Talk:Peer-to-peer/Archive 2

Archive 1

Archive 2

Arun Kumar Srivastava

For the record, I am also a regular at WP:RM (see my move log if you doubt me), and I think 199.125.109.88's close was correct. The arguments for and against the move are essentially of equal strength, one editor was convinced to change his mind from support to oppose, and (last and least) there were twice as many editors opposing the move as were supporting it. So there was clearly a consensus to keep the current title, or at most no consensus to support a move. I will say that Kbrose was wrong to close the discussion; editors should never close discussions they are involved in, unless the outcome is unanimous. I always avoid closing discussions I've participated in, even if I've only offered a comment and not a formal !vote. Also, Kbrose was clearly incorrect in stating that a simple vote count was enough; this is clearly not what WP:Consensus says.--Aervanath (talk) 04:33, 8 July 2009 (UTC)

Note: My comment above was cross-posted from Wikipedia_talk:Requested_moves#.22close.22_of_Peer-to-peer_move_request.--Aervanath (talk) 04:36, 8 July 2009 (UTC)

Arun Kumar

Cleanup

I did not put the {{cleanup}} tag on the article but I agree it should be there. I think the article needs an extensive rewrite and then a move. Its topic and scope are not clearly defined. I added text saying peer-to-peer is in contrast to client-server and master-slave, and someone else deleted my text as irrelevant. This is absurd. In some computing environments, peer-to-peer and master-slave are alternative architectures. In supercomputing, for example. Either this is an article that covers it all, in which case it has to accomodate any computing environment the reader may have in mind; or the article has some other, narrower scope, not yet stated. I am thinking this article is an example of Blind men and an elephant. Also, the incoming links need disambiguating. I am a frequent participant in Wikipedia:Disambiguation pages with links, but I will not disambiguate incoming links to to this article. If the article is moved to, eg, Peer-to-peer (computing) and the disambiguation page is put here where I think it belongs, I will then gladly help to disambiguate the incoming links. --Una Smith (talk) 16:40, 8 July 2009 (UTC)

It is painfully clear that there are at least two discrete topics in this article, and three groups of editors. See for example Kbrose's statement that The concepts of 'P2P' as well as 'client-server' computing have nothing to do with 'network architecture', other than that both are Application Layer service delivery models in IP networks that can be transacted over any network architecture that IP supports.[1] Some editors here want this article to be about P2P network architechture, others want it to be about P2P applications, still others want it to be about both ("anything relating to peer-to-peer whatever and computers"). --Una Smith (talk) 16:49, 8 July 2009 (UTC)

No, these are not separate topics. There is only one major aspect of 'peer-to-peer' from which all others, even the social networking meme map, are derived. In fact, the definition given in peer-to-peer (meme) is a very good, encompassing one, however using the language of philosophy or sociology. *This* article defines it in terms of computer technology, which is where the concept of peer-to-peer originated. Surely, modern western societies have developed to a point at which the concepts of equal peers and peering are very relevant and appealing and therefore society has provided a 'breeding ground' for the manifestations of peer-to-peer systems. The Internet has been an enabler of the paradigm, originally envisioned in the beginnings of the ARPANET (e.g.,the end-to-end principle is a component of the idea) and the peer-to-peer systems are the exploitation of the technical tools and societal conditions.

'P2P networking' and 'P2P computing' are one and the same thing. They are a form of application architecture, specifically, distributed application architecture. Distributed applications are applications built on top of multiple computing elements that, therefore, have to communicate with one another, which leads to them being networked and forming a network of nodes. This is not a fundamental, native, or physical networking system, but a concerted interworking of application components, and therefore it's not a 'network architecture', but a distributed application architecture. It could well be implemented without a 'network', such as in a massively parallel supercomputer connected with a high-speed backplane (not very different really from a network). The collection of nodes of peers may well be called a 'P2P network', just as the collection of franchises of a given 'brand name' product may be called a 'brand name' network of stores. You wouldn't characterize it 'brand name (network architecture)'. Such an application network may have an architecture, which is expressed in the various topologies of P2P systems (so far poorly represented in the body of this article). This may be called a P2P network architecture, but one has to understand the proper association, namely P2P-network architecture, and not P2P network-architecture. This is why it's improper to categorize this article as 'network architecture'. All P2P system may use the same network architectures of IP, with its subnets, routers, switches, whatever. And so does the client-server model, which is another form of dividing an application into cooperating partner components, without sharing between equal partners. The concept of master-slave is usually different, actually can have multiple interpretations, in computing, usually it is a relationship of replication of service or application to provide redundancy in which the master is the 'primary' controller.

