Researchers at Microsoft's computer science lab in Cambridge have developed a peer-to-peer filesharing system that they say overcomes the scheduling problems associated with existing distribution protocols such as Bit Torrent.
The researchers claim download times are between 20-30 per cent faster, using their network coding approach, than on systems that only code at the server, and between 200 and 300 per cent faster than distributing un-encoded information.
View: Full Article @ The Register
View: Avalanche Whitepaper
The researchers claim download times are between 20-30 per cent faster, using their network coding approach, than on systems that only code at the server, and between 200 and 300 per cent faster than distributing un-encoded information.
View: Full Article @ The Register View: Avalanche Whitepaper
With Microsoft now licensing parts of Nvidia’s technology from the Xbox, it now looks as though the deal will see Microsoft continuing to make payments to Nvidia for several years.
They make a successful product, takes the world by storm and have the biggest market share and whenever they do something new, people start whinging that they are 'abusing their position'.
Grow up.
ItsOnlyMonday, why don't you try studying the sitution with Microsoft before you start espousing Anti-Microsoft rhetoric.
No they aren't. Practically they are a monopoly. Technically there are other solutions available, and there are is no discrimination intentionally done. However, in practice, many are forced to used Microsoft products because of compitability issues with all the others who use them.
Perhaps this is something they will implement as a Windows service, perhaps an update to "Background Intelligent Transfer Service"... It could save them a lot of money on hardware for their servers powering Windows Update.
This sort of "Bit Torrent" technology is fantastic if you create files that you want to distribute but can't afford the server bandwidth. Just provide a torrent link on your site, and offer the file yourself. When lots of people come to download your file at once, they all share the bandwidth by providing the file to themselves. bah, you all know how bit torrent works!
LMAO
Why not just take your pirated DVD's to your local police station and try to sell them there...
Maybe a service to take off load from servers and popular files?
Don't just think about piracy when you hear P2P.
P2P is a modern network model like the server/client model, nothing else. You can design apps around P2P where it's not even POSSIBLE to distribute illegal files. Think Skype for example. It's using P2P to find routes to people to talk to.
Copyright on the music, yes, but not on the recorded performance of it. Symphonies support themselves by selling the CDs.
and it looks impressive. I doesn't look like this format has any specific features that will
make tracking of content easier. So far its just a great idea that will hopefully make it to
being a great p2p protocol one day.
P.S. DRM has nothing to do with filesharing protocols. Its independent.
That's right, there is something out there already why try to make a better protocol? That line of thinking would mean we would be using some crude slow network. (I can communicate with other machines on my at 1MB why research doing it at 100MB?)
Or what about service packs.. XP SP2 was 220megs? Server 2003 SP1 is 330megs... If a service like Bittorrent was built into the OS for more efficient distribution of content like this, it could be a very cool thing.
Just my theory, but it all seems to add up.
AJ
Windows XP Activation/Registration is less about home users/casual piracy, and more about companies selling Windows XP-installed systems illegally (re: the recent suits againt 4 fly by nights).
1. See your XP SP2 released in .torrent and notice effectiveness...
2. Squash the release...
3. Build your own Bit Torrent
4. ???
5. Profit (well, more of it from redueced server loads)
I do not condone the use of P2P, nor do I use P2P. It's bad.
That is like saying Client/Server is bad, ethernet is bad, ect. Pirating software/music/movies is bad yes but it doesn't take a genius to do the same thing via client/server. (IRC FTP sites, Usenet, ect.)
Bit Torrent will still be the best for a range of download material
WHAT'S JOKE!
How many times MS has claimed about "i have the definitive solution of everything!".. and people still trust in MS's claims?.
For bittorrent, the only optimization allowed is about the minimize the bandwidth lossed in transactions, there are not more optimization!. For example, think in 2 pc sharing using bittorrent, each one have 1giga/1giga connection. Theorically the max speed for one pc is 1gigas but the facts can show only a connection of 0.99gigas, losing a minimus percent (less that 1% in many cases) in transaction, parity datas and such.
