss the rest of the theoretical cost savings goodbye.
This is why Windows still dominates the desktop. When you look at the total cost model, there isn't a lot of savings to be had while there are plenty of potential headaches.
To my mind the biggest risk MSFT has taken in the last decade is the set of changes to the UI in Office 2007. Since 1994, there has been essentially no learning curve to an Office upgrade and now there is one. Suddenly, the non-MSFT office suite is the one with better legacy skill support.
Why? Because he's a convict still serving his sentence (that's why he's under home confinement). If he doesn't like the terms of home confinement, he can always go back in the slammer and have even more restricted access.
If you come in with a Treo, no one is going to say anything.
Not necessarily true. We had tickets to some random movie preview 10 months ago. As always, I had my phone with me as required for work (if the servers go down...). The goon at the door saw that my phone had camera capability and denied me entrance. After an extended, polite, kafkaesque conversation, my wife watched the movie while I drank coffee and read a book next door.
Good point - you could probably apply this thinking to suburbia as well. If the real problem is the future rise of petroleum costs, people in the 'burbs will expect to be able to drive shorter distances to get things done, while maintaining the same standard of living. That alone dictates more localized food production, storage and sale than what we have now.
I agree. My point is that this won't make high-density urban areas self-sustainable. Attempting to build them in the city cores would still require a reduction in population density. Assuming that transportation costs don't stay stable and that people won't be inclined to move, then it makes more sense to build these is suburban areas near the core (i.e. across the river in New Jersey or on Long Island instead of Manhattan). It still dramatically reduces transport costs but doesn't suffer from the insane land prices in the city cores.
It's also worth noting that world history since the Roman Republic has in no small part been shaped by cities trying to feed themselves. When transportation costs or reliability have driven prices up people have starved and/or wars have started.
While vertical factory farms would be pretty nifty, how useful would they really be?
Let's consider a medium-sized produce farm parcel as 640 acres (one square mile) and should be able to feed around 1000 people (order of magnitude estimate). To get that amount of space, Manhattan would need to sacrifice a square block and build to 100 stories. This seems to fail the cost-benefit test unless they can produce orders of magnitude more food/acre than traditional farming.
How many do you need? Let's look at Manhattan (population 1.5M). Assuming that a vertical farm is 10 times as productive as a traditional farm, you need around 150 of them to support the island. This would be around 1.5 square miles of building footprint (about 15% of the island).
Livestock produce less food/acre than plants unless you truck in all of the food. Using it for animal "crops" (as the article suggested) seems to be a worse idea than just growing fresh green stuff and trucking the meat in.
"All of the water in the entire complex would be recycled" isn't possible. A bunch of that water is being turned into food and shipped outside the farm. Remeber the Lunar Revolution? Presumably, they are expecting to process the urban sewage and use that for water and fertilizer.
No diseases or parasites? BioSphere2 didn't manage to stay that isolated. Besides, where do they expect to get the seeds/grafts, fertilizers, water, and staff? Do they expect the entire facility to operate as a clean-room?
Ultimately, if you want to reduce transport costs (money, fuel, etc.) the people need to be closer to the food production. This seems like an idea better suited to lower-density, urban sprawl (where you can grab relatively large areas without consuming a large percentage of the available space) rather than in the middle of compact urban areas.
Like I said, there may be legitimate academic uses for p2p. However, p2p is just a data transfer technique and there are many to choose from. For example, a course may need to distribute linux ISO's, but that does not imply that they need bittorrent; an FTP server on the campus network may be just as (or perhaps more) efficient.
Personally, I keep hoping someone will ask me for a gopher server to support one of their classes.
The policing side of the DMCA is a mess. However, if I read the University's press release, they are avoiding that issue by banning a bandwidth-hogging set of applications rather than looking for offending content.
I'm actually more curious if the DMCA is even relevant for student owned computers. The DMCA describes a client-server environment where ISP's were afraid of being held responsible for content on their server hardware. It seems like a stretch that any ISP could be held responsible for the content on equipment they neither own nor control. This is the test case I keep hoping to see go to trial.
A university's business is generally education, research, and outreach (the priority and list varies by institution). Despite (in most cases) a lack of profit motive, the allocation of limited resources is still a business decision.
