If you download the mcmd file and look at it in a text editor it's just an XML playlist file. The rtsp:// streams play fine with Quicktime on Windows/Mac, having a few problems with mplayer on linux.
The more interesting thing is if you go to the URL in a Treo browser, it launches the SprintTV application and lets you stream all of them to your phone. Hell of a lot better than $5/month/channel.
I think this shows the future of where, IMHO, the music industry, or at least individual artists, should be going: convenience and patronage. People are realizing that the inherent value of a CD, and especially of a downloaded mp3, is pretty close to $0. The main reason to spend money on an inherently worthless mp3 is for convenience: $1 and 1 minute to itunes, or $0 and searching the pirate bay / mucking with bittorrent. The other reason is because you genuinely like the music and want to support the band, so give them money for the sake of giving them money.
This NIN experiment shows it clearly: there's $0 of inherent value in the songs themselves, as they are CC licensed and can legally be copied. For the convenience factor $5 or $10 gets you the mp3s or 4 CDs - pretty hard to beat (ignoring NIN's site being hammered the last few days). The $75 set is clearly patronage; you get the shiny book and some extra CDs with it, but you're really spending the money because you want to give NiN the money. The $300 level is an odd one, as it's a combination of patronage and market speculation for resale.
The current projected price for an LA to SF conventional high-speed train is on the order of $30billion. That's for 500 miles and only going through the fairly small mountains around San Francisco.
NY-LA is 5x as long, and has the freaking Rocky Mountains in the way. How exactly do they figure the $70bil price, even if it was a conventional high speed and not an exotic maglev?
Erm, anyone have a link to anything that's actually worth reading, not a short press release? You know, maybe with some PICTURES of their image processing...
No. Everything about their data centers is a closely guarded secret. Number and location of data centers, what each one does, how many cpus / ram / storage are in each one, etc. You can obviously piece bits together from news stories like these and various estimates given both by them and other people, but you'll never find hard numbers. They consider their computing power one of their competitive advantages, and treat information about it as such.
Everyone's known for a long time that overly cheap western food has destroyed African farming, perpetuating the cycle of poverty. African countries are finally realizing that if they tell the US to screw off, they can do a lot better themselves.
It does have the hoursepower to do complete PS2 emulation in real time, especially if you use the Cell and RSX chips to help. The plan originally was to be 100% software-based backwards compatable. Then it was a combination of software and some real PS2 chips. Now they're just dropping it all.
Quoth the blogger: "With hundreds of languages and API's out there, is anyone really dumb enough to think "yet another one" will fix our parallel programming problems?"
Yet Intel touts its Threading Building Blocks library as just such a fix to many parallel programming problems. Now, TBB is a very nice product, and in many ways it is superior to a lot of existing libraries, APIs, and languages, but one gets the sense that maybe the left hand doesn't know what the right hand is doing at Intel.
Not only TBB, but they're also currently working on a NEW language called Ct that's mentioned in the blog. If "yet another one" is such a bad thing, then why is his own group working on one?
I've am also actively involved with the OpenMP shared memory programming API. I've worked on all the OpenMP specifications "out there" today and I am CEO of the corporation that "owns" OpenMP (the OpenMP Architecture review Board).
It's kind of amusing looking at the languages he lists. MPI and OpenMP are by far the most-used environments, but pthreads and java should probably be next not at the end of the list. Ct, intel's new parallel language, hasn't even been formally announced yet let along there being any released documentation / code for it. CUDA however, Nvidia's competing parallel language, isn't even mentioned though it's been released for months now.
First google result for bittorrent interdiction is a resume from a former MediaSentry (a competitor of MediaDefender) director. The juicy bit (in case it goes away):
Director of Interdiction Development MediaSentry Div of SafeNet (Public Company; 501-1000 employees; SFNT; Computer & Network Security industry) September 2004 -- November 2005 (1 year 3 months) Lead team of software developers and systems engineers developing interdiction solutions for P2P networks. Designed and deployed new Linux based 300+ host distributed infrastructure for p2p decoy distribution with automated command, control and monitoring. Designed and deployed network of filtered eDonkey servers. Managed roll out of new BitTorrent interdiction infrastructure. Implemented multiple p2p file trading clients on hosts utilizing VMware.
