And of course ALL the other expenses of an ISP are constant per month. Electric, salaries, rent, equipment loan interest, private line interconnects, property taxes, etc...
Some of those are proportional to the amount of equipment, and ISPs have to install more equipment to provide more bandwidth.
Level 3 is surely incurring the cost of the vast majority of the long-haul bandwidth, since, well, they're bloody Level 3.
Not necessarily; if Level 3 is using hot-potato routing then they'll pass traffic to Comcast close to the source and Comcast will have to carry it cross-country. I don't think that "bit-mile" argument is relevant, though, because most of the cost appears to be in the last mile; Comcast just wants Level 3/Netflix to pay for Comcast's last mile upgrades.
I thought that digital cable-boxes were basically just a streaming device, with the channel numbers being a code to tell the cable company what stream to show instead of an actual frequency marker.
No, it's the opposite. The cable company is streaming all channels at all times, and the box just tunes into the channel you choose. That's how it normally works, but in the future with SDV the headend will only transmit channels that somebody is watching.
This isn't a Stanford network; it's a Google network so it will probably connect to the GoogleBone. If you thought Google was fast now, just imagine - with this network you probably get search results before you type the query.
The J2ME VM is not free; phone vendors pay a lot of money to Snoracle for it. Just multiply the price of J2ME by the number of phones that have Dalvik instead of J2ME.
Maybe to save the costs of SAS HBA (at least 200-300$)
That's the reason. OCZ found some really cheap obsolete Silicon Image PCI-X RAID controllers and PCIe-to-PCI-X bridge chips in a warehouse somewhere and decided to kludge together some "SSDs".
The Sandy Bridge GPU is still weak by Apple's standards, but they can't keep using Core 2 forever. As long as OS X can compile OpenCL into AVX code I think Sandy Bridge will work OK in future MacBooks.
If universities had some kind of blanket exception to ITAR, you can bet that they would fill up with foreign students who would get access to restricted hardware (for "educational" purposes, natch) which would then get "lost" and find its way to Iran. ITAR is probably more restrictive than it needs to be, but I don't think it can be fixed by adding loopholes.
The FCC grants them a temporary experimental license because they can't cause much interference out in the middle of the desert. If you fire up OpenBTS anywhere in civilization you're probably breaking the law. Fortunately the equipment is a bit more expensive than CB radio and the carriers have a real incentive to crack down on interferers, so I doubt there will be too many problems in the real world.
I stopped reading right where it said Dalvik is Java based. It doesn't even run Java byte codes...
No, it runs "different" bytecodes that just happen to have exactly the same semantics as Java bytecodes. Dalvik can correctly execute Java programs, therefore it is a Java VM; everything else is syntactic hair-splitting.
What's next, Oracle going to sue GCJ for compiling Java to native?
Don't give them any ideas. Oracle's new policy appears to be "use OpenJDK or pay up"; since GCJ falls in neither category, it could be in trouble.
Re:I'm a bit out of my depth here, so I'm asking..
on
The Case For Oracle
·
· Score: 1
Those editions are supposed to be standardized by the Java Sham Community Process, but when it comes to Android Google didn't participate in the JCP.
Slashdot Mirror
Nothing would stop them. They'd make money, other miners would make slightly less money, and Bitcoin would go on.
AFAIK that loophole was closed decades ago; legal fees have to be reasonable in the eyes of the judge.
Even Apple and D-Link don't support DS-Lite and won't provide free firmware upgrades. Basically nothing on the market today has adequate IPv6 support.
The XO 1.0 supposedly cost $188, so $165 isn't surprising. Unless you think they've been lying all along and no OLPC customer has outed them.
Plug all regular consumers behind a IPv4 NAT and give the servers the remaining IPv4 addresses. Also give everybody IPv6 addresses.
This is such a good idea that it's called DS-Lite and many ISPs are planning to implement it. http://en.wikipedia.org/wiki/Dual-Stack_Lite (looks like this could be improved)
And of course ALL the other expenses of an ISP are constant per month. Electric, salaries, rent, equipment loan interest, private line interconnects, property taxes, etc ...