"Nail in the coffin of internet freedom" is a bit of an overstatement. There's no free lunch. Dealing with DMCA takedown notices is a huge burden on campus IT staff (our campus has a network security officer who has spent most of his tenure chasing movies and music) which cannot be ignored without the risk of losing the campus's protection under the DMCA safe-harbor provisions. Further, campuses don't have a magically free internet connection. Most pay into a state-wide consortium for Internet2 access then pay an additional, metered rate for commercial internet traffic. Why should universities spend limited resources to subsidize torrent traffic?
Now before anyone talks about the legitimate p2p use, even that is a questionable use of university resources. Ideally p2p shares bandwidth costs so that everyone gets something for a minor contribution. This doesn't necessarily work out to the benefit of universities since their fat, low-latency pipes take priority over the narrow, slow-upload-speed DSL and cable-folks. Ultimately, the universities have to allocate resources to support university business and this policy must be seen as a business decision. If it is necessary for an aspect of university business, I suspect an exception will be allowed as soon as a faculty member makes the request. If the students are miffed, they can pay for commercial wireless access (like most cell phone companies offer) for on campus or use xDSL or cable at home.
Please read what I said. False advertising is never acceptable. However, expecting enterprise class service for mass-market, commodity prices is unreasonable.
I am not about to defend the ISP practice of advertising "unlimited" downloads, then dropping customers using more than their "fair share". That's even stupider than presuming that high levels of bandwidth use is an indication of copyright infringement.
However, if you are paying $30/month for DSL or cable, you should keep in mind that that is comparable to AOL/CompuServe/Genie/Prodigy rates 15 years ago. For that you got a minimumal connections speed (at the time typically 9600 or 14400 baud) into their system and then shared access to their internet gateway (if any) and either caps on usage or extreme overage fees. You could pay for higher speed, busness-class services (higher speed modem pools, ISDN, frame-relay, etc.).
I would argue that the current, minimally-acceptable speed in the US is somewhere around a symmetric 384k based on the currently available services and typical usage patterns. At the low price points, the ISP will be focusing on meeting this minimum need across their subscriber base. Their ToS should clearly establish what your usage limits are and the termination clause, upgrade process, or price/MB for exceeding that limit.
On the other hand, if you know that the minimum plan won't meet your need, don't whine or subscribe then bitch about it. Go find a plan that will either through a competitor (if available) or via a business class service. For example, Cox Business Services division will sell cable-modem service to a residence. You just have to call a different phone number. You also get a more tech-friendly ToS agreement.
Too true. Pilots even give permission to call out from time to time. Two years ago I had the good fortune to be on a flight benefiting from a 90 kt. tail wind. As we started to descend (20 minutes before touching down) the pilot came on the PA and suggested that we all break out our cell phones and let people know we were going to be about 45 minutes early. She surely wasn't worried about navigational interference...
For 50 years we've pretended that things were different in space; everyone would ignore national rivalries and history and stare with awe at the daring feats of cosmonauts and astronauts. It was a nice fantasy and flew in the face of reality. The Apollo missions grew out of a fear of sleeping "under a communist moon."
Here's the reality check. The US Navy exists to do a few things:
Project power ashore (i.e. shoot and bomb things that don't float or fly)
Guarantee US access to sea lanes of control
Deny access to SLoCs to US enemies
Both the US Air Force and US Navy have space commands and with good cause. Clearly access to orbit is as critical now as access to the seas were 100 years ago. It is in every nation's self-interest to guarantee its access to orbit. It's not much of a leap to get from there to seeing that having technologies to deny that access to enemies is a strategic advantage. How many lives (on either side of a conflict) might be saved by neutralizing an enemy's communications and recon satellites? It's a no-brainer policy. (Insert Bush joke here...)
Red Hat is not competing with "Big Iron." They are a software service company not a hardware outfit. The biggest of the Big Iron operators, IBM, sells and supports Red Hat. They are competing in the server space with other commercial Linux support outfits, Solaris, *BSD, Win2K3, and a few others. They aren't really competing at all in the desktop space.
That said, Linux is not as full featured as some of the commercial OS offerings. For example, some of the debugging tools available for Sparc Solaris rely on features that the x86 hardware platform simply doesn't implement. Is that a Red Hat issue? No. It's a limitation chosen by the user when they selected a more limited hardware platform.