In several instances here the comment has been made that P2P file sharing (e.g.) is not an application, but that the client program is the application. This is wrong from a computer science perspective, the client is merely a program, a computer code. The application is 'file sharing', without any other peers (or servers in c/s computing) the client program is absolutely useless, and not an application of anything at all.

This article is indeed poorly written. The body of the article needs complete rework, doesn't even reflect the architectural aspects of peer-to-peer system. It's just a random accumulation of pieces of text. The correlation of them is left up to expert readers who should already know it to make sense of this. But that is not unusual on Wikipedia. Wait and watch for more changes, but don't confuse the issues even further by renaming and disambiguating what already has the proper title. In general I don't like articles to have names that really are just adjectives, like 'peer-to-peer', it should be 'peer-to-peer something', but the term has come to have pseudo-noun status in much of the literature. Kbrose (talk) 18:38, 8 July 2009 (UTC)

Sorry, but what you say an application is, is doubtless a correct definition, but:

1. (computing) A computer program or the set of software that the end user perceives as a single entity as a tool for a well-defined purpose. (Also called: application program; application software.)

The iPhone application VR+ connects to major social networks.

is another one.- (User) Wolfkeeper (Talk) 19:48, 8 July 2009 (UTC)

I am not disputing this definition, if the 'tool' fulfills a purpose by itself. Here is not the place of definitive or exhaustive definitions. I addressed a specific example, in which the tool does not make a complete application. When it doesn't there is no application at all. After all, an 'application' is the deployment of a method to solve a problem, accomplish a goal. If you don't accomplish a goal, you merely made a mess. Note your (comment): 'application program/software' which clarifies the situation and is the complete and better description. Kbrose (talk) 20:08, 8 July 2009 (UTC)

I agree that peer-to-peer (meme) has a superior definition of what "peer-to-peer" means. In fact, many links to Peer-to-peer probably should go to Peer-to-peer (meme) instead. Look also at the redirects to Peer-to-peer. Ignoring the permutations on "peer-to-peer" itself there exist:

• Peer (Networking)
• Peer-to-peer communications
• Peer-to-peer (computing)
• Peer-to-peer computing
• Peer-to-peer network
• Peer-to-peer (network architecture)
• Peer-to-peer networking
• Peer-to-peer networks
• Peer-to-peer protocol
• Peer-to-peer sharing
• Peer-to-Peer Streaming Systems and Incentive Mechanisms

I also agree that "peer-to-peer" is an adjective, and almost always qualifies something. (Rarely it is used alone and the reader must infer from context what is being qualified; in the case of some book titles that is difficult.) So I typed "peer-to-peer" into the Wikipedia search box, used the Search (not Go) button, and looked at the context of that string in the first few pages of search results. Ignoring the context that was clearly social (lending, "social networks", etc.), I got:

• file sharing 11
• networks 5
• protocol 3
• applications 2
• systems 2
• distributed application
• distribution system
• network architecture
• package
• software application
• software package
• technology
• websites

Once again, incoming links to Peer-to-peer are in need of disambiguation, which supports my view that the disambiguation page should be at this ambiguous base name (an adjective). Also, from that small test alone it would appear the primary topic (if any) of "peer-to-peer" is not Peer-to-peer but rather Peer-to-peer file sharing. This is similar to the situation with Weymouth, formerly an article about Weymouth, Dorset, and now a dab page with many entries related to that town. After Weymouth was made a dab page, disambiguating links very often involved changes such as "Weymouth Beach" to "Weymouth Beach". Similarly, we should be changing many instances of "peer-to-peer file sharing" (2 links) to "peer-to-peer file sharing" (1 link). See? --Una Smith (talk) 03:36, 9 July 2009 (UTC)

