This is a purely FALSE statement!If EVERBODY used BitTyrant, performance would in fact INCREASE!

I’ll explain why:

Imagine everyone using BitTyrant, including you. Your BitTyrant client uploads more to clients who uploads the most to you, as will everyone elses. As a result, slow upload equals slow download speed, fast upload equals fast download speed. You can prove mathematically that in swarms of only BitTyrant clients, every user would have a download speed equal to his or her own upload speed. This is not the case today though. Using traditional bittorrent clients, slow uploaders normally has a higher download than upload speed while the opposite goes for fast uploaders. There are at least two reasons why this is not ideal. First of all, thoose who upload the most doesn’t get proportional compensation. The second reason is that the current allotment will slow down average swarm speed.

BitTyrants protocol is increasing the probability of high speed clients being able to seed. This will increase average swarm speed since high speed users obviously upload faster than slow speed users. As a side effect, BitTyrant also uploads much less to leeching clients (ie increased download speed for honest users). In fact, if everyone used BitTyrant, leaching clients would probably not exist today.

The scenario with increased swarm speed holds true in all swarms with only BitTyrant clients. This is also true in mixed swarms IF everyone who is using BitTyrant has a share ratio of at least 1.00!

BUT, and there is a but. In swarms where only a few of the clients are BitTyrant there is one potential problem. BitTyrant clients will usually finish downloading the torrent faster in mixed swarms as a result of its more advanced protocol, probably with a share ratio of less than 1.00 . This will provide BitTyrant users with the option to upload less by removing the downloaded torrent from the seeding list as soon as it is finished. This would in turn hurt average swarm speed.

Conclution
Increase or decrease in swarm speed will depend on the honesty of BitTyrant users as long as the BitTyrant tit-for-tat protocol is implemented in a minority of clients. If a majority of conventional bittorrent clients would adopt this protocol, the only reasonable effect would be an increased average swarm speed.

He states:
”
Selfishness might work for a single person, but if everybody starts to use BitTyrant, performance will decrease. So, as the makers of the client put it: “When all peers behave selfishly, average performance degrades for all peers, even those with high capacity.”
”

If he bothered to read the paper properly he would have noticed that the client does not work in “selfish” mode by default. The client you download works in strategic mode, which does not cap excess upload capacity.

No wonder there are so many idiotic comments on here when the article isn’t even reading the paper correctly.

Also, as seen on the front page of their website:

Q: Won’t BitTyrant hurt overall BitTorrent performance if everyone uses it?

This is a subtle question and is treated most thoroughly in the paper. The short answer is: maybe. A big difference between BitTyrant and existing BitTorrent clients is that BitTyrant can detect when additional upload contribution is unlikely to improve performance. If a client were truly selfish, it might opt to withhold excess capacity, reducing performance for other users that would have received it. However, our current BitTyrant implementation always contributes excess capacity, even when it might not improve performance. Our goal is to improve performance, not minimize upload contribution.

but you could’ve read that on the site.

that fact is, LEARN HOW SWARMS WORK. Countless I hear friends saying, “ugh, I’m dl at 2-3kb/s I’m blocking my upspeed”

Give and ye shall recieve.

I hope that trackers don’t restrict bittyrant.

and I hope that users dont block bittyrant through plugins.

just share more, I have a ratio of over 5.5

and to be honest, bittyrant isn’t updated much, I wish it was. countless times I’ll be trying to connect to peers and their choked.

simple fact is:

I still seed till my upspeed’s saturated regardless of my client

Older P2P systems (gnutella, kazaa) did nothing almost to encourage uploading, and many users selfishly decided to leech.

Bittorrent’s tit-for-tat scheme rewards users that upload with fast download speeds, and so users find it to be in their own interest to upload as much as possible while downloading.

It’s possible however, to cheat the system by uploading small amounts of data (and giving other lousy download speeds) while downloading much faster.

The new version, BitTyrant, solves the problem by uploading to those that contribute the most (instead of just those that contribute at all).

It may be true that if everybody used the new system with their current upload and download speed limits set, then performance would be degraded, but if everybody used this algorithm, then people would increase their upload speeds to get the better download speeds.

The only people that seem to be harmed by this are those that are trying to maximize their download rate while minimizing the upload rate.

To look at this another way, people are selfish. To say the people should not be selfish and ban this client is the wrong approach. A better strategy is to change the rules of the game so that the strategy which optimizes download speed also maximizes upload speed. That way when a person acts in his own interest, he unintentionally benefits everybody else in the swarm.

