Talk:BitTorrent/Archive 5

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Archive 4 Archive 5 Archive 6


In the article indexing is better described as in BitTorrent index. merging or splitting it? mabdul 0=* 14:11, 24 October 2008 (UTC)

List of file transfer protocols

should we add bittorrent and other P2P protocls to this list? mabdul 0=* 16:55, 4 February 2009 (UTC)

I think the BitTorrent protocol would fit in there. The other P2P clients are a bit more complex, as FastTrack and Gnutella both use simple HTTP for the actual file transfer. Rurik (talk) 14:47, 5 February 2009 (UTC)

The section Implementations should be updated based on information about the unofficial fork of torrentflux, torrentflux-b4rt. The functions "ZIP compression, RAR decompression and Playstation 3 streaming & download support" are in torrentflux-b4rt, along with other functions. —Preceding unsigned comment added by (talk) 03:50, 18 February 2009 (UTC)

ratio - how it exactly works?

There is "ratio" mensioned in article (in "The leech problem" subsection) but nowhere is described how trackers count it. Is client application sending information about how many data it down- and uploaded? So why it sholdn't report fake amount of uploaded data? So maybe trackers take information about one client from other peers? —Preceding unsigned comment added by (talk) 21:37, 23 May 2009 (UTC)

When clients "check in" with the tracker, usually every 30 minutes, they send information about how many total bytes they've uploaded and downloaded. Rurik (talk) 22:59, 23 May 2009 (UTC)
Ok, so maybe you write this to article? (talk) —Preceding undated comment added 23:42, 23 May 2009 (UTC).

elaboration needed on article

explain the "swarm", it is still a rather unknown concept without description. Murakumo-Elite (talk) 08:44, 25 June 2009 (UTC)

The swarm is the collection of peers that are participating in a single torrent. I can't see a good place to put this in the article, though. Theymos (talk) 18:03, 25 June 2009 (UTC)
I ask about this because there are times where I run a torrent with 0(96) and 20(187)[or relative numbers] and I have NEVER once gotten a clear lock on the complete files, weeks would go by and still no chance at completing it. This was always something I needed to understand to curtail any frustration. Murakumo-Elite (talk) 07:11, 26 June 2009 (UTC)
The tracker is sending you the locations of peers in the swarm, but they are not able to connect to you because your firewall/router/proxy is blocking them. I can't be more specific than that without more information about your configuration, and this talk page is for discussing the article. Ask at the reference desk if you can't figure it out. Theymos (talk) 07:55, 26 June 2009 (UTC)

checksum or SHA-1 hash code?

Not consistent (talk) 22:38, 26 June 2009 (UTC)

It is a SHA-1 hash code but it is only used as a unique identifier for the swarm. It isn't used to verify the contents of the data, so technically not a checksum. Rurik (talk) 00:07, 27 June 2009 (UTC)
Each 64-4096MB piece also gets a SHA-1 hash, which is used to verify the data's integrity. Theymos (talk) 00:59, 27 June 2009 (UTC)
You're right, I misread the content. In terms of consistency, the two aren't really interchangeable. The checksum is the 'what', the SHA-1 is the 'how'. Rurik (talk) 01:29, 27 June 2009 (UTC)
I agree. The article is fine as is and there is no consistency problem. Theymos (talk) 02:05, 27 June 2009 (UTC)


Many files can not be downloaded due to lack of seeding. —The preceding unsigned comment was added by Ergn (talkcontribs) 01:06, August 23, 2007 (UTC).

To address the problem, section "Content unavailability" was added. —Preceding unsigned comment added by (talk) 15:45, 27 January 2010 (UTC)

Are Animated gifs really a good idea?

This is the second time in a few days I've had to put my hand over part of the screen so I don't have an irritating flashing image making the article really difficult to read. Can't there just be a static image linked to the animation? I'm not sure the animation really illustrates much anyway. See, all that flashing has made me bad tempered! —Preceding unsigned comment added by (talk) 00:04, 26 April 2008 (UTC)

