Slashdot Mirror
Rosen, Valenti Warn Colleges About P2P
fini writes "The RIAA and MPAA just sent a letter to 2,300 colleges or so, asking to crack down on P2P. Juicy nugget: 'Not only is piracy of copyrighted works illegal, it can take up a significant percentage of a university's costly bandwidth.' Also mentioned, some quasi-FUD on security issues. Six higher-ed honchos also sent a concurring letter. From the RIAA website, here's the story and the letters (PDF only). Mentioned as examples of model policies: Drake University, UNC Chapel Hill and University of Michigan . Interestingly enough, there is no threatening 'or else' stuff in those letters. Not yet..."
...but little do Rosen, Valenti, and the rest of the Consumer Control Cartel know that most college students trade amoungst themselves. Such was the case at my school and my friend's school.
> College is where people are taught to turn off their minds and subscribe to politically-correct orthodoxy, so shearing the sheep at the shearing station is the right tack for Valenti et al.
I don't get what you're saying here. College is where I learned to think for myself instead of letting the Republican party and other associated nazis make my mind up for me.
www.phynd.net is a great solution to P2P. I'm sure the RIAA and MPAA will hate it just as much as P2P, but both colleges and college students love it. Here at UConn, someone has kindly donated the use of their Linux box to run Phynd, which scours the network and catalogs all types of shared files (not just mp3/ogg or movies). In a college with thousands of on campus residents, this saves hunge amounts of internet bandwidth [money] by keeping file sharing traffic entirely on campus. The students are happy because there are almost never any dead links, and files transfer at full speed.
Before this was implemented, P2P programs tied up HUGE amounts of bandwitdth. UConn was forced to administer a bandwidth quota per student, but fortunately that's only for off campus traffic, not local traffic.
But the best thing about it is that the students solved the problem all by themselves. And UConn loves it because it's saving them vast amounts of money.
"No fair, you changed the outcome by measuring it!" - Professor Hubert J. Farnsworth
I don't know about some of the Universities that some /.'ers attend, but I would like to give an account of how mine handles p2p stuff. For obvious reasons, I would like to keep the name of the school anonymous. I work for my school's computer department on Linux clustering and research oriented computing. I have been with the department for about 8 months (almost since I started school). One thing that I really like about my school is how our network admins handle p2p. We have no 'real' policy on it. Basically, we leave it up to the users to determine what is right and wrong. There is a reason for this. We consider our network resources to be 'public domain'. They are paid for, in part, by the university endowment, but mostly it is paid for by tax payer dollars.
Now, since the government of my state has not placed a ban on p2p networks of any type, we are in no position to deny our users the right to use them. We are, however, allowed to throttle their traffic so that more of our bandwidth goes to university-related causes. Really, our department tries as much as possible to turn a blind eye to the p2p situation. We don't want to impede on our students abilities to use the internet in the way they see fit. The university will not, however, back a student who has been busted by the RIAA for illegally possessing copyrighted material.
I'm not sure if this is truly P2P but it is probably close enough (and was to a certain extent inspired by Napster) background.
The major biological databases (EMBL, GenBank, Swissprot etc.) are repositories for sequence data, the information that describes the order of the DNA or proteins (depending on the database). This is collected and curated by a relatively small number of people compared to the size of these databases.
This information is relatively useless without annotation. Annotation is the description of the biological role of the sequence and which bits are important. Unfortunately annotation is difficult and time consuming for people who are non experts to maintain. THis means that many of the entries in the databases are either poorly annotated (poor), have out of date annotation (poor) or blatently incorrect annotation (really bad).
A system of P2P sharing of annotation data has been devised where an expert working on gene Xyz can make available his own annotations without having to burden the overworked people at GenBank/EMBL/DDBJ to make updates to the central database. Interested parties can access this data in a P2P manner (ie a query on 'what does anyone know about Xyz').
One of the main protagonists of DAS (Distributed Annotation System) is Lincoln Stein at Cold Spring Harbor Laboratory (yes, of CGI.pm fame). It will also be presented at the Bioinformatics Open Source Conference in July this year (where I hope to find out a lot more about it too..)
This sounds like a perfect example of productive P2P.Have a look at http://stein.cshl.org/das/ for more information. I know that at least one of the authors on the paper referenced has been guilty of reading Slashdot in the past so maybe he would comment.
And, why not? Rosen and Valenti are right: bandwidth isn't cheap. At my last year of college, there were times the Internet was unusable because of Napster users. Making P2P filesharing more difficult is in the best interests of every college, from both a legal and financial standpoint.
- A.P.
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
Heck yes, they pay. My university ended up doing some router configuration tweaks after the student population consistently used far, far more bandwidth (a few times as much or something like that) than the university's contract with the ISP provided for... with P2P a major culprit.
Only the dead have seen the end of war.
There's plenty of public domain stuff out there. Last time I used a P2P app it was to collect some Christmas music for a party my wife threw. None of it was copyrighted to my knowledge.
While I completely agree that p2p piracy (last time I used it was to get some music by a friend who distributes that way) it is, alas, more than likely that those songs were indeed under copyright. The musicial composition itself is most probably public domain but the particular recordings might not have been. Remember, if it was recorded after the early 20s it is still under copyright unless it either lapsed through neglect or was intentionally placed into the public domain by the author/artist.
jello.
aka aron.
