So you're argument is that if there was a space race to the moon, and the first person on the moon notices that the surface is quite loose, so they quickly send back that data and make some "snow shoes" that better support their next space crafts, they should be granted a patent on said "snow shoes" because they were the first to the moon and to make an overwhelming obvious observation?
I argue that 99% of patented solutions to problems are obvious to nearly anyone with common sense, but what makes it seem like the patent is special is that the problem was special.
GPUs are primarily being used because everyone has one. FPGAs are about 10x faster and consume about the same amount of power as a GPU for stuff like bitcoin. the problem is the up-front cost and trust issues from the companies that make these pre-built FPGA boards.
Many USA Unis are run as corps, but not all. My state Unis mostly let in state companies use patents free of charge. Similar with tuitions. In-state tuition is nearly free and we turn a profit on out of state. By offering high quality education, we can draw in students from around the world, who help supplement out in-state costs. While the local populace drives in used Ford or taking public transportation while living with their parents in the suburbs, the out of state students drive around Audis, BMWs, Mercedes that they purchased locally, and pay $2k-$5k/month to have a nice apartment near the Uni.
Since our Unis and Hospitals rank as some of the highest in the world when it comes to R&D, we tend to make a lot of money by selling services or intellectual property. We get people from other countries coming here for leading cancer treatments, and many people from around the USA. Out of state gets charged a lot more than in-state.
It's a great source of income for the state as a whole.
Now, if you think that there's some supernatural method by which arbitrary numbers of telecommunications companies could provide service to every home,
Connection from the home to an exchange point. Use a different VLAN for every ISP that wants to connect to each customer.
At my home I could then connect to multiple ISPs at the same time if I so choose, just by making use of VLANs. Most customers wouldn't need this and they could have a default VLAN set to their port, but the option would be there.
With this setup, you could have up to 4,000-ish ISPs.
Those legal bills are so BS, they should just have a vote. A city could do something like pass an ordinance stating an ISP that does not have a popular vote does not get right-of-way access to private property without the owner's permission. And not having signed permission while having infrastructure or equipment on their property is trespassing.
No, because it was all in Amazon. Who needs tape when you have the cloud, right?
A rule of thumb that I've heard was "It's not backed up until on at least 2 different media types, at least 2 different file systems, and stored in at least 2 different physical locations".
At really high sampling rates, you can "see around corners" using math and looking at photons one at a time. I'm sure someone will think of of something cool to do with high video sampling some time in the future as better techniques and optics become cheap.
Just to give an example of how classrooms aren't the best. I learned more Japanese from a few months of Anime than I learned German from 1 year of college German.
This topic is about "AP CS" classes, which means college credit. If a college has an 80% drop out in a CS 101 class for good reason, they're not going to want fresh high school students to skip the class because they got credit for it in class they were "guaranteed" to pass.
At least where I live, the University must accredit the high school AP courses, so it's the Universities name at stake when they claim students who pass are of their standard.
According to Level 3's write-up on the subject, TW Telcom has a very large metro fiber network that spans many key cities. They offer some fairly high end services up to 100g: http://www.twtelecom.com/telec...
Nothing is ever "free". Even if Google absorbed all of the costs, it means something somewhere would have to get more expensive, and follow all the connections and the true cost just gets passed on. After saying that, the customer is getting the service for free and is even getting the installation at a below-cost price.
In one way, being that the customer is getting charged below costs, you could say Google is paying them. It does require perverting normal logic a bit.
My cousin said his CS 101 class had about an 80% drop out rate because it was too hard for most. I wonder if the HS classes would be of high enough standard to have the same. I guess I could see the benefit of teaching CS just for "fun", but I would hope the HS doesn't give the children a false sense of hope for their college expectations.
Sounds simple enough, yet no company offers 400gb ports that can support QoS at even 50% line rate. This includes all the big companies like Cisco, Juniper, and Alcatel-Lucent. It was a year ago that I saw benchmarks of 8 different brands that supported 400gb or faster ports, and all of them drops speeds dramatically when QoS was enabled. Maybe it has gotten better since then.
If the whole internet was built up to the same standards as the PSTN network is (no congestion whatsoever allowed during typical peak hours)
If everyone who has the Internet was connected 24/7 like they are now, I'm sure we'd be getting that message all the time. Back during dial-up days, the Internet wasn't very popular and few had 2 lines.
A change in ownership doesn't change the fact that the "company" stated they won't sue over certain patents. I can't invite you over, then bet you with a baseball bat, claiming self defense against a trespasser.
I was under the impression that while humans mostly cannot hear ultrasonic sounds, the existence of them can be perceived as a kind of "texture" to other sounds that we can hear. Removing these frequencies all together from all sounds sources can make stuff sounds more artificial.
I have not researched the subject a lot, but these are what I have read across the many years of the discussion reemerging.
Personally, if I listen to an 256kbit MP3, then switch back to FLAC, I hear a slight difference, but it's hard to pinpoint. But if I have one ear listening to the MP3 and the other ear listening to the FLAC, the distinction is HUGE. A lot of the difference is in the highs, and that's with my crappy integrated sound card and semi-decent headphones.
