Universities Developing Internal, Controlled P2P System
sukottoX writes "Penn State along with MIT and the University of British Columbia are developing a P2P application (called LionShare in the PSU incarnation) to be used only by students, faculty and staff. According to this article at the Penn State Daily Collegian, the file-sharing program, which wouldn't be completed until 2005 at the earliest, would log each transaction, allowing illegal use of the network to be traced. The purpose of this is to lessen the load on servers for tasks such as professors sending files to students, thereby decreasing the amount of manpower necessary to administer them. Funding will come in part by a grant from the Mellon Foundation, as well as from the students' information technology fee."
If P2P isn't being used for illicit file-swapping, is it really that much more efficient or useful for university students and professors than e-mail attachments and the various online course management software packages that are already out there?
DecafJedi
my weblog: apropos of something
Just operate their own tracker with controlled access on who can add things. Standard apache logs will tell you who was downloading which torrents, and you'll ease the load on servers.
Where's your "no legitimate uses" argument now, RIAA?
Freedom isn't free; its price is the well-being of others.
Comment removed based on user account deletion
as well as from the students' information technology fee.
I always loved that about going to a University. I was paying for them to keep up their Windows addiction.
I should think so. This is oddly top-heavy. How is it going to cut down on traffic if students are using Lionshare for class AND Kazasterwire for their friends?
One man's -1 Flamebait is another man's +5 Funny.
not all universities are connected to i2. even then, that i2 connection is only super-fast to OTHER i2-connected institutions.
for example, the dorms at my old school (URI) could get 300k/sec on a transfer from an NYU dorm. however, the connection from the dorms to the rest of the world was a single T1. no good for 5000 students all sharing files.
(fortunately i lived in a house on the 'academic' network that had 10 t3's shared with other RI colleges)
How is this Peer to Peer if students can't send data to each other? Because only profs can post data on the network, it is more like a master-to-serf transaction.
10 Bits= $.25
100 Bits= $.50
110 Bits= $.75
1000 Bits= 1 byte
The funding is to controll what is exchanged with "bittorent".
http://ebgp.net/ccc/
When it comes to brains, they've got the lion's share.
...
But when it comes to brute strength (in number of users)
I'm afraid their network's at the shallow end of the gene pool.
i am a soviet space shuttle
... because we all know how difficult it is to write a P2P system. [/sarcasm]
Methinks it is time to switch careers!
The initial version of LionShare used the LimeWire open source codebase. Good to see open code and open networks being used to build new applications....
smd4985
dozens of university networks(student housing) run local directconnect hubs for anything. helps cut the traffic out of the network too..
and as for legit uses, there's practically no need for such system. local data transfers are CHEAP, and so is hd, it's no biggie to have fast networks that make the need for it(p2p) practically nil(actually it makes more sense to have most such data few centralised servers).
oh well, i guess some universities run their storage servers off from dsl or something.
world was created 5 seconds before this post as it is.
Is it just me, or wouldn't BitTorrent work to ease the load on the servers when professors distribute material? I realize that there are times when BitTorrent isn't the ideal solution, but I don't see why it couldn't be used here.
The only reason I can think of is that what they are really trying to do is institute the transaction logging, and the file distribution is the official reason given.
The program is being designed as a way for students, faculty and staff to exchange personal and academic materials on a sanctioned, secure peer-to-peer network. Another advantage is that large files, which would be impossible to send via e-mail or another method, can be shared.
While it all sounds nice and warm inside, how long will it be before it becomes abused. Now wait before you think it's trolling of me to say this, think about how lax security is at colleges.
Problems aren't with p2p they're with the users of it, and while some may think sharing a file or two isn't a crime, the fact is, it adds up. So for this to work think about the kind of boolean settings someone is going to have to program to search for illegalities.
What is staff going to do when snoop|grep -i *.mp* doesn't work because users decided to rename files to madonna.zip or madonna.sda? It's just something to contend with when indeed they do get these p2p programs out. So while it all sounds nice, and the intentions are good, these 'foundations', schools, and business shouldn't advertise or rather expect no shady dealings to go on using p2p on their networks. Sure it'll be closed to the outside world for a minute or two before someone figures out how to use something like datapipe to break that theory.
Controlled? Sorry never heard of the word
MoFscker
Leave it to Academia to spend way too much money to reinvent something with less functionality than so many other *free* products out there.
Idiots.
Bittorrent, as it is, is poorly suited for serving large numbers of relatively small files, to a population in which only a small percentage of users will want any one file. This solution probably won't cut it.