MS claim that they can obtain a overspeed of +20% and it's IMPOSSIBLE. Why?, because there are no more compresion for compressed datas. Because MS still need to check parity datas and this datas must be send...
I believed that MS tried to make a "filter-torrent" where they play the role of big-brother.
Oh, the horror. Someone theorized a better mousetrap. If it were any company but Microsoft, you'd actually think this was like cold fusion.
This is a white paper, it's like an essay saying "here's an idea how to do something" that, on paper, is better than Bit torrent.
It's not a piece of software, it's R&D. ANYONE can use this paper as a reference to build thier own protocol like bit torrent, heck the next generation P2P networks could use it and be totally independant from microsoft.
This is research, nothing more.
You dont say, eh?
"It's not a piece of software, it's R&D."
"This is research, nothing more."
Really?
http://www.itworld.com/Net/4087/050616msbittorrent/
"Microsoft has developed a prototype of Avalanche and is testing it by using it to distribute software applications to several thousand of its software beta testers, according to a research engineer demonstrating the software in Cambridge. The company has distributed a 4G-byte application in as little as a day, down from about two weeks when it sends a program directly, he said."
So, would software, qualify, as software? Perhaps.. "something more"?
But hey, the actual fact that they have actual software that is currently actually being beta tested by disseminating an actual large application, clearly, is "nothing more than research".
Alas, who can really blame you? After all, you use as your news source, a linked story at one silly ass "tech news site", drawn from another really silly ass "tech news site", whom most likely lifted the story from another (non silly ass) news site/service (without proper attribution), but, in the haste of their sixth-grading-writing-style-and-emotive-reaction capacity, forget to include pertinent aspects of the story.
Indeed.
Some people are indeed, idiots.
Dumbassism - It's Everywhere, All The Time.
Basically, my point was that anyone could make a P2P program from the same kind of technology and have nothing to do with Microsoft.
Considering that the news post is about Avalanche (its in the title of the news article), and not about the white paper released describing the underlying ideas that Avalanche is based upon, your claim here, make no sense. The news article doesnt even mention the existence of the white paper, until the very last line. It does however, mention Avalanche - right away, in the title.
Avalanche isnt the the name of the white paper. Avalanche isnt the name of the basic ideas presented in the early portions of the white paper. Avalanche is the name of the actual software system that Microsoft developed based on the ideas presented.
"Based on the system presented in this paper, we have implemented Avalanche, a real system using network coding. Through Avalanche, we are currently investigating the
benefits of using network coding to distribute very large files to a large number of users in realistic settings. During the design and implementation of Avalanche we have identified various practical issues related to network coding."
Again, the news article refers to Avalanche. Avalanche, is actual software. Avalance is an actual system. It is actually extant, and actually being used. It not "abstract theory". Avalanche is the name of an actual peer to peer file sharing system with actual software in actual use that Microsoft developed based on the basic ideas presented in early portions of the white paper.
"Avalanche" is its code name at this time, much like Microsoft gives almost all of its software (and other projects) that it develops for later public release. Its actually currently being developed for release, hence the "beta testing".
That white paper, was written, describing the underlying ideas of its Avalanche software and file sharing system. Notice the name of the article: "Enter Avalanche.." Notice the white paper itself, mentions the development and existence of Avalanche, based upon the early basic ideas presented in the paper. Notice that the name of the white paper isnt titled "Avalanche".
The news story is about Avalanche. Avalanche is an actual software system that actually exists and is actually in use.
Remember when you said "its not a piece of software"? Remember that? Remember when you said "this is research, nothing more"? Remember that?
I correct your misapprehension with fact, and you still wont acknowledge it.
This is where you, in good humor and grace, say something like "mea culpa" or "oops, i stand corrected" or "you are right" etc etc.
"This is research, nothing more."
"It's not a piece of software"
Want to try again?
"Basically, my point was that anyone could make a P2P program from the same kind of technology and have nothing to do with Microsoft."
