Some universities do offer assistance to primary care givers. For example, Columbia University has "parental workload relief" programs (http://worklife.columbia.edu/parental-leave-policies-resources) that provide faculty with a semester of no-teaching and a delayed tenure clock. Still does not make it easy to combine kids and faculty careers. The policy is gender-neutral and there's a small fraction of new fathers who have used the program.
The article fails to mention that in both Korea and Japan, the government played a major role: In Japan, by forcing the incumbent to allow multiple IP operators to use the fiber at very low rates and in Korea, by construction subsidies. This isn't really technology competition as the article calls for - all the countries with cheap broadband use fiber-to-the-home or fiber-to-the-apartment-complex, with CATV playing a relatively small role. (There's very limited WiBro, the local version of WiMax, deployment in Korea, which plays almost no role.) And, as far as I know, all those countries have deployed such cheaper and more advanced infrastructure without violating network neutrality.
The argument about population density might explain the absence of DSL and fiber in Montana, but doesn't exactly explain the high cost of FiOS in New Jersey (or its limited availability). The population density of New Jersey is very similar to that of Korea, at around 400-500 people/sq km.
http://www.cs.columbia.edu/~hgs/rtp/history.html lists early VoIP and voice-over-packet work, dating back into the 1970s. The closest is probably the ITU G.764 standard, which describes packet transmission to interconnect voice systems. These were typically used for trans-oceanic links, to save bandwidth.
If these "cameras" were in the British Airlines/United terminal at JFK, you were likely looking at an electronic light show. It's mounted at the entrance, with about 50 of these things. They can indeed swivel 360 degrees, but have nothing to do with security.
We already have two such standards: Power over Ethernet (PoE) and USB. Neither was designed as a power delivery mechanism, but they have become quite popular for that. There are even USB-powered reading lights. Many VoIP phones and 802.11 base stations can run on Ethernet power, as running power into some places where phones and APs are (suspended ceilings, for example) is quite expensive. Unfortunately, the deliverable power in both cases is small (16W for PoE).
The draft and others will be discussed at the IETF meeting in Seoul next week. One possibility is that instead of a nationwide emergency number, we will get a global emergency identifier, 'sos'.
A large number of networking and OS-related research-oriented code is developed by graduate students, who are almost always funded by taxpayer money, e.g., through the National Science Foundation or ONR (Office of Naval Research). However, universities now have the right (Baylor act) to restrict the dissemination of such code or to claim patents on the development. Many biomedical patents, in particular, are thus paid for twice: once by taxpayer money through the NIH and once again by taxpayers through higher drug prices. (European and Canadian taxpayers save on both counts...)
Many reputable engineering schools run off-campus versions of their Master's programs. (It's less common for undergraduate degrees.) Students on-campus take the class in the old-fashioned way; the class is videotaped and distributed via VHS tapes and FedEx or, more recently, via RealAudio/Video. Off-campus students are held to the same homework and exam schedules as on-campus students. It does require commitment since it's easy to fall behind when work projects interfere. Also, for courses with programming assignments, students sometimes have difficulty replicating the right setup. (Most of our programming assignments are on POSIX-compliant OS.) Usually, there are mailing list and bulletin boards, but students can also email the instructor or call him up during office hours.
There's even a 'virtual' university, NTU (http://www.ntu.edu), that bundles courses from major engineering schools. You end up with an NTU degree in that case.
I teach almost all my graduate classes in this hybrid approach. The local video staff is *very* sensitive to student complaints and won't hesitate to call the dean to have a word with the instructor should the instructor be slow in answering student email, for example.
This is generally not cheap, but you get a real degree with name recognition and faculty that are (mostly) accountable for their behavior.
Most students are enrolled through their companies, who also pay the bill, but I don't think this is required.
Germany has arcor.de, a 192 number (kind of like 900, minus the porn associations) that charges 1c/minute. Compared to the typical rate of $1/minute for per-hour Internet access (e.g., AT&T's low-usage plan), it's cheap. No contract required. From my experience, it worked well - I never got a busy tone and had 50+ kb/s almost all the time. No need to enter local dialing information, either, as the same number works all across Germany. (In general, phone costs in Germany are now significantly *lower* than in the US, for both mobile and land-line phone calls. DSL is also cheaper, at around Euro 33/month. Flat-rate modem dial-up is still rare, so there, the US rates compare favorably if you're online more than an hour a day.)
Also, the article author's are clearly behind on their reading. Plain text passwords (known in HTTP as 'basic' authentication) are explicitly disallowed in RFC 3261, the current SIP spec.
Also, most SIP proxy servers support TLS for secure connections at least between proxies.
The security problems are real, but it doesn't help anybody (except consultants, maybe) to spread myths.
http://www.cs.columbia.edu/sip lists implementations. There aren't many for Unix-related systems, but our CINEMA sipc tool does run on all common Unix/Linux platforms and supports audio, video and other conferencing functionality. It is not free software.