However, the underlying protocol would work for this just fine. (In fact, you could probably just get away with changing the client; servers could probably be the same, though I don't know all that much about the internals.) I'd say that a customized Bittorrent client would probably be fairly effective.
I do wonder exactly how much use this would be. After all, the main reason why P2P is used is to distribute the bandwidth needed. At a typical university, most of that bandwidth is going to be the university's anyway. It seems that this is more a job for multiple servers than P2P technology; of course, that wouldn't involve any buzzwords.
How can we continue to believe in a just universe and freedom to eat crackers if we have no ale?
What's the advantage of this over a course web page and personal web pages? I mean, isn't that what web pages were designed for, file sharing and information sharing among individuals? And we even have URLs we can include in mail messages to point people at things. And while for MP3 sharing web pages may not work and not everybody has web hosting, evey university student should have web hosting and the skill to put up a web page somehow.
Bets on how long until someone puts together a "gateway" program to connect this network to an outside p2p network?
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
I'm sure this will be popular. It will easily displace every other P2P system in existance.
OTOH, it could make grading easy in the future. To wit:
Got caught trading illegal files -- F
Didn't get caught trading illegal files -- A
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
mv Britney_Spears.mp3 Professor_Boring's_Presentation.pdf
"Luck is the residue of design" -- Branch Rickey
Some days you get the Lion.
Some days the Lion gets you.
But always dress for the hunt (in your RIAA-proof underwear, no doubt).
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
BitTorrent trackers keep logs of all downloads, just like Web servers do.
I get 170K/s from a measly T1 to a Debian mirror on the other side of the state.
I take it URI's connection was already under high load at the time?
tasks(723) drafts(105) languages(484) examples(29106)
Logging assumes some central point where logs are kept. And it's detrimental to privacy.
Why not have clients simply check documents' fingerprints and digital signatures, and only share/download "legit"/"authorized" content? That's the truly distributed way to do this.
I wonder how overloaded those webservers can get from a handful of students downloading some PDF'ed powerpoint presentations though..
SCO employee? Check out the bounty
Oh logging? You mean as in DHCP logging? Sure but what happens when $USERJOE decides to hang out in $USERJOHN's room and oops uses his connection while there to share files with someone else? Aside from that, what happens when(if) users start doing stupid little things like hijacking addresses, arp hijacks, yadda yadda to try to circumvent the p2p programs, then you have an altogether other nightmare on your hands.
There are too many variables to contend with, and personally I see it as another method of someone 'getting even with the system' to an extent.
MoFscker
If it's not p2p... and only profs will be able to post content... and there's auth and central server...
They need a grant to make a website? Sheesh!
Why not a client that like Groove or something that allows more than file sharing? File sharing can only go so far. Since you are in a university setting, you can setup servers to host things. But oh wait, if professors need to share things with the students let's say....they can always use web server. So why P2P again?
Apparently I'm missing something really obvious here. I mean you're going to have a P2P system to distribute college material? Maybe these people have never heard of file uploads through http. Are you going to trust the english department to properly share their files? Are the clients going to do md5 checksums on the files to make sure they haven't been tampered with? Are their servers really suffering that bad from "legitimate" academic traffic?
I mean you're telling me that you couldn't take like 2 Linux/BSD Guru's + around 5 CGI coders and come up with a simple web based solution? Now I have to wait for a client for MY OS of choice which may be a PDA which hasn't even been invented for all they know. I seriously can't see what this thing does at all! What you get with "decentralization" and "lessening the bourdon of administration" you're probably going to get screwed twice over with playing traffic cop and showing people over and over and over how to use the system. I also don't see how "logging every transaction" will help anything. We all know college students have worm ridden machines already, like I couldn't figure out how to hijack someone else machine and start injecting porn in place of the philosophy departments files.
Hell give me just 1 million (10% discount guys!!!) and I could do it myself - I'll even through in the hardware of a couple servers =P.
So you think p2p would be the answer? How about load balancing the servers for the admins there. Look I don't want to sound grinchy or anything but p2p is definitely not the answer to the problems you're mentioning.
This would be a way to ease up on that. Plus, a well-done system would have very good classification of material and no spoofs (no porn instead of lecture notes),
That's a big if. Considering the pranksters running around such schools, think about someone getting into one of the machines and changing everything you read to something else. Wouldn't that be horrible. It's opening up another Pandora's box. As for spoofs, it could also lead to people cheating by sharing answers, etc, which one would have to normally research, which is one of the pros about getting an education. Wouldn't you want to learn it as opposed to being spoon fed it?