Basically, you dont want to have to acknowledge that your assertions stated above, are, indeed, wrong.
Of course anyone could code a peer to peer program and have nothing to do with Microsoft - its already been done, many times. But this article isnt about any peer to peer program coded by anybody - the article is about Avalanche, coded by Microsoft.
And while its true that "network coding" (described in the white paper) could be implemented by anybody, the article isnt about "network coding" implemented by anybody - the article is about Avalanche, a peer to peer file transfer software system coded by Microsoft.
Look at the title of the article if you are confused.
Its mea culpa time for you.
ALL I was trying to point out was that there were a bunch of idiots saying **** like "this will just lead you into the Hands of the RIAA" and that Microsoft will be able to catch you more easily if you use this or a system based off of it and stuff, yet here you come weiting out essays thinking that you have some big almighty point to disprove whatever my point was, when in actual fact my point was a simple one: The techniques described in this article could be implimented into a P2P system that has nothing to do with Microsoft at all.
That is my point and that is what you missed completely.
Concession works. Try it. For example "you are right, avalanche IS software, i was wrong when i insisted it wasnt" or "you are right, the article is about Avalanche, which is software, i was wrong when i said avalanche wasnt software and the article was about a white paper and that avalanche was research and nothing more"
When you make false claims, and its demonstrated they are wrong, its best to acknowledge you were wrong. You were quite adamant avalanche wasnt software, "research, nothing more"? You even insisted the article was about the white paper, and not avalanche itself (which you still refuse to acknowledge is software). Remember that son?
Look up the word "concede". Look up the phrase "mea culpa".
"..yet here you come weiting out essays thinking that you have some big almighty point to disprove whatever my point was,.."
Points like "its not software" "its research, nothing more"? "the article was about the white paper"...points like those?
"my point was a simple one: The techniques described in this article could be implimented into a P2P system that has nothing to do with Microsoft at all."
That may have been one of your points, and while that is true, that isnt the only "point" you made. Remember when you insisted "its not software" and "its research, nothing more"? Remember those "points"?
Remember when you insisted the "news post refers to a paper"?
None of those "points" is correct, yet, you still cant seem to acknowledge it.
The article is about Avalanche (not the white paper). Avalanche is software (not just research), developed and coded by Microsoft. Software that is currently being beta tested. Its more than research.
Think you can concede any of those "points", contrary to your assertions and avoidance?
"That is my point and that is what you missed completely."
I addressed the whole issue of a non-Microsoft coded P2P app, even one using "network coding":
"Of course anyone could code a peer to peer program and have nothing to do with Microsoft - its already been done, many times. But this article isnt about any peer to peer program coded by anybody - the article is about Avalanche, coded by Microsoft.
And while its true that "network coding" (described in the white paper) could be implemented by anybody, the article isnt about "network coding" implemented by anybody - the article is about Avalanche, a peer to peer file transfer software system coded by Microsoft."
Remember when i said that? its right up there....
I couldnt "miss it completely", if, in fact, i specifically addressed it.
You just cant help making one false assertion after another, can you? Nor will will you probably ever concede they were wrong/incorrect/false, eh?
You can do it..say "i was wrong when i insisted avalanche wasnt software" "i was wrong when i said it was nothing more than research" "i was wrong when i said the article was about a white paper" "i was wrong when i said you completely missed my point"
..........................
Last edited by 76885 on 18 Jun 2005 - 06:18
Also, there's an entire library of classical music and stuff out there that is not copyrighted and readily available on P2P networks, show me a website that has all of that music available for free.
That's not to say that P2P networks aren't used for piracy, of course they are, but so is IRC, newsgroups, even some websites.
http://research.microsoft.com/~pablo/paraload.aspx
Let me know when MS comes up with something original. This "swarming" method has been around since before Bittorrent. MS copies someone else's innovation, makes a big deal out of it through PR and expects everyone to be amazed? Give credit to the people who invented parallel downloading, not MS's mediocre programmers.
Wether that's true or not remains to be seen, however...