...and students are always finding creative ways to degrade the performance of the campus network and residence hall network by playing "amateur network engineer" or installing the latest fad in recreational software.
When students call us complaining that the network is slow or unusable, and we discover that a student has (for example) connected a hub and ten of his friends to the network, half of whom are across the street in a frat house and downloading pron (or whatever), and someone in the frat house thinks it's funny to connect both ends of a crossover cable to the frat house hub...
Wireless access points are particularly egregious because you can't follow the wires and find out what's really connected to your network.
When the admin loses all control like this, no amount of "This is academic!" or "I paid for this!" or "I expect ISP-level service!" is going to make a damn bit of difference. Admin has to provide usable service within the budget. This applies to any network anywhere, not just a residence hall network at a university.
It's extremely important to remember that the network connections provided to students in dormitories are designed and intended for academic use, on a single computer, by the authorized student, only. They are provided so that the student may reliably access university resources (e-mail, library card catalog, etc.) and also, as an added benefit, for periodic high-speed internet access. They are not designed to allow students to provide a 24-hour file service to the internet (a la Napster), to play games, or to constantly saturate the network with (invariably non-academic) multimedia downloads. A network designed to support this would be prohibitively expensive: compare with what you might pay a managed hosting company for a single port, for example.
Heck I am still waiting for folks to find a licensing and copyright statement in the human genome.
Is anyone looking?
Seriously: we can get the raw data, right? Has there been any concerted effort to find any meaning in DNA at all other than the blueprint for life? We've known about mother nature's most reliable data store for decades, now. Are we sure, yet, that the complete works of the great society of 10^n years ago are not just waiting to be found?
Due to the small samples such a machine would process, the error margin is bound to be huge. This is elementary statistics, folks; if you want milligrams per deciliter of blood cholesterol, or any sort of statistic about a body fluid, the more of a sample available to the process, the more accurate it's going to be. Compare this to Nielsen surveying only twenty people, or Gallup only a hundred.
I've been wrong before; maybe a biochemist could chime in and let us know how much blood or urine constitutes a true statistical sample?
I don't own a television. People who watch television are continuously being told what to buy and like, who to love and hate, and how to think.
I firmly believe I'm a happier person without a TV.
I might do a Douglas Adams marathon tonight. Have you read a good book lately?
This is a social problem, not a technical problem. Obviously, you'll need to store the actual billing information in a database, in cleartext, and access it on a regular basis when recurring billing needs to take place; this affords every opportunity for abuse of the data.
The goal, then, becomes motivating your colleagues and/or marketing-and-management-types to use this data responsibly. My suggestion would be to insist that the customer be given unconditional recourse in the event their data is mishandled as a term of the (recurring) sale. I can't say I know of any company that actually does this, though. The ubiquitous "terms of service" and/or "privacy policy" often give a company wide latitude with your data.
You could compare this problem to convincing the management that spamming is a Bad Idea; often they just Don't Get It.
This month's issue of Communications of the ACM features a reader forum titled "How to Address the Global IT Worker Shortage."
Ralph Castain of Fort Collins, Colorado wrote the following, which began on the same page (12) as Google's ad, "Google Seeks Expert Computer Scientists"
As someone who has been involved in industry and academia over the last few years, I find the current IT "shortage" to be somewhat self-inflicted due to several factors:
Super-specialized requirements. Many companies express no interest in investing in employee training. Hence, we are treated to job listings requiring a combination of skills unlikely to be held by anyone outside that specific company. For example, a recent local job ad required candidates to have at least five years of experience with a highly specialized, industry-specific software package; three years of experience with a specific software-development package; and one to three years experience in that particular industry. When contacted, the human resource manager blamed the IT shortage for her difficulty filling the position.
Low salaries/limited experience. A scan of recent job ads at one popular Web site showed an average salary range (where quoted) well below industry standards in the U.S. In addition, experienced programmers (for example, those with more than five years) report being rarely contacted for interviews, even when meeting all other knowledge requirements. This raises the question of whether there really is a shortage within the industry, or a shortage only of entry-level personnel willing to work for below-standard wages.
