[484] in resnet
Re: Keeping traffic internal
daemon@ATHENA.MIT.EDU (Tyler)
Mon Dec 3 03:30:13 2001
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Message-ID: <Pine.LNX.4.33.0112021446370.25114-100000@hal.rescomp.berkeley.edu>
Date: Mon, 3 Dec 2001 00:29:26 -0800
Reply-To: Resnet Forum <RESNET-L@listserv.nd.edu>
From: Tyler <tyler@RESCOMP.BERKELEY.EDU>
To: RESNET-L@listserv.nd.edu
In-Reply-To: <45672E99BBEDD411A4FC00A0C9CFDE1B5856A8@housing.ucr.edu>
On Sat, 1 Dec 2001, Hsu, Mike wrote:
>
> Yes Gnutella is a tool, but as you an I both know the extent to which
> Gnutella has been used as simply a tool for legitimate purposes is .000001%
> of it's actual use. What percentage of the time do you use the web to
> search for porn? It's main use by most everyone is as an information
> resource. IRC is used mainly for chatting. The only people who would see
> irc as primarily used for warez, divx movies, and kiddie porn would be those
> who participate in such. I use IRC to talk to friends and assist people in
> #slackware and to share knowledge.
though i give you props for your subtle attack on my character (i'll take
the 5th on all your questions and insinuations about porn, warez, and
divx), you can't have it both ways. you know of functional uses for the
web and for IRC, so you say these tools are legitimate. you know of none
for Gnutella, so you say that tool is illegitimate.
now, obviously, gnutella (and the other p2p apps) are currently used
almost exclusively for trading copyrighted media. however, this is beside
the point, because gnutella is just a tool.
resnets are ISPs. while most resnets were created for and continue to
officially be for "academic use",
1. residents have a life outside academia, and will want to use the
network for their outside interests.
2. often times, residents are a captive subscriber base. many schools
include some sort of resnet charge in a resident's housing fees, meaning
the student pays for the network whether or not she uses it. at many
schools, including UC Berkeley, it isn't possible for residents to get any
kind of broadband connection due to technical or contractual constraints.
3. it is therefore unreasonable to say that residents may only use their
resnet connection for "academic purposes". (even given a non-captive
subscriber base, it's at least a little drastic to say, "academic use
*ONLY*".)
academic use should certainly have priority on these networks, and in
the case where p2p traffic is making academic use impossible or very
painful, steps must be taken to protect the academic usability of the
network.
but this is where things get tricky, and you start down a slippery slope.
what kinds of tools are "for academic use"? what about IRC? while i have
heard of people meeting on IRC to discuss projects and have meetings, this
is not the common use. what about IM software?
the point i'm trying to clarify is that tools are tools. providing a local
gnutella server is providing better access to a tool, just as providing
local email service is providing better access to a tool. when you start
concering yourself with which tools are academic and which are not, not
only do you open up all kinds of difficult policy decisions, you move a
lot further away from the "common carrier" protections you have as an ISP,
and you start mucking with academic freedom, the freedom of your network,
and your users' right to privacy.
by the way, note that the provision of tools is not exclusive with making
your network run as well as it can. rate-limiting by itself does not
interfere with what your users can and cannot do, and makes the academic
use portion of your network much happier.
it also bears pointing out that p2p is a very new technology, and there's
still a lot of research on what it can do. so while gnutella may be used
only for mp3 trading now, who knows where it will be in 1|5|20 years. look
at a project like Freenet, which could certainly be used to trade illegal
materials, but is also a powerful tool against censorship and repression
of information. given legislation and procedures that the US government
has adopted recently (DMCA, increased wiretapping abilities for the FBI)
and is considering currently (this business about security tools needing
to be produced by "authorized" sources), freedom of information is going
to become even more crucial in the immediate future.
> I think you read me wrong on this one and I probably phrased it wrong
> too and I apologize for that. What I meant is that, Yes education can be
> effective. But with the size of the networks that we administor and the
> popularity of p2p applications, education is not nearly effective enough.
i already conceded this point. specifically, i wrote: "of course, in the
real world, you have to supplement education with rules and limits."
> It's the same as with drug and alcohol awareness, there is education up
> the ass for drugs and alchohol. And everybody knows that it isn't the
> healthiest thing, yet billions of people participate everyday. Drinking and
> driving is another thing. Everyone and their mothers know that drinking and
> driving can be fatal, but millions of americans do it everyday. People do
> what is pleasing to them. Some people listen to reason, a majority do not.
i was purposefully evading the drug argument, but now the cat's out of the
bag :).
in fact, the current state of alcohol and drug abuse shows exactly why
education is the only solution, and the establishment of rules and laws is
insufficient.