In the aftermath of the raid, members of the Left and Moderate parties in Sweden have proposed scrapping last year's law that criminalized illegal file-sharing, reported the Local, an English-language newspaper in Sweden.
It looks like a reporter has a hard time distinguishing between legal jurisdictions. I doubt that the Swedes would have wasted time criminalizing something that was already illegal. This is a perfect example of the fuzzy thinking that most people bring to this (admittedly complex) issue.
I've been to LISA a few times over the years and it is easily the best training environment I've run across. Their tutorial sessions (which run over the days before the actual conference kicks off) are every bit as solid as the training you would get from Sun, HP, or Oracle. (I can only speak for the vendor training I've attended.) That said, I also think well of Sun's Solaris admin training and have heard good things first-hand of Red Hat's.
The most important thing, though, is to spend time on it after training. Make it your primary home OS if you are not resposible for an installation at the office. Training on your resume might get you the interview, but knowing how to make it dance will land you the job.
Microsoft originally targeted a 2005 launch for the new Windows, then pushed the release out to 2006 before announcing in March that Vista would again be delayed to improve the product's quality.
Am I the only one that remembers that "Longhorn" was supposed to follow XP about three years? I went a Googling and found plenty of chatter back in 2002 about how pissy customers would be if their new, expensive Software Assurance didn't include an upgrade to the new OS within three years. One of Microsoft's VPs even suggested MS would "do something" if the date slipped that far. It seems that the reporters don't remember anything preceding the original, official release date of 2005.
It really isn't that complicated. There are really only two types of computer/network use allowed by policy: use in direct support of our organization's business and incidental personal use. While inexpensive voice connections to other sites could meet a business need, that isn't what Skype does. Once you are a supernode, your bandwidth is being used in support of other people so the "business use" doctrine doesn't apply.
That leaves "incidental personal use." Here again, there is a lot of bandwidth being used to facilitate communications for unknown persons. This fails the personal use in two ways. First, the amount of traffic is high enough that it can't really be classed as "incidental." Secondly, since our user does not know any of the other people involved, it isn't really "personal" use.
Skype is a great technology, but as a business plan it has problems. Essentially, the only way it works is for Skype to convince other people to provide the most expensive component of the service. Our organization does not consider underwriting the expenses of a commercial entity an appropriate use of our resources.
Even senior management can't break through the barrier.
IT operations are not independent. There is always a senior-enough management level that can decide to override an IT organization's policies. After all, if the IT group isn't meeting the business need, then it is failing no matter what the uptime statistics say.
It sounds like you are really running up against a difference in organizational priorities. For whatever reason, your side of the issue is deemed less important than the arguments advanced by the other group. It may be that there are technical issues you are unaware of. It also may entirely be a political issue: competition between groups for limited resources.
Here are a few of your options:
Give up and find another job.
Go numb and ignore it. After 5:00 it's not your problem
Rail pointlessly at the unfairness of the universe losing sleep and friends along the way.
Engage and try to understand the organization better. Use that understanding to make more focused and effective supporting arguments for your recommendations.
One of these will probably suit your temperament...
If you read the EULA, you are agreeing to allow Skype access to your bandwidth should your machine become a supernode. Well, employees don't own the business's bandwidth and so are generally not in a position to accept those terms. In our case, they are absolutely not in a position to accept those terms.
Since our users cannot agree to the EULA, our organization has banned Skype. While I dislike the traffic, the deciding issue for administration was that the license was totally inappropriate.
Come on people, let's pay attention to the article. Contrary to the poster's headline, nothing in it even hints that using Linux would violate Sarb-Ox. Sarb-Ox is supposed to make investing a bit safer by forcing companies to audit their practices and disclose potential problems.
If someone is building products on GPL code (like, say broadband router/NAT boxes based on Linux) then they are supposed to disclose that tidbit to their investors. The important part is that they don't own all of the intellectual property for that product and investors should know since that could change the company's value. If they fail to disclose the data, then they have violated Sarb-Ox.