Inflated job listings. Companies have become notorious for the placement of job listings on popular Web sites they have no real intention of filling. For example, a large multinational company recently placed more than 100 IT job openings on a Web site over the course of a two-week period. At the same time, the company announced massive layoffs, salary reductions, and forced vacations due to declining sales. How likely is it that any of these listings will ever be filled? Yet the IT shortage studies include such listings in their statistics, thereby distorting the overall picture.
As opposed to a shortage of IT professionals in the U.S., the recent experience of a local headhunter is much closer to the norm. After placing a job listing on a Web site for a SQL server DBA, the headhunter was deluged with more than 300 resumes and 100 phone calls within a 24-hour period. A similar experience followed another listing for a C++ programmer.
The fact is many of us in the corporate world outsource our IT needs to foreign companies and professionals simply because this practice is cheaper. Programmers from foreign countries are willing to work for much lower wages than their U.S. counterparts, especially if they stay overseas. The inefficiencies caused by such remote operations are more than covered by the savings in compensation.
IT shortage? Rather than trusting questionable statistics, I'd recommend asking the people in the job market and the head-hunters.
It's obvious to me that the Cue folks are hoping to get better than a reasonable statistical sample of viewing habits. Data this accurate about TV watching habits is extremely valuable to marketing departments at television stations/networks as well as advertisers, and I'd bet they'd pay Cue big money for the info. Bonus: The data is instant, costs near-zero to collect, and can be broken down by individual advertisement.
As a business person, I'd be more likely to buy ad time from a station that can show me "Cue says we have a 65% market share" versus one that shows me "Nielsen's survey says we have a 65% market share". I'd be more likely to buy ad time from a station that can tell me "Your ad was presented in 65,536 homes" versus one that tells me "According to Nielsen, the sponsored show was viewed at least partially in 65,536 homes"
Some may think this is all about tracking individuals, and perhaps it is, also. The data this generates, however, is likely to be most valuable to track entire advertising markets. We've already established that Cue knows where you live (by ZIP code) and that's all this would need.
Whatever. I don't use Windows, and I don't own a television set. I wonder if Nielsen even has a blank for "respondent doesn't own a television set"... Cue certainly doesn't.
Lookup discs in a database (stored where?) as they are loaded into the player
Automatically reconfigure the region.
Back to no fuss no muss DVD playing.
OR
2. Buy up to six DVD players, one for each region you own discs for, and a video switcher/audio mixer, and lots of cables...
Certainly a permanent solution, but very expensive. (methinks with all the groaning this might be worth it...)
For as much as the research, development, and marketing of this product must have cost, they could have just purchased all of their customers a trunked pair of telephone lines (112k total; dual-56k) and gotten a bulk deal on extra 56k modems. This copout is even a few kilobits/sec better than the top speed for "AuDSL", and isn't at all software-intensive. Window$ or Linux or whatever.
In this case, software emulation and stock hardware makes more sense for ham radio, where you otherwise end up purchasing an expensive, specialized modem anyway, and you certainly don't expect speeds like 768kbps from radio (most amateur packet radio is still 1200 baud)
This is a great candidate for some graduate student's research project, but a really hard sell for some poor marketing department.
Re:While we're at it: car mountable, anyone?
on
Tiny Little Computer
·
· Score: 1
but I've yet found a board that takes just a single 12V.
I doubt you'll find one. Converting DC voltages wastes a lot of energy to heat, something you can't usually afford in a portable.
I'd be more worried about the auto's power supply... take a look at the kind of power filtering required for mobile CB or hi-fi, for example. If you're going to purchase/build such a filter, it would be simpler to just be sure it has the output voltages and capacity your application requires.
Slashdot Mirror
When students call us complaining that the network is slow or unusable, and we discover that a student has (for example) connected a hub and ten of his friends to the network, half of whom are across the street in a frat house and downloading pron (or whatever), and someone in the frat house thinks it's funny to connect both ends of a crossover cable to the frat house hub...
Wireless access points are particularly egregious because you can't follow the wires and find out what's really connected to your network.