drugs and alcohol (especially alcohol) are available to anyone who wants
them, in spite of laws restricting their use, just as p2p apps are
available regardless of how much you block or rate-limit them. given this
environment, how can you make a person not use drugs and alcohol? you
can't (assuming some semblance of free society). the best you can do is
provide that person with information and encourage him to make an
intelligent decision.
in the view you have presented, you say that education is insufficient to
prevent drunk driving (which is empirically provable by the large and
tragic number of deaths caused by drunk driving). the implication is that
the way to solve this problem is either with more rules, or by taking
alcohol away. since taking alcohol away is impossible in this society for
numerous reasons (cf. Prohibition), the only path left is stricter laws.
while i think that American laws against drunk driving are ridiculously
weak, i don't see strengthening these laws as the magic solution to drunk
driving. even though use of various other narcoits has brought with it
harsher and harsher penalties, these narcotics continue to be used. why
should alcohol, which is deeper entrenched and more socially acceptable
than "harder" drugs, be any different?
instead, a combination of education and laws is called for (as i stated
before), but the education is the really important part. if i may argue
anecdotally: do you remember what the laws are about DUI/DWI? do you know
what the legal consequences of drunk driving are?
now, do you remember watching Red Asphalt when you took Driver's Ed? do
you remember watching a highway patrolman scooping the brains of a drunk
driver off the pavement?
do you think education or law has a stronger effect on the choices you
make about drunk driving, or really about anything at all? if drunk
driving were suddenly not illegal, would you start driving while
intoxicated?
> I speak to residents everyday about the matter, here's the usual
> response, "F that, I want my music/movies/pr0n." Go out to a rave or
> something and try to educate people about ecstacy and see what kind of
> response you get. p2p clients are a drug for people, that's why they use
> it.
convincing someone on E that they shouldn't be on E is perhaps the most
intractable problem ever suggested. "you know how you feel really
amazingly happy and fulfilled right now? well you should stop that."
a better time to try and change a drug user's behavior would be when that
person is crashing off her high. i know a lot more people who wake up The
Morning After with a raging hangover and moan, "i'll never drink again,"
than i know people who swear off alcohol in the midst of a night of
debauchery and warm fuzzies.
to bring this back to the point at hand, how resnets choose to educate
their residents is a big part of how those residents will choose to use
their network. if you can say, "you, huge bandwidth hozer, are the reason
the network is slow," that might hit a resident where he lives. even
better, as UCB rescomp discussed (mostly facetiously), would be to say to
a resident's flooramtes, "George Dorn, in 508, is the reason the network
is slow." we suspected that once that news got out, George Dorn would no
longer be a bandwidth problem :).
> "this is an interesting, if strange, point. i'm equally embarrassed to
> see other resnets stomping on their users' privacy and blithely censoring
> vast sections of the internet. the best i can hope to do is to educate those
> resnets (in the form of presentations at resnet symposia or
> stridently-worded emails). i don't think any single resnet's actions
> reflects on the whole world-wide concept of ResNet, though. this is a
> bit like saying, "the US government is waging an unjust war against poor,
> starving, innocent civillians; therefore, the notion of Government comes
> down one notch." while Government may be inherently evil (i've been
> reading the _Illuminatus!_ trilogy), the actions of any single
> government doesn't necessarily reflect the whole.""
>
> The main argument I have against the whole Stanford running a gnutella
> server is not related to privacy and censorship, I'm talking about the fact
> that by running the server it is condoning the spread of copyrighted
> material. Sure it keeps some of the traffic internal but it's saying hey
> use this to download copyrighted material, it's faster and costs the
> university less money.
the points about privacy and censorship were just to show that other
resnets do things i find reprehensible, but that these resnets don't
reflect on me or the resnet i've worked for any more than stanford's
gnutella server reflects on you or your resnet. here's what i wrote:
"i'm equally embarrassed to see other resnets stomping on their users'
privacy and blithely censoring vast sections of the internet. [...] i
don't think any single resnet's actions reflects on the whole world-wide
concept of ResNet..."
stanford isn't condoning anything except more efficient network usage.
first, as you quoted in your initial mail, stanford's AUP very clearly
says that using the net to trade copyrighted materials is prohibited.
second, gnutella is just a tool, as i've argued above.
> Sorry for making this soo long,
a discussion of philosophical underpinnings is desperately needed at many
institutions, so i think this discussion, though somewhat verbose
(</understatement>), is at least warranted. besides, seeing something on
this list that can't be easily answered with a google search or by reading
the archives is a nice change of pace.
this really is very long though. i feel like i should be offering
powerbars and gatorade to anyone who made it all the way to the end.
tyler
--
"E190 - Midterm.doc": 2,309 characters (an approximate value)"
--Microsoft Word, on meaningful approximation
___________________________________________________
You are subscribed to the ResNet-L mailing list.
To subscribe, unsubscribe or search the archives,
go to http://LISTSERV.ND.EDU/archives/resnet-l.html
___________________________________________________