Slashdot Mirror
I think you may have missed my point. When in London or Paris, an eight mile commute means jumping on the Tube/Metro for a quick trip then a walk from the station. Unless you are in an area protected from urban sprawl or that has had a substantial amount of federal highway funds dumped into mass transit (not common in the States' political climate, it normally goes into widening roads or building bypasses), then mass transit is not a good option
As an aside, it's unreasonable to say that people (most of whom are neither young nor fit) should commute under their own power for distances of more than a a mile or so. Take a look at American demographics. Saying "man up" shows a lack of consideration for the larger problem.
If we are discussing the US, then this is a misleading statement. A more accurate one would be "In certain urban locations and a few suburban locales where geography has prevented urban sprawl, there is mediocre to barely acceptable public transportation." The best the States have to offer (San Francisco or Washington D.C.) compares poorly with, say, London or Paris. (I've used all four of these and won't speak to those I haven't.) I'm not willing to convert my 8-mile/13-minute commute into an 75 minute bus ordeal because it is "environmentally friendly."
While I don't disagree with most of the conclusions in the article, the financial reasoning is wrong. The purchase cost of a computer is the smallest part of the lifetime costs. If you take his theoretical 100 PC environment, you can expect to need a full-time support position (industry averages are between 100 and 200 systems per support tech). If you pay him/her peanuts ($15/hour) then your cost just went up by $30k/year for the (let's say) four year life expectancy of the equipment. That makes the numbers $177k vs $234k; the percentage savings just dropped through the floor.
It gets worse if you look at the real cost of support people. Windows support techincians are easy to find. Linux support technicians are not (I say this as someone who has tried to hire dozens of support staff at levels from Mac-Hand-Holder-3rd-Class to Unix-Minor-Deity). If you factor in the cost differences of the two skill sets, the savings gets even slimmer.
Finally, the analysis ignores the reality that people like to use the software with which they are familiar. I've run experiments with family and while I (mostly) like OpenOffice, it makes some people crazy because it is different. Non-geek users don't want to learn new software, they want to get their target task completed then go do something else. If you add the additional training/hand-holding to the mix you can pretty much kiss the rest of the theoretical cost savings goodbye.
This is why Windows still dominates the desktop. When you look at the total cost model, there isn't a lot of savings to be had while there are plenty of potential headaches.
To my mind the biggest risk MSFT has taken in the last decade is the set of changes to the UI in Office 2007. Since 1994, there has been essentially no learning curve to an Office upgrade and now there is one. Suddenly, the non-MSFT office suite is the one with better legacy skill support.
Why? Because he's a convict still serving his sentence (that's why he's under home confinement). If he doesn't like the terms of home confinement, he can always go back in the slammer and have even more restricted access.
Anyone care to guess how many of these wre AOL coasters?
Not necessarily true. We had tickets to some random movie preview 10 months ago. As always, I had my phone with me as required for work (if the servers go down...). The goon at the door saw that my phone had camera capability and denied me entrance. After an extended, polite, kafkaesque conversation, my wife watched the movie while I drank coffee and read a book next door.
I agree. My point is that this won't make high-density urban areas self-sustainable. Attempting to build them in the city cores would still require a reduction in population density. Assuming that transportation costs don't stay stable and that people won't be inclined to move, then it makes more sense to build these is suburban areas near the core (i.e. across the river in New Jersey or on Long Island instead of Manhattan). It still dramatically reduces transport costs but doesn't suffer from the insane land prices in the city cores.
It's also worth noting that world history since the Roman Republic has in no small part been shaped by cities trying to feed themselves. When transportation costs or reliability have driven prices up people have starved and/or wars have started.
While vertical factory farms would be pretty nifty, how useful would they really be?
Ultimately, if you want to reduce transport costs (money, fuel, etc.) the people need to be closer to the food production. This seems like an idea better suited to lower-density, urban sprawl (where you can grab relatively large areas without consuming a large percentage of the available space) rather than in the middle of compact urban areas.
Like I said, there may be legitimate academic uses for p2p. However, p2p is just a data transfer technique and there are many to choose from. For example, a course may need to distribute linux ISO's, but that does not imply that they need bittorrent; an FTP server on the campus network may be just as (or perhaps more) efficient.
Personally, I keep hoping someone will ask me for a gopher server to support one of their classes.
The policing side of the DMCA is a mess. However, if I read the University's press release, they are avoiding that issue by banning a bandwidth-hogging set of applications rather than looking for offending content.