Most SIP tools allow direct communications. Some may need a proxy server. A proxy server is somewhat similar to an H.323 gatekeeper. The VOCAL set includes this, but there are many others, too, listed at the URL above.
Except that Vonage doesn't use the Cisco call manager. The same phone (Cisco 7960) can run either Cisco CallManager or SIP (RFC 2543). Vonage runs SIP, as the TIME article states. We have one in our lab, too.
See http://www.cs.columbia.edu/sip for a description of the technology and other implementations and service providers.
From http://www.cs.columbia.edu/~hgs/internet/advertisi ng.html (bottom), the advertising revenue per viewer is about 0.6c/ad/person (rounding up for inflation). If an hour program has 20 ads, this means that each lost viewer costs the TV station and/or network 12c. They could charge a quarter for an ad-free version delivered on demand and be well ahead of the game, purely financially.
In fairness, it should be noted that many IETF
standards are also encumbered by patents. See http://www.ietf.org/ipr
Some patent holder grant the rough equivalent of
RF licenses, others RAND licenses. Only the latter
is a requirement according to RFC 2026.
One interesting difference, however, is that one
needs at least two independent, interoperable
implementations, both of which have to have exercised the licensing policy, to advance a document from Proposed to Draft Standard stage.
(For reasons unrelated to IPR issues, most recent
IETF RFCs are Proposed Standards, not Draft
Standards.)
There is a pretty comprehensive list of VoIP and multimedia applications at
http://www.cs.columbia.edu/~hgs/rtp and
http://www.cs.columbia.edu/sip (under 'implementations')
These applications are all interoperable, at either the media or call set up level.
Among others, our sipc application runs on Linux (and Solaris, FreeBSD, etc.).
You may find the listings of software using RTP and SIP of interest, as both are widely-accepted Internet standards, with software running on just about any platform. The two sites list some of these Internet standards-compliant tools. For example, our own tool (sipc) integrates audio, video, whiteboard, screen sharing and text chat, using existing media tools.
Slashdot Mirror
Some universities do offer assistance to primary care givers. For example, Columbia University has "parental workload relief" programs (http://worklife.columbia.edu/parental-leave-policies-resources) that provide faculty with a semester of no-teaching and a delayed tenure clock. Still does not make it easy to combine kids and faculty careers. The policy is gender-neutral and there's a small fraction of new fathers who have used the program.
The article fails to mention that in both Korea and Japan, the government played a major role: In Japan, by forcing the incumbent to allow multiple IP operators to use the fiber at very low rates and in Korea, by construction subsidies. This isn't really technology competition as the article calls for - all the countries with cheap broadband use fiber-to-the-home or fiber-to-the-apartment-complex, with CATV playing a relatively small role. (There's very limited WiBro, the local version of WiMax, deployment in Korea, which plays almost no role.) And, as far as I know, all those countries have deployed such cheaper and more advanced infrastructure without violating network neutrality.
The argument about population density might explain the absence of DSL and fiber in Montana, but doesn't exactly explain the high cost of FiOS in New Jersey (or its limited availability). The population density of New Jersey is very similar to that of Korea, at around 400-500 people/sq km.
http://www.cs.columbia.edu/~hgs/rtp/history.html lists early VoIP and voice-over-packet work, dating back into the 1970s. The closest is probably the ITU G.764 standard, which describes packet transmission to interconnect voice systems. These were typically used for trans-oceanic links, to save bandwidth.
Some related resources of mine are at http://www.cs.columbia.edu/~hgs/etc/writing-style. html, based on experiences editing student papers. See also http://www.cs.columbia.edu/~hgs/etc/writing-bugs.h tml.
Unfortunately, it is trivial for the phisher to remove the hundreds of thousands of requests that all come from the same IP address.
http://mice.cs.columbia.edu/getTechreport.php?tech reportID=363&format=pdf provides additional performance data that indicates that PHP is a pretty reasonable choice from a performance perspective, even without the Zend engine.
If these "cameras" were in the British Airlines/United terminal at JFK, you were likely looking at an electronic light show. It's mounted at the entrance, with about 50 of these things. They can indeed swivel 360 degrees, but have nothing to do with security.
We already have two such standards: Power over Ethernet (PoE) and USB. Neither was designed as a power delivery mechanism, but they have become quite popular for that. There are even USB-powered reading lights. Many VoIP phones and 802.11 base stations can run on Ethernet power, as running power into some places where phones and APs are (suspended ceilings, for example) is quite expensive. Unfortunately, the deliverable power in both cases is small (16W for PoE).
The draft and others will be discussed at the IETF meeting in Seoul next week. One possibility is that instead of a nationwide emergency number, we will get a global emergency identifier, 'sos'.