The three groups (network architecture, P2P applications, both) are incomplete. My position is not that peer-to-peer refers to any of these, but rather that "p2p" and "peer-to-peer", as far as our technology-illiterate readers are concerned "have somethin to do with, like, napster and kazaa, right?". That is, most people don't know (and don't care) what p2p "really" means - they just think it has something to do with file sharing. File sharing is not an application of p2p, it's a type of information/content (not data) exchange that p2p applications have made very easy. Because the content matters, we've seen some very important legal cases, millions (likely billions) of dollars expended, the creation of political parties, and so on which have nothing to do with p2p networks in general. It's important to get our concepts straight - we can combine them all into a single article later, if we want:

• (By "peer" I mean a general purpose computer - so not one specially configured as a server - typically used for word-processing, browsing the internet, and so on.)
• Shared resources: the ability to access files on other (typically peer) computers
• Peer-to-peer computing: distribution of computational tasks across peers (cf. folding@home)
• Distributed computing: an application that spans several servers (not peers)
• Peer-to-peer networking: when the applications running on all sides are the same - usually this means 'same handshake'
• Peer-to-peer: "participant equipotency"
• File sharing: the electronic exchange of creative works
• Peer-to-peer networks made and used for file sharing.

This article should be about peer-to-peer networking, with very little focus on file sharing, and no vague theoretical fluff. File sharing technologies belong in file sharing. I don't want to see the word "equipotent" unless it's defined beforehand. What the hell is a "service peer"? It "partitions tasks"? No, that's just distributed computing.
We should first remove poor composition and fluff. Then we can get rid of the excess file sharing junk. Then we can settle on a definition. Let's try to avoid reverting each other (offer corrections instead, or go to talk), and ensure that we spend just as much time improving the article as we do talking here. The size of this article should be going down, not up.    M   19:31, 14 July 2009 (UTC)

OLTP

The article implies that P2P is a fairly recent invention utilizing the Internet. However, I recall first hearing the term "peer-to-peer" (before 1990) in the context of OLTP systems (primarily mainframe systems, such as CICS). Peer-to-peer was meant to replace the client-server model of programming, eliminating the restrictions placed on each process, so that it could operate on one context as a client and in other context as a server when communicating with multiple other processes. The article makes no mention of this history. — Loadmaster (talk) 19:10, 13 July 2009 (UTC)

You are certainly right about earlier usages of these concepts and some are mentioned in the article (earlier today I tried to emphasize that in a new section). It will take some time to rework the article and balance out some of the bias that's presented. Contributing reliable sources would help the effort. Kbrose (talk) 19:30, 13 July 2009 (UTC)

Peer

Peer are the persons who are same like you (like your age, abilities, rank or position) —Preceding unsigned comment added by 125.60.243.87 (talk) 12:03, 14 July 2009 (UTC)

Right, and a server is a person who brings you food. I guess we should more explicitly define peer, though.    M   20:57, 20 July 2009 (UTC)

Mesh and MANET articles

Mesh networking and Mobile ad hoc networks are two closely related concepts. These ad hoc networks do implement P2P principles, however since they are often embedded in hardware and deal with actual physical networks and not overlay networks they are seen as distinct from P2P.

In a sense, all digital networks are overlay networks, we just need a layer diagram to situate all of them... Anyone got a good one to share?

And so where should Ad hoc network redirect to? Ooskapenaar (talk) 21:30, 12 January 2010 (UTC)

Comments in the article

I've rewritten (a bit) the first section, trying to semplify it. I've found some HTML comments, criticizing some sentence in the article. I don't know who wrote them, but I think they make some confusion. I list them here, with my opinion about them. My opinion is not explained with citations, if you think I'm wrong feel free to say it here.

Comment: Most DHTs are not 'structured'.

My opinion: I know that DHTs are structured by definition. The use itself of DHTs makes a network structured.

Comment: huh? all of them [Unstructured P2P systems] do [have an algorithm for organization of network connections], except direct connect/darknet-like networks, where people select their own peers. And is not this just ad-hoc?