Plus, think of the sharing potential. One could share class notes (I have a friend who takes his class notes using a pda, writes straight to latex. The resulting .dvi files were VERY much sought after), material between universities, get data from a course I don't remember and I need to remember *right now*, etc.
Again sorry to sound trollish if I do, but if you don't remember apparently you need to do a better job for yourself instead of being too dependent on technology. In some ways I often think that the older generation scholars were much more smarter than we are. Mainly because of the abuses in technology, and this to me is sort of abusive. Learn something, study it, go out to the library get some air. Why take the chance of so called sharing when 1) information can be changed so easily so make sure that shared network is secured to the tee. 2) Wouldn't you rather socialize with someone instead of being crammed up on a machine?
Just my opinion...
MoFscker
Sowhat happend to plain email and html pages where you can download files? What wait until 2005 when you can do the same today ;)
Internet2 is essentially a high-bandwidth yet limited-admission major backbone provider...
During its heyday, a good chunk of Napster's traffic flowed over I2 because at the time each school routed all traffic headed to another I2 school (by IP space) over the I2 link rather than the main Internet link. However, once university officials got wise to this they either excluded dorm room connects from accessing or did port-level routing.
I2's concept is essentially to be what ARPAnet was meant to be before it got diverted into commercial uses, a closed-admission club where abuse by authorized users overstepping their bounds wasn't such of a threat because such a troublemaker would lose access for good and most likely lose other security clearances and their job as well...
Also, how long will it take before the students develop something that encrypts the name of the file as it is transferred, but offers another server somewhere to rehash the names? (I don't know the technical details of this so please forgive any ignorance on the matter).
Buy Steampunk Clothing Online!
No, a place of learning is the last place anyone should tolerate censorship. Without free speech, there is no accademic freedom. Whithout accademic freedom you don't get an education, you get an indoctrination. A free state depends on real education. Without it, we are slaves. Censorship of electronic publishing is tantamount to book burning. Shame on any University that implements this shit.
Friends don't help friends install M$ junk.
That's easy, you expell the students who run Kaza and what not. It's not right, but it can be done. If your mac address does not give you away, a remote call through IE to get your prcessors UID will. My fear, soon to be realized, is that only "approved" applications will be alowed on campus networks.
Anonymous speech is integral to free speech. If there is no anonymous electronic publishing, there is no free elctronic speech. Campus networks were the last bastion of reasonable networking. Moves like this eliminate the last hope of a free internet over owned wires.
Time to move to free air.
Friends don't help friends install M$ junk.
Accept that the first release is going to be the "Spoofed Client" Programming Contest.
Accept that you have just given students a great kicker to explore embedding seriptious content in the containing school-oriented files . Suddenly, the school is put in the position to declare random bits as being intelligent content or not. My personal container fav: That crazy TIFF
By forming a closed system, you have effectively removed spyware for the RIAA to inject into. Once in, this system is golden. I like it.
mug
Oh, he writes directly to LaTeX source code... (it sure WOULD be a wet dream if a PDA could translate handwritten equations to LaTeX)
The funding is to block out all but the spyware they are building. That way they can track the little nits and crush the ones with double plus ungood thoughts. You don't want people at a University thinking, do you? This makes Carnivore look innocent.
They are destroying what the internet should be. By placing a central inteligence at the core and forbidding alternate services, they are creating one giant collection of dumb terminals. Nice work Penn! Shame on you MIT for letting youself get tarred by association with this kind of junk.
Friends don't help friends install M$ junk.
Illegal P2P networks? What law is their against Gnutella? Bridget, you got some 'spaining to do...
like i said, 5000 students. all running napster.
all of them.
not a pretty sight.
By definition *only* a government can do that.
Also, when you sign up at a school you willingly give up some of your rights to free expression and privacy. Same as when you accept a job offer..
They didnt take it, you gave it up.
Sure it might suck, but they are 100% within their rights to do so. If you dont like it, you dont goto school there. ( or work there, etc )
---- Booth was a patriot ----
How are the transfer rates to other computers in the same dorm? If I understand the article correctly, this is an internal thing. Still, it's good to know that the students get i2 access in their dorms.