When the admin loses all control like this, no amount of "This is academic!" or "I paid for this!" or "I expect ISP-level service!" is going to make a damn bit of difference. Admin has to provide usable service within the budget. This applies to any network anywhere, not just a residence hall network at a university.
It's extremely important to remember that the network connections provided to students in dormitories are designed and intended for academic use, on a single computer, by the authorized student, only. They are provided so that the student may reliably access university resources (e-mail, library card catalog, etc.) and also, as an added benefit, for periodic high-speed internet access. They are not designed to allow students to provide a 24-hour file service to the internet (a la Napster), to play games, or to constantly saturate the network with (invariably non-academic) multimedia downloads. A network designed to support this would be prohibitively expensive: compare with what you might pay a managed hosting company for a single port, for example.
Is anyone looking?
Seriously: we can get the raw data, right? Has there been any concerted effort to find any meaning in DNA at all other than the blueprint for life? We've known about mother nature's most reliable data store for decades, now. Are we sure, yet, that the complete works of the great society of 10^n years ago are not just waiting to be found?
I used to get six-figure "offers" for my domain (diamond.org) back in 2000.
None of them ever came through, though.
The cash woulda been nice, too
I've been wrong before; maybe a biochemist could chime in and let us know how much blood or urine constitutes a true statistical sample?
I might do a Douglas Adams marathon tonight. Have you read a good book lately?
The goal, then, becomes motivating your colleagues and/or marketing-and-management-types to use this data responsibly. My suggestion would be to insist that the customer be given unconditional recourse in the event their data is mishandled as a term of the (recurring) sale. I can't say I know of any company that actually does this, though. The ubiquitous "terms of service" and/or "privacy policy" often give a company wide latitude with your data.
You could compare this problem to convincing the management that spamming is a Bad Idea; often they just Don't Get It.
Ralph Castain of Fort Collins, Colorado wrote the following, which began on the same page (12) as Google's ad, "Google Seeks Expert Computer Scientists"
It seems that VA is not so subtly implying that "Open Source" and "Enterprise" are now mutually exclusive!?!
Why on earth is a "Linux" company doing this to the community?
Couldn't VA have simply sold "Enterprise Support" for the open product (a la Red Hat)?
Does VA have a stake in Microsoft these days, or something?
As a business person, I'd be more likely to buy ad time from a station that can show me "Cue says we have a 65% market share" versus one that shows me "Nielsen's survey says we have a 65% market share". I'd be more likely to buy ad time from a station that can tell me "Your ad was presented in 65,536 homes" versus one that tells me "According to Nielsen, the sponsored show was viewed at least partially in 65,536 homes"
Some may think this is all about tracking individuals, and perhaps it is, also. The data this generates, however, is likely to be most valuable to track entire advertising markets. We've already established that Cue knows where you live (by ZIP code) and that's all this would need.
Whatever. I don't use Windows, and I don't own a television set. I wonder if Nielsen even has a blank for "respondent doesn't own a television set" ... Cue certainly doesn't.
- Implement CDDB-like recognition of discs
- Lookup discs in a database (stored where?) as they are loaded into the player
- Automatically reconfigure the region.
Back to no fuss no muss DVD playing.OR
2. Buy up to six DVD players, one for each region you own discs for, and a video switcher/audio mixer, and lots of cables...
Certainly a permanent solution, but very expensive.
(methinks with all the groaning this might be worth it...)
In this case, software emulation and stock hardware makes more sense for ham radio, where you otherwise end up purchasing an expensive, specialized modem anyway, and you certainly don't expect speeds like 768kbps from radio (most amateur packet radio is still 1200 baud)
This is a great candidate for some graduate student's research project, but a really hard sell for some poor marketing department.
I doubt you'll find one. Converting DC voltages wastes a lot of energy to heat, something you can't usually afford in a portable.
I'd be more worried about the auto's power supply... take a look at the kind of power filtering required for mobile CB or hi-fi, for example. If you're going to purchase/build such a filter, it would be simpler to just be sure it has the output voltages and capacity your application requires.
This site has been selling Mac aquariums for a quite a while now. And they buy used Mac cases