My opinion: Unstructured P2P are systems where there is no algorithm for organization of topology and/or resource location. Think of launching sensors off a plane to a forrest. Every sensor is connected to other sensors within its range. The resulting P2P network can use P2P techniques (such as flooding) to operate, but its topology is clearly unstructured.

Comment: these categories are much too vague to be formally defined by us.

My opinion: Ok, "pure", "hybrid", "centralized", ... categories are a bit vague and sometimes overlapping, but we should give a yet vague definition, basing on how the terms are often used in scientific articles. —Preceding unsigned comment added by 4nT0 (talk • contribs) 15:06, 25 August 2010 (UTC)

I concur with your comments. These comments where added HERE after the article was cleanup previously. Kbrose (talk) 15:20, 25 August 2010 (UTC)

Thanks for the feedback! Hope to have some time in the near future to cleanup the whole page completely. --4nT0 (talk) 12:26, 28 August 2010 (UTC)

Very good. This article topic was taken over at some point by file-sharing zealots with limited real knowledge who would like it that P2P has only that meaning. Kbrose (talk) 17:42, 28 August 2010 (UTC)

I agree to your comments, 4nT0. Whoever made them didn't know much about the background of this topic...

mfg, OldDeath - 15:08, 31 August 2010 (UTC)

Proposed edits

Course Assignment

We will be editing this page over the next few weeks as part of our Online Communities course assignment at Cornell University. The course site can be found here.

Here are some potential plans that we currently have in mind for the article:

• Some sections have already been designates as "requiring expansion". We intend to expand the "hybrid models" subsection and the "current research" section.
• We want to add sections such as the "Social Impacts" & "Political Impacts" and subsections like "Legal Issues". The addition of these sections will make the article more connected as subsections such as "Incentivizing resource sharing and cooperation" would more appropriately fit under the Social Impacts, or under a more general subsection entitled "online collaboration.
• Many of the subsections can be further developed and expanded, using some of the sources which we have outlined below.
• We plan to implement some of the suggestions of other wikipedians in this talk page, and will work with other editors to improve this article.

Some sources which we would like to include to support our article include:

• Networked Publics by Kazys Varnelis - a secondary work about new media trends which we see today in
• Lawrence Lessig's work surrounding Remix Culture and its implications. His book will provide much support for discussing the paradigm shift that peer-to-peer collaboration has caused in modern society. We are considering including a video file of his TED talk, if we are able to get permission to post in on Wikipedia.
• Zizi Papacharissi has several recent research papers which can help us develop the sections on the "Social Effects" and the "Political Implications"
• There are multiple research papers available to help us as we develop the sections designated as "Requiring Expansion". Specifically, we are looking at papers such as Yang's (ref: http://ilpubs.stanford.edu:8090/727/1/2001-36.pdf) to expand the Hybrid Models subsection.

Christian will oversee a restructuring of the article, and modification of the organizational structure in compliance with Wikipedia guidelines. He will also assist the rest of the group in finding high quality sources, and assure the article's neutrality.

Josh will primarily create the new sections that we plan to add and citations to support the newly generated content.

Alyssa will primarily expand the existing sections in the article, ensuring that each subsection fits into the hierarchy of the article.

We look forward to working with you all! Please contact us with any feedback as we update the article!

• CBCompton
• J grider65
• AlyssaG92
• Cespo4

CBCompton (talk) 05:41, 24 September 2013 (UTC)

DYK Nomination

Just wanted to let you all know that we recently nominated this article for Did You Know Status. Please help us to continue to make this article great!

CBCompton (talk) 12:11, 3 October 2013 (UTC)

Hey guys, I'm gonna be editing the page in a few days. I was wondering what you thought of my edits right now. Feel free to modify them as you like. Cp123127 (talk) 00:00, 20 March 2013 (UTC)

Niteshgoyalwiki (talk) 15:09, 1 October 2013 (UTC)Feedback from Niteshgoyalwiki (talk)

• Move the Template for Educational Assignment to the top of the Talk Page. Done. Good Job.
• I'd advise being more proactive and talking to other users on the talk page by commenting back to them suggesting how you would like to incorporate their comments and suggestions. This should improve chances of communication.
• Try to write content in an understandable language when you discuss some of the content (research papers etc.) Also discuss why these papers in particular are more important to be covered here than the rest of the research work.
• Remember to nominate your article to did you know by the end of class on Oct 1.