The concept of the animated gif doesn't really bother me, but it would be a lot more useful as an article supplement if it went slower. Right now, I feel like it cycles through too quickly to be a good illustration. I have no idea if this is something that can be changed or not, though. Anneagain (talk) 22:24, 30 September 2009 (UTC)
Hitting the escape key in the Firefox browser halts any animated gifs. Quite handy. :) (talk) 17:18, 2 April 2010 (UTC)
Good idea - this works in IE & FF. It's an accessibility issue. I've added "To stop animation, click browser Stop or hit ESC key." to the Image box caption. --Lexein (talk) 00:39, 3 April 2010 (UTC)
Anneagain, you're not meant to 'understand' all the transfers - most of them are random. It's just to show the idea that once the file is 'out there', the original server has little to do while others exchange parts of the file, and many users each end up with a full copy. --Nigelj (talk) 21:02, 2 April 2010 (UTC)
I'm not sure this is the best solution, as not all users are going to be using IE/FF, and some don't use a browser with a 'click-able stop button', some don't have a mouse at all, etc. I really think that this thing is an eyesore, like the first poster, and I think that it should be added as a small thumbnail of the 8 frames shown in 'film roll style', as in one below another, in small thumbnail size of course, which upon clicking leads to the animation. If there isn't some way to have this picture be non-animated, at least to start with, I don't feel it should be on this page, it doesn't add enough to the article.
If anyone agrees/has any more thoughts, I'm perfectly fine with doing all of the legwork with the image and article editing, seeing as there are several opinions, all equally valid, maybe I'll wait for a response or the very least, can I at least shrink it? Nigtv (talk) 22:04, 7 April 2010 (UTC)
Edit the image and make the cycles bigger so that the image at the end is slower. Wonder why it should be so fast: in my opera it is really slow (a bit to slow for me)... But i wouldn't remove the animated image! mabdul 14:38, 8 April 2010 (UTC)
As soon as I have a free second or two, I'm going to try that, as well as lowering the number of clients and pieces in the image. This image would work much better with only 3-4 clients and 3-5 pieces. kifo 09:25, 15 April 2010 (UTC) —Preceding unsigned comment added by Nigtv (talkcontribs)

What goes inside the torrent

I think this article could be improved if an expert can add information regarding what goes inside a .torrent file. I was looking it up myself but is really complex, but I am sure someone familiar can add this information so more people could get knowledge on what is inside the torrent file itself.--Camilo Sanchez (talk) 17:23, 5 August 2009 (UTC)

I've started working on this: User:TomFitzhenry/.torrent
On the talk page I've posted a todo list and some things that I'm unsure how to present: User_talk:TomFitzhenry/.torrent -- TomFitzhenry (talk) 10:29, 21 October 2009 (UTC)

Does a protocol really have a logo? You quite sure it isn't the logo of the BitTorrent INC or perhaps of the official client (also by the same name)? (talk) 20:24, 20 September 2009 (UTC)


The text says SHA1 is used for the hashes, however this is demonstrably incorrect. Create a file with a single null in it (SHA1 hash=5ba93c9db0cff93f52b521d7420e43f6eda2784f), create .torrent file, open in client (i.e. Transmission), click on properties and note the hash given is different. This should be explained here. —Preceding unsigned comment added by (talk) 17:16, 12 October 2010 (UTC)

Really? You guessed at the behavior (getting it wrong) and then changed the article instead of, say, looking up one of the many copies of the spec? Here's one copy. Here's a Google query you could have used. Brian Geppert (talk) 10:15, 20 November 2010 (UTC)


I'm probably going to cut out the Sandvine part for inaccuracy.

Sandvine has CLAIMED for a long time that their products effectively identify and block Bittorrent traffic. While this statement is technically true, it also blocks huge swaths of other applications because of false positives. The cited article used this equation for traffic shaping uploads only:

For a period of time(x), from on single IP, fixed UDP port -> many destination IP(y), fixed or random UDP ports

Of course, this is not what Sandvine was doing. Sandvine ONLY blocks TCP, as there is no equivalent to the TCP RST command in UDP. So they used something like the above equation for TCP traffic. And lots and lots of legitimate TCP traffic (XBOX Live, for example) uses this pattern. So when deployed on Comcast, Sandvine initially blocked or throttled LOTS of applications.

This kind of pattern analysis is desirable because (as per the article) it's much cheaper than deep packet inspection(DPI). The problem is that it doesn't work in practice. —Preceding unsigned comment added by (talk) 20:13, 15 March 2010 (UTC)


I note the ACS:Law article is marked as semi-orphaned. Seems a link from bittorrent would be highly appropriate. ;-) —Preceding unsigned comment added by (talk) 11:36, 10 April 2010 (UTC)

Legal Issues

The article states that Mininova has been shut down. This is not entirely true. While Mininova has removed all content that infringes copyright in a manner not pursuant with the law, they still distribute torrents using a content-distributor model using a verification system to make sure the uploader of the torrent does in fact have legal rights to distribute said content [1]. A good majority of the content is available, if I'm not mistaken, under some sort of open-content licensing (such as some Creative Commons license). Alex (talk) 10:48, 1 February 2011 (UTC)

  • The Legal_issues_with_BitTorrent page reflects that Mininova hasn't been shut down but has redirected its efforts. The legal issues section of this article needs someone more articulate than I to summarize that for this article so that both are in agreement.Alex (talk) 12:41, 3 February 2011 (UTC)