I'm actually more curious if the DMCA is even relevant for student owned computers. The DMCA describes a client-server environment where ISP's were afraid of being held responsible for content on their server hardware. It seems like a stretch that any ISP could be held responsible for the content on equipment they neither own nor control. This is the test case I keep hoping to see go to trial.
A university's business is generally education, research, and outreach (the priority and list varies by institution). Despite (in most cases) a lack of profit motive, the allocation of limited resources is still a business decision.
"Nail in the coffin of internet freedom" is a bit of an overstatement. There's no free lunch. Dealing with DMCA takedown notices is a huge burden on campus IT staff (our campus has a network security officer who has spent most of his tenure chasing movies and music) which cannot be ignored without the risk of losing the campus's protection under the DMCA safe-harbor provisions. Further, campuses don't have a magically free internet connection. Most pay into a state-wide consortium for Internet2 access then pay an additional, metered rate for commercial internet traffic. Why should universities spend limited resources to subsidize torrent traffic?
Now before anyone talks about the legitimate p2p use, even that is a questionable use of university resources. Ideally p2p shares bandwidth costs so that everyone gets something for a minor contribution. This doesn't necessarily work out to the benefit of universities since their fat, low-latency pipes take priority over the narrow, slow-upload-speed DSL and cable-folks. Ultimately, the universities have to allocate resources to support university business and this policy must be seen as a business decision. If it is necessary for an aspect of university business, I suspect an exception will be allowed as soon as a faculty member makes the request. If the students are miffed, they can pay for commercial wireless access (like most cell phone companies offer) for on campus or use xDSL or cable at home.
Please read what I said. False advertising is never acceptable. However, expecting enterprise class service for mass-market, commodity prices is unreasonable.
I am not about to defend the ISP practice of advertising "unlimited" downloads, then dropping customers using more than their "fair share". That's even stupider than presuming that high levels of bandwidth use is an indication of copyright infringement.
However, if you are paying $30/month for DSL or cable, you should keep in mind that that is comparable to AOL/CompuServe/Genie/Prodigy rates 15 years ago. For that you got a minimumal connections speed (at the time typically 9600 or 14400 baud) into their system and then shared access to their internet gateway (if any) and either caps on usage or extreme overage fees. You could pay for higher speed, busness-class services (higher speed modem pools, ISDN, frame-relay, etc.).
I would argue that the current, minimally-acceptable speed in the US is somewhere around a symmetric 384k based on the currently available services and typical usage patterns. At the low price points, the ISP will be focusing on meeting this minimum need across their subscriber base. Their ToS should clearly establish what your usage limits are and the termination clause, upgrade process, or price/MB for exceeding that limit.
On the other hand, if you know that the minimum plan won't meet your need, don't whine or subscribe then bitch about it. Go find a plan that will either through a competitor (if available) or via a business class service. For example, Cox Business Services division will sell cable-modem service to a residence. You just have to call a different phone number. You also get a more tech-friendly ToS agreement.
Too true. Pilots even give permission to call out from time to time. Two years ago I had the good fortune to be on a flight benefiting from a 90 kt. tail wind. As we started to descend (20 minutes before touching down) the pilot came on the PA and suggested that we all break out our cell phones and let people know we were going to be about 45 minutes early. She surely wasn't worried about navigational interference...
For 50 years we've pretended that things were different in space; everyone would ignore national rivalries and history and stare with awe at the daring feats of cosmonauts and astronauts. It was a nice fantasy and flew in the face of reality. The Apollo missions grew out of a fear of sleeping "under a communist moon."
Here's the reality check. The US Navy exists to do a few things:
- Project power ashore (i.e. shoot and bomb things that don't float or fly)
- Guarantee US access to sea lanes of control
- Deny access to SLoCs to US enemies
Both the US Air Force and US Navy have space commands and with good cause. Clearly access to orbit is as critical now as access to the seas were 100 years ago. It is in every nation's self-interest to guarantee its access to orbit. It's not much of a leap to get from there to seeing that having technologies to deny that access to enemies is a strategic advantage. How many lives (on either side of a conflict) might be saved by neutralizing an enemy's communications and recon satellites? It's a no-brainer policy. (Insert Bush joke here...)Red Hat is not competing with "Big Iron." They are a software service company not a hardware outfit. The biggest of the Big Iron operators, IBM, sells and supports Red Hat. They are competing in the server space with other commercial Linux support outfits, Solaris, *BSD, Win2K3, and a few others. They aren't really competing at all in the desktop space.