100ms is buffer bloat, unless you're communicating with someone around the globe. After 20ms of jitter, I get packet-loss; And that's how it should be.
I didn't realize there was an issue with jitter on the Internet.
PING vlan80.csw3.Frankfurt1.Level3.net (4.69.154.190): 56 data bytes
64 bytes from 4.69.154.190: icmp_seq=0 ttl=52 time=115.855 ms ...
--- vlan80.csw3.Frankfurt1.Level3.net ping statistics ---
10 packets transmitted, 10 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 114.518/115.397/115.923/0.441 ms
4,300 mile one way, 8,600 mile round trip and only 0.441 ms of jitter based on std-dev. The best "Long Distance" 2,200 mile round trip route I've seen was to LA, and I got about 0.1ms of std-dev jitter.
Get this, my ISP is currently fixing a problem with latency and jitter. That 0.5ms of jitter to Frankfurt should come down a bit once they figure out the issue. But I guess that's what I get when I pay an expensive $100/month for a 50/50 dedicated unlimited line. 2ms of first hop latency with sometimes jitter in the 1ms range is horrible! But really, my ISP is working on this and I'm currently in email with their network supervisor. I used to have 0.1ms first hop latency.
A TB spread perfectly even over 30 days is only 3.08mb/s, which runs about $1.5 for transit. But Comcast said 99% of their bandwidth is not transit, as most comes from CDNs and the like, so it's pretty much free. So depending on your source of data, 1TB can be between $0 and $1.5.
You've never seen a 100tb/s router with 400gb ports. QoS is nearly impossible. There just isn't enough time to do anything other than shuffle around the electrons.
With all the new DWMD tech, we may be better off handling critical stuff like 911 calls at the Layer 1 or 2. Carve up their own dedicated bandwidth and have separate logical routes. Maybe even support dynamic route bandwidth allocation. Have something like 10gb on reserve, but allow it to scale up to a full channel width.
Slashdot Mirror
So you're argument is that if there was a space race to the moon, and the first person on the moon notices that the surface is quite loose, so they quickly send back that data and make some "snow shoes" that better support their next space crafts, they should be granted a patent on said "snow shoes" because they were the first to the moon and to make an overwhelming obvious observation?
I argue that 99% of patented solutions to problems are obvious to nearly anyone with common sense, but what makes it seem like the patent is special is that the problem was special.
And 50% of the USA would suddenly become unemployed when they could no longer get to work.
GPUs are primarily being used because everyone has one. FPGAs are about 10x faster and consume about the same amount of power as a GPU for stuff like bitcoin. the problem is the up-front cost and trust issues from the companies that make these pre-built FPGA boards.
Many USA Unis are run as corps, but not all. My state Unis mostly let in state companies use patents free of charge. Similar with tuitions. In-state tuition is nearly free and we turn a profit on out of state. By offering high quality education, we can draw in students from around the world, who help supplement out in-state costs. While the local populace drives in used Ford or taking public transportation while living with their parents in the suburbs, the out of state students drive around Audis, BMWs, Mercedes that they purchased locally, and pay $2k-$5k/month to have a nice apartment near the Uni.
Since our Unis and Hospitals rank as some of the highest in the world when it comes to R&D, we tend to make a lot of money by selling services or intellectual property. We get people from other countries coming here for leading cancer treatments, and many people from around the USA. Out of state gets charged a lot more than in-state.
It's a great source of income for the state as a whole.
Now, if you think that there's some supernatural method by which arbitrary numbers of telecommunications companies could provide service to every home,
Connection from the home to an exchange point. Use a different VLAN for every ISP that wants to connect to each customer.
At my home I could then connect to multiple ISPs at the same time if I so choose, just by making use of VLANs. Most customers wouldn't need this and they could have a default VLAN set to their port, but the option would be there.
With this setup, you could have up to 4,000-ish ISPs.
Those legal bills are so BS, they should just have a vote. A city could do something like pass an ordinance stating an ISP that does not have a popular vote does not get right-of-way access to private property without the owner's permission. And not having signed permission while having infrastructure or equipment on their property is trespassing.
That would make things interesting.
No, because it was all in Amazon. Who needs tape when you have the cloud, right?
A rule of thumb that I've heard was "It's not backed up until on at least 2 different media types, at least 2 different file systems, and stored in at least 2 different physical locations".
At really high sampling rates, you can "see around corners" using math and looking at photons one at a time. I'm sure someone will think of of something cool to do with high video sampling some time in the future as better techniques and optics become cheap.
Just to give an example of how classrooms aren't the best. I learned more Japanese from a few months of Anime than I learned German from 1 year of college German.
This topic is about "AP CS" classes, which means college credit. If a college has an 80% drop out in a CS 101 class for good reason, they're not going to want fresh high school students to skip the class because they got credit for it in class they were "guaranteed" to pass.
