Wired Interviews Bram Cohen, Creator of BitTorrent

ZP-Blight writes "Wired has posted an in-depth five page interview with Bram Cohen, the creator of the popular Peer-2-Peer software, BitTorrent."

Article describes eDonkey2000

[Baldrson]

The article's description of Cohen's "invention" is a description of the way eDonkey2000 [edonkey2000.com] works:

Paradoxically, BitTorrent's architecture means that the more popular the file is the faster it downloads - because more people are pitching in. Better yet, it's a virtuous cycle. Users download and share at the same time; as soon as someone receives even a single piece of Fokkers, his computer immediately begins offering it to others. The more files you're willing to share, the faster any individual torrent downloads to your computer.

Re:Hey...wait a second....

[jmcmunn]

Actually, the previous Wired article was not an interview with him. It was an editorial on the "Dark Pyramid" of the pirating underground. There was no interview with Cohen in it if I remember.

I can't seem to get to this new one just now (thank you /.) but it sounds entirely different.

Azureus client is the best

[tedgyz]

If you have not tried Azureus [sourceforge.net], you have not felt the full power of bittorrent.

Re:Could some smart person explain

[Anonymous Coward]

The Official site [bittorrent.com] has everything in detail.

Re:Not a 5 page article

[deglr6328]

It's calld a "printer friendly version [wired.com]".

Re:Article describes eDonkey2000

[Java Pimp]

Given that quoted paragraph and the following, it is apparent that even after interviewing the creator, the author has absolutely no idea what bit torrent is for or how it works.

He [Cohen] sketched out a protocol: To download that copy of Meet the Fokkers...

Yeah, I'm sure that's what he was thinking when he created the protocol...

a user's computer sniffs around for others online who have pieces of the movie

No, trackers keep track of who is downloading or seeding the file, there is no sniffing around. Infact, there is no search capability that I am aware of...

The more files you're willing to share, the faster any individual torrent downloads

Are you kidding me? No... the more people downloading/seeding an individual torrent, the faster it downloads. More files have nothing to do with anything.

Re:bittorrent weakness

[314m678]

go to www.google.com type in FileIwant filetype:torrent Press search.

Article author either misleading or misinformed

[josecanuc]

I read the article (no, really!) and found it to be mildly interesting. What bothered me, though, are the statements that, basically, "the more clients that are uploading pieces, the faster the download gets"

That's all fine and dandy, but the author makes it sound like this gets around the limitation of one's own pipe to the Internet. If you're on a modem, there's no way you are going to cut down a 500MB download from hours to a few minutes, yet the article has a paragraph that implies that an hours-long Kazaa download is cut down to a few minutes with BitTorrent.

Obviously, if the limiting factor is the source pipe, then more sources equals faster download at the destination. This kind of writing bugs me since it doesn't mention such obvious limitations -- it all sounds "miraculous" (or "marketish"?).

1 page version

[Eslyjah]

It would be nice if the submitter or "editors" had linked to the printer-friendly 1 page version [wired.com].

Re:Money

[elFarto the 2nd]

Bram is employed by Valve, the developers of Half Life.

Regards
elFarto

Re:Speed Bittorrent v. Kazaa

[bman08]

I believe that Bittorrent is faster because it links the speed of your download to how much you're uploading in the much vaunted tit-for-tat system. Too many Kazaa users turn of uploading which crushes the network whereas BT punishes you for so-doing... At least that's how the official client works. Some of the more pirate friendly clients might have 'solved' this problem, I don't know.

High School Memories

[echocharlie]

I went to High School with Bram Cohen. He was brilliant back then too. The article paints a pretty good picture of what I remember of him. We went to Stuyvesant, a specialized HS in NYC with a standardized test to get in. Basically, it's a school for uber-nerds.

Found a picture of the Math team back in 1993 [mojo-working.com]. Bram's the guy with the bushy hair in the back row near the center next to the tall asian guy and another guy with a hat. He was the co-captain of the team that year, if I remember correctly. I think he ended up in the State University of New York at Buffalo. That always bothered me for some reason. He definitely was smart enough to make it into a better school. Why did he choose to go to Buffalo?

Re:Why isn't BitTorrent defeatable?

[keytoe]

No idea if there's anything to permanently ignore a client that's pumping out nothing but junk, though - but on a busy tracker, it would get drowned out by all the others. Anyone know?