Article Name

There needs to be an article about the actual BitTorrent protocol which is different from the BitTorrent article, much the same as the WWW article is different from the HTTP article. This article BitTorrent_(protocol) should be renamed to simply BitTorrent, or possibly BitTorrent_network, so that a second article BitTorrent_Protocol can be created. Also, the use of the word "protocol" is inconsistent and ambiguous in the article. (talk) 19:13, 23 December 2010 (UTC)

BitTorrent (and the Bittorrent redirect) is the disambiguation page, which distinguishes protocol, company, and software clearly, in my opinion. Your point about clarity in this, the protocol article, may have merit. If you have specific suggestions, feel free to make them here. Because it's a highly technical subject, there are two conflicting goals: comprehensibility for general readers, and technical thoroughness. It can be difficult to harmonize these goals, since Wikipedia is WP:NOT a tutorial, nor is it a reference manual. --Lexein (talk) 02:06, 24 December 2010 (UTC)
My point is that this article BitTorrent_(protocol) is not actually about a protocol. It's actually misnamed. This article is actually about the software and network. It never goes in to any detail about the actual protocol. look at the [[2]] which covers the architecture, network, and software. Then look at [[3]] which is about the protocol of the www. This article here is like the WWW page. There should be a second article about the protocol. This article here should hold the [[4]] then an actual protocol article at [[5]] should be created. (talk) 18:33, 11 January 2011 (UTC)
Well, this article actually does discuss the protocol in the context of the behavior of a generic bittorrent client, which I find meets Wikipedia's goal of being comprehensible to a general audience. More detail might be good, if reliably sourced, but a too-narrowly technical "pure protocol" article of the sort which satisfy programmers and academics would likely diverge from that goal, into reference manual land. There probably will never be any protocol article on Wikipedia which is complete enough to actually code any client, so it's probably best to let that go.
Some time ago, the once-unified "BitTorrent" article was split. The divisions semi-settled on company, first client, and protocol/generic-client behavior/performance/adoption/cultural impact/etc. Renaming isn't much of an option, because a search for "BitTorrent" should land on the disambiguation page. I don't think splitting further is a good option, nor is re-merging, really, so maybe careful rewriting and re-organization is the best option. --Lexein (talk) 22:52, 11 January 2011 (UTC)

I agree with mister IP address—this is a BitTorrent overview, and is written in a very accessible, unassuming manner. This is not about the BitTorrent *protocol*. — Preceding unsigned comment added by (talk) 21:04, 16 August 2011 (UTC)

User accounts per IP

"At any given instant of time BitTorrent has, on average, more active users than YouTube and Facebook combined. (This refers to the number of active users at any instant and not to the total number of unique users.)[5][6][dubious – discuss]"

It would be nice to remove the "dubious" tag, obviously, but it is going to take a little work. It can be empirically observed that the statement is true, but it could be misleading. It has to be true, because there is no such thing as no IP address for a positive number of active user accounts, and it is so improbable as to be impossible that there are no IP addresses with more than one user account. On the other hand, is the ratio significant enough to mention? If BitTorrent has numbers for IP addresses, then someone should get them. Please. Anarchangel (talk) 23:26, 16 December 2011 (UTC)


Piracy at least deserves a mention. It's not a secret that BitTorrent's main use-case is movie piracy by college students and anime fan dubs. — Preceding unsigned comment added by (talk) 23:31, 22 August 2011 (UTC)

This is the article about the protocol which is (should be) a purely technical article, explaining how it works. Piracy is mentioned elsewhere, see "See also", and the topnote. --Lexein (talk) 23:40, 22 August 2011 (UTC)

If it were purely technical it wouldn't have an adoption section. Piracy should have it's own bullet point there, probably the first one. (talk) 23:09, 29 August 2011 (UTC)

In fact, there isn't even a link to the Piracy article Fionakiwi (talk) 21:15, 27 February 2012 (UTC)

Tit for Tat/Reform?

"Downloading torrents and sharing files" section makes a reference to a "tit for tat scheme" Bittorrent clients following the specification do not operate via tit for tat whatsoever, and I believe this should be removed. If you read the spec, you'll find that the actual behavior is something else. This is from memory and very informal- A given torrent will have a number of upload slots. These upload slots are given to the peers which are giving you the most upload speed. There is no other shaping done. I think this article needs to be completely reformed as it includes very large amount of relatively specific and slightly wrong statements. It's also has very mixed information. I'm too lazy to take up any reform effort, but I will complain =p. It should be noted that bittorrent is commonly misunderstood and even so-called reliable sources may be incorrect. ( (talk)) —Preceding undated comment added 21:49, 1 June 2011 (UTC).