That said, Linux is not as full featured as some of the commercial OS offerings. For example, some of the debugging tools available for Sparc Solaris rely on features that the x86 hardware platform simply doesn't implement. Is that a Red Hat issue? No. It's a limitation chosen by the user when they selected a more limited hardware platform.
You still can't put them on a dimmer. For people liking digital lighting conditions that's fine. However, I'm more a fan of analog light controls...
This is precisely the issue the author failed to make clear. It isn't clear from the article which of the following is correct:
Not having any experience with Swedish law, I don't know which is correct or any of the supporting details and TFA doesn't provide them.
It looks like a reporter has a hard time distinguishing between legal jurisdictions. I doubt that the Swedes would have wasted time criminalizing something that was already illegal. This is a perfect example of the fuzzy thinking that most people bring to this (admittedly complex) issue.
I've been to LISA a few times over the years and it is easily the best training environment I've run across. Their tutorial sessions (which run over the days before the actual conference kicks off) are every bit as solid as the training you would get from Sun, HP, or Oracle. (I can only speak for the vendor training I've attended.) That said, I also think well of Sun's Solaris admin training and have heard good things first-hand of Red Hat's.
The most important thing, though, is to spend time on it after training. Make it your primary home OS if you are not resposible for an installation at the office. Training on your resume might get you the interview, but knowing how to make it dance will land you the job.
Am I the only one that remembers that "Longhorn" was supposed to follow XP about three years? I went a Googling and found plenty of chatter back in 2002 about how pissy customers would be if their new, expensive Software Assurance didn't include an upgrade to the new OS within three years. One of Microsoft's VPs even suggested MS would "do something" if the date slipped that far. It seems that the reporters don't remember anything preceding the original, official release date of 2005.
It really isn't that complicated. There are really only two types of computer/network use allowed by policy: use in direct support of our organization's business and incidental personal use. While inexpensive voice connections to other sites could meet a business need, that isn't what Skype does. Once you are a supernode, your bandwidth is being used in support of other people so the "business use" doctrine doesn't apply.
That leaves "incidental personal use." Here again, there is a lot of bandwidth being used to facilitate communications for unknown persons. This fails the personal use in two ways. First, the amount of traffic is high enough that it can't really be classed as "incidental." Secondly, since our user does not know any of the other people involved, it isn't really "personal" use.
Skype is a great technology, but as a business plan it has problems. Essentially, the only way it works is for Skype to convince other people to provide the most expensive component of the service. Our organization does not consider underwriting the expenses of a commercial entity an appropriate use of our resources.
IT operations are not independent. There is always a senior-enough management level that can decide to override an IT organization's policies. After all, if the IT group isn't meeting the business need, then it is failing no matter what the uptime statistics say.
It sounds like you are really running up against a difference in organizational priorities. For whatever reason, your side of the issue is deemed less important than the arguments advanced by the other group. It may be that there are technical issues you are unaware of. It also may entirely be a political issue: competition between groups for limited resources.
Here are a few of your options:
One of these will probably suit your temperament...
If you read the EULA, you are agreeing to allow Skype access to your bandwidth should your machine become a supernode. Well, employees don't own the business's bandwidth and so are generally not in a position to accept those terms. In our case, they are absolutely not in a position to accept those terms.
Since our users cannot agree to the EULA, our organization has banned Skype. While I dislike the traffic, the deciding issue for administration was that the license was totally inappropriate.
Come on people, let's pay attention to the article. Contrary to the poster's headline, nothing in it even hints that using Linux would violate Sarb-Ox. Sarb-Ox is supposed to make investing a bit safer by forcing companies to audit their practices and disclose potential problems.
If someone is building products on GPL code (like, say broadband router/NAT boxes based on Linux) then they are supposed to disclose that tidbit to their investors. The important part is that they don't own all of the intellectual property for that product and investors should know since that could change the company's value. If they fail to disclose the data, then they have violated Sarb-Ox.