Major edits

• The Distributed hash tables section is much too technical. This should be left to the actual article to explain. On the peer to peer article, it should only serve to put peer to peer in more context. Additionally there is already a very good mention of distributed hash tables in Structured systems and I'm not sure whether that should serve as the only mention of DHTs in the article. Perhaps this section should be moved to the actual DHT page.
• The Unstructured systems section in marginally inaccurate, as it says little about the bandwidth inefficiencies of flooding the network with a query, and makes no mention of Gossip protocol, though it is very relevant.
• Social and economic impact is a little lacking and needs to be expanded.
• Applications needs to be updated with new, relevant software using p2p, because there are not many new pieces of software there now.

Minor edits

• The Advantages and Weaknesses section needs more structuring, and needs to be separated out into individual advantages and weaknesses.
• The introduction lacks certain citations, uses unprofessional words: "the first P2P killer application", basic grammar mistakes, etc.
• Perhaps there should be a new section under Architecture of P2P systems named "Hybrid systems", as after the portion on unstructured systems, it names hybrid systems but does not create a separate section for them.
• Many of the projects referred to in Applications are not notable enough to warrant inclusion.
• The article contains a couple of citation needed tags which need to be cleaned up.
• Rename Peer-to-Peer (meme) to Social peer-to-peer processes, as it fits better

Feedback from Prof. Leshed

You are on the right path to making substantial improvements to this article. Good job! Here are two suggestions to ensure that your project is successful:

• Reach out to other editors of this article to get their help, advice, and feedback on changes you are making or proposing to make. This means, going to their talk pages and posting a message there.
• Ensure at least one of the group members is comfortable with the wiki markup language and is getting familiar with the Wikipedia guidelines and standards. This will make your contributions more substantial, and avoid mistakes such as User:CBCompton instead of CBCompton or Christian.
• Sign all posts on the talk page with four tildes ~~~~ to avoid "unsigned" posts.

Happy editing! LeshedInstructor (talk) 14:50, 30 September 2013 (UTC)

Looks good. You should post your planned outline on this talk page. It's always best to get some discussion before you make major changes to any article. That being said, your structure looks good. Also be sure to include information that's already in the page, making any corrections as necessary; restructuring a page should not remove anything of value, but should present existing (as well as new) info in a more readable way. Also be aware of the more general editing guidelines for Wikipedia. — Loadmaster (talk) 16:11, 30 September 2013 (UTC)

Hierarchy Changes

This is the new template we were considering implementing

 1. Historical Development

 2. Current Applications
   a. Communications
       i. Other P2P Applications
   b. Content Delivery
   c. File Sharing Networks
       i. Streaming Media

 3. Architecture
   a. Routing and Resource Recovery
       i. Hybrid Models
       ii. Structured Networks
       iii. Unstructured Networks
   b. Security and Trust
       i. Routing Attacks
       ii. Corrupted Data and Malware
   c. Creating more resilient and scalable computer networks
   d. Distributed storage and search

 4. Social Implications
   a. Demographics and Usage stats
   b. Incentivizing resource sharing and cooperation
   c. Privacy and Anonymity
   d. Economic Implications
       i. Music/Film

 5. Political Implications
   a. Network Politics
   b. Network neutrality
   c. Intellectual Property law and illegal sharing

 6. Current Research
   a. Future Trends

This change was implemented on 10/1 after consulting several Wikipedia editors who had previously made edits on the page. Feedback is greatly appreciated for our course project!

CBCompton (talk) 14:26, 8 October 2013 (UTC)

Social Peer-to-peer processes integration?