"These upload slots are given to the peers which are giving you the most upload speed." Sounds exactly like a tit for tat scheme. (talk) 11:18, 6 June 2012 (UTC)

Distributed Trackers

The section on Distributed Trackers should be more thoroughly cited:

  • Virtual torrents - little to no explanation of how this works by searching. Need a citation or explanation.
  • Anatomic P2P -

Also, the section itself could possibly be restructured to better inform. Changes should be outlined regarding the difference between Kademlia and BitTorrent Mainline DHT.

Liamzebedee (talk) 11:51, 14 July 2012 (UTC)

Inaccessible language

I have done my best with general copyediting, but a more knowledgeable copyeditor will need to review the article for technical language.--Soulparadox (talk) 09:54, 24 September 2012 (UTC)

Briefly describe the outstanding advantage over centralization using big-O or graph?

It is tagged that the article is too technical. It bundles the encryption and DHT stuff, I think the article is also too long. What i would like to change is that the reader would say "aha!" immediately out of visualization of p2p network, a svg graph for example, to explain what bittorrent is, and its difference between centralized server. --Mylittleanon (talk) 06:46, 7 April 2013 (UTC)


The way bittorrent works is analogous to the way big ideas are transmitted between people. Is there a science that looks at this in more detail? I think the world can learn a lot from computer science. — Preceding unsigned comment added by (talk) 21:55, 25 June 2012 (UTC)

I agree but the talk is too big. Firstly, what you really want to see should be written in p2p article, whereas bittorrent the protocol, no matter how common, doesn't change the fact that it is an implementation of the greater concept you describe, we have plenty p2p networks for different things, btw. I don't know how it is related to CompSci, the idea is highly trivial for CompSci-ists. My guess is that you are referring to sociology, what you feel so amazing, so powerful about p2p is that, if people have lower connectivity among each other, then people(the clients) have higher tendency to rely on certain clients(aka. centralized server). This happens everywhere before bittorrent, when it gets on news some brands it as illegal, we also have new words like piracy, not strictly relevant though.
I am not sure how this kind of sociology can fit in, but you better find some reliable sources first. --Mylittleanon (talk) 07:07, 7 April 2013 (UTC)


The GIF detailing how BitTorrent works is mindlessly fucking confusing, and this is for someone who knows how the protocol works. I can only imagine how bewildering it is for someone who doesn't understand BitTorrent at all. — Preceding unsigned comment added by (talkcontribs) 16:38, 5 July 2011

Specific recommendations would be helpful. There was discussion about it as it was created and modified. In my opinion, a Javascript/HTML5 image player with a controllable timeline would be an improvement, but I don't know of one offhand. --Lexein (talk) 23:51, 22 August 2011 (UTC)
Second that; Diagram really annoying. Speaking as someone who does not know the protocol its useless and annoying. (talk) 07:39, 14 February 2013 (UTC)
Uh, yeah, got it, you don't like it. Does that mean "stop the animation" or "delete the image" or "change the image" or "make it not repeat" or what? I vaguely recall asking for specific recommendations. --Lexein (talk) 08:40, 14 February 2013 (UTC)
I completely disagree with the above, the current FildelningMedBitTorrent.png is much less useful than the previous Torrentcomp small.gif; I suggest a revert to the latter. Buhman (talk) 21:33, 12 May 2013 (UTC)
I agree - I always found the animated GIF to be a good illustration. The present one misses just about everything. I think people thought there was going to be a test, and rather than just watching the flow and understanding the principle, some thought they had to understand and remember the exact order that each computer talked to which other, and what colour each arrow was. Or something. --Nigelj (talk) 22:44, 12 May 2013 (UTC)

Relevance of college to article

I reverted this twice and I would prefer to have other feedback before doing so again. I see no relevance to listing Cohen's college in this article at all. It is extremely rare for this to be done (outside of cases like Google where Page & Brin started the search engine at Stanford). Cohen didn't start BitTorrent at Buffalo, didn't receive any assistance/funding/etc. from Buffalo, and didn't even graduate from there. If I'm the only one who thinks this, I will leave it alone but I would like to see other editors opinions on this.Caidh (talk) 12:31, 9 August 2013 (UTC)

Hey I am unsure if pages like should be listed in BitTorrent#Broadcasters. I am not sure what the policy is for a listing. functionas as a bittorrent webseed supported tracker for podcasts. --Sk!d (talk) 12:29, 2 January 2014 (UTC)

My apologies - I inadvertently removed your comment on the talk page but I restored it. Sorry about that! Caidh (talk) 19:50, 2 January 2014 (UTC)