At least where I live, the University must accredit the high school AP courses, so it's the Universities name at stake when they claim students who pass are of their standard.
According to Level 3's write-up on the subject, TW Telcom has a very large metro fiber network that spans many key cities. They offer some fairly high end services up to 100g: http://www.twtelecom.com/telec...
Nothing is ever "free". Even if Google absorbed all of the costs, it means something somewhere would have to get more expensive, and follow all the connections and the true cost just gets passed on. After saying that, the customer is getting the service for free and is even getting the installation at a below-cost price.
In one way, being that the customer is getting charged below costs, you could say Google is paying them. It does require perverting normal logic a bit.
My cousin said his CS 101 class had about an 80% drop out rate because it was too hard for most. I wonder if the HS classes would be of high enough standard to have the same. I guess I could see the benefit of teaching CS just for "fun", but I would hope the HS doesn't give the children a false sense of hope for their college expectations.
Use an email provider that honors flagging a sender as spam.
Sounds simple enough, yet no company offers 400gb ports that can support QoS at even 50% line rate. This includes all the big companies like Cisco, Juniper, and Alcatel-Lucent. It was a year ago that I saw benchmarks of 8 different brands that supported 400gb or faster ports, and all of them drops speeds dramatically when QoS was enabled. Maybe it has gotten better since then.
If the whole internet was built up to the same standards as the PSTN network is (no congestion whatsoever allowed during typical peak hours)
If everyone who has the Internet was connected 24/7 like they are now, I'm sure we'd be getting that message all the time. Back during dial-up days, the Internet wasn't very popular and few had 2 lines.
A change in ownership doesn't change the fact that the "company" stated they won't sue over certain patents. I can't invite you over, then bet you with a baseball bat, claiming self defense against a trespasser.
I was under the impression that while humans mostly cannot hear ultrasonic sounds, the existence of them can be perceived as a kind of "texture" to other sounds that we can hear. Removing these frequencies all together from all sounds sources can make stuff sounds more artificial.
I have not researched the subject a lot, but these are what I have read across the many years of the discussion reemerging.
Personally, if I listen to an 256kbit MP3, then switch back to FLAC, I hear a slight difference, but it's hard to pinpoint. But if I have one ear listening to the MP3 and the other ear listening to the FLAC, the distinction is HUGE. A lot of the difference is in the highs, and that's with my crappy integrated sound card and semi-decent headphones.
100ms is buffer bloat, unless you're communicating with someone around the globe. After 20ms of jitter, I get packet-loss; And that's how it should be.
...
...
I didn't realize there was an issue with jitter on the Internet.
PING vlan80.csw3.Frankfurt1.Level3.net (4.69.154.190): 56 data bytes
64 bytes from 4.69.154.190: icmp_seq=0 ttl=52 time=115.855 ms
--- vlan80.csw3.Frankfurt1.Level3.net ping statistics ---
10 packets transmitted, 10 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 114.518/115.397/115.923/0.441 ms
8:48p local
Start: Wed Jun 11 20:48:47 2014
HOST: pfsense.localdomain Loss% Snt Last Avg Best Wrst StDev
3. 4.71.x.x 0.0% 10 13.6 14.0 11.6 20.3 3.1 Level 3 Chicago
8. 4.69.137.61 0.0% 10 106.0 106.9 105.4 109.1 0.9
9. 4.69.143.145 0.0% 10 118.2 116.9 115.2 121.4 1.6
10. 4.69.154.190 0.0% 10 114.8 116.1 114.6 122.2 2.2
4,300 mile one way, 8,600 mile round trip and only 0.441 ms of jitter based on std-dev. The best "Long Distance" 2,200 mile round trip route I've seen was to LA, and I got about 0.1ms of std-dev jitter.
Get this, my ISP is currently fixing a problem with latency and jitter. That 0.5ms of jitter to Frankfurt should come down a bit once they figure out the issue. But I guess that's what I get when I pay an expensive $100/month for a 50/50 dedicated unlimited line. 2ms of first hop latency with sometimes jitter in the 1ms range is horrible! But really, my ISP is working on this and I'm currently in email with their network supervisor. I used to have 0.1ms first hop latency.
A TB spread perfectly even over 30 days is only 3.08mb/s, which runs about $1.5 for transit. But Comcast said 99% of their bandwidth is not transit, as most comes from CDNs and the like, so it's pretty much free. So depending on your source of data, 1TB can be between $0 and $1.5.
Running the fiber to the home is only 30% of the over all cost. Density has almost nothing to do with it.
You've never seen a 100tb/s router with 400gb ports. QoS is nearly impossible. There just isn't enough time to do anything other than shuffle around the electrons.
Sounds like someone never got an "all circuits are busy" message before. And that was well before 24/7 usage like the Internet.
With all the new DWMD tech, we may be better off handling critical stuff like 911 calls at the Layer 1 or 2. Carve up their own dedicated bandwidth and have separate logical routes. Maybe even support dynamic route bandwidth allocation. Have something like 10gb on reserve, but allow it to scale up to a full channel width.
Not China anymore.