Some of those are proportional to the amount of equipment, and ISPs have to install more equipment to provide more bandwidth.
Did you not hear about the Hackintosh movement and Psystar? Plenty of people have complained about OS X being tied to Macs.
Level 3 is surely incurring the cost of the vast majority of the long-haul bandwidth, since, well, they're bloody Level 3.
Not necessarily; if Level 3 is using hot-potato routing then they'll pass traffic to Comcast close to the source and Comcast will have to carry it cross-country. I don't think that "bit-mile" argument is relevant, though, because most of the cost appears to be in the last mile; Comcast just wants Level 3/Netflix to pay for Comcast's last mile upgrades.
I thought that digital cable-boxes were basically just a streaming device, with the channel numbers being a code to tell the cable company what stream to show instead of an actual frequency marker.
No, it's the opposite. The cable company is streaming all channels at all times, and the box just tunes into the channel you choose. That's how it normally works, but in the future with SDV the headend will only transmit channels that somebody is watching.
As it is, IPv6 must be run in dual-stack mode
That was true a few years ago, but now we have DS-Lite so you can free up those IPv4 addresses.
Wow, I can connect to the Internet from "inside" my eyeballs? Sign me up!
You don't want to go there; just think of the spam.
I suspect that a lot of Akamai's data transfer is either not counted or not attributed to Akamai because it comes from "inside" the eyeball ISPs.
This isn't a Stanford network; it's a Google network so it will probably connect to the GoogleBone. If you thought Google was fast now, just imagine - with this network you probably get search results before you type the query.
This network is for houses on Stanford's campus where faculty and staff live. The students will have to be content with only 100 Mbps in the dorms.
The J2ME VM is not free; phone vendors pay a lot of money to Snoracle for it. Just multiply the price of J2ME by the number of phones that have Dalvik instead of J2ME.
Think about it; narrow SAS 2.0 is 6 Gb/s and wide is four lanes.
Maybe to save the costs of SAS HBA (at least 200-300$)
That's the reason. OCZ found some really cheap obsolete Silicon Image PCI-X RAID controllers and PCIe-to-PCI-X bridge chips in a warehouse somewhere and decided to kludge together some "SSDs".
A wide SAS 2.0 cable is 2.4 GB/s ful duplex, which blows away anything in this article.
The other completely botched thing was providing no way for an IPv6-only host to ever talk to an IPv4-only host.
This has finally been fixed with NAT64+DNS64.
HP and Hynix are doing memristors, while the entire rest of the industry is doing phase-change memory.
The Sandy Bridge GPU is still weak by Apple's standards, but they can't keep using Core 2 forever. As long as OS X can compile OpenCL into AVX code I think Sandy Bridge will work OK in future MacBooks.
If universities had some kind of blanket exception to ITAR, you can bet that they would fill up with foreign students who would get access to restricted hardware (for "educational" purposes, natch) which would then get "lost" and find its way to Iran. ITAR is probably more restrictive than it needs to be, but I don't think it can be fixed by adding loopholes.
The FCC grants them a temporary experimental license because they can't cause much interference out in the middle of the desert. If you fire up OpenBTS anywhere in civilization you're probably breaking the law. Fortunately the equipment is a bit more expensive than CB radio and the carriers have a real incentive to crack down on interferers, so I doubt there will be too many problems in the real world.
I stopped reading right where it said Dalvik is Java based. It doesn't even run Java byte codes...
No, it runs "different" bytecodes that just happen to have exactly the same semantics as Java bytecodes. Dalvik can correctly execute Java programs, therefore it is a Java VM; everything else is syntactic hair-splitting.
What's next, Oracle going to sue GCJ for compiling Java to native?
Don't give them any ideas. Oracle's new policy appears to be "use OpenJDK or pay up"; since GCJ falls in neither category, it could be in trouble.
Those editions are supposed to be standardized by the Java Sham Community Process, but when it comes to Android Google didn't participate in the JCP.