It seems like it's basically a distributed network of clients that feed garbage data, trying to slow down everyone's downloading. Sadly for them it seems that uTorrent defeated their work:
After more in-depth analysis...we've determined that the new version DOES affect our interdiction in a negative way. They've added a new "bt.ban_ratio" field that takes into consideration how many good pieces a client has uploaded. [....] We still see a lot of hash_check fails...but now the only peers getting banned are ours. This also affects MediaSentry's interdicted torrents. They are no longer effective on the newest version either.
The title is misleading (of course). A quick glance at Google's jobs site clearly shows they have an entire Hardware category they're hiring for. Google is very interested in making hardware, but for internal consumption. All their servers, their racks, their power, everything in the data center is custom designed for/by them. I've even heard rumors that they're working on custom NICs and switches. Of course there's also the google search appliance.
The point of the question was end-user; talking about the iPhone and PCs. Now we'll see if the gPhone happens and controdicts his answer, but he's saying they don't want to make the device that gives you google's services.
It's kind of amusing how the original research demo is in Java, so it runs on anything. The microsoft demo of course is Windows XP/Vista only. At least they ported the plugin from ActiveX to work in Firefox.
At least on sprint you can disable all the voicemail crap for people calling you. When someone calls me all they get is "Hey, leave a message *beep*". Just call up the voice mail and walk through a bunch of the preferences, it's not too hard to find.
I don't know if anyone still uses it, but I sure as hell remember my UIN even though I haven't touched ICQ since something like 2000. It's great how I can remember the now-usless 8-digit number for so long and not remember my parents birthdays.
If you look at the price list for this chip it states that "Prices include MP3 license of Thomson Multimedia."
If you actually read the price list, you'll see that the VS1000 isn't included on there. All the other chips they produce are MP3 playback, so have to pay the MP3 license. Presumably when they update the price list to include the VS1000, they'll modify the wording.
Sorry to burst your conspiracy theory, but data mining the root name servers would be next to useless. These are the Root name servers and as such all they know about are TLD (top level domains). You ask one of the roots "who is in charge of.com" or.edu or.uk, and they respond. The only data you could ever get from them is distribution among TLDs. Now add caching name servers into the equation (99.999999% of boxes on the internet are behind one) and the statistics becomes even more useless. The records returned by the roots have a lifetime of 2 days. This means it doesn't matter if there's 1 client or 1 million clients behind a particular caching name server, it's only going to ask about.com every 2 days.
>We really need to move to a more formalized structure that reinforces the long-term continuation of the good system we have today. And who's going to run that formalized structure? Hrm, maybe some "good individuals and organizations" would be willing to do it?
Slashdot Mirror
GMane is a *far* easier interface to read than whatever nanog's official archive uses:
http://thread.gmane.org/gmane.org.operators.nanog/54752
If you download the mcmd file and look at it in a text editor it's just an XML playlist file. The rtsp:// streams play fine with Quicktime on Windows/Mac, having a few problems with mplayer on linux.
The more interesting thing is if you go to the URL in a Treo browser, it launches the SprintTV application and lets you stream all of them to your phone. Hell of a lot better than $5/month/channel.
http://qtv.mobitv.com/sprintTVlive.mcd
FAIL.
you can't have extra spaces in a CSV, unless you drive a " Porsche".
I won't be impressed until Excel can pull of something as simple as a flight simulator.
I think this shows the future of where, IMHO, the music industry, or at least individual artists, should be going: convenience and patronage. People are realizing that the inherent value of a CD, and especially of a downloaded mp3, is pretty close to $0. The main reason to spend money on an inherently worthless mp3 is for convenience: $1 and 1 minute to itunes, or $0 and searching the pirate bay / mucking with bittorrent. The other reason is because you genuinely like the music and want to support the band, so give them money for the sake of giving them money.
This NIN experiment shows it clearly: there's $0 of inherent value in the songs themselves, as they are CC licensed and can legally be copied. For the convenience factor $5 or $10 gets you the mp3s or 4 CDs - pretty hard to beat (ignoring NIN's site being hammered the last few days). The $75 set is clearly patronage; you get the shiny book and some extra CDs with it, but you're really spending the money because you want to give NiN the money. The $300 level is an odd one, as it's a combination of patronage and market speculation for resale.
The current projected price for an LA to SF conventional high-speed train is on the order of $30billion. That's for 500 miles and only going through the fairly small mountains around San Francisco.
NY-LA is 5x as long, and has the freaking Rocky Mountains in the way. How exactly do they figure the $70bil price, even if it was a conventional high speed and not an exotic maglev?