In all this talk about tracking every file transfer to protect the rights of music publishers I think I'd have more sympathy for them if they'd disclose what percentage went to whom. One is much less likely to pirate a song if one can know the impact it has on a beloved artist. So how about it RIAA, as a good will jesture. Show us how much of these royalties actually get back to the artist. My back bristles at paying some slick production company $$$$ when Chuck Berry gets $.00000002. Why don't they launch a similar tracking project to show us that we really impact artists not just record companies? I didn't think so...
Bittorrent would, however, be good for those vaguely plausible 100 mebibyte powerpoints. I think that Waste would be a good idea for students to use if they get fed up with the whole "know who gets what files while still using the term 'P2P'" thing. After all, the university would have some trouble stopping them.
You can build a whole lot of file servers. And more importantly, back them up, so that when one of the less literate professors starts opening random email attachments and loses their hard drive they don't lose all their course materials along with it. Seems like it would be much easier (and cheaper) to teach people how to map a network drive.
Hell, I could have almost exactly what they want using dchub in a week or two.
it was 10baseT, and on campus transfers were pretty fast iirc (the frequent network troubles aside)
i think they're trying to deploy wireless throughout the campus instead of constantly rewiring everything. they spent a lot of money wiring up every room on campus only to see 100basetx get real cheap, and then wireless goes and takes off. oops.
This doesn't make me particularly happy as a PSU student. PSU already logs the amount of bandwidth we use here everyday, it's not too much of a stretch to have them give us our own P2P so they can track that. I guess it wouldn't be so bad, but to force all students to pay for such a service is wrong. And, be weary of anything you read in the Daily Collegian.
Isn't there a reason Bittorrent was developed? Why are they reinventing the wheel to lessen the load of those oh so large PDF files that professors distribute? I don't remember the last time my professor wanted to send me a DIVX movie. Sounds pointless to me.
-----Zephyre
I guess knocker foundation was already taken.
Run your own student network in the dorms. Tell the administration to go back to their RIAA masters.
A few access points comprising an isolated network with authentication using a secure file transport client would be undetectable. When I was in university, we ran unofficial and against policy ethernet and cable lines with little difficulty. Wireless should make it a snap.
..don't panic
Now university students can download lecture notes that suddenly cutoff after a few pages, and repeat over and over again. Or they will download what they think is course materials, only to find it is just a file that repeatedly says "what the fuck do you think you're doing?"
#!/
-Carter
The problem with this implementation is that it is a "sneaky" way to do it which allows people to still get in trouble only to be caught at a later time. The job of a school is to teach not punish. They should just teach students not to abuse the school network. Whats so hard about that?
Clickety Click
Why not just setup a system like strangesearch and let people use windows/SAMBA sharing?
autopr0n is like, down and stuff.
This would be so easy to get around. All you have to share legal files...but embed the "tsk tsk" stuff inside of them with something like Camouflage ;)
Not enough if you want to track / log / audit data exchange going on ...
Schools are blocking P2P programs all over the country, but I would have to believe that freeing up bandwidth is NOT their first priority. Look here at Universiy of Miami, where everything P2P we have tried, they have blocked. Even BitTorrent doesn't work, and I am wanting to use it for LEGAL downloading. Ah, but I walked into a Biology computer lab the other day, and what do I find? SETI@home running on practically every computer.
A much better use of their time and lessen the network load, let's teach some of the college professors how to use Acrobat (or whatever) to create PDF's more efficiently.
I worked in the computer commons in college and I can't tell you how many times we had to reset printer spools because of students printing 100 meg pdf's from intranet sites (yes students were taught to download then print).
It is true that graphics will increase the size of any document, however if I can get a 600-700 page apache tutorial in pdf and it takes up only a nominal amount of filespace, what is the deal with the 100 meg 5-10 page pdf!?
:-( --- argh. Despair, I owe again.
As one of the developers of LionShare and a /. addict, I feel obligated to shed a little light on this story, unfortunatly I don't have time at the momement. I managed to post a quick journal entry to dispel some of the myths. In a few weeks, when we have detailed information on the web, I will see if the /. editors would be kind enough to post an accurate followup submission.
The funding is to federalize the authentication used in Lionshare. See Internet2's "Shibboleth" project and imagine that applied to a higher education filesharing project.
The idea is to create federations of like minded users (ie a bio professor at PSU has more in common with a bio professor at Ohio State than they do with a physics professor at PSU). Unfortunatly the current auth realms are set up as islands of universities, not islands of professions. Hopefully federated authentication can solve that and filesharing has been choses as the first non web application to try this with. Also with this fine grained acls can be applied to objects to handle authorization based on user attributes.
BitTorrent is a great app, but does not address any of these issues.
Finkployd