Squid solves the problem either by setting the brower up to use it as a proxy, or by setting up the a router to your upstream provider to transparently re-direct traffic to port 80 to a local squid server. So if anyone on your downside link attempts to hit the same page twice, you'll have no traffic leaves the network. Now your upstream ISP does the same thing. Now the upstream provider to them can do the same thing. The upstream provider from them can do the same thing, all the way to the backbone providers. So essentially, you pass thru various levels of proxying servers to get the content you want. Ummm, this sounds like a lot of Web Servers clusters passing around pages from other servers near them so you don't have to go to directly to the site. Which I'll bet money is paraphrasing your ideal P2P setup for web page delivery. Deployment of squid servers located on every Tier 1,2, and 3 bandwidth providers would look precisely like your P2P solution unless I miss my guess. This would mean when you asked for content, you'd only go upstream to the place it has been close to previously.
It's identical the caching layers in a CPU. First you look in the L1 cache, then you look in the L2 cache, then you go L3 cache. Now you look in RAM, if it isn't there you look on disk. If you've got HSM (heirarchical storage management), you look on tape.
How does this make a site cheaper... Well Akaima is cheaper then enough bandwidth to serve the pages yourself (if your big enough). It's cheaper, because you don't need nearly the bandwith, and Akaima already is huge, thus having economy of scale so they turn a profit on it, while saving you money.
Assuming everyone runs a Squid Cache at the various levels as described above, you'll only get 1 hit per page on your website ever until the cached copy expires on your backbone providers Squid Cache. You have the 11 backbone providers talk directly to each others squid caches. When provider A wants a page from provider B it asks the squid cache of provider B, the squid cache goes to your site gets the page and caches it. From now one, anyone who wants your page will get it from the backbone provider who routes you onto the internet at large. Stop and think about it, you could be getting viewed by ever slashdotter in the world, and see a single hit.
As long as the core caches have enough disk, so you don't get flushed out, you only need enough bandwith to do just that. That's it. You only need enough bandwidth to ensure that your pages can get to the core squid cache quickly enough for the first view not to die of bordom, and that's it.
Of course you run a pretty dull site, beings that you only have a static site. It doesn't work for dynamic content that depends on user level information, but then again neither would P2P.
Actually, I'm not enough of a squid expert to say that for certain. You might get 11 hits per page, one for each backbone provider now that I think about it. Still not an overwhelming amount of traffic.
Now it's time to talk reality... The backbone providers don't want to do that.... It'll save you money, and cost them money because you pay them. However, there is nothing to stop a group of people running clusters of Squid caching servers to do this for themselves. So it's doubtful it'll ever get deployed. It however is a solved problem.
As an aside, bandwidth is expensive because it's expensive. Running a website will *ALWAYS* be expensive. The core providers will always charge a premium price for it because they are providing a rare service you can't get from very many people. They charge a premium price because you'll pay it,and nobody will offer it to you for less. It's that simple... Build all the tools you want, and running a site will still cost the same amount.
Kirby
> For the first time in my scholastic career, I had a history class that went beyond "We had a revolutionary war in 1776. We had a civil war in 1860.
Yeah, a good class would get the dates right.
> Yes, I've heard stories of political correctness being forced on people at universities, but it's not at every one, and even at those universities, you'll find an amazing diversity of opinion if you actually talk to the students and teachers.
Teasing aside, the main point of your post is certainly correct.
Sheesh, evil *and* a jerk. -- Jade
I work tech support for a small midwestern liberal arts college. We've got a 6 megabit outgoing. We had the subnets for KaZaa, WinMX, etc blocked. The first week of classes the connection was great. Then word got out that Morpheus was still working. Within a day the outgoing had slowed to a crawl. I like p2p as much as the next Slashbot but darn it, the network can't take that kind of abuse. We continue to allow LAN file sharing and AIM file transfers because they don't suck bandwidth, but the major p2p apps are just too wasteful...
~Chazzf
No statement is true, not even this one.
Then last week, they basically send an e-mail saying that they didn't care if we downloaded stuff, as long as we didn't upload stuff.
That's even worse. That would only decrease usability of P2P networks and users' ability to find quality content, therefore damaging the network as a whole..
We at least remember what it was like to have to *buy* a cd!
Slashdot lame reply response (TM) almost kicked in here. I was about to say "Hey, I still buy CD's. In fact, I've bought CD's of bands I've found through P2P networks!", but then I realized, No. No I haven't. The last CD I bought was Reel Big Fish - Cheer Up!, which came out late last summer. I buy about a CD every 6 months.
Yeah, I don't buy CD's anymore, unless they're from a band at a show - Now that's something I have done - gone to a show of a band I found on P2P. But, you know, fuck the RIAA.
~Will
sig?
First of all this policy is not used to police ANYTHING that happens on campus. As I looked at the network in the dorm I was in there were no fewer than 200 different MP3's, 50 or so different animie movies, and a handfull of different games. One personal friend of mine used the network to download every Gameboy, SNES, Genisis, DreamCast, PS game that he could find, along with EVERY ISO game he could find. We called him our "supplier". Also before I graduated from Drake last year there were really two seperate networks on campus. One was for the dorms of Goodwin-Kirk and Stalnaker, and the other network was for the rest of the dorms. The administrators have NO idea what is going on for the network. An example of this is my roommate couldn't get on the network (he was running Windows ME..go figure). Anyway I showed him my network and that I could see all the other computers. He didn't believe that I could do that! How is that for informed administrators?
So yeah, RIAA and MPAA might quote Drake's policy as being "good", but if it isn't enforced than what good is it?
Drake CS Grad 2002
Even if the music itself is public domain, you own your performance of it. That radio stations can rebroadcase is because they have a special exemption allowing them to not pay the artist for the performance right. That is because radio is seen as advertising - so much so that studios have developed a payola system on behalf of their performers.