As part of our course assignment, we were considering merging this article on Social peer-to-peer processes into this one. Please let us know your thoughts! CBCompton (talk) 15:21, 8 October 2013 (UTC)

Information on the process can be seen at Wikipedia:Merging.. You can also propose a merger of the two pages by place the template {{merge|Social peer-to-peer processes|discuss=Talk:Peer-to-peer#Social Peer-to-peer processes integration?|date=May 2024}} at the top of each page - that links back to this page and this tlak. -- Moxy (talk) 17:24, 8 October 2013 (UTC)

In general, there are two approaches: 1. Combine the two articles, and 2. Reference one article from the other. It appears that the second approach is best, adding some material and links in the "Social implications" section in the p-to-p article, but leaving the two as separate articles. If you find too much duplication, and you think the two articles should be combined, then you should start a new discussion topic here to get other opinions. (Follow the process, as Moxy pointed out.) A cursory reading of the two articles, though, seems to indicate that they are different enough that they should be separate articles (to me, anyway). Others may feel differently, especially if they can find a logical way to combine the two concepts into a single article. — Loadmaster (talk) 17:55, 8 October 2013 (UTC)

Confusion among multiple meanings of peer-to-peer

There are multiple meanings of "peer-to-peer" in the computer networking world, and this article creates confusion by lumping them together. A similar observation was noted in 2009 on this Talk page (see section "OLTP", above). The disambiguation page points out some of the multiple meanings, the first one (the one that links to this article) being "a distributed computing architecture." While that isn't very specific, it seems to refer to the notion of peer-to-peer in terms of the architecture of the network itself (e.g., network layer), not the applications that run on it. This is what was achieved in the early days by (if I'm not mistaken) network architectures such as ARPANET (Internet technology), X.25, XNS, and DECnet -- as opposed to, for example, IBM's SNA, which was hierarchical rather than peer-to-peer. This article fails to make this distinction. Note that both the DECnet and SNA articles link to this page, because those articles are using the term in the context of contrasting peer-to-peer vs. hierarchical network architectures.

Therefore, this article should focus mainly on "peer-to-peer" in the network architecture sense, and the other senses can be dealt with in other articles (like peer-to-peer file sharing). Either that, or this article should disambiguate the different meanings. As it stands, this article gives credit to such things as Napster and PeerLogic for being among the "earliest" peer-to-peer systems (which may be true in the sense of application architecture), when in reality, the earliest peer-to-peer systems were those network technologies being developed and used in the 70s and 80s (DECnet, etc.). -- HLachman (talk) 03:58, 25 September 2014 (UTC)

Content should be re-framed

Wikipedia articles seek to add meaning or improve understanding to works that are referenced in the article. This article doesn't seem to do so. Rather, it minces misunderstandings. A friend working as a Software Scientist for Hewlett-Packard explained his view of a server that's stayed with me because it's simple and complete: A server is a service that manages resources.

Another term that is some is used is "daemon". An example many have seen in email, is "mailer-daemon", which handles incoming email. "Server" and "Client" may be used regardless of the architecture. Examples are Client-Server and Peer-to-Peer architectures. In an ideal Client-Server network, a client and a server are each an individual node. In Peer-to-Peer, each node has a client and a server. Kernel.package (talk) 01:32, 17 December 2014 (UTC)

And what is your problem? --Kgfleischmann (talk) 06:52, 17 December 2014 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified 6 external links on Peer-to-peer. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20040910154927/http://www-db.stanford.edu/~manku/phd/index.html to http://www-db.stanford.edu/~manku/phd/index.html
• Added archive https://web.archive.org/web/20131006022409/http://www.p2pnews.net/2012/06/14/art-thou-a-peer/ to http://www.p2pnews.net/2012/06/14/art-thou-a-peer/
• Corrected formatting/usage for http://www.p2pna.com/glossary.html
• Corrected formatting/usage for http://www.sciencedirect.com/science/issue/5624-2008-999689997-678759
• Corrected formatting/usage for http://p2pna.com/
• Added archive https://web.archive.org/web/20070928224509/http://www.igi-pub.com/books/details.asp?ID=4635 to http://www.igi-pub.com/books/details.asp?ID=4635

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 16:20, 26 July 2017 (UTC)