Bittorrent is better than anything that came before it because it rewards users who upload. The new strategy provides greater rewards for those who upload, specifically those who upload large amounts of data. To me BitTyrant seems to be something that everybody should adopt for the good of the network.

Also note the difference between the “strategic” method and the “selfish” method (p. 11). “Strategic” means preference for clients with good upload rates. “Selfish” means the same, upstream bandwidth is conserved when it will not benefit the uploader. It seems obvious that adoption of the strategic method will encourage users to increase their upstream bandwidth. The study also notes that this method improves network performance.

Quoted directly from the BitTyrant site:

“Although the evaluation in our paper focuses on users with slightly higher upload capacity than is typically available from US cable / DSL providers today, BitTyrant’s intelligent unchoking and rate selection still improves performance for users with less capacity. All users, regardless of capacity, benefit from using BitTyrant.”

For everyone else. This is a client that is to be used on a good connection. I live in Tokyo. I have fiber. My connection is insane.

Before I switched, I used Azereus. I’d always DL pretty fast and UL a little under that.

With Tyrant

Things are insanely fast. I’ve never hit these speads before with any client. Not even my UL.

By the time I’m done with most torrents, I have a 2 ratio completed. It’s almost obscene.

So pay attention and stop talking about things you only partially understand.

Tyrant is brilliant for the people it was created for.

And it seeds everyone in preference to their UL stats. If everyone is a bastard or has a crap connection, then everyone gets all of that blazing UL speed.

I love it how people get on here and bash things they have never used properly.

The site clearly states that this is NOT for most US users.

It IS a selfish client (it is very stingy with its bandwidth allocations — and on the surface, that definitely does seem like it would give you better speed), but I haven’t seen any benefit whatsoever from that behavior — in fact, just the opposite, much to my surprise.

I’ve been testing it for about 2 weeks now, and I *consistently* got worse download speeds with BitTyrant than when I use my regular, identically-configured Azureus client. And I do mean consistently, as in every single torrent I’ve tried it on, on both public and private trackers. Every single one, bar none, has performed worse with BitTyrant.

I’m not sure where they got their statement “We found that BitTyrant improves performance for all peers that use it”, because I’ve found that to be 100% demonstrably untrue. It wasn’t even _remotely_ close to matching, much less beating, the download performance of my regular Azureus client. And I’ve spent much time tweaking the settings one way or another to try and get better results — absolutely nothing helped.

Whether or not that poor performance was because decent sharers out there have banned it, I couldn’t say, and it doesn’t really matter. What matters are results, and it failed to deliver.

What’s more, after observing it in action first-hand, I’ve ditched the client and added the string “tyrant” to my Stuffer plug-in’s filter.

In summation:
This client just doesn’t deliver. It was a good idea but, for whatever reason, it utterly fails to work as advertised. BIG thumbs down.

All u complainers prbly have shit ratios..

And people with high bandwidth won’t drop off from seeding the torrent as Horde suggested. I don’t care if I upload 5X or more… it only takes a couple hours on a high speed connection,… but I want to download quickly and not sit on it for days at a time.

If BitTyrant was only used by people with high bandwidth then wouldn’t it acutally benefit the swarm as people with high bandwidth usually seed much more than people with low bandwitdh?

Anyhow,… that’s my 2 cents worth…

http://elcheapo.atspace.com

amazing speeds ? lol
how about 1.5MB/s (thats megabytes) up AND down ?
go figure out how much that is pr.hour
You try and see how all those BitComet users leech on that !
I use normal Azureus for years and I have tried this bittyrant. I STILL get raped
by all you asymmetric guys . Only difference is that I get to complete the torrent
a little faster myself , and to see that my downspeed can be almost as good as my upspeed .you guys are still leeching more of me than you are giving back
and my share ratio is still something you have wet dreams about achieving ..

It doesn’t cheat in any way, if there is available bandwidth it will upload to anybody,
it actually LIKES to upload because that increases the chances of getting something back.
I would like if someone could tell
me exactly how it is that it breaks the bit-torrent protocol ..

People have some weird ideas .. it’s “unfair” that someone has a good downspeed,
they are taking too much out of the swarm blablabla
maybe it’s because we also have a good upstream … you think its dsl-users with 128kbit/s uploads who makes bit-torrent as fast as it is, even for those who constantly download at a rate that’s 8-16X their upload?