A large number of networking and OS-related research-oriented code is developed by graduate students, who are almost always funded by taxpayer money, e.g., through the National Science Foundation or ONR (Office of Naval Research). However, universities now have the right (Baylor act) to restrict the dissemination of such code or to claim patents on the development. Many biomedical patents, in particular, are thus paid for twice: once by taxpayer money through the NIH and once again by taxpayers through higher drug prices. (European and Canadian taxpayers save on both counts...)
Many reputable engineering schools run off-campus versions of their Master's programs. (It's less common for undergraduate degrees.) Students on-campus take the class in the old-fashioned way; the class is videotaped and distributed via VHS tapes and FedEx or, more recently, via RealAudio/Video. Off-campus students are held to the same homework and exam schedules as on-campus students. It does require commitment since it's easy to fall behind when work projects interfere. Also, for courses with programming assignments, students sometimes have difficulty replicating the right setup. (Most of our programming assignments are on POSIX-compliant OS.) Usually, there are mailing list and bulletin boards, but students can also email the instructor or call him up during office hours.
There's even a 'virtual' university, NTU (http://www.ntu.edu), that bundles courses from major engineering schools. You end up with an NTU
degree in that case.
I teach almost all my graduate classes in this hybrid approach. The local video staff is *very* sensitive to student complaints and won't hesitate to call the dean to have a word with the instructor should the instructor be slow in answering student email, for example.
This is generally not cheap, but you get a real degree with name recognition and faculty that are (mostly) accountable for their behavior.
Most students are enrolled through their companies, who also pay the bill, but I don't think this is required.
Germany has arcor.de, a 192 number (kind of like 900, minus the porn associations) that charges 1c/minute. Compared to the typical rate of $1/minute for per-hour Internet access (e.g., AT&T's low-usage plan), it's cheap. No contract required. From my experience, it worked well - I never got a busy tone and had 50+ kb/s almost all the time. No need to enter local dialing information, either, as the same number works all across Germany. (In general, phone costs in Germany are now significantly *lower* than in the US, for both mobile and land-line phone calls. DSL is also cheaper, at around Euro 33/month. Flat-rate modem dial-up is still rare, so there, the US rates compare favorably if you're online more than an hour a day.)
Also, the article author's are clearly behind on their reading. Plain text passwords (known in HTTP as 'basic' authentication) are explicitly disallowed in RFC 3261, the current SIP spec.
Also, most SIP proxy servers support TLS for secure connections at least between proxies.
The security problems are real, but it doesn't help anybody (except consultants, maybe) to spread myths.
http://www.cs.columbia.edu/sip lists implementations. There aren't many for Unix-related systems, but our CINEMA sipc tool does run on all common Unix/Linux platforms and supports audio, video and other conferencing functionality. It is not free software.
Most SIP tools allow direct communications. Some may need a proxy server. A proxy server is somewhat similar to an H.323 gatekeeper. The VOCAL set includes this, but there are many others, too, listed at the URL above.
Except that Vonage doesn't use the Cisco call manager. The same phone (Cisco 7960) can run either Cisco CallManager or SIP (RFC 2543). Vonage runs SIP, as the TIME article states. We have one in our lab, too.
See http://www.cs.columbia.edu/sip for a description of the technology and other implementations and service providers.
From http://www.cs.columbia.edu/~hgs/internet/advertisi ng.html (bottom), the advertising revenue per viewer is about 0.6c/ad/person (rounding up for inflation). If an hour program has 20 ads, this means that each lost viewer costs the TV station and/or network 12c. They could charge a quarter for an ad-free version delivered on demand and be well ahead of the game, purely financially.
See our 7DS for an implementation of a closely related concept.
In fairness, it should be noted that many IETF
standards are also encumbered by patents. See http://www.ietf.org/ipr
Some patent holder grant the rough equivalent of
RF licenses, others RAND licenses. Only the latter
is a requirement according to RFC 2026.
One interesting difference, however, is that one
needs at least two independent, interoperable
implementations, both of which have to have exercised the licensing policy, to advance a document from Proposed to Draft Standard stage.
(For reasons unrelated to IPR issues, most recent
IETF RFCs are Proposed Standards, not Draft
Standards.)
There is a pretty comprehensive list of VoIP and multimedia applications at http://www.cs.columbia.edu/~hgs/rtp and http://www.cs.columbia.edu/sip (under 'implementations') These applications are all interoperable, at either the media or call set up level. Among others, our sipc application runs on Linux (and Solaris, FreeBSD, etc.).
You may find the listings of software using RTP and SIP of interest, as both are widely-accepted Internet standards, with software running on just about any platform. The two sites list some of these Internet standards-compliant tools. For example, our own tool (sipc) integrates audio, video, whiteboard, screen sharing and text chat, using existing media tools.
See here for some background on this issue, including the letters from Pillsbury.