Erm, anyone have a link to anything that's actually worth reading, not a short press release? You know, maybe with some PICTURES of their image processing...
No. Everything about their data centers is a closely guarded secret. Number and location of data centers, what each one does, how many cpus / ram / storage are in each one, etc. You can obviously piece bits together from news stories like these and various estimates given both by them and other people, but you'll never find hard numbers. They consider their computing power one of their competitive advantages, and treat information about it as such.
Here's a perfect article today to go along with this story:
http://news.yahoo.com/s/ap/20071209/ap_on_re_af/rethinking_africa_from_the_ground_up
Everyone's known for a long time that overly cheap western food has destroyed African farming, perpetuating the cycle of poverty. African countries are finally realizing that if they tell the US to screw off, they can do a lot better themselves.
I'm looking for an apartment in the area. Thanks for the (negative) pointer =)
It does have the hoursepower to do complete PS2 emulation in real time, especially if you use the Cell and RSX chips to help. The plan originally was to be 100% software-based backwards compatable. Then it was a combination of software and some real PS2 chips. Now they're just dropping it all.
Not only TBB, but they're also currently working on a NEW language called Ct that's mentioned in the blog. If "yet another one" is such a bad thing, then why is his own group working on one?
A note from Tim's bibliography:
A slight conflict of interest maybe?
It's kind of amusing looking at the languages he lists. MPI and OpenMP are by far the most-used environments, but pthreads and java should probably be next not at the end of the list. Ct, intel's new parallel language, hasn't even been formally announced yet let along there being any released documentation / code for it. CUDA however, Nvidia's competing parallel language, isn't even mentioned though it's been released for months now.
It seems like it's basically a distributed network of clients that feed garbage data, trying to slow down everyone's downloading. Sadly for them it seems that uTorrent defeated their work:
The title is misleading (of course). A quick glance at Google's jobs site clearly shows they have an entire Hardware category they're hiring for. Google is very interested in making hardware, but for internal consumption. All their servers, their racks, their power, everything in the data center is custom designed for/by them. I've even heard rumors that they're working on custom NICs and switches. Of course there's also the google search appliance.
The point of the question was end-user; talking about the iPhone and PCs. Now we'll see if the gPhone happens and controdicts his answer, but he's saying they don't want to make the device that gives you google's services.
It's kind of amusing how the original research demo is in Java, so it runs on anything. The microsoft demo of course is Windows XP/Vista only. At least they ported the plugin from ActiveX to work in Firefox.
HD versions:v er_been/apple-iphone-never_been_848x496.movw _to/apple-iphone-how_to_848x496.movl amari/apple-iphone-calamari_848x496.mov
http://movies.apple.com/movies/us/apple/iphone/ne
http://movies.apple.com/movies/us/apple/iphone/ho
http://movies.apple.com/movies/us/apple/iphone/ca
At least on sprint you can disable all the voicemail crap for people calling you. When someone calls me all they get is "Hey, leave a message *beep*". Just call up the voice mail and walk through a bunch of the preferences, it's not too hard to find.
I don't know if anyone still uses it, but I sure as hell remember my UIN even though I haven't touched ICQ since something like 2000. It's great how I can remember the now-usless 8-digit number for so long and not remember my parents birthdays.
Thanks for the useless link. Anyone with a link to an actual advisory, LKML post, lwn, etc that might have some actual information in it?
Here's the original NY Times article with a half dozen pictures of the contraption and on a single page:u siness/17gazshoes.html?_r=1&oref=slogin&pagewanted =all
http://www.nytimes.com/2007/03/17/business/worldb
Sorry to burst your conspiracy theory, but data mining the root name servers would be next to useless. These are the Root name servers and as such all they know about are TLD (top level domains). You ask one of the roots "who is in charge of .com" or .edu or .uk, and they respond. The only data you could ever get from them is distribution among TLDs. Now add caching name servers into the equation (99.999999% of boxes on the internet are behind one) and the statistics becomes even more useless. The records returned by the roots have a lifetime of 2 days. This means it doesn't matter if there's 1 client or 1 million clients behind a particular caching name server, it's only going to ask about .com every 2 days.
>We really need to move to a more formalized structure that reinforces the long-term continuation of the good system we have today.
And who's going to run that formalized structure? Hrm, maybe some "good individuals and organizations" would be willing to do it?
I want one. Anyone managed to find one and put it up on ebay?