The BT protocol is designed to leave that choice up to the clients. Pretty much all the clients out there will shun a peer after a certain number of corrupt pieces. If there aren't enough peers, it may try again - but it's up to the client implementation.

The interesting part is that the protocol (or trackers) don't have to deal with those kind of decisions. The clients can each behave however they want, but you are rewarded for playing nice with better connectivity. Badly behaved clients end up with no peers willing to send them data.

There is nothing gained by writing a BT client that is an asshole to it's peers and nothing stopping you from trying. It will simply be ignored by the other peers that aren't assholes.

Wired's reasons for lacking technicality

[drunken dash]

I think Wired purposefully lacked many details of the BT protocol for obvious reasons: their target audience.

I think Wired tried to dumb that article down (waaaaay down!) for the sake of moron's who cant understand what a tracker is, or the concept of optimistic (un)choking (but then again, do you really wanna read about all that stuff in a Wired article??)

seems like everybody's really picking on the little technical details (although I guess they should have reworded it to sound not-quite-as-WRONG).

Re:WJR 760

[NoMoreNicksLeft]

My tax dollars pay for their enforcement, unfortunately. Mind you, originally the legal burden would have been on them to sue in civil court. So, I can't not support them. In other countries, they're even allowed to tax blank media and playback devices.

So, there are plenty of laws in place, that let them use my money. Was it not just a few months ago, where we read how the DOJ would even prosecute civil claims on their behalf?

Re:High School Memories

[Anonymous Coward]

I think he ended up in the State University of New York at Buffalo. That always bothered me for some reason. He definitely was smart enough to make it into a better school. Why did he choose to go to Buffalo?

Money?

SUNY Buffalo was/isn't a bad school by any means, but its prestige factor is definitely lacking for areas outside of New York State (I was originally an Upstater, btw). I had friends who attended UB and got into their Honors program; free ride, tons of perks. Since Stuyvesant is a public school, I'm surmising that he didn't necessarily come from a privileged background. Hence, maybe it was a financial decision?

There's something to be said about being the big fish in a small pond, too. Albeit a frozen pond in the case of UB.

Re:Azureus client is the best

[nadadogg]

It evaluates any torrent that you are on, and if the seed:peer ratio is not good, you will begin seeding it again. That way, if you are on one with 100 seeds and 4 peers, you won't use your bandwidth on it, instead you would be seeding the torrent with 20 seeds and 10 peers.
ps, who modded my previous post a troll for extolling the virtues of a damn good torrent client?

Re:Bram is cool

[garbletext]

Oh, and last on my distant-future wishlist: A financial-incentive packet bartering priority boost. I.e., anyone can download, but if you contribute money to the authors of the content you're downloading (this would require a centralized server, no way around it), you get a faster download rate. The more you contribute, the faster your downloads go; your donation distribution could be handled automatically.

This is done by Soulseek [slsknet.org] and is the reason why I believe them to be doomed in the long run. Soulseek, while not perfect, is an extremely nice music p2p app that allows for the downloading of complete albums (as long as people organize their albums into seperate folders, which most soulseek users do), mostly at high bitrates. Literally most any music you might want can be found on slsk, but lines to download are often 30-1000 users long. Nothing leaving the program on overnight won't fix, but for a non-recurring donation of $5, you will be added to a list of "privelaged users" who jump ahead of any non-donators in line. this makes it possible to download an absolutely obscene amount of good (i.e. not commercial) music in short order. But it also means that soulseek is profiting in a very real sense off of copyright infringement. I don't know where they are located, and doubt that it's the USA, but no matter where they are I'm sure that as soon as they hit a critical mass, they'll be doomed to litigation hell. Maybe I shouldn't post this. oh well.

(disclaimer: even though it might seem so, I have no affiliation with soulseek or its developers. I just like the software.)

Re:Azureus client is the best

[Anonymous Coward]

It seems to me that the word "gay" is already adopting a third meaning. Most folks seem to use it so mean "crap"

Unfortunately the reason for this may stem from childish homophobia, but nonetheless - it's changing and I have heard at least one homosexual male friend use it to mean "crap"

Re:The best thing about bit torrent

[reverius]

yes, read about it here: http://slashdot.org/articles/04/03/17/0210237.shtm l?tid=126&tid=187&tid=95

it links to a